The Goals of Private Law 9781472560629, 9781841139098

Citation preview

PREFACE

This book originated in The Fourth Biennial Conference on the Law of Obligations, which was held at the National University of Singapore (NUS) in July 2008. The conference was convened by the editors of this book and co-hosted by the NUS Law School, Melbourne Law School and the Singapore Academy of Law. Sixty-eight papers were presented over the three days of that conference and we have assembled here a collection of those papers focused on the conference theme of the goals of private law. We are very grateful to Dean Tan Cheng Han of NUS Law School, former Dean of Melbourne Law School, Professor Michael Crommelin and present Dean, Professor James Hathaway for supporting the conference. We thank Serene Wee, Grace Lee-Kok and Alexis Ong from the Singapore Academy of Law for their financial support and assistance with the organisation of the conference. We also owe thanks to Guy Spooner and Lye Kah Cheong from Norton Rose, Singapore for their financial support. We are particularly grateful to Connie Yew and Khairiyah binti Abu, who ran the conference extremely well and took enormous care with every detail. Their superb organisation was greatly appreciated. We are also grateful to the NUS student helpers, especially Remy Choo Zheng Xi and Li Fangyi, for taking care of the conference speakers. We would finally like to thank all of the speakers, chairs and delegates for the great contributions they made to the conference. Financial support for the production of this book was provided by Melbourne Law School, NUS Law School, Singapore’s Ministry of Education and the Singapore Academy of Law. We are very grateful to Kevin Tan and Dimitri Ternovski for their editorial assistance. Particular thanks go to Dimitri Ternovski for his extraordinarily careful reading of the manuscript and his eagle-eyed attention to detail. We are also very grateful to the editorial team at Hart Publishing for their careful editing of the manuscript. Andrew Robertson and Tang Hang Wu April 2009

LIST OF CONTRIBUTORS

Roderick Bagshaw is a Fellow and Tutor in Law at Magdalen College, Oxford and a CUF Lecturer in Law, University of Oxford. Erika Chamberlain is an Assistant Professor at the Faculty of Law, University of Western Ontario. Hanoch Dagan is Dean and Professor of Law at the Buchmann Faculty of Law, Tel Aviv University. Anthony Duggan holds the Hon Frank H Iacobucci Chair at the Faculty of Law, University of Toronto and is a Professorial Fellow at the Faculty of Law, University of Melbourne. Matthew Harding is a Senior Lecturer in Law at the University of Melbourne. Steve Hedley is a Professor at the Faculty of Law, University College Cork. William Lucy is Professor of Law, Manchester University. Mayo Moran is Dean and James Marshall Tory’s Dean Chair of the Faculty of Law, University of Toronto. Donal Nolan is a Fellow and Tutor in Law at Worcester College, Oxford and a CUF Lecturer in Law, University of Oxford. Stephen Perry is the John J O’Brien Professor of Law and Professor of Philosophy, University of Pennsylvania Law School. Andrew Robertson is a Professor of Law at the University of Melbourne. Craig Rotherham is a Professor at the School of Law, University of Nottingham. Emily Sherwin is Professor of Law, Cornell University. Stephen Smith is James McGill Professor in the Faculty of Law, McGill University.

x

List of Contributors

Robert Stevens is Professor of Commercial Law, University College London. Yock Lin Tan is Professor at the Faculty of Law, National University of Singapore. Tang Hang Wu is an Associate Professor at the Faculty of Law, National University of Singapore. Graham Virgo is Professor of English Private Law, University of Cambridge. Charlie Webb is a Lecturer in the Law Department at the London School of Economics and Political Science.

1 Introduction: Goals, Rights and Obligations ANDREW ROBERTSON*

I. Introduction It was once taken for granted that the identification of the purposes of legal rules and assessment of the adequacy of the rules in light of those purposes was good legal scholarship.1 70 years ago Lon Fuller and William Perdue began their famous essay on contract damages with the observation that greater attention needs to be paid to the purposes of legal rules.2 They suggested that, although the notion that law is a means to an end is a commonplace idea, legal writers too often lose sight of it in the practice of legal analysis. The idea that private law might usefully be thought to pursue goals has, however, become controversial. We have recently been warned of the dangers of seeing law as a means to an end.3 Moreover, Ernest Weinrib has famously argued that private law, like love, has no goals.4 While Professor Weinrib is by no means alone in adopting an anti-instrumentalist understanding of private law,5 he acknowledges that this does not reflect ‘the standard view in contemporary scholarship’.6 Professor Weinrib is right to suggest that most scholars and judges in the common law world accept that goals are relevant to private law, but few would subscribe to what Professor Weinrib characterises as ‘the dominant contemporary view’ that ‘private law is—and can be nothing but— the legal manifestation of independently justifiable goals’ (emphasis added).7 *

I am very grateful to Anne Orford for her helpful comments on a draft of this chapter. G Williams, ‘The Aims of the Law of Tort’ [1952] Current Legal Problems 137, 137. 2 LL Fuller and WR Perdue, ‘The Reliance Interest in Contract Damages: 1’ (1936) 46 Yale Law Journal 52, 52. 3 BZ Tamanaha, Law as a Means to an End: Threat to the Rule of Law (Cambridge, Cambridge University Press, 2006). 4 EJ Weinrib, The Idea of Private Law (Cambridge, Massachusetts, Harvard University Press, 1995) 3–6. 5 See, eg, A Beever, Rediscovering the Law of Negligence (Oxford, Hart Publishing, 2007); R Stevens, Torts and Rights (Oxford, Oxford University Press, 2007) and R Stevens, ‘The Conflict of Rights’ ch 6 of this book. 6 Weinrib, above n 4, at 2. 7 ibid, at 5. 1

2

Andrew Robertson

Rather, for most judges and scholars, the consideration of questions of purpose in private law is neither heretical nor apt to yield anything like a complete answer to the question of what private law is all about. Much contemporary scholarship adopts a mixed view, which recognises the importance of function in private law without subscribing to a wholly instrumentalist account. Analysis of the goals of private law in contemporary scholarship is undertaken at different levels. At the broadest level, scholars seek to identify the goals of the entire field of private law in order to understand the nature of the field as a whole, and sometimes to consider whether and how it might be distinguished from public law. A second level of analysis concerns the goals of particular branches within private law, such as contract and tort. A third and more well-established tradition in the common law seeks to identify the goals of particular rules so that those goals might influence the development of legal doctrine. While the analysis of the goals of private law as a whole, of its constituent fields and of particular doctrines may be undertaken for different purposes, they are by no means separate questions. The higher-level analysis tends inevitably to descend into a consideration of the goals of particular doctrines. When Professor Weinrib denies that private law has goals, he clearly means that neither the constituent parts of private law nor particular doctrines should develop by reference to external goals. When we ask what purposes are served by contract law or tort law, we expect that the answers will help us to understand particular doctrines of contract and tort and will influence their development. The goals of a field of law are thus most commonly understood as the aspirations that inform the explication and development of the law in that field.8 A consideration of the goals of private law, understood as the objectives or purposes that might be said to influence the development of private law rules, does not involve the anthropomorphisation of private law, and nor does it involve any speculation about the ‘actual workings of the judicial mind’.9 Lord Steyn has observed that, although the real purpose of a rule is often debatable, it is possible to identify that purpose through ‘rational argument, and a judgment by an impartial judge’ and it is then ‘an important and legitimate matter to enquire whether the rule as formulated fulfils that purpose’.10 The endless common law process of searching for the goals of legal rules and re-shaping the rules in light of those perceived goals was explained by Oliver Wendell Holmes as follows: A very common phenomenon, and one very familiar to the student of history, is this. The customs, beliefs, or needs of a primitive time establish a rule or a formula. In the course of centuries the custom, belief, or necessity disappears, but the rule remains. The reason which gave rise to the rule has been forgotten, and ingenious minds set themselves to inquire how it is to be accounted for. Some ground of

8

cf R Bagshaw, ‘Tort Law, Concepts and What Really Matters’ ch 10 of this book. Fuller and Perdue, above n 2, at 57. 10 J Steyn, ‘Contract Law: Fulfilling the Reasonable Expectations of Honest Men’ (1997) 113 Law Quarterly Review 433, 433. 9

Introduction: Goals, Rights and Obligations

3

policy is thought of, which seems to explain it and reconcile it with the present state of things; and then the rule adapts itself to the new reasons which have been found for it, and enters on a new career. The old form receives a new content, and in time even the form modifies itself to fit the meaning which it has received.11

What, then, are the goals that private law might be said to serve? It might be useful to begin with some basic examples from contract and tort. Although none of the essays in this collection takes the goals of contract law as a primary focus, contract law is often analysed by reference to its functions. Roger Brownsword takes the ‘core purpose’ of contract law to be the provision of ‘a regulatory framework within which exchange can take place’.12 Professor Brownsword goes on to identify three subsidiary functions of contract law: facilitating exchange; protecting vulnerable parties and the public interest against misuse of the contract apparatus; and providing a mechanism for the resolution of contractual disputes.13 Other attempts to identify the purposes of contract law have yielded broadly similar sets of objectives.14 An economic perspective yields a related but slightly different set of goals for contract law: preventing opportunism where exchanges are non-simultaneous; reducing transaction costs by supplying default rules that can operate in the absence of contrary provision; filling gaps in inadvertently incomplete agreements; and providing a set of excuses that permits efficient exchanges while preventing the enforcement of inefficient exchanges (eg, those involving involuntariness, imperfect information or negative externalities).15 For identification of the goals of tort law, contributors to this collection have turned to Allen Linden, for whom tort law could most satisfactorily be understood by focusing on its functions.16 Justice Linden suggests that tort law can usefully be understood to serve the following goals, even though it serves none perfectly and each by itself could more effectively be pursued by other means. First, it is said that tort law aims to provide a mechanism by which those who suffer loss through another’s fault can obtain compensation. While providing effective compensation to victims of harm may be better regarded as an effect of tort law, rather than one if its goals,17 it is nevertheless commonly seen by judges 11

OW Holmes, The Common Law (Boston, Little Brown & Company, 1881) 5. R Brownsword, Contract Law: Themes for the Twenty-First Century (London, Butterworths, 2000) 23. 13 ibid, at 24–5. 14 See, eg, RA Hillman, The Richness of Contract Law (Dordrecht, Kluwer, 2007) 272–3 and Steyn, above n 10, at 434: ‘The function of the law of contract is to provide an effective and fair framework for contractual dealings. This function requires an adjudication based on the reasonable expectations of parties. It is right to acknowledge, however, that the reasonable expectations of parties cannot always prevail. Sometimes they must yield to countervailing principles and policies. For example, other values enshrined in law and public policy may render the contract defeasible’. 15 MJ Trebilcock, The Limits of Freedom of Contract (Cambridge MA, Harvard University Press, 1993) 16–17. 16 AM Linden, Canadian Tort Law, 5th edn (Toronto, Butterworths, 1993) ch 1. 17 J Stapleton, ‘Duty of Care Factors: A Selection from the Judicial Menus’ in P Cane and J Stapleton (eds), The Law of Obligations: Essays in Celebration of John Fleming (Oxford, Clarendon Press, 1998) 63; D Nolan, ‘Causation and the Goals of Tort Law’ ch 7 of this book. 12

4

Andrew Robertson

and commentators as a goal.18 Secondly, by imposing liability on industrial and commercial enterprises (which can include such costs in their prices) and on activities commonly covered by insurance, tort law pursues the goal of relieving individuals from the financial burden of their injuries, and distributing the losses across sections of the community that are engaged in the relevant activity.19 Thirdly, tort law seeks to deter harmful conduct by requiring wrongdoers to pay damages for loss caused by their behaviour. The deficiencies of deterrence are also well known: some people may be ignorant of the prescribed standards of behaviour, unwilling to comply with them or unable to do so, and liability insurance commonly robs tort law of much of its deterrent effect.20 Fourthly, tort law pursues the goal of minimising harmful conduct through market deterrence: it makes harmful activities more expensive and provides an incentive to switch to safer alternatives.21 Fifthly, it is said that tort law aims to educate people about appropriate standards of behaviour and respect for others, both in general and in particular contexts. Sixthly, tort law has what Glanville Williams described as an ‘appeasement’ function: it aims to keep the peace by assuaging the desire for revenge felt by someone who has been wronged.22 Finally, tort law may aim to perform an ombudsman function by applying ‘pressure upon those who wield political, economic or intellectual power’ to use that power responsibly.23 In her contribution to this volume, Erika Chamberlain explores the ombudsman function of tort law in its application to police.24 Justice Linden suggests that the ombudsman function of tort law is not limited to government agencies, but extends to the professions and to the corporate sector in fields such as product liability and environmental litigation.25 Acceptance of the idea that private law has identifiable goals, and the notion that those goals might influence the development of private law rules, does not require the adoption of a wholly functionalist account of private law or any of its constituent parts. Complex questions are involved in the identification of relevant goals, the choices to be made between competing goals and the relationship between the pursuit of identified goals and other strong influences on private law—such as the formal values of certainty, consistency and coherence and the need to do justice to each of the parties to particular disputes. Goals are generally relevant only where the law or its application to particular facts is uncertain or where, in an appellate court, the validity of an established 18 See, eg, (in relation to vicarious liability) Bazley v Curry [1999] 2 SCR 534 [30], citing B Feldthusen, ‘Vicarious Liability for Sexual Torts’ in N Mullany and AM Linden (eds), Torts Tomorrow: A Tribute to John Fleming (Sydney, LBC Information Services, 1998) 224 and see also (in relation to tort law as a whole) Williams, above n 1, at 151–72. 19 Linden, above n 16, at 4–6. 20 ibid, at 7–9. 21 ibid, at 18. 22 Williams, above n 1, at 138–40; Linden, above n 16, at 15. 23 Linden, above n 16, at 20. 24 E Chamberlain, ‘Negligent Investigation: Tort Law as Police Ombudsman’ ch 12 of this book. 25 Linden, above n 16, at 22–4.

Introduction: Goals, Rights and Obligations

5

rule is under challenge. Neil MacCormick, for example, emphasises the role of consequentialist arguments ‘in problem cases’ (emphasis added).26 Brian Tamanaha warns against the dangers of an excessively instrumentalist approach to law, but acknowledges that the development of the common law has long been influenced by considerations of convenience and social utility.27 Moreover, he accepts that the instrumental dimension of law has a positive role to play, provided it is integrated within a system with a strong commitment to the ideals of justice and the rule of law.28 The mainstream view seems to be that policy goals have a legitimate but limited role to play in private law. The prevalent view is perhaps that the goals of legal rules are relevant only in exceptional cases and, even in those cases, the pursuit of goals needs to be balanced against a range of other considerations and will only occasionally have a decisive influence in moulding private law rules.29

II. Private Law and Public Goals This collection begins by considering the goals of private law from the point of view of the relationship between public and private law. The focus of this book on the goals of private law inevitably directs attention to the public dimension of private law. A number of chapters in this collection are concerned with the identification and legitimacy of community welfare concerns, public values and even public law values in private law. In her essay Mayo Moran observes that, from the point of view of public law, private law is commonly seen as a mechanism for the realisation of important public law values such as those entrenched in bills of rights.30 As is well known, this view is not without foundation: Professor Moran discusses a number of examples of cases in which constitutional norms such as equality, non-discrimination, the right to privacy and freedom of expression have shaped the development of private law rules and influenced the outcomes of private law cases. For some private law scholars, the intrusion of these values 26 N MacCormick, Rhetoric and the Rule of Law: A Theory of Legal Reasoning (Oxford, Oxford University Press, 2005) 118. 27 Tamanaha, above n 3, especially at 34. Tamanaha also seems to accept the impossibility of ‘constructing a plausible, functional non-instrumental view of law’: (at 132). 28 ibid, especially at 249, where he identifies the necessary constituent ideals as follows: ‘that the law is a principled preserver of justice, that the law serves the public good, that legal rules are binding on government officials (not only the public), and that judges must render decisions in an objective fashion based upon the law’. Tamanaha’s real complaint seems not to be with the idea of law as a means to an end, but with the idea that ‘law is an empty vessel to be filled as desired’, the notion that ‘legal rules are tools to be manipulated to achieve ends’ and with ‘the rampant manipulation of the law for particular ends’: ibid, at 1, 249, 250. This point about Tamanaha’s book is made by Leslie Greene, ‘Law as a Means’ in P Cane (ed), Hart-Fuller at 50 (Oxford, Hart Publishing, forthcoming); Oxford Legal Research Paper 8/2009 available at http://ssrn.com/abstract=1351304, accessed on 25 April 2009. 29 See further A Robertson, ‘Constraints on Policy-Based Reasoning in Private Law’ ch 11 of this book and E Sherwin, ‘The Rules of Obligations’ ch 18 of this book. 30 M Moran, ‘The Mutually Constitutive Nature of Public and Private Law’ ch 2 of this book.

6

Andrew Robertson

is to be resisted, since they distract attention from private law’s proper, and properly exclusive, focus on the interactions between the parties. Professor Moran argues that these views overlook the dynamic and positive interaction between the two bodies of law. While the impact of public law values on private law has been well documented, Professor Moran suggests that the influence has not been uni-directional. Her essay explores the way in which the legal person who is the subject of constitutional guarantees is constituted through private law rights, notably the right to bodily integrity and the right to autonomy, which is manifested through the freedom to contract and hold property. While Mayo Moran’s chapter focuses on the connections between public and private law, William Lucy asks whether, for the purpose of the law’s ‘conceptual cartography’, it is possible to draw a boundary between the two fields.31 Professor Lucy suggests that what is distinctive about private law might be that its purpose is to provide a framework within which individual goals and projects can be pursued. Public law, on the other hand, is said to be concerned with facilitating the pursuit of public or collective projects such as maintaining or changing social or economic structures. As many of the chapters in this collection show, private law routinely answers questions about the limits of private project pursuit by reference to public or collective goals. Professor Lucy’s distinction is ‘a philosophicalnormative’ distinction, which does not purport to reflect historical fact. Moreover, he says, we can expect constant tension between the two realms, and the proposed distinction between them can offer little assistance in resolving hard cases.

III. Rights and Goals The essays in the second part of this book focus on the rights-based or corrective justice-based approach to private law, which takes private law to be founded on moral duties and correlative rights arising out of the interactions between individuals. The rights-based approach encompasses a range of views, some of which recognise that community welfare concerns can have a limited influence on private law and some of which insist that they can have none at all. In chapter four, Stephen Perry examines the element of duty in a rights-based theory of negligence.32 Professor Perry sees tort law as predominantly rights-based, in the sense that the basic structure of tort law is grounded in moral duties and their correlative rights. He suggests, however, that tort law may be understood instrumentally, in the sense that it serves external moral ends rather than possessing the kind of internal coherence that Professor Weinrib claims characterises private law. Moreover, Professor Perry accepts that collective goals such as deterrence and 31

W Lucy, ‘What’s Private about Private Law?’ ch 3 of this book. S Perry, ‘The Role of Duty of Care in a Rights-Based Theory of Negligence Law’ ch 4 of this book. 32

Introduction: Goals, Rights and Obligations

7

loss-spreading can come into play as second-tier considerations to fill gaps that inevitably arise in a rights-based structure.33 The two-stage Anns v Merton London Borough Council34 test for establishing a duty of care exemplifies this pluralist understanding of the role of collective goals in a rights-based law of tort. That test, Professor Perry suggests, gives priority through the first stage to rights and correlative duties arising out of the interactions between the parties, while allowing community welfare considerations to come into play ‘in a secondary or ancillary manner’ through the second stage. In chapter five Stephen Smith develops a pluralist account of a different sort to explain private law rights.35 His essay offers a taxonomy of private law rights that distinguishes between ordinary rights, action rights (rights to court orders) and court-ordered rights (rights arising from court orders). Professor Smith argues that, of these three categories, only ordinary rights might plausibly be explained by the principle of corrective justice. Ordinary rights themselves rest on pluralist foundations in the sense that a plurality of qualitatively different values are required to explain them. Professor Smith does not see either of these insights as an embarrassment for corrective justice theory. On the contrary, Professor Smith argues that corrective justice is more complex than is sometimes assumed and does not need to explain the entirety of private law in order to make a significant contribution to our understanding of the subject. The author of chapter six, Robert Stevens, would presumably subscribe to value pluralism of the kind identified by Stephen Smith, since he accepts that rights arise for a range of different reasons.36 In chapter six, Professor Stevens firmly rejects a pluralist approach that draws on a multiplicity of different policy goals or a combination of moral rights and instrumental considerations.37 He insists that ‘Private law is simply about the rights we have one against another’. Professor Stevens’ concern in his essay is with the question of what happens when rights come into conflict with one another. The law does not, Professor Stevens argues, resolve these conflicts by reference to some higher social policies or goals, but through a balancing process that establishes a hierarchy of rights. Donal Nolan considers in chapter seven whether tort law’s treatment of problems of causal indeterminacy can be reconciled with a strict rights-based or corrective justice approach.38 From a corrective justice viewpoint, causation is fundamental to liability in private law because it establishes the connection between the plaintiff and defendant that is essential to justify liability. Mr Nolan shows that it is very difficult to reconcile any relaxation of the rules of causation in cases of causal indeterminacy with the principle of corrective justice. Cases 33 S Perry, ‘Professor Weinrib’s Formalism: The Not-So-Empty Sepulchre’ (1993) 16 Harvard Journal of Law and Public Policy 597, 618–19. 34 Anns v Merton London Borough Council [1978] AC 728 (HL) 751. 35 S Smith, ‘The Rights of Private Law’ ch 5 of this book. 36 Stevens, ‘Torts and Rights’, above n 5, at 329–37. 37 R Stevens, ‘The Conflict of Rights’ ch 6 of this book. 38 D Nolan, ‘Causation and the Goals of Tort Law’ ch 7 of this book.

8

Andrew Robertson

such as Fairchild v Glenhaven Funeral Services Ltd,39 in which defendants have been held liable for loss that cannot strictly be shown to have been caused by their wrongdoing, can only be explained on the basis of either a concern to produce just outcomes in individual cases or, as seems to be the case, the pursuit of social goals such as loss-spreading and deterrence. Relaxation of the rules of causation to further social goals is undesirable, Mr Nolan argues, because it produces incoherence and has no logical stopping point.

IV. The Role of Goals in Private Law The essays in the third part of the collection embrace the idea that community welfare goals have an important role to play in private law. Steve Hedley draws a distinction between internalist and externalist perspectives on private law.40 He identifies four overlapping strands of internalist thinking in private law scholarship: taxonomy, rights theory, corrective justice theory and interpretive theory. Professor Hedley suggests that these strands of thinking are, broadly speaking, hostile to the idea that the institutions of private law have purposes. In Professor Hedley’s externalist camp are both those who write about private law from an external perspective (such as history, sociology or economics) and those who accept the relevance of external or public concerns within it. From a corrective justice perspective, public concerns are ‘external’ to private law in the sense that they are external to the relationship or interaction between the parties that structures and gives content to private law. Professor Weinrib, for example, rejects the idea that private law can be understood to pursue ‘goals’ or ‘purposes’, in part because goals and purposes are external referents and private law can only be understood from within.41 Professor Hedley argues that both internal and external perspectives are necessary elements of a proper understanding of private law and advocates a more engaged dialogue in private law scholarship between the internalists and the externalists. In chapter nine, Charlie Webb explores the contribution made by legal taxonomy to the goal of consistency, or ‘treating like cases alike’.42 The main focus of Dr Webb’s essay is on dispositive, rather than expository classification. He is principally concerned with taxonomies that seek to influence the content of legal rules, help decide cases and influence the future development of the law. Dr Webb argues that the development of a dispositive classificatory scheme cannot be value-neutral. It necessarily involves identification of the features of particular categories of cases that will inform their classification. This leads to 39

Fairchild v Glenhaven Funeral Services Ltd [2002] UKHL 22; [2003] 1 AC 32. S Hedley, ‘Looking Outward or Looking Inward? Obligations Scholarship in the 21st Century’ ch 8 of this book. 41 Weinrib, above n 4, especially at 3–6. 42 C Webb, ‘Treating Like Cases Alike: Principle and Classification in Private Law’ ch 9 of this book. 40

Introduction: Goals, Rights and Obligations

9

the conclusion that whether cases are ‘alike’, and thus belong in the same category in a dispositive scheme, requires consideration of the goals and functions of the law and attention to the question of how best to achieve them. Roderick Bagshaw suggests that we can analyse the ‘goals’ of a legal institution from a different perspective.43 A goal could be understood as a yardstick by which we can assess whether a legal institution is worth having. Following this approach, in ‘Tort Law, Concepts and What Really Matters’, Mr Bagshaw takes the goals of tort law to be the reasons why we should continue to tolerate its existence. From this perspective, Mr Bagshaw suggests, the primary goal of tort law is to make the world a better place than it would be without tort law. The focus of chapter ten is on the competition between different forms of concepts in tort law. The author’s particular concern is with the way in which the choices between different forms of concepts affect the capacity of tort law to meet its goal of making the world a better place. In chapter eleven I make an argument for the legitimacy of community welfare goals in private law.44 Anti-instrumentalist corrective justice scholars have argued that judges who take account of community welfare considerations exceed their judicial role and improperly act as legislators. Instrumentalist law-and-economics writing also commonly proceeds as though judges were unconstrained in the pursuit of policy goals. Chapter eleven represents the beginnings of an attempt to develop an account of the constraints on the pursuit of community welfare goals in private law. It explores the via media between unconstrained instrumentalism and the policy-free anti-instrumentalism of corrective justice and rights-based theories. Judges can legitimately take account of policy considerations if they act in a constrained judicial capacity and not in an unconstrained legislative capacity. The anti-instrumentalist attack on the legitimacy of policy is misguided because it ignores those constraints. A purely functionalist analysis of private law doctrines, on the other hand, is unhelpfully incomplete to the extent that it fails to take account of relevant constraints.

V. Community Welfare Goals in Private Law Doctrines The essays in the fourth part of this book look more specifically at the role of some particular community welfare goals in relation to particular private law doctrines. For the most part the authors in this section also unhesitatingly embrace the idea that policy goals have an important role to play in private law. In chapter twelve, Erika Chamberlain takes up Allen Linden’s idea of the ombudsman

43 44

R Bagshaw, ‘Tort Law, Concepts and What Really Matters’ ch 10 of this book. A Robertson, ‘Constraints on Policy-Based Reasoning in Private Law’ ch 11 of this book.

10

Andrew Robertson

function of tort law in its application to police.45 Justice Linden has suggested that tort law can operate as an instrument of social pressure on government institutions by exposing poor practices and abuses of power and provoking change to policies and practices. Tort law is said to provide psychological as well as political empowerment both to individual victims of police failures and disadvantaged groups. Professor Chamberlain’s particular focus is on negligence by police in the course of their investigation of crime, which can affect victims of crimes that might have been prevented, as well as the wrongly accused. Professor Chamberlain’s chapter explores the tension that exists in the negligent investigation cases between the ombudsman function of tort law and other policy goals articulated by the courts, such as ensuring that tort law does not unduly interfere with police discretion or encourage police to adopt defensive practices. Deterrence is perhaps the most widely discussed goal in private law. It plays a particularly prominent role in tort and fiduciary law as well as illegality doctrines and in the analysis of gain-based remedies. As Yock Lin Tan observes in chapter thirteen, the role of deterrence is a question that must be considered in any serious study of the goals of private law.46 Professor Tan argues that deterrence cannot be seen as a principle of distributive justice and is better understood as a tax imposed on certain activities. If deterrence is correctly understood as a tax, Professor Tan argues, then deterrence in private law ought to be structured as such. Professor Tan takes fiduciary law as the testing ground for his ideas about the role of deterrence. He argues that the deterrence rationale for the ‘no profit’ and the ‘no conflict’ rules in fiduciary law ‘is redundant and unconvincing’ and has marred the development of the law in this area. The goals of fiduciary law and the role of deterrence are also considered by Matthew Harding.47 The focus of Dr Harding’s chapter is on the disgorgement of profits resulting from a breach of fiduciary duty, and in particular on the justification for allowing fiduciaries to retain a proportion of the profit where that profit is partly attributable to the fiduciary’s own ‘skill, efforts, property and resources’.48 Dr Harding rejects suggestions that fiduciary allowances can be justified on the basis of causation or remoteness, and argues that they are best understood to be based on the goal of ‘satisfying desert’, or ‘giving to people what they deserve’. In his essay Dr Harding explores the tensions between the broader, forward-looking goal of deterrence that underlies the account of profits and the more specific, backward-looking goal of satisfying desert that underlies the making of allowances to defaulting fiduciaries. Anthony Duggan offers a different perspective on the role of deterrence in the granting of gain-based remedies for breach of fiduciary duty and other wrongs.49 45

E Chamberlain, ‘Negligent Investigation: Tort Law as Police Ombudsman’ ch 12 of this book. YL Tan, ‘Deterrence in Private Law’ ch 13 of this book. 47 M Harding, ‘Justifying Fiduciary Allowances’ ch 14 of this book. 48 Warman v Dwyer (1995) 182 CLR 544 (High Court of Australia) 561. 49 A Duggan, ‘Gain-Based Remedies and the Place of Deterrence in the Law of Fiduciary Obligations’ ch 15 of this book. 46

Introduction: Goals, Rights and Obligations

11

Professor Duggan suggests that the unsettled state of Canadian law on proprietary and personal gain-based remedies for wrongs is at least partly the product of ambivalence about the respective roles of corrective justice and deterrence in this area of the law. For those who accept that a gain-based remedy can be justified on the basis of deterrence in a case where the defendant’s enrichment does not occur by subtraction from the assets of the plaintiff, bankruptcy considerations complicate the policy analysis. On the one hand, bankruptcy of the defendant may blunt the deterrent effect of the remedy. On the other hand, a rule that requires a different remedial outcome in the event of bankruptcy may create undesirable incentives. Craig Rotherham uses the context of gain-based relief in nuisance to analyse the normative foundations of restitution for wrongs.50 Professor Rotherham argues that formal or internal justifications for restitution for wrongs, which have their basis in the nature of the underlying rights or the ‘immanent doctrinal logic of particular torts’, rest on the drawing of distinctions that do not stand up to close scrutiny. He argues that restitution for wrongs must be regarded as resting on a more complex and largely pragmatic set of considerations, centred on the goals of punishment and deterrence.

VI. The Goals of Unjust Enrichment Law The chapters in the last part of the book focus on the goals of the law of unjust enrichment. In chapter seventeen Hanoch Dagan considers the question of what distinguishes just from unjust enrichments.51 Professor Dagan argues that these distinctions are made on the basis of the values of autonomy, utility and community, which underpin the law’s approval of just enrichments and its disapproval of those enrichments considered unjust. Professor Dagan does not argue that these public values alone structure private law. He agrees with Professor Weinrib that correlativity between the plaintiff ’s entitlement and the defendant’s obligation requires private law to ‘be able to justify to defendants both the identity of the beneficiary of any liability imposed on them, and the exact type and degree (or magnitude) of that liability’. Professor Dagan nonetheless insists that ‘the reasons for the parties’ entitlements’ need not be—and often are not—‘wholly internal to the parties’ relationship’. Thus, he concludes that private law is neither ‘an autonomous realm with its own inner intelligibility, isolated from the social, economic, cultural and political realms’ nor a territory governed by the forces of ‘brute instrumentalism’.

50 C Rotherham, ‘The Normative Foundations of Restitution for Wrongs: Justifying Gain-Based Relief for Nuisance’ ch 16 of this book. 51 H Dagan, ‘Just and Unjust Enrichments’ ch 17 of this book.

12

Andrew Robertson

Emily Sherwin’s chapter explores the relationship between the rules of private law and the rationales or social values that underlie them.52 While Professor Sherwin does not disagree with Professor Dagan’s identification of the values underlying the law of unjust enrichment, she questions whether Professor Dagan’s analysis accords too great a role for those values in legal decision-making, at the expense of the authority of the rules that instantiate them. Professor Sherwin’s view is that, once a rule is in place, the goals or values that underlie the rule have completed their work. She takes issue with the idea that a judge applying a legal rule should ensure that the outcome produced by the application of the rule is not significantly at odds with its purpose. In chapter nineteen Tang Hang Wu analyses the stories with which Peter Birks illustrated his writing on unjust enrichment.53 Dr Tang suggests that Birks’ stories ‘inform and shape many of the doctrines found in the law of unjust enrichment’. There is, Dr Tang suggests, an important relationship between Professor Birks’ stories and the goals of unjust enrichment law, particularly furthering autonomy, promoting security of receipt and promoting rationality and coherence. Dr Tang seeks to unsettle some of Professor Birks’ more important stories by offering counter-narratives and, in doing so, seeks to force a reassessment of some of the assumptions underlying the law of unjust enrichment. In the final essay in the collection Graham Virgo seeks to demolish Professor Birks’ pyramidic ‘absence of basis’ theory of the law of unjust enrichment.54 Professor Virgo argues that whether there was an effective basis for the transfer in question is an important issue to be considered in an unjust enrichment claim, and this is well established in the case law. It is, however, better understood as an adjunct to the ‘unjust factors’ approach rather than as an alternative to it. Whether there was a basis for the transfer can, Professor Virgo argues, be accommodated within the traditional linear approach to the unjust enrichment inquiry, along with the question of whether the claimant has taken the risk that there might not be a basis for the transfer.

VII. Conclusion The essays in this collection contribute to a number of important debates about goals and functions of private law. They address questions about whether private law should be seen as having goals, about whether any such goals are particular to private as opposed to public law, about the legitimacy of the pursuit of community welfare goals in private law, about the place of goals and purposes in private

52

E Sherwin, ‘The Rules of Obligations’ ch 18 of this book. HW Tang, ‘Storytelling in the Law of Unjust Enrichment’ ch 19 of this book. 54 G Virgo, ‘Demolishing the Pyramid—The Presence of Basis and Risk-Taking in the Law of Unjust Enrichment’ ch 20 of this book. 53

Introduction: Goals, Rights and Obligations

13

law scholarship and about the relationship between goals and the other influences that shape private law. Perhaps most significantly, they contribute to debates about the role that particular policy goals do and should play in particular private law doctrines and about the relationship between community welfare goals and considerations of interpersonal morality arising from the interactions between individuals. Together the essays in this collection show that, while the idea of purpose in private law is controversial, attention to these questions about goals casts significant light on the nature of private law and its constituent doctrines.

2 The Mutually Constitutive Nature of Public and Private Law MAYO MORAN

I. Introduction There has been a surge of interest in the relationship between public and private law. The literature on this topic, formerly quite thin, has suddenly blossomed. The resulting debate, which extends across much of the common law world and beyond, is preoccupied with how to understand the increasingly complex nature of the relationship between public and private law. Though there are of course many nuances in this debate, much of the discussion is dominated by an approach focused on the implications of public law for private law. It is only the occasional private law enthusiast that suggests that private law may be something more than convenient machinery for public law values. It is true that the debate has, as a general matter, done much to illuminate what is at stake in the relationship between public and private law. Nonetheless, the dominant alternatives share a critical deficiency. That deficiency proceeds from the shared assumption that the flow of values through public and private law is essentially uni-directional. That is, either public law or private law is the critical locus of value but not both. But this underlying assumption, I shall argue, ought to be rethought. In fact the relationship between public and private law is better understood as deeply integrated and mutually constitutive. Since much of the literature has been generated by public law enthusiasts, keen to ensure that the new constitutionalism has full reign in the legal system, it is unsurprising that the dominant approach in the contemporary debate is the ‘public law positive’ approach to the relation between public and private law. Underlying this view is the idea that there is a hierarchy in law which locates the significant source of values and principles at the top of that hierarchy, preeminently in constitutional and international bills of rights. According to this view, which has been in the ascendancy in thinking on this subject, private law is useful in providing the machinery for enforcing or giving institutional support to key legal values. Thus, in this context we see the idea that important legal values can be delivered or realised through private law. Private law is thus understood

18

Mayo Moran

as a mechanism for channelling important legal values from above through to individual interactions. But what is critical to note, in my view, is that this approach does not see private law itself as the source of important legal values. The values that really matter instead come from above, typically in the form of guarantees of rights. Although recent developments in constitutional law, particularly in the growing number of jurisdictions that have adopted constitutionally entrenched bills of rights, have ensured that this ‘public law positive’ approach dominates, there is also a faint note of resistance. Some commentators do insist that the structure and integrity of private law are significant and important in their own right.1 Thus, they suggest that there are reasons to be cautious about any significant pressure—including constitutional pressure—from beyond its confines. Such pressure, some critics suggest, threatens to distort private law and distract from its proper focus on the interaction of the parties. So this counterpoint to the enthusiasm for constitutional imperialism is relatively resistant to the idea that public law ought to affect the articulation of rights and duties in private law. If the public law approach imagines law as a hierarchy of values with constitutional values emanating down through the legal system and into the otherwise banal corners of private law, the most that private law devotees seem to hope for is that private law will be relatively insulated from the intrusions of public law. What I want to suggest here, however, is that both of these views miss something important. Indeed, as I shall argue, this is no coincidence because, although they differ in many things, at bottom the two views share an impoverished understanding of the ways in which legal norms interact with each other across the legal system. Thus, although each view captures something useful, they also obscure as much as they reveal. An alternative more adequate account of the public-private law relation ought instead to attend to the interaction between different elements of the legal order. Thus, as I shall suggest, the key values of our legal system critically depend not simply on contributions from both public and private law but also on a distinctive kind of interplay between those domains. My argument here accordingly proceeds along the following lines. First, I briefly explore the dominant understanding of the public-private law relation, noting what is usefully captured but also what is obscured. We can glean something important, I suggest, by attending to what seems wrong with this approach. This then helps to provide a foundation for an alternative account that posits that public and private law can best be viewed as, at bottom, mutually constitutive. This approach is consistent with the idea that public and private law do each have their own distinctive structures and can hence rightly be viewed as irreducible to 1 LE Weinrib and EJ Weinrib, ‘Constitutional Values and Private Law in Canada’ in D Friedmann and D Barak-Erez (eds), Human Rights in Private Law (Oxford, Hart Publishing, 2002) 43; A Barak, ‘Constitutional Human Rights and Private Law’ also in Human Rights in Private Law, at 13.

Public and Private Law

19

each other. So the truth in the ‘separate spheres’ idea is found in the insight that neither public nor private law is subservient to the other. But this truth often reads distinctiveness as separateness and hence insists on the impermeability of private law. This, I shall suggest, is a mistake. Private law not only tolerates but indeed requires a certain kind of openness to the larger legal system and it is this dimension that the separate spheres view seems incapable of properly rendering. This ought not to suggest, however, as it is so often taken to, that the really important locus of value is found in the overarching principles of constitutional and international law. While the nature of those influences on private law may suggest normative priority, the meaning of that priority cannot be understood without reference to the very private law over which such influence is exerted. Indeed, recognising how this is so points up the weakness in the public law-focused hierarchy approach and points instead towards an approach which recognises the mutually constitutive nature of public and private law and the much more interactive quality of the relation between them. One virtue of such a view, I will suggest, is the way that it recognises the integrity of public and private law while simultaneously noting how the dynamic relation between them can, at its best, enhance each in its own terms and in so doing contribute to the positive development of the legal system.

II. Hierarchy of Values: Understanding the Dominant Approach It is hardly surprising that the hierarchy of values approach dominates the understanding of the public-private law relation. The same enthusiasm for the extension of constitutional human rights that has fuelled the expansion of constitutionalism has also had an impact in the private sphere. Thus, the argument goes, constitutional guarantees ought not to be limited to the realm of state action but ought also to exert an influence over private law. This argument derives much of its force from the fact that the recent impetus to constitutionalise basic human rights has typically been a response to flagrant violations of those rights by public and private actors alike. So it seems plausible that constitutional provisions might be thought to be the solution not only to invidious state action but also to private law’s openness to discrimination and other ills. And so for this reason, there is considerable enthusiasm for the idea that democratically or politically articulated values ought to flow into private common law in order to correct its errors. This approach does have much to recommend it from both a normative and a jurisprudential point of view. There is no doubt, for instance, that private law has often worked against the basic ideals of equal human dignity that underlie constitutional human rights. And there is force in the view that courts in constitutional regimes should not exercise their authority in a way that undercuts

20

Mayo Moran

the basic commitments of the constitution.2 It is largely for these reasons that courts in constitutional democracies have adopted the view that constitutional rights do indeed have significant implications for private and common law. There are a variety of possible positions here and commentators have debated the nuances in great detail. However, the basic approach is clear—in most newly constitutionalised regimes courts have held that constitutional rights have at least a significant ‘horizontal’ effect.3 So notwithstanding the absence of state action in its classical form, constitutional rights do nonetheless have significant implications for how courts develop and apply a whole array of private law rights. In most newly constitutionalised jurisdictions, the impact of constitutional rights in the private sphere takes the form of a mandatory influence also referred to as indirect horizontal effect. Elsewhere I have argued that this is a distinct model of legal authority best thought of as ‘influential’.4 The influential authority of constitutional norms can perhaps best be understood by contrast with its two extreme alternatives. The first alternative is found in its most pristine form in the United States. This ‘vertical’ view holds that constitutional rights require state action and that such rights are entirely separate from and unrelated to private law.5 Contrasting with this is the ‘direct application’ or fully horizontal approach. This approach, which seems to be a unique feature of the Irish Constitution, allows a person, at least in some circumstances, to seek redress against private parties for the infringement of certain constitutionally protected rights.6 However in most newly constitutionalised regimes, courts have not embraced either of these alternatives. They have instead adopted the view that constitutional rights serve as a distinctive kind of mandatory influence that shapes the articulation of private law. This is in essence what is often termed indirect horizontal effect and it prevails in Canada, South Africa, Israel and in the United Kingdom under the Human Rights Act 1998, as well as elsewhere. This approach holds that constitutional human rights do not generally apply directly to private interactions between purely private parties, nor do they apply 2 A Barak, ‘Constitutional Human Rights and Private Law’ (1996) 3 Review of Constitutional Studies 218, 254–7 and 279–80. I explore this idea in more detail in M Moran, ‘Authority, Influence and Persuasion: Baker, Charter Values and the Puzzle of Method’ in D Dyzenhaus (ed), The Unity of Public Law (Oxford, Hart Publishing, 2003) and M Moran, ‘Inimical to Constitutional Values: Complex Migrations of Constitutional Rights’ in S Choudhry (ed), The Migration of Constitutional Ideas (Cambridge, Cambridge University Press, 2006). 3 For the leading article on this issue see M Hunt, ‘The Horizontal Effect of the Human Rights Act’ [1998] Public Law 423. 4 Moran, ‘Authority, Influence and Persuasion’, above n 2. 5 While the American model does insist on a stricter separation than that adopted in the new constitutional models, there are also important qualifications to the American state action doctrine: Barak, above n 2, at 254–7; and Moran, ‘Inimical to Constitutional Values’, above n 2, at 247–51. 6 Hunt, above n 3, at 428 quoting Walsh J in Meskell v Cors Iompair Eireann [1973] IR 121 (Supreme Court of Ireland) 132–3.

Public and Private Law

21

directly to the common law.7 This means that, unlike in Ireland, individuals cannot claim damages against other private individuals for infringements of constitutional rights. Instead, constitutionalised human rights affect private law in a distinctive way. Although constitutional rights do not apply directly and hence have no force in private law, they nonetheless exert a distinctive kind of mandatory interpretive pressure. This pressure means that constitutional values not only should be but must be respected in the application and development of private law. One example of this is found in the approach that courts in Canada, England and South Africa have taken to the question of the impact of the law of defamation on the right to freedom of expression where the allegations of libel or defamation concern public figures.8 Although in all cases the courts have insisted that constitutional rights of freedom of expression do not apply directly in private law, they have simultaneously required that the private law of libel and defamation be developed in light of the overarching importance of the constitutional guarantee of freedom of expression. Similarly, the House of Lords ruled that the Article 8 right to privacy under the European Convention on Human Rights was relevant to the law of confidence and privacy. Again, although the Convention right was not directly applicable to private relations, it did give rise to a ‘reasonable expectation of confidence’ that expressed itself through the traditional cause of action in private common law.9 In a like manner, the right to equality may be relevant to determining the kinds of agreements a court might be called upon to enforce or to the extent of the deference granted to the exercise of political choices by public authorities.10 This then is the jurisprudential basis on which the dominant approach to the relationship between public and private law is founded. Given the mandatory language of these judicial pronouncements, it is not surprising that commentators have embraced the view that private law stands to be corrected by public law. Indeed, in a landmark ruling on the application of constitutional rights to private and common law by the Constitutional Court of South Africa, a number of judges specifically noted that were this not the case, the danger would be that the common law would ossify and hence enshrine previously existing discrimination.11 And in fact, following these landmark rulings, courts have clearly applied

7 See, for instance, in Canada, Dolphin Delivery [1986] 2 SCR 573 and Hill v Church of Scientology [1995] 2 SCR 1130; in South Africa, Du Plessis v De Klerk (1996) 3 SA 850 (CC); and in England, Campbell v MGN Inc [2004] 2 AC 457 (HL) and Douglas and Zeta-Jones v Hello! Magazine [2001] QB 967 (CA). 8 See, for example, Hill v Church of Scientology [1995] 2 SCR 1130. In other jurisdictions, see Du Plessis v De Klerk (1996) 3 SA 850 (CC); Douglas and Zeta-Jones v Hello! Magazine [2001] QB 967 (CA); A v B [2003] QB 195 (CA). The foundational case here is, of course, the US Supreme Court decision in New York Times v Sullivan, 376 US 254 (1964). 9 Campbell v MGN Inc [2004] 2 AC 457 (HL). 10 Canada Trust Co v Ontario Human Rights Commission (1990) 69 DLR (4th) 321 (Ontario CA); Jane Doe v Toronto (Metropolitan) Commissioners of Police (1989) 58 DLR (4th) 396 (Ontario High Court of Justice). 11 Du Plessis v De Klerk (1996) 3 SA 850 (CC) [85]–[86] (Mahomed DP).

22

Mayo Moran

and shaped private law rights to maximise consistency with constitutional values. Since the Canadian Charter of Rights and Freedoms is one of the most developed of the new wave of post-war constitutions, it is perhaps unsurprising that Canada has a particular wealth of cases on this subject. To illustrate the extent of the impact of public law values on private law, let me give a few illustrations. In the law of tort, there are a number of cases that illustrate this impact but the Jane Doe case stands out as particularly important.12 In Jane Doe, a woman brought an action against the Toronto police for negligence when she was attacked by a serial rapist. The police had identified her and some other women in the area as likely targets of the rapist but decided not to warn them because they thought that the women might panic and alert the rapist. The rapist was arrested after he assaulted Jane Doe. She claimed that in failing to warn her, the police had used her as bait. In their response, the police invoked the second stage of the Anns13 test, arguing that policy reasons precluded a duty of care from arising. They insisted that to find otherwise would be to allow courts to intrude into the politically sensitive realm of policy choices. However, the Ontario courts refused to adopt this line of reasoning, instead awarding Jane Doe the right to sue the police. The reasoning in Jane Doe can be taken to illustrate one important effect that constitutional norms and values have on private law—such values limit or restrain the kinds of policy reasons that can serve to negate the duty of care in tort. So, in a regime of constitutionalised human rights, although political authorities have considerable discretion to make choices in furtherance of their activities, they cannot act on reasons that are at odds with basic constitutional values. Thus in Jane Doe’s case it was not open to the police to invoke stereotypes about women as the basis for determining how to proceed with an investigation. Another illustration of the significance of background constitutional values in the law of negligence can be found in the controversial Canadian Supreme Court decision in Dobson v Dobson.14 This case involved a preliminary determination of whether there was a cause of action in a case where a child sued its mother for pre-natal injuries suffered as the result of the mother’s allegedly negligent driving. The arguments before the Supreme Court of Canada stressed how the history of the legal regulation of reproduction had discriminated against and oppressed women. A majority of the Supreme Court refused to find a duty of care on policy grounds, specifically noting this history and its impact on the autonomy of women. In light of this threat, the Court held, any such cause of action ought not to be created judicially but instead must be generated by a legislative act which could ensure that it was properly curtailed to respect the autonomy interests of pregnant women. So here the background equality values were again funnelled 12 Jane Doe v Metropolitan Toronto Police (1989) 58 DLR (4th) 396 (Ontario High Court of Justice). 13 Anns v Merton London Borough Council [1978] AC 728 (HL). 14 Dobson v Dobson [1999] 2 SCR 753.

Public and Private Law

23

through stage two of the Anns test. But while in Jane Doe constitutional values limited the kinds of policy reasons that public authorities could invoke, in Dobson the effect was more dramatic—attentiveness to constitutional values precluded a duty of care from arising at all. Background equality values also played an extremely important role in Re Eve.15 That case involved an application by a parent to the court to exercise its parens patriae jurisdiction to authorise the non-therapeutic sterilisation of a mentally disabled teenage girl. The argument in favour of sterilisation rested on reasons that were primarily social and pragmatic, such as the challenges of childbirth, the vulnerability of the girl, the difficulty that Eve would have caring for a child and the like. However, the Supreme Court found that these were not sufficient reasons to justify the deprivation of such a basic human right. The impact of background constitutional values of equality was again crucial to the Court’s finding. Like in Dobson, the Court noted the danger of discrimination. La Forest J stressed that courts must proceed with caution in cases where history—including jurisprudential history—alerted them to the danger of undervaluing the liberty interests of those with mental disabilities. Here, this dictated that the Court must adopt an approach which placed the highest value on those very liberty interests which it had historically been too quick to discount. Accordingly the Court refused to authorise the sterilisation.16 In addition to these examples from the law of tort, areas of law involving voluntary agreements such as contracts, wills, and trusts have also been affected in a range of ways by the impact of constitutional values. As in the law of torts, sometimes these values play an interpretive role, directing the courts, for example, to broader readings of value terms such as unconscionability, or to hybrid actions that blend those principles with fiduciary obligations.17 However, on occasion constitutional values may play a more dramatic role, which I refer to elsewhere as estoppel-like, precluding the court from enforcing otherwise formally valid contracts or trusts where the substance is fundamentally at odds with constitutional values.18 Although courts in common law jurisdictions did traditionally enforce racist or other discriminatory agreements, after the end of World War II they began to question this practice. Across several common law jurisdictions, courts began to resile from the prospect of implicating themselves in the enforcement of discriminatory private agreements. The 1945 decision of the Ontario High Court in Re 15

Re Eve (1986) 31 DLR (4th) 1 (Supreme Court of Canada). Note the contrast with the English case Re B (A Minor) (Wardship: Sterilisation) [1987] 2 All ER 206 (HL), which relied on the same reasons that the Supreme Court of Canada held were not adequate and were a source of concern on equality grounds. 17 Norberg v Wynrib [1992] 2 SCR 226 is an example here. 18 M Moran, ‘Time, Place and Value’ in D Dyzenhaus and M Moran (eds), Calling Power to Account: Law, Reparations and the Chinese-Canadian Head Tax Case (Toronto, University of Toronto Press, 2005); M Moran, ‘Influential Authority and the Estoppel-Like Effect of International Law’ in H Charlesworth et al (eds), The Fluid State (Sydney, Federation Press, 2005); Moran, ‘Inimical to Constitutional Values’, above n 2. 16

24

Mayo Moran

Drummond Wren is a particularly dramatic example of this.19 In that case, Mackay J concluded that a racially restrictive covenant that prohibited the transfer of the subject land ‘to Jews, or to persons of objectionable nationality’ was void because it was contrary to public policy.20 Though this case pre-dated the constitutional entrenchment of human rights in Canada, the decision is a clear precursor of and inspiration for later courts that draw on constitutional values.21 In Drummond Wren, MacKay J suggests that in addition to enshrining discrete rights and obligations, the legal system also contains certain mandatory values which figure in, amongst other things, what kinds of private agreements a court can enforce. As sources for such values, he specifically points to the United Nations Charter, which Canada had signed and ratified, as well as to the Atlantic Charter. He also identifies a number of domestic sources, which, although they lack force, are nonetheless relevant because of how they infuse the meaning of public policy.22 And it is ultimately on public policy grounds that he rests his finding that a racist contract cannot be enforced because it offends the most fundamental commitments of the contemporary legal order. In 1948, three years after Drummond Wren was decided, the United States Supreme Court was also confronted with a group of cases involving racially restrictive covenants. Shelley v Kramer,23 which joined a number of cases before the Supreme Court, involved neighbours who sought to prevent the sale of property to black purchasers and who sought the eviction of black property owners. Vinson CJ noted that unlike ordinary restrictive covenants, these covenants ‘are directed toward a designated class of persons’, a class ‘defined wholly in terms of race or color’.24 Equality in the ownership and enjoyment of property, he also pointed out, was an essential pre-condition to the realisation of other basic rights which the Fourteenth Amendment was designed to protect.25 Nonetheless, since the Amendment, like the rest of the Constitution, was directed only to state and not private action, the Court concluded that ‘the restrictive agreements standing alone cannot be regarded as a violation’ of any Fourteenth Amendment rights. Thus, to the extent that compliance with the restrictive covenants was accomplished by voluntary adherence, the Fourteenth Amendment could not reach it. In these cases, however, the individuals invoked the machinery of the state courts when they sought judicial enforcement of the covenants. In each case, Vinson CJ noted, the judicial enforcement of the restrictive covenants through the common law of the state ‘bears the clear and unmistakable imprimatur of the State’.26

19

Re Drummond Wren [1945] 4 DLR 674. ibid, at 675. For an elaboration of the link between the constitutional variant of such an approach and the older public policy jurisdiction see Moran, ‘Inimical to Constitutional Values’, above n 2. 22 Re Drummond Wren [1945] 4 DLR 674, 677. 23 Shelley v Kramer 334 US 1 (1948). 24 Shelley v Kramer 334 US 1 (1948) 10. 25 ibid. 26 ibid, at 20. 20 21

Public and Private Law

25

So the Supreme Court held that the Fourteenth Amendment prevented state enforcement of racially restrictive covenants.27 In Canada, the basic approach of Shelley v Kramer was adopted in Canada Trust Co v Ontario Human Rights Commission.28 There, post-Charter Ontario courts had to consider the contemporary validity of the ‘Leonard Foundation Trust’, an explicitly racist charitable trust established in 1923. The trust provided educational scholarships but excluded from its benefit those who were ‘not Christians of the White Race’. It was formally valid and had been in operation for many decades. However, its terms gradually began to seem troubling and when it finally came before the courts in the late 1980s, the primary question was whether it violated public policy. At the time of the formation of the trust, the terms ‘would have been held to be certain, valid and not contrary to any public policy’.29 On this ground, the trial judge upheld the trust, noting that it suffered from no positive defect at the time of its formation. However, the Ontario Court of Appeal unanimously held that a court’s ability to enforce a trust is limited by ‘current principles of public policy under which all races and religions are to be treated on a footing of equality’ (emphasis added).30 The reasoning invoked the provincial human rights code, other legislation and policy, international covenants and especially the provisions of the Charter of Rights and Freedoms that guarantee equality rights and multiculturalism. On this basis, the Court of Appeal found that the terms of the trust must be altered to avoid conflict with contemporary constitutional values. Thus in a range of cases across jurisdictions, courts have drawn on mandatory constitutional values to profoundly affect the shape and meaning of private law. As even these few examples illustrate, in contemporary constitutional orders, public law values exert an extremely important effect on private law. And though courts and commentators sometimes question the exact configurations of the obligation, they largely agree that constitutional values serve as a distinctive kind of mandatory source that exerts constitutional demands on private law. Thus, whether constitutional values infuse the meaning of important value terms such as public policy, reasonableness and other terms or whether constitutional values appear in their more dramatic estoppel-like guise, they are accepted to be of vital importance to the law of private relations. Because it is so clear that constitutional values, when relevant, do exert a distinctive kind of mandatory effect on private law, the dominant view of the public-private law relation is that it is a hierarchy. Thus, the understanding that comes from the jurisprudence and the commentators is largely that private law can be left alone so long as it does not run afoul of constitutional law values. On this hierarchical view, public law alone is the important locus of value. 27 The legacy of Shelley is undoubtedly complex and many American commentators now regard it as wrong. See my detailed discussion in Moran, ‘Inimical to Constitutional Values’, above n 2, with references to the very extensive American literature on the subject. 28 Canada Trust Co v Ontario Human Rights Commission (1990) 69 DLR (4th) 321 (Ontario CA). 29 ibid, at 336. 30 ibid, at 335 (Robins JA).

26

Mayo Moran

III. An Alternative View of The Public-Private Law Relationship The dominant position regarding the public-private law relation in this sense focuses on how values flow from public law to private law. This is a very important perspective, well established in the jurisprudence of many contemporary constitutional regimes. Thus, it seems undeniable that public law values do exert a very significant effect on private law. However, because both the metaphors and the underlying ideas invoke a hierarchy of norms, this approach all too often casts private law simply as a convenient means of ‘delivering’ the really important values which come, presumptively, from public law. As noted above, there are some other strands in this debate. Thus, an occasional commentator will insist on the integrity of private law—on its distinctive conceptual structure and aims—and on that ground will resist extensive application of constitutional norms as alien to or incompatible with private law. Similarly, there are those who see something valuable in the relation between public and private law even when viewed in light of their distinctive aims and structures. Because of the overarching values that permeate both, private law is legitimately open to and influenced by public law values in distinctive ways. The language and metaphors of these commentators are more integrative, explicitly noting how public law values can legitimately be woven into the fabric of private law.31 And while this approach does move closer to an integrative understanding of the relation of public and private law, there is still a feature that it leaves unexplored. It is that feature which I would like to elaborate here. All of the views of the public-private law relationship seem to take for granted the basic idea that to the extent that there is a flow of values, that flow is downward, from public law norms such as constitutional rights and human rights, into private law. Commentators do differ regarding how open private law is to public law values as well as on the extent to which those values resonate with the basic commitments of private law. However, what remains unexplored is the idea that the relationship between public and private law may not be uni-directional. Instead, private law may also contribute something of vital importance to public law guarantees. So, on this view, while public law values do exert critical pressure over private law, private law rights also turn out to be vital to understanding the scope of constitutional guarantees. If public law helps to determine the shape of private rights, private rights in turn often fill out the content of constitutional guarantees, particularly where fundamental human rights are concerned. This

31 See Weinrib and Weinrib, above n 1 (more focused on the private law dimension); Barak, above n 2 (more focused on the points of opening in private law to public law); S Waddams, Dimensions of Private Law: Categories and Concepts in Anglo-American Legal Reasoning (Cambridge, Cambridge University Press, 2003) (noting how public law values have always been part of private law reasoning).

Public and Private Law

27

in turn suggests that the relationship between public and private law is not uni-directional—instead it may be more apt to suggest that these fields inform and even constitute each other in very fundamental ways. An important illustration of the interactive quality of public and private law is found in the relationship between constitutionalised human rights and the private law idea of the person. Indeed, though seldom noticed as such by contemporary commentators, the concept of the person that is so critical to the entire framework of constitutionalised human rights is, in very fundamental ways, a construct of private law. One implication of this is the fact that constitutional guarantees cannot be fully understood without reference to the content that derives from the private law notion of the person. Thus, noting how the idea of legal personhood cuts across public and private law helps to reveal a very important missing feature of the relationship between these two domains. Some elements of the public-private law relationship have been explored quite thoroughly. This is particularly true, as noted above, of the impact of public law values on private law. Further, as we have also seen, there are commentators who have suggested a more integrative understanding of public and private law. However, the dimension of the picture that has not been similarly explored is the idea that private law is not simply an apt vehicle for public law values but that it actually contributes something very important to public law. In order to explore this missing dimension of the public-private law relation, I will examine the relationship between the legal content of the person and the constitutional guarantee of equality. It turns out, I will suggest, that the (evolving) private law idea of the legal person provides vital input without which the meaning of constitutional equality guarantees cannot be understood. Indeed, the failure to recognise the extent to which private law rights constitute the person that is protected by constitutional guarantees of equality and non-discrimination has arguably resulted in unnecessary difficulties in the articulation of what many commentators count as the most foundational constitutional guarantee. Thus, in the equality context, the underlying belief that the legal person is essentially a public law construct has, among other things, led courts to invoke highly subjective and problematic equality tests. Attending to the bundle of rights, including prominently private law rights which have historically made up what we count as the full legal person, would, by contrast I argue, provide a securer foundation for analysis under the equality guarantee. In order to explore this idea, it is vital to have a sharper sense of who exactly is the person in law. What makes up the status that we call full personhood? This very question may strike us as odd in no small part because the concept of personhood is so foundational that we tend not to think of what actually constitutes it. Because we can often glean much about what something is by paying attention to what it is not, I will seek to flesh out at least some of the vital legal content of personhood by looking at a notoriously problematic case of personhood— the treatment of women, particularly married women. By asking what women who married lost when they lost the status of person in law, we can begin to see what exactly it meant in legal terms to be a person. What was lost when women

28

Mayo Moran

who married slipped out of the category of person? Noting in some detail exactly how married women lost legal personhood and what specific legal rights this implied will help to cast light on the inadequacy of thinking of the legal person protected by constitutional equality guarantees as essentially a creature of public law. Instead, as we shall see, the shifting status of women who married—their movement into and then eventually out of a property-like status—is in large measure composed of changes in the rules of private law that governed their civil relations. Married women’s loss of legal personhood was not accomplished through a shift in their public law rights but rather through their exclusion from the bundle of civil rights, including the right to bodily integrity, the right to hold property and the right to contract, which together compose the incidents of legal personhood and at bottom distinguish persons from things.

A. Public Law Equality and the Private Law Person Given the recent rise of human rights-oriented constitutionalism, it seems unsurprising that contemporary accounts of equality are preoccupied with the force of public law. This is revealed, as noted above, in the fact that many of the discussions about the relation of public and private law take for granted the idea that most of the central normative content, particularly with regard to human rights, derives from public law and can then be funnelled into private law. This predisposition to emphasise constitutional or public law as central to struggles around human rights is also apparent in the discussions of the human rights revolution and the gradual progress of the law towards greater attentiveness to the individual legal person. Indeed these discussions tend to focus on constitutional law rights and guarantees to the almost complete exclusion of any attention to private law rights. An illustration can be found in the approach taken in many of the histories of women’s rights. What is particularly noteworthy for our purposes is the emphasis that many accounts place on the struggle for the right to vote along with other rights associated with public participation. Of course, the struggle for suffrage was an extremely vital and critical effort in the extension of rights to women, and it did go, in profound ways, to their status as full members. But it is also the case that the right to vote constitutes citizenship, and is in that sense distinguishable from the rights implicated in legal personhood. Being a citizen, associated with rights of political participation, military service and the like, is thus not identical to being a person in law. In fact, the example of the historical treatment of married women illustrates that natural persons could possess various forms of status, only one of which could be called full legal personhood.32 So it is significant that 32 I use the term personhood to distinguish my discussion from the question of legal personality, which is typically seen as the means by which entities which are not natural persons can nonetheless come to be legal persons (typically through incorporation). What I am interested in here, by contrast, is how natural persons come to be legal persons. The categories of legal and natural person have, of course, not historically been co-extensive and many critics would argue that they are not yet co-extensive today.

Public and Private Law

29

the movement of married women through the nineteenth and twentieth centuries from a property-like status to something more like full personhood took place largely through the common law of private relations. In fact, shifts in this area form a distinct and important trajectory that relates to, but is not co-extensive with, efforts to change the constitutional rights of women to enable them to participate as citizens. Noting this distinction between the public law idea of the citizen and the private law idea of the person turns out to be critically important. In an era of postwar constitutionalism premised on the innate human dignity of all individuals, the idea of the person is the cornerstone of much constitutional human rights analysis. But because so much of the idea of the person is articulated through private law and because the gaze of courts and commentators is fixed so steadily on public law, it has been difficult to elaborate the meaning of the equality guarantee. One result is that courts and commentators have largely been unable to get beyond the abstractions of the idea of innate human dignity and to find a way to assess the significance to legal personhood of an alleged violation of constitutional equality. The argument here suggests that there is a reason for this difficulty. Public law rights create the citizen, not the legal person. However, through the guarantee of equality and elsewhere, public law also enshrines a commitment to the idea that natural persons are presumptively also full legal persons. And in so doing, my argument goes, public law guarantees of equality necessarily draw in the content of legal personhood, created in no small measure through the common law of private relations.33 I think it is arguable that, although unrecognised, some features of this connection may be implicit in certain elements of equality analysis.34 However, the failure to recognise the significance of this private law content has

33 The underlying idea of the public-private law relation that we see is at bottom premised on an older account of public and private law, the pre-egalitarian account perhaps best exemplified by the US Bill of Rights, pre-the Fourteenth Amendment. The legal revolution worked by the Fourteenth Amendment was found in the fact that it seemed to touch on issues well beyond the ordinary statecitizen relation and to reach into areas that were traditionally the purview of states, predominantly civil rights. Though this has been the most controversial aspect of the Fourteenth Amendment, expressed in part through the extremely complicated state action doctrine and the like (see Moran, ‘Inimical to Constitutional Values’, above n 2, for a discussion of how this plays out in the context of the public-private law relation), it is generally accepted as a feature of modern equality guarantees (either through radiating values or some other means). But my argument here suggests that the account of the public-private law relation has not caught up with this profound change in the nature of public and private law. Instead, more contemporary accounts have tended, by and large, to simply privilege the public over the private, in contrast to the traditional model which privileged the private by creating a small discrete sphere of constitutional relations composed of primarily negative rights (zones of freedom or non-intrusion rather than entitlement). What I suggest here instead is that, at least since the egalitarian turn in constitutionalism, a fundamentally different model is needed, rather than an inversion of the traditional model. 34 One example that comes to mind here is found in how the assessment of historical disadvantage in equality jurisprudence tends to draw in a history in which private law discrimination often looms large. Canadian examples include Eldridge v BC [1997] 3 SCR 624 and Laws v Canada [1999] 1 SCR 497 itself.

30

Mayo Moran

meant that courts have struggled to figure out how to approach assessing whether there is a violation of a constitutional guarantee of equality. This is because they largely miss the fact that they must look to private law in order to have a full picture of the legal person, to see the detailed legal content of the idea of the person and, with that backdrop, to assess the significance of alleged violations. Instead, in an effort to give some content to the abstract idea of equal human dignity, they have invoked the reaction of the equality complainant. But this in turn has created all sorts of other difficulties and too often seems to miss the essential point. Looking to the private law construction of the person may provide at least a partial solution to this difficulty. By drawing on the vital private law content of the legal person, the significance of alleged denials of equality can be more fully understood without recourse to the highly problematic idea that the touchstone of discrimination is found in how someone feels. But what exactly is the ‘private law content of the person’? The treatment of women who lost vital elements of legal personhood upon marriage and the century-long efforts to regain that personhood illustrate how much of the content of what it meant to be a person in law was actually fashioned through the law of private relations. Thus, as the case of married women illustrates, the institutions of property, tort, contract and criminal law provide much of the content of the legal personhood that was lost and eventually regained. This suggests that while the extension of the right to vote could make women citizens, that alone would not be enough to make them persons. To accomplish that vital transition required profound transformations in all of the key institutions of private law—property, tort, and contract and, of course, in criminal law as well. Let us briefly examine how this occurred.

B. The Private Law Content of Legal Personhood: The Case of Married Women Although today we tend to ignore the fact, until quite recently natural persons were not simply by virtue of that feature also full legal persons. History is replete with examples of categories of natural persons who were historically not persons in law but who instead occupied a status more akin to property. A notorious and extreme example is of course found in the antebellum American law of slavery. In that regime, African-American slaves were not only non-citizens and hence without political rights, they were also non-persons and without civil rights.35 For example, the Slave Code of the District of Columbia explicitly distinguished 35 For an insightful overview, see A Davis, ‘The Private Law of Race and Sex: An Antebellum Perspective’ (1999) 51 Stanford Law Review 221; also see MV Tushnet, The American Law of Slavery 1810–1860: Considerations of Humanity and Interest (Princeton, NJ, Princeton University Press, 1981); TD Morris, Southern Slavery and the Law, 1619–1860 (Chapel Hill, NC, University of North Carolina Press, 1996) (especially pt II, ‘Slaves as Property’). The property-like status of the slave was the subject of extensive legal regulation in slave codes that included many provisions designed to keep the slave as property.

Public and Private Law

31

between natural and legal persons, defining a slave as ‘a human being, who is by law deprived of his or her liberty for life, and is the property of another’.36 As a form of property, slaves were the object of contract and thus subject to being bought and sold. They were accordingly incapable of being the authors of contract, including contracts for the value of their labour. Similarly, as property, slaves were unable to sue for violations of their physical integrity and were denied most of the other incidents of civil personhood expressed through a wide array of civil rights. Legal history also contains many examples of other groups that were viewed as incompetent—children, those with mental disabilities or mental illnesses and, of course, women. The history of any one of these groups would provide interesting insight into what rights one has to acquire to become a person in law. For the purposes of this exploration however, I will focus on the legal personhood of women, particularly married women, as an example of how the categories of legal person and non-person are constituted through the central doctrines of private law. The legal treatment of married women and the dramatic—though still probably not complete—shift it has undergone serve as a window into the rights implicated in legal personhood and, in particular, the significant role of the rights of private relations. It is well documented that married women were not full legal persons as a formal matter until quite recently in Anglo-American common law legal regimes. Their status was understood in terms of the legal fiction of ‘merger’, also known as coverture. In the eighteenth century, Blackstone’s Commentaries on the Laws of England outlined this idea and its implications for the personhood of married women: By marriage, the husband and wife are one person in law; that is, the very being or legal existence of the woman is suspended during the marriage, or at least is incorporated or consolidated into that of her husband, under whose wing, protection, and cover, she performs everything.37

The doctrine of merger thus explicitly suspended the legal existence or personhood of women during marriage. Indeed, the ‘pure English Common Law’, as Sir Henry Maine described it in Ancient Law, enforced with ‘rigorous consistency’ the ‘complete legal subjection on the part of the wife’ to the husband.38 Merger suspended the legal personhood of women through the fiction of a single identity for the married couple—the legal identity of course being that of the man. The ramifications of this suspension of women’s legal personhood were felt throughout private law, illustrating the centrality of the bundle of private law rights to the very idea of the person in law. Thus it is telling that the suspension of the 36 Slave Code for the District of Columbia, 1860, available from The Law Library of Congress, ‘Slaves and the Courts: Slave Code for the District of Columbia’ lcweb2.loc.gov/ammem/sthtml/stpres02. html, accessed on 18 November 2008. 37 W Blackstone, Commentaries on the Law of England, W Carey Jones (ed) (San Francisco, Bancroft-Whitney, 1915–16) vol 1, *442. 38 H Maine, Ancient Law (London, JM Dent & Sons, 1972) 93–4.

32

Mayo Moran

legal personhood of married women was accomplished through their exclusion from most of the key private law rights, including the right to physical integrity, the right to hold and deal with property and the right to enter into contractual relations.

1. Protection for Physical Integrity The single most important interest associated with personhood is the right to physical integrity, expressed through the fact that persons, unlike things, have the right to autonomy and hence to vindicate violations of physical integrity. This right is enshrined both in tort law and in criminal law and is the most highly protected interest in both of those areas of law.39 In tort law the interest in physical integrity is protected both by intentional torts, especially the torts of assault and battery, as well as by unintentional torts such as negligence. The right to physical integrity is also the most important interest protected by the criminal law through many of its central prohibitions. One of the most fundamental ways that the doctrine of merger shifted married women out of the category of persons was by precluding married women from vindicating violations of physical integrity through tort law. The idea of merger and the unitary legal personality that it entailed meant that all legal actions on behalf of women had to be at the instigation of the husband and in his name. Thus if a woman was wrongfully injured she could not bring an action against the person who injured her. Instead, it was up to her husband to decide whether to take any action and if so, what form that action might take.40 The underlying principle was that it was the husband’s right to his property, in this case his wife, that was violated, even in the case where the wife was physically injured. This idea that the wrong to the wife was an injury to the husband’s property rather than to the wife’s physical integrity was also reflected in the fact that any damages collected as a result of injuries to the wife were the property of the husband, not the wife.41 However, if the husband was wrongfully injured, the wife had no ability to sue, since the doctrine of merger meant that in the eyes of the law she had suffered no legally cognisable injury.42 One very important effect of this dimension of the doctrine of merger was that women could not maintain actions against their husbands because husband and wife were treated in law as one unified legal person. There were a number of implications for married women. However, this was especially significant for 39 S Perry, ‘Protected Interests and Undertakings in the Law of Negligence’ (1992) 42 University of Toronto Law Journal 247. 40 Blackstone, above n 37, at *431. 41 L Holcombe, ‘Victorian Wives and Property: Reform of the Married Women’s Property Law, 1857–1882’ in M Vicinus (ed), A Widening Sphere: Changing Roles of Victorian Women (Bloomington IN, Indiana University Press, 1977) 6, cited in HD Lord, ‘Husband and Wife: English Marriage Law from 1750: A Bibliographic Essay’ (2001) 11 Southern California Review of Law & Women’s Studies 1, 43. 42 Lord, ibid, at 43, quoting AH Manchester, A Modern Legal History of England and Wales 1750– 1950 (London, Butterworths, 1980) 369.

Public and Private Law

33

their physical integrity, particularly in light of the fact that women were—and still are—most likely to suffer grievous physical harm at the hands of their husbands or intimate partners. Yet at common law women were unable to sue their husbands for personal injury.43 Indeed, courts routinely pointed to the fact that the wife’s legal personality was subsumed into that of the husband as a reason to refuse to permit a wife to sue a husband.44 This meant that married women were completely precluded from vindicating the most common and deadly violations of their physical integrity.45 In fact, some courts went so far as to assert that the husband as guardian could put ‘gentle restraints’ on his wife’s liberty.46 Married women were also unable to sue their husbands for other severe harms such as the transmission to them of venereal diseases.47 As one commentator put it, in early common law the combined effect of various features of marriage made it impossible for one spouse ‘ever to be held civilly liable as a tortfeasor, in any situation, and without exception, to the other for any act, antenuptial or during marriage, causing personal injury which would have been a tort but for the marriage’.48 Similarly, other courts and commentators across a range of common law jurisdictions supported the general position that even battery by the husband could not give rise to a right of action to recover damages.49 The criminal law of the period also precluded married women from vindicating violations of physical integrity. Under common law, the legal rights of the head of the household to control his children, servants and wife included that of ‘moderate correction’. This was the basis for the right of ‘restraining her by domestic chastisement’ which included moderate beating: ‘The Husband hath by Law Power and Dominion over his Wife, and may keep her by Force within the Bounds of Duty, and may beat her, but not in a violent or cruel Manner’.50

43 C Tobias, ‘Interspousal Tort Immunity in America’ (1988–89) 23 Georgia Law Review 359, 364–5. 44 Abbott v Abbott, 67 Me 304 (Supreme Judicial Court of Maine, 1877) 306, cited in Tobias, ibid, at 364. 45 Indeed, many commentators would argue that this is still the case despite the demise of the doctrine of merger. Though formally women can sue their husbands or partners for assault, in practice, courts and other agencies of the state including police are extremely reluctant to ‘intervene’ in what they often take to be private matters. See SM Buel, ‘Access to Meaningful Remedy: Overcoming Doctrinal Obstacles in Tort Litigation Against Domestic Violence Offenders’ (2004) 83 Oregon Law Review 945; C Hathaway, ‘Gender Based Discrimination in Police Reluctance to Respond to Domestic Assault Complaints’ (1986) 75 Georgetown Law Journal 667; A Brenneke, ‘Civil Rights for Battered Women: Axiomatic and Ignored’ (1992) 11 Law & Inequality 1, 22. 46 Abbott v Abbott, 67 Me 304 (Supreme Judicial Court of Maine, 1877). 47 Schultz v Christopher, 65 Wash 496; 118 P 629 (Supreme Court of Washington, 1911). 48 WE McCurdy, ‘Personal Injury Torts Between Spouses’ (date) 4 Villanova Law Review 303, 307, cited in Tobias, above n 43, at 365. 49 T Reeve, The Law of Baron and Femme, 1st edn (New Haven CT, Oliver Steele, 1816) 65. See also Haglund, ‘Tort Actions Between Husband and Wife’ (1939) 27 Georgetown Law Journal 697, 704 cited in Tobias, above n 43, at 371. 50 HA Kelly, ‘Rule of Thumb and the Folklaw of the Husband’s Stick’ (1994) 44 Journal of Legal Education 341, 361, quoting M Bacon, A New Abridgement of the Law, 4th edn (London, W Strahan & M Woodfall, 1778) vol 1.

34

Mayo Moran

Wives who were badly beaten could on occasion seek assistance from the Court of Chancery or from ecclesiastical courts. As Blackstone notes by 1765, ‘this power of correction was confined within reasonable bounds; and the husband was prohibited to use any violence to his wife’ notwithstanding the mythical ‘rule of thumb’.51 Despite changing attitudes towards domestic violence, the courts were slow to intervene. However, beginning in the mid-eighteenth century, legislative changes began to protect women and children from violence at the hands of the head of the household.52 Finally in R v Jackson,53 the English Court of Appeal decided that a husband was not entitled to confine his wife to enforce restitution of conjugal rights, and in so doing abolished the common law’s defence of chastisement and confinement.54 Nonetheless, in many common law jurisdictions, the exemption that criminal law created for marital rape remained intact until the late twentieth century.55 This brief sketch reveals how the key right of legal personhood—physical integrity—was constituted largely through certain vital private law and criminal law rights. Thus, when women married and their legal personhood was suspended, this suspension was accomplished in large part through their inability to access the private law rights that protect physical integrity and autonomy interests. This occurred, as we have seen, because the doctrine of merger stripped married women of legal standing in favour of their husbands. Because they were thus precluded from being legal actors, they no longer had the ability to invoke private law actions for assault, battery, wrongful imprisonment and other associated wrongs. And when they lost these private law rights, they also lost a critical element of what it means to be a person, rather than property, in law. Going hand in hand with the exclusion from the private law rights, as we have seen, was the fact that they were also unable to invoke the parallel criminal law protections that are normally available in cases of assault and other crimes involving the person.

2. Property and Contract The role of private law in both the creation and the suspension of the legal person is also revealed through other implications of the doctrine of

51

Blackstone, above n 37, at *432. On the rule of thumb see Kelly, ibid, at 344–50. Aggravated Assaults on Women and Children Act 1853 (16 Vict c 30); Matrimonial Causes Act 1878 (41 & 42 Vict c 19) s 4; Summary Jurisdiction (Married Women) Act 1895 (58–60 Vict c 39) s 5. See also the decision in R v Lister (1795) 93 ER 645, 646 setting out the basis of limits to the power of correction. 53 R v Jackson [1891] 1 QB 671 (CA). 54 M Doggett, Marriage, Wife-Beating and the Law in Victorian England (Columbia, South Carolina, University of South Carolina Press, 1993) ix. 55 See, eg, in the United States: NJ Stat Ann § 2C: 14-1 (abolishing New Jersey’s marital exemption in 1978) and People v Liberta, 474 NE 2d 567 (New York CA, 1984) (holding there to be no rational basis for distinguishing between marital rape and non-marital rape); in the UK: R v R [1992] 1 AC 599 (HL) (eliminating the long standing marital rape exemption at common law); in Australia: R v L (1991) 174 CLR 379 (High Court of Australia) (rejecting the marital rape exemption). 52

Public and Private Law

35

merger. Many of these implications have to do with the relationship between personhood and autonomy or agency. When women married and lost legal personhood, this loss was expressed through several interrelated private law doctrines that affected the law of property and the law of contract. What happened to women who married was, in essence, a category shift from person to property. Story J captured the idea that the status of married women at common law was property-like when he stated ‘a man has as good a right to his wife as to the property acquired under a marriage contract. He has a legal right to her society and her future’.56 The results of this view of married women as property were far-reaching. Historically women’s loss of legal personhood as a consequence of marriage was perhaps most dramatically revealed in the role that contract played in the ‘wife sale’. This practice starkly captures the fact that married woman were understood as the property of their husbands. The practice of wife sale probably strikes contemporary readers of Thomas Hardy’s Mayor of Casterbridge as shocking and almost unimaginable but in fact it was a commonplace practice in rural England in the nineteenth century. Commentators estimate that between 1780 and 1850 about 300 wife sales took place in England. Samuel Pyeatt Menefee’s famous study of British popular divorce suggests that in rural districts wife-sale was a mutually agreeable means for the peasantry to dissolve dysfunctional marriages.57 Whatever socially useful function the practice may have had however, it was clearly not a matter of bilateral entitlement. Instead, as we see in The Mayor of Casterbridge, it was a transaction that could be instigated only by a husband, the result of which was the transfer of the wife. This practice, paralleling infamous examples in the American slave trade, illustrates perhaps more profoundly than any other how merger suspended the legal personhood of married women, transforming them from legal actors or agents who were capable of contracting to something that was instead the object of contract— that is, to property. Since women became a form of property upon marriage, it is not surprising that there were dramatic consequences for their ability to hold property. The transformation of women upon marriage from at least partially capable legal persons to a property-like status necessitated many shifts in how the law of property applied to married women. The common law, as far as it concerned wives, spent enormous energy defining the property rights that women surrendered to their husbands when they married. A husband gained legal possession or control of all property that belonged to his wife upon marriage as well as any property that might come to her during the marriage.58 Husbands were entitled to their 56 Trustees of Dartmouth College v Woodward, 17 US 518 (US Supreme Court, 1819) 696–7, quoted in Tobias, above n 43, at 370–1. 57 SP Menefee, Wives for Sale: An Ethnographic Study of British Popular Divorce (New York, St Martin’s Press, 1981). 58 AH Manchester, A Modern Legal History of England and Wales 1750–1950 (London, Butterworths, 1980) 369–70.

36

Mayo Moran

spouse’s personal property once they had possession, and were also entitled to the management and control of their real property if it was held under legal title.59 The fact that the common law and equity often had distinct rules for real and personal property meant that a complex set of doctrines was developed to govern the husband’s entitlement to his wife’s property. For instance, during marriage, the husband controlled his wife’s real property and the income from it. However, there were restrictions on his ability to alienate it, as well as complex rules regarding the reversion of property on his death.60 The power of the husband over personal property was far more extensive. In fact it was ‘the sole and absolute property vest[ed] in the husband, to be disposed of at his pleasure’.61 Thus, upon marriage, the woman lost effective control, in her lifetime, of both her real and personal property. It may be the case that, as noted above, commentators today focus primarily on the public law dimension of women’s rights, but reformers a century ago recognised that women could not be legal persons without fundamental changes to the law of private relations. Indeed, the amount of time and energy that reformers devoted to pressing for legislative amendment to the common law of property is a powerful illustration of the centrality of private law to legal personhood. In 1857, one of many attempts to reform private law with specific attention to the proprietary rights of married women was brought before Parliament.62 That legislation would have ensured that married women were ‘as capable as unmarried women of acquiring, holding and disposing of property both real and personal’, and it would have ‘allowed them to make contracts and to sue and be sued’.63 However, the Bill was defeated and the issue was not raised again for a decade. Even then, the changes were relatively minor and it was not until the Married Women’s Property Act 1882 that English law enshrined the principle that ‘married women should have the same property rights as unmarried women’. Although reformers who had pressed for an enactment of equal property rights for men and women were not fully satisfied, the 1882 Act brought about a major change in the legal personhood of married women, ending what was referred to as the ‘virtual slavery’ of marriage.64 Even once these reforms were enacted, at common law a wife could not sue her husband with regard to her property. Instead, she had to enlist the help of a man, usually a male relative, to bring a suit on her behalf. With the passage of the Married Women’s Property Acts 1870 and 1882, a wife gained a statutory right 59 RH Chused, ‘Married Women’s Property Law: 1800–1850’ (1983) 71 Georgetown Law Journal 1359, 1361; N Basch, In the Eyes of the Law: Women, Marriage and Property in Nineteenth Century New York (Ithaca NJ, Cornell University Press, 1982) 26–7, cited in Tobias, above n 43, at 370. 60 Blackstone, above n 37, at *433. 61 ibid. 62 Matrimonial Causes Act 1857 (20 & 21 Vict c 85). 63 Lord, above n 41, at 35, quoting Holcombe, above n 41, at 11. 64 Lord, above n 41, at 37, quoting Holcombe at 47.

Public and Private Law

37

to sue and be sued with respect to her separate property, just as she had gained the right to enter into contracts.65 However, the Married Women’s Property Act 1882 generally precluded husband and wife from suing each other in tort to protect property.66 This aspect of the Act was not touched by the 1935 Law Reform (Married Women and Tortfeasors) Act67 and it was not until 1962 that this bar was completely abolished.68 Another parallel and significant change in the legal personhood of married women was worked through the private law of contract. Legal personhood entails the power to act as a subject and as such is expressed through the ability to contract. Perhaps the most vital link between contract and legal personhood is found in the ability of legal persons to secure the right to the fruits of their labour, a right that is most commonly realised through the ability to enforce a contract for wages. In addition to the ability to secure value in exchange for labour, contract is fundamental in other ways to legal personhood. Without the ability to contract there is no power to deal with property through, for instance, alienating, improving and acquiring assets. As we have noted, the exclusion of married women from the rights of personhood was graphically illustrated by the fact that in the wife-sale she was the object of contract. Yet even when not the object of contract, the property-like status of married women was ensured through a web of exclusions from the ordinary powers that legal persons had to contract. The significance of these contractual entitlements to the content of legal personhood becomes apparent when one examines what happened to women’s power to contract when they lost legal personhood upon marriage. Under common law, a wife generally had no power to enter any contract or incur any debt. Although she had some limited powers to act as her husband’s agent or to pledge his credit for necessary supplies for instance, her ability to contract in his name was limited by his authority and that authority or consent could be withdrawn at any time.69 Because of the doctrine of merger, there could be no contract between husband and wife, as this would be tantamount to the husband contracting with himself.70 Though the courts of equity sometimes afforded more latitude to married women, enabling them to contract with their husbands

65 Married Women’s Property Act 1870 (33 & 34 Vict c 93); Married Women’s Property Act 1882 (45 & 46 Vict c 75). 66 The legislation of 1935 reaffirmed the rule contained in the Married Women’s Property Act 1882 that a wife could sue her husband in tort to protect her property, although he did not have a reciprocal right to sue her: Lord, above n 41, at fn 325, quoting Holcombe. 67 Law Reform (Married Women and Tortfeasors) Act 1935 (25 & 26 Geo 5 c 30). 68 Lord, above n 41, at 44, citing Holcombe, above n 41, at 225. The Law Reform (Husband and Wife) Act 1962 finally established that each spouse has a right to sue the other in tort as if they were not married. As Lord points out, the main reason for this was to enable spouses to collect insurance damages for injuries suffered through the fault of the other, such as injuries resulting from an automobile accident. 69 Blackstone, above n 37, at *430–31; Manchester, above n 58, at 371. 70 Blackstone, ibid, at *430.

38

Mayo Moran

with regard to their separate property, the validity of such contracts was controversial. Indeed, in 1919 the House of Lords held in a leading case that mutual promises made between husband and wife were not legally binding contracts.71 Again, if contemporary commentators tend to focus on personhood as largely a creation of public law, reformers of the day did not. Instead, they addressed their efforts to achieving legislative change to the common law. Much of this initiative also took place through efforts to reform the property rights of married women. This was because under common law, at least some of women’s ‘contractual incapacity lay in the fact that she had no property which she could bind’.72 The Married Women’s Property Acts meant that married women gained control over their separate property, and with this power they regained the right to enter into contracts. However, this contractual capacity was not tantamount to full legal personhood for it was not personal, but proprietary: that is, a wife’s contracts did not bind her personally but bound only her separate property.73 Again, there were a number of amendments over time concerning the power to devise property by will. In this way, the power of married women to contract was gradually dissociated from its proprietary roots. However, it was not until 1935 that the Law Reform (Married Women and Tortfeasors) Act rendered a married woman responsible for all her torts and contracts.74

C. Personhood and the Bundle of Private Law Rights Noting the shifting private law rights of women who married in the eighteenth and nineteenth centuries serves to illuminate some of the key strands in the bundle of rights that together make up what we call legal personhood. As noted above, the rise of constitutionalism and the focus on the transformative and egalitarian power of public law have tended to distract attention from the nature of the legal personhood that is protected by constitutional guarantees. One result is that contemporary courts and commentators seem to assume that being a full person in law is largely a matter of drawing on and properly developing public law guarantees. However, the efforts to proceed in this manner have proved problematic. Even a brief review of the shifting legal personhood of married women serves to remind us of what would have been obvious in an earlier legal era—the bundle of rights that together make up what we think of as the essence of legal personhood is largely constituted not by public law rights but rather by private 71 Balfour v Balfour [1919] 2 KB 571 (CA). See also L Holcombe, Wives and Property: Reform of the Married Women’s Property Law in Nineteenth-Century England (Toronto, University of Toronto Press, 1983) 221. 72 Manchester, above n 58, at 371. 73 ibid. 74 Law Reform (Married Women and Tortfeasors) Act 1935 (25 & 26 Geo 5 c 30).

Public and Private Law

39

law rights. Being a full person in law (as opposed to being a form of property or occupying some other status associated with incompetence or incapacity) is a status composed of a set of rights to physical integrity and to autonomy. It is thus no accident that the movement to emancipate the black slaves in the United States was referred to as the ‘civil rights movement’. The slave codes that made human beings into property did so in large measure through exclusion from all of the basic civil rights that together constitute the legal person. And, accordingly, the movement to gain the status of legal person for those individuals was focused on the acquisition of fundamental civil or private law rights. Similarly, as the situation of women reveals, when they were ‘suspended’ as legal persons upon marriage, the loss of that status was accomplished through a web of exclusions from all of the most critical private law rights, as well as from vital criminal law rights. The private law rights that were lost by women when they lost their legal personhood included the right to sue for violations of physical integrity, the right to demand value for labour, to grant or withhold consent through contract and other actions, the right to hold and deal with property and many other private law rights too numerous to detail here. But the basic point is clear and was clear to reformers at the time—to have the status of a full legal person is in large measure to have certain fundamental private law rights. Like the major efforts at reform in the aftermath of chattel slavery in the United States, these efforts were rightly understood as ‘civil rights’ struggles—struggles, in other words, to work fundamental reform not in the public sphere but rather in the law of private relations. As this reminds us, the legal person cannot be understood solely through the lens of public law, for it is to a very significant extent a creation of private law.

D. Why the Private Law Person Matters to Public Law Even if it is the case that the legal person is largely composed of private law rights, one might ask why this matters to the relationship between public and private law. It matters because noticing the private law composition of the legal person that is the subject of public law concern illustrates the error of thinking of these areas as unrelated domains. This connection, moreover, alerts us to the fact that private and public law not only draw on but actually need each other in fundamental ways. While these bodies of law are conceptually distinct and hence not reducible to each other, they nonetheless require each other to fully articulate and develop their own doctrines. It has been recognised for some time now that public law values shape private law in important ways. But what has escaped recognition is the fact that public law guarantees themselves cannot be properly articulated without a deep understanding of the role and contribution of private law. What I would now like to do briefly is to illustrate how attending to the private law composition of the legal person helps to resolve some of the challenges faced by the public law guarantee of equality.

40

Mayo Moran

Modern constitutions tend to view as their centrepiece an equality guarantee that enshrines the overarching importance of equal human dignity.75 The central thrust of this guarantee has been to insist that, at least presumptively, natural persons are full legal persons—this is the essence of equal human dignity. This emphasis has had the effect of creating increasingly robust jurisprudence around the question of what it means to guarantee equality. The challenges inherent in this question have faced courts and commentators with extremely difficult problems, leading to the characterisation of equality as ‘the most difficult right’.76 Most modern guarantees of constitutional equality use discrimination as the touchstone of at least one dimension of equality. So the analysis typically proceeds by asking first, whether there is a constitutionally significant distinction in treatment and second, whether that distinction is discriminatory. This is because it is typically only discriminatory distinctions that raise constitutional concern. The result is that the discrimination inquiry generally forms the heart of modern constitutional equality analysis. Yet there is no equality issue so sensitive and as challenging as the question of what counts as discrimination. Courts in contemporary constitutional regimes have struggled to come up with an adequate account of discrimination. In so doing, they have typically invoked the importance of the underlying value of human dignity. But while this is undoubtedly important, it has proved challenging to move beyond a general invocation of the importance of dignity to the equality inquiry. A more finely textured assessment of what exactly discrimination means in the context of a concrete constitutional challenge has proved elusive. Courts in regimes with robust equality guarantees have identified this as a major problem with contemporary equality jurisprudence and have struggled to deal with it. They have adopted a range of different approaches to try to respond, but none to date has seemed satisfactory. To explore the promise of a more integrated understanding of public and private law, I would like to examine how the Supreme Court of Canada has suggested analysing discrimination for the purposes of constitutional equality. Noticing the difficulties created by the current approach suggests it may be more helpful to look in another direction—towards private law. So a more integrative understanding that attends to how the legal person is constituted in private law may actually help us to respond more adequately to a central conundrum of public law. 75 See M Moran, ‘The Reasonable Person and the Discrimination Inquiry’ in S Tierney (ed), Accommodating Cultural Diversity (London, Ashgate Publishing, 2007); LE Weinrib, ‘The Postwar Paradigm and American Exceptionalism’ in The Migration of Constitutional Ideas, above n 2; D Reaume, ‘Discrimination and Dignity’ in C McCrudden (ed), Anti-Discrimination Law, International Library of Essays in Law and Legal Theory—2nd Series (Hampshire, Ashgate Publishing, 2004). 76 B McLachlin, ‘Equality—The Most Difficult Right’ (2001) 14 Supreme Court Law Review (2d) 17. For an analysis of some of the core challenges with equality jurisprudence see S. Moreau, Equality and the Promise of Law v. Canada. “University of Toronto Faculty of Law Journal, Volume 57.2 (2007) pp. 415–430, and The Wrongs of Unequal Treatment”. University of Toronto Faculty of Law Journal, Vol. 54.3 (2004) pp. 291–326.

Public and Private Law

41

Like other courts with robust modern equality guarantees, the Supreme Court of Canada places great emphasis on the role of discrimination in identifying which distinctions are constitutionally suspect on equality grounds. Thus, it is not sufficient for a claimant to point to a distinction that disadvantages her; she must also establish that that distinction is discriminatory. The discrimination inquiry itself has increasingly focused on the question of dignity. Thus, the court will ask whether the distinction is one that demeans the essential human dignity of the claimant. In order to answer this question the Supreme Court of Canada has invoked a reasonable person test. That test asks whether the alleged violation of equality would be viewed by a reasonable person as violating her inherent human dignity.77 This approach, however, has proved to have very serious difficulties. Elsewhere, I explore these in some detail but for the moment let me briefly mention two. First, the test puts the court in a difficult position—a position that is particularly problematic in the context of equality claims. This arises because any person who brings an equality challenge will quite naturally believe that the alleged violation infringes her dignity. So the test clearly does not, and indeed cannot, rely on the plaintiff ’s actual or subjective sense of whether her dignity is infringed, for if it did so every distinction that came to court would count as discriminatory. For this reason, the Supreme Court has described the test as a combination of subjective and objective elements, which it clearly must be. However, the fact that the Court analyses whether discrimination exists by asking whether the reasonable person would believe that his dignity was demeaned means that the denial of a claim implies that the claimant’s response was not reasonable. In addition to the fact that this seems a rather weak justification, it poses particular problems in the context of equality claims brought forth by those who are marginalised or disadvantaged. The fact that the members of the court, who tend not to be disadvantaged, end up justifying their holding by telling those who are marginalised that their reactions to differential treatment are unreasonable seems likely to diminish rather than enhance respect for the decision. There is however a second and more fundamental problem with this approach. A test that turns on the sense of injury to a dignitary interest, that focuses on the idea of feeling demeaned and debased, is a misleading and problematic way to depict the wrong of discrimination. No amount of ‘disciplining’ the subjective basis of this approach with objective reasonableness will eliminate its fundamental difficulty. This is because it suggests that the essential injury of equality is a matter of ‘hurt feelings’—damage that has as its touchstone the subjective response of the complainant rather than an objective wrong.78 But the guarantee of equality does not occupy the place it does in so many modern constitutions because it protects people from feeling bad or injured; it is the touchstone because it expresses a commitment to the idea of the equal human worth of all persons. 77 78

Laws v Canada [1999] 1 SCR 497. Moran, ‘The Reasonable Person and the Discrimination Inquiry’, above n 75.

42

Mayo Moran

Yet an analysis that does not draw on the objective legal content of what it means to be a person in law will inevitably fall back on generalities or poor substitutes, like the reasonable person’s sense of being demeaned. And to draw more of the legal content of personhood into the discrimination inquiry, one must look to private law. It may seem surprising that courts have not drawn more on the law itself to flesh out the discrimination inquiry, but in many ways it is not. To do so would be to recognise that public law needs private law and not simply that private law needs public law. In order to explore just how drawing on the private law of personhood could assist in the analysis of constitutional equality, I will briefly discuss a couple of examples that will help to flesh out the idea. Though it has not been discussed by commentators in any detail, that ought not to be taken to suggest that this idea is entirely new. Indeed, there are many glimmers of recognition that something like this connection between public and private law is necessary and important. One such glimmer can be found in the analysis of equality in Eldridge v British Columbia.79 That case involved a question about whether the failure by the British Columbia health-care plan to fund interpreters for the deaf constituted a violation of the equality guarantee under the Charter. Although the lower courts held that it did not, the Supreme Court of Canada unanimously held that the funding failure did constitute a violation of section 15 and, moreover, that this violation was not saved as a reasonable limitation under section 1. Interestingly for our purposes here, the Court looked explicitly to private law in order to assess the constitutional significance of the failure to provide deaf patients with sign language interpreters. In particular, the Court suggested the relevance of the private law doctrine of informed consent in tort to the equality guarantee. It reasoned in the following way. The law of assault and battery, which protects the essential interest the person has in physical integrity, holds that a patient only gives consent where that consent is informed.80 However, consent cannot be informed in the absence of proper communication between the doctor and the patient. Thus, the critical dividing line between treatment and assault depends on appropriate communication between doctor and patient. The fact that effective communication is so central to authorising treatment in private law enabled the Court to conclude that the failure to provide the means of such communication to the deaf violated their public law right to equal treatment under the provincial health-care plan. Thus here judicial awareness of the interlocking nature of private and public law was central to the Court’s reasoning on equality. The background effect of the recognition of the relationship between core private law rights of personhood and the constitutional guarantee of equality can also be seen in another equality controversy, one that did not end up in litigation. The issue arose in the wake of an important private law case called Muir 79 80

Eldridge v British Columbia [1997] 3 SCR 624. Reibl v Hughes [1980] 2 SCR 880; Hopp v Lepp [1980] 2 SCR 192.

Public and Private Law

43

v Alberta.81 That landmark case awarded one million dollars in damages to a woman who had been wrongfully sterilised because she was believed to be mentally disabled. In fact, she had been wrongly classified under the eugenics regime of the Province of Alberta and on that basis she won a very large damage award. Subsequent to the decision in Muir, the Alberta government became concerned about the fiscal impact of other cases. Accordingly, it proposed legislation. Bill 26 proposed a formula the effect of which would have been to limit the damages that courts could award in cases involving involuntary sterilisation. For our purposes here, what is noteworthy is that the government apparently recognised the interrelationship between public and private law. Indeed, it feared that the curtailment of private law rights of action for assault and battery would violate the equality guarantee of the Charter. As a result, Bill 26 expressly invoked the override provision of the Charter (section 33), a measure which would have enabled the legislation to survive despite non-compliance with the Charter. What is relevant here is the awareness that limiting private rights of action in certain ways may implicate the equality guarantee. In this case, what seems most noteworthy is the fact that the Bill limited a right that is associated with the most fundamental interest of the person protected by private law; that is the right to the integrity of the person, protected in this instance through access to court to vindicate violations of bodily integrity. No doubt it was also significant that the right to vindicate violations of physical integrity was not limited generally but only for those whose violations were related to their assumed mental disabilities, a prohibited ground of distinction under the Charter. This type of limitation on such a key interest would thus have been discriminatory because it reinforced the historic exclusion of persons with disabilities from the ordinary protections that private law affords to key elements of personhood. In this sense then, in this instance, the critical link between the private law person and the public law guarantee of equality was of vital importance to the constitutional concern. Indeed, the concern was such that when Premier Klein proposed this legislation, there was significant public outcry. Accordingly, the Bill was retracted and Alberta negotiated a settlement with the victims.82 Despite these glimmers of recognition of a relationship between the private law person and public law rights, it must be admitted that most equality cases fail to notice the significance of the background of private law rights. An example can be found in Children’s Foundation,83 a case involving the constitutionality

81

Muir v Alberta (1996) 132 DLR (4th) 695 (Alberta QB). On 11 Mar 1998, the Alberta Government retracted Bill 26, acknowledging that victims were entitled to fair compensation. On 4 June 1998, the 504 victims represented by the Limited Trustee reached an out-of-court settlement in the amount of CAN$75,000 per claimant, with an additional CAN$25,000 payable in three years to those who live outside of the province’s three mental health institutions: See L Jacobs, ‘Reconciling Tort and Administrative Law Concepts of Justice: The Case of Historical Wrongs’ (2007) 57 University of New Brunswick Law Journal 134, 153–5. 83 Canadian Foundation for Children, Youth & the Law v Canada (2004) 234 DLR (4th) 257 (Supreme Court of Canada). 82

44

Mayo Moran

of the chastisement defence to assault that criminal law typically provides for parents and teachers. In Canada, this defence was part of the Criminal Code and was challenged on equality grounds by the Canadian Foundation for Children, Youth and the Law (The Children’s Foundation). A majority of the Supreme Court of Canada found that the so-called spanking law did not violate the equality guarantee of the Charter (section 15). The nub of the Court’s reasoning is found in its discrimination analysis. The Court acknowledged that the Criminal Code provisions did indeed create a distinction between children and adults, and it further agreed that that distinction disadvantaged children because it denied them, on the ground of their youth, protection for physical integrity that is extended to all other people. However, the majority held that this distinction was not discriminatory because a reasonable child who was hit by his parents would not feel his dignity was demeaned. The minority strongly dissented on this point, arguing instead that being subject to physical violations without recourse to the criminal law would be experienced as a denial of dignity by a reasonable child. The nature of the analysis in Children’s Foundation points up some of the weaknesses noted earlier with the fictionalised person as the touchstone for discrimination. In contrast, drawing on what it means to be a legal person and situating the significance of the spanking defence in that context seems likely to yield a far more illuminating analysis. Thus for example, if the analysis were to focus on the fact that the equality guarantee enshrines the idea that natural persons are presumptively to be treated as full legal persons and to examine the significance of the alleged violation in that context, it would look quite different. Approaching the issue in this way would not lead a court to ask whether a reasonable child would view her dignity as demeaned, but rather to ask about the significance to legal personhood of the distinction in question. In the case of the defence to assault, as the examination of the treatment of married women reminds us, the ability to vindicate violations of physical integrity, through private and criminal law, is one of the most critical rights in the bundle that makes up legal personhood. The right to physical integrity and to vindicate violations thereof was a critical right lost by women who married, transforming them into property rather than persons. If a core reason for the robust equality guarantees in modern constitutions is to correct the errors of a discriminatory history in which natural persons had the status of things, then any denial of physical integrity to a natural person ought to awaken very intense equality scrutiny. The denial of a right to protect one’s physical integrity when viewed in the context of legal personhood treats children as the property of their parents, not as beings with their own autonomy and dignity interests. That this exclusion perpetuates a property-like status is apparent in the fact that the exclusion from the criminal law protections is not general but rather takes the form of parental immunity, exactly replicating the traditional understanding of children as property. The fact that this is the case does not necessarily suggest that no reason could justify the denial, but given the significance of such an exclusion to legal personhood, it would have to be a very good one.

Public and Private Law

45

IV. Conclusion This essay suggests that there are compelling reasons to attend to the interaction between public and private law. While, as noted above, much attention has been paid to the impact of public law on private law, similar attention has not been paid to the impact of private law on public law. However, as we have noted, the legal person who is the subject of constitutional concern, particularly in the guarantee of equality, is largely composed of a bundle of private law rights. Paying attention to this is significant for a number of reasons, not the least of which is the fact that it suggests that the relationship between public and private law is not characterised by at most a uni-directional flow of value from public law down into private law, as many commentators assume. Instead, it suggests that public and private law are best understood as complementing, supplementing and in some instances correcting each other in a manner that can best be described as mutually constitutive. This should not be taken to suggest, however, that public law ought to simply reflect the content of private law rights, any more than that private law rights ought to automatically change to echo constitutional rights. Instead, the interrelationship between public and private law is best understood as a more dynamic one. This means that there is no easy answer to how to read them together but it also suggests that legal reasoning will be stronger and more satisfying when that admittedly difficult task is undertaken. For, though conceptually distinct, public and private law need each other, they speak to each other and they help to create each other.

3 What’s Private About Private Law? WILLIAM LUCY*

I. One Size Doesn’t Fit All When lawyers turn to other disciplines in the social sciences and humanities for guidance, they usually do so in pragmatic spirit: they want answers to particular difficult questions. This pragmatic spirit might be both mistaken and philistine: mistaken if it assumes greater determinacy in other disciplines than exists in law, and philistine because this mistake, and the pragmatic spirit in which it is made, shows little or no appreciation of the disciplines invoked. It is a form of intellectual voyeurism.1 If we nevertheless persist with this seemingly pragmatic and possibly philistine approach, bringing it to bear on one old and apparently irresolvable issue, some determinate guidance is in this instance available. The issue, which is really a nest of issues, can be captured by a disarmingly simple question: Is there any significant and fruitful way of distinguishing private from public law? The guidance from other disciplines, as lawyers often find, is on this issue somewhat unhelpful. For, although the content of the guidance is clear—there is no single meaningful distinction between ‘public’ and ‘private’, there being instead a manifold set of distinctions, drawn for quite different purposes and thus having quite different contours—it is unhelpful for lawyers’ usual pragmatic purposes.2

* Thanks to Mindy Chen-Wishart, Mike Feintuck, Mayo Moran, John Murphy, Stephen Smith and the participants at ‘The Goals of Private Law’ conference, National University of Singapore, July 2008, for helpful questions and comments. 1 B Leiter, ‘Intellectual Voyeurism in Legal Scholarship’ (1992) 4 Yale Journal of Law and Humanities 79. 2 Three fine essay collections confirm this: MP d’Entreves and U Vogel (eds), Public and Private: Legal, Political, and Philosophical Perspectives (London, Routledge, 2000); J Weintaub and K Kumar (eds), Public and Private in Thought and Practice: Perspectives on a Grand Dichotomy (Chicago IL, University of Chicago Press, 1997); SI Benn and GF Gaus (eds), Public and Private in Social Life (London, Palgrave MacMillan 1983). Also richly informative is R Geuss, Public Goods, Private Goods (Princeton NJ, Princeton University Press, 2001). P Cane is one of the few jurists to grasp this feature of the public/private distinction and to consider its implications: see his ‘Public Law and Private Law: A Study of the Analysis and Use of a Legal Concept’ in J Eekelaar and J Bell (eds), Oxford Essays in Jurisprudence: Third Series (Oxford, Clarendon Press, 1987); and ‘Accountability and the Public/Private Distinction’ in N Bamforth and P Leyland (eds), Public Law in a Multi-layered Constitution (Oxford, Hart Publishing, 2003). That the domain of private law can be drawn more or less broadly (or narrowly), depending on one’s purposes, is well noted in N MacCormick, Institutions of Law (Oxford, Clarendon Press, 2007) 224–6.

48

William Lucy

These pragmatic purposes are so named because they revolve around the resolution of particular legal-doctrinal questions upon which ‘the’ (or, better, ‘a’) public/private distinction seems to bear. But the resolution of particular doctrinal questions as they occur in specific cases is but one purpose the lawyer might have. Another, broader purpose could well be described as that of getting a clear view of the law’s conceptual cartography, of being able to see how the various branches and components of the legal system fit together (if at all). This broader purpose is animated by the idea, perhaps brought to the fore by critics of the common law yet also important for civilian lawyers, that the law should not be a disorderly, dissonant and incoherent jumble of particular doctrines and decisions. The law should be a system, on this view, with different component parts, all of which should function together as a coherent whole. It seems obvious that there is no necessary connection between these two different kinds of purpose. In particular, the pragmatic purpose of discovering solutions to specific legal-doctrinal questions need not be directly determined by broader considerations of conceptual cartography. The judicial resolution of a particular case might, of course, mediate the relationship between public and private law. But the broader issue of the relationship between public and private law surely need not be resolved first, in order for the resolution of the particular case to be legitimate or otherwise respectable. Similarly, my decision to buy my morning coffee from Mega-Corp rather than Small Cafe, because Mega-Corp’s coffee tastes better, may ultimately have consequences for the retail make-up of my local high street. Yet clearly I do not need to take a stance on the benefits and disadvantages of global versus local capitalism in order to determine which coffee tastes better. It also appears odd to go about constructing any general distinction or set of distinctions such as that between contract and tort, say, or that between private and public law, with a specific case or doctrinal issue in mind. Broad distinctions can be serviceable even when they offer little or no guidance in specific cases. It is still helpful to know the general differences between rugby and wrestling, even though this knowledge is inert when faced with the question of the legitimacy of using wrestling-style tackles in rugby. The argument of this essay might seem ultimately to deny the truth—that there is no single all-purpose distinction between private and public that can be invoked to straightforwardly distinguish private and public law—it purports to take seriously. For, somewhat perversely in light of the previous paragraphs, the argument is that there is a significant and plausible distinction that can be drawn between public and private law. This argument is developed once two problematic attempts to provide a significant and plausible distinction between private and public law are set aside, in section II. The fulcrum of the argument, outlined in section III, is nevertheless congruent with the claims already made about the plural nature of the public/private distinction(s). This is because the version of the distinction defended here is neither universal nor multipurpose nor binary. It is not universal because it functions to distinguish from one another some, but

Duty of Care in APrivate Rights-Based TheoryLaw? of Negligence What’s About Private

49

certainly not all, areas of law often labelled either public or private. It is therefore compatible with the claim that the public/private distinction must be drawn differently for different (including legal) purposes. It is not multipurpose because it operates only at a very general level; it does not purport to give dispositive answers to pragmatic doctrinal legal questions as well. Nor is the distinction defended here binary, generating ‘either/or’ answers whenever and wherever applied.3 The distinction holds as a matter of degree and in some contexts but not all. This, it is argued, is no reason for embarrassment, nor is the fact that the distinction may be of relatively little use for pragmatic doctrinal purposes. But can such a qualified and not obviously pragmatically useful distinction fulfill the task of all genuine distinctions: Can it provide a reliable means of marking a real difference (or set of real differences)?4 The argument that follows offers an affirmative answer.

II. Two False Starts The history of legal thought is littered with attempts to distinguish public from private law. While this is no place for a survey of these attempts, it is worth noting two particularly interesting and very different efforts to distinguish the two. Their weaknesses are quite different but both are rooted in ambition: the first effort is regarded by many jurists as being insufficiently ambitious, while the problems of the second arise from a surfeit of ambition. The first effort can be labelled the ‘legal-doctrinal’ distinction and, in English law at least, seems both undeniable and unproblematic. This means of distinguishing public and private law consists of highlighting the various doctrinal and procedural differences between the two domains. For much of the common law’s history in England the remedies for public law wrongs, the rules of standing, as well as the doctrinal requirements for establishing such wrongs and obtaining remedies, have been for the most part different from the wrongs, remedies and doctrinal requirements embodied in private law.5 Furthermore, there is now an administrative court in England, thus reinforcing a public law/private law divide.6 3 The term ‘binary’ and its substance belong to Cane. See P Cane, ‘Accountability and the Public/ Private Distinction’ in N Bamforth and P Leyland (eds), Public Law in a Multi-Layered Constitution (Oxford, Hart Publishing, 2003) 269. 4 I thus ignore, for reasons that become explicit in the remainder of the essay, one of the two conditions Duncan Kennedy sets for the distinction: see his ‘The Stages of the Decline of the Public/Private Distinction’ (1982) 130 University of Pennsylvania Law Review 1349, 1349. This essay was subject to unsubtle parody in D Shapiro, ‘The Death of the Up-Down Distinction’ (1984) 36 Stanford Law Review 465. 5 For an overview of administrative law remedies and related issues, see PP Craig, Administrative Law, 5th edn (London, Thomson, 2003) pt 3 and W Wade and C Forsyth, Administrative Law, 9th edn (Oxford, Clarendon Press, 2004) pt VII. 6 From 1981 until 2000 the Crown Office List ensured that only judges with public law experience heard applications for judicial review; as a result of a practice direction of 20 July 2000 the list was renamed ‘The Administrative Court’.

50

William Lucy

This set of doctrinal, remedial and procedural differences between public and private law is not, of course, the only possible set; other jurisdictions draw the distinction in rather different ways.7 But this set of differences undoubtedly adds up to a significant distinction between the two domains and that, surely, is more than sufficient for most lawyers’ pragmatic-doctrinal purposes. Some jurists find this conclusion unsatisfying, without being perfectly explicit as to why. They are content to note the legal-doctrinal distinction between public and private law as just stated, yet they then proceed as if it is in need of further explanation and justification.8 What, then, is the worry here? Perhaps this: the legal-doctrinal distinction is insufficiently ‘deep’ or, what likely amounts to the same thing, altogether too contingent. Thus the distinction as currently embodied in English law might simply be an historical accident rather than a well-founded and valuable means of distinguishing private and public law. Espousing this view does not require great scepticism of the jurist or lawyer, but simply awareness that the law, either in the hands of judges, legislators or both, can take wrongturnings. These turnings can be wrong in legal, moral or political terms. A statute, judicial decision or line of decisions can inhibit desirable doctrinal development, or impact adversely on some aspect of commercial, social or cultural life, as well as embodying morally and politically objectionable distinctions or suppositions.9 This awareness inhibits the tendency to regard all legal-doctrinal features and developments as always prima facie desirable and justified; it is part of the process of ‘demystifying the law’.10 One way, particularly appealing to contemporary jurists, in which legaldoctrinal features can be given greater—more than ‘merely’ contingent—depth is by providing them with a normative foundation. Of the two principal candidates currently competing for the role of the normative foundation of private law, only corrective justice is likely to be able to make sense of the distinction between public and private law.11 Efficiency, the main rival to corrective justice and lodestar

7 For a warning that the distinction as currently embodied in English law is a worrisome legal transplant, see JWF Allison, A Continental Distinction in the Common Law (Oxford, Clarendon Press, 1996). A contemporary overview of the distinction in French and English law is provided by the essays in M Freedland and J Auby (eds), The Public Law/Private Law Divide: Une Entente assez Cordiale? (Oxford, Hart Publishing, 2006). 8 See Cane, above n 3, at 248–9 for some interesting observations on this issue. 9 A common-law list of shame usually includes Lochner v New York, 198 US 45 (US Supreme Court, 1905) and Bartonshill Coal Co v Reid (1858) 3 Macq 266, among others. The list should also contain those English criminal law cases pre-dating R v R [1992] 1 AC 599 (HL). 10 Possibly initiated or at least made prominent, in the Anglophone world, by Jeremy Bentham. See HLA Hart, Essays on Bentham (Oxford, Clarendon Press, 1982) ch 1. 11 There are, of course, many attempts to articulate the normative content and context of public law, as well as a number of attempts to undermine that project, none of which are discussed here. In England some of the principal contributions are TRS Allan, Law, Liberty, and Justice: The Legal Foundations of British Constitutionalism (Oxford, Clarendon Press, 1993) and Constitutional Justice (Oxford, Clarendon Press, 2001); PP Craig, Public Law and Democracy in the United Kingdom and the United States of America (Oxford, Clarendon Press, 1990); M Loughlin, Public Law and Political Theory (Oxford, Clarendon Press, 1992) and The Idea of Public Law (Oxford, Clarendon Press, 2003).

Duty of Care in APrivate Rights-Based TheoryLaw? of Negligence What’s About Private

51

of both positive and normative economic analysis of law, rarely registers the legal-doctrinal features that non-economically inclined lawyers regard as significant. Economic analysis of law’s view of the public law/private law distinction is therefore likely to be as heretical, and as challenging to the normal lawyerly view, as is its view of tort liability.12 Corrective justice is conceived by most of its current proponents as a moral or political idea, its guiding precept being that those who wrong others must do something about that wrong.13 Stated thus, the idea has obvious intuitive appeal but says little. It requires considerable unpacking, for although part of the notion’s appeal is that it regards the relationship between victim and wrong-doer as inherently morally significant, many aspects of this bipolar relation must be further illuminated. So, for example, an account of what does and does not count as a wrong for the purposes of corrective justice is necessary. Some illumination of the ways in which wrongs can and cannot be brought about is also required, as is some indication as to what must be done in response to wrongs. Furthermore, a full and jurisprudentially respectable account of corrective justice must elucidate the basis of the obligation to correct, giving it a foundation in addition to its ‘bare’ legal existence. What normative reasons are there, beyond the existence of a ‘mere’ legal obligation, to correct some of the wrongs one does? An account of wrongs and wrongdoing must also of necessity occupy itself with general notions of causation and responsibility. Not all existing accounts of corrective justice satisfy these requirements or, indeed, regard each of them as genuine requirements, but that is not important here.14 What is significant is that the two principal contemporary accounts of corrective justice are offered in order to fulfill the role of the normative basis of private law. These accounts may explain some areas of private law better than others and may even need amendment or expansion, including embracing values other than corrective justice itself, in order to provide a foundation for all areas of private law.15 Yet both of them assuredly regard private law as the domain of corrective justice, although they differ as to what other values are, or should be, in play there. Jules Coleman, for instance, seemingly allows interplay in private

12 It is misleading to say that lawyer-economists will see no differences between private and public law; the differences they do see, however, are likely to be superficial. A typically trenchant statement of the approach is: ‘Practices, institutions, and bodies of law that are wholly unrelated when viewed through the lens of orthodox legal analysis are seen to involve ... identical economic issue[s]. Whole fields of law are interchangeable when viewed through the lens of economics’ (R Posner, The Frontiers of Legal Theory (Cambridge MA, Harvard University Press, 2001) 40). See also R Posner, Economic Analysis of Law, 7th edn (New York, Aspen, 2007) pt VII. 13 On the variety of views among our current corrective justice theorists as to whether the notion is moral, political or some hybrid, and the difference this might make, see my Philosophy of Private Law (Oxford, Clarendon Press, 2007) 259–60. 14 I have offered a more general statement of these requirements, alongside a more sustained analysis of the two principal contemporary accounts of corrective justice in Philosophy of Private Law, ibid, ch 8. 15 This is made vividly clear in Stephen Smith’s contribution to this volume.

52

William Lucy

law between corrective and distributive justice considerations, while maintaining that the former animates large areas of private law doctrine.16 Ernest Weinrib, by contrast, holds that corrective and distributive justice are different and incompatible forms of understanding the social world which cannot be combined.17 It is Weinrib’s view of the relationship between corrective and distributive justice, appropriately labelled ‘the separation thesis’, which can provide a binary distinction between public and private law: the two incompatible forms of justice animate the two different branches of law. The qualification that Weinrib’s account can provide a distinction between public and private law is important, since this is not an issue to which he gives a great deal of explicit attention. Rather, a version of the distinction can be implied from some of his express comments. This process of implication might be less robust than could be hoped. Weinrib’s account of private law and its very different relation to corrective justice, on the one hand, and distributive justice, on the other, is the second way of separating public and private law under consideration here. It seems beyond doubt that this approach can provide normative depth to the first, the historically contingent current legal-doctrinal distinction between private and public law. The vital issue, then, is whether or not this distinction between the domains of corrective and distributive justice is itself a plausible and robust way of distinguishing public and private law. If intended as a binary distinction between public and private law—a distinction which can tell us, without fail, that some legal-doctrinal issue is either a public or private law matter—then there is an immediate problem. For if this way of distinguishing private and public law relies, as it surely must, on the claim that private law is exclusively the realm of corrective justice, then that claim seems prima facie false. This is because, according to some judges and jurists, distributive justice plays a significant role within private law. Both groups could presumably be mistaken about this, either misidentifying corrective justice concerns as distributive justice concerns or wrongly importing the latter into private law.18 But supporting either of these two claims is not easy for Weinrib. This is because his account of private law purports to take seriously ‘juristic experience’19 and the law’s ‘self-understanding’20, the latter being in part given by ‘the experience of those who are lawyers’.21 Doing this must entail taking the views of participants seriously and, while this does not entail that such views 16 See his Risks and Wrongs (Cambridge, Cambridge University Press, 1992) pt I and ch 20; and his The Practice of Principle (Oxford, Clarendon Press, 2001) 44 and 53. 17 EJ Weinrib, The Idea of Private Law (Cambridge MA, Harvard University Press, 1995) 61 and 71–4. Weinrib’s thinking has not stood still: for developments see his ‘Correlativity, Personality, and the Emerging Consensus on Corrective Justice’ (2001) 2(1) Theoretical Inquiries in Law (online edition) Art 4; ‘Corrective Justice in a Nutshell’ (2002) 52 University of Toronto Law Journal 349. 18 For discussion of some of the cases see R Mullender, ‘Corrective Justice, Distributive Justice, and the Law of Negligence’ (2001) 17 Professional Negligence 35; and P Cane, ‘Distributive Justice and Tort Law’ (2001) New Zealand Law Review 401. 19 Weinrib, The Idea of Private Law, above n 17, at 3. 20 ibid, at 14. 21 ibid, at 9.

Duty of Care in APrivate Rights-Based TheoryLaw? of Negligence What’s About Private

53

be regarded as incorrigible, it requires that departure from them, or any subset of them, must be justified. And, of course, the reasons for overlooking or reclassifying the views of those participants in the institution of private law who regard it as being a domain of both corrective and distributive justice must be consistent with the other commitments of Weinrib’s theory. Whether or not Weinrib’s theory has the resources to do this is an open question.22 The opposite side of this argumentative coin is also significant. It holds that there is no good reason supporting the claim, vital to this way of distinguishing public and private law, that public law is a corrective-justice-free zone. Indeed, most components of theories of corrective justice—the ideas of wrong, wrongdoing, the duty to correct wrongs, as well as causation and responsibility—seem equally appropriate means of characterising both public law and private law disputes. If I am denied a licence to operate a cab by my local authority on inappropriate grounds such as, for instance, my ethnicity, religion or sexuality, it does not seem bizarre to regard this as a wrong to me, done by a representative person or body of the local authority, which they have a duty to correct.23 Of course, the wrong is unlike some private law wrongs, but this observation can occlude the fact that private law wrongs themselves are not uniform. Moreover, it might seem that the usual (but obviously not the only) remedy for private law wrongs—an award of monetary compensation—is an inappropriate response to some public law wrongs. Again, this is surely correct, but the point obscures two important issues. First, that there is a range of remedies for private law wrongs, some of which have something in common with public law remedies. And, secondly, that corrective justice theories must, if they are to have explanatory power, either license more than one form of private law remedy or show why remedies other than compensatory damages are unjustifiable. It might be objected that three vital and closely related features of Weinrib’s corrective justice analysis of private law have been overlooked. As a result, our claim that public law is as amenable to explanation in terms of corrective justice as is private law is mistaken. The features ignored are three of the five which constitute the bipolar nature of corrective justice. The latter, for Weinrib, embraces, inter alia, a conception of injustice as a violation of quantitative equality; a … conception of damage as a loss by the plaintiff correlative to the defendant’s gain; … and a … conception of the remedy as the annulment of the parties’ correlative gain and loss’.24

Since this third feature depends for its plausibility upon the truth or plausibility of the second, only the first and second features are discussed in what follows. 22 I have argued elsewhere that it is difficult for Weinrib’s theory to do this, consistent with its other aims: see my ‘Method and Fit: Two Problems for Contemporary Philosophies of Tort Law’ (2007) 52 McGill Law Journal 605. 23 The scenario is only very remotely related to R v Liverpool ex p Liverpool Taxi Fleet [1972] 2 QB 299 (CA). 24 Weinrib, The Idea of Private Law, above n 17, at 65–6.

54

William Lucy

Weinrib’s view that corrective injustice constitutes a violation of a quantitative equality is a direct derivation from Aristotle. The image Aristotle uses to describe corrective injustice is that of shifting an amount from one equal line to another, so what was once equal (namely, A and B’s quantities represented by equal length lines) is now unequal (A, let us assume, having taken ‘something’ from B so that A’s line is now longer than B’s).25 Aristotle and Weinrib hold that, provided some other conditions are satisfied, A has a duty in corrective justice to correct the wrong done to B. As is obvious, the damage that results from this wrong (insofar as they differ) is well characterised by the second feature of private law’s bipolarity: the damage or loss to B is correlative (which in one sense is a synonym for ‘equal’) to the gain to A. These two features of private law’s bipolarity present an immediate difficulty for the argument that public law is a realm in which, among other values, corrective justice is realised. It seems, first, difficult to argue that the parties to a public law dispute are equal in any realistic empirical sense. For there is surely always and ever a significant inequality of power between the parties to a public law dispute, since one is usually a representative or manifestation of the state while the other is not. Secondly, it is also very difficult to see how, in a public law dispute like the hypothetical one noted above, the defendant local authority’s gain (in refusing to issue a cab licence on inappropriate grounds) is in any way equal or correlative to the claimant taxi driver’s loss. Is the argument that corrective justice plays a role in public law thus stillborn? Not necessarily. One difficulty with Weinrib’s Aristotlelian account of corrective justice is that of unpacking the sense in which the parties to a private law dispute are equal. Weinrib accepts that the equality of the parties, both in terms of their standing qua claimant and defendant, and in terms of the gains and losses instantiated in their private law dispute, is primarily notional (or normative) and not factual.26 Thus, the alleged ‘gains’ I achieve from carelessly running you down are not, first and foremost, ‘real’, ‘empirical’ or ‘tangible’ gains which exactly equal your loss. Indeed, whatever ‘real’ losses you suffer from being struck by my negligently driven car (physical injury and the related pain and suffering, as well as the losses associated with medical expenses, time off work etc) seem impossible to correlate with whatever gains, if any, I achieved from my momentary lack of concentration. Your real losses seem to far outweigh whatever real gain I received. Two questions must be answered about scenarios such as these. In what sense are (i) the parties and (ii) their gains and losses equal? There is a reasonably clear sense in which the parties can be regarded as normatively equal: each is taken to be a free agent with the same requirements or needs for manifesting their agency. 25 Nicomachean Ethics, Book V, paras 1132a25–1132b21 in J Barnes (ed), The Complete Works of Aristotle (Princeton NJ, Princeton University Press, 1984) vol 2. See Weinrib, above n 17, at 65. 26 Weinrib, The Idea of Private Law, above n 17, at 62–3, 76–83; and chs 4 and 5. Henceforth I will speak only of normative equality. Weinrib moves between the two (above n 17, at 62–78), but eventually settles on ‘normative’ equality (above n 17, at ch 5). For our and Weinrib’s purposes, the significant similarity of meaning between the two terms is this: both mean ‘non-factual’.

Duty of Care in APrivate Rights-Based TheoryLaw? of Negligence What’s About Private

55

To realise myself in the world, the very least I need is bodily integrity, some freedom of movement and some minimal level of well-being. All other agents, if they are to realise their agency, require the same. Thus all agents are equal in the abstract sense of needing the same ‘things’, and thus the same spheres of protection, in order to be agents. Agents are therefore equal in standing, insofar as all either are, or are presumed to be, capable of agency, and equal in what they require, in general and minimal terms, to be agents (bodily integrity, freedom and well-being). This account of the equal standing of the parties can be extended to give an account of equality of gains and losses in corrective justice. A corrective justice wrong could be understood as conduct by one party that undermines or removes one or more of the conditions of another agent’s agency. The transgressor obtains an advantage in undermining my agency correlative to the amount my agency is undermined: his freedom and well-being, we might say, is extended to the same degree to which mine is undermined. This extension is not without strain, for the exact nature of the correlativity (or equality) between wrongdoer’s gain and victim’s loss is, to say the least, somewhat oblique. But this at least gives us a sense of how the parties’ equality of standing might inform an equality of gains and losses. Weinrib, too, uses an account of the parties’ equality of standing to provide an account of the correlativity of gains and losses in corrective justice. And because the account of the parties’ equal standing is normative, so too is the account of correlativity (or equality) in gains and losses. Weinrib’s account is not, however, like the one just offered: his account unfolds via Kant’s account of right, the precise details of which need not detain us here.27 It is important to note, though, that—as the account offered in the previous paragraph makes clear—Kantian right is not the only way in which the normative equality of the parties and the normative equality of their gains and losses could be unpacked. True, that is the path Weinrib takes. It is not, however, the only one available and some reason needs be offered in its favour over and above its fit with Aristotle’s (and Weinrib’s) account of corrective justice. Such a reason is required if the normative account of equality is to inform and explain, rather than be informed and explained by, corrective justice. Equally important is a point as obvious as it is significant: the doubled-sided notion of equality animating Weinrib’s account of corrective justice is normative not factual. The significance of this is that some conception of normative equality, in terms of both equality of standing and equality of gains and losses, is surely as applicable to public law as to private law disputes. That there is often—perhaps always—a real imbalance of power between the parties to a public law dispute, a factual inequality, is of no more significance from the perspective of Weinrib’s or Kant’s normative conception of equality than is a related imbalance of power in a private law dispute. Nor can it be argued that the factual difficulty of equating 27 See Weinrib, The Idea of Private Law, above n 17, ch 5. The account offered in the previous paragraph is broadly Gewirthian. The starting point is A Gewirth, Reason and Morality (Chicago IL, University of Chicago Press, 1978).

56

William Lucy

the licensing authority’s gain with the taxi driver’s loss means that this public law dispute cannot be conceived in terms of corrective justice. It might, however, be argued that the licensing authority’s normative gain cannot be correlated with the taxi-driver’s normative loss in terms of Kantian right. But the question then arises as to the explanatory power of Kantian right. Why, given that there are other possible normative accounts of the parties’ equality of standing and of gains and losses, does the latter dominate? That Weinrib’s account of corrective justice invokes a normative conception of equality does not justify the judgement that all factual or real inequalities between the parties, either of standing or in terms of their gains and losses, are irrelevant. Factual inequalities are sometimes also markers of normative inequalities. Moreover, a reminder that the obvious factual inequalities of standing and of gains and losses in public law disputes are prima facie irrelevant, while helpful, does nothing to show that the required normative equality of standing and of gains and losses is actually present in such disputes. This could only be successfully shown on a case-by-case basis but, pending such an analysis, it is helpful and possibly prescient to note the presence of both aspects of normative equality, understood in terms of Kantian right, in the most unpromising of private law cases. Thus, according to Weinrib, this equality exists not only in standard negligence cases, but also presumably in cases of breach of contract in the absence of pecuniary loss.28 If normative equality of gains and losses can stretch this far, then is there any obvious block to extending its range to public law gains and losses? It is also, surely, apposite to note an obvious normative-cum-juridical equality of standing in public law, particularly in many common law jurisdictions. Here, the state or Crown has legal personality in exactly the same form as do citizens; the ordinary rules of private and criminal law apply as much to the Crown as to any other citizen, unless specific legal provisions grant Crown immunity. Both Crown (or state) and citizen are supposedly equal before and under the law.29 It is not, then, immediately obvious that public law disputes cannot be accommodated within Weinrib’s account of corrective justice. In particular, there is no a priori reason why the normative conception of equality that underlies the equal standing of the parties and the equality of gains and losses in private law cannot include public law as well. And, even if there is such a reason, it cannot circumvent questions about the legitimacy of Weinrib’s Kantian conception of equality: why choose that path rather than others? There might, nevertheless, be an a priori reason why corrective and distributive justice cannot be combined. For, although it might be possible to conceive of public law disputes in corrective justice terms, 28 See Weinrib, The Idea of Private Law, above n 17, at 136–40 and ch 6. A scrupulous but understated analysis of Weinrib’s treatment of contract and related issues is L Smith, ‘Restitution: The Heart of Corrective Justice’ (2001) 79 Texas Law Review 2115, 2129–35. 29 For an overview of the Crown’s legal status in the UK see M Sunkin and S Payne (eds), The Nature of the Crown: A Legal and Political Analysis (Oxford, Clarendon Press, 1999). An expansive analysis of the law in Canada is PW Hogg and P Monahan, Liability of the Crown, 3rd edn (Toronto, Carswell, 2000).

Duty of Care in APrivate Rights-Based TheoryLaw? of Negligence What’s About Private

57

it is a mistake to do so, since this is to combine two notions—corrective and distributive justice—which cannot be coherently combined. This claim provides a warrant for discounting the views of those jurists and practitioners who regard not just private law, but also public law, as realms in which both corrective and distributive justice are in play. It shows how a theory of private law that purports to take the views of participants seriously can nevertheless discount some of those views. What, then, is this reason and what weight does it have? The reason is the separation thesis itself. In Aristotle it has two principal components, the first identifying the two forms of justice, the second articulating their relationship. As to identity, Aristotle tells us that distributive and corrective justice are recognisable and distinguishable because they embody different mathematical operations—the former consists of geometrical equality, an equality of ratios (or proportions), while the latter is arithmetical equality or an equality of quantities.30 It seemingly follows from this characterisation that the two cannot therefore be intelligibly combined: their relationship is one of incompatibility.31 Weinrib accepts the thesis, including this strong incompatibility claim, wholesale. The strength of the latter claim should not be overlooked. It becomes vivid when we note that few who regard corrective and distributive justice as different also accept that they cannot be combined.32 The basis of the separation thesis in Aristotle seems to be the logic of the concepts themselves, but he provides little by way of argument to support this claim. Even if we accept that the two different terms should or must refer to different concepts, this clearly does not show that these concepts are necessarily incompatible. Nor, of course, does it show that one should displace or dominate the other in any particular or even in all conceivable contexts. It is clear that corrective and distributive justice can be understood in the way Aristotle suggests: the claim that they embody different mathematical operations is not radically implausible, as would, for example, be the claim that they instantiated different colours. But we surely need more than a bare possibility and absence of radical implausibility to accept the claim. This is not just because Aristotle’s way of conceiving of corrective and distributive justice seems, within our current intellectual context, somewhat odd. That there is a rich tradition of thought about distributive justice that understands it neither in the formal manner of Aristotle’s mathematical operation, nor in Aristotle’s substantive terms, solely as a matter of desert, is not a conclusive reason to reject Aristotle’s view.33 More worrying than 30

Nicomachean Ethics, above, n 25, paras 1131b12–1131b13 and 1132a32–1132a33. But see K Marc-Wogau, Philosophical Essays: History of Philosophy. Perception. Historical Explanation (Copenhagen, Ejnar Munksgaard, 1967) ch 2. 32 See Coleman, above n 16. An interesting view of the relationship between the forms of justice is J Finnis, Natural Law and Natural Rights (Oxford, Clarendon Press, 1980) 177–93. For expansive views of the components of distributive justice see M Walzer, Spheres of Justice (Oxford, Blackwell, 1983) and D Miller, Social Justice (Oxford, Clarendon Press, 1976) and his Principles of Social Justice (Cambridge MA, Harvard University Press, 1999). 33 The high point of this tradition being J Rawls, A Theory of Justice, Revised edn (Oxford, Clarendon Press, 1999). 31

58

William Lucy

this is the fact that strong reasons supporting the impermeable boundary between corrective and distributive justice are simply absent from Aristotle’s own treatment. The distinction, judged solely within the boundaries of the Nicomachean Ethics itself, is plainly little more than a stipulation.34 It is thus not a particularly rationally robust foundation upon which to build a strong, binary distinction between private and public law.

III. A Distinction with a Difference The grounds for regarding Weinrib’s separation thesis as generating a distinction between public and private law include not just his invocation of Aristotle’s distinction between the forms of justice, but also some admittedly fleeting comments on public law itself.35 The grounds for regarding this attempt to distinguish private and public law as problematic include not only the doubts about the plausibility of the separation thesis itself, but also a suspicion that the hallmarks of bipolarity displayed by private law disputes are also displayed by public law disputes. If this effort to give normative depth to the legal-doctrinal distinction between public and private law is therefore set aside, where else might such depth be sought? What follows is an account of another possible foundation for the distinction between public and private law, although it is not one that claims to map, in any exact way, onto the current legal-doctrinal distinction. Legal doctrine in the common law world is a malleable medium, providing a fairly inhospitable habitat for rigorous and rigid distinctions of any kind, be they purely legal, purely moral, purely political or some admixture of the three.

A. Private and Public Project Pursuit Put brutally, the distinction is this: it is between the value we place upon individuals having the freedom to formulate and pursue their own projects without warrant from other citizens or the state, and the value we place upon our collective power, primarily through law and politics, to change and maintain the social and related structures in which we live.36 If it is true that we value these two

34 The distinction is outlined in Nicomachean Ethics, above n 25, at paras 1131a10–1132b21. For elaboration see EJ Weinrib, ‘Aristotle’s Forms of Justice’ in S Panagiotou (ed), Justice, Law and Method in Plato and Aristotle (Edmunton, Academic Printing and Publishing, 1987) particularly at 134–42. 35 Weinrib, The Idea of Private Law, above n 17, at 8 and 48. 36 This is a paraphrase of NE Simmonds, ‘The Possibility of Private Law’ in J Tasioulas (ed), Law, Values and Social Practices (Aldershot, Dartmouth, 1997) ch 6, 144. See also his ‘Justice and Private Law in a Modern State’ (2006) 25 University of Queensland Law Review 229–52. Cane’s account of the public/private distinction runs along broadly similar lines but has a very different intellectual lineage. See Cane, above n 3, at 273–5.

Duty of Care in APrivate Rights-Based TheoryLaw? of Negligence What’s About Private

59

realms, and liberal polities must by definition do so, then it needs be established that they are indeed (i) distinct and (ii) similarly important or equally worth upholding, yet (iii) capable of undermining one another. Only if these conditions are satisfied do we have a distinction marking a difference truly worth upholding. It seems that these realms are indeed distinct. It is both physically and normatively possible for me to devote all my leisure hours to bird-watching or collecting and drinking fine wine or playing amateur football. I need not seek any one’s or any body’s permission to engage in the former pursuits, although for the latter I will need a group of friends or will have to join a club. The costs and benefits of these possible uses of my time do not have to be determined before I am permitted to continue. Indeed, it might well be the case that my pursuit of these hobbies harms me: I could, as a result of relentless bird-watching or wine drinking, become even more sadly misanthropic than I already am, while constant football playing might give me arthritis. But even here, if I do not harm others, my hobbies are regarded as no one’s business but my own. These are just some of the projects I am free to formulate and pursue. Others go beyond the realm of leisure. My choice of employment or career is, as most reliable liberals would agree, a matter for me and no one else, except those I might consult or whose interests I value. The assignment of careers or modes of employment by a centralised agency pangs not only of the workhouse, but also of the forced labour camp. Similarly the choice to have a family or remain childless is one we assume almost all adults have the right to make; this is not a realm in which government intervention is either welcome or appropriate. Many consumption choices are, we think, also entirely our own: whether I drink coffee or tea, or whether I live frugally or extravagantly, is the legitimate concern of no one but myself. These choices need not be validated by others, either individually or in some collective form, before I am allowed to act upon them. Nor are individuals’ choices about the nature of their emotional and sexual lives subject to such unpalatable ‘interference’. This is assuredly part of the realm of private and not public project pursuit. The importance of individuals being both allowed to author large aspects of their own lives independently of the directives of the community, and being protected in that self-authoring, is a theme well-explored by many liberal thinkers. It might be argued that being able to create many significant aspects of their lives for themselves, including not just general matters like a ‘life-plan’ but relatively small aspects of daily life like particular consumption choices, allows human beings to approximate more closely a fine form of human existence and flourishing. Private project pursuit, on such a view, is one feature of what a good human life looks like. Yet the goodness of such a life does not necessarily arise from the content of the actual choices particular agents make, for these might in some or many cases turn out to be bad choices. Rather, the value of such a life might be found in the fact that such choices can be and are made, whether they turn out to be good or bad. It is the making of these choices, the opportunity to create tracts of one’s own life for oneself, which is truly significant here.

60

William Lucy

The value of private project pursuit need not, however, be rooted in a perfectionist picture of what a good human life looks life.37 It could instead be based upon scepticism about humankind’s ability to know what such a life looks like, and a worry about the use of political power to realise and maintain such a picture. Human history could well suggest that allowing individuals to author large aspects of their own lives creates far less suffering and pain, or far more happiness or preference satisfaction, than when their lives are in large part created for them. But the basis used to articulate the value—or, more likely, values—of private project pursuit does not matter for present purposes. The significant point is that there is a high degree of convergence among both philosophers and citizens that private project pursuit is indeed valuable, even though there may be disagreement as to why, exactly, this is so. Equally clear is the commitment that flows from valuing individuals’ ability to author large parts of their own lives, namely, valuing and protecting the means of such self-authoring. At the very least, this entails a commitment to protecting bodily integrity and freedom of movement, as well as other conditions of individual autonomy. The realm of public project pursuit could—and probably should—be defined as all that is not private project pursuit. While not particularly helpful, this definition at least avoids the trap of claiming that public project pursuit is about nothing more than the realisation of what economists and social choice theorists call ‘public goods’. Such goods are non-rival and non-excludable, but their provision, while being of benefit to many, cannot be protected against free-riders.38 It is thus thought that, in many cases, such goods are best provided by government action of some kind—usually in the form of non-consensual taxation—since this is often the most effective way of reducing or even eliminating free-riding. All citizens, it is assumed, can legitimately be forced to contribute to the realisation of goods from which they benefit. National defence, the system of property and related rights, public order and law enforcement are but three benefits commonly regarded as good examples of (fairly) pure public goods. It is also argued that constitutions, bills of rights, a reliable law of contract as well as good government are relatively pure public goods.39 The problem with the private goods/public goods taxonomy for present purposes is obvious from this list: private law itself is a public good. The private good/public good distinction cannot therefore be used to distinguish private and public law.

37 The most interesting recent version of perfectionist liberalism was offered by J Raz, The Morality of Freedom (Oxford, Clarendon Press, 1986). There is a good deal of controversy over the extent to which, if at all, JS Mill’s account of the advantages of individualism is perfectionist: see ch III of ‘On Liberty’ in his Utilitarianism, On Liberty and Considerations on Representative Government (London, Dent Publishers, 1972). For an explicitly non-perfectionist liberalism, see J Shklar, ‘The Liberalism of Fear’ in NL Rosenblum (ed), Liberalism and the Moral Life (Cambridge MA, Harvard University Press, 1989) ch 1. 38 The classic treatment of public goods is P Samuelson, ‘The Pure Theory of Public Expenditure’ (1954) 36 Review of Economics and Statistics 387–9. 39 The discussion in J Buchanan, The Limits of Liberty (Chicago IL, University of Chicago Press, 1975) chs 1–4 and 7, is a significant starting point.

Duty of Care in APrivate Rights-Based TheoryLaw? of Negligence What’s About Private

61

But what does the realm of public project pursuit include, beyond some public goods? A slightly more informative answer than the first one is this: all that activity which is conducted through the various instruments of the state. In most liberal democracies, the provision of elementary level education and health-care is organised through national bodies supported by taxation and controlled, at some level, by national government or by a combination of national and local government. The actual raising of taxation itself is surely an instance of public project pursuit, as in most societies is the provision and upkeep of national defence, transportation networks and public spaces. There is, of course, nothing necessarily ‘public’ in the provision of these services in the sense that ‘public’ (governmental) bodies need not necessarily provide them. They can often be provided by ‘private’ (non-governmental) bodies. One point of significance, though, is that although the service provider need not be public in this sense, the payer for the service almost always is: it is central and local governments who fund such services from local and national taxation, rather than ‘private’ subscribers. That public project pursuit includes activity conducted through various instruments of the state hints at something else that should be included in that class. Are the various instruments of the state constrained in any way in their pursuit of objectives? Are there even some objectives that are and should be beyond their range? The answer to both questions, in the liberal democracies at least, is affirmative. The principal source of constraints on the ends that can be pursued, as well as upon the ways in which permissible ends can be pursued, are the systems of constitutional and administrative law in these societies. Indeed, it is these areas of law that are taken to be paradigmatically ‘public’ because they mediate the relationship between state (and its various instruments), on the one hand, and the citizen, on the other. The value of public project pursuit at its most general is the value of coordinating and organising various aspects of our collective life together. While such collective endeavour might in some sense be inherently valuable, it is often regarded as valuable only insofar as it brings about more specific values, goals or states of affairs. It is thus not wrong to say that in some instances the value of public project pursuit depends upon the particular goals pursued and achieved. Some public projects—the elimination of inflation, for example—are both short-term and clearly not inherently valuable. The elimination or reduction of inflation is valuable only insofar as it advances other valuable goals or states of affairs: higher employment, an increased standard of living, or a better balance of payments. The effort to reduce inflation will be promptly abandoned once it ceases to achieve those goals or states of affairs. Similarly, investment in particular areas of public transport, or aspects of health-care or education can at any point be scaled back or advanced, depending upon the specific aims of such investment. For, while few would deny that health-care, education and public transport are goods worth achieving and developing in a society, it is hard to argue that investment in each should be unlimited. A once vital public project can thus become only marginally significant or completely insignificant. There is, however, a component of

62

William Lucy

public project pursuit that remains fairly statically valuable. This component is unconcerned with possibly transient policy or community goals; it is that aspect of public project pursuit that consists of the values (and constraints) in accordance with which such goals are pursued. Those values are now most obviously found in constitutions and bills or charters of rights and freedoms, taken in conjunction with the values of procedural fairness or administrative due process. The ‘value’ of public project pursuit is thus, like the value of private project pursuit, a complex question. A more discriminating approach to both would at least begin with a cartography and analysis of the various particular values realised by these bluntly conceived macro-values. It could proceed to an evaluation and comparison of each particular putative value within each more general class. While we cannot say in advance what the results of such an analysis might be, we can surely say that public and private project pursuit, and the values constitutive of and realised by each, are both very important. They might, indeed, be equally important, although this idea plainly implies a single metric upon which both can be ranked, each option or set of options having the same weight. If both are genuinely of equal weight, how shall we choose between them? The choice surely cannot be compelled by reason. If it were, that would undermine the claim that each option is of equal weight. Furthermore, priority rules might resolve a choice between equally ranked options, but if the options are indeed genuinely equal then the priority rules clearly cannot reflect the merits of the options. If they did, the priority rules would either fail to resolve the tie, since the merits are ex hypothesi equal, or they would show that there is not after all a tie between equals here. Priority rules must therefore be based on considerations other than those which give value to the options in play. Such rules might, for example, be based solely upon considerations of efficiency in decision-making, insisting that one option be chosen on random grounds if necessary. Recourse to randomness shortcircuits what is likely to be a long and relatively fruitless deliberative process. It is also a means of circumventing decisional paralysis, which, of course, is particularly important if the decision has consequences for the conduct of others. My inability to decide between Earl Grey and Breakfast Tea is of little moment to anyone but myself, whereas an appellate court’s decision as to the validity of a particular construction industry liquidated damages clause could affect the whole industry. An equal ranking of the respective value of public and private project pursuit is not the only conceptual possibility here. This much was hinted at by the curious locution ‘similarly important’ in point (ii), at the beginning of this section. For the relation between these two values—or, more accurately, congeries of values—might be better understood as that between incommensurables, even though the practical consequence of this is much the same as if the values were equally ranked. Two options are incommensurable if one is neither more nor less valuable than the other, nor are they equal in value.40 A common shorthand

40

See Raz, above, n 37, at 322.

Duty of Care in APrivate Rights-Based TheoryLaw? of Negligence What’s About Private

63

characterisation holds that incommensurability marks the fact that there is no common metric by which two or more allegedly incommensurable values can be compared. This characterisation is unexceptionable provided it is not taken as the beginning of a quest for such a metric, for those who believe incommensurability to be a hallmark of our values usually also believe that no such metric will or can be found.41 For these thinkers, incommensurability is in no sense a failing in our moral, political and legal value systems but, rather, an undeniable and significant truth about them. The significance of this truth is manifest in at least two ways. First, choices between incommensurables cannot be compelled by reasons, in the sense that the choice of one genuine incommensurable X, over another, Y, is a clear requirement of rationality. Or, put in slightly different terms, the choice of one genuine incommensurable over another genuine incommensurable cannot ever be a rational mistake, based upon an erroneous assessment of the reasons pro and con. Yet, although choices between genuine incommensurables cannot be compelled by reasons, such choices can nevertheless be based upon reasons.42 Secondly, the existence of incommensurability supposedly explains our intuitive reactions in many choice situations. On the assumption that these reactions are incorrigible, incommensurability is necessary in order to account for moral (and related) experience. It makes sense of something that is otherwise oblique and this, it is thought, gives the idea validity and salience. If incommensurability is a feature of some of our values, then choices between incommensurables take much the same form as choices between exactly equally ranked values. The choice between the latter is, in the absence of priority rules, one that is very hard and sometimes seemingly impossible to make. The difficulty, the perceived impossibility, results from a lack of compelling rational guidance: it is simply hard to find grounds for the choice. Exactly the same can and, we might expect, would always be true of choices between incommensurables. There will be reasons pro and con, none of which are obviously rationally more compelling than their competitors. The choice is thus not one we are completely comfortable making; the choice process is anguished, one in which we may often change our minds before eventually coming to a resolution. The resolution is much more an instance of picking rather than choosing.43 It is obvious that the value or values embodied in private project pursuit, and that or those involved in public project pursuit, can undermine one another. While it might be almost practically impossible to reduce the realm of private project pursuit to zero, since some few individuals could presumably always successfully evade surveillance, it can be massively reduced simply by requiring permission for almost any form of human conduct. Were permission granted

41 Raz, above, n 37, at 327. See also the essays by E Anderson, C Taylor, M Stocker and J Finnis in R Chang (ed), Incommensurability, Incomparability, and Practical Reason (Cambridge MA, Harvard University Press, 1997). 42 Raz, above, n 37, at 339. 43 E Ullman-Margalit and S Morgenbesser, ‘Picking and Choosing’ (1977) 44 Social Science Research 757.

64

William Lucy

only if the conduct in question contributed to the common good, then the space for private project pursuit would all but disappear, since one could pursue one’s private projects only if they also contributed to public projects.44 At this point, it is appropriate to wonder whether the term ‘private project pursuit’ has any genuine meaning. That private project pursuit can undermine public project pursuit might surprise some. Yet it is surely the case that sufficient individuals’ incentives could be unintentionally structured so as to prevent the provision of public goods properly so called. A call for contribution towards, for example, the provision of street-lighting in a neighbourhood would fail if those living there could each provide lighting and security for themselves at lower cost. Furthermore, the provision of community goods over and beyond public goods might well be thwarted by attitudes engendered by the promotion of private project pursuit. It seems, for instance, that the valorisation of private project pursuit among a population can undermine that population’s interest in, and willingness to support, various forms of community activity and participation.45 If the differences between public and private project pursuit, and the congeries of values both implicate, are genuine, then what significance does this have for law? A tempting and, in this context, unsurprising answer is that this set of differences can be parlayed into a distinction between public and private law. The distinction between public and private project pursuit simply acts as a framework by which to distinguish the two areas of law: private law is a means of securing and facilitating the realm of private project pursuit, while public law is a means of upholding and facilitating public project pursuit. Since both realms are either equally valuable or incommensurable, and since any one instance of conduct can often have both private and public aspects, we can expect a constant tension between them. Both realms make equally insistent or incommensurable claims upon us and the resolution of these claims, either at large, or within the legal context, is likely to be neither neat nor easy. This way of distinguishing private and public law might be different from many previous attempts but it does not, of course, bring that distinction into being. It does not make visible that which was invisible. Since most mature legal systems at the very least display the sequelae of a distinction between public and private law, and since they also grapple, often inconclusively, with the task of adequately distinguishing these two realms, the current version of the distinction can claim only to make better sense of these sequelae and that process.46 Of course, whether or not this particular version of

44 The argument that property is a form of sovereignty over fellow human beings (a classic instance of which is M Cohen, ‘Property and Sovereignty’ (1927–28) 13 Cornell Law Quarterly 8, 11–14 and 27–30) can lead to the conclusion that it should be subject to exactly the same constraints as any exercise of government power. One such regularly invoked constraint is, of course, the common good. 45 See RD Putnam, Bowling Alone (New York, Simon & Schuster, 2000), a focal point of the contemporary debate on this and related issues. 46 Efforts to formulate a distinction can be traced back at least to Justinian’s Institutes. An interesting historical and comparative treatment is A Tay and E Kamenka, ‘Public Law—Private Law’ in Benn and Gaus (eds), above n 2.

Duty of Care in APrivate Rights-Based TheoryLaw? of Negligence What’s About Private

65

the distinction between private and public law is indeed actually better than the alternatives remains to be seen. Before any particularly robust judgement can be arrived at on that issue, it is important to appreciate what is not being claimed on behalf of this way of distinguishing public and private law. The distinction does not purport to be a historical one: it is primarily normative. The historical genesis and cultural context of any particular version of the distinction between public and private law is undoubtedly an interesting issue, but it is not one upon which this essay has anything to say, save that history would indeed be kind if it embodied all of our significant normative distinctions. Nor is it being claimed that this way of distinguishing private and public law will be a great boon in deciding particular cases. Yet, by viewing private and public law through the prism of a normative distinction, the argument offered here is open to an obvious challenge. It can be captured with this question: Why should this philosophical, normative distinction take priority within the law? Furthermore, by eschewing any great instrumental role in the decision of particular cases, the distinction between public and private law recommended here is surely redundant. These two objections are the fulcrum of the following two sections.

B. The Priority of Philosophy? The distinction between private and public law just elucidated is not a direct inference from or strict implication of some authoritative legal source, nor is it explicitly and fully enshrined in the traditional texts and commentaries of the law. It is a philosophical-normative distinction, provided this rather grandiose term is taken to indicate nothing other than the output of somewhat trite and low-level philosophical reflection.47 Yet even if this relative superficiality is overlooked, there is another more obvious problem faced by the distinction: Of what weight is it, or can it be, in law? Why should judges or lawyers attend to it? It has no authority in the technical legal sense and therefore cannot be binding upon courts. Moreover, surely this distinction merely replaces a small and relatively shallow legal swamp with a large philosophical quagmire. Shifting the site of discussion from the pragmatic legal-doctrinal terrain of particular cases to the realm of broad philosophical distinctions is a prima facie unpromising way of resolving a puzzle and, furthermore, risks delaying further the resolution of particular cases. The distinction seems to invite judges to navel-gaze rather than decide cases. In addition, how can it ever be legitimate for judges’ philosophical commitments, or their views as to the precise contours and implications of broad philosophical distinctions, to determine their legal decisions? 47 The distinction can, however, be given a grander philosophical framework such as that outlined in A Brudner’s unjustly overlooked The Unity of the Common Law (Berkeley CA, University of California Press, 1995) ch 1. The tension between private and public project pursuit is subject to a Hegelian synthesis by Brudner: at 14 (a sketch of the apparently paradoxical nature of this task).

66

William Lucy

This two-sided difficulty can be labelled ‘the philosophy problem’, one dimension being that of decisional inefficiency, the other raising the issue of legitimacy. As will be seen, both dimensions of the problem are exaggerated. That enmiring judges in philosophical controversies will not help them decide cases is a sub-theme in Judge Richard Posner’s work. Posner is specifically concerned with a strand of moral philosophy he calls ‘academic moralism’ and disagrees with jurists, such as Ronald Dworkin and others, whose jurisprudential advice to practitioners culminates in the claim that judges are (and should be) moral philosophers.48 Whether Posner does indeed accurately characterise the position of his opponents on this issue, and the merits of that position more generally, are not of concern here.49 Rather, our interest is in the decisional inefficiency that Posner thinks arises from judicial immersion in moral-philosophical discourse. If Posner is right about this, the problem might extend to judicial immersion in any kind of philosophical discourse, even such a minor aspect of it as articulating and mediating the distinction between private and public project pursuit. Posner’s ‘weak thesis’ is that academic moral-philosophical positions have not impacted upon the judicial resolution of particular important cases, combined with the claim that such positions actually cannot contribute to that process.50 The latter claim appears to have a double foundation. Posner argues, first, that the outputs of academic moralists’ arguments are significantly inert and that they therefore have little impact on the deliberative processes of either judges or citizens at large.51 Secondly, he argues that even if the issues raised in some legal cases could be tackled through the concepts, techniques and arguments of academic moralism, they will not in practice be so tackled by judges. This is in part because judges, in Posner’s view, are just not comfortable with such arguments, and in part because there is usually a plethora of other kinds of argument that judges can and will use in such cases.52 Do these objections to academic moralism also apply to the judicial use of broadly philosophical concepts or distinctions? This seems unlikely, if only

48 R Posner, The Problematics of Moral and Legal Theory (Cambridge MA, Harvard University Press, 1999) chs 1 and 2. A slightly shorter version of the book was published, under the same title, in (1998) 111 Harvard Law Review 1638. 49 Perhaps the best and shortest statements of this position are R Dworkin et al, ‘Assisted Suicide: The Philosophers’ Brief ’ (1997) 44(5) New York Review of Books 41 and R Dworkin, ‘Must Our Judges be Philosophers? Can They Be Philosophers?’ New York Council for the Humanities Scholar of the Year Lecture (2000) at www.nyhumanities.org/soylecture2000html, accessed on 29 April 2003. Three instalments in what became a somewhat ill-tempered exchange between Dworkin and Posner on the issues raised in Posner, ibid, are R Dworkin, ‘In Praise of Theory’ (1997) 29 Arizona State Law Journal 353; ‘Darwin’s New Bulldog’ (1998) 111 Harvard Law Review 1718; ‘Philosophy and Monica Lewinsky’ (2000) 47(4) New York Review of Books 48–52; and R Posner, ‘Conceptions of Legal “Theory”: A Response to Ronald Dworkin’ (1997) 29 Arizona State Law Journal 377; ‘Reply to Critics of The Problematics of Moral and Legal Theory’ (1998) 111 Harvard Law Review 1796, 1796–1806; ‘An Affair of State: An Exchange’ (2000) 47(7) New York Review of Books 60–65. 50 Posner, above, n 48, at 1639 and 1693–1709. 51 ibid, at pts I and II. 52 ibid, at 1698–1708.

Duty of Care in APrivate Rights-Based TheoryLaw? of Negligence What’s About Private

67

because the latter process seems neither directly nor indirectly reducible to academic moralism. The path from one to the other is far from clear. Moreover, such a path is assuredly barred if judicial use of broadly philosophical concepts is, as Posner himself recognises, unavoidable insofar as concepts such as (inter alia) causation, responsibility and intention are the fulcrum of many areas of legal doctrine.53 Furthermore, the specific failing that besets academic moralism, according to Posner, is not necessarily a general failing of all philosophical positions. While Posner might be right that academic moralism rarely motivates conduct and thus is relatively inert in practical deliberation,54 the truth of this claim entails nothing about the effect that general philosophical distinctions have upon either conduct or deliberation. Posner’s attack, remember, strikes only at a very limited sub-class of the broader category of philosophical concepts and discourse. Objecting to a particular species of discourse cannot, without more, undermine the whole genus (assuming the latter is not a single species class). It is also a mistake to think that this version of the private/public distinction is solely or purely philosophical; for the distinction is also embodied in the political culture of the mature liberal democracies, and it is entirely appropriate, as well as absolutely necessary, for judges to operate within and uphold crucial aspects of that political culture. That the distinction is part of our political culture does not guarantee its ultimate normative significance. Being immanent within a particular political culture is not of itself a demonstration of the genuine value of some particular putative value or set of values. Such a demonstration is far beyond our current task. Rather, our immediate aim is to show that public and private project pursuits are each embedded and valued in our political culture. If they are, then judges have much more to work with than a ‘merely’ philosophical distinction. But is the distinction indeed ‘there’, truly rooted in our political culture? Showing this requires, at the outset, an answer to a prior question: What is ‘political culture’? A famous political science account holds that it is, inter alia, ‘the political system as internalized in the cognitions, feelings, and evaluations’ of the population of a particular nation state.55 While not completely unhelpful for current purposes, this statement misses much that is of significance by saying nothing about the institutional manifestation of these values, beliefs and attitudes. If, for example, the value of individual freedom is said to be part of the 53 Posner denies that the role these notions play in the law makes academic moralism relevant: see ibid, at 1693–7 and ‘Reply to Critics of The Problematics of Moral and Legal Theory’, above n 49, at 1804–5. 54 ‘[T]he arguments pro and con are too flaccid to induce a sensible person to change his beliefs or behaviour’: ‘Reply to Critics of The Problematics of Moral and Legal Theory’, above n 49, at 1801. ‘Every move in normative moral argument can be checked by a countermove. The discourse of moral theory is interminable because indeterminate’: (at 1802). 55 G Almond and S Verba, The Civic Culture (Princeton NJ, Princeton University Press, 1963) 13. They also have a less digestible version: ‘The political culture of a nation is the particular distribution of patterns of orientation toward political objects among members of the nation’: (at 13). Almond and Verba’s view of political culture has not changed significantly: see G Almond and S Verba (eds), The Civic Culture Revisited (Boston, Little Brown, 1980) ch I (at 26–8) and ch X (at 395–7).

68

William Lucy

political culture of some nation state, it is not unreasonable to expect that value to be evidenced not just in a tradition of discussion about its form and nature, but to exist also in that state’s formal repositories of public values. Some obvious repositories would include not just that state’s constitutional documents, international legal obligations and associated case law, but also the explicit policy goals, or the parameters upon such goals, set by the government of the day and the political parties, and other sources like reports of parliamentary and related committees. Furthermore, it is not unusual to find local government bodies, as well as large corporations and semi-public/private bodies in liberal democracies, committing, either through charters or other statements of values and goals, to carry out their functions consistent with rights of all citizens of the polity. Now, although a scrupulous empirical study of these sources cannot be undertaken here, listing them (even non-exhaustively) is surely sufficient to show where any state’s expressed commitments to private and public project pursuit are likely to be found. The core claim is, I hope, non-controversial: that the political culture of the mature liberal democracies includes not just the attitudes, values and beliefs of existing citizens, but also the history of public elaboration and institutionalisation of those attitudes, values and beliefs. That these many and varied sources of public elaboration and institutionalisation exist is beyond doubt; more speculative, however, is any claim about their precise content. This essay relies on one such speculative claim, namely, that it will not be hard to find among these sources copious evidence of a commitment to both the goals and values of public and private project pursuit.56 The distinction, then, most likely has an institutional life or embodiment and this might well calm those who worry about the prospect of judges mediating, or grappling with, a supposedly vague or intractable (merely) philosophical distinction. The distinction is not just philosophical: it is ‘ours’ collectively, a part of our current polity and a feature, albeit mediated and contested, of our political history and tradition. This particular worry thus seems quite easily dissipated, although it might quickly be replaced with another. This other worry can be highlighted by a question: Ought judges to be drawing upon their country’s political culture in interpreting and applying the law? The issue here might seem to be that of the legitimacy of using this politicalcum-moral distinction to structure the law. But simply posing the issue in these terms serves to show that it is bogus, for what is being questioned here—the legitimacy of judges working within and upholding the values that constitute our political culture—is something unavoidable and, indeed, necessary. Judges cannot, in fact, do otherwise than interpret the law within the political culture of which it is part because—obviously!—it is part of that culture. This is undeniable unless we believe that those features of our legal systems, like the rule of law, bills of rights, natural justice and the separation of powers, cannot also be part of our 56 Even a starting point confined to purely legal sources furnishes much: see the European Convention on Human Rights, Arts 2–5, 8–12 and 14 and the Human Rights Act 1998 ss 1–6.

Duty of Care in APrivate Rights-Based TheoryLaw? of Negligence What’s About Private

69

political culture. The claim that these features are not part of the political culture of the mature liberal democracies seems prima facie very puzzling, apparently assuming a bright-line distinction between easily separable ‘political’ and ‘legal’ values and ideas. As a matter of intellectual and cultural history, this assumption is plainly unfounded: that these ideas developed within, and have subsequently had, very different spheres of influence is false. The fact that these key juristic values were (perhaps only partially) realised in legal systems as the result of political change and struggle which, in Europe at least, was part of the change from feudal systems of governance and economic production to bourgeois governance and capitalist economic production, demonstrates this. The distinction could only be insisted upon by one already committed to the view that law and its key juristic values must exist in a vacuum, somehow utterly unrelated to broad matters of political principle. This is clearly a normative commitment, usually founded in what has come to be called ethical legal positivism.57 While this commitment might or might not be normatively well founded, it is inaccurate as a description of many contemporary legal and political systems. Moreover, all we know about rule-interpretation and application suggests it is a value-laden process. This further bolsters the point that judges deciding cases are often unavoidably immersed in judgements about our political culture. In the face of ambiguity or contestation about the correct scope of a rule, recourse to some account of its point, purpose or value is almost immediate. This, of course, is well known to lawyers and jurists, but it is worth labouring the obvious to remind ourselves of the constraints upon this process.58 If the search for the point, purpose or value of some or other proposition of law is not exhausted by its doctrinal or legislative history, where else might lawyers look? One potential source of guidance is the general scheme of values or goals supposedly embodied in, or animating the general area of, doctrine within which the specific dispute has arisen. Another source is the broader scheme of values or goals inherent in the law as a whole: it might, for example, be said that since gender equality is a broad, animating principle of contemporary English law, then that value ought to be upheld in each and every area of the law. Now it does not seem odd, as a descriptive matter, to regard some such principles (or ‘goals’, or ‘purposes’) as not only embedded in our law, but also as part of our political culture. Judicial reference to such features of our political culture is thus indispensible, provided the application and interpretation of propositions of law involves recourse to accounts of the point, purpose or value of such propositions. 57 An exemplar of which is T Campbell, The Legal Theory of Ethical Positivism (Aldershot, Dartmouth Publishing, 1996). 58 It was LL Fuller who reminded modern lawyers and jurists what they already knew, but sometimes forgot, about the purposive nature of rule interpretation and application: ‘Positivism and Fidelity to the Law—A Reply to Professor Hart’ (1958) 71 Harvard Law Review 630, 661–9. The reminder is probably as old as legal thought itself: for an earlier instance see S Pufendorf, De Jure Naturae et Gentium Libri Octo, C and W Oldfather (trans), 1688 edn (Oxford, Clarendon Press, 1934), vol 2, book V, ch XII (On Interpretation).

70

William Lucy

C. Redundancy and Other Embarrassments ‘[I]t says nothing, it distinguishes nothing’. J Donne, Sermon XV, 8 March 1621.59

The most obvious alleged embarrassment for the argument offered here appears serious. It is that the proposed distinction between private and public law is practically useless, since it will not help to resolve cases with which we are familiar—the much discussed, usually appellate court cases, which seemingly turn on a distinction between public and private law.60 Yet, rather than being an embarrassment, this point must instead be embraced because it contains a significant truth. For one reason why this particular distinction between private and public law—or any other version of the distinction, for that matter—does little work for us in the usual range of reported cases is because those cases are hard. Hard cases are complex and require judgement. They cannot be solved by judges in anything like an automatic way by, for instance, an ‘algorithmic’ invocation of some or other version of the public/private distinction. This requires elucidation. Hard cases are those ‘in which reasonable lawyers disagree’ and ‘where no settled rule dictates a decision either way’.61 A more helpful but still abstract statement of the hallmarks of hard cases is offered by Neil MacCormick. Such cases usually present one or more of three possible doctrinal issues. First, they might raise a question as to which interpretation, from a range of two or more available interpretations of an agreed proposition of law, applies to the case at bar. Secondly, they arise from doubt as to which proposition of law, from a range of two or more incompatible propositions of law, applies to the case at bar. Finally, they might raise the question of whether or not any proposition of law applies to the case at bar.62 These cases thus require not just a statement of the correct applicable proposition of law, but also argument justifying that precise statement of the law. MacCormick shows that in the United Kingdom judges typically seek to justify their doctrinal choices in hard cases by three different kinds 59

J Donne, LXXX Sermons (London, M Flesher for R Royston and R Marriot, 1640) 148. See Cane, above n 3, at 249–61, for a typically sure-footed treatment of some of the principal cases. Those inclined to doubt that legal scholars expect the public/private distinction to do pragmatic doctrinal work should see: Kennedy, above n 4, at 1350–54 (the distinction is problematic in particular cases and thus does not do the work it should); the essays by, inter alios, HJ Friendly and KE Klare (1982) 130 University of Pennsylvania Law Review 1289 and 1358 respectively (Klare holds that the distinction ‘is devoid of significant, determinate analytical content’ (at 1360) when invoked in the cases, this being taken as an indictment); C Harlow, ‘ “Public” and “Private” Law: Definition without Distinction’ 43 (1980) Modern Law Review 241, 242–50 (purported demonstration that the distinction does not do any substantive work in the English cases, this being regarded as baleful). It is senseless to complain that the distinction does little or no work in the cases if one thinks, as this essay argues, that it can do little or no work in the cases. 61 R Dworkin, Taking Rights Seriously (London, Duckworth, 1978) xiv and 83 respectively. 62 This typology of the features of hard cases, while in some respects compatible with many aspects of Dworkin’s account (ibid, chs 4 and 5), is derived from N MacCormick, Legal Reasoning and Legal Theory, revised edn (Oxford, Clarendon Press, 1993) chs V–VIII. For elaboration of this position see N MacCormick, Rhetoric and the Rule of Law (Oxford, Clarendon Press, 2005) chs 3–10. 60

Duty of Care in APrivate Rights-Based TheoryLaw? of Negligence What’s About Private

71

of argument. Two of the three kinds of argument are intra-systemic, involving considerations internal to the legal system. Of these two, one kind—arguments from consistency—embody ‘a fundamental judicial commandment: Thou shalt not controvert established and binding rules of law’.63 Arguments from consistency are narrow in the sense that they focus solely on propositions of law in the immediate vicinity of the dispute in question, holding that no argument will be acceptable if (i) it is clearly incompatible with a closely contiguous proposition of law; and (ii) is unsupported by one of the remaining two kinds of argument. The focus of the second kind of argument—from coherence—is still upon considerations internal to the legal system, but is nevertheless wider than the focus of arguments from consistency. This is because considerations of coherence either test, reject or commend an argument in a hard case by reference to its resonance (or lack of it) with principles and values of the legal system as whole, rather than just the area of doctrine within which the case has arisen. Such arguments, says MacCormick, rest on the assumption ‘that the multitudinous rules of a developed legal system should “make sense” when taken together’.64 The focus of consequentialist arguments—the third type of argument invoked by judges to justify their decisions in hard cases—is extra-systemic, looking to the effects of a hard case ruling one way or another on society as a whole. Rather than being concerned with what makes sense within the legal system, they are concerned with ‘what makes sense in the world’.65 What, then, is the criterion of sense here? It consists of evaluating the consequences of a decision one way or another. It is a matter of ‘choosing between rival possible rulings in a case [and] involves choosing between what are to be conceived of as rival models for, rival patterns of, human conduct in society’.66 It seems to be the case that consequentialist arguments are often the strongest kind of argument in this trio.67 This account of how judges do and should decide hard cases can be contested but, for non-sceptics at least, the space for dispute is limited. Just about all nonsceptical jurists accept a picture of what hard cases look like which is very similar to that offered by MacCormick; the same jurists also agree that the arguments MacCormick finds judges actually using when deciding cases are indeed appropriate considerations for judges to use.68 Disagreement arises, however, as to which of these kinds of consideration should dominate: Ronald Dworkin, for instance, 63

Posner, above n 48, at 195. MacCormick, Legal Reasoning and Legal Theory, above n 62, at 152. 65 ibid, at 103. 66 ibid, at 104. 67 ibid, at 127–8. 68 A smattering of relatively recent studies (all, curiously, published in English within four years of one another) illustrates the point: R Alexy, R Adler and N MacCormick (trans), A Theory of Legal Argumentation: The Theory of Rational Discourse as Theory of Legal Justification (Oxford, Clarendon Press, 1989); SJ Burton, Judging in Good Faith (Cambridge, Cambridge University Press, 1992); K Günther and J Farrell (trans), The Sense of Appropriateness: Application Discourses in Morality and Law (Albany NY, State University of New York Press, 1993); J Levin, How Judges Reason: The Logic of Adjudication (New York, Peter Lang, 1992). 64

72

William Lucy

has little truck with MacCormick’s claim that consequentialist arguments do and should dominate other arguments. Consequentialist arguments are, of course, far too similar to arguments of policy in the Dworkinian schema and, as we all now know, arguments of principle trump arguments of policy.69 However interesting they might be, the details of this internecine jurisprudential dispute are not germane here. Our point, remember, is the claim that most reported cases in which the (or ‘a’) public/private distinction features are hard cases. Being such, they raise the three broad doctrinal issues, and are resolved by any combination of the three distinct kinds of argument just noted. What, then, is the moral of this story? That expecting any version of the distinction between public and private law to be of use in judicial decision-making is a mistake, if ‘of use’ is taken to mean ‘conclusively dispositive of any particular hard case’. Hard cases are not, and can never be, so easily resolved. Even the most cursory glance at the cases confirms this. Consider just one example, the House of Lords decision in Aston Cantlow v Wallbank.70 This case did indeed raise a question about the nature of public and private functions, since one of the issues the court had to address was whether or not a parochial church council of the Church of England was a ‘public body’ for the purposes of section 6 of the Human Rights Act 1998 (HRA). It might thus be regarded as an instance of a public/private distinction being invoked to conclusively determine the decision in the case. But the public/private issue was but one of a range of issues the court had to tackle, the remaining ones including (i) the current state of the law on chancel repairs; (ii) the retrospective applicability of the HRA; and (iii) the question of whether or not the parochial church council’s order to lay rectors was compatible with the lay rectors’ rights under Article 1 of the First Protocol of the ECHR. At least three of the four issues in this case—the law on chancel repairs being relatively unambiguous—are easily subsumable under one or more of the abstract hallmarks of hard cases. So, for example, the issue of the applicability of the HRA raises the question of which interpretation of an agreed proposition of law applies (the interpretative choice being between (i) the HRA does apply retrospectively and (ii) the HRA does not apply retrospectively). Determining the compatibility issue (ie was the parochial church council’s order compatible with Article 1?) was a matter of either choosing between different propositions of law or different interpretations of an agreed proposition of law. It is undeniable that two of the issues in Aston Cantlow were closely related: the question of incompatibility with Article 1 only becomes live if the parochial church council is a public body under section 6. But it is simply wrong to regard the public/private issue as the only or even the most significant matter in the case. Furthermore, it is no surprise to find the judges resolving the question of the public or private status of a parochial church council via arguments of coherence 69 R Dworkin, Law’s Empire (London, Fontana, 1986) 244. See also Taking Rights Seriously, above n 61, ch 4. 70 Aston Cantlow v Wallbank [2004] 1 AC 546 (HL).

Duty of Care in APrivate Rights-Based TheoryLaw? of Negligence What’s About Private

73

and consistency. The judgments of Lords Hope, Hobhouse and Rodger include not just arguments of consistency, in which they addressed the English cases on the legal standing of the Church of England, but also arguments of coherence, in which they considered the European Court of Human Rights decisions on, inter alia, the status of Greek monasteries and the Swedish church.71 These strands of legal doctrine were less significant in Lord Scott’s judgment, perhaps because his view, unlike that held by Lords Nicholls, Hope, Hobhouse and Rodger, was that the parochial parish council was exercising a public function.72 The nature of public bodies under legislation such as the Scotland Act 1998 was also discussed by Lord Nicholls. Interestingly, consequentialist arguments featured in the judgments only fleetingly. They took the form of discontent about the potential harshness of the law on chancel repairs and played a role in bolstering the view that parochial church councils should not be regarded as public bodies.73 There are two additional reasons, both entirely compatible with that just articulated, which suggest that the public/private distinction will do little pragmatic doctrinal work. One reminds us of a feature that the values of private and public project pursuit might have: they might be incommensurable with one another. If this is so, then the process of mediating the boundary between the two realms will be far from simple; decisions to rank one over the other are not compelled by reason, although they are based on reasons. This means the decisions arrived at by some as to which, if any, value should prevail in a particular case will not always and necessarily be regarded as correct and compelling, or even as helpful and influential, by others. The fact that decisions one way or another are not obviously rationally compelling could also explain another odd feature of the discussion about the parameters of public and private law in the common law world. This oddity is the acceptance of two apparently contradictory views. One view is that public law does and should dominate private law in the sense of ‘constitutionalising’ it: thus the rights and protections in bills and charters of rights and freedoms must be extended into the sphere of private law. The other view holds that private law controls the operation of constitutional and public administrative law, there being no special law for the state. Public or state bodies are subject to the same law of the land as all citizens, and that law includes private law. Hence we find constitutional scholars expressing the view that private law is the constraining context in which public law exists.74 The final reason which suggests it is plausible to think that any version of the public/private distinction is unlikely to do much pragmatic work can be highlighted with a question. It is this: how many other equally broad or even much narrower legal distinctions are ever dispositive of particular cases? The distinction

71 ibid, at [42], [48]–[53], [59]–[64] (Lord Hope); [86]–[88] (Lord Hobhouse); [153]–[157] and [159]–[164] (Lord Rodger). 72 ibid, at [130]–[132]. 73 ibid, at [15] and [17] (Lord Nicholls). 74 Allan, Law, Liberty, and Justice, above n 11, at 127–30.

74

William Lucy

between mens rea and actus reus in the criminal law is often just as contested and as unhelpful—as in need of mediation and elaboration—as that between public and private law. The distinction, which all common lawyers take for granted, between contract and tort is equally tricky and hardly ever dispositive; nor is the fact that it rarely resolves particular hard (or perhaps even easy) cases ever regarded as a reason for rejecting the distinction. Rather, this fact is taken to indicate something meaningful about the nature of the distinction itself. Lawyers’ expectations about the pragmatic power of these distinctions are apparently nowhere near as high as the expectations we have of the public/private distinction. There are, of course, two quite different inferences that can be drawn from this observation. One is that we are wrong to expect so little of these other distinctions. The other is that low expectations are justified, because distinctions of this kind are of little use when faced with the fine detail and broad range of issues presented by hard cases. Obviously, the arguments already advanced in this essay give reason to favour the latter, rather than the former, inference. Other potential embarrassments for the distinction between public and private law as so far elucidated are probably too numerous to consider here. Yet it should be noted that one of the most tempting putative embarrassments, over and beyond its relative lack of pragmatic utility, is not an embarrassment at all. Thus, those tempted to point out that the distinction offered here is of limited range are simply confirming the argument rather than undermining it. Furthermore, the limited range of the argument is doubled-sided, for while one of its obvious implications is that the public/private distinction sketched here must live alongside (and possibly in conflict with) other versions of the public/private distinction within the same legal system, there is also a more oblique implication. It is that different versions of the public/private distinction will be found in different political and legal cultures. It is thus conceivable that the version of public/ private distinction articulated here will not exist in some, presumably non-liberal, political cultures.

IV. Conclusion What, then, are the goals of private law? For Weinrib, ‘the sole purpose of private law is to be private law’.75 Many have found this statement rather gnomic and mystifying, but there is a sense in which it is undoubtedly true, at least from the perspective of the argument offered in this essay. The goal or purpose of private law as a whole is to facilitate private project pursuit—its aim is to allow and enable all citizens to achieve those of their goals that are consistent with a like power being enjoyed by all. Private law is thus a framework within which many

75

Weinrib, The Idea of Private Law, above n 17, at 8.

Duty of Care in APrivate Rights-Based TheoryLaw? of Negligence What’s About Private

75

different, and to some degree incompatible, individual goals can be pursued, not only through its obviously facilitative elements like contract law and trusts, but also through those elements, such as tort and property law, that protect some of the conditions of individual autonomy. Saying that private law’s goal is to allow individuals to pursue many of their goals is clearly not to say that it allows or facilitates the pursuit of any or all goals citizens might have. It does not facilitate the breaching of contracts, nor does it provide a means by which some citizens can physically attack others; it also safeguards against many forms of deception and interference with holdings. That private law is not neutral on all aims citizens might have does not show that it cannot be a facilitative body of legal rules. The rules of association football contain many prohibitions but they also undeniably facilitate a particular form of group conduct. Over and beyond this, the attribution of what (in the Weinribian vernacular) might be called an ‘external’ goal (or goals) to private law appears dubious. Of course, there can be no complaint about jurists and practitioners insisting that private law embody generic rule of law virtues, such as consistency, intelligibility, predictability and non-retrospectivity. Yet these virtues, as their collective name suggests, are not unique to private law. They are most likely internal to the very idea of law itself, if it is to approximate anything like the ideal of subjecting human conduct to the governance of rules. Nor can there be any objection to holding particular branches of private law to their implicit or internal goals: that jurists should evaluate doctrinal developments in the law of contract, for example, with a eye to their consistency with the scheme of values embodied in that area of law seems both obvious and a valuable feature of legal scholarship. The scheme of values embodied in contract law is assuredly complex for, in addition to providing a means of creating legally-enforceable transactions, the law also sets standards as to how transacting parties must behave, these standards sometimes obtaining only in the absence of explicit directives from the parties and sometimes despite or contrary to such directives.76 It seems both bizarre and unhelpful to deny that these values can be understood as the goals of contract law, yet this should cause Weinrib no embarrassment. For these goals are significantly ‘internal’ to the law. They are either implicitly or explicitly embodied in the publicly stated rules of the law of contract, serving to make normative sense of, and to systematise, those rules. And so it is, presumably, with other areas of private law. Is there any additional space for meaningful talk about the goals of private law? Certainly, but this space is primarily the domain of those social scientists and other scholars interested in the side-effects or consequences of the whole (or some area) of private law. It is not a domain in which many private lawyers and jurists are equipped to sport themselves without succumbing to the vice of intellectual voyeurism.

76 An admirable treatment of contract law and its animating values is SA Smith, Contract Theory (Oxford, Clarendon Press, 2004).

4 The Role of Duty of Care in a Rights-Based Theory of Negligence Law STEPHEN PERRY

I. Introduction In 1996, when I took up a new position at the University of Pennsylvania Law School, I thought I knew something about the law of torts. I had been lucky enough to study torts at the University of Toronto with Ernie Weinrib, and I had been teaching the subject for a little over 10 years at McGill University. In my first year of teaching at Penn I chose a standard American casebook, which in its general outline was very similar to the casebook, edited by Ernie, that I had been using in Canada. At least at first glance the casebook seemed to be organised so as to reflect what I took to be the basic doctrinal principle that the tort of negligence consists of a number of distinct elements, usually rendered as duty of care, breach or standard of care, legally cognisable injury, cause in fact, and proximate cause. All of this was very familiar to me, except that I was used to the label ‘remoteness’ for what in the United States is called ‘proximate cause’. When I came, however, to study the chapter in the American casebook on the duty of care, I was, to say the least, puzzled. It was clearly the generally accepted view in Anglo-Canadian law that the duty of care is an element of fundamental importance in the tort of negligence. The most celebrated torts case in modern times was Donoghue v Stevenson, in which Lord Atkin had enunciated the famous ‘neighbour’ principle as being, in his words, ‘the general conception of relations giving rise to a duty of care’.1 In Ernie Weinrib’s class we had also closely studied the landmark American cases in which Benjamin Cardozo had laid out his understanding of the duty element. Lord Atkin himself had cited Cardozo CJ’s opinion in MacPherson v Buick Motor Co,2 and Palsgraf v Long Island Railroad3 1

Donoghue v Stevenson [1932] AC 562 (HL) 580. MacPherson v Buick Motor Co, 217 NY 382 (New York CA, 1916), cited in Donoghue v Stevenson [1932] AC 562 (HL) 598–9. 3 Palsgraf v Long Island Railroad, 248 NY 339 (New York CA, 1928). 2

80

Stephen Perry

was perhaps the most famous case in the common law world. But in the chapter on duty in my American casebook there was, in the introductory section, only the briefest of passing mentions of MacPherson, and no mention at all of Palsgraf. The very first case dealt with the common law’s failure to recognise a duty of easy rescue, and the remainder of the chapter dealt with those exceptional situations in which the law imposes an affirmative obligation of some kind. To be sure the casebook did contain both MacPherson and Palsgraf, but at much later points than the chapter on duty. MacPherson was to be found in a chapter entitled ‘Products Liability’, and Palsgraf was contained, to my mind completely bizarrely, in the chapter on proximate cause. Hadn’t Cardozo CJ explicitly stated that ‘[t]he law of causation, remote or proximate, is … foreign to the case before us’,4 and wasn’t it only in Andrews J’s dissenting opinion that there was to be found any discussion of proximate cause? And hadn’t even Andrews J devoted the initial part of his opinion to laying out a different understanding of the duty of care from Cardozo’s? At the time that I first began teaching in the United States I thought that the failure of this particular casebook to discuss the notion of a general duty of care was an idiosyncrasy, and accordingly I adopted a syllabus that jumped around somewhat so as to begin the section on duty with MacPherson and Palsgraf. The result of this and various other changes that I introduced was that, without realising it, I taught my first class of American law students a version of tort law that was, in spirit, closer to Anglo-Canadian law than to its American counterpart. Although my students were very good-natured about this even when they and I subsequently discovered that American tort law, or at least the dominant academic understanding of American tort law, was a very different creature from what I had represented it to be, I’ve always thought that they were entitled to a partial rebate on their tuition fees for that year. Even so, however, I’ve continued in my torts course to begin the section on duty of care with Palsgraf. But nowadays when I teach the case I am careful to accompany it with various warnings and disclaimers, which, as we all know, can do wonders in heading off liability and all manner of other troubles. At least with the caveats, I do not think my somewhat unorthodox approach to teaching duty of care has been seriously misleading, since even within American tort theory there is a vigorous minority view of both negligence law in general, and of the duty of care in particular, which is very different from the dominant view and much more in line with what I still take to be the general Commonwealth understanding. It is, moreover, by no means unreasonable to think that some version of this minority view provides the best, or at least a very decent, overall interpretation of American case law, since the courts themselves have often proved recalcitrant in adopting the prevailing academic understanding of what they are doing. It is commonplace to say that theories of tort law in the common law world, both in the United States and in the Commonwealth, divide into two main camps.

4

ibid, at 346.

Duty of Care in a Rights-Based Theory of Negligence

81

Theories in the first camp, which are identified by a wide range of labels including instrumentalist, functionalist, pragmatist, welfarist, utilitarian, consequentialist, and economic, represent the dominant strand of theorising about torts in the United States. Such theories hold that the point or purpose of tort law as a whole is to achieve certain kinds of moral goals, such as the maximisation of welfare or the promotion of economic efficiency, which can broadly be characterised as collective or aggregative in character. Following widespread usage I will refer to this entire category of theories by the label ‘instrumentalist’, although for reasons that I will explain, I do not think the term is an entirely apt one. Theories in the second camp again go by a wide range of labels, which variously make reference to such notions as rights, duty, responsibility, fairness, reciprocity, corrective justice, noninstrumentalism, and deontology.5 My own scholarly writing about torts falls into this second camp, and, like Peter Benson, Jules Coleman, Arthur Ripstein, Ernest Weinrib, and several others, I have tended to identify my work by using the term ‘corrective justice’. Theorists who have adopted this term have come in for a certain amount of criticism, most but not all of which is to my mind misplaced, from others in the second camp for, among other things, supposedly concentrating on remedy rather than substance.6 The very use of the term ‘corrective’ is said to emphasise after-the-fact rectification or compensation rather than the primary rights and duties which, the objection continues, lie at the core of tort law. Although, as I say, I think much of this criticism has been misplaced, I agree with the basic point that the core of tort law concerns certain fundamental moral duties and their correlative rights. There is, moreover, so much disagreement about the nature of corrective justice that I prefer now not to use the term at all.7 In the remainder of this chapter I will accordingly refer to this entire category of theories as ‘rights-based’, and I will say something further later as to why that term is an apt one. This chapter has its origins in a paper presented at a conference on ‘The Goals of Private Law’. There are some rights theorists who think that it is a mistake even to ask the question, what are the goals (or functions or purposes) of private 5 Prominent theorists who fall into this camp include Peter Benson, Jules Coleman, George Fletcher, John Gardner, John Goldberg, Tony Honoré, Gregory Keating, Arthur Ripstein, Robert Stevens, Ernest Weinrib, Richard Wright, and Benjamin Zipursky. 6 See, eg, J Goldberg and B Zipursky, ‘The Moral of MacPherson’ (1998) 146 University of Pennsylvania Law Review 1733, 1739; J Goldberg, ‘Twentieth-Century Tort Theory’ (2003) 91 Georgetown Law Review 513, 576; R Stevens, Torts and Rights (Oxford, Oxford University Press, 2007) 327–8. 7 In an earlier article I argued that corrective justice should not be understood, as some theorists have claimed, as a moral principle that is concerned with maintaining a just distribution of entitlements, but rather that it should be understood as a moral principle that is concerned with repairing the harm that one person has caused another. See S Perry, ‘On the Relationship between Corrective and Distributive Justice’ in J Horder (ed), Oxford Essays in Jurisprudence: Fourth Series (Oxford, Oxford University Press, 2000). I do not think that the view I expressed there was mistaken, but I have come to think that it was misleadingly incomplete. As I say in the text, I now prefer not to use the term ‘corrective justice’ at all, in part because the term is often associated with very specific theoretical views of the moral principles underlying tort law, such as Weinrib’s Kantian account. I would now prefer to say that the moral principles underlying tort law are concerned, first, with primary rights not to be harmed and, second, with secondary or remedial rights that arise when a primary right is violated, where the most important such remedial right is a right to repair.

82

Stephen Perry

law or of tort law, because to do so is to commit oneself to an instrumentalist account of these institutions. Ernest Weinrib, for example, has famously stated that ‘the only purpose [tort law] can have as tort law [is] simply to be tort law’.8 Elsewhere he has elaborated on this idea in the following way: ‘An instrumentalist account understands private law as a means to something else, whereas a noninstrumentalist account construes law as being internally intelligible and thus requiring no reference to purposes external to itself ’.9 Weinrib maintains that one does not avoid being an instrumentalist in his sense simply because one is committed to a non-instrumentalist account of morality, because even an account of private law that draws on a morally non-instrumental notion such as fairness will be instrumentalist in his sense if it attributes to private law a purpose that is ‘external to itself ’ and ‘alien to its nature’.10 Weinrib here draws an implicit distinction between non-instrumentalist accounts of morality and non-instrumentalist accounts of social institutions such as tort law or private law. On his view, a non-instrumentalist account of private law in the second, institutional sense must not only adopt a non-instrumentalist account of morality, but must adopt such an account that construes tort law as ‘internally intelligible’ and as possessing a certain distinctive kind of unity and coherence.11 I have criticised this aspect of Weinrib’s theory of private law elsewhere,12 and will not repeat those criticisms here. For present purposes I simply wish to leave open the possibility that the moral theory of rights that informs the correct or most satisfactory rights-based account of tort law is of a character to make it appropriate, or at least not inappropriate, to speak of tort law as a means for achieving the moral end of enforcing certain private rights and their correlative duties.13 Although I will not defend the claim here, I believe that any plausible moral account of coercive social institutions such as tort law will involve this kind of means-end reasoning, and will accordingly be characterisable as instrumentalist in something like Weinrib’s second sense. It is for that reason that it may be somewhat inapt to restrict the term ‘instrumentalist’ to the first category of theories that I identified earlier, since those theories regard tort law as a means for pursuing only collective or aggregative moral goals, whereas I am suggesting that many rights-based theories can also properly be called instrumentalist. However, the more limited use of the term to identify a category of theories which regard tort law as an instrument for pursuing collective or aggregative goals like 8

EJ Weinrib, ‘Understanding Tort Law’ (1989) 23 Valparaiso University Law Review 485, 491. EJ Weinrib, The Idea of Private Law (Cambridge, Massachusetts, Harvard University Press, 1995) 49–50. 10 ibid, at 50. 11 ibid, at 22–55, 75–6. Weinrib draws together these various strands of his theory of private law under the heading of ‘formalism’. 12 S Perry, ‘Professor Weinrib’s Formalism: The Not-So-Empty Sepulchre’ (1993) 16 Harvard Journal of Law and Public Policy 597. 13 Or, perhaps, the end of giving persons a form of recourse that enables them to enforce their rights if they so choose. See B Zipursky, ‘Rights, Wrongs and Recourse in the Law of Torts’ (1998) 51 Vanderbilt Law Review 1. 9

Duty of Care in a Rights-Based Theory of Negligence

83

efficiency is quite widespread, and so long as one bears in mind that the term is being used in this particular sense, there should be no confusion. It will be obvious from what I have said so far that rights-based theories of tort law draw on accounts of morality that make room for individual rights, whereas instrumentalist theories of tort law draw on accounts of morality that make room for the collective pursuit of collective goals such as efficiency or the maximisation of welfare. Accounts of morality of these two kinds need not be mutually exclusive, and the distinction between them does not necessarily line up neatly with, say, the distinction between purely deontological accounts of morality and purely consequentialist accounts. Thus I do not mean to rule out the possibility that the best moral theory of rights has a consequentialist dimension. Discussion of these aspects of moral theory is well beyond the scope of the present chapter. The point I wish to focus on for present purposes is that rights-based and instrumentalist theories of tort, as these have actually been put forward and defended in the literature, tend to take very different approaches to both the explanation and the justification of the institution of tort law. Rights-based theories can be said, as a general matter, to claim to take seriously what Coleman has called the ‘structure’ of tort law14 and Weinrib has called its ‘self-understanding’.15 These notions of structure and self-understanding refer, at least as an initial matter, to the doctrinal articulation of the tort of negligence as consisting of the distinct elements of duty of care, standard of care, injury, cause in fact, and proximate cause, and further structural elements that come into play at the stage of remedy. Instrumentalist theories, by contrast, tend to view tort doctrine in a much more plastic and malleable manner, and hence as being subject to perhaps quite radical reform in the name of better serving relevant collective goals. Thus, for example, Guido Calabresi has offered a notoriously revisionist, probabilistic account of the doctrinal requirement of cause in fact.16 Relatedly, defenders of rights-based theories of tort law have frequently criticised efficiency-based theories as simply not possessing either the conceptual or the normative resources to offer an adequate explanation or justification of the requirement of cause in fact as it actually figures in the case law. In the remainder of this chapter I wish to explore the appropriate role, within a general rights-based approach to tort law, of the doctrinal element of a duty of care. As I mentioned earlier, when I first began teaching in the United States I discovered that the idea of a general duty of care has almost disappeared from the prevailing American academic understanding of negligence law. This historical development has been comprehensively and perceptively discussed by John Goldberg and Benjamin Zipursky.17 As they persuasively argue, the reasons for the

14

JL Coleman, Risks and Wrongs (New York, Cambridge University Press, 1992) 374–5. Weinrib, The Idea of Private Law, above n 9, at 50. G Calabresi, ‘Concerning Cause and the Law of Torts’ (1975) 43 University of Chicago Law Review 69. 17 Goldberg and Zipursky, above n 6. 15

16

84

Stephen Perry

development have to do with the sway of legal realism within the American legal academy and an associated and growing acceptance, over the course of the twentieth century, of an instrumentalist understanding of tort law that has been strongly influenced by the work of William Prosser but that can ultimately be traced back to the writings on the common law of Oliver Wendell Holmes. This has given rise to a predominant academic understanding of the nature of tort law in the United States which is strikingly different from the understanding that I believe still prevails in the Commonwealth. As a scholar with a foot in both camps, I try in what follows to elucidate the element of duty of care in negligence law by examining its underpinnings in moral theory, with a view to reinforcing the traditional understanding of the duty element and, in the American context, to contributing to its rehabilitation. As will emerge, I do not think that the concept of a duty of care is as clear-cut or as uncomplicated as rights theorists sometimes seem to assume. My hope is that clarification of the moral character of the duty element will lend at least indirect support for a general understanding of negligence law, and ultimately of tort law as a whole, which is predominantly rights-based.18

II. Duty of Care and Rights On any understanding of tort law, instrumentalist or rights-based, the tort of negligence subjects persons to a mandatory norm of conduct. In doctrinal terms this is of course the standard of reasonable care, which requires that one act so as not to subject others to unreasonable or excessive risks. A mandatory legal norm is, by its very nature, obligatory, which is simply to say that one has a duty to comply with it. So negligence law cannot do without the concept of duty in at least this limited sense, and this point is of course recognised by instrumentalist theories. Thus the proposed final draft of the American Law Institute’s Restatement of the Law of Torts, Third: Liability for Physical Harm,19 which incorporates and indeed effectively codifies a predominantly instrumentalist understanding of torts, does not ignore the concept of duty entirely. Section 7(a) says that ‘[a]n actor ordinarily has a duty to exercise reasonable care when the actor’s conduct creates

18 I say ‘predominantly’ because, as I have argued elsewhere, I think that a pluralist understanding of tort law, which sees it as having goals of both a rights-based and an instrumentalist character, is ultimately the most satisfactory view. See Perry, ‘Professor Weinrib’s Formalism’, above n 12, at 618–19. On the view I advocate there, rights-based considerations would have priority over policy-based considerations such as deterrence and loss-spreading. On a strong version of such a view, ‘priority’ would mean ‘lexical priority’. 19 Restatement of the Law of Torts, Third: Liability for Physical Harm (Proposed Final Draft No 1, 2005). According to the American Law Institute’s website at www.ali.org/index.cfm?fuseaction= publications.ppage&node_id=53, accessed on 18 November 2008, the Proposed Final Draft No 1 has met with the Institute’s final approval, with the exception of some comments to § 27 and § 28. It has not, however, been published in final form because the project has been expanded to include chapters on emotional harm and landowner liability.

Duty of Care in a Rights-Based Theory of Negligence

85

the risk of physical harm’. Although § 7 is entitled ‘Duty’, and § 7(a) states conditions under which a person has a duty, this is clearly just a formulation of the familiar standard of reasonable care. It states a mandatory legal norm of conduct, and of course there is always a duty to comply with a mandatory norm. There is absolutely nothing in the Restatement that corresponds with the element of duty of care as it figures in Commonwealth law or, indeed, in Justice Cardozo’s classic opinions in MacPherson and Palsgraf. Legal academics who advocate a rights-based understanding of tort law have mounted a vigorous campaign within the American Law Institute to have the concept of duty of care in something like its traditional understanding introduced into the draft Restatement,20 but at this stage it seems clear that the battle has been lost. But what is it, exactly, that the critics were objecting to and wanted to have changed? While it is true, as we shall see, that the Restatement’s discussion of duty downplays the notion of reasonable foreseeability which figures so prominently in the classic cases, this has not, for the most part, been the concern of the Restatement’s critics. Their focus lies elsewhere. Almost without exception, apart from some early work by Jules Coleman21 and George Fletcher,22 rights theorists have characterised tort law as being ‘relational’ in character, a feature which has also been variously referred to as ‘interactional’, transactional’, ‘bipolar’, ‘bilateral’ and ‘correlative’.23 What rights theorists have in mind in employing these terms is quite straightforward, and involves two separate but related points. The first is that tort law in general, and negligence law in particular, regulates certain kinds of interactions or transactions between persons, who in the paradigmatic case are two in number. The second point, which is ultimately the more important of the two, concerns the normative character of the duty of care. While negligence law 20 See, eg, JCP Goldberg and B Zipursky, ‘The Restatement (Third) and the Place of Duty in Negligence Law’ (2001) 54 Vanderbilt Law Review 639; and EJ Weinrib, ‘The Passing of Palsgraf?’ (2001) 54 Vanderbilt Law Review 803. 21 See, eg, JL Coleman, ‘Corrective Justice and Wrongful Gain’ (1982) 11 Journal of Legal Studies 421, reprinted in JL Coleman, Markets, Morals, and the Law (New York, Cambridge University Press, 1988) 184–201; JL Coleman, ‘Tort Law and the Demands of Corrective Justice’ (1992) 67 Indiana Law Review 349. 22 G Fletcher, ‘Fairness and Utility in Tort Theory’ (1972) 85 Harvard Law Review 537, 542 and 553–4. It is worth noting that neither Coleman nor Fletcher appeared to deny the correlativity of what I refer to later in the text as primary rights and duties; their concern was, rather, with the supposed independence of secondary, or remedial, rights and duties, and more particularly with the right of a plaintiff to receive compensation for harm caused and the obligation of a defendant to pay such compensation. 23 See, for example, Weinrib, ‘Understanding Tort Law,’ above n 8, at 494 (‘bipolar’); JL Coleman, Risks and Wrongs (New York, Cambridge University Press, 1992) 311–8 (‘relational’); P Benson, ‘The Basis of Corrective Justice and its Relationship to Distributive Justice’ (1992) 77 Iowa Law Review 515, 533 (‘relational,’ ‘intelligible as interaction’) and 569 (‘corresponding’ duty and right); S Perry, ‘The Moral Foundations of Tort Law’ (1992) 77 Iowa Law Review 449, 507 (‘correlative rights and duties’); Weinrib, The Idea of Private Law, above n 9, at 114–44 (‘correlativity’); Goldberg and Zipursky, above n 6, at 1744 (‘relational’); JL Coleman, The Practice of Principle: In Defence of a Pragmatist Approach to Legal Theory (New York, Oxford University Press, 2001) 13–24 (‘bilateral’); A Ripstein, ‘Tort Law in a Liberal State’ [2007] 1(2) Journal of Tort Law, Art 3 www.bepress.com/jtl/vol1/iss2/art3, accessed on 18 November 2008, at 2 (‘bilateral’) and at 14–15 (‘transactional’).

86

Stephen Perry

on any view involves a duty to comply with a mandatory norm, the traditional understanding of the duty of care, and the understanding advocated by rights theorists, is that of a duty which is owed to someone.24 Duties to comply with mandatory norms do not necessarily, or even usually, have this feature. Laws making it an offence to litter or jay-walk, for example, create mandatory legal norms which impose duties not to litter or jay-walk. But such duties are not ordinarily thought to be owed to anyone, or, at the very least, they are not thought to be owed to other citizens. To say, however, that a duty is owed to another person is to say that the other person has a correlative right. Duties can exist without rights, but rights, at least in this context, cannot exist without duties. That is the main reason I prefer to label the second category of theories of tort law that I identified in the Introduction by using the term ‘rights-based’ rather than one of the other terms in current use, such as ‘corrective justice’. It is a controversial matter in the philosophy of interpersonal morality whether rights or duties are the more fundamental concept, and I do not mean my use of the term ‘rights-based’ to suggest otherwise. The term simply helps to signal what is distinctive about and common to the entire family of theories in question. No theoretical account of tort law can do without the concept of duty, but it is further true of the family of theories I am discussing that they cannot do without the concept of a correlative right. It is helpful to further distinguish between ‘primary’ and ‘secondary’ rights and their correlative duties.25 A primary right is one that prohibits others from invading, in some specified way, a specified protected interest or set of such interests. A secondary right is a right to a remedy that arises upon the violation of a primary right. The most important secondary right in negligence law involves the remedy of compensatory damages. My concern in this chapter is, however, with primary rights. The most important primary right in negligence law is usually taken to be one that protects fundamental interests in personal security and tangible property, and the correlative duty is taken to be, or at least to be partially specified by, the doctrinal element of duty of care. The precise relationship between the duty of care and the protected interests of personal security and tangible property is not entirely straightforward, however, since it is possible to breach the duty of care without invading either of these interests. How this can be so is one of the puzzles about the duty of care that this chapter is intended to address. Before turning to a discussion of the moral character of the duty of care, it will be convenient to revisit for a moment the issue of instrumentalist versus

24 See S Perry, ‘Comment on Coleman: Corrective Justice’ (1992) 67 Indiana Law Journal 381. In that article I criticised Coleman’s early version of a corrective justice theory on the grounds that it did not accept the correlativity of the plaintiff ’s right and the defendant’s duty at the level of remedy (‘secondary’ rights and duties). Coleman subsequently accepted this criticism, and adopted a fully relational understanding of both corrective justice and tort law. See Coleman, Risks and Wrongs, above n 23, at 311–18. 25 See, eg, Coleman, ‘Tort Law and the Demands of Corrective Justice’, above n 21, at 367–8; Perry, ‘Comment on Coleman,’ above n 24, at 388–9; Perry, ‘The Moral Foundations of Tort Law,’ above n 23, at 507; Coleman, Risks and Wrongs, above n 23, at 316–17; and Stevens, above n 6, at 327–8.

Duty of Care in a Rights-Based Theory of Negligence

87

non-instrumentalist theories of institutions. At one point in The Idea of Private Law, Weinrib seems to suggest that the mere introduction of the notion of correlative moral rights is enough to make a theory of tort law noninstrumentalist in his second, institutional sense. Thus he writes that ‘[u]nless we take account of [the feature of correlative rights] we will be treating private law—paradoxically—merely as a means to a non-instrumental morality’.26 But the acknowledgment of correlativity is surely too weak a premise to establish Weinrib’s conclusion. The difficulty is that most moral theories of rights, and not just Weinrib’s favoured Kantian theory, recognise the correlativity of rights and duties. These include, for example, Lockean theories and variants on the general interest theory of rights. Many tort theorists who draw on such nonKantian theories of rights find nothing untoward in speaking of tort law, in the manner I discussed earlier, as serving the goal or function of enforcing, or giving private citizens the power to have enforced, certain private rights that persons hold against one another.27 It bears re-emphasising that the widespread use of the label ‘instrumentalist’ to refer to a category of theory that takes the function of tort law to be the pursuit of collective or aggregative moral goals should not blind us to the possibility—and, in my view, the appropriateness—of viewing tort law as a means to a very different kind of moral end.28 Instrumentalist theories, in the sense of instrumentalism which regards the function of tort law as the pursuit of collective or aggregative moral goals, take a variety of forms, but one of the most influential variants sees tort law in general, and negligence law in particular, as a means for promoting economic efficiency. The general idea is that actors, both potential defendants and potential plaintiffs, should be given incentives to conform their behaviour to the standard of reasonable care, which is then typically defined in cost-benefit terms along the lines of Judge Learned Hand’s famous formula.29 The core of the efficiency-based understanding of tort law is thus the standard of reasonable care, understood as a mandatory norm which all persons have a duty to comply with. Even though the content of this norm is concerned with the imposition of inefficient levels of risk on other persons, the associated duty is simply a duty to comply with the norm, in exactly the same way that one has a duty to comply with legal norms

26

Weinrib, The Idea of Private Law, above n 9, at 50. See, eg, Perry, ‘Professor Weinrib’s Formalism’, above n 12; Coleman, Risks and Wrongs, above n 23, at 209–11. 28 Arthur Ripstein, who follows Weinrib in rejecting talk of the function of tort law and reserving the label ‘non-instrumentalism’ for a Kantian theory of torts, justifies these claims on the ground that tort law ‘is an ineliminable component of the system of private ordering that a liberal state must have’: Ripstein, ‘Tort Law in a Liberal State’, above n 23, at 7. As I understand his view, the claim is that it is in the very nature of the relevant rights that they must be coercively enforced by the (liberal) state, and that they do not have normative force except in this context. This view does indeed offer a more stringent understanding of what non-instrumentalism in the institutional sense requires. 29 US v Carrol Towing Co, 159 F 2d 169 (US CA, 2d Cir, 1947) 173. According to Hand J, an actor is negligent if, for that actor, B < PL, where B is the burden or cost of precautions, L is the magnitude of the loss were it to occur, and P is the probability of its occurrence on any given occasion. 27

88

Stephen Perry

which prohibit jaywalking or littering. The duty is not one that is owed to anyone else. On the pure efficiency-based approach, then, there is neither need nor room for the full-blooded notion of a duty of care, understood as a duty which is owed to persons whose interests an actor’s activities have put at risk, and who can therefore be said to have a correlative primary right. What matters, for the efficiency-based view, is simply that persons conform their conduct to the standard of reasonable care. It is thus often said that, on this view, tort law is best viewed simply as a variant of a general regulatory approach, where regulation happens to take a private form.30 In light of the influence and pervasiveness of the economic interpretation of tort law in the United States, it is not in the least surprising that the element of duty of care in negligence law has for many years been de-emphasised by instrumentalist theorists and that now, with the advent of the new Restatement, seems poised to disappear almost entirely from the most influential formulation of American tort doctrine. To the extent that the formal idea of a duty owed to another has survived in some vestigial form in American instrumentalist accounts of negligence law, it is often understood, either explicitly or implicitly, in terms of Holmes’s notion of a duty ‘of all the world to all the world’.31 For instrumentalists who are more pragmatic or more pluralist than advocates of, say, the pure efficiency-based approach, the existence of a duty is sometimes understood as always depending directly— either on a case-by-case basis or at least on a category-of-case by category-of-case basis—on the balancing of a miscellany of policy considerations, where the notion of ‘policy’ is generally understood to refer to a wide range of various collective

30 As Richard Posner put the point many years ago, on the economic view the plaintiff is given a cause of action after the fact—ie after having suffered harm as a result of the defendant’s breach—not because he or she has a moral right of any kind, but simply to harness the energies of self-interest so as to ‘enlist the participation’, as Posner put it, of the plaintiff and the plaintiff ’s lawyer in the operation of the system: R Posner, ‘A Theory of Negligence’ (1972) 1 Journal of Legal Studies 29, 33. 31 ‘The Theory of Torts’ (1873) 7 American Law Review 652, 662. This article has no attributed author but, according to the Draft Third Restatement, Holmes is widely credited as having written it. The Restatement cites this article in support of what it calls the ‘the general rule’ of § 7(a) that ‘[a]n actor ordinarily has a duty to exercise reasonable care when the actor’s conduct creates a risk of physical harm’. See Reporter’s Note to comment a, § 7. Clearly, Holmes’ notion of a duty ‘of all the world to all the world’ is reflected in Andrews J’s proposition, in his dissent in Palsgraf, that ‘[e]very one owes to the world at large the duty of refraining from those acts that may unreasonably threaten the safety of others’: Palsgraf v Long Island Railroad, 248 NY 339 (New York CA, 1928) 350. As Goldberg and Zipursky point out, Andrews J’s understanding of duty is not best interpreted as a universal duty that is correlative of a universal right: Goldberg and Zipursky, above n 6, at 1821. What Andrews J had in mind, rather, is best understood as simply a completely general duty to conform one’s conduct to the norm of reasonable care. Notice that the formulation of the ‘general duty’ in § 7(a) of the Restatement involves no reference even to the formal idea of a duty that is owed to another person. § 7(b) states that ‘[i]n exceptional cases, when an articulated countervailing principle or policy warrants denying or limiting liability in a particular class of cases, a court may decide that the defendant has no duty or that the ordinary duty of reasonable care requires modification’. It should be noted that the ‘exceptional cases’ referred to in § 7(b) are ones in which an articulated principle or policy warrants denying or limiting liability, as opposed to a relationship of correlative right and duty. This point is underscored in § 7, comment j: ‘A no-duty ruling represents a determination … that no liability should be imposed on actors in a category of cases’.

Duty of Care in a Rights-Based Theory of Negligence

89

and usually aggregative moral goals.32 On such a ‘general balancing’ approach, analysis does not begin with anything like the Restatement, Third’s ‘general rule’ of an ordinary duty of reasonable care, which is then subject to being restricted by various considerations of policy.33 On the ‘duty to all the world’ approach which is adopted, at least in the Reporters’ Note, by the Restatement, Third, there is obviously no room for the restrictive requirement of reasonable foreseeability that figures so prominently in the classic cases of Donoghue v Stevenson and Palsgraf. A similar point holds true of at least some versions of the ‘general balancing’ approach.34

32 The economic approach in its purest form envisages negligence law as a means for achieving the goal of economic efficiency, which is often taken to be a proxy for the goal of maximising utility, simply through the creation of appropriate incentives for conduct. Pluralist instrumentalist theories allow not only for such deterrence-oriented, forward-looking goals but also for the backward-looking goal of loss-spreading, or some related but more amorphous goal of ‘compensation’. Both loss-spreading and compensation are clearly best understood by reference to some more fundamental consequentialist goal, such as the maximisation of utility. The same can be said of the policy concern, often invoked in American cases, of avoiding overly extensive liability in a given context, for example in cases of pure economic loss. Presumably the same point holds of more specific policy considerations, such as the concern often expressed in the context of social host liability that imposing liability on social hosts who serve alcohol to guests who then injure others will, in the words of the Draft Restatement, have a problematic impact ‘on a substantial slice of social relations’: Restatement, Third § 7, comment a. In the Reporters’ Note to comment a, § 7, the Restatement, Third draws attention to a ‘subtle’ distinction between principle and policy, where ‘[t]he latter constitutes an instrumental justification for a legal rule, while the former is a moral reason in support of a moral rule’. Without noting that some writers, most notably Ronald Dworkin, have characterised the notion of ‘policy’ by reference to ‘collective goals of the community as a whole’ and the notion of ‘principle’ by reference to arguments which respect or secure an individual or group right—see R Dworkin, Taking Rights Seriously, revised edn (London, Duckworth, 1978) 82—the Reporters’ Note states that ‘[t]he use of principle or policy in this Section communicates that the sum total of rationales is available in determining whether it is appropriate to recognise an exception to the general rule of a duty of ordinary care’. It would thus seem that, according to the Draft Restatement, Third, rights-based considerations might, at best, figure among the considerations to be balanced in justifying a no-duty exception to ‘the general rule of a duty of ordinary care’. They most certainly are not accorded priority of any kind, let alone priority in the sense envisaged by rights-based theories, according to which a duty of care is necessarily correlative of a right on the part of some person or category of persons. 33 See nn 31–32 above. 34 In 532 Madison Ave Gourmet Foods Inc v Finlandia Center Inc, 96 NY 2d 280 (New York CA, 2001) 288–9, the New York Court of Appeals reiterates its support for a general balancing approach, and cites a number of relevant policy considerations, such as ‘the proliferation of claims’ and ‘the likelihood of unlimited or insurer-like liability’, but does not include foreseeability in the list. The Court then goes on to observe that ‘[a]s we have many times stated, foreseeability of harm does not define duty’. The California Supreme Court also accepts a version of the general balancing approach, but not only does it include foreseeability of harm to the plaintiff as one of the considerations to be balanced, it states that foreseeability is the most important such consideration: see, eg, Tarasoff v The Regents of the University of California, 17 Cal 3d 425 (Supreme Court of California, 1976) 434. The Draft Restatement, Third explicitly rejects the California approach: ‘So long as the actor’s conduct created a risk of harm—the predicate for a duty under this Section—foreseeability has no role under this Section and Restatement in a determination that a duty exists vel non’ (§ 7, Reporters’ Note, comment j). Comment j itself states that ‘foreseeable risk is an element in the determination of negligence [by the fact finder]’, ie it is an element in the determination of whether or not the defendant exercised reasonable care. Comment j goes on to add that no-duty rulings, which are determinations by the court that ‘no liability should be imposed on actors in a category of cases’, should be explained and justified by reference to reasons of policy and principle, where such reasons ‘should be articulated directly without obscuring references to foreseeability’.

90

Stephen Perry

As has often been observed, on a pure efficiency-based understanding of negligence law there is not only neither need nor room for a robust, rights-oriented conception of the duty of care, there is also no need in principle for either of the elements of legally cognisable injury or cause in fact, since the goal of efficiency is served by discouraging inefficient risk-taking, whether harm occurs or not. The same is true of any instrumentalist understanding of tort law which focuses primarily on deterrence or the creation of incentives. Of course, as a practical matter, the elements of injury and cause in fact are not about to disappear from the law of negligence any time soon, and it is thus the element of duty of care that has turned out to be the main doctrinal battleground in the United States between instrumentalists and rights theorists. To the extent that instrumentalist views maintain some formal or vestigial idea of a duty owed to another, this is clearly not understood, in either the ‘duty to all the world’ or the ‘general balancing’ variants, as corresponding to an underlying moral conception of correlative rights and duties. Rights theorists, on the other hand, see the element of duty of care as the feature of negligence law that captures its true relational character, which means precisely that it corresponds to an underlying moral duty that is correlative of a moral right. While the duty element is not, in the contemporary case law of either the United States or the Commonwealth, usually formulated explicitly in the language of rights, it is easy enough to discern the relational and rights-oriented character of the duty of care in the classic cases. Thus, in Donoghue v Stevenson Lord Atkin spoke of the duty of care as ‘a general conception of relations’,35 and in Palsgraf Cardozo CJ was even more explicit that what is at stake is a right on the part of the plaintiff: ‘What the plaintiff must show is a “wrong” to herself, i.e., the violation of her own right, and not merely a wrong to someone else, nor conduct “wrongful” because unsocial, but not a “wrong” to anyone’.36 Modern Commonwealth case law on the duty of care stems from Anns v London Borough of Merton,37 which laid out a two-stage test for determining the existence of a duty of care. According to the first stage, which builds on Lord Atkin’s ‘neighbour principle’, one asks whether there is a sufficient relationship of proximity or neighbourhood such that, in the reasonable contemplation of the [alleged wrongdoer], carelessness on his part may be likely to cause damage to the [person who has suffered damage]—in which case a prima facie duty of care arises.

35

Donoghue v Stevenson [1932] AC 562 (HL) 580. Palsgraf v Long Island Railroad, 248 NY 339 (New York CA, 1928) 343–4. The misinterpretation of Cardozo CJ’s opinions in MacPherson and Palsgraf in the service of various instrumentalist causes has been superbly documented by Goldberg and Zipursky, above n 6. For a particularly egregious instrumentalist misconstruing of Palsgraf, see Calabresi, ‘Concerning Cause and the Law of Torts’, above n 16, at 91–102. For an excellent defence of the view that Cardozo CJ in fact held a systematically rights-oriented understanding of tort law which can be discerned across the full range of his opinions, see P Benson, ‘The Problem with Pure Economic Loss’ (2009) 60 South Carolina Law Review (forthcoming). 37 Anns v London Borough of Merton [1978] AC 728 (HL) 751–2. 36

Duty of Care in a Rights-Based Theory of Negligence

91

At the second stage, one asks whether there are any considerations which ought to negative, limit, or reduce the scope of a prima facie duty that arises at the first stage. This two-stage test differs from both of the American approaches I have distinguished—ie, the ‘duty to all the world’ and the ‘general balancing’ approaches—in that it expressly emphasises the relationship between the parties, and it does so in a way that preserves the central role of reasonable foreseeability in defining that relationship. As McLachlin CJ interprets the first stage of the Anns test in Cooper v Hobart, ‘reasonable foreseeability must be supplemented by proximity’ where proximity is clearly intended to connote that the circumstances of the relationship inhering between the plaintiff and the defendant are of such a nature that the defendant may be said to be under an obligation to be mindful of the plaintiff ’s legitimate interests.38

McLachlin CJ goes on to state that the factors relevant to the determination of proximity include ‘expectations, representations, reliance and the property or other interests involved’.39 It is worth emphasising that these factors are, of course, among the most important considerations that bear on the question of whether one person has, as a moral matter, a right as against another person that the latter be mindful of the former’s legitimate interests.40 It is only at the second stage of the Anns test that policy considerations of the sort that figure directly in the onestage American general balancing test—in McLachlin CJ’s words, considerations that ‘are not concerned with the relationship between the parties, but with the effect of recognizing a duty of care on other legal obligations, the legal system, and society more generally’41—come into play. This two-stage, structured approach to the existence of a duty of care is very naturally interpreted as giving priority to considerations that bear on the existence of a duty on the defendant’s part which is correlative of a right on the part of the plaintiff, but which also allows policy considerations in the broader sense to come into play in a secondary or ancillary manner.42 38

Cooper v Hobart [2001] 3 SCR 537, 551 and 552. ibid, at 552. 40 See S Perry, ‘Protected Interests and Undertakings in the Law of Negligence’ (1992) 42 University of Toronto Law Review 247. 41 Cooper v Hobart [2001] 3 SCR 537, 551 and 554. 42 This understanding of the duty of care is consistent with the kind of pluralist account of tort law that I advocate in Perry, ‘Professor Weinrib’s Formalism’, above n 12, at 618–19, in which rights-based considerations take priority over policy considerations. As I observe in n 18 above, a strong version of such a pluralist account would treat the relevant notion of priority as lexical priority. Since the Anns test characterises the duty which arises at the first stage as a prima facie one only, which can then be negatived or limited at the second stage by considerations of policy, the underlying conception of priority is presumably not lexical. It should be pointed out that, in subsequent cases, the House of Lords has to some extent distanced itself from the Anns two-stage test. See Peabody Donation Fund v Sir Lindsay Parkinson & Co Ltd [1985] AC 210 (HL) 239–41 (Lord Keith); Yuen Kun-Yeu v A-G of Hong Kong [1988] AC 175 (PC) 190–94 (Lord Keith); Rowling v Takaro Properties Ltd [1988] AC 473 (PC) 501 (Lord Keith); Hill v Chief Constable of West Yorkshire [1989] AC 53 (HL) 60 (Lord Keith); Caparo Industries plc v Dickman [1990] 2 AC 605 (HL) 617–18 (Lord Bridge), 633 (Lord Oliver) and 658 (Lord Jauncey). These developments are insightfully discussed by Robert Stevens in R Stevens, ‘Law of Torts’ 39

92

Stephen Perry

III. The Moral Character of Risk I said in the preceding section that rights theorists see the element of duty of care as the feature of negligence law that captures its true relational character. That this is so is not an entirely straightforward matter, however, at least insofar as one of the conditions of adequacy of a rights-based theory is that it take seriously what I earlier referred to, following the helpful terminology of Coleman, as the ‘structure’ of tort law. The potential difficulty is that it is not entirely clear that the duty of care does, in fact, correspond to an independent moral duty. This is so for two different but related reasons, both of which concern the fact that the content of the duty of care would appear to be, at the level of doctrine, a duty to refrain from creating unreasonable or excessive risks to others, including most importantly risks of physical harm to the person and to tangible property. As Cardozo CJ memorably put the point in Palsgraf, ‘[t]he risk reasonably to be perceived defines the duty to be obeyed, and risk imports relation; it is risk to another or to others within the range of apprehension’.43 As was noted in the preceding section, it is possible, certainly as a doctrinal matter, to breach one’s duty of care—which means, essentially, to subject another person to an unreasonable level of risk—without causing that person physical harm. The first of the two reasons for thinking that the duty of care might not correspond to an independent moral duty is that it is very plausible to think that our most fundamental moral rights are, quite simply, rights not to be harmed in various ways, rather than rights that others not subject us to the risk of harm. The second reason is that even if our most fundamental rights are rights not to be harmed, it is not clear that there is room within the appropriate moral framework for a secondary or ancillary right not to be subject to unreasonable risks. Putting these two concerns together, the worry is that there is simply no such thing, morally speaking, as a right to be free of risk.44 To address these problems we must say something about the morality of risking, and to say anything interesting about the morality of risking we must say something about the nature of risk. In order to bring the concept of risk into in L Blom-Cooper, G Drewry and B Dickinson (eds), The Judicial House of Lords 1876–2009 (Oxford, Oxford University Press, forthcoming 2009). Stevens argues that the two-stage test in Anns went wrong in defining duty of care in terms of loss (damnum) instead of right (injuria) and, as a result, ‘the duty is not a real duty, with a correlative right held by specific others’. This may be correct, and that may well be why the House of Lords has since treated the Anns test with caution. Even so, however, I believe that a properly-formulated two-stage test is defensible, and I think the Supreme Court of Canada was on the right track in Cooper. In this regard it is worth drawing attention to the fact that in Yuen Kun Yeu, the Privy Council adopted a view of the Anns two-part test that is roughly equivalent to the view of the Supreme Court of Canada in Cooper, stating, among other things, that the second branch of the Anns test will seldom come into play: ([1988] AC 175 (PC) 193). 43

Palsgraf, NY 339 (New York CA, 1928) 344. The view that there is no such thing as a right to be free from risk is argued for in HM Hurd, ‘The Deontology of Negligence’ (1996) 76 Boston University Law Review 249; and in HM Hurd and MS Moore, ‘Negligence in the Air’ [2002] 3(2) Theoretical Inquiries in Law, Art 3 www.bepress.com/ til./default/vol3/iss2/art3, accessed on 18 November 2008, at 17–21. 44

Duty of Care in a Rights-Based Theory of Negligence

93

sharper moral focus, I will develop an example based on the English case of Hotson v East Berkshire Area Health Authority.45 Hotson was not a duty of care case, but rather a case concerning the requirement in negligence law that the plaintiff have suffered a legally cognisable injury. The plaintiff ’s claim in Hotson was that he had, as a result of the defendant’s negligence, lost a chance of avoiding an adverse physical outcome, and that this lost chance could properly be characterised as harm in its own right, of a kind that was capable of sustaining a negligence action. A number of prominent tort theorists have also argued for this view,46 and in the United States several courts have recognised such claims in the medical malpractice context.47 To say, however, that a person has suffered a reduced or lost chance of avoiding an adverse physical outcome is simply to say that the person has been subjected to an increased risk of suffering that same outcome. A ‘lost’ chance is the special case in which the chance of avoiding the adverse physical outcome has been reduced to zero, or, to put the point in terms of risk, in which the risk has been increased to a certainty. The plaintiff ’s claim in Hotson was, in essence, that he had been subjected to a risk, and that this risk was itself a legally cognisable injury.48 On a rights-based understanding of tort law, to say that a person has incurred a legally cognisable injury is to say that she has suffered harm, which means, in turn, a setback to a morally significant interest. Injuries to the person or to tangible property are paradigmatic instances of harm in this sense. If risk is to be understood as harm, it must be possible to specify an appropriate interest which has been set back. If it is possible to specify such an interest, then we would be well on the way to responding to the worry that there can be no such thing, morally speaking, as a right to be free of unreasonable risks. Such a right would simply be a particular example of a right not to be harmed. The facts in Hotson were that the plaintiff had suffered a fracture of his left femoral epiphysis, which the defendant health authority negligently failed to 45

Hotson v East Berkshire Area Health Authority [1987] AC 750 (HL). See, eg JH King, ‘Causation, Valuation, and Chance in Personal Injury Torts Involving Preexisting Conditions and Future Consequences’ (1981) 90 Yale Law Journal 1353; J Stapleton, ‘The Gist of Negligence, Part 2’ (1988) 104 Law Quarterly Review 389. 47 See, eg, Alberts v Schultz, 975 P 2d 1279 (Supreme Court of New Mexico, 1999); Beswick v City of Philadelphia, 185 F Supp 2d 418 (US District Court, Pennsylvania, Eastern District, 2001). The Draft Third Restatement seems cautiously to endorse these cases: Restatement Third Torts, § 26, comment n. In my view these cases often do not take seriously the claim that the lost or reduced chance is harm in itself. At times this idea seems to serve, rather, as a kind of fiction that permits the court to avoid applying the traditional standard of proof (preponderance of the evidence) to the cause in fact requirement, but without having to state explicitly any justification in policy or principle for doing so. This fiction also frees the court from having to offer any well-grounded reason for limiting the doctrine of recovery for a lost or reduced chance to the medical malpractice context. If a reduced or lost chance of avoiding an adverse physical outcome is harm in this kind of case, the same should be true in many other kinds of cases well. 48 Strictly speaking, since the case involved breach of an affirmative obligation based on the physician–patient relationship, the plaintiff ’s claim was that the defendant had failed to take reasonable steps to prevent a risk, which the plaintiff already faced, from increasing. This distinction is not important for present purposes. 46

94

Stephen Perry

diagnose. The trial judge found that at the time the plaintiff arrived at the defendant’s hospital there was a 75 per cent chance that even with proper treatment a serious condition called avascular necrosis would have occurred, but that the delay in treatment which resulted from defendant’s negligent diagnosis made the development of this condition a certainty. The plaintiff argued, and the trial judge accepted, that the defendant had caused the plaintiff to lose a 25 per cent chance of avoiding avascular necrosis, and that this lost chance should itself be treated as damage compensable in tort. In the House of Lords Lord Mackay put the plaintiff ’s argument this way: [Counsel] who appeared for the plaintiff … said that in relation to the facts of this case as found by the trial judge what was meant by a chance was that if 100 people had suffered the same injury 75 of them would have developed avascular necrosis … and 25 would not. This, he said, was an asset possessed by the plaintiff when he arrived at the authority’s hospital … It was this asset which [counsel] submits the plaintiff lost in consequence of the negligent failure of the authority to diagnose his injury properly.49

The plaintiff ’s argument, as summarised by Lord Mackay, is a crystal-clear statement of the idea that a risk or probability should be understood by reference to the notion of relative frequency. The 25 per cent chance of avoiding avascular necrosis which the plaintiff claimed to have lost is expressed as a relative frequency of the occurrence of avascular necrosis within a reference class of 100 persons who are said to have suffered the ‘same’ injury. This chance is said, moreover, to be an asset which the plaintiff possessed at the time he arrived at the defendant’s hospital. If this claim can be sustained, and if it can be further shown that in an appropriate sense the defendant caused the defendant to lose this asset, then it is at least arguable that the defendant set back an interest of the plaintiff ’s in a way that might be said to constitute harm. It is, however, notorious within the philosophy of probability that for any given situation involving probability or risk in the relative frequency sense, there is no determinate way to specify a reference class which can be regarded as the canonical or correct one. This gives rise to problems for the claim that risk is a harm in itself, which we can bring out by further reflection on the facts of Hotson.50 As mentioned, the reference class that figures in the plaintiff ’s argument is that of ‘100 people [who] … suffered the same injury’. But no two people ever suffer exactly the same injury, and we can therefore imagine a variety of different reference classes that might be formulated in a case like Hotson. The immediate cause of the avascular necrosis was a reduction in the supply of blood reaching the epiphysis, and one factor among others that affected the degree of reduction in blood supply was the proportion of blood vessels that had ruptured at the time of the original fracture. In the plaintiff ’s case it was not known how many blood 49

Hotson [1987] AC 750 (HL) 783. cf S Perry, ‘Risk, Harm, and Responsibility’ in D Owen (ed), Philosophical Foundations of Tort Law (Oxford, Clarendon Press, 1995) 330–39; S Perry, ‘Risk, Harm, Interests, and Rights’ in T Lewens (ed), Risk: Philosophical Perspectives (New York, Routledge, 2007) 193–8. 50

Duty of Care in a Rights-Based Theory of Negligence

95

vessels had originally been ruptured, but suppose that this fact was known. We could then formulate a different reference class consisting of 100 persons who had suffered a fracture of the left femoral epiphysis and who had also ruptured such-and-such a proportion of vessels supplying blood to the epiphysis. Suppose the proportion of ruptured blood vessels was relatively low. We could then expect to find, as an empirical matter, that the relative frequency of persons who will develop avascular necrosis even with proper treatment is, within this reference class, something less than 75 per cent. It might be, say, 40 per cent. Alternatively, suppose we knew that the proportion of ruptured blood vessels was relatively high. Then we could in parallel fashion formulate yet another reference class of persons, and within this reference class the relative frequency of persons who will develop avascular necrosis even with proper treatment would likely be higher than 75 per cent. It might be, say, 90 per cent. And we can further imagine coming to know more and more about the causes of avascular necrosis so as to be able to formulate, at least in principle, a reference class consisting of persons who fractured their left femoral epiphysis and who also possessed certain other characteristics, call them X, such that the relative frequency of developing avascular necrosis even with proper treatment was 100 per cent. Similarly, we can imagine that increased empirical knowledge would allow us to formulate a reference class consisting of persons who fractured their left femoral epiphysis and who also possessed certain other characteristics, call them Y, such that the relative frequency of developing avascular necrosis even with proper treatment is, within this reference class, 0 per cent. And of course, if we knew enough about the plaintiff ’s particular injury so as to be able to say either that he possessed the X characteristics or that he possessed the Y characteristics, then we would be in a position to say that his chance of developing avascular necrosis even with proper treatment was either 100 per cent or 0 per cent. As, I hope, the example begins to make clear, both the existence and the magnitude of a so-called lost chance of avoiding physical harm of some kind—or, what comes to the same thing, a risk of suffering that kind of physical harm—is a function of our current state of knowledge about both the general type of injury and the specific plaintiff ’s particular condition. Perhaps it would be more accurate to say that the existence and magnitude of the risk is a function of our current state of ignorance about these matters. But whichever way we put the point, it is a very odd form of harm the very existence of which depends in this way on the state of our knowledge.51 Notice that, in the example based on Hotson, the accumulation 51 The fact that a risk of suffering harm (or, equivalently, a reduced probability or chance of avoiding harm) is always a function of some specified level of knowledge has led some theorists to say that risk is by its very nature a purely epistemic phenomenon which has no existence in the physical world. From the perspective of an omniscient being, the risk is always either 0% or 100%, and therefore these are the only risks that can, according to these theorists, actually be said to exist. See, eg Hurd, above n 44, at 263–4; Hurd and Moore, above n 44, at 19. But, as the discussion in the text illustrates, this is not quite accurate. Risk or probability is simply defined as a relative frequency within a specified reference class, and for any given reference class, the relative frequency within it of some specified attribute

96

Stephen Perry

of greater empirical knowledge which permits us to define new reference classes does not justify the conclusion that statements of relative frequency that were made in a state of greater ignorance were erroneous. Thus a relative frequency of 75 per cent might be accurate for the admittedly rather vaguely identified reference class that figures in the quote from Lord MacKay, and the 40 per cent figure might be accurate for the different reference class that I described in my variant hypothetical. Both statements of relative frequency are (or at least might be) accurate, for the reason that a statement of relative frequency is always made relative to the specification of a reference class, and the plaintiff in Hotson belongs to both of these reference classes. One way of putting the significance of the fact that the plaintiff in Hotson belongs to many reference classes is to say that a probability, understood as a relative frequency, is a property of specified reference classes and not a property of the individuals who happen to fall into those classes. Any given individual will fall into an indefinitely large number of reference classes, where the relative frequency of harm will be, for each one, different. For this reason, counsel for the plaintiff in Hotson was wrong to argue that the plaintiff ’s supposed 25 per cent chance of avoiding avascular necrosis if proper treatment had been administered was an asset that the plaintiff possessed at the time that he entered the hospital. Even if we assume that the 25 per cent figure was accurate for some reference class that could be identified with some reasonable degree of specificity, the plaintiff also belonged, as a matter of objective fact, to indefinitely many other reference classes for which the relative frequency of avoiding the physical harm was different. In one of the directions in which I developed the hypothetical, the plaintiff simultaneously belongs to reference classes for which the relative frequency of avoiding the harm if proper treatment had been administered is 25, 60 and 100 per cent. But surely the plaintiff did not simultaneously possess assets of a 25 per cent chance, a 60 per cent chance, and a 100 per cent chance of avoiding the harm if appropriate treatment had been rendered. This is just another way of saying that a decreased chance of avoiding a physical harm—or, equivalently, an increased risk of suffering that harm—cannot plausibly be viewed as itself being a form of harm.

such as the occurrence of physical harm is, or at least is capable of being, an objective fact about the world. Of course, as a general matter, and certainly in the tort context, we will ordinarily try to discover and employ the narrowest causally relevant reference class that we can, but in other contexts that may not be true. In epidemiology, for example, there may be good reasons to focus on, say, the 5% risk of disease occurring within a general population under certain conditions, and this may be true even if we have the knowledge to state relative frequencies of the disease for various more precisely specified sub-classes within the general population. For further discussion of these and related matters, see S Perry, ‘Responsibility for Outcomes, Risk, and the Law of Torts’ in GJ Postema (ed), Philosophy and the Law of Torts (New York, Cambridge University Press, 2001) 98–9 (2001); Perry, ‘Risk, Harm, Interests, and Rights’, above n 50, at 193–8.

Duty of Care in a Rights-Based Theory of Negligence

97

IV. Duty of Care and Risk In the preceding section I argued that the House of Lords was correct to reject the plaintiff ’s argument in Hotson that risk is a form of harm. Of course Hotson was not a duty of care case, and it might seem that we have strayed rather far from our original questions about the true moral character of the defendant’s duty and the plaintiff ’s correlative right. But we have not strayed so far as all that, because in the course of this digression we have learned something about the nature of risk, and the morality of risking is the central concern of the doctrinal requirement of duty of care. I have argued elsewhere that a statement to the effect that an agent A imposed a risk of harm on another person B must, for purposes of both morality and the law, always be made relative either to A’s actual epistemic perspective or to a constructed epistemic perspective that is, for moral and legal purposes, attributable to A.52 An epistemic perspective involves a specification, either implicit or explicit, of the knowledge of the relevant facts that A either possessed (actual epistemic perspective) or ought to have possessed (constructed epistemic perspective) at the time of acting. Such a specification will in turn involve, usually as an implicit rather than an explicit matter, the idea that a certain description of the kind of action engaged in by A—and, hence, that a certain characterisation of a background reference class of types of actions or effects, together with an associated relative frequency of a specified kind of harm within that reference class—is the appropriate or correct description for purposes of morally and legally evaluating A’s actual act of risk imposition. The notion of the reasonable person in negligence law involves, among other things, a constructed epistemic perspective to which defendants are held regardless of their actual state of knowledge. If the argument that the plaintiff advanced in Hotson were correct, it is very difficult to see how to avoid the conclusion that every risk that one person imposes on another is an instance of harm.53 The proposition that every risk imposition is harm in itself no doubt seems absurd on its face. Some rights theorists have, however, argued for the proposition that every risk imposition is a rights violation, and that our most fundamental moral rights are precisely rights not to be subject to any risk of physical harm whatsoever, whether known or unknown. The moral

52 See Perry, ‘Risk, Harm, Interests, and Rights’, above n 50, at 192–3, 197–8. See also Perry, ‘Responsibility for Outcomes, Risk, and the Law of Torts’, above n 51, at 97–101. 53 There are two aspects to this generalisation. First, as was observed in n 47 above, the idea that a reduced or lost chance is harm in itself has for the most part only succeeded in gaining a foothold in the case law in medical malpractice cases. But the reasons for implicitly limiting the lost chance doctrine to the medical malpractice context are almost certainly policy-based, rather than grounded in some principled understanding of which risks are to be regarded as harm and which are not. Second, there is no obvious reason for limiting the lost chance doctrine to lost or reduced chances (or increased risks) which result from negligence. Lord Hoffmann essentially recognized both points in the House of Lords’ recent revisiting of the loss of chance issue in Greg v Scott [2005] 2 AC 176 (HL) 195–200. The English case law on loss of chance and related issues is illuminatingly discussed in D Nolan, ‘Causation and the Goals of Tort Law’ ch 7 of this book.

98

Stephen Perry

philosopher David McCarthy has argued for this view, for example.54 McCarthy maintains that, first, many risky actions are permissible even though they infringe rights, because their benefits outweigh their costs,55 and secondly, that while all rights-infringing risky actions in principle require compensation, one acceptable way to meet this requirement is to guarantee that the rights-holder will be compensated after the fact if the risk materialises and physical injury occurs. This latter, rather ingenious move, which McCarthy calls the Natural Lottery Rule of compensation, permits him to argue that, contrary to appearances, damages in tort law really compensate for risk rather than for physical harm. He contrasts the Natural Lottery Rule with the Direct Payment Rule, which holds that each and every risk imposition prima facie requires the payment of compensation, whether or not the risk is known and whether or not physical injury occurs. Since it is not, to say the least, a straightforward matter to compensate for unknown risk, it is just as well that McCarthy has an alternative conception of compensation available to him. The Natural Lottery Rule brings his thesis that every risk imposition is a rights violation into line with at least one fundamental feature of negligence law, which is that, as a general matter, it is a requirement of liability for unintentional harm that the plaintiff have suffered physical injury. But McCarthy’s view nonetheless differs radically from our current system in that the associated standard of liability would be absolute or close to absolute, meaning that, in a broad range of cases, liability would be imposed simply for causing harm. McCarthy says little about how he thinks the concept of risk should be analysed, but it seems evident that he, like most rights theorists who discuss risk, implicitly has in mind a relative frequency theory of the kind that I discussed in connection with the loss-of-chance cases. Thus he gives the example of going for a drive in the country and thereby imposing a risk of one in a million on Jones, who lives near the road. McCarthy clearly assumes that there just is some determinate fact of the matter, which is independent of anyone’s state of knowledge, as to what risk the driver has objectively imposed on Jones; he argues, for example, that one of the advantages of an after-the-fact compensation rule is that it permits us to take moral account of even unknowable risks. The arguments I offered earlier in connection with loss-of-chance cases like Hotson show, however, that there is no such determinately objective fact of the matter. Jones belongs, as a matter of objective fact, to indefinitely many reference classes which differ from one another with respect to how, exactly, the surrounding circumstances are characterised—was the driver simply ‘driving’, did she drive ‘negligently’, did she ‘swerve’, was she drunk, did the car have a perhaps unknown defect, how far was the driver from the edge of the road, how far was Jones from the edge of the road, was Jones drunk, and

54 See D McCarthy, ‘Rights, Explanation, and Risks’ (1997) 107 Ethics 205; D McCarthy, ‘Liability and Risk’ (1996) 25 Philosophy and Public Affairs 238. 55 Thus McCarthy understands rights theory as having a consequentialist dimension, a theoretical possibility that I allowed for earlier in the text.

Duty of Care in a Rights-Based Theory of Negligence

99

so on—and with respect to each distinct reference class the relative frequency of harm to someone in Jones’s situation is different.56 While McCarthy’s thesis that every risk imposition infringes rights might seem on its face to be completely implausible, it at least has the virtue of giving a relatively straightforward answer to a notoriously difficult question in rights theory, namely: How do we distinguish between risks that violate rights and risks that do not? McCarthy says, quite simply, that all risks violate rights. But if we reject that view, as most rights theorists do, we are left, as I say, with a very difficult question in moral theory. The difficulty, in brief, concerns whether and under what circumstances we are to balance the risk to the plaintiff (and possibly to third parties) against the costs to the defendant of reducing or eliminating the risk. In negligence law this issue arises, doctrinally speaking, as a concern with the content of the standard of reasonable care. The law says that we must act as a reasonable person would, but this simply pushes the difficulty back a step. What is reasonable care? Which risks is it reasonable to impose on others and which not? English law has grappled with these questions in a line of cases in which Bolton v Stone57 and Wagon Mound (No 2)58 figure prominently, but in my view neither the law nor moral philosophy has so far given us a truly satisfactory solution. Nor will I try to remedy that deficiency in this chapter.59 For present purposes I will simply offer the following two observations. First, instrumentalist tort theories do seem to have at least an initial advantage here, since they formulate the standard of reasonable care in fairly precise terms along the lines of the Learned Hand test or some similar form of cost-benefit analysis.60 Secondly, once we have rejected McCarthy’s thesis that all risk impositions violate rights, we have no obvious reason to adopt a modified version of his view, according to which our most fundamental moral rights are still rights not to be subject to risk, but in which the risk impositions that are said to violate rights are, quite simply, those that involve a breach of the duty of care in negligence law. This line of thought is pursued by the legal theorists Toby Handfield and Trevor

56 I offer a more detailed critique of McCarthy’s thesis in Perry, ‘Risk, Harm, Interests, and Rights’, above, n 50, at 203–5. 57 Bolton v Stone [1951] AC 850 (HL). 58 Wagon Mound (No 2) [1967] 1 AC 617 (PC) 641. 59 I have offered a preliminary attempt to answer these questions in S Perry, ‘The Impossibility of General Strict Liability’ (1988) 1 Canadian Journal of Law & Jurisprudence 147, 169–71; and Perry, ‘Responsibility for Outcomes, Risk, and the Law of Torts’, above n 51, at 108–15, but I am by no means satisfied with the proposals advanced in those articles. For excellent discussions of some of the difficulties that must be confronted in formulating the content of rights not to be subject to risk, see R Nozick, Anarchy, State, and Utopia (New York, Basic Books, 1974) 54–87; D McKerlie, ‘Rights and Risk’ (1986) 16 Canadian Journal of Philosophy 239. 60 However, as Hurd and Moore correctly point out, the cost-benefit analysis that a correct application of the Learned Hand test actually requires under an economic or utilitarian understanding of negligence law is far more complicated than most discussions of the issue, including those offered by many instrumentalist theorists, usually assume. See further note 88 below.

100

Stephen Perry

Pisciotta.61 They accept what I take to be the intuitively unappealing aspect of McCarthy’s view—that what tort law is really doing, despite all appearances to the contrary, is compensating for the imposition of risk rather than for the causation of physical harm—while at the same time rejecting what I take to be the main virtue of McCarthy’s approach, which is that it offers a fairly straightforward answer to the question of which risks violate rights. The answer that Handfield and Pisciotta offer to the question of which risks violate rights is the same unsatisfying answer that negligence law gives, namely, those risks that are associated with conduct that falls below the standard of reasonable care. As I have suggested, this is not really an answer at all, but at best only the beginnings of an answer. If our most fundamental moral rights are not rights not to be subject to (certain) risks, what are they? The obvious alternative, and indeed the most intuitively appealing answer to this question, is that they are, quite simply, rights not to be physically harmed. This issue in tort theory has an analogue in criminal law theory. The distinction between duties not to cause harm and duties to try not to cause harm is the crux of the fundamental distinction between so-called objectivist and subjectivist theories of the criminal law. Framing the issue in terms of duties rather than rights, John Gardner characterises the basic question in moral theory as a distinction between a duty to succeed, meaning a duty to succeed in not causing harm, and a duty to try, meaning a duty to try to avoid causing harm.62 Adopting language that is more amenable to the issue as it arises in tort law, Arthur Ripstein and Benjamin Zipursky draw a similar distinction between what they call duties of ‘non-injury’ and duties of ‘non-injuriousness’.63 It is, as I say, intuitively appealing to suppose that our most fundamental rights are rights not to be harmed, as opposed to rights that others try not to harm us or rights that they not subject us to certain risks of harm, and this is certainly the view that seems to be embedded in ordinary language and in the language of the law. Criminal law, for example, follows ordinary language in defining most offences in terms of causally complex acts like killing, wounding and stealing, and the subjectivist has his work cut out for him in showing that what the criminal law is really up to is prohibiting acts of trying to kill or wound, or acts that risk killing or wounding.64 Despite the clear intuitive appeal of the view that our most fundamental moral duties are duties not to cause harm, and that our 61 T Handfield and T Pisciotta, ‘Is the Risk-Liability Theory Compatible with Negligence Law?’ (2005) 11 Legal Theory 387, 395. 62 J Gardner, ‘Obligations and Outcomes in the Law of Torts’ in P Cane and J Gardner (eds), Relating to Responsibility: Essays in Honor of Tony Honoré on His 80th Birthday (Oxford, Hart Publishing, 2001) 120–25. The language of a duty to try to avoid causing harm seems more at home in the criminal law than in the law of torts. Gardner nonetheless argues that the duty of care in negligence law involves a duty to try, in his relatively technical sense of that term. 63 See A Ripstein and B Zipursky, ‘Corrective Justice in an Age of Mass Torts’ in GJ Postema (ed), Philosophy and the Law of Torts (New York, Cambridge University Press, 2001) 219. 64 For a distinguished recent defence of the subjectivist view in criminal theory, see SJ Morse, ‘Reason, Results, and Criminal Responsibility’ [2004] Illinois Law Review 363.

Duty of Care in a Rights-Based Theory of Negligence

101

most fundamental moral rights are correlative rights not to be caused harm, this is nonetheless a very difficult issue in moral philosophy, which I cannot fully address here.65 But we must venture at least some way into this difficult philosophical terrain in order to arrive at a proper understanding of the role of the duty element in negligence law. On one view, which has been defended by the moral philosopher Judith Jarvis Thomson66 and, in a slightly different form, by the libertarian legal theorist Richard Epstein in his early, rights-based work on torts,67 our most fundamental moral right is a right not to be harmed tout court, which would translate legally into a standard of absolute liability. I have argued elsewhere against this view on the grounds that, inter alia, any such right would be unavoidably indeterminate in content and morally extremely unattractive.68 I will not repeat those arguments here. The point I wish to draw attention to for present purposes is that a right not to be harmed need not be a right not to be harmed full stop, but can instead be—to adopt the useful terminology of Ripstein and Zipursky69—qualified in various ways. Thus, the most fundamental moral right underlying negligence law is, very roughly, a right not to be harmed as a result of someone else acting negligently towards one. The correlative moral duty is a duty not to harm others as a result of acting negligently towards them. Since the idea of acting negligently towards another involves, among other things, subjecting the other to an unreasonable degree of risk, we can describe the qualified underlying moral right as, very roughly, a right not be harmed as a result of having being subjected to a certain unacceptable degree of risk. I say this is ‘roughly’ what the underlying moral right is because the idea of acting negligently ‘towards’ another needs to be further unpacked, so as to take account of the distinctive role played by the notion of reasonable foreseeability. This is an essential aspect of the duty of care which was famously addressed by Cardozo CJ in Palsgraf and by Lord Atkin in Donoghue v Stevenson. I will take up this aspect of what it means to act negligently towards another in the following section. In the meantime, I will simply note that the manner in which the fundamental moral right underlying negligence law is qualified entails, among other things, that an actor does not violate her correlative moral duty unless she causes harm as a result of engaging in conduct that falls below the standard of reasonable care. The standard of reasonable care is, as we saw earlier, a

65 For philosophical arguments in favour of the intuitively appealing view, see the valuable discussions in H Hurd, ‘What in the World is Wrong?’ (1994) 5 Journal of Contemporary Legal Issues 157; and in Gardner, above n 62, at 134–43. 66 JJS Thomson, The Realm of Rights (Cambridge, Massachusetts, Harvard University Press, 1990) 227–48. 67 See, eg, R Epstein, ‘A Theory of Strict Liability’ (1973) 2 Journal of Legal Studies 151. 68 See S Perry, ‘The Impossibility of General Strict Liability’, above n 59; S Perry, ‘Libertarianism, Entitlement, and Responsibility’ (1997) 26 Philosophy & Public Affairs 351. 69 Ripstein and Zipursky, above n 63, at 219–20.

102

Stephen Perry

mandatory legal norm. It is thus a necessary condition of being held liable for negligence that one have breached a mandatory norm, and for that reason negligence is quite properly characterised as a fault-based tort.70 But the manner in which a moral right not to be harmed can be qualified does not necessarily require the breach of a mandatory norm, and for that reason there can be, certainly as a conceptual matter, true torts of strict liability.71 In American law, vigorous versions of strict liability can be found in the doctrine of abnormally dangerous activities72 and in the American understanding of the rule in Rylands v Fletcher.73 In the case of both doctrines, the qualification of the relevant right not to be harmed simply involves, in the first instance, the fact that the defendant engaged in a certain activity or type of activity.74 In both cases the right is further qualified so as to require that the relevant type of harm be reasonably foreseeable to someone in the defendant’s position, but, differently from the tort of negligence, the role of reasonable foreseeability is not, in the case of these doctrines, to help define a duty of care. This is because, in the case of strict liability, compliance with the duty does not, unlike negligence law, require compliance with a mandatory norm of reasonable care; the duty that corresponds to the right is, quite simply, a duty not to cause reasonably foreseeable, sufficiently proximate harm as a result of engaging in the specified activity or type of activity. A number of prominent rights theorists, including Ernest Weinrib and Alan Brudner, have argued that, doctrinal appearances to the contrary notwithstanding, most so-called torts of strict liability really involve a fault standard, and that any other understanding would be morally incoherent or otherwise morally

70 It should be borne in mind that, on a rights-based view of negligence law, the standard of reasonable care, while obviously a legal norm in doctrinal terms, is also a moral norm. According to the rights-based view, the law aspires to translate moral rights, duties, and norms more or less directly into legal rights, duties and norms. It is because the translation is supposed to be more or less direct that we face the puzzle of how it is possible as a doctrinal matter to breach one’s duty of care without causing harm when the underlying moral duty is, on the view I am entertaining, precisely a duty, albeit a qualified one, not to cause harm. 71 Strict liability is of course very different from absolute liability, which involves a duty not to cause harm tout court. Although in his early rights-based work on torts Richard Epstein said that he was arguing in favour of a general standard of strict liability, it is more accurate, and certainly less confusing, to say that he was actually arguing in favour of a form of absolute liability. 72 Restatement, Second: Torts §§ 519–20. 73 Rylands v Fletcher (1868) LR 3 HL 330. For the modern American understanding of Rylands, see Restatement, Second: Torts, § 520, Comment i. For the modern English understanding, see, eg, Cambridge Water Co v Eastern Counties Leather plc [1994] 2 AC 264 (HL) 306, in which Lord Goff suggested that the rule in Rylands v Fletcher should be regarded ‘essentially as an extension of the law of nuisance’. 74 In the case of the doctrine of abnormally dangerous activities, § 520 of the Second Restatement sets out criteria for determining which activities are to be considered abnormally dangerous and which are not, but case law has determined that certain activities in particular, the paradigmatic example of which is blasting, always fall into this category. The rule in Rylands v Fletcher is generally understood by reference to the criterion set out in § 520(d), which looks to ‘the extent to which the activity is not a matter of common usage’. In effect, then, the rule in Rylands has been absorbed into the more general doctrine of abnormally dangerous activities.

Duty of Care in a Rights-Based Theory of Negligence

103

problematic.75 Although I cannot take up the issue in detail here, I believe that they are mistaken in this regard.76 It is important to emphasise that arguments for the view advocated by Weinrib and Brudner must be made on moral grounds, and once it is accepted that our fundamental moral rights are rights not to harmed in certain qualified ways, it is easy to see that the tort of negligence and true torts of strict liability have much more in common than is often thought to be the case.77 Put crudely, the difference between them simply concerns the manner in which fundamental moral rights not to be harmed are qualified. Contrary to what is sometimes suggested, it is not the case even in American tort law, which is dominated by instrumentalist thinking, that torts of strict liability can only be defended on instrumentalist grounds. Returning to the tort of negligence, it has been suggested that Cardozo CJ, in adopting in Palsgraf a relational and hence rights-based understanding of the duty of care, committed himself to the position that the fundamental moral right underlying negligence law is a right not to be subjected to certain risks, rather than, as I have been suggesting, a right not to be caused harm as a result of another person acting negligently towards one. Thus Heidi Hurd and Michael Moore summarise what they take to be Cardozo’s chain of inferences in the following fashion: [N]egligence implies (in the sense of presupposes) wrongdoing; wrongdoing implies a rights-violation; a rights-violation occurs only when the rights-holder is unreasonably risked, not when the rights-holder is caused harm. This latter is true because a victim’s rights (with respect to non-intentional interference) are rights against being placed at risk of harm.78

This is, in my view, an uncharitable and forced interpretation of Cardozo CJ’s reasoning in Palsgraf. Although he may not have explicitly stated that the fundamental underlying moral right at stake in negligence law is a qualified right

75 Weinrib, The Idea of Private Law, above n 9, at 1284–303; A Brudner, The Unity of the Common Law: Studies in Hegelian Jurisprudence (Berkeley, University of California Press, 1995) 190. Ripstein also accepts that the rule in Rylands v Fletcher and related doctrines of strict liability are best understood as involving ‘wrongs of fault’, but I believe that he understands the notion of fault in a broader and more abstract manner than either Weinrib or Brudner. See Ripstein, above n 23, at 26–7. 76 Let me very briefly address one of Weinrib’s arguments. Weinrib maintains that ‘[t]o ascribe liability to an action, regardless of culpability, for whatever harmful effects it has simply because they are its effects, is to hold the agent liable for being active’. Since, Weinrib continues, strict liability judges action by its effects in just this way, it ‘treats the defendant’s agency as an incoherent normative phenomenon’: Weinrib, The Idea of Private Law, above n 9, at 181. It is a mistake, however, to think that strict liability ‘hold[s] the agent liable for being active’. The correct normative characterisation of strict liability is that it holds persons to an appropriately qualified duty not to cause harm of some specified type. Liability for violation of such a duty is thus liability for causing harm as a result of acting in the appropriately qualified manner; it is not liability for simply ‘being active’. If one wishes to speak in terms of ‘ascribing liability to an action’, the appropriate characterisation of the action will be a causally complex description that includes the effect in question (for example, an act of killing). For a related criticism of Brudner’s claim that strict liability involves a form of extreme self-effacement, see Gardner, above n 62, at 111–6. 77 This point is made particularly well by John Gardner. See Gardner, above n 62, at 123–5. 78 Hurd and Moore, ‘Negligence in the Air’, above n 44, at 16.

104

Stephen Perry

not to be harmed, rather than a right not to be subject to certain risks, the former understanding of rights clearly informs his thinking throughout his opinion, as for example when he states that ‘bodily security is protected, not against all forms of interference or aggression, but only against some’.79 Cardozo CJ made the point explicitly and very clearly in his earlier decision in MacPherson: ‘A poison falsely labelled is likely to injure anyone who gets it. Because the danger is to be foreseen, there is a duty to avoid the injury’.80 Lord Atkin made the point just as clearly in Donoghue v Stevenson, when he famously wrote that ‘[t]he rule that you are to love your neighbour becomes in law, you must not injure your neighbour’.81 It is true that Cardozo CJ’s doctrinal focus in Palgraf was on the duty of care, and it is of course also true that an actor can breach her duty of care even without causing physical injury. How that fact about doctrine can be reconciled with the view that the most fundamental moral right underlying negligence law is a right that others not cause one physical harm by acting negligently towards one, rather than a right that others simply not subject one to an unreasonable degree of risk of physical harm, is an issue that must be addressed, and I will attempt to address it below. For now, the important question is this: Why do Hurd and Moore attribute to Cardozo CJ the view that not just the legal right but also the underlying moral right is, quite simply, a right not to be risked, as opposed to a right not be physically harmed? Why do they think that Cardozo CJ must be understood as holding this view? We can begin to answer this question by observing that Hurd and Moore maintain that the concepts of wrongdoing, rights-violation and duty-violation are ambiguous between two different views of their objects. According to one view, ‘our rights are rights against being caused harm of certain kinds’, whereas according to the second view, ‘our rights are rights against having others intend or risk such harms’.82 They then suggest, without purporting to offer a full defence of this conclusion, that the first view is the correct one, essentially because it is more consistent with the categorical force of our most fundamental moral duties.83 Hurd and Moore go on to argue that ‘risk is an epistemic notion’,84 by which they mean, roughly, that judgements of risk depend on the knowledge, actual or constructed, of the person making the judgement, in much the same way that I concluded earlier, in connection with the discussion of Hotson, that assessments of risk must always be offered from a specified epistemic perspective. They then suggest that ‘[e]pistemic failure by persons is the locus of their culpability, not their wrongdoing’.85 Cardozo CJ, having failed to grasp the moral implications 79 80 81 82 83 84 85

Palsgraf v Long Island Railroad, 248 NY 339 (New York CA, 1928) 345. MacPherson v Buick Motor Co, 217 NY 382 (New York CA, 1916) 385. Donoghue v Stevenson [1932] AC 562 (HL) 580. Hurd and Moore, above n 44, at 16 (emphasis in original). Hurd defends this view at much greater length in Hurd, ‘What in the World is Wrong’, above n 65. Hurd and Moore, above n 44, at 20. ibid.

Duty of Care in a Rights-Based Theory of Negligence

105

of the epistemic character of risk, opted for the second, mistaken view of what the objects of the concepts of wrongdoing, rights-violation and duty-violation should be taken to be. Andrews J, on the other hand, saw that ‘rights, duties and wrongs presuppose … “a relationship between [a person] and those whom he does in fact injure”’.86 I have already indicated that I am sympathetic to the view expressed by Hurd and Moore that our most fundamental moral rights are indeed rights not to be harmed, as opposed to rights that others not try to harm us or that they not subject us to the risk of harm. And while I would not put the point in quite this way,87 I have also indicated that I am sympathetic to the view that risk must be understood in epistemic terms. Where, then, do Hurd and Moore go wrong in their interpretation of Cardozo CJ, and, ultimately, in their view of the moral character of risk? It seems to me that their arguments go astray in the following way. While they accept, correctly, that our most fundamental moral rights are rights not to be caused harms of various kinds, they apparently do not accept that these rights can be subject to qualifications of the sort I discussed earlier, for example the qualification that the right is only violated if the actor acted negligently towards the person harmed. Their view is instead, if I understand them correctly, that rights violations do not have moral or legal implications unless they are accompanied by some form of culpability, which in the case of the criminal law is usually one of the standard forms of mens rea and that in the case of negligence, whether in the criminal law or in tort law, involves a breach of the standard of reasonable care. The right not be caused harm to one’s person is, thus, a right not to be caused harm to one’s person, full stop. The right is violated whenever one individual causes personal injury to another individual, and this is true even though the rights-violation does not have moral or legal implications unless it is accompanied by an appropriate form of culpability. While I cannot discuss the issue in detail here, I believe that Hurd and Moore are mistaken in thinking that rights and their correlative duties cannot be qualified along the lines I have been discussing. Their error is, in effect, to adopt a conception of rights which is too closely oriented to the distinction in criminal law between actus reus and mens rea. The correct view of the most fundamental 86 ibid, at 17, quoting Andrews J in Palsgraf, 248 NY 339 (New York CA, 1928) 102. Andrews J, it is worth pointing out, does not actually say that our ‘rights, duties and wrongs’ presuppose a relationship between a person and those whom he in fact injures, but rather that negligence presupposes such a relationship. Andrews J, while not conceding Cardozo CJ’s claim that risk is relational, does nonetheless think that negligence involves unreasonable risk-taking. Andrews J’s opinion is confusing in a number of respects, and extracting a clear and precise statement of his moral views on rights, risk and harm is no easy task. While I cannot undertake that task here, let me state that I think the correct conclusion is that he does not actually hold a rights-based view of negligence at all; the duty he says is owed ‘to the world at large’ is not a duty owed to others, but rather a duty simply to conform one’s conduct to the legal norm prohibiting unreasonable risk-taking. See further n 31 above and n 91 below. 87 See n 51 above.

106

Stephen Perry

moral right underlying negligence law, and the view I believe that Cardozo CJ was in effect endorsing, is a right that others not cause us reasonably foreseeable physical harm by acting negligently towards us, or, put differently, that they not cause us reasonably foreseeable physical harm by subjecting us to an unreasonable degree of risk. It is true, as I acknowledged earlier, that determining which acts can be said to subject us to an unreasonable degree of risk is a difficult problem in moral theory, and solving that problem requires, among other things, constructing an appropriate epistemic perspective which can be used to assess actors’ conduct regardless of their actual and possibly quite idiosyncratic beliefs about the riskiness of their actions. But we do not have to solve this problem in its entirety to see that a solution involving the idea of a qualified right is along the right lines. And whatever the complete details of the ultimate solution might be, we can see that there can be categorical (and not just consequentalist) qualified duties involving risk by acknowledging that, intuitively, it is, at the very least, categorically wrong to cause injury to others by subjecting them to substantial risks of physical harm.88 Hurd and Moore’s discussion of risk does, nonetheless, raise an interesting question about rights-based theories of negligence law. As I have already noted, it is true that, as matter of doctrine, one can breach one’s duty of care without committing the tort of negligence, for the simple reason that one’s negligence did not cause anyone, or at least the plaintiff, any physical harm. As a matter of doctrine, then, the duty of care does not correspond to what I am suggesting is the fundamental moral right underlying negligence law, namely, a qualified right not to be caused physical harm. Is this a significant problem for a rights-based understanding of negligence? Although I cannot undertake the task here, I believe that it can be shown that, once we accept that our most fundamental moral rights are qualified ones, such as the right that others not cause us reasonably foreseeable physical harm by behaving negligently towards us, there can be derivative moral rights, such as a right that others not behave negligently towards us, full stop. Intuitively, there does not seem to be any reason why there cannot be such derivative or ancillary rights, and elsewhere

88 cf Lord Reid’s judgement in Bolton v Stone [1951] AC 850 (HL). Later in their article Hurd and Moore argue that there can be, and are, consequentialist duties not to create certain risks, but these duties are non-relational; as the title of their article suggests, they accept that, contrary to Cardozo CJ’s view of the matter, negligence can be ‘in the air’: Hurd and Moore, above n 44, at 18, 25–33. In the course of developing their view they suggest, I believe correctly, that a correct consequentialist understanding of risk imposition cannot be limited to a simple cost-benefit analysis of the risks that one particular defendant happened to impose on one particular plaintiff; it must involve, rather, a cost-benefit analysis that takes account of all the risks that a given defendant’s action created, and this will involve a calculation that is exceedingly more complicated than a simple application of the Learned Hand test to two persons and two persons only. Thus the apparent simplicity that efficiency theorists tend to attribute to acceptance of the Learned Hand test as the appropriate formulation of the negligence standard is illusory. Of course this is not a problem for a properly-formulated rights theory because the only risks that matter, so far as the rights of this particular plaintiff are concerned, are the risk that the defendant imposed upon her, and not the risks that the defendant might also have happened to have imposed on others. That is the core of Cardozo CJ’s insight in Palsgraf.

Duty of Care in a Rights-Based Theory of Negligence

107

I have tried to show in more detail how this can be so within the general framework of an interest theory of rights.89 However, even if I am wrong about this, I do not think that the resulting lack of correspondence between the doctrinal duty of care and the fundamental moral right would pose a serious problem for a rights-based theory. It would simply become necessary explicitly to recognise that the basic elements of the tort of negligence—duty, breach, injury, cause in fact and proximate cause—would amount, in essence, to an artificial separation of different aspects of what is, morally speaking, a single, complex, portmanteau right of a certain kind, namely, a right not to be caused reasonably foreseeable physical harm as a proximate result of another person’s engaging in unreasonably risky conduct towards one. It is true that the fact that this separation is an artificial one would place the standard claim of rights theorists to take seriously the doctrinal structure of tort law in a somewhat curious light, since the duty of care in law would not then directly correspond to a distinct duty in morality. But it would correspond to one aspect of the actual moral duty, and that seems to me to be sufficient to sustain a plausible, rights-based interpretation of negligence law. From the practical perspective of the litigation process, there is obviously an enormous amount to be said in favour of artificially separating different aspects of the single right into distinct doctrinal requirements that the plaintiff must meet one by one. That is a form of instrumentalism, but in my view it is instrumentalism of a kind that should be completely acceptable to a rights theorist.

V. The Role of Reasonable Foreseeability I said in the preceding section that the fundamental moral right underlying negligence law is, roughly, a right not to be caused reasonably foreseeable physical harm as a proximate result of another person’s engaging in unreasonably risky conduct towards one. I also said that this rough approximation of the underlying moral right needs to be further unpacked so as to take account of the distinctive role played by the notion of reasonable foreseeability. Foreseeability of not just the type of harm but also of a person or class of persons placed at risk by the defendant’s conduct was central to both Lord Atkin’s and Cardozo CJ’s view of the relational nature of the duty of care. Thus the proposition that the defendant engaged in unreasonably risky conduct ‘towards’ the plaintiff must be further unpacked so as to include the idea that the defendant ought to have foreseen that he or she was acting so as to create an unreasonable risk of physical harm for a reasonably foreseeable class of persons that included the plaintiff. It is, of course, 89 See; Perry, ‘Responsibility for Outcomes, Risk, and the Law of Torts’, above n 51, at 76–7; Perry, ‘Harm, History, and Counterfactuals’ (2003) 40 San Diego Law Review 1283, 1304–9; Perry, ‘Risk, Harm, Interests, and Rights’, above n 50, at 202–4.

108

Stephen Perry

true that an actor can only take account of the potentially harmful effects of her actions if she has the capacity to foresee that her actions might, in fact, cause harm to another person. But instrumentalists often argue that foreseeability is appropriately dealt with in the calculus of risk that is, they claim, built into the standard of care. The Draft Third Restatement endorses this view, and explicitly rejects the idea that foreseeability has a role to play at the doctrinal level of duty.90 What is perhaps surprising is that some rights theorists essentially agree with the instrumentalists on this point. Even such prominent defenders of a rights-based, relational understanding of tort law as Goldberg and Zipursky appear to downplay the role of reasonable foreseeability in determining a defendant’s duty of care, and they interpret Cardozo CJ’s landmark opinions in MacPherson and Palsgraf as doing likewise. Goldberg and Zipursky argue, correctly, that as a conceptual matter, even an understanding of the duty of care that is relational, in the sense of being rightsbased, could be universal, in the sense that the duty is owed to every other individual in the world. In the famous language of Andrews J in his dissenting opinion in Palsgraf, the duty, while relational, would nonetheless be ‘owed to the world at large’.91 Goldberg and Zipursky argue that, doctrinally and morally, the duty of care is not, generally speaking, universal in this sense, and they interpret Cardozo CJ as recognising as much in Palsgraf. But they nonetheless suggest that, in the core cases of causing physical injury or property damage to others, duties of care are ‘universal … in the relational but unrestricted sense’: ‘These are duties of care each person owes to each other person’.92 There is a tension here, however, at least insofar as they are attributing this view to Cardozo CJ, because Cardozo CJ clearly regarded Palsgraf as just such a core case. Goldberg and Zipursky suggest that, in Palsgraf, ‘[c]learly the railroad did owe a duty of care to its customer and there was no need for any discussion of reasonable foreseeability in order to establish this conclusion’.93 While it is certainly true that, as a general doctrinal matter, the special relationship of common carrier

90 See above n 34. Recall that the Draft Third Restatement implicitly rejects the idea of a relational duty, and does not even employ the traditional terminology of a ‘duty of care’. Duty is envisaged, rather, simply as a general duty to conform one’s conduct to the standard of reasonable care. See above n 31. Thus the Restatement’s view about the relationship between duty and foreseeability is, strictly speaking, the view that foreseeability should not be taken into account in so-called ‘no-duty rulings’; it should only be taken into account in determinations by the trier of fact that the standard of reasonable care has been breached. 91 Goldberg and Zipursky, above n 6, at 1821–2, correctly recognise that Andrews J himself was not defending a relational or rights-based understanding of tort law, but was rather drawing on the instrumentalist views that even in 1928 were already strongly influential in the United States. Hurd and Moore, above n 44, mistakenly interpret Andrews J as defending the view that the duty of care is both relational and universal. See further nn 31 and 86 above. 92 Goldberg and Zipursky, above n 6, at 1833. It is an interesting question why Cardozo CJ did not take account of the fact that Mrs Palsgraf was the railroad’s customer. Possibly he wanted to take the opportunity to emphasise that, in all types of negligence cases, reasonable foreseeability is a necessary determinant of the duty of care. 93 ibid, at 1820.

Duty of Care in a Rights-Based Theory of Negligence

109

and passenger gives rise to a duty of care, the fact remains that this consideration played absolutely no role in Cardozo CJ’s reasoning in the case. He clearly would have reached exactly the same conclusion as he did even if Mrs Palsgraf had been a stranger and a bystander to the defendant, for example if she had happened to be injured as a result of flying debris that hit her while standing outside the station. Cardozo CJ held against Mrs Palsgraf because it was not reasonably foreseeable to the defendant that she was, when considered either as an individual or as a member of a class of persons, at risk of being physically harmed by its negligent action, and on this basis he concluded that the defendant did not owe her a duty of care.94 Goldberg and Zipursky argue that the significance of the fact that harm to Mrs Palsgraf was not foreseeable was not that she was not owed a duty of care, but rather that the defendant simply had not, as a matter of law, breached the duty it owed her.95 The ‘subtle’ duty issue raised by the case is, they claim, ‘why it should matter that the railroad did not breach its duty to Mrs Palsgraf, given that the railroad was already presumed to have breached the duty of care it owed to the other passenger’.96 The answer, they maintain, is simply that the notion of duty should be viewed relationally rather than non-relationally. Cardozo CJ took the first view of the matter, while Andrews J took the second. This is correct, so far as it goes, but it is simply irrelevant to the further question of whether or not reasonable foreseeability has a substantive role to play in determining the existence of a duty of care in the relational sense. In my view Cardozo CJ is most naturally interpreted as holding, in both MacPherson and Palsgraf, that the answer to this question is ‘Yes’. Cardozo CJ’s general view was that it is a necessary condition for the existence of a duty of care in negligence law that it be reasonably foreseeable to the defendant that the defendant’s contemplated action place

94 Other rights theorists go much further than Goldberg and Zipursky in rejecting a role for reasonable foreseeability in the determination of the duty of care. See, for example, DA Esper and GC Keating, ‘Putting “Duty” in Its Place: A Reply to Professors Goldberg and Zipursky’ (2008) 41 Loyola of Los Angeles Law Review 1225, 1243–55. Esper and Keating defend what they call a ‘weakly relational’ view of duty, which they contrast with the ‘strongly relational’ view that they attribute both to Goldberg and Zipursky and to Cardozo CJ in Palsgraf. According to Esper and Keating, a strongly relational view mistakenly views duty as ‘personal’, in the sense of involving an obligation that ‘runs from this named defendant to this named plaintiff ’; in their view, ‘particularizing’ duty in this way has the effect, among other things, of eroding the line between judge and jury: (at 1250–51). They reject outright Cardozo CJ’s view of duty of care, arguing that ‘duty is owed to everyone … and it is ordinarily triggered simply by acting in a way that poses a “reasonably foreseeable” risk of harm to anyone at all’: (at 1255). Esper and Keating’s defence of their view is complex, and cannot be dealt with properly here. The only point I would make for present purposes is that it is a mistake to think that the difference between a duty of care that is relational and universal and one that is relational but non-universal is that in the latter case, but not the former, the duty is ‘personal’. Rights and duties in the relational sense are always personal, but it is sufficient to establish this that the plaintiff belonged to a class of persons who were placed at risk by the defendant’s conduct. Neither Cardozo CJ nor Goldberg and Zipursky can be plausibly interpreted as holding the view that duty of care must run from a named defendant to a named plaintiff. 95 Goldberg and Zipursky, above n 6, at 1819–20 and 1821–4. 96 ibid, at 1820.

110

Stephen Perry

the plaintiff, either individually or as a member of a class of persons, at risk of suffering harm. In ordinary misfeasance cases involving physical harm between strangers, reasonable foreseeability will generally be not just a necessary, but also a sufficient condition for the existence of a duty of care. In cases involving some special feature—for example, an affirmative rather than a negative duty, or the fact that the harm was of a non-standard kind, such as economic loss or emotional distress—reasonable foreseeability of harm to the plaintiff will still be a necessary condition for the existence of a duty of care, but it will not generally be a sufficient condition. For example, there is, traditionally, no general right in negligence law not to be caused pure economic loss. The fact that it was reasonably foreseeable to the defendant that the plaintiff was at risk of suffering such loss is thus not sufficient to establish a duty of care; some further factor, such as the fact that the defendant gave the plaintiff an undertaking of an appropriate kind, is required.97 In English and Commonwealth law, it has been clear since Donoghue v Stevenson that reasonable foreseeability is always a necessary condition for the existence of a duty of care, and subsequent case law has emphatically confirmed that answer. As was noted at the end of section II above, the clear tendency in modern English and Commonwealth law, as evidenced by the decision of the Privy Council in Yuen Kun Yeu and of the Supreme Court of Canada in Cooper v Hobart, has been to add further restrictions, generally of a rights-based character, to the class of persons to whom a duty of care is owed, both in particular cases and in categories of cases.98 But allegiance to the basic requirement that reasonable foreseeability is always a necessary condition of the existence of a duty of care has remained steadfast. It would be exceedingly odd if an appropriately rights-based interpretation of American law differed so starkly on such a fundamental point from English and Commonwealth law, which, as noted in section II, has generally continued to lend itself much more straightforwardly to interpretation in rights-based terms. Doctrine in American tort law is complicated by the fact that civil cases in the United States are still generally tried to juries, and almost certainly one reason that many commentators support a shift of the role of reasonable foreseeability from the duty of care to the standard of care or to the element of proximate cause is a common belief, often shared by proponents of quite different theoretical standpoints, that the foreseeability question should be decided by juries and not by judges.99 This belief may well be correct, but it does not settle the answer

97

See eg Benson, above n 36; Perry, above n 40, at 262–302; Stevens, above n 42. Goldberg and Zipursky of course do not deny that such factors as ‘expectations, representations, reliance, and … property … interests’, to borrow the language of McLachlin CJ in Cooper v Hobart, play a role in determining the precise character of the duty of care both in particular cases and in categories of cases. See their helpful discussion of such issues in Goldberg and Zipursky, above n 6, at 1825–36. 99 This is the view explicitly adopted in the Third Restatement: § 7, comment j. The very complicated set of questions concerning the proper role of judge and jury in negligence cases, including in particular the proper delineation of the duty element, is illuminatingly debated in the following two 98

Duty of Care in a Rights-Based Theory of Negligence

111

to the following question: As a theoretical matter, is reasonable foreseeability an appropriate consideration to be taken into account in determining, to borrow the language of McLachlin CJ in Cooper v Hobart, whether ‘the circumstances of the relationship inhering between the plaintiff and the defendant are of such a nature that the defendant may be said to be under an obligation to be mindful of the plaintiff ’s legitimate interests’?100 After all, even if the determination of reasonable foreseeability should generally fall to juries, juries can be charged conditionally. Why might one think that reasonable foreseeability should not only play a role, but play the central role, in determining the existence of a duty of care in negligence law? In Donoghue v Stevenson Lord Atkin famously wrote that I owe a duty of care to persons who are so closely and directly affected by my act that I ought reasonably to have them in contemplation as being so affected when I am directing my mind to the acts or omissions which are called in question.101

He presented this proposition neither as self-evident nor as following from received legal doctrine, but rather as an answer to the abstract question: ‘Who, then, in law is my neighbour?’ The idea that we are neighbours not just in a temporal or spatial sense, but in a moral sense as well, is a powerful metaphor for the relational conception of rights and duties. As Lord Atkin recognised, the morality of tort law is not the morality of love but rather the morality of rights, and he was therefore correct to emphasise that the lawyer’s question ‘receives a restricted reply’. As moral neighbours, we do not owe one another a universal duty not to cause one another harm, and that is so even if harm is reasonably foreseeable to someone other than the person actually harmed. I only owe you a duty of care if you are among the persons foreseeably placed at risk by my action. This answer is the correct one because unless I can foresee the possibility that I might injure you, there is no way that I can take steps to avoid injuring you; to impose a duty on me even in the absence of a capacity to avoid injuring you would not be a proper expression of our general relationship as moral

articles: DA Esper and GC Keating, ‘Abusing “Duty” ’ (2006) 79 Southern California Law Review 265; and JCP Goldberg and B Zipursky, ‘Shielding Duty’ (2006) 79 Southern California Law Review 329. The debate is not easily summarised, and both sets of authors make a number of insightful points. I think that Esper and Keating are correct in observing that at least sometimes ‘[w]hen all the risks of physical injury created by some conduct are so remote that they are not reasonably foreseeable, no tort duty of care is owed to anyone’: Esper and Keating (at 283–4). Elsewhere, however, Esper and Keating appear to reject this view: see Esper and Keating, above n 94. In the latter article, they defend the view that Cardozo CJ was mistaken to make reasonable foreseeability a central determinant of the duty of care, and part of the reason for this is that doing so ‘erodes the line between judge and jury’: (at 1250). In their view reasonable foreseeability is most appropriately taken into account at the stage of proximate cause, and most American casebook editors are therefore correct to treat Palsgraf as a proximate cause case and not a duty of care case: (at 1253–5 and 1261–4). 100 101

Above n 38, at 552. Donoghue v Stevenson [1932] AC 562 (HL) 580.

112

Stephen Perry

neighbours.102 It was Holmes, in a brilliant passage in The Common Law, who gave most eloquent expression to this thought: The requirement of an act is the requirement that the defendant have made a choice. But the only possible purpose of introducing this moral element is to make the power of avoiding the evil complained of a condition of liability. There is no such power where the evil cannot be foreseen.103

There is a particular irony to the fact that this powerful formulation of the moral centrality of foreseeability to a relational conception of tort law should have been offered by the most influential architect of the prevailing view in the American legal academy that tort law should be understood in instrumentalist terms.

102 I have tried to show at greater length elsewhere that possession of a capacity to avoid a certain type of outcome is a necessary moral condition of the existence of a relational duty to avoid bringing about an instance of that outcome: see Perry, ‘Responsibility for Outcomes, Risk, and the Law of Torts’, above n 51, at 91–108. 103 OW Holmes and M Howe (ed), The Common Law (Cambridge MA, Belknap Press, 1963) 77.

5 The Rights of Private Law STEPHEN A SMITH

I. Introduction The title of this collection will no doubt raise eyebrows. It appears to assume what many regard as heresy—namely that private law has instrumental aims. Even instrumentalists may take umbrage. The most important instrumental theory of private law (the economic theory) asserts that private law has, or at least should have, just one goal, not many. It is indisputable, however, that private law has rights, and that there are many of them. There are private law rights to the performance of contracts, to the return of money paid by mistake, to the payment of damages for losses arising from a wrong, and to the reasonable enjoyment of property, to name only a few. The question—for both instrumentalists and their critics—is how to explain these rights. Broadly speaking, three kinds of explanations are possible. ‘Monist’ explanations suppose that the various private law rights all derive from a single goal, value, right, or principle (hereafter ‘value’). A clear example (and, if this essay’s arguments are correct, probably the only example) is the above-mentioned economic theory, which argues that the entirety of private law is grounded in the value of maximising welfare.1 The second kind of explanation, a ‘sceptical’ explanation, supposes that the various private law rights can only be explained (insofar as they can be explained at all) by a myriad of contrasting and conflicting values.2 In the sceptics’ view, any attempt to impose a rational order on private law, much less a monist order, is bound to fail. The third and final category comprises pluralist explanations. Like the sceptic, the pluralist accepts that private law cannot be explained by a single master value. But unlike the sceptic, the pluralist sees an

1 It is difficult today to find scholars willing to defend a full-blown economic theory of private law, but in his earlier writings Richard Posner came very close: see, eg, Economic Analysis of Law, 1st edn (Boston MA, Little Brown & Co, 1973). It has recently been argued that, whatever the foundations of the positive law, private law should be concerned exclusively with welfare maximisation: see L Kaplow and S Shavell, ‘Fairness v Welfare’ (2001) 114 Harvard Law Review 961. 2 This is broadly the view of private law defended in SM Waddams, Dimensions of Private Law: Categories and Concepts in Anglo-American Legal Reasoning (New York, Cambridge University Press, 2003).

114

Stephen A Smith

order in the relationship between the different values that underlie private law. According to the pluralist, different private law rights address qualitatively different kinds of questions and thus are grounded in different values. The values underlying private law are not, therefore, in tension with one another. In the pluralist’s view, the judges’ task in a difficult case is neither to balance conflicting values, nor to search for a unifying master value; it is to identify the value that is appropriate for the specific issue at hand. This essay has two main aims. The first is to show that the most important non-instrumental theory of private law—the corrective justice theory—is compatible with (though it does not require) a pluralist account of private law. More specifically, it will be argued that a corrective justice theory is compatible with two kinds of pluralist accounts of private law: (a) an account that holds that certain parts of private law cannot be explained on the basis of corrective justice (‘external pluralism’); and (b) an account that holds that the parts of private law that corrective justice can explain are grounded in qualitatively different kinds of corrective justice values (‘internal pluralism’). The second aim is to demonstrate that any plausible corrective justice account of actual private law, at least as private law is conventionally understood in common law jurisdictions, must be pluralist in both these senses. These arguments will not prove that the best account of private law is a pluralist account, much less that it is a corrective justice version of such an account. Those conclusions follow only if it is already accepted that private law is best explained, at least in general terms, on the basis of corrective justice. Many scholars hold just this view. It is hoped, however, that the essay’s arguments will be of interest to other scholars as well. The most important division within private law scholarship is not between instrumentalists and non-instrumentalists; it is between those who think it is possible to make sense of—to render intelligible—private law (the ‘believers’)3 and those who think this is impossible (the ‘sceptics’). Insofar as it is assumed that the only alternative to scepticism is monism, this division seems impossible to bridge. There is little, if any, common ground between the monists and the sceptics. Pluralism stands between monism and scepticism. The possibility of a pluralist account of private law opens up the possibility of common ground between the believers and their strongest critics.

II. Corrective Justice and Pluralism Corrective justice is usually described in terms of a duty to repair or correct a wrong or injustice: ‘Corrective justice theory is based on a simple and elegant idea: when one person has been wrongfully injured by another, the injurer must

3

Both economic theorists and corrective justice theorists fall within this first category.

Duty of Care in Rights-Based Theory TheA Rights of Private Law of Negligence

115

make the injured party whole’.4 This way of describing corrective justice has led some commentators to conclude that the concept has nothing to say about non-rectificatory (or ‘non-corrective’) duties.5 All that corrective justice can tell us, they believe, is what should happen when a wrong or injustice has occurred; it cannot tell us what counts as a wrong or injustice. If this is correct, then the concept of corrective justice can at most explain only a small part of private law. Contractual duties, duties not to trespass, duties not to cause a nuisance, duties not to defame, and many other private law duties have nothing to do with rectifying wrongs or injustices.6 Unsurprisingly, few of the scholars who currently defend corrective justice accounts of private law understand the concept’s scope in this way. Rather, what they believe, in broad terms, is that corrective justice can explain both duties to rectify wrongs and duties not to commit wrongs. Specifically, they believe that the idea of rectification contains within it, or at least is predicated upon, a particular understanding of wrongs and injustices and thus of the kinds of non-rectificatory duties the law should uphold.7 In the words of the best-known defender of corrective justice accounts of private law, Ernest Weinrib, ‘[c]orrective justice not only rectifies injustices in transactions; by structuring the justificatory considerations relevant to transactions, it is also regulative of the notion of injustice that is applicable to them’.8 In this view, the central features of corrective justice accounts of rectificatory duties (ie, that rights and duties are correlative, that they express a concept of equality between persons, and that they are grounded in the agency or interest of the parties rather than a broader social goal or value) apply equally to legal duties to perform contracts, not to cause injuries, not to trespass, and so on.9 Corrective justice can thus provide an account, at least in broad terms, of the kinds of duties, rectificatory or not, that we owe other individuals in their capacity as individuals. In Allan Beever’s words, ‘ “corrective justice” refers to an area of interpersonal morality 4 B Zipursky, ‘Civil Recourse, Not Corrective Justice’ (2003) 91 Georgetown Law Journal 695, 695; see also EJ Weinrib, ‘Corrective Justice in a Nutshell’ (2002) 52 University of Toronto Law Journal 349: ‘Corrective Justice is the idea that liability rectifies the injustice inflicted by one person on another’ (at 349); M Stone, ‘On the Idea of Private Law’ (1996) 9 Canadian Journal of Law & Jurisprudence 235: corrective justice is the idea of ‘one person [being] responsible for the harmful effects of her actions on another’ (at 253); and JL Coleman, The Practice of Principle (New York, Oxford University Press, 2001): corrective justice is the idea that ‘individuals who are responsible for the wrongful losses of others have a duty to repair the losses’ (at 4). 5 See, eg, P Cane, ‘Distributive Justice and Tort Law’ (2001) New Zealand Law Review 401. 6 Duties to reverse unjust enrichments are omitted from this list because while such duties are not duties to rectify a wrong they can plausibly be understood as duties to rectify an injustice. I briefly discuss this view in section IV.C below. 7 See, eg, EJ Weinrib, The Idea of Private Law (Cambridge MA, Harvard University Press, 1995) especially 133–44; JL Coleman, The Practice of Principle, above n 4, at 34; RW Wright, ‘Right, Justice, and Tort Law’ in DG Owen (ed), Philosophical Foundations of Tort Law (Oxford, Oxford University Press, 1995) 159. A Beever provides a clear explanation of the broader view and of its roots in Aristotle and Aquinas in Rediscovering the Law of Negligence (Oxford, Hart Publishing, 2007) 41–69. 8 Weinrib, above n 4, 351. 9 Thus Weinrib notes that within this framework rights cannot ‘be understood, for instance, simply as bundles of welfare’ because welfare is not correlatively structured: above n 4, at 353.

116

Stephen A Smith

that both defines [the] rights persons possess against each other as individuals and elucidates how one should respond to violations of those rights’.10 As Beever acknowledges, this understanding of ‘corrective justice’ might better be labelled (following Aquinas) ‘commutative’ justice or simply ‘interpersonal’ justice. This is the conception of corrective justice examined in this essay. The idea that corrective justice has nothing to say about non-rectificatory duties can therefore be set aside. The concept is, however, limited in two other respects. First, corrective justice does not provide anything like a general theory of morality, legal or otherwise. As we have just seen, corrective justice is concerned exclusively with interpersonal morality, that is to say, with the area of morality concerned with how individuals qua individuals (and not as members of a wider community or as representatives of the state) should act towards each other.11 It follows that constitutional law, tax law, international law and many other legal fields are not amenable, even in principle, to corrective justice explanations. Corrective justice theories are not meant to explain how the state should act towards its citizens, much less towards other states. The contrast with economic and most other monist theories is stark. In principle, the entirety of the law is amenable to an economic explanation (and in practice its entirety has been explained in just this way). Corrective justice theories do not purport, even in theory, to be capable of explaining the entirety of the law. It follows from this first limitation that the best corrective justice explanation of private law might be what I have described as an externally pluralist explanation. It might be an explanation, in other words, that leaves parts of private law to be explained using non-corrective justice values. There is no a priori reason to assume that private law is constituted exclusively by the kinds of rules that are amenable, even if only in principle, to a corrective justice explanation. Of course, if private law is defined in advance as the law that governs legal duties owed by individuals to other individuals, then this possibility can be ruled out. The definition equates private law with the kinds of rules that corrective justice is in principle capable of explaining. Yet, while such a definition might be an appropriate conclusion to an exploration of corrective justice and private law, it cannot be the starting point. Any attempt to understand private law must begin with the conventional understanding of private law’s subject-matter. That conventional understanding does not adopt the above criterion or any other substantive criterion for identifying private law; it simply identifies private law with the rules conventionally discussed in books on contract, tort, unjust enrichment, trusts, remedies, etc. In theory, it is entirely possible that some of those rules are the kinds of rules that corrective justice accounts are not intended, even in principle, to explain.

10

Beever, above n 7, at 56. Corrective justice is also unconcerned with interpersonal morality insofar as it extends to what are sometimes called ‘supererogatory duties’ or ‘duties of virtue’, such as duties to give to charity. 11

Duty of Care in Rights-Based Theory TheA Rights of Private Law of Negligence

117

The second way in which corrective justice is limited as an explanatory principle is that it only provides an abstract framework for thinking about the nature of non-rectificatory rights and duties. The concept of corrective justice tells us that wrongdoers have specific duties to rectify the wrongful injuries that they have caused. It also tells us that wrongful injuries arise from breaching certain kinds of rights (ie, rights that are correlative to duties owed by other individuals and which are grounded in the agency or interests of those individuals rather than the community or in any other social value or goal). It does not tell us what the latter set of rights and duties actually are. It (merely) tells us that, whatever they are, they must be consistent with the aforementioned constraints. In The Idea of Private Law,12 Ernest Weinrib turns to Kant for an account of our non-rectificatory rights and duties. In principle, other accounts are possible. As just noted, the rights that corrective justice explains must have a certain character or structure; they can be neither group or social rights nor rights to a share of communal resources. To this extent, they must resemble in broad outline the kinds of rights that Kant believes the law protects. Yet within that broad outline different kinds of rights are possible. The concept of corrective justice does not tell us, for example, whether a promise or a trust or a will should give rise to legal rights; nor does it tell us whether there ought to be legal rights not to be defamed or misled or injured. Most importantly, the concept of corrective justice does not tell us whether our legal rights are derived from a single master value or from heterogeneous foundations. It follows that a corrective justice account of private law is consistent in principle with what I have called internal pluralism. It is consistent, in other words, with the idea that the values that explain our non-rectificatory legal rights are of qualitatively different kinds. It is also consistent with the possibility that those values are different from the value or values that explain rectificatory duties. The conclusion to be drawn from these observations is that corrective justice is both limited in range (allowing for external pluralism) and consistent with the possibility that the rights and duties it can explain are grounded in different values (allowing for internal pluralism). This does not mean that corrective justice theories of actual private law must be pluralist in either of these senses. The actual private law might be comprised exclusively of the kinds of interpersonal rights and duties that corrective justice can in principle explain. It is also possible that these interpersonal rights and duties are all derived from the same value. It is possible, in other words, that the most persuasive corrective justice theory of the actual private law is a monist theory. The evidence, however, suggests otherwise. As I will show in the next section, in order to account for the actual private law in common law jurisdictions, a corrective justice explanation must be both externally and internally pluralist.

12

Weinrib, above n 7, at 84–113.

118

Stephen A Smith

III. External Pluralism: Ordinary Rights, Action Rights, and Court-Ordered Rights The legal rights discussed in books on property, tort, contract, unjust enrichment and other subjects conventionally regarded13 as falling within private law (eg, remedies, equity, trusts, company law) fall into three broad categories.

A. Ordinary Rights Ordinary legal rights are the legal rights that individuals going about their daily business hold against other individuals, going about their daily business. Examples include rights to the performance of a contract, to the exclusive possession of chattels, to the reasonable enjoyment of property, to the return of money paid by mistake, and to the payment of damages for consequential losses arising from a tort or breach of contract. These rights are ‘ordinary’ because they arise from everyday events or sets of facts (eg, making a promise, acquiring property, paying money by mistake, suffering an injury), rather than from an act of the courts,14 and because they are rights that one individual holds against another individual (the promisor, the recipient of a mistaken payment, etc) as opposed to the court or other state entity.15 When landowners tell trespassers to get off their property they are asserting or ‘standing upon’ their ordinary rights (in this case to exclusive possession of their property). The same is true of contracting parties when they ask their counterparts to perform their contractual promises and of mistaken payors when they ask mistaken payees to return a mistaken payment.

B. Action Rights The second category of rights, action rights, is comprised of the rights that individuals stand upon when they ask courts to make orders. The foundation of a typical civil suit is of course an allegation that the defendant has infringed or (more rarely)

13 The qualifier ‘conventionally’ is added because a possible rejoinder to this argument is that the rights in question are not true private law rights. I consider this rejoinder later. 14 The typical source or explanation for the generic rights that all of us have to the performance of contracts, to the return of money paid by mistake, and so on, is, of course, an act by a court (ie, a judgment) or a legislature (ie, legislation). This is true for each of the three kinds of rights discussed above. The distinguishing feature of ordinary rights is not the explanation of the generic right, but the explanation of the event or set of facts that gives rise to a particular individual possessing a specific instance of the generic right. Thus, while we all have the generic right to the performance of contacts made with us (because a court, legislator, or other source of law has said so), a specific claimant’s right to the performance of a specific contract arises if, and only if, a certain kind of ordinary event (such as making an agreement) has occurred. 15 ‘Individual’ here includes artificial persons, such as corporations, as well as state entities when such entities are regarded in law as ordinary individuals.

Duty of Care in Rights-Based Theory TheA Rights of Private Law of Negligence

119

is about to infringe one of the claimant’s ordinary rights.16 Yet claimants do not go to court merely to communicate to defendants or to the court that the defendant has done something wrong. Actions for declaratory relief aside, a lawsuit is fundamentally a request by a claimant that the court do something—namely to make an order against the defendant. Thus, a claimant who has proven a breach of contract might ask the court to order the defendant to perform the contract or to pay a sum of money sufficient to obtain alternative performance. Alternatively, or additionally, the claimant might request that the defendant be ordered to pay a sum equal to the consequential losses suffered as a result of the breach. In yet other cases, the claimant might simply request that the defendant be ordered to pay a nominal sum. It is arguable that some court orders are available only as a matter of the court’s discretion. Requests for specific relief (eg, specific performance, injunctions) are traditionally described in this way. This description is controversial,17 but insofar as it is accurate, claimants who request specific relief are not strictly asserting a right. They are merely asking the court to exercise its discretion in their favour. Many court orders, however, are governed by fixed rules. Orders to pay a debt or damages or to return money paid by mistake are examples. Where claimants request such orders they are asserting a right to the order, and where such orders are made they are made because of that right. Such orders are available ‘as of right’ if the relevant debt, loss, or unjust enrichment has been established. The label ‘action rights’ is, therefore, an abbreviation of ‘cause-of-action rights’. ‘Cause of action’ is the label conventionally given to the set of facts that a claimant must establish in court to obtain a court order.18 Like ordinary rights, action rights arise from everyday events. The right to a court order that a debt be paid arises because a debt was incurred and not satisfied; similarly, the right to an order that damages for consequential loss be paid arises because the claimant suffered consequential loss as a result of a tort or breach of contract. The difference between ordinary rights and action rights is that action rights are rights that individuals hold against courts rather than (as in the case of ordinary rights) other individuals. When claimants argue that they have a right to a particular ‘remedy’ from the court, they are arguing that the law requires that the court do something—namely, make an order against the defendant.

C. Court-Ordered Rights Assuming that the parties have not reached a settlement, a typical lawsuit ends with the court’s decision to make (or not make) a court order. If an order is made, 16 The exception is where the claimant is seeking a declaration. In these cases, the claimant asserts that her ordinary rights vis-à-vis the defendant are uncertain. 17 See, eg, A Burrows, Remedies for Torts and Breach of Contract, 2nd edn (London, Butterworths, 1994) 457. 18 ‘A cause of action is simply a factual situation the existence of which entitles one person to obtain from the court a remedy against another person’: Letang v Cooper [1965] 1 QB 232 (CA) 242–3 (Diplock LJ).

120

Stephen A Smith

it gives rise to a third kind of private law right—a ‘court-ordered right’. Courtordered rights usually (though not invariably) have a close relationship to the claimant’s ordinary and action rights. Thus, where a court order is made because the claimant had an action right to the order (eg, an order that the defendant pay a debt), the order merely gives effect to the action right. Further, where the action right is itself a right to an order that the defendant perform an ordinary duty (as is true of an action right to an order to pay a debt), the content of the court-ordered right is the same as the content of the claimant’s ordinary right (eg, a right to payment of the debt). The court order merely compels the defendant to do whatever he or she was already obliged to do. Nonetheless, even in these cases, claimants’ court-ordered rights are distinct from whatever action rights and ordinary rights they may have enjoyed. It is the order, not the action right or ordinary right that is the source of the court-ordered right. The legal effect of a court order is to extinguish whatever ordinary rights the claimant was relying on to obtain the order and to replace them with the rights articulated in the order. When claimants complain to the sheriff or bailiff that defendants subject to court orders have failed to do what they should have done, they stand on their court-ordered rights, not their ordinary rights. The law that governs court-ordered rights has two parts. As is true of all rights, there are rules that establish the existence and meaning of court-ordered rights. Nothing further needs to be said about these rules; there are few of them and they are not perceived, even in the conventional view, as part of private law. The second part, which is distinctive and important, comprises the rules that govern the availability and content of those orders. This second part incorporates the law of action rights because to the extent that a claimant has an action right to an order, the explanation for that order is the action right. Yet in theory it is possible—and in practice it almost certainly happens—that courts sometimes grant court orders in cases where the claimant has no action right or only a weak action right to the relevant court order. As already mentioned, according to the traditional view claimants have no right to specific relief orders; they can merely request that the court exercise its discretion to make a specific relief order in their favour. Insofar as this description is accurate, the court orders made on this basis cannot be explained simply by reference to the claimant’s action rights. Rather, they can only be explained (insofar as an explanation is possible at all) on the basis of those rules that give courts the discretion to make the orders in question. These are rules about court-ordered rights.

D. Ordinary Rights, Action Rights, Court-Ordered Rights, and the Scope of Corrective Justice The rules that govern ordinary rights are in principle amenable to being explained on the basis of corrective justice. Ordinary rights are rights that individuals enjoy against other individuals. As such, they appear, on the surface anyway, to be

Duty of Care in Rights-Based Theory TheA Rights of Private Law of Negligence

121

precisely the type of interpersonal rights that corrective justice theories are meant to explain. This does not prove that these rights are, in the end, best explained on the basis of corrective justice. The point is merely (though importantly) that these are the kind of legal rights that in principle corrective justice theories are meant to explain. The rules that govern action rights and court-ordered rights are different. Neither is amenable, even in principle, to a corrective justice explanation; they have the wrong structure. Both the law of action rights and the law of courtordered rights are comprised of rules that tell courts how they should act.19 Neither set of rules is inconsistent with corrective justice. Yet they lie outside corrective justice’s concerns. As we have seen, the concept of corrective justice is meant to explain how individuals should act towards one another. It is not meant to explain how the state, acting in its capacity as the state, is supposed to act. This conclusion, note, is consistent with believing that what courts should do is to order defendants to do what corrective justice requires. It is also consistent with believing that this is what courts actually do in practice. The point is that the concept of corrective justice does not explain why courts should or would order defendants to do what corrective justice demands. Corrective justice is a theory of how individuals should act, not courts. Defenders of corrective justice theories might respond to these observations by arguing that the law governing action rights and court-ordered rights is strictly speaking a part of public, not private, law. For the purposes of this essay, it does not particularly matter what position is taken on this issue. This is not an essay about the private/public divide. The essay’s aim is (merely) to show that any plausible corrective justice account of the body of law for which such accounts are conventionally assumed to provide an explanation must be a pluralist account. In the conventional understanding, private law includes the law governing action and court-ordered rights. In passing, however, it is worth mentioning that as a matter of first principles it is not obvious whether the law governing action rights and court-ordered rights is public or private.20 As we have seen, this law consists essentially of directions to courts, that is to say, of directions to state entities. This might be thought sufficient grounds for regarding them as public law rules.21 Yet, viewed from a different perspective, these rules are quintessentially private. Action rights arise from what are, by any definition, private law disputes. They are rights that individuals enjoy precisely because their ordinary rights were infringed or threatened by other individuals and precisely in order to protect or vindicate 19 As mentioned earlier, I ignore the part of the law of court-ordered rights dealing with the validity and interpretation of court orders. These rules are directed at citizens. 20 I discuss this issue in S Smith, ‘The Law of Damages: Rules for Citizens or Rules for Courts?’ in R Cunnington and D Saidov (eds), Contract Damages: Domestic and International Perspectives (Oxford, Hart Publishing, 2008). 21 See, eg, M Loughlin, The Idea of Public Law (Oxford, Oxford University Press, 2003) 153; G Slapper and D Kelly, The English Legal System, 6th edn (London, Routledge-Cavendish, 2006) 5; JF McEldowney, Public Law, 3rd edn (London, Sweet & Maxwell, 2002) 6.

122

Stephen A Smith

those rights. Further, when courts make orders, the result is to affirm or create a private law relationship between the parties. An order to pay damages compels the defendant (not the state) to pay the claimant (not the state) a sum of money. Thus, it is impossible to explain either action rights or court-ordered rights without at the same time explaining the ways in which they arise from, and in turn give rise to, rights and duties between individuals. So central is the concept of an action right to our understanding of a civil suit that some scholars regard it as the core of private law.22

E. Weinrib’s Explanation of Court Orders A different view of court orders and their relation to corrective justice is defended by Ernest Weinrib. Although Weinrib does not argue that corrective justice directly explains court orders, his theory of private law binds the two tightly together. For Weinrib, both corrective justice and court orders are explained and justified by Kant’s ‘doctrine of Right’.23 In Weinrib’s words, ‘Kantian right allows one to trace corrective justice back to its normative roots in self-determining agency and forward to the values and institutions of a coherent legal order’.24 Those institutions include courts that have the authority to make coercive orders. Weinrib provides what appear to be two Kantian-based explanations of court orders. In examining them I will assume that the idea of corrective justice can be explained, as Weinrib believes, on the basis of Kant’s doctrine of Right. The question I will focus on here is whether, as Weinrib further believes, this doctrine (as presented in Weinrib’s work25) can also explain the necessity of court orders. The first explanation of court orders, which is set out in Weinrib’s book The Idea of Private Law, presents court orders (and adjudication generally) as a necessary consequence of private law’s exclusive focus on the external manifestations of human behaviour. In Weinrib’s words: ‘Because acting out of virtue [ie, acting for the right motives] is irrelevant to the externality of the relationships governed by law, an external authority must be present to enforce upon the actor the external requirements of practical reason’.26 This argument appears to rest 22 See, especially, Zipursky, above n 4; B Zipursky, ‘Philosophy of Private Law’ in JL Coleman and SJ Shapiro (eds), The Oxford Handbook of Jurisprudence and Philosophy of Law (Oxford, Oxford University Press, 2002). 23 For this reason, it is arguably more accurate to describe Weinrib’s theory of private law as a Kantian theory than a corrective justice theory. 24 Weinrib, above n 7, at 114. Earlier in The Idea of Private Law Weinrib remarks that ‘Kant connects corrective justice to the institutions of a functioning legal order’: (at 84). 25 The discussion that follows focuses exclusively on the account of court orders that is presented in Weinrib’s writings. The account is explicitly based on Kant’s work, but I make no attempt to assess the accuracy of Weinrib’s interpretation. 26 Weinrib, above n 7, at 99. The same argument is repeated in other places in The Idea of Private Law: ‘Although the parties may spontaneously observe the requirements of right either by forbearing from wrong or, once a wrong has occurred, by making or extracting proper amends, these possibilities have no juridical standing since they presuppose in the parties an external virtue foreign to the externality of legal relations. Therefore, the public significance of wrong can be signalled only by the

Duty of Care in Rights-Based Theory TheA Rights of Private Law of Negligence

123

on two propositions. The first is that in the Kantian conception of private law all that matters, when assessing the legality of an action, is how the action appears from an external or objective perspective. Thus, when evaluating a claim in negligence it does not matter whether the defendant thought she was acting carefully and with full regard for the plaintiff ’s rights. All that matters is whether she did in fact act with proper care, etc; the defendant’s motivations are irrelevant. The derivation of this idea of ‘externality’ from the doctrine of Right involves a number of steps, but the proposition itself is relatively uncontroversial. Although not everyone would agree that motivations should never be relevant in private law, few would dispute that in general they are, and should be, irrelevant. Nothing further needs to be said about this first proposition. The second proposition—the idea that the externality of private law explains the need for court orders—is, however, far from straightforward. The only conclusion that appears to follow from the proposition that private law should be unconcerned with individuals’ motivations is that the tests for determining whether a legal right has been infringed should ignore motivations. The further question of what the state should do when a legal right has been infringed is an entirely different issue. To be sure, the externality of private law explains why court orders are not inconsistent with a Kantian-based moral explanation of private law. If acting morally required individuals to act from the right motives, then the practice of making court orders would pose a problem for any attempt to explain the law on moral grounds. Individuals often perform contracts, refrain from trespassing, and so on only because they fear a court order. The practice of making court orders does not pose a problem for Kantian legal theories because Kantian morality (or at least the part that is meant to explain private law) makes motivation irrelevant. All that this shows, however, is that making court orders is not inconsistent with Kantian morality. It does not provide a positive reason for court orders. At one place in The Idea of Private Law Weinrib appears to suggest that the link between the externality of law and coercive orders lies in the fact that only actions that are motivated by fear of external sanctions are consistent with legal rights. In describing the difference between the realms of law and ethics, he writes: ‘in law the actor responds to the prospect of external coercion, whereas in ethics the idea of duty itself motivates the action’.27 This cannot mean what it appears to mean. As Weinrib repeatedly observes, the courts do not care why we keep our contracts, etc; all they care about is whether we keep them. Needless to say, many, perhaps most, people keep their contracts, etc, not because they fear legal sanctions but because they believe that this is the right thing to do. availability of coercion that represents the external operation upon the parties of the concept of right’ (at 107); and ‘Since the point of right is to hold the external aspect of action to the external demands of practical reason, law must posit an external force capable of determining the actor’s will, that is, capable of acting as a deterrent’ (at 109). 27

Weinrib, above n 7, at 110.

124

Stephen A Smith

Finally, it is clear that Weinrib’s point about externality is not that court orders are necessary merely to induce bad people to act in conformity with the external standards by the law: ‘in Kantian legal theory, the presence of the judge is not merely a phenomenologically observed element of private law, but is conceptually necessitated by the mutual externality of the interacting agents’.28 For Weinrib, like Kant, court orders would be necessary even in a ‘world of angels’ (where everyone obeyed the law). In the end, it is not clear what the externality of private law has to do with court orders. Weinrib’s second explanation for court orders is set out primarily in two essays, each published after The Idea of Private Law, in which he discusses and defends a Kantian account of property rights.29 The explanation is complex, but the basic idea is that a judiciary with the authority to coercively enforce legal rights is required in order to make property rights legitimate. The full explanation proceeds in three steps. In the first step, Weinrib introduces the foundational Kantian idea of the doctrine of ‘Right’, which ‘requires that that one person’s actions be able to co-exist with the freedom of everyone under a universal law’.30 In lay terms, the doctrine of Right is roughly the idea that each individual’s freedom of choice, or more strictly, capacity for choice, should be equally protected from interference or domination by others. Every individual has the right to be his or her own master. This ‘universal’ doctrine contains within it the more specific idea that we all have, merely by virtue of being alive, an ‘innate right to freedom’. Innate right ‘consists of the independence of one’s actions from constraint by the actions of another, insofar as such independence is consistent with the freedom of everyone else’.31 In short, I have the right, merely by virtue of my existence, that others not interfere with my body or my bodily movements (except insofar as I am interfering with their body or bodily movements). The idea that each of us has an innate right to freedom is the basic explanation for legal rights to bodily security and integrity (eg, the right not to be assaulted). It can also explain why it is wrong for you to take something from me that I have in my physical possession, for example, an apple in my hand. To take the apple you would need to forcibly move my fingers, thus infringing my right to bodily security. This innate right to freedom does not explain, however, continuing rights to external objects—that is to say, property rights. So far as the innate right to freedom is concerned each of us has an equal right to take and use external objects. Others cannot move or manipulate our bodies to take those objects from us. Yet once we have physically released the objects, others are physically free to take them. 28 ibid, at 84. Elsewhere Weinrib writes that the ‘necessity of such an authority is a conceptual one’ (at 99). 29 The explanation also applies to contractual rights, but for simplicity I will focus (as Weinrib does) on property rights. 30 EJ Weinrib, ‘Poverty and Property in Kant’s System of Rights’ (2003) 78 Notre Dame Law Review 795, 803. 31 ibid, at 803.

Duty of Care in Rights-Based Theory TheA Rights of Private Law of Negligence

125

It is at the second stage of Weinrib’s account that the idea of property rights is introduced, though at this point property rights exist only in what he describes as a ‘provisional’ sense. They are provisional because although from one perspective they are consistent with and indeed required by the doctrine of Right, according to another they are inconsistent with the doctrine. They are consistent, Weinrib says, because it follows from the doctrine of Right that, ‘everything that is distinct from that to which one person has an innate right is available for another’s use as an external object of choice’.32 Further, property rights are required by the doctrine of Right: because one’s capacity for choice cannot, consistently with one’s freedom, be limited to what one actually chooses (in this context, what one holds in one’s physical possession), then one must postulate rights that maintain things in a state of being possible objects of choice, even when not in one’s possession.33

At the same time, property rights are inconsistent with the doctrine of Right because such rights come into being through unilateral acts, for example, by an individual taking first possession of land. This is inconsistent because it means that one individual can unilaterally impose obligations on other individuals (such as obligations to respect an owner’s newly created property rights). As Weinrib explains, property-right creation ‘should be beyond the rightful power of one person’s unilateral will, for it is inconsistent with innate equality of all that the acquirer should be able to subordinate others to his or her purposes’.34 In the third step, this tension in the idea of property rights is resolved, Weinrib explains, through the introduction of a state that provides ‘duly authorized institutions of adjudication and enforcement’.35 With the introduction of the state, the coercion that secures each person’s rights is no longer private but emanates from a public lawful regime … this notional union of all wills transforms the external acquisition of unowned things from a merely unilateral act on the part of the acquirer to an omnilateral act, to which everyone as possible owners of property implicitly consents’.36

Crucially, the introduction of courts with the authority to make coercive orders is, according to Weinrib, required by the concept of a property right that was introduced in the second stage of the argument. It is everyone’s duty, Weinrib says, to bring into being a state with the authority to make property-right creation legitimate: ‘all are under a duty to create the conditions under which those proprietary entitlements [ie, those supposed in the second phase] are fully rightful’.37 Thus, the introduction of courts with the authority to make coercive orders

32 33 34 35 36 37

ibid, at 806. ibid, at 806. Weinrib, above n 30, at 807. ibid, at 808. ibid, at 809. ibid, at 810.

126

Stephen A Smith

is, for Weinrib, both required by, and part of the justification for, the concept of a property right. Even for readers broadly sympathetic to Kantian legal theory, this explanation raises a number of questions. Perhaps the first is why it is thought to be necessary that individuals be able to acquire property rights. If ‘freedom’ consists in leading a fulfilling life, achieving valuable ends, having a wide range of choices, controlling the shape of one’s life, or something similar, then the connection to property rights is clear. The institution of property rights allows us to do many things we could not otherwise do. Yet the Kantian notion of freedom that Weinrib defends is strictly negative; all that matters is that others do not unilaterally infringe my capacity for choice. Innate right, as we saw earlier, ‘consists of the independence of one’s actions from constraint by the actions of another’.38 Freedom in this negative sense might be consistent with property rights, assuming that such rights can be otherwise justified, but it is not obvious why it requires property rights. Weinrib’s explanation, recall, is that because one’s capacity for choice cannot, consistently with one’s freedom, be limited to what one actually chooses (in this context, what one holds in one’s physical possession) … one must postulate rights that maintain things in a state of being possible objects of choice, even when not in one’s possession.39

The explanation is difficult to understand. An inability to own external objects does not preclude a capacity to make choices that involve such objects. It just means that such choices will sometimes be frustrated. Admittedly, if I already have a property right in something, say a bicycle, which you then tie a rope around, you will have interfered with my freedom just as if you had tied a rope around my legs. Yet this is to assume the property right that we are trying to prove. If we start from a situation of non-ownership, then there is no interference. In any event, even if it is accepted that the capacity for making choices is greater in a world with property rights than in one without them, it is not clear why, from Weinrib’s perspective, we should be concerned about this. Freedom for Weinrib is a matter of independence from others, not a matter of increasing one’s capacity to make choices. All that the second step in Weinrib’s explanation appears to have shown is that if property rights could (on other grounds) legitimately be created, then their existence would be consistent with the idea of equal freedom.40

38

ibid, at 803. ibid, at 806. 40 It is interesting, therefore, that Weinrib often describes the relationship between property and freedom in terms of consistency rather than necessity: ‘Because property is consistent with the freedom of all, it is rightly secured and protected by the law’s coercive powers’ (ibid, at 801). In an earlier passage Weinrib describes the link in even weaker terms: ‘having something as one’s own, although not analytically contained in the universal principle of Right, marks a connection to external things that matches the capacity for choice characteristic of self-determining action’ (at 802). As explained above, these sentences accurately describe the relationship between property and freedom. From Weinrib’s perspective, however, they are problematic because they disavow the only positive reason he offers for state recognition of property rights (ie, that they are required for individuals to be free). The third step 39

Duty of Care in Rights-Based Theory TheA Rights of Private Law of Negligence

127

But let us accept for the sake of argument that the possibility of property rights is required, at least provisionally, by the idea of equal freedom. As Weinrib makes clear, there remains the problem that property rights are brought into existence by unilateral acts and as such appear to contradict the Kantian principle that no one may unilaterally place obligations on others. In Weinrib’s story, this problem disappears with the introduction of the state: ‘this notional union of all wills transforms the external acquisition of unowned things from a merely unilateral act on the part of the acquirer to an omni-lateral act, to which everyone as possible owners of property implicitly consents’.41 Again, it is difficult to understand the argument. If everyone actually, and not merely implicitly, consents to a stateauthorised regime of property rights, then such authorisation is consistent with Kantian morality. Yet absent such consent, which clearly does not exist, it is hard to see how a regime of state-authorised property rights can be consistent with the foundational idea that individuals have a right to be free from interference or domination by others. However property rights are described and whatever ‘postulates’ are invoked in their justification, the inescapable reality is that their creation is a unilateral act that results in new obligations being imposed on others. A ‘notional’ union of the wills is just that, notional. Implicit consent is not consent. This is not to deny that perfectly respectable justifications can be given for state-authorised regimes of property rights. For instance, the state action may be necessary, even in a world of angels, to establish precisely what kinds of acts should be recognised as property-right creating acts. None of these arguments, however, resolve the problem, for Weinrib’s account, that the recognition of property rights allows individuals to unilaterally impose obligations on others. Finally, even if it is accepted that the above argument explains the necessity of a property-right authorising legal regime, there remains the question of why it is thought to be necessary that courts have the authority to coercively enforce property rights.42 As already mentioned, Weinrib expressly disavows any argument based on the need to induce recalcitrant individuals to respect others’ rights. Coercive orders are a necessity even in a world where everyone acts rightly. The only distinct argument that is given for coercive orders (aside from the ‘externality’ idea discussed earlier) is that they are required for ‘assurance’. Weinrib writes: ‘[b]ecause no one is obligated to respect the entitlements of others unless assured that everyone else will do so, the state’s coercive power is required to guarantee what everyone owns’.43 Leaving aside the fact that the state’s coercive power in his argument is not meant to provide such a positive reason, but merely to rebut the objection that property-right creation is a unilateral act. 41

Weinrib, above n 30, at 809. A further objection is that the argument does not appear to explain court orders that protect innate rights. 43 Above n 30, at 809. See also EJ Weinrib, ‘Deterrence and Corrective Justice’ (2002) 50 University of California at Los Angeles Law Review 621, 637: ‘for [property] rights to be secure, however, they must be guaranteed by state institutions which give them determinate form and provide the apparatus of coercive enforcement’. Weinrib also makes the same argument, albeit only in passing, in The 42

128

Stephen A Smith

provides no guarantee that others will actually respect my rights, the suggestion that I have no obligation to respect others’ rights simply because there is a chance that they will not respect mine sounds very odd. Surely it is not acceptable for me to take others’ property merely because there is a risk, and a merely theoretical risk at that (‘theoretical’ because the argument is meant to apply in a world of angels), that they might take my property. Taking others’ property only makes it more likely that someone will try to take mine in the future. More to the point, how does refraining from interfering with others’ property make me subservient to them?44 Why is my independence not fully protected by taking whatever steps are necessary (merely) to protect my property rights. Admittedly, if others’ entitlements are genuine entitlements (as Weinrib implies when he says that the state’s powers are required to guarantee what everyone ‘owns’), then I ought to respect them, at least insofar as this is consistent with my rights to self-defence, etc. This obligation, however, would exist regardless of whether these rights were recognised and enforced by the state. On the other hand, if the entitlements that the introduction of the state is meant to secure are not genuine entitlements (as the existence of the argument’s third step implies), then it is not clear why the state or anyone else should worry about them. This is not to deny, again, that coercive state-enforcement of property rights may be justified on various grounds, such as the need to force the recalcitrant to respect others’ rights. It is just to deny that those grounds can be explained, as Weinrib attempts to do, as logical consequences of the idea of independence from others. In the end, it is unclear how Weinrib’s Kantian-inspired argument is meant to explain property rights, much less the authority of courts to make coercive orders. This conclusion should not be surprising. The normative ideal on which the entire argument rests, namely freedom in the sense of independence from the interference or domination of others, is essentially negative. Yet the legal rules that Weinrib is attempting to explain are power-conferring rules (the power to create property rights, the power to make coercive orders). Like other powerconferring rules, they are essentially positive: they confer powers to create new rights and duties. As a matter of basic logic, it is not clear how such powers could ever be derived from rights to non-interference. The powers to create property

Idea of Private Law: ‘since it is inconsistent with the equality of free wills that one should refrain from wronging someone who might not exercise an equal restraint, everyone is to be coerced into a public regime of law as a way of guaranteeing in advance the equal security of everyone’s freedom’ (above n 30, at 108). This statement appears in a discussion of rights in general, not merely rights to external things, thus suggesting that it is acceptable in the state of nature not just to take other’s property but to hit, maim or even kill others merely if there is a risk that they might trivially injure you, for example by bumping into you in a crowd. 44 According to the Kantian account of rights that Weinrib defends, right-holders always have the authority to coerce others in defence of their rights (see above n 30, at 807). This authority can arguably be derived from the idea of independence (because self-defence is nothing more than preventing others from infringing one’s independence), but it does not explain why individuals should be able to engage in non-defensive coercion. Nor does it explain why the state should enjoy a similar authority.

Duty of Care in Rights-Based Theory TheA Rights of Private Law of Negligence

129

and make court orders that Weinrib is attempting to explain are, in essence, powers to interfere.

F. The Rubber-Stamp View of Court Orders The conclusion that the law governing action rights and court-ordered rights is not open, even in principle, to a corrective justice explanation highlights an important limit on the potential scope of such an explanation. The practical significance of this conclusion, however, depends on the extent to which private law (as conventionally understood) is concerned with action rights and court-ordered rights. In theory, the law of action rights and court-ordered rights might be very small—just two rules. This would be the case if the ‘rubber-stamp’ view of court orders were correct. According to this view, every court order can be explained by an action right, and every action right is a right to an order confirming an ordinary right. If the rubber-stamp view is correct, the law of court orders and the law of actions would each be constituted by just one rule. The law of court orders would be the rule that courts will make orders if and only if claimants have action rights to such orders. The law of actions, in turn, would be the rule that claimants whose ordinary rights have been infringed or threatened have the right to orders compelling defendants to act in accordance with their ordinary rights (assuming this is still possible). Needless to say, these would be important rules. An account of the law governing private disputes would need to explain them. They would not, however, take up much space in textbooks.

G. Rules Governing the Availability of Court Orders It is difficult to tell how much support the rubber-stamp view enjoys amongst lawyers and academics.45 But it finds little support in the positive law. There are two groups of rules that are flatly inconsistent with the rubber-stamp view. The first is comprised of the many rules that set conditions on the availability of court orders. Included in this group are the rules governing limitation periods and the rules that establish the pre-conditions for obtaining specific relief orders, such as specific

45 I am not aware of any author who has explicitly defended the rubber-stamp view, but the almost complete absence in the common law literature of any suggestion that there exists a distinct body of law dealing with action rights and court-ordered rights suggests this view is widely, if perhaps unconsciously, assumed. Peter Birks, who reflected as deeply as anyone on the structure of private law, explicitly recognised court orders as a formal source of rights: ‘Rights, Wrongs, and Remedies’ (2000) 20 Oxford Journal of Legal Studies 1, 27. Yet Birks never acknowledged that the courts were doing anything other than rubber-stamping ordinary rights. His description of legal rights in the Introduction to English Private Law, 1st edn (Oxford, Oxford University Press, 2000) xxxvi, as rights that ‘people are able to realise in courts’ leaves open the possibility, in theory anyway, that a right might be ‘realised’ other than by its enforcement, but the more natural interpretation is that Birks regarded legal rights as rights that courts will enforce directly.

130

Stephen A Smith

performance, injunctions, and constructive trusts.46 These rules describe a range of situations in which courts are allowed or (in some cases) required to refuse a claimant’s request that they order the defendant to perform an ordinary duty that the defendant has thus far failed to perform. Thus, courts regularly refuse to order defendants to perform valid non-monetary contractual obligations. There are entire books devoted to the law governing the availability of specific performance and other forms of specific relief. Even orders available ‘as of right’, such as orders to pay an outstanding debt or money otherwise due, are typically available only if requested within a certain time period.47 The conclusion to be drawn from these and other rules is that, even if one assumes that court orders always enforce ordinary rights, there exists a large body of law that governs when they are able to do this.48 This law, which belongs to the law of action rights and court-ordered rights,49 is not open, even in principle, to a corrective justice explanation.

H. Rules that Require or Permit Courts to Create New Rights The second group that contradicts the rubber-stamp view comprises rules that permit or require courts to make orders that create new rights, that is to say, to make orders that have a different content from the parties’ ordinary duties. The existence and size of this second category is controversial.50 The idea that court orders can create new rights conflicts with traditional assumptions about what courts should be doing. It may be one thing to say that courts are allowed, from time to time, to refuse to enforce a claimant’s ordinary rights, but it is another thing to suppose that the courts should be allowed to create entirely new rights. To the extent that they assume the latter role, courts are no longer acting as mere umpires. 46 The declaration of a constructive trust is strictly speaking not an order to do something. The effect of the declaration is, however, to give the claimant an ordinary property right and, thereby, to place others under ordinary duties to respect that right, In the case of the defendant, this typically means handing over the relevant property. 47 With rare exceptions, limitation periods do not extinguish the claimants’ ordinary rights; they merely preclude courts from making orders that enforce (directly or not) that right: Ketteman v Hansel Properties [1987] AC 189 (HL) 219. 48 It is an interesting question whether there exist ordinary rights for which affirming court orders are truly available ‘as of right’, that is, for which there exist no pre-conditions (including limitation periods) to making an order, aside from proof that the right has been infringed or threatened. In practice, the right to exclusive possession of one’s property appears to meet these conditions as it is not subject to equitable bars or limitation periods. The defence of adverse possession might, however, be regarded as equivalent to a limitation period in that it was traditionally said (though the matter is controversial) that adverse possession does not extinguish the owner’s title but merely the right to enforce the title. 49 It is not possible to conclude that a particular rule belongs to the law of action rights as opposed to the law of court-ordered rights without first determining whether, in the case where the claimant obtains the relevant order, the claimant had a ‘right’ to the order as opposed to having merely benefitted from the court’s discretion. To the extent that the court order is available only at the court’s discretion, the rules governing its availability are part of the law of court-ordered rights rather than action rights. 50 Or at least it appears to be controversial—‘appears’ because the possibility of this category is not so much denied as simply not discussed.

Duty of Care in Rights-Based Theory TheA Rights of Private Law of Negligence

131

The other reason this category is controversial is that it is not easy to know if and when court orders create new rights rather than confirm existing rights. The task is particularly difficult in a common law system. Courts rarely say whether they regard their orders as confirming existing rights or creating new rights. Indeed, it often appears to be assumed that the only possible proof that an ordinary right exists is that courts are willing to make orders that confirm the alleged right. This cannot be correct. If it were we would have to conclude (with Holmes) that there is no duty to perform most contracts because specific performance orders are unavailable for most contracts.51 It is nonetheless a common assumption. The conventional view that there is no duty to return property transferred by mistake (only a duty to return its monetary value) is often defended on the grounds that the courts of law did not entertain claims for specific restitution.52 It may well be that there is no duty to make specific restitution, but unless one is also willing to adopt the Holmesian view of contracts, the fact that the courts of law refused to order specific restitution does not establish this. The courts of equity were willing to make orders to return mistakenly transferred goods in roughly the same circumstances as they were willing to make orders to deliver contractually promised goods.53 These comments are not meant to deny the obvious truth that many, probably most, court orders do nothing more than confirm ordinary rights. Court orders to pay debts, to perform contracts, to quit land, to cease a nuisance or trespass, to make restitution of a monetary enrichment, and to pay compensation for consequential losses arising from a tort or breach of contract all appear to confirm ordinary rights. In each of these cases, the rubber-stamp view is correct to say that court orders merely compel defendants to do what they were already legally obliged to do. The content of such orders is therefore explained by the law of ordinary rights. That law is simply incorporated by the ‘action-right rule’ that the claimant has, in the circumstances, an action right to the enforcement of her ordinary right (or by the ‘court-ordered-right’ rule that the court has a discretion to enforce the ordinary right). It follows that the content of such orders (though not their availability) is in principle amenable to a corrective justice explanation. Yet it also seems clear that court orders sometimes create new rights. An example is an order to pay costs. Before a court order to pay costs is made, the party subject to the order cannot sensibly be said to be under a duty to make such a payment. Another example is an order to pay punitive damages. A defendant is not under a duty, prior to a court order, to pay punitive damages to the claimant. Punitive damages only achieve their punitive aim if they have been ordered by a court or other authority. Yet another example is an order that a defaulting party pay damages under a still valid contract. If the contract is valid, then the defendant’s 51 ‘The duty to keep a promise at common law means a prediction that you must pay damages if you do not keep it—and nothing else’: OW Holmes, ‘The Path of the Law’ (1897) 10 Harvard Law Review 457, 462. 52 P Birks, English Private Law, 1st edn (Oxford, Oxford University Press, 2000) 24. 53 The equitable jurisdiction to order ‘delivery up’ of goods is now enshrined in the UK in the Torts (Interference with Goods) Act 1977 (c 32).

132

Stephen A Smith

duty (until the order is made) is to perform the contract. That is exactly what it means to say that a contract remains in force. If a court then refuses to enforce the contractual duty, and instead enforces a different duty (ie, to pay damages), the inescapable conclusion is that the court has created, by its order, a new duty. It remains an open and little-studied question how many other examples of right-creating orders exist.54 But it is likely that there are others. The above examples point to at least three reasons that a court might find it necessary or desirable to create, by its order, a new right. One reason is that the court wants to make, by its order, a symbolic statement. In addition to punitive damages, orders that arguably fall into this category include nominal damages, damages for pain and suffering, and a range of other so-called vindicatory awards. In all these cases, the symbolic meaning of the relevant duty is only realised if it is created by a court order.55 A second reason that a court might reasonably want to create a new right is that there are institutional or other problems associated with direct enforcement of the relevant ordinary right. This seems to be the best explanation for most (though not all) of the voluminous and complex law that sets out the pre-conditions for specific relief orders. As I alluded to earlier, it might also explain the rule that an order compelling restitution in kind is only exceptionally available to claimants who bring actions in unjust enrichment. What these and other rules appear to allow for is the (perfectly reasonable) substitution of monetary awards for administratively complex and (in some cases) symbolically loaded56 non-monetary orders. A third reason is that it is not possible to define the relevant right’s content or exigibility (against whom it may be asserted) prior to making a court order. A right to costs falls into this category; a litigant who is able and willing to pay costs in the event that she loses a lawsuit will not know that she has lost until the court delivers its judgment. Arguably another example is an order that one partner in a failed relationship hold specific property on trust for the other partner.57 In ‘failed-partnership’ cases, we might want to say, prior to the order, that the claimant has a right of some kind over parts of the defendant’s property. But there might be no way of determining in advance what sort of right or which property it should be over. Further, if third parties are involved (eg, creditors), then it might be impossible for the parties themselves to create the kinds of property interests the court will feel are appropriate. The court order solves these questions. 54 The question was explicitly addressed in Rafael Zakrzewski’s pioneering work, Remedies Reclassified (Oxford, Oxford University Press, 2005). I touch briefly on the question in a review of Remedies Reclassified in (2006) 122 Law Quarterly Review 164 and in more detail in ‘The Law of Damages: Rules for Courts or Rules for Citizens?’ in Cunnington and Saidov (eds), above n 20, arguing in each case that the category of right-creating orders is larger than Zakrzewski concluded. 55 Authors who have recently argued that courts regularly make vindicatory awards include R Stevens, Torts and Rights (Oxford, Oxford University Press, 2007) 59–92 and D Pearce and R Halson, ‘Damages for Breach of Contract: Compensation, Restitution, and Vindication’ (2008) 28 Oxford Journal of Legal Studies 73. 56 I refer to the symbolism attached to court orders that compel an individual perform a service. 57 Peter v Beblow [1993] 1 SCR 980.

Duty of Care in Rights-Based Theory TheA Rights of Private Law of Negligence

133

These brief comments only brush the surface of a large and complex issue, but they are sufficient to show that the group of legal rules that require or permit courts to create, by their orders, new legal rights is substantial. Although these rules are conventionally assumed to belong to private law, they are qualitatively different from the kinds of rules that corrective justice theories are meant to explain. If corrective justice theorists want to explain these rules, then they must look to values other than corrective justice.

IV. Internal Pluralism The above arguments have sought to show that the only part of private law, conventionally understood, that is in principle amenable to a corrective justice explanation is the part that governs ordinary rights. This section discusses the kind of explanation that a corrective justice account will need to provide if it is to satisfactorily explain the law of ordinary rights. The conclusion will be that any plausible corrective justice account of the law of ordinary rights must be an internally pluralist explanation. It must be an account, in other words, that invokes a range of qualitatively different values or interests. The law of ordinary rights can be divided in different ways. The tripartite scheme presented below aims to identify structural differences that, viewed from the perspective of corrective justice, reflect a concern for qualitatively different kinds of interests or values. It may be useful to say at the outset that I will not attempt to explain or defend the actual rights in question. The purpose of this exercise is merely to identify the features of these rights that make their explanation on the basis of a single value unlikely. Finally, I do not argue that the scheme is exhaustive (although it might be). For present purposes, it is sufficient to show that at least some ordinary rights are qualitatively different from other ordinary rights.

A. Innate Rights Innate rights are those rights that citizens enjoy simply by virtue of residing within a particular jurisdiction. In broad outline, these are rights to our ‘person’, that is to say, rights to bodily integrity and to freedom of movement or action. Examples include rights against assault, battery, negligent infliction of injury, false imprisonment and deceit.

B. Created Rights Created rights are rights that individuals bring into existence or ‘create’ by performing acts that have this effect and where part of the reason they have this effect is that they are understood, in the normal case anyway, as manifestations

134

Stephen A Smith

of an intention to create such rights.58 Within the realm of private law, the main examples of created rights are contractual rights and property rights. Contractual rights arise when and because individuals communicate intentions to undertake obligations. There are other requirements (eg, consideration, capacity, etc), but the essential feature of contractual obligations is that those bound by them are bound, at least in part, because they intended to create such an obligation. Exceptional (and dubious) cases aside, it is not possible to make a contract by accident.59 Property rights, as we have already seen, are similar. Like contractual rights (and unlike innate rights) property rights are in the first instance60 created rights. The ‘first ownership’ of objects of property occurs when individuals perform certain acts with certain intentions. Thus it is possible (though nowadays rare) to take ownership of unoccupied land, animals, or things on land by performing acts that manifest an intention to take control of the land, animal or thing. Similarly, and more commonly today, individuals can create new property rights by (intentionally) creating new or transformed physical objects or, in the case of intellectual property rights, intellectual objects. Both created rights and innate rights are, in principle, amenable to corrective justice explanations. They are each based on important individual interests (bodily integrity, freedom of movement, the performance of promises, control of property) and the duties correlative to them can, in each case, be explained on the basis that treating others as equals means respecting these interests. Yet the interests themselves are very different. Innate rights are based on interests in our persons, while created rights are based on interests in things external to us. The explanation for why innate rights exist will therefore have a different structure from the explanation for created rights. For contractual and property rights to 58 On the importance of the actor’s awareness that what he is doing communicates an intention to create a right, see J Raz, The Morality of Freedom (Oxford, Oxford University Press, 1985) 80–88. 59 I discuss this in more detail in S Smith, Contract Theory (Oxford, Oxford University Press, 2004) 174. 60 I intentionally do not discuss the many rules governing the transfer (as opposed to the creation) of property rights and other rights. The proper classification of these rules raises difficult and, as yet, unresolved issues. According to what is probably the dominant understanding, the ability to transfer a right (or not) is part of the meaning of the right. Thus it is commonly said that the right of ownership includes the right to transfer what one owns to someone else: see, eg, T Honoré, ‘Ownership’ in AG Guest (ed), Oxford Essays in Jurisprudence (Oxford, Clarendon Press, 1961) 104. If this is correct, then it explains why the common law does not recognise the ‘law of transfer’ as a distinct legal category. It seems more reasonable, however, to regard the concept of transferring something to another person as distinct from the concept of (merely) owning that thing. The fact that I own a car means I have exclusive control over the car and so can generally do what I want with the car. Subject to others’ property rights, environmental legislation and so forth, I can abandon my car or burn it or grind it up for scrap. Transferring ownership to another is different from each of these acts. Most obviously it involves another person, who is not a mere passive recipient, but, like the transferor, must consent to the transfer. A transfer, like most contracts, is a (bilateral) intentional juridical act, that is to say, it is an act that has the legal effects it has in part because of the intentions of the relevant actors. At the same time, a transfer is different from the acts that bring into being contractual rights and property rights because rather than creating a new right it merely transfers an existing right. It seems likely, therefore, that the rules governing the transfer of rights are distinct from the three categories of rules discussed above.

Duty of Care in Rights-Based Theory TheA Rights of Private Law of Negligence

135

exist it must be possible for individuals to create rights where none existed before. For some scholars, this is a logical impossibility; the legal equivalent of creating something out of nothing. This view explains the frequent attempts to explain contractual obligations as either obligations not to let down those whom you have induced to rely61 or obligations to respect the transfer of already existing rights.62 These attempts are not necessary. But the fact that they have been thought necessary highlights the distinctive nature of created rights. As moral philosophers have long acknowledged, the concept of a self-imposed obligation is distinctive and complex. The same is true of the concept of ownership. By contrast, the idea that individuals have rights to bodily integrity is basic or primary within every moral system. Anyone who can understand the plea ‘leave me alone’ or ‘don’t touch me’ can understand innate rights. Neither property rights nor contractual rights are basic in this sense. As was explained in the earlier discussion of Weinrib’s Kantian theory of property, there is no direct route from ‘don’t touch me’ to ‘you promised’ or ‘that is mine’.63 Any plausible explanation of property rights and contract rights must, at some point in the story, say something about why it is important for individuals to be able to create such rights—about why the power to do this should exist at all. The explanation need not be instrumentalist in a crude sense, but it must at a minimum contain certain forward-looking or ‘positive’ normative ideas. Any attempt to explain contract and property must take account of the fundamental fact that such institutions provide opportunities that do not exist without them.

C. Remedial rights Remedial rights, in the sense intended here, are rights to the reparation, cure, or remedy of a problem—typically a loss or an injury. The clearest example is the right enjoyed by the victim of a tort or breach of contract to compensation from the wrongdoer in respect of consequential losses suffered as a result of the wrongdoing (‘compensatory’ damages). It is an interesting question (which I 61 See, eg, PS Atiyah, The Rise and Fall of Freedom of Contract (Oxford, Oxford University Press, 1979); PS Atiyah, Promises, Morals, and Law (Oxford, Oxford University Press, 1981). 62 See, eg, P Benson, ‘The Idea of a Public Basis of Justification for Contract’ (1995) 33 Osgoode Hall Law Journal 273; RE Barnett, ‘A Consent Theory of Contract’ (1986) 86 Columbia Law Review 269. 63 As noted earlier, Weinrib accepts that created rights (which he calls ‘acquired’ rights and which include, like the category of created rights, both property and contract) are distinctively different from innate rights. Indeed, understanding why they are different is the starting point for his explanation of created rights. Thus we saw that while innate rights exist in the state of nature, created rights are only fully realised in civil society. Consistent with this distinction, Weinrib states that the idea of property rights, unlike the idea of innate rights, is not ‘analytically contained in the universal principle of Right’ (above n 30, at 802). At the same time, Weinrib explains created rights as being derived (in a difficult-to-describe sense) from the same Kantian doctrine of Right that explains innate rights. In the end, it is not entirely clear to what extent Weinrib would accept that created rights have different normative foundations from innate rights. Nevertheless, it is clear that insofar as he would deny this conclusion, he would do so on the basis of the explanation of property rights that was discussed—and rejected—earlier in this essay.

136

Stephen A Smith

discuss below) whether there are other remedial rights. For present purposes, the important point is that remedial rights are in principle amenable to a corrective justice explanation. Indeed, for those who adopt the narrow view that equates corrective justice with duties to repair, the law of remedial rights appears to be the only part of private law that corrective justice can explain. The narrow view having been rejected earlier, the question to be considered now is not whether duties to repair can be explained by the broader view—clearly they can—but whether such duties are different in significant ways from duties to respect innate rights and created rights. The evidence points to an affirmative answer. Duties not to trespass, to perform contracts, and not to defame others, are just that. They are duties not to trespass, to perform contracts, and not to defame. They are not duties to repair the losses that might arise from trespassing, breaking a contract, or defaming another. It is not just the substance but the grounds of these duties that are different. My duty not to trespass over your land is grounded in your interest in exclusive control of your land. Your right to compensatory damages for losses caused by my trespass is grounded in your right to be compensated for wrongs done to you. A victim’s right to compensatory damages is distinct from, and must be explained on different grounds to, the victim’s right that the tort or breach of contract not happen in the first place. The correlative duty is a duty to make things right rather than a duty not to make things wrong. Against this view, Weinrib has argued that the duty to pay damages is (merely) a transformed version of the original duty not to trespass, etc: ‘Like Descartes’ famous piece of wax that remains the same piece of wax despite its different qualities when heated and cooled, the same duty is continuously correlative to the same right before and after the tort’.64 There are indeed certain kinds of damages that are naturally explained in just this way.65 ‘Cost of cure’ damages, which are routinely awarded in both contract and tort, provide claimants with sums sufficient to allow them to repair, replace, or purchase a substitute for whatever goods or services were damaged, destroyed, lost or not provided as a result of the tort or breach of contract—even if the sum is greater than the goods’ or services’ pecuniary value to the victim.66 Though not identical to the defendant’s original

64 EJ Weinrib, ‘Two Conceptions of Remedies’ in CEF Rickett (ed), Justifying Private Law Remedies (Oxford, Hart Publishing, 2008) 13; see also Weinrib, above n 7, at 134–44. Ironically, if this view is accepted then the one thing that ‘corrective justice’ is not concerned with is ‘correcting’ injustices. In Weinrib’s view, damages orders are similar to specific performance in that they are designed to enforce the original duty. If this is right, then, while the occurrence of a wrong might trigger the duty to pay damages, the wrong is not the justification for that duty: the justification is the same (non-corrective) justification that explained the original duty. 65 As I have argued elsewhere: S Smith, ‘Substitutionary Damages’ in CEF Rickett (ed), Justifying Private Law Remedies (Oxford, Hart Publishing, 2008). 66 Except where the sum is ‘out of all proportion’ to the value of the loss: Ruxley Electronics & Construction Ltd v Forsyth [1996] AC 344 (HL). This exception is consistent with the idea that cost of cure awards are a form of substitute specific performance because specific performance may be refused on similar grounds.

Duty of Care in Rights-Based Theory TheA Rights of Private Law of Negligence

137

duty, the duty to pay for a cost of cure can reasonably be understood as a form of substituted performance, and, as such, can be justified on the same grounds as the original duty.67 Damages for consequential losses are different. A vendor’s contractual duty in a contract of sale is nothing more nor less than a duty to deliver the contractually specified goods. The vendor has no duty to ensure that the buyer obtains whatever financial benefits he or she reasonably expects to obtain from those goods. The obligation is to deliver the goods, not to deliver the value of the goods. The rules governing claims for pure economic loss make this clear. The buyer has no claim for compensation if the vendor—or anyone else—reduces the expected value of the goods to the buyer, say by negligently destroying other property that the buyer was going to use in conjunction with the goods. The distinction between things and their value is even clearer in tort cases. I have an ordinary right that my property not be damaged through another’s carelessness. But I have no freestanding ordinary right to the value of the property. You can reduce the value of my property in a myriad of ways; the only time you will be liable for that reduction is if it arises from an injury to the property. The law of ordinary rights does not protect value per se.68 The significant body of law that exists on remoteness and the mitigation of damages is further confirmation that the duty to pay damages for consequential loss is distinct from duties to respect status and created rights. If the duties were the same there would be no need to take into account mitigation and remoteness. The content of the original duty would fully establish the content of the substitute duty. This conclusion does not deny that the duty to repair a loss is linked to the duty not to cause the loss. An important part of the explanation for why a defendant is liable to pay compensatory damages is that the relevant loss was caused by the defendant’s infringement of the claimant’s right. To this extent, the law of compensatory damages incorporates the law of ordinary rights. But the relationship between primary duties and compensatory duties is normative, not logical. A legal system that refused to award compensatory damages might not be desirable, but it would not be illogical. An important question for anyone trying to produce a complete typology of private law rights is whether there exist other remedial rights in addition to the right to damages for consequential losses suffered from a tort, breach of contract, or other wrong. It was already noted that rights to cost of cure damages probably do not qualify as remedial rights. Nominal damages, punitive damages, damages for pain and suffering and other kinds of vindicatory damages also do not qualify.

67 In contract cases where the original duty is still valid up to the moment of judgment, a cost of cure award, though substitutive in content, appears to create a new right rather than to give effect to an existing ordinary right. 68 Nor could it coherently do so: J Penner, ‘Value, Property, and Unjust Enrichment: Trusts of Traceable Proceeds’ in R Chambers, C Mitchell and J Penner (eds), Philosophical Foundations of the Law of Unjust Enrichment (Oxford, Oxford University Press, forthcoming, 2009).

138

Stephen A Smith

Although such rights are distinct from status and created rights, they are not remedial because they are not intended to repair or cure a problem. As was suggested earlier, these rights are probably not ordinary rights at all (they are courtcreated rights), but if they are, then they belong to a separate category. The most likely candidates for inclusion in the category of remedial rights alongside rights to compensatory damages appear to be the rights that are given effect in restitutionary and ‘near-restitutionary awards’ (eg, awards arising out of claims based on quantum meruit, restitution for wrongs, restitutionary damages, salvage, general average, necessitous intervention, and necessity). Though such rights are not rights to be compensated for a wrong, they are structurally similar to rights to compensatory damages.69 Like rights to compensation (and unlike innate rights, created rights, and vindicatory and other court-created rights) their general purpose is to cure or remedy a problem or injustice. The correlative duties are duties to make things right, not duties not to make things wrong.

V. Conclusion The general argument of this essay is that private law is complex. It is complex because it comprises two fundamentally different kinds of rules—rules directed at citizens and rules directed at courts—and, further, because the rules in the first category (ie, the rules directed at citizens) are themselves of qualitatively different kinds. An important consequence of this complexity is that any attempt to explain private law on the basis of individual rights and duties must be both externally and internally pluralist. This conclusion has two implications for corrective justice explanations of private law: (a) some private law rights cannot be explained by corrective justice (external pluralism); and (b) the private law rights that corrective justice can explain rest on qualitatively different kinds of values (internal pluralism). The essay’s conclusion is thus in part a criticism and in part a defence of corrective justice (and other rights-based) approaches to private law. It is a criticism to the extent that it rejects the view that corrective justice theories provide a simple model for explaining the entirety of private law. The law is more complex than this view assumes. At the same time, the conclusion is a defence of corrective justice approaches to the extent that it rejects the view that such approaches are worthless unless they can show that corrective justice explains the entirety of private law. The concept of corrective justice is more complex than this view assumes.

69 See S Smith, ‘Unjust Enrichment: Nearer to Tort Than Contract’ in R Chambers, C Mitchell and J Penner (eds), Philosophical Foundations of the Law of Unjust Enrichment (Oxford, Oxford University Press, forthcoming, 2009).

6 The Conflict of Rights ROBERT STEVENS*

I. Judgement Many years ago, I was a nervous young pupil-barrister at a commercial set of chambers. Like all those who go through that unhappy and unnatural stage, I was anxious as to whether I would be awarded tenancy. Being ‘taken on’ would lead to prosperity and, I thought, happiness. One day I overheard a conversation about me between my pupil-master, who was overseeing my work, and the head of chambers, a leading silk. ‘He is clever’. ‘I don’t doubt it, but does he have judgement?’

What on earth did that mean, I thought? Why didn’t the silly old fool think that my good degree and obvious intelligence were enough? This chapter concerns the importance of judgement.

II. Introduction At one time it was popular to try to understand private law generally, and the law of torts in particular, through the goals or social policies it could be seen as furthering.1 The human condition is one of constant striving for an end, and it is natural to assume that private law is similarly trying to achieve something. A once fashionable form of this approach was that advocated by some lawyer-economists who sought to show that ‘the common law of torts is best explained as if the judge who created the law through decisions operating as precedents in later cases were

* I am grateful to all who commented on the earlier draft of this chapter, in particular to Dr Allan Beever, Nick McBride, Professor Jason Neyers, Professor Steve Smith and Dr Charlie Webb. All errors are mine, as are most of the good bits. 1 Eg G Williams and BA Hepple, Foundations of the Law of Tort (London, Butterworths, 1976). The great jurist Glanville Williams was a committed utilitarian.

140

Robert Stevens

trying to promote efficient resource allocation’.2 There are, as has been pointed out several times, a number of insuperable problems with this sort of purely functionalist account.3 Perhaps more widespread nowadays are ‘mixed’ or ‘pluralist’ views of private law in general, and the law of torts in particular.4 On this view we cannot see the law as pursuing any single goal. All considerations of justice from any perspective are in play. A mixed account leaves room for (moral) rights, which are, like love and art, functionless and valued for themselves.5 This approach of mixing has four obvious attractions. First, it has the apparent virtue of being polite. It concedes to, say, both economists and corrective justice theorists that there is some truth in what they claim. Secondly, it leaves all possible rational reasons for the resolution of a dispute in play, making the subject sexy and interesting. Thirdly, it avoids difficult choices between, say, utilitarianism, other forms of consequentialism, Kantian deontology and Aristotelian virtue ethics. Lawyers are not philosophers or economists, and so it is comforting to conclude that it is unnecessary to think very hard about such matters. Fourthly, it enables the lawyer to dismiss purely theoretical accounts as inconsistent with the complexities of the law. This comforts lawyers (like me) as it means we have a set of skills and a body of knowledge that the theorist does not possess. It is a realist or pragmatic conception of law. Unfortunately, whilst the ‘mixed’ or pragmatic view of law is popular it cannot be true, or at least not true in relation to a legal system which makes any sense. Law is not like minestrone soup. One cannot simply add together a number of disparate ingredients and hope to get a satisfactory result. In many cases, seeking to construct a rule which will lead to the maximisation of wealth (or aggregate choice, or pleasure, or utility, or anything else) will give a quite different result from a rule based upon our moral rights one against another. By what further criterion is a judge to determine whether considerations of interpersonal justice can be trumped by concerns for the general welfare, or vice versa, when these concerns do not exist on the same plane? One cannot be a bit of a utilitarian or a half-hearted Kantian, and it is certainly impossible to try and be a combination

2 WM Landes and RA Posner, The Economic Structure of Tort Law (Cambridge MA, Harvard University Press, 1987) 1. 3 See eg EJ Weinrib, The Idea of Private Law (Cambridge MA, Harvard University Press, 1995); JL Coleman, The Practice Of Principle: In Defence of a Pragmatist Approach to Legal Theory (Oxford, Oxford University Press, 2001). 4 Eg G Schwartz, ‘Mixed Theories of Tort Law: Affirming Both Deterrence and Corrective Justice’ (1997) 75 Texas Law Review 1801. 5 WH Auden, ‘Law Like Love’ in Collected Shorter Poems, 1927–1957 (London, Faber, 1966) 154: ‘ “Law”, says the judge as he looks down his nose, Speaking clearly and most severely, “Law is as I’ve told you before, Law is as you know I suppose, Law is but let me explain it once more, Law is The Law” ’.

The Conflict of Rights

141

of the two. Attempting to form a view as to what the law of torts specifically or private law in general is about is unavoidable. Why would they be worth considering as distinct topics if they are not ‘about’ something or other?6 A ‘mixed’ theory is not a theory at all. Another temptation is to try and give an internally-coherent account of the law without subscribing to any particular view as to why we have it. However, coherence is not the only virtue law can possess. If we were to accept the premises of an abhorrent belief system, the conclusions which could then be deduced may well be internally consistent, coherent and non-contradictory. They would be no less abhorrent for having these characteristics. A legal system is very expensive to run, and without some morally-defensible theory as to why we have it the calls for its abolition in whole or in part will be irresistible. If the only thing to be said in favour of a body of law is that it is coherent, it should be done away with. Coherence alone cannot make law valuable, but is rather a necessary condition that both the law and theories about it must satisfy in order to qualify as such. Further, although in most cases we can work out what the result should be without having to subscribe to any grand theory of law, at least in hard cases it is impossible to say what the result on a particular point will be without having formed a view as to why the law, at a higher level of generality, makes sense. This view has recently begun to be described as an ‘interpretive’ approach to law.7 However, a purely descriptive account of the law, which ignored the question of why the law makes moral sense, would be impossible, save in relation to the core application of certain basic rules. Such an account would also be exceptionally tedious. By contrast, a wholly normative approach does not have any necessary relationship with the law as it currently is. Interpretation is just what lawyers do, and have always done. The attraction of seeing the law as pursuing a single particular goal, such as wealth maximisation or overall happiness, is that we are presented with a single yardstick by which to resolve all forms of conflict between persons. By contrast, the attraction of the mixed view is that there never is any necessarily right answer, it is just a matter of learning the rules at any given moment and unmasking the reasons the judges have for making their political choices, with a conflict ongoing and never ending. For someone, such as myself, who puts himself forward as a believer in a rights-based approach, but who also considers that such rights arise for different reasons, the position is rather different. The task is at one level made simple because the range of reasons in play in resolving disputes is relatively restricted. Private law is simply about the rights we have one against another. We cannot

6 cf T Weir, An Introduction to Tort Law, 2nd edn (Oxford, Oxford University Press, 2006) ix: ‘Tort is what is in the tort books and the only thing holding it together is the binding’; J Murphy, ‘Rights, Reductionism and Tort Law’ (2008) 28 Oxford Journal of Legal Studies 393. 7 Eg A Beever and C Rickett, ‘Interpretive Legal Theory and the Academic Lawyer’ (2005) 68 Modern Law Review 320.

142

Robert Stevens

expect it to end world poverty, reverse global warming, make our lunch or, more generally, make the world a better place. The task is not one of trying to reconcile a number of incommensurable policy goals. One of the central points of law is to provide determinate rules when moral rights are, at the margin, underdetermined. Choices therefore still have to be made, but at least they are easier to make. However, how are our rights to be ranked when they conflict? How can such ranking be determined without appealing to a higher-level single goal or policy which the law is pursuing? Giving an account of private law which explains its complexity but which is also coherent and morally defensible is not easy. The topic of the conflict of rights raises issues from the mundane to the fundamental. So, can I cut back the branches of my neighbour’s apple tree when they overhang my land? Can I sacrifice another’s life in order to save my own? This chapter is a conservative one, and the following substantive part seeks to achieve three modest things. First, to analyse whether and how rights can conflict with one another. Secondly, to differentiate different sorts of conflict and to consider how they are resolved. Thirdly, to reflect upon the hierarchy of rights in the common law.

III. Rights Defined Compared with the work of many modern-day legal theorists, Wesley Newcomb Hohfeld’s work reveals a refreshing engagement with the law that lawyers deal with on a daily basis. His examples range from unpaid repairer’s liens to the powers of factors to confer good title upon a bona fide purchaser. Hohfeld was a great lawyer, and he was seeking to show how his work was of relevance to law in practice. Unfortunately, the only people some modern day theorists seem to be engaged with are other theorists. I have heard Hohfeld described as ‘old hat’. I could not disagree more strongly. Unlike those who subsequently examined the usage of ‘rights’ in ordinary language,8 Hohfeld was, in part, prescriptive. He sought to escape from the confusion in everyday usage by separating out different meanings of ‘rights’, rather than trying to find an underlying unity to all usages. Hohfeld was appalled by the sloppiness of ordinary language, not inspired by it. Of course, Hohfeld’s prescriptions have a basis in ordinary usage, in particularly the usage of lawyers, but it is strictly speaking meaningless to say that Hohfeld’s prescriptions are right or wrong, or to claim that there are gaps in his account.

8 Eg HLA Hart, Essays on Bentham: Studies in Jurisprudence and Political Theory (Oxford, Clarendon Press, 1982) ch 7.

The Conflict of Rights

143

The key distinction is between a claim right and a liberty. I have a claim right that you do (or refrain from doing) an action if and only if you have a duty to me to do (or refrain from doing) that action.9 I have a liberty with respect to you to do (or refrain from doing) an action if and only if you have no claim right that I should not so act (or refrain from acting). Once we have a snapshot of my claim rights, the negative is the liberty of all others with respect to me. The two other senses of ‘right’ Hohfeld identified—powers and immunities— are always merely components of larger statements concerning claim rights and liberties. So, I may have a claim right that you do not use my car because the factor who conveyed the car to me had the power to confer good title upon an innocent purchaser. Powers and immunities have no free-standing existence separate from the claim rights and liberties to which they relate. The same is not necessarily true of claim rights and liberties, which are our basic building blocks.10

IV. Can Rights Conflict? Within legal reasoning, rights live a double life. First, specified rights are used as conclusions to legal questions. Secondly, rights are used in a more general sense as premises from which answers to legal questions may be deduced. Only rights used in the latter sense can conflict. Hohfeld’s strictures are sometimes difficult and inconvenient to adhere to, as is reflected in the fact that not all holders of the Wesley Newcomb Hohfeld Chair of Jurisprudence have employed them.11 Formally, as a matter of positive law, specified rights in Hohfeld’s sense cannot conflict within a legal system if it is to be coherent. So, it is impossible for both of the following two statements about rights and liberties to be true within a single legal system: 1. A has a claim right that B does not drive his car negligently so as to injure A. 2. B is at liberty with respect to A to drive his car in any manner B so chooses. If A possesses the claim right posited in the first statement, B does not have the liberty posited in the second. There can be no conflict; either A has the claim right, which excludes B’s liberty, or he does not.

9 It is sometimes argued that claim rights do not always have correlative duties: eg N MacCormick, ‘Children’s Rights: A Test Case for Theories of Right’ (1976) 32 Archiv für Rechts und Sozialphilosophie 305; N MacCormick ‘Rights in Legislation’ in PMS Hacker and J Raz (eds), Law Morality and Society: Essays in Honour of HLA Hart (Oxford, Clarendon Press, 1977). These arguments are without substance: see generally NE Simmonds, ‘Rights at the Cutting Edge’ in MN Kramer, NE Simmonds and H Steiner (eds), A Debate Over Rights: Philosophical Enquiries (Oxford, Clarendon Press, 1998) 113. 10 It is possible to have claim rights against other rights, for example the trust. 11 Eg R Dworkin, Taking Rights Seriously (London, Duckworth, 1977). Dworkin usually employs ‘rights’ in the second general sense, whilst Hohfeld is discussing specified rights.

144

Robert Stevens

Similar unqualified broad statements about the rights persons have may be logically incompatible with one another. So, it is impossible for these two statements, without further qualification, to both be true within a legal system: 1. All of us have a right that no other may touch us without our prior consent. 2. All owners of land may use reasonable force to exclude others who enter their property without their consent. If the second statement is true, it must, at least, provide for a qualification of the first. The second statement concerns a privilege that the landowner has. A privilege is an exceptional form of liberty. It is a liberty which persons generally do not possess. (Similarly, if I have an immunity with respect to you, you, exceptionally, have no power with respect to me). Exceptionally, the trespasser has no claim right that the landowner does not use reasonable force to compel his ejection. The right the landowner has against all others to exclude them from his premises confers upon him a privilege to use reasonable force to remove trespassers. Normally the use of such force would be a battery. Again, therefore, as completely accurate statements which operate as conclusions as to what our legal rights are, it is impossible within a rational system for there to be conflicts of this kind. Therefore, when used in our first sense as specific conclusions in answer to legal questions (eg, Do I have a liberty with respect to you to do x?) rights cannot be permitted to conflict within a coherent system. However, unqualified statements as to what our prima facie rights are can, and do, conflict. Similarly, the reasons underlying why one person may have a right as a matter of positive law may conflict with the reasons underlying the right of someone else. When used in a general sense from which we try to deduce the answer to legal questions, rights commonly conflict. These conflicts then have to be resolved. The conflict of prima facie rights creates privileges, such as that above of the landowner, which would not otherwise exist. Some other conflicts whilst logically possible cannot sensibly be allowed. So, if X has a right to exclude all others from Blackacre, and Y also has such a right, nobody has a liberty with respect to all others to use Blackacre. X can exclude Y, and all others, and so can Y. Such a stultification of the use of land is obviously intolerable. Priority disputes in relation to the right to things are a common example of rights in conflict.

V. Legal and Moral Rights Whilst there are many necessary connections between law and morality, without further explanation the maxim lex injusta non est lex is patently false. Bad law is still law, just as bad art is still art. So, in the antebellum United States the law gave recognition to legal rights over other people, denying large numbers of people

The Conflict of Rights

145

freedom of movement, a right to bodily safety and the freedom to choose for themselves how to live their lives. The law itself supported and made possible the ‘peculiar institution’. The legal rights created by chattel slavery were a grotesque violation of the moral rights of those enslaved. It is beyond dispute that rights as a matter of positive law and moral rights can conflict with one another. Of course, we should not try to understand what law (or art) is or why we value it by focusing on the worst or more marginal examples of it. The legislature has the power to create legal rights for any reason at all, unconnected with the moral rights we have one against another. Instrumentalist justifications generally underlie intellectual property rights, for example. Trademarks enable purchasers of goods to make educated choices in the market and, over time, provide consumers some assurance as to quality. Patents encourage research by securing the financial reward of innovation. It is no accident that such rights were created by legislation, rather than being judge-made. By contrast, the judiciary, in creating and changing over time our common law, sourced our legal rights one against another in our moral rights. I have a moral right that you do not punch me on the nose, and we consider such a right to be important. As a result I also have a legal right to the same effect. The courts have the same relationship to justice that art galleries have with art: the role of the former is to promote the latter as a good in itself. Neither the moral right nor the legal right giving it concrete form exist in order to fulfill any further policy or function, although any number of incidental benefits may follow. The law of torts is founded upon the negative formulation of the Golden Rule.12 We should not do unto others what we would not want done unto ourselves. This is the cornerstone of both interpersonal morality and law. Within the law of torts, therefore, judge-made law is legitimately concerned with conduct in relation to others which is malum in se. Torts which are only such because of the law itself, malum prohibitum, are possible, but are the province of the legislature. Not all of our moral rights either are or should be given the force of law, just as not all of the things capable of being described as art belong on the walls of public galleries. Insults, for example, wrong others but we do not consider the right violation to be sufficiently important so as to be actionable in a court. The ‘sledging’ committed by the cricketers Shane Warne and Glenn McGrath against their opponents throughout their careers was disgraceful and morally wrong, but it was not actionable, at least not when they were playing in England or Australia.13 There is no conflict between law and morality created by the law’s leaving some moral rights outside of its scope. The law’s silence does not facilitate or constitute approval of such conduct. 12 JR Atkin, ‘Law as an Educational Subject’ (1932) Journal of the Society of Public Teachers of Law 27, 30. For a fine review of the Golden Rule in law see N Duxbury, ‘Golden Rule Reasoning Moral Judgement and Law’ (2009) 84 Notre Dame Law Review (forthcoming). 13 It is possible that there is an action under South African law.

146

Robert Stevens

VI. ‘Human’ Rights The modern idiom for moral rights is human rights, but the usage of this term has become detached from its original meaning. Properly so-called, a human right is a right we have simply by virtue of being human. Such rights are exigible against all other persons. A human right is a moral right which holds good regardless of the society we find ourselves in. Examples include the right to bodily safety and the right not to be told lies which we believe.14 That we acquire such a right by virtue of our status as human beings does not necessarily mean that the right is particularly important, nor that it ranks higher in importance than rights recognised for other reasons. The label ‘human rights’ has over time acquired quite a different meaning. ‘Human rights’ have come to mean rights that individuals have against the State that the State secure a particular human good or interest. Such rights place limits upon a State’s sovereignty. It is this sense of ‘human rights’ which is employed by the European Convention for the Protection of Human Rights and Fundamental Freedoms. The title of the Convention makes this clear. I may have a right to a freedom because the State in which I live has a duty to secure that freedom for me. Against persons generally I just have a freedom, not a right that they secure it for me. The nature of the Convention is shown by the recognition of a right to education.15 Clearly, this right requires positive steps be taken by the State, upon which the correlative duty is placed, to secure this good. It is obvious that this is not a moral right which arises simply by virtue of being human. I do not have a right against all persons generally arising upon birth that they take steps to educate me, nor does the Convention require the creation of a right to education with ‘horizontal effect’ against persons other than the State itself. The ‘right to education’ is contingent upon there being a State in which I live against which it is exigible. The right to freedom of expression is similar: it is a right against the State only, not persons generally. At common law, I generally have a liberty to speak as I choose with respect to you, but there is no right either at common law, or in morality, specific to this liberty. There is nothing peculiarly ‘human’ about these rights against the State when compared to any other right. If it is objected that ‘human rights’ merely mean rights capable of being held by humans, the qualifier loses all meaning, as this is true of all rights. The modern label ‘human rights’ is a misnomer for this category, a better alternative would be ‘citizens’ or ‘civil’ rights, which emphasises the party against whom such rights are exigible (the State).

14 ‘An unbelieved lie is like a shot which misses’. cf R Stevens Torts and Rights (Oxford, Oxford University Press, 2007) 8. 15 European Convention for the Protection of Human Rights and Fundamental Freedoms, Protocol No 1, Art 2.

The Conflict of Rights

147

When it is said that rights are ‘trumps’ it is usually the case that what are being referred to are the (‘human’) rights individuals have specifically against the State. These rights of the individual have peremptory force so that a State cannot simply override them because this makes sense in social policy terms. So, the State cannot torture to obtain information, or detain suspects indefinitely without trial in order to combat potential terrorist threats. It is unacceptable as a matter of morality to use individuals in this way as a means to an end, however laudable the goal. There is no ‘conflict of rights’ properly so-called here, although politicians sometimes rhetorically claim that there is when they speak of the rights of potential victims. Rather there is a potential conflict between the rights of individuals and overall welfare. ‘Human rights’ in this sense confer upon individuals not only negative rights—that the State refrain from acting in certain ways—but positive rights as well. So, for example, the right to life imposes a positive duty on the State to investigate killings.16 These ‘human rights’ impose duties upon States that they secure a minimum level of protection of certain basic goods for the individual. However, at least at the margin, that some goods and not others are conferred protection appears to be a matter of political choice rather than natural reason. Housing and healthcare may be thought to be just as important as education, but it is only the latter which is protected by the European Convention. The label ‘human rights’ has a rhetorical pull, as it implies that all such rights are, like our right to bodily safety, not matters of political choice. No doubt the label ‘human rights’ is employed in order to give the impression that the list of goods protected is self-evident. Some of the reasons which can be given for the recognition of a legal right against the State, restraining its ability to act in certain ways, are concerned to further certain policies or goals to promote the common good, rather than to protect the interests of the individual. So, we might consider that the ‘right to freedom of expression’ is justifiable because it facilitates a market in ideas, which is a price worth paying for any short-term harm caused by tolerating the expression of offensive or dangerous views.17 This is a justification independent of any interest of the right-holder. Similarly, a rule-utilitarian might argue that an absolute right not to be tortured has a prophylactic effect. If ‘good’ tortures were allowed this would undermine the general prohibition on torture, which would lead to more ‘bad’ tortures. Of course, if such policy goals are to be the justification for these constraints on State action, it would require a theory as to why the courts, as opposed to, say, a democratically elected legislature, are better at judging how the common good is best pursued. It must be doubtful whether such a theory is possible. It is much easier to understand how a court has special competence to judge whether a majoritarian decision is inconsistent with the important interests of an individual, which deserve protection as a good in themselves. 16

Kelly v United Kingdom (App No 30054/96) 4 May 2001. Abrams v United States, 250 US 616 (US Supreme Court, 1919) 627–8 (Holmes J); Whitney v California, 274 US 357 (US Supreme Court, 1927) (Brandeis J). 17

148

Robert Stevens

It is possible for our domestic common law rights against one another to conflict with rights against the State that certain liberties be respected. This is true both of liberties contained in international conventions such as the European Convention on Human Rights and those enshrined in a constitution such as that of the United States. If, for example, the English right not to be defamed were unqualified, so that there was no defence of fair comment, this would violate the right that citizens have against the State to secure freedom of expression. The Convention would require the domestic law right not to be defamed to be cut back. Consequently, there is potential for the right to the liberties enshrined in the European Convention on Human Rights to have ‘horizontal effect’, as the existence of certain claim rights may be a violation of Convention liberties. However, the courts have repeatedly held that the English law of defamation is compatible with the Convention.18 We do have defences of fair comment and qualified privilege, which leave space for freedom of speech. However, the defence of qualified privilege was expanded by the House of Lords in Reynolds v Times Newspapers Ltd.19 The defence of qualified privilege now applies where the publication involved a ‘duty’ to the public to disclose the information. As discussed below, this did not involve, as had previous examples of qualified privilege, a genuine conflict between the claimant’s right to reputation and a third party’s moral or legal right to be told the statement made. Specific individuals have no right in morality or law that a newspaper prints particular information. Rather, Reynolds involved the public interest in freedom of speech. In the United States, the constitutional right to freedom of expression led to the emasculation of the common law right to reputation.20 Although Reynolds has cut back the common law right, there is no significant indication that the American fate awaits the law of defamation in England, and the House disclaimed any direct reliance upon the Convention. Convention rights to rights, by contrast, are adequately protected through a remedy against the State itself, in the United Kingdom now provided by the Human Rights Act 1998.21 There is no conflict between a Convention right to privacy against the State and the absence of any domestic private right to privacy against persons generally. Before the Human Rights Act was brought into force, there was a potential for conflict because the absence of a remedy in the domestic courts against the State, vindicating a Convention right violation, could itself be a Convention Right violation: the right to an effective remedy under Article 13. The common law was the only way of filling the gap to ensure Convention compliance. Today there is no need to develop the common law in order to ensure compliance 18 Eg Derbyshire County Council v Times Newspapers Ltd [1993] AC 553 (HL) 551 (Lord Keith); Reynolds v Times Newspapers Ltd [2001] 2 AC 127 (HL) 203 (Lord Nicholls); McCartan Turkington Breen v Times Newspapers Ltd [2001] 2 AC 277 (HL) 300–301 (Lord Cooke). 19 Reynolds v Times Newspapers Ltd [2001] 2 AC 127 (HL). 20 New York Times Co v Sullivan, 376 US 254 (US Supreme Court, 1964): ‘actual malice’ required where a claim is brought by a public official. 21 Human Rights Act 1998 ss 6–8.

The Conflict of Rights

149

with Article 13, as the Human Rights Act itself constitutes an adequate remedy for these purposes, which is why Article 13 was not incorporated as a protected right under the legislation. The condition of English domestic law itself does not constitute an infringement of, say, the right not to be tortured. If someone is in fact tortured, the existence of a domestic law right not to be tortured will not prevent there being a violation of Article 3 of the Convention. Conversely, if nobody is tortured, the lack of a domestic law right not to be so treated is not in itself a violation of Article 3. The Convention’s potential impact on domestic private law is very limited, but what impact it has will be to cut back rather than to expand our rights at common law. It is unclear what those who consider that the substantive articles of the Convention themselves require the creation of parallel domestic common law rights think the purpose of Article 13 is. That the impact of the enshrinement of the European Convention on Human Rights into domestic English law by the Human Rights Act 1998 reduces the significance of the Convention so far as domestic private law is concerned has been obscured by the coincidence that at the time of the enactment of the Human Rights Act a number of decisions were handed down by the European Court of Human Rights where the United Kingdom was found to be in violation of its duties under Article 13. These cases concerned the legal position before the Act came into force.22 The Court’s decisions on violations of other Articles, such as Articles 3 and 8, would have been precisely the same regardless of the position under United Kingdom domestic law. In private law we are normally referring to the rights which we have one against another, not to the rights we have against the State and which we do not have against others, and this will be the focus of the remainder of my chapter.

VII. Rights and Liberties It is difficult to imagine a world where our liberties were absolute. Freedom for the pike is death for the minnow,23 but we are all both pikes and minnows. Absolute freedom of expression would allow the shouting of ‘Fire!’ in a crowded theatre, and absolute liberty of movement would allow us to stab one another to death. Conversely, the rights that we have which are exigible against all others are not absolute in scope, nor could they be. We cannot all have rights good against all others that are without limit. This has implications for how strict liability can be justified within the law. A tort is a civil wrong. A civil wrong is a breach of a duty owed to another. A breach of such a duty is an infringement of a right the claimant has against the 22 Eg Z v United Kingdom (App No 29392/95) 10 May 2001; Wainwright v United Kingdom (App No 12350/04) 26 September 2006. 23 I Berlin, Two Concepts of Liberty (Oxford, Clarendon Press, 1958).

150

Robert Stevens

defendant. Prior to breach it is sometimes possible to obtain an order from a court that the claimant’s right is to be respected. In contract this may take the form of an order to positively confer a benefit upon the claimant—an order for specific performance in the case of a service, and an action for the agreed sum in the case of money. Our rights against persons generally are, however, of a negative kind and so the ordinary order will be an injunction restraining the defendant from acting in a particular way. Properly speaking, such court orders are not remedies for torts or wrongs which have been committed. Rather, such orders enforce the claimant’s primary right, whether it is to exclude someone else from their land, or that others do not libel them or, for prophylactic purposes that they do not approach the claimant.24 The court order confers upon the right-holder further powers of enforcement, as failure to comply will be a contempt of court. Specific relief of this kind is demonstrably not a remedy for a wrong, as proof of a wrong is unnecessary. It is definitionally impossible to order specific relief as a response to liability for negligence, as this is a situation where the wrong has already occurred. It is too late. Once the primary right is violated, a secondary right to pay damages may arise. Payment of damages is the next best thing that can be ordered after the right has been infringed. Some wrongs (such as breach of contract) are actionable without proof of financial loss, whilst others (such as, generally, slander) are not. Where proof of consequential loss is required, this goes to actionability, not to whether there is a wrong. A wrong which is not actionable is still a wrong, with some legal effects, just as an unenforceable obligation is still an obligation, with some legal effects, and not a nullity. In the light of the above, there are limits upon what can constitute a tort, based upon the inevitable limits of our primary rights. First, not all claims for compensation can plausibly be explained as based upon a civil wrong. One example in the common law is the claim to general average. When part of a ship or cargo is sacrificed whilst at sea in order to save the rest, all parties (ship, cargo and freight) proportionately share any loss from a voluntary and successful sacrifice. The carrier commits no wrong against anyone if it sacrifices part of the vessel, nor does one cargo-owner wrong another when the carrier sacrifices part of the cargo to save the rest. The general average claim for compensation for loss cannot plausibly be based upon a tort. Secondly, a regime of general strict liability for injury caused to others could not be part of the law of torts. If we were each under a duty not to engage in any conduct which posed any risk, however slight, to anyone else, each of us would be under a duty not to engage in any conduct at all. All action poses some risk to someone else, however remote. We cannot all be obliged to lie rigidly in bed all day long. (Indeed even choosing to stay permanently in bed imposes risks on others: it isn’t very healthy). Once we accept that the law of torts concerns the breach of real duties owed to others—with the infringement of real

24

Eg Burris v Azadani [1995] 1 WLR 1372 (CA).

The Conflict of Rights

151

correlative rights—and that such duties are not merely a ‘control device’25 or ‘incidence rule’26 of liability, the imposition of liability for conduct causing harm cannot, without more, be described as based upon a tort. Scepticism about the reality of private law duties and their correlative rights has a long pedigree, and is probably the dominant view amongst legal realists in the United States.27 This has fuelled the view that general absolute liability without fault is doctrinally possible within the law of torts. Where are the limits of the rights we all have which are potentially exigible against the rest of the world? This is perhaps the most difficult question private law must answer. The duty of others to take care not to injure me is objective because I have the same right against all others. The law draws a line where my right ends and the liberty of everyone else with respect to me begins. Perhaps confusingly, there is no single place where the border between your right and my liberty is to be found. Where my conduct is intentional I may be a wrongdoer even where wholly without fault. Demonstrating the intention to cause harm or to commit a wrong is unnecessary. If I kiss you, honestly and reasonably believing that I have permission when I do not, I am a tortfeasor. We cannot take liberties we do not have. All cases of trespass, whether to land, goods or person, involve the breach of a duty, just as much as does liability for negligence.28 The impossibility of torts of absolute liability does not and has not precluded torts where proof of fault is unnecessary. More common, or at least easier to prove, are unintentional transgressions. When is my engaging in conduct which poses some risk of injury to others potentially tortious? Economists have argued that the law employs a cost/benefit analysis,29 most famously captured in the formula of Learned Hand J: B < PL, where P is the probability of an injury, L is the magnitude of the injury, and B is the burden of cost that would be borne in avoiding the activity, including the benefits forgone if ceased.30

25 J Fleming, ‘Remoteness and Duty: The Control Devices in Liability for Negligence’ (1953) 31 Canadian Bar Review 471; Home Office v Dorset Yacht Co Ltd [1969] 2 QB 412 (CA) 426 (Lord Denning MR); JD v East Berkshire Community Health NHS Trust [2005] 2 AC 373 (HL) [94] (Lord Nicholls). 26 J Stapleton, ‘Evaluating Goldberg and Zipursky’s Civil Recourse Theory’ (2006) 75 Fordham Law Review 1529, 1532–5. 27 See for example OW Holmes, ‘The Path of Law’ (1897) 10 Harvard Law Rev 457, 471–2; L Green, ‘The Duty Problem in Negligence Cases’ (1928) 28 Columbia Law Review 1014, 1025: ‘a pious aphorism’; PH Winfield ‘Duty in Tortious Negligence’ (1934) 34 Columbia Law Review 41; W Buckland, ‘The Duty to Take Care’ (1935) 51 Law Quarterly Review 637; J Stone, The Province and Function of Law: Law as Logic, Justice and Social Control (Sydney, Maitland, 1950) 181–2; WL Prosser, ‘Palsgraf Revisited’ (1953) 52 Michigan Law Review 1, 15: ‘duty is merely a word with which we state our conclusion that there is or is not liability, it necessarily begs the question’. For further discussion see S Perry, ‘The Role of Duty of Care in a Rights-Based Theory of Negligence Law’ ch 4 of this book. 28 A v Hoare [2008] UKHL 6; Stingel v Clark [2006] HCA 37. 29 RA Posner, Law and Legal Theory in England and America (Oxford, Clarendon Press, 1996) 39–41. 30 United States v Carroll Towing Co, 159 F 2d 169 (US CA, 2d Cir, 1947) 173.

152

Robert Stevens

In fact, this stark utilitarian approach neither reflects the law nor should it do so. The barmy-ness of the utilitarian approach is reflected in the fact that taking excessive precautions is just as ‘blameworthy’ and worthy of sanction as taking too few. Both too many and too few precautions are wasteful and reduce overall welfare. There are many other ways of demonstrating the Hand formula’s inadequacy,31 but one will be given here. If accepted, the Hand Formula would allow us to treat others as means to our own ends. If I calculated that the expected benefit to me of an activity was greater than the expected harm to others, I would not be negligent in carrying it out. So, a manufacturer of a product or the provider of a service who calculated that the profits for himself from an activity would be worth more to him than the losses to third parties—for example from their personal injuries— would escape liability on this test. In fact, such a cynical disregard for the interest of others by imposing costs for free upon them is a reason for the imposition of punitive damages, not a reason for relief from liability. The benefits of an activity are relevant in deciding whether carrying on that activity can constitute a breach of a duty with respect to the plaintiff, but only if the plaintiff is a member of the class of persons who ex ante could potentially benefit from it. If this is so, the plaintiff is not being used as a means to an end. Merely taking into account the potential harm to the plaintiff, whilst ignoring the potential benefits, would be to ignore half the picture in judging what is fair as between plaintiff and defendant. So, for example, all of us as members of the public potentially benefit from the provision of emergency vehicles such as fire engines and ambulances. It is fair therefore to allow such vehicles to be driven, in an emergency, in a manner which imposes higher risks upon bystanders than would be the case if they were driven for a commercial purpose. An ambulance can be driven in a different manner from a delivery van.32 This is not, as has been suggested,33 solely because of the flashing lights, special livery and highly competent drivers employed. Identical vehicles driven for delivering apples to the shops and heart-attack victims to hospital may be driven in different ways. At the margin, the difficulty is in assessing whether the potential benefit to the plaintiff is too remote to be taken into account. We conceal this question, and others, behind the label of reasonableness. Perhaps ironically, the legal meaning of ‘fault’, the border of our rights, attracts more attention in the United States, where it is a

31 For my account of a number of other difficulties, see R Stevens, Torts and Rights (Oxford, Oxford University Press, 2007) 92–7. See further RW Wright, ‘The Standards of Care in Negligence Law’ in DG Owen (ed), Philosophical Foundations of Tort Law (Oxford, Oxford University Press, 1995); KW Simons, ‘Deontology, Negligence Tort and Crime’ (1996) 76 Boston Law Review 273; RW Wright, ‘Hand, Posner and the Myth of the “Hand Formula” ’ (2003) 4 Theoretical Inquiries in Law 14; R Perry, ‘Re-Torts’ (2008) 59 Alabama Law Review 987; BC Zipursky, ‘Sleight of Hand’ (2007) 48 William & Mary Law Review 1999. 32 Watt v Hertfordshire County Council [1954] 1 WLR 835 (CA); Marshall v Osmond [1983] QB 1034 (CA); Compensation Act 2006 s 1. 33 A Beever, Rediscovering the Law of Negligence (Oxford, Hart Publishing, 2007) 110.

The Conflict of Rights

153

jury question, than it does in England and other Commonwealth countries, where the issue is resolved by a judge. This is perhaps because the appropriate test is the subject of a direction to the jury.

VIII. Contract and Other Rights At common law, it is possible to waive all of our rights. So if I step into a boxing ring wearing gloves, my opponent commits no battery by hitting me. My guests are not trespassing when I invite them into my home, and I will not be falsely imprisoning them if I lock the door behind them. If a carrier has contracted to deliver goods to me on 1st April but I state that it is fine if they deliver the next day, they will not be in breach of contract if they deliver on the 2nd. Waiver is effective without the necessity of proving a contract or an estoppel. Waiver needs to be differentiated from alienation or transfer. In none of these examples has there been a transfer of a right. Rather, the right-holder’s manifestation of consent has conferred a privilege. This is sometimes put in Latin: ‘volenti non fit injuria’. ‘Volenti’ is not a defence to a wrong in the same way that contributory fault is. Rather, once the right is waived there will be no wrong to commit. Some rights can be transferred, including, for example, many, but not all, rights to things and some contractual rights. The rights protecting bodily safety, or free movement, or reputation cannot be transferred, but they can be waived. That a right is inalienable does not, however, necessarily tell us anything about its importance. A contractual right to a debt may be made inalienable by its terms;34 it is not thereby transformed into a right with a different order of importance. An interesting question, and one which sheds light on the common law’s hierarchy of rights, is whether it is possible by contract to bind oneself so as to be incapable of withdrawing consent to a right violation. If I contract to have a painful experiment carried out upon my person, should a subsequent change of mind on my part be recognised? Similarly, if I contract with a television company to stay in the Big Brother House for the entire run of the series and I enter the house saying nothing, then the programme-makers are not, at that point, guilty of false imprisonment. What if I later change my mind and demand to be released? The answer is that the contractual right that I have conferred upon the surgeon or the television company conflicts with rights I still have; in the first case with my right to bodily safety and in the second with my right to freedom of movement. These rights have not been transferred or destroyed by my agreement not to exercise them. The rights conflict with one another. If the patient withdraws his consent, it would be a battery to continue with the experiment. Where the

34

Linden Gardens Trust Ltd v Lenesta Sludge Disposal Ltd [1993] 3 WLR 408 (HL).

154

Robert Stevens

plaintiff has voluntarily placed himself in a position from which the occupier would need to take positive steps to free him, the plaintiff cannot demand immediate release. A miner working underground cannot demand an immediate return to the surface before his shift is finished.35 However, it would be false imprisonment to continue to detain the contestant in the Big Brother House after the point where he can be readily set free.36 We cannot contractually bind ourselves so that we cannot withdraw our consent in these cases. A modern-day Shylock could not have enforced Antonio’s promise of a pound of flesh, and he would be committing a battery if he exercised ‘self-help’, as a promisee’s rights do not override all others. This analysis provides one significant reason, but not the only one, why specific performance of contractual obligations will not be ordered.37 Where what has been promised is the provision of a service, a court order compelling the carrying out of the service would deny the promisor the right to freedom of movement. We do not specifically enforce contracts for the provision of services because we do not allow the irrevocable waiving of this right. Slavery is not permitted, even when contracted for. The contractual right conflicts with another right which has priority over it. The remedial right to specific performance is not available as a result. It is unusual to consider the liberty to break a contract without being compelled to keep it to be a (qualified) privilege, but this is what it is. The right to performance itself has not been lost, so that a breach of contract is still a wrong, but actual performance is excused. The privilege to break a contract is not, of course, absolute. The contract breaker comes under a secondary obligation to pay damages. The conflict of the two rights does not render one a nullity. An employee’s privilege to break his promise is qualified. Requiring the payment of money as damages does not create an institution of slavery. If, however, the contract stipulated for a penalty upon breach so as to effectively leave the employee no choice but to work, this would abrogate the employee’s right to freedom of movement just as ordering specific performance would. Allowing contracting parties absolute freedom to contract by the insertion of penalty clauses would conflict with the freedom of their future selves in exercising rights we consider to be of greater importance. Attempts to constrain the future liberty to withdraw, through penalty clauses, are struck down for the same sort of reasons why specific performance is not ordered. Where, by contrast, the promise is to pay a sum of money on the satisfaction of a condition, such as the performance of a service by the claimant, the court will ‘specifically enforce’ the primary obligation through the most common order made by courts in relation to a contract: an order compelling payment of the agreed sum. Compelling payment does not conflict with any other right of the promisee. 35 36 37

Heard v Weardale Steel, Coal & Coke Co Ltd [1915] AC 67 (HL). But see the difficult case of Robinson v Balmain New Ferry Co Ltd [1910] AC 295 (PC). The most common is that it is impossible because the time for performance has passed.

The Conflict of Rights

155

This is not a remedy for a wrong, and if the sum is earned it will be awarded as of right.38 It suffices that the sum is earned. When property rights come into conflict with contract rights the result is rather different. A contractual licence to use property, whether land or goods, confers no proprietary interest in the thing which is the subject-matter of the contract. A licence, unlike a lease of land, involves no transfer of rights. A licensee as such has no standing to sue in nuisance or trespass because he has no property right good against third parties.39 If the contractual licence is not by its terms revocable, the owner of the land40 or goods41 cannot withdraw his licence to use. No claim in conversion or trespass is possible against the licensee who refuses to leave the land or return the goods. The contractual right of the promisee trumps the proprietary right of the owner to exclude all others from the use of the thing. Perhaps more surprising is the common law’s conferral of a privilege to retain possession upon an unpaid seller42 or an unpaid repairer of goods. The lien of the unpaid seller or repairer is unusual, as it is a form of un-bargained-for security. There is not, on its face, any direct conflict between the right to be paid of the repairer and seller and the property right of the owner to exclude all others. Perhaps it is thought to be basic to the bargain that the owner will not get his goods back before he has paid, just as it is implicit that goods sold are of satisfactory quality, so that there is a term implied into the bargain conferring a licence to retain the goods. More difficult is where the ‘thing’ over which possession is sought is a person. Anyone who has travelled in a black taxi-cab will have seen that the doors remain locked by the driver until the fare has been paid. If the passenger insists upon being released without payment, can the driver detain him against his will? If what I have claimed to be true about the contestants of the Big Brother house is true, that even if they have expressly promised to remain it would be a tort not to allow them to leave, surely the same would follow a fortiori of the taxi’s passenger, as any promise of the passenger to remain will at best be implicit and not express? The answer appears to lie in the criminal law. Making off without payment for a service received is a criminal offence43 and any person may arrest someone they suspect with reasonable cause to be attempting to commit the offence.44 This is a statutory privilege, just as police ‘powers’ to arrest persons and seize goods would more accurately be called privileges. The taxi driver consequently commits no tort in refusing to unlock the door, at least in relation to the party obliged to pay.

38

White & Carter (Councils) Ltd v McGregor [1962] AC 413 (HL). Hunter v Canary Wharf Ltd [1997] AC 655 (HL). Hurst v Picture Theatres Ltd [1915] 1 KB 1 (CA). See RP Meagher, JD Heydon and MJ Leeming, Meagher, Gummow and Lehane’s Equity Doctrines and Remedies, 4th edn (Sydney, Butterworths, 2002) 759–66. 41 Re Cosslett (Contractors) Ltd [1998] Ch 495 (CA). 42 Sale of Goods Act 1979 s 41. 43 Theft Act 1978 s 3(1). 44 Theft Act 1978 s 3(4). 39 40

156

Robert Stevens

Contract rights can conflict with one another where bargains are successively entered into by parties where the terms conflict with one another. The solution is to conclude that the last in time prevails, the latter varying the former. The reason for the existence of the right explains the priority rule. Where a single agreement is entered into whose terms are inconsistent with one another, the court must either construe the terms in such a way that they can be reconciled or, where this is impossible, rectify the document that embodies the parties’ agreement so that it accords with what the parties intended as a matter of fact. It is tempting to conclude from the above discussion that contract rights trump property rights, but that rights to bodily safety and freedom of movement trump contract. It would be tempting to say that it follows a fortiori that property rights are trumped by rights to bodily safety and freedom of movement. As we will see below, however, the hierarchy is slightly more complex.

IX. Property Rights If my football is kicked into my neighbour’s garden, can I get it back? Normally I cannot enter the land of someone else unless they confer a licence upon me. My neighbour cannot make use of my football; it is mine and if he decides to use it he commits a tort. The law responds to this conflict between my right to the football and my neighbour’s right to exclude others from his land by conferring upon me a privilege, or licence, to enter to re-take the ball. So long as the presence of the ball on my neighbour’s property is not my fault, he cannot exclude me from his land.45 However, this licence to enter is qualified. If the football has broken a window of my neighbour’s house, I cannot enter to re-take the ball without paying for the damage. My neighbour may retain the ball until he has been paid for the damage caused. He commits no tort in refusing to hand it over. This qualification of the privilege to re-enter to take back the property is called the right to distress damage feasant.46 We see a similar conflict when the branches of a tree growing on my land overhang my neighbour’s property. I have a right that others do not cut down my tree, whilst my neighbour has a right that others do not trespass on his land, including the space above the ground. This conflict is resolved by the law conferring a privilege upon my neighbour to cut off the overhanging branches,47 so long as he returns them to me.48

45 RFV Heuston and RA Buckley, Salmond and Heuston on the Law of Torts, 21st edn (London, Sweet & Maxwell, 1996) 573–4. 46 Harry Street once invoked this privilege when a corporation bus crashed into his father’s garden. 47 Lemmon v Webb [1895] AC 1 (HL). 48 Mills v Brooker [1919] 1 KB 555 (KB).

The Conflict of Rights

157

These privileges are commonly called ‘self-help remedies’. In a loose sense this is accurate. Just as a secondary right to damages is generated by the infringement of a primary right, the privilege is generated by the conflict between the privilegeholder’s primary right and that of someone else. However, they are ‘remedies’ in quite a different sense from the right to claim damages. There is no obligation on anyone to do anything. More difficult examples of privileges, because they do not involve the direct conflict of one right with another, are those arising out of circumstances of necessity. A famous example, which has resulted in controversy, is Vincent v Lake Erie Transportation Co.49 The defendant’s ship had just finished unloading at the plaintiff ’s dock when an unusually severe storm arose. The plaintiff withdrew the permission to tie up, but the crew continued to lash the ship to the dock. The ship repeatedly banged against the dock causing US$500 of damage, which the plaintiff successfully claimed for. There is no direct conflict of rights here; what is in conflict is the dock owner’s right to exclude others from the dock and the ship owner’s interest, in the broad sense, in his vessel. The circumstances of necessity conferred a privilege to use the dock. Prima facie, although completely blameless, the defendant’s actions in intentionally tying up to the dock constituted a trespass, as they no longer had a licence to do so after the withdrawal of consent. Normally in such a case of intentional trespass the dock owner would have the privilege to untie the boat’s cables, but this privilege did not arise because of the emergency, so that if the cables had been untied the dock owner would have been committing conversion, and potentially liable for the full value of the ship if lost.50 A judge at dockside would not have restrained the boat’s owner from tying up, but the privilege to do so was qualified, so that compensation was payable calculated by reference to the loss suffered from the use. The defendant was only within the privilege, and not committing a wrong, if compensation representing the loss suffered was paid. The cause of action is not, therefore, trespass, but there would be a trespass if the money for the loss suffered was not paid. Here the defendant’s actions are justified not merely excused. Some commentators have sought to explain the result as based upon the defendant’s unjust enrichment in saving his ship at the expense of the claimant’s dock.51 This is doubtful. First, the court in Vincent awarded US$500 as damages; no attempt was made to calculate the gain the defendant made. Although, it might

49 Vincent v Lake Erie Transportation Co 109 Minn 456, 124 NW 221 (Supreme Court of Minnesota, 1910); cf Mouse’s Case (1608) 12 Co Rep 63, where the sacrifice of the claimant’s property to save that of third parties gave rise to no claim. 50 Ploof v Putnam, 81 Vt 471, 71 A 188 (Supreme Court of Vermont, 1908). 51 RE Keeton, ‘Conditional Fault in the Law of Torts’ (1959) 72 Harvard Law Review 401, 411 and 418; D Laycock, ‘The Scope and Significance of Restitution’ (1989) 67 Texas Law Review 1277, 1285–6; A Kull, ‘Rationalising Restitution’ (1995) 83 California Law Review 1191; L Smith, ‘Restitution: The Heart of Corrective Justice’ (2001) 79 Texas Law Review 2115, 2146–8; Weinrib, above n 3, at 196–203; D Friedmann, ‘Restitution for Wrongs: The Measure of Recovery’ (2001) 79 Texas Law Review 1879, 1888.

158

Robert Stevens

be replied that the claimant is only entitled to claim the gain made to the extent that he has suffered loss, such a limitation would mean that where the loss he suffers is greater than the value of the vessel, he would not be able to recover in full. Secondly, if the ship had sunk despite its being lashed to the dock, the defendant would have made no gain from their conduct, but they should still be obliged to pay for the damage caused by tying up. Vincent may be usefully contrasted with the Canadian decision in Munn & Co v ‘Sir John Crosbie’.52 The defendant’s ship was moored at the plaintiff ’s wharf to discharge coal when a hurricane-force wind forced it against the wharf, causing substantial damage. In the absence of proof of negligence, the claim failed, the Exchequer Court of Canada correctly distinguishing Vincent on the basis that the permission to moor had not been withdrawn and the defendant had at no stage deliberately tied up without permission, so as to be potentially liable in trespass. Where there is a conflict of rights in relation to a single item of property, the law is faced with a priority dispute. The basic rule here is simple: first in time prevails. Whilst, therefore, a squatter on land or a thief of goods acquires a right to a thing, this right is exigible against all persons save someone with prior title. If I seize something before you do, I have a superior right to that which you have. The reason for the existence of the right explains the priority rule. Intuitively, the easiest sort of conflict to resolve would be that between one person’s property right and another’s right to bodily safety. One would not expect a mere property right, justified at base by possessing something before someone else or deriving title from someone who did, to trump bodily safety—surely intuitively our most important right. Shockingly, this was precisely the position adopted for many years by the common law. The House of Lords decision in Robert Addie & Sons (Colliers) Ltd v Dumbreck53 is, to a modern reader, a startling example of the prioritising of the right to land over another’s right to bodily safety. Children used a field on the defendant’s land as a playground. The defendant knew of their presence, but they were trespassers. The defendant operated a haulage machine nearby, which was unguarded and easily accessible by the children. The machine was started by an employee, who did not check to see whether a child was playing upon it. A four-year-old boy was crushed to death by its operation. The House of Lords held that no duty of care was owed to the child. Liability would only be imposed if the injury had been intentionally or recklessly inflicted. The child’s commission of the wrong of trespass should not have (partially) deprived him of his right to bodily safety in this way. The principle ‘ex turpi causa

52 Munn & Co v ‘Sir John Crosbie’ [1967] 1 Ex CR 94 (Exchequer Court of Canada). See also Lords Bailiff-Jurats of Romney Marsh v Trinity House Corp (1870) LR 5 Ex 204 (Court of Exchequer); Whalley v Lancashire and Yorkshire Railway Co (1884) LR 13 QBD 131 (CA); Cope v Sharpe (No 2) [1912] 1 KB 496 (CA); Esso Petroleum Co Ltd v Southport Corp [1956] AC 218 (HL). 53 Robert Addie & Sons (Colliers) Ltd v Dumbreck [1929] AC 358 (HL).

The Conflict of Rights

159

non oritur actio’ does not apply because the child’s trespass does not form a necessary part of his cause of action.54 In many cases, the fact that the claimant is a trespasser may make his presence unforeseeable to the defendant, so that no duty of care will be owed to him in respect of activities on the land, but this was clearly not the case in Dumbreck. The case was not one where the defendant had merely failed to confer upon the plaintiff the benefit of protection from harm—the defendant had by his actions injured the plaintiff. If the activity had been carried out on the public highway, there clearly would have been liability. The result was a disgrace, and was in principle wrong. It belonged to an era of extraordinary deference to the interests of landowners in being able to exclude others from their property. Because the law confers upon the landowner the right to exclude others from the premises, it is fair to require him to take positive steps to protect others from harm. If a fire starts on my land, it is fair to require me to take steps to put it out if it poses risks to my neighbour; as I can exclude others from entering, the quid pro quo is that I must protect others from this danger. If I do not wish to be under the positive duty, I can choose to give up the right. The positive duty is not particularly onerous; a burglar injured by child’s roller-skate left at the base of stairs has no claim. However, today, ‘occupation of premises is a ground of liability and is not a ground of exemption’.55 The High Court of Australia, much to its credit, simply refused to follow Dumbreck.56 In England we had to wait for the legislature to finally remove the error made.57 Even today, it may be that if a trespasser is negligently injured by the occupier of premises his claim is subject to a different legal regime than it would have been if he had been injured by a third party.58 In principle, the claimant’s status as trespasser should be irrelevant in such a case. We cannot, however, conclude that the right to bodily safety always trumps property rights. A landowner does not have to tolerate trespassers on his land: they may be evicted, using reasonable force if necessary.59 It is not accurate to say, therefore, that we have a straightforward hierarchy, with property rights always at the bottom.

X. Third-Party Rights If D owes a duty to C1 to do x, and a duty to C2 to do not-x, D cannot act in compliance with both duties. C1’s right against D conflicts with C2’s right against D. One solution is to allow D to perform the duty owed to one party, and to pay 54 55 56 57 58 59

Tinsley v Milligan [1994] 1 AC 340 (HL). Commissioner for Railways v McDermott [1967] 1 AC 169 (PC) 186 (Lord Gardiner LC). Hackshaw v Shaw (1984) 155 CLR 614 (High Court of Australia). Occupiers Liability Act 1984. But see Revill v Newbery [1996] QB 567 (CA) 576 (Neill LJ). Dean v Hogg and Lweis (1834) 10 Bing 345, 131 ER 937 (Court of Common Pleas).

160

Robert Stevens

damages to the other. We are familiar with this position in contract law. So, if I have separately contracted to teach next week C1 in London and C2 in Glasgow, I am not relieved from paying damages to either for breach merely because it is impossible for me to fulfill both of these contractual obligations.60 However, where the inconsistent obligations have not been voluntarily assumed this may not be a satisfactory conclusion. Further, it may be that one obligation is more important than the other, and that it would be undermined if fulfilling it would render the defendant liable to pay damages to another. Although there may not be any logically necessary conflict, the potential for conflict may confer a privilege not to respect one of the rights. In some circumstance therefore, the right of one party against D may override the prima facie right of another against D, conferring a privilege upon him. For example, if an agent induces a breach of contract he can escape liability under the rule established in Said v Butt.61 A theatre critic was refused a ticket to a first night. The critic bought a ticket through a friend but was refused entry. The theatre manager was sued for inducing breach of contract. McCardie J held that an agent cannot be liable for the tort of inducing breach of contract by his principal. Although no persuasive reason for the rule is given in the judgment,62 and it was merely given as an alternative basis for the decision—as the contract was said to be vitiated due to mistake—it has been assumed to be correct on numerous occasions. For example, the Court of Appeal has applied the rule, albeit reluctantly, to a claim against a receiver who procured a breach of contract by the company over whose assets he had been appointed.63 The rule prevents the law imposing conflicting duties upon the agent. An agent is under a (fiduciary) duty to act in the best interests of his principal.64 Holding an agent liable for procuring a breach of a contract of his principal may conflict with this duty where a refusal to perform a contract is in the principal’s best interests. An agent cannot be under a duty to his principal not to proceed with a bad bargain and a concurrent duty to the counter party not to procure the breach of the contract.65 The fiduciary duty, the obligation of the agent to subordinate his interests to those of his principal, is particularly important and vulnerable to abuse. This explanation has the advantage of explaining the exceptions to the rule in Said v Butt. The privilege only applies where the agent procures a breach of contract by his principal, not where he procures a breach by a third party. Similarly, 60 Eg J Lauritzen AS v Wijsmuller BV [1989] 1 Lloyd’s Rep 148 (QB), affirmed [1990] 1 Lloyd’s Rep 1 (CA). 61 Said v Butt [1920] 3 KB 497 (KB). 62 Welsh Development Agency v Export Finance Co Ltd [1992] BCLC 148 (CA) 173. 63 ibid. 64 FMB Reynolds, Bowstead and Reynolds on Agency (London, Sweet & Maxwell, 2001) 166; Restatement of the Law, Second, Agency (St Paul MN, American Law Institute, 1957) § 387. 65 Adga Systems International Ltd v Valcom Ltd (1999) 168 DLR (4th) 351 (Ontario CA) 357. Compare the reluctance of the courts to impose a duty on a receiver to a subsequent debenture-holder where this could conflict with his duty to the debenture-holder who appointed him: Downsview Nominees Ltd v First City Corp Ltd (No 1) [1993] AC 295 (PC).

The Conflict of Rights

161

where the agent is also in breach of his duty to his principal in procuring the breach, no conflict arises and the privilege does not apply.66 There is some authority for the proposition that an agent will lose his privilege if acting in bad faith,67 but it is unclear whether this adds anything to the proposition that it does not apply where the agent is in breach of his duty to his principal in procuring the breach.68 The rule in Said v Butt is today best seen as a species of the defence of justification, which applies generally to claims for inducing breach of contract. This defence was applied by the Court of Appeal in Edwin Hill v First National Bank69 in relation to pressure applied by a bank upon a debtor not to proceed with a contract. A similar approach could be taken where the decision not to proceed is taken by a receiver appointed by the bank. The High Court of Australia in Zhu v Treasurer of New South Wales70 adopted this approach in seeking to explain the defence of justification.71 It may be queried, however, whether ‘the defence of justification’, which is the label given to this privilege, is confined to cases of conflicts between legal rights. In the famous Brimelow v Casson72 the persuasion of theatre proprietors by a performers’ protection society to break their contracts with a theatrical manager was justified on the basis that the wages paid to the chorus girls by their manager were so low that they were obliged to supplement their income through prostitution. If, following Christ, we adopted the positive formulation of the Golden Rule, we might accept that there was a moral obligation to come to the aid of the chorus girls which, whilst not enforceable in law, sufficed to generate a privilege to procure the breach.73 However, to ask the court to rule on this issue of moral compulsion is arguably too uncertain a base for a defence, and it has been doubted in obiter dicta.74 Support for the approach in Brimelow may be sought in the application of the same idea in the area of defamation. Where the defendant is under a duty to the recipient of a communication to make it, he will have a defence of qualified privilege. So, an employee may be obliged to communicate the misconduct of another employee to his employer, so that no claim in defamation is possible. The privilege is qualified, because if the defendant is motivated by malice it will not apply. In such circumstances there can be no conflict of rights.

66

Ridgeway Maritime Inc v Beulah Wings Ltd (The Leon) [1991] 2 Lloyd’s Rep 611 (Com Ct). Lathia v Dronsfield Bros [1987] BCLC 321 (QB) 323. 68 Ridgeway Maritime Inc v Beulah Wings Ltd (The Leon) [1991] 2 Lloyd’s Rep 611 (Com Ct). 69 Edwin Hill v First National Bank [1989] 1 WLR 225 (CA). 70 Zhu v Treasurer of New South Wales (2004) 218 CLR 530, [2004] HCA 56. 71 ibid, at [144]. 72 Brimelow v Casson [1924] 1 Ch 302 (Ch). 73 Pritchard v Briggs [1980] Ch 338 (CA) 416 (Goff LJ). 74 Eg South Wales Miners’ Federeation v Glamorgan Coal Co Ltd [1905] AC 239 (HL) 245 (Lord Halsbury LC). 67

162

Robert Stevens

However, this qualified privilege in the law of defamation is not limited to cases of conflicts of rights. Most obviously a former employer will have a privilege in providing a reference, although he was under no legal duty to the recipient to provide it. Moral obligations to speak appear to suffice to confer the privilege. The difficulties in this approach are articulated by Scrutton LJ: Is the judge merely to give his own view of moral and social duty, though he thinks a considerable portion of the community hold a different opinion? Or is he to endeavour to ascertain what view ‘the great mass of right-minded men’ would take?75

The former would be the better view. If all of our enforceable legal obligations are sourced in our moral rights one against another, there is nothing necessarily objectionable to some moral rights having some legal effects, even if not enforceable per se. Within the law of negligence conflicts of rights of this kind also occur but are commonly dealt with by finding that there was no duty of care. In JD v East Berkshire Community Health NHS Trust76 claims were brought by parents of young children against public authorities for psychiatric harm suffered as a result of false allegations of physical abuse made by doctors. If it is accepted, as it has been in England for many years,77 that we, each of us, have a free-standing general right not to suffer such injury, the issue was the scope of the authority’s privilege. If there had in fact been abuse, there would be no doubt a privilege to inflict such injury on the parent. Where no abuse actually took place and the authority was negligent in thinking that it had, the issue is more difficult. A majority of the House of Lords78 concluded that the need to put the interests of children first coupled with the possibility that interests of parents and children may conflict, as they obviously do where the parent is the potential abuser, meant that no duty was owed to the parents by the doctors in their investigation and diagnosis. The potential conflict between the statutory duty for the protection of the child and the parent’s private right overrode the latter. A similar problem, to which different answers are given in different jurisdictions, concerns the advocate’s ‘immunity’ from suit. The advocate frequently owes a duty to the court to behave in a way contrary to the interests of his client. Does this potential for conflict justify the absence of a duty of care to the client in these circumstances, a privilege to be careless? Certainly, the policy arguments which can be adduced in favour of advocates’ ‘immunity’ are not particularly compelling. It may be doubted whether the behaviour of advocates will change according to whether there is or is not a duty of care owed to the client, nor will public confidence in the administration of justice be adversely affected by upholding a claim against the advocate; if anything the reverse is true. The issue

75 76 77 78

Watt v Longsdon [1930] 1 KB 130 (CA) 144. JD v East Berkshire Community Health NHS Trust [2005] UKHL 23, [2005] 2 AC 373. McLoughlin v O’Brian [1983] 1 AC 410 (HL). Lord Nicholls, Lord Steyn, Lord Rodger and Lord Brown; Lord Bingham dissenting.

The Conflict of Rights

163

is whether the potential for conflict is such that the prima facie right of the client is overridden. In England, advocates’ ‘immunity’ has been abolished,79 whilst in Australia it has been retained.80

XI. Conclusion Most of this chapter has concerned privileges. If, maddened at this chapter, you lunge at me with a carving knife, I have a privilege to use reasonable force in defending myself. Here my right to bodily safety conflicts with yours, resulting in an exceptional situation where you have no right to stop my application of reasonable force. As recently established by the House of Lords,81 the scope of this privilege in private law is not identical with self-defence in the criminal law. The belief that such force in self-defence is necessary must be reasonably held, not just honestly. This is a result of the difference between the sorts of justice private law and the criminal law concern. The criminal law is concerned with sanctions for behaviour which is wrongful with respect to society. (Although this is shaped by what is wrongful with respect to one another.) Private law is concerned to ‘identify and protect the rights that every person is entitled to assert against, and require to be respected by, others’.82 Where one right conflicts with another the law must ‘strike a balance’83 between the two in a way which is unnecessary within the criminal law. As we have seen, the conflict of rights is not the only reason for the conferral of a privilege. Circumstances of necessity may confer upon me a privilege to make use of your property, and the legislature has created various privileges for disparate reasons, such as, subject to conditions, those of the police to arrest persons and seize property, or of trade unions to procure a breach of contract. One unfortunate aspect of modern legal discourse is that the language of rights in general, and human rights in particular, has been taken to mean a particular sub-set of rights: those that citizens have against the State which ‘trump’ concerns of overall social welfare. This has had the unhappy effect of upping the rhetorical ante whenever rights are invoked. Properly speaking, ‘human rights’ are both more pervasive and more humdrum than modern discourse allows. The range of human goods that in theory we could have a right that the State secures for us are as many and varied as the wit of human imagination permits. There is nothing peculiarly ‘natural’ or ‘human’ about them, unlike, say, the right not to be told lies we believe or the right not to be punched.

79

Arthur JS Hall & Co v Simons [2002] 1 AC 615 (HL). D’Orta-Ekenaike v Victoria Legal Aid (2005) 223 CLR 1 (High Court of Australia). McHugh J’s judgment is particularly strong. 81 Ashley v Chief Constable of Sussex Police [2008] UKHL 25. 82 ibid, at [18] (Lord Scott). 83 ibid. 80

164

Robert Stevens

To obtain a better understanding of rights, their scope and how conflicts are resolved in the law, it is better to focus upon the more basic examples, which have long been the subject-matter of our judge-made common law. Rights are not an innovation introduced after the Second World War by international conventions, nor were they dreamt up by the framers of the Constitution of the United States. Although the Greeks literally didn’t have a name for ‘rights’,84 all of us at all times and places have them. Resolving the conflicts between the rights of one another does not depend upon wider social policies or goals, as rights do not take the justification for their existence from such concerns. As we have seen, some rights are more important than others, one right may be adequately vindicated by an award of damages when another is not, one right-holder may be at fault for the conflict whilst the other is not, where of the same order the reason for the right’s existence may explain the order of priority, and so on. The fact that reasons of policy may not be, and are not, used to resolve the inevitable conflict between the rights of individuals does not mean that no reasons at all may be brought to bear. We should not conclude that because the solution to these conflicts is underdetermined that there is something else, for example a hidden unstated policy goal, underlying the choices made by the judges resolving them. Inevitably, there is not always necessarily one demonstrably right answer as to how these conflicts are to be settled. Irreducibly, judgement is required.

84

cf A MacIntyre, After Virtue: A Study in Moral Theory (London, Duckworth, 1985), 69.

7 Causation and the Goals of Tort Law DONAL NOLAN*

I. Introduction In this chapter, I explore the relationship between causation and what some people think of as the ‘goals’ of tort law. My contention is that it is impossible to understand or to resolve difficult questions concerning causation unless it is appreciated that these issues go to the heart of what tort law is about. The scope of the chapter is quite limited. In the first place, my analysis is limited to causation in fact—the question of whether or not there is a historical connection between the wrongful conduct of the defendant and the damage suffered by the claimant. The analytically distinct question of ‘legal causation’ (or ‘proximate cause’)—whether the historical connection between the defendant’s wrongful conduct and the damage suffered by the claimant is strong enough to justify the imposition of liability—gives rise to its own problems, but to my mind these are less closely connected to the question of what tort law is about than the issues I explore in this chapter. Secondly, within the realm of causation in fact, I am concerned only with problems of uncertainty, or proof. The difficult issue of over-determined causes (or ‘multiple causation’) therefore lies outside the scope of the chapter, since in these cases we know everything there is to know about what happened, and the problems lie elsewhere. Again, my justification for excluding these types of case is that I think that they tell us less about the ‘goals’ of tort law than do cases of evidential uncertainty.1 The scope of my chapter is also limited by the supposed goals of tort law I consider. In his very helpful analysis of the goals of tort law, Allen Linden

* I am grateful to the participants in the discussion of a draft of this chapter at the Obligations IV conference for their comments. Special thanks are also due to John Mee for helping me to clarify some of the issues raised. The usual caveat applies. 1 Though cf LA Alexander, ‘Causation and Corrective Justice: Does Tort Law Make Sense?’ (1987) 6 Law & Philosophy 1, 12–13 (arguing that the urge to allow recovery in cases of causal over-determination is grounded in retributive and distributive concerns).

166

Donal Nolan

lists six,2 some of which appear more plausible than others, but none of which appears absurd. Nevertheless, in this chapter, I limit my analysis to only two of the goals which Linden identifies: compensation and deterrence. The justification for this is that these are the two goals which are most frequently relied upon by those who adopt an instrumentalist approach to tort law, and that when judicial reasoning in tort cases takes an instrumentalist turn, it is to these two goals that reference is most often made. (As we shall see, this tendency is apparent in causation cases, although in this context judicial reliance on deterrence reasoning tends to be the more transparent.) The idea of tort law as a deterrence mechanism is easy enough to understand, though it should be made clear that in this chapter references to deterrence as a goal of tort law should be taken as referring to ‘specific’ deterrence rather than ‘market’ deterrence. A word or two about ‘compensation’ as a goal of tort law is necessary, however. At first blush it seems odd to think of compensation as an aim of tort law, since compensation is a large part of what tort law does. Compensation therefore looks like a means to an end, rather than the end itself, with the result that the claim that compensation is a goal of tort law seems nonsensical: one of the purposes of compensating is to compensate. Needless to say, however, this is not what is meant when compensation is described as a goal of tort law. Rather, ‘compensation’ is in this context shorthand for the spreading of one person’s loss throughout society, this being regarded as a legitimate societal goal because it is conducive to the overall welfare (the combination of small losses being thought to be less damaging than a single catastrophic one) and consistent with prevailing notions of social solidarity. In other words, for ‘compensation’ read ‘loss-spreading’. I contrast instrumentalist accounts of tort law with the idea of tort law as a system of corrective justice. When I use the term ‘corrective justice’ in this chapter, I intend the phrase to encompass any account of tort law which justifies tort obligations by reference to the bipolar relationship between the parties, as opposed to external social goals such as deterrence and compensation. This clearly encompasses the work of corrective justice theorists such as Ernest Weinrib,3 as well as the important recent rights-based accounts of tort law and negligence law given by Robert Stevens4 and Allan Beever.5 There are, of course, many points of disagreement between these writers, but for the purposes of this chapter what they have in common is more important. The remainder of this chapter is divided into three sections. In the first section, I consider the problem of causal indeterminacy in tort law by reference to the leading English (and, in one case, Scottish)6 cases on the subject. I begin by looking at 2

A Linden, Canadian Tort Law, 5th edn (Toronto, Butterworths, 1993) ch 1. EJ Weinrib, The Idea of Private Law (Cambridge MA, Harvard University Press, 1995). 4 R Stevens, Torts and Rights (Oxford, Oxford University Press, 2007). 5 A Beever, Rediscovering the Law of Negligence (Oxford, Hart Publishing, 2007). 6 It so happens that the leading Commonwealth cases on causal indeterminacy have arisen before the English courts, though the Supreme Court of Canada’s decisions in Cook v Lewis [1951] SCR 830 and Lawson v Laferriere (1991) 78 DLR (4th) 609 are also important landmarks, and there is some 3

Causation and the Goals of Tort Law

167

the cases themselves in some detail, and then ask whether these cases are consistent with accounts of tort law as a system of corrective justice. In the second section, I look at the loss of a chance theory, according to which some cases of causal indeterminacy in tort law can be resolved by re-characterising the claimant’s injury as the loss of a chance of avoiding injury, as opposed to the injury itself. Again, this is done by looking first at the two leading English cases in which this analysis has been put forward, and then by asking whether or not the theory holds water. I should make clear that in my view there is no fundamental distinction between the loss of a chance cases and the other cases of causal indeterminacy which have come before the courts, as is demonstrated by the fact that many of the problems with the loss of chance theory have direct counterparts in other approaches to the resolution of the causal indeterminacy problem. Nevertheless, it is, I think, convenient to separate out the loss of chance cases for the purposes of analysis. In the final section of the chapter, I look at the relationship between causal indeterminacy and the ‘goals’ of deterrence and compensation. In particular, I argue that recent developments in the English law on causation are only explicable by reference to these two goals, and that they also underlie the push towards recovery for lost chances. I finish by identifying some of the drawbacks of modifying tort law in this way in response to instrumentalist concerns.

II. Causal Indeterminacy The cases I consider in this section of the chapter are all cases in which it is impossible to know for certain whether or not there was a causal link between the defendant’s negligent conduct and an injury to the claimant which, broadly speaking, falls within the risks created by that conduct. If, notwithstanding a degree of causal indeterminacy, the claimant can still establish on the balance of probabilities that there is a causal connection, recovery is of course permitted. It is the cases where this is not possible which are problematic. There are now three such cases in which the House of Lords has allowed recovery notwithstanding the claimant’s inability to establish causation in the orthodox way: McGhee v National Coal Board,7 Fairchild v Glenhaven Funeral Services Ltd8 and Barker v Corus (UK) Ltd.9 I begin by analysing these three cases in some detail, before asking whether or not these departures from causal orthodoxy can be justified within a framework of corrective justice, or whether instead they can be explained only by reference to external goals of deterrence and compensation. recent Australian jurisprudence on loss of a chance, including the decision of the New South Wales Court of Appeal in Rufo v Hosking [2004] NSWCA 391. In the United States there is, of course, extensive case law in this area, and reference is made to this where appropriate. 7 8 9

McGhee v National Coal Board [1973] 1 WLR 1 (HL) (McGhee). Fairchild v Glenhaven Funeral Services Ltd [2002] UKHL 22, [2003] 1 AC 32 (Fairchild). Barker v Corus (UK) Ltd [2006] UKHL 20, [2006] 2 AC 572 (Barker).

168

Donal Nolan

A. The Case Law The pursuer in McGhee v National Coal Board10 was required by his employers to clean out brick kilns, in the course of which he was exposed to clouds of abrasive brick dust. Had his employers provided adequate washing facilities at his workplace, the pursuer could have washed off this grime at the end of his working day, but instead he remained caked in it during his cycle home from work. After a few days working in the kilns, the pursuer contracted dermatitis, and he sued his employers in negligence. The medical evidence indicated that the dermatitis was caused by the brick dust, and that the risk of contracting dermatitis had been increased by the failure to provide washing facilities. However, while the Court of Session had found that the failure to provide washing facilities had been negligent, it also held that the pursuer had not shown on the balance of probabilities that this would have prevented the dermatitis. The House of Lords allowed the pursuer’s appeal, with the majority of their Lordships holding that a finding that the defender’s breach of duty had materially increased the risk of injury amounted, for practical purposes, to a finding that it had ‘materially contributed’ to it, which was itself a sufficient causal connection according to the earlier decision of the House in Bonnington Castings Ltd v Wardlaw.11 This conflation of material increase in risk and material contribution was a fairly blatant judicial sleight of hand, which has since been disapproved.12 The one member of the House who did not endorse this reasoning was Lord Wilberforce, who described the majority’s approach as ‘something of a fiction’,13 and relied instead on the ‘sound principle’ that where a person had negligently created a risk, and injury had occurred within that risk, he should be liable unless he could show that the injury had some other cause.14 In other words, it was the negligent defendant, rather than the innocent claimant, who should bear the consequences of the ‘inherent evidential difficulty’.15 Clearly, the House of Lords was determined to hold the defenders in McGhee liable, even if to do so required either reliance on a fiction or an unprecedented reversal of the burden of proof. Clues as to their Lordships’ motivation are provided by Lord Simon and Lord Salmon. According to the former, not to impose liability ‘would mean that the respondents were under a legal duty which they could, in the present state of medical knowledge, with impunity ignore’.16 Lord Salmon expressed himself in almost identical terms: the approach of the courts below ‘would mean that in the present state of medical knowledge … an employer would be permitted by the law to disregard with

10 11 12 13 14 15 16

McGhee [1973] 1 WLR 1 (HL). Bonnington Castings Ltd v Wardlaw [1956] AC 613 (HL). See Fairchild [2002] UKHL 22, [2003] 1 AC 32 [65] (Lord Hoffmann). McGhee [1973] 1 WLR 1 (HL) 7. ibid, at 6. ibid. ibid, at 9.

Causation and the Goals of Tort Law

169

impunity his duty to take reasonable care for the safety of his employees’.17 In the light of these remarks, it seems plausible to suppose that the rationale for the departure from orthodox causation principles in McGhee was the perceived need to deter employers from exposing their employees to unreasonable risks in cases where it would be scientifically impossible for an employee to establish a link between that exposure and his or her subsequent injury. The second of the three cases, Fairchild v Glenhaven Funeral Services Ltd,18 involved a variant on the classic ‘two hunters problem’ raised by the Californian case of Summers v Tice.19 Each of the two defendants in Summers at or about the same time shot at a quail, and in so doing carelessly fired towards the plaintiff, who was struck by shot. Although it was clear that one of the two defendants was responsible, the plaintiff was unable to establish from which of the two guns the shot had come. The Californian Supreme Court held that the plaintiff could recover against either defendant. The burden of proof as to causation should be shifted to the defendants, as otherwise the outcome—to deny liability altogether—would be unacceptable. A couple of years after Summers v Tice, the same problem confronted the Supreme Court of Canada in Cook v Lewis.20 The majority of the court agreed that if the jury found themselves unable to decide which of the two hunters had shot the plaintiff because both had shot negligently in his direction, then both defendants should have been found liable. The principle underlying these decisions was subsequently expressed in the Second Restatement of Torts in the following terms: Where the conduct of two or more actors is tortious, and it is proved that harm has been caused to the plaintiff by only one of them, but there is uncertainty as to which one has caused it, the burden is upon each such actor to prove that he has not caused the harm.21

The issue in Fairchild was whether, if an employee was employed at different times by two employers, both of whom breached their duty to take reasonable care to prevent him inhaling asbestos dust, the employee could recover damages from the employers if he contracted mesothelioma as a result of this exposure, but was unable to establish on the balance of probabilities whether this was due to the exposure caused by one employer, the other employer, or the two together.22 Before addressing the legal issues raised by this question, Lord Bingham identified precisely what was and was not known about the causal link between the inhalation of asbestos dust and the onset of mesothelioma.23 The more dust that was inhaled, the more likely it was that mesothelioma would develop. However, it

17 18 19 20 21 22 23

ibid, at 12. Fairchild [2002] UKHL 22, [2003] 1 AC 32. Summers v Tice 199 P 2d 1 (Supreme Court of California, 1948). Cook v Lewis [1951] SCR 830. Restatement of the Law, Second: Torts (St Paul MN, American Law Institute, 1977) § 433B. See Fairchild [2002] UKHL 22, [2003] 1 AC 32 [2] (Lord Bingham). ibid, at [7].

170

Donal Nolan

was equally likely that the condition was caused by a single fibre, by a few fibres, or by many fibres, and the condition, once caused, was not aggravated by further exposure. It followed that, where an employee was exposed to asbestos dust by two employers in succession, his mesothelioma might have been caused by inhalation of a single fibre during one or the other period of employment, or by inhalation of a number of fibres during both periods of employment, and no one of these possibilities was any more probable than the others. The scenario in Fairchild was therefore slightly different from the ‘two hunters’ scenario, the difference being that while in the two hunters case it was definitely only one of the defendants who caused the harm, in Fairchild it could have been both. The House of Lords held that in this situation, the conduct of the two employers in wrongfully exposing the employee to the risk of mesothelioma should be treated as making a material contribution to the contracting of that condition.24 It did not matter if one of the two wrongdoers was not before the court, and the employee’s entitlement against either was to full damages, though the usual contribution rules would apply. According to Lord Hoffmann, the purpose of the causal requirement rules was ‘to produce a just result by delimiting the scope of liability in a way which relates to the reasons why liability for the conduct in question exists in the first place’.25 Although one could generalise about causal requirements to some extent, where the application of such a generalisation would produce a result at odds with the underlying rationale of the liability in question, then an exception to that generalisation should be recognised. In his Lordship’s opinion, this was such a situation, since the application of the general ‘but-for’ rule in this case would rob the duty to protect employees against mesothelioma of all content. In such circumstances, it was therefore sufficient for the claimant to establish that the breach of duty ‘contributed substantially to the risk that [he] would contract the disease’.26 Lord Hoffmann’s policy analysis was echoed in the other speeches. Lord Bingham accepted that it could be said to be unjust to impose liability on a party who had not been shown to have caused the claimant’s injury, but felt that this possibility of injustice was outweighed by the fact that otherwise an employer exposing his workers to asbestos dust could ensure his complete immunity from mesothelioma claims by ensuring that she employed only those previously exposed to excessive quantities of asbestos dust.27 Similarly, Lord Rodger said that if the law imposed a standard of proof that no claimant could ever satisfy, then ‘the substantive duty of care would be emptied of all practical content’ and employers could negligently expose their employees to the risk of mesothelioma with impunity.28

24 25 26 27 28

ibid, at [34] (Lord Bingham). ibid, at [56]. ibid, at [47]. Fairchild [2002] UKHL 22, [2003] 1 AC 32 [33]. ibid, at [155].

Causation and the Goals of Tort Law

171

The House of Lords felt that the outcome in Fairchild was consistent with that in McGhee.29 It was true that, unlike in Fairchild, in McGhee the defenders were the only possible wrongdoers, but Lord Rodger did not think the distinction material: The important point is that in both cases the state of scientific knowledge makes it impossible for the victim to prove on the balance of probabilities that his injury was caused by the … defendants’ wrongdoing rather than by events of a similar nature which would not constitute wrongdoing on their part.30

On the other hand, the House went out of its way to distinguish its decision in Wilsher v Essex Area Health Authority,31 where a doctor had negligently administered excess oxygen to a premature baby who had then developed a serious eye condition. Although the excess oxygen was one possible cause of this condition, on the balance of probabilities it was the result of one of a number of other possible causal agents, none of which were connected to the doctor’s negligence. In these circumstances, the House of Lords had applied the but-for test and denied the claimant damages. In Fairchild, Lord Bingham distinguished Wilsher on the grounds that while in Fairchild only a single noxious agent was involved, in Wilsher there were a number of noxious agents which could have caused the damage.32 Lord Rodger agreed that the claimant would usually have to prove that his injury was caused, ‘if not by exactly the same agency as was involved in the defendant’s wrongdoing, at least by an agency that operated in substantially the same way’, since it was essential that the claimant prove that his injury ‘was caused by the eventuation of the kind of risk created by the defendant’s wrongdoing’.33 By contrast, Lord Hoffmann believed that no principled distinction could be drawn between a single-agent and a multiple-agent case. Why should it matter if the claimant had been exposed to asbestos dust and some other dust, both of which had created the risk of cancer? Nevertheless, his Lordship agreed with the result in Wilsher on the ground that that was not a situation in which the duty in question—to take reasonable care of patients—would have been virtually drained of content unless the causal rules had been modified.34 It is noteworthy that, once again, several of their Lordships in Fairchild expressly relied on what looked very much like a deterrence rationale for the relaxation of the causation rules, namely the argument that in cases of scientific uncertainty a strict application of orthodox causation principles would empty the duty of care owed by the defendant of all substantive content. On the other hand, it also seems fairly clear that the House of Lords considered that to refuse an employee recovery in these circumstances would be unjust, although this position was not really 29 30 31 32 33 34

See, eg, ibid, at [21] (Lord Bingham). ibid, at [153]. Wilsher v Essex Area Health Authority [1988] AC 1074 (HL) (Wilsher). Fairchild [2002] UKHL 22, [2003] 1 AC 32 [22]. See also [118] (Lord Hutton). ibid, at [170]. ibid, at [69].

172

Donal Nolan

fully articulated, no justification being given for the assertion that irredeemable evidential uncertainty should operate to the detriment of the negligent defendants rather than the blameless employee. In the third case, Barker v Corus (UK) Ltd,35 the House of Lords was required to tidy up some unfinished business arising out of the Fairchild decision. The two issues that fell for decision in Barker were (1) whether the Fairchild principle applied if one element of the claimant’s overall exposure to asbestos came about through a period of self-employment; and (2) whether, if liability were imposed on a defendant under the Fairchild principle, he was entitled to an apportionment of his liability to reflect the extent of the overall exposure attributable to his breach of duty. Their Lordships answered both these questions in the affirmative, although Lord Rodger dissented on the second issue, and was only prepared to go along with the majority’s reasoning on the first issue in the light of their conclusion on the apportionment question. The leading opinion for the majority was delivered by Lord Hoffmann, who dealt with the scope of the Fairchild ruling first. The significance of the fact that some of the exposure took place during a period of self-employment was that, unlike in Fairchild, it could not be said that the mesothelioma was definitely the result of wrongful conduct by somebody. However, his Lordship did not believe that the Fairchild principle had been intended to be so limited, not least because in Fairchild the House had taken the view that the facts of McGhee fell within the Fairchild principle. This was despite the fact that in McGhee the alternative source of exposure to the brick dust had not been the negligence of a third party, but the non-negligent conduct of the defendant in exposing the claimant to brick dust during his working day. Instead, Lord Hoffmann considered the real limit on the scope of the Fairchild exception to be the requirement that ‘the impossibility of proving that the defendant caused the damage arises out of the existence of another potential causative agent which operated in the same way’.36 Departing from his position in Fairchild, Lord Hoffmann was therefore now prepared to accept the validity of the ‘single-agent’ limitation, albeit only in the more nuanced form put forward by Lord Rodger, with the result that it did not matter that the alternative potential causal agent differed from the agent for which the defendant was responsible in some causally irrelevant respect, as long as the mechanism by which it could have caused the damage was the same. The decision on this issue was more or less inevitable once it was accepted—as it had been in Fairchild—that McGhee fell within the Fairchild principle. As we shall see, though, it has been argued that there is a particularly strong case for displacing the normal rules of factual causation where the harm is undoubtedly the result of wrongful conduct, and the extension of the Fairchild principle beyond this type of case means that a justification for it must be found elsewhere. Furthermore, once it is accepted that the exception is not so limited, there is a 35 36

Barker [2006] UKHL 20, [2006] 2 AC 572. ibid, at [24]. See also at [64] (Lord Scott).

Causation and the Goals of Tort Law

173

danger that it will run out of control, and as a result it was necessary to adopt a different limitation, the ‘single-agent’ rule. However, it is difficult to see any logic in this limitation, and Lord Hoffmann’s criticisms of it in Fairchild appear justified. On a practical level, the single-agent limitation does ensure that the Fairchild principle will not swallow up the rule to which it forms an exception, but it also seems likely to lead to significant uncertainty, as there are bound to be doubts as to whether or not two different agents have operated ‘in the same way’.37 The apportionment question in Barker arose because mesothelioma is an indivisible injury, with the result that the defendants in a case like Fairchild appear to be concurrent joint tortfeasors, who would then, applying well-established principles, be expected to be jointly and severally liable for the entirety of the claimant’s damage. Since in asbestos-related litigation many of the defendants are likely to be untraceable or insolvent, there was every likelihood that the combination of the Fairchild exception and the principle of joint and several liability would lead to the burdening of a single solvent defendant with a very considerable liability, despite its being responsible for only a small proportion of the claimant’s overall exposure to asbestos. In Barker, Lord Hoffmann solved this problem by holding that in cases coming within the Fairchild principle, the damage which the defendant should be regarded as having caused was not the disease itself, but the creation of a risk of the disease coming about. And since risks are infinitely divisible, this meant that each defendant would be liable only for the share of the claimant’s exposure attributable to his negligence. Baroness Hale agreed that liability should be apportioned, but since in her view the damage was the disease (and not the risk of disease), her Ladyship could only arrive at this result by carving out an exception to the rule that liability is in solidum where the damage is a single, indivisible injury. Although the majority’s desire to apportion the liability in Barker was understandable, both the routes taken to arrive at that result were problematic. Baroness Hale’s analysis represented a clear departure from wellestablished principles on joint and several liability, while Lord Hoffmann’s recognition of risk imposition as injury is open to a number of objections, and is also difficult to reconcile with the ‘loss of a chance’ cases discussed in the third section of this chapter. Before we turn to look at these issues in more depth, however, we should note that the decision in Barker on the apportionment issue has since been reversed by Parliament, although the relevant provision (section 3 of the Compensation Act 2006) applies only to claims for mesothelioma, and hence the proportional liability approach will continue to apply in other cases. 37 B Coote, ‘Chance and the Burden of Proof in Contract and Tort’ (1988) 62 Australian Law Journal 761, 765 neatly captures the essence of the single-agent restriction when he distinguishes between McGhee and Wilsher on the grounds that in McGhee the defendant increased the intensity of the risk from a single cause whereas in Wilsher the defendant increased the number of potential causes (see also Accident Compensation Corporation v Ambros [2007] NZCA 304 [31], where Glazebrook J describes the distinction as being between cases where the breach ‘increases an existing risk factor’ and cases where it adds a new risk factor). However, we are still left with the difficult task of distinguishing single causes from multiple potential causes in borderline cases.

174

Donal Nolan

B. Corrective Justice Explanations In this sub-section of the chapter, I ask whether these three decisions of the House of Lords can be justified by reference to corrective justice considerations. I group the possible explanations into three categories: straightforward equity-based arguments for the imposition of liability; arguments based on the idea that in these causal indeterminacy cases the defendant has caused the claimant ‘evidential damage’; and finally the argument that subjection to an increased risk of injury should itself count as actionable damage in negligence. Many commentators take the view that, in at least some cases of causal indeterminacy, justice demands that a negligent defendant be held liable. This is particularly apparent in discussions of the ‘two hunters’ scenario, where it is frequently said that it would be quite wrong for the claimant to ‘fall between two stools’ because unable to establish causation against either one or other of the hunters. For example, Jane Stapleton has written that as between the innocent victim and the hunters, both of whom were careless and one of whom must have fired the fatal bullet, it might be said that the injustice to the victim in applying the orthodox common law was very great.38

Similarly, John Fleming considered the rationale of Summers v Tice to be that ‘the equities between an innocent plaintiff and two negligent defendants, each of whom could have caused his injury, favour placing the risk of proof of uncertainty on the latter’.39 There is no doubt that this position has intuitive appeal, and it seems likely to have had at least some impact on the thinking of the House of Lords in Fairchild, even if only Lord Nicholls and Lord Rodgers make express reference to the equities as between the parties as a factor in favour of the decision.40 The flaw in the argument, however, is that it refers not to each defendant as an individual but to the defendants as a collective entity. And while it seems reasonable to say that, as between the claimant and the defendants, the equities favour the claimant, unless there is some good reason why we should ‘collectivise’ the defendants in this way, we still do not have a justification for the imposition of liability as between the claimant and each individual defendant. Of course, tort law does sometimes treat defendants as a collective entity, under the rubric of the principle of joint liability, but this requires an agreement to engage in conduct fraught with unreasonable risk, an agreement which was found to be absent in Summers v Tice41 and which was clearly missing on the facts of Fairchild. Where two defendants do not act in concert one with the other, they are entitled to be dealt with as individuals, and it follows that, from a corrective justice viewpoint, we need an explanation of why this defendant ought to be liable to this claimant. In the 38

J Stapleton, ‘Lords A’leaping Evidentiary Gaps’ (2002) 10 Torts Law Journal 276, 289. J Fleming, ‘Probabilistic Causation in Tort Law’ (1989) 68 Canadian Bar Review 661, 665. Fairchild [2002] UKHL 22, [2003] 1 AC 32 [39], [155]. 41 Summers v Tice, 199 P 2d 1 (Supreme Court of California, 1948). It was also held that there had been no such concerted action in Cook v Lewis [1951] SCR 830. 39 40

Causation and the Goals of Tort Law

175

two hunters scenario, such an explanation is difficult to find. From the individual defendant’s point of view, it seems irrelevant that the other possible cause of the claimant’s injury is the negligence of another person, as opposed to non-negligent conduct or a natural event.42 And yet unless the fact that all the possible causes of the injury are tortious is significant, there is nothing to distinguish the two hunters scenario from any other case of causal indeterminacy, at least as far as the basic equities between the parties are concerned. In any case, even if the equities do favour the claimant in the two hunters scenario, this explains only one of the three House of Lords decisions we have looked at, the Fairchild case. In neither McGhee nor Barker were all the possible causes of the claimant’s injury tortious, and the attempt by the defendants in Barker to limit the Fairchild decision to such situations failed. If we turn to look at cases of causal indeterminacy in general, and we set aside for now arguments which reformulate the damage suffered by the claimant, the only remaining justification for the imposition of liability within a corrective justice framework is that, as between the innocent claimant and the admittedly negligent defendant, it is the defendant who should lose out when causation is impossible to establish. However, underlying this argument rests the assumption that somehow fault matters more than causation (without this assumption, there is no more reason to say that because the defendant is at fault, the claimant need not prove causation than there is to say that because the defendant admittedly caused the injury, the claimant need not prove that the defendant was negligent). And within a corrective justice framework this assumption is simply false, for, as Allan Beever points out, ‘from the perspective of corrective justice, the defendant is entirely innocent (with respect to the claimant) unless the claimant establishes all [the elements of the cause of action] against the defendant’.43 In any case, this argument once again fails to explain the English case law, since in both Wilsher44 and in Hotson v East Berkshire Health Authority45 (one of the loss of a chance cases discussed below) it was impossible for the claimant to prove that the admittedly negligent doctor had caused his condition, and yet recovery was denied. The only other way of justifying the English cases within a corrective justice framework is by reformulating the damage that the claimant has suffered. Two ways of doing this have been suggested. The first, which we can call ‘evidential damage’, is best explained by reference to the two hunters scenario. In this scenario, we know that one of the two defendants shot the claimant, but we don’t know which one. However, we also know that by his negligent act in shooting in the claimant’s direction, the defendant who did not shoot the claimant (D2) obscured the fact that the other defendant (D1) did. In other words, if we take D2’s negligent 42 Stevens, above n 4, at 151. See also B Hogan, ‘Cook v Lewis Re-examined’ (1961) 24 Modern Law Review 331, 344 (imposing liability in the two hunters scenario ‘fails to give to the interests of the defendant the same consideration which it readily accords the plaintiff ’). 43 Beever, above n 5, at 446. 44 Wilsher [1988] AC 1074 (HL). 45 Hotson v East Berkshire Health Authority [1987] AC 750 (HL) (Hotson).

176

Donal Nolan

conduct out of the equation, the claimant would have a straightforward claim for damages against D1. It has been argued that D2 has therefore caused the claimant ‘evidential damage’, and that this fact justifies either imposing liability on both defendants in the two hunters scenario, or at least reversing the burden of proof.46 This is an ingenious argument, which goes some way towards explaining why we might justly treat a defendant differently where the other possible cause or causes of the claimant’s injury are wrongful conduct. Ultimately, however, the argument fails to provide a justification for the English case law. As with the ‘between two stools’ argument, it works only if all the possible causes of the injury are tortious, and so the only case it might explain is Fairchild, but even there it falls down. If we revert for a moment to the two hunters scenario, the evidential damage argument must provide a justification either for reversing the burden of proof or for a stand-alone cause of action against D2. The former works fine if there are indeed two hunters, but if there are more than two it fails, since each defendant will be able to establish on the balance of probabilities that he or she was not responsible, and hence to discharge the reversed burden of proof. But the House of Lords in Fairchild did not limit its decision to cases involving only two employers (at least one of the appeals concerned more than two) and nor did it reverse the burden of proof, and hence give each employer the opportunity to exculpate itself. More generally, the fact that reversing the burden of proof works only in cases involving two defendants significantly undermines its utility as a mechanism for achieving just results in this context, there being no obvious distinction in principle between a two-hunter and a three-hunter scenario.47 Unfortunately, the evidential damage doctrine runs into even greater problems if used to justify a stand-alone cause of action. Let us suppose that in the two hunters case the claimant is able to establish that in the case of both D1 and D2 either the defendant shot him or the defendant caused him evidential damage in such circumstances as would ground a cause of action. Let us suppose further that 46 See A Porat and A Stein, Tort Liability Under Uncertainty (Oxford, Oxford University Press, 2001) esp ch VI. A similar argument was relied upon by Rand J in Cook v Lewis [1951] SCR 830. According to his Honour (at 832–3), the burden of proof should be shifted to the defendants as they had ‘violated not only the victim’s substantive right to security, but ... also culpably impaired the latter’s remedial right of establishing liability’. For related arguments that centre on the creation of uncertainty as the basis of liability, see A Ripstein, Equality, Responsibility and the Law (Cambridge, Cambridge University Press, 1999) 78–80; and A Beever, ‘Cause-in-Fact: Two Steps Out of the Mire’ (2001) 51 University of Toronto Law Journal 327. 47 A possible way round this would be to require the defendants to discharge the reversed burden of proof without recourse to purely statistical evidence. In other words, it would not be enough for a defendant in a three hunters case to point only to the statistical improbability of his having shot the claimant—he would have to adduce other, more particularistic, evidence (for example, that the markings on the bullet show that it did not come from his gun). Although there is as yet no sign of the English courts discounting purely statistical evidence in causation cases, the adoption of such an approach would give more bite to the reversal of the burden of proof technique, an advantage of which is that it may serve to flush out evidence the defendants might otherwise withhold (see Summers v Tice, 199 P 2d 1 (Supreme Court of California, 1948) 4 (Carter J): ‘Ordinarily defendants are in a far better position to offer evidence to determine which one caused the injury’). I am grateful to Stephen Perry for alerting me to these possibilities.

Causation and the Goals of Tort Law

177

we are prepared to accept that the damages in the evidential damage action would be the same as the damages in the personal injury claim, as in principle they ought to be. In these circumstances, it seems clear to me that we ought to allow the claimant to recover in full from either defendant, since we know for sure that one way or another he has a valid claim against that defendant of value £x.48 The trouble is, however, that the case for such a stand-alone claim for evidential damage seems weak. As has been pointed out,49 it is unlikely that a claim would exist for the negligent loss of hospital or police records essential to a cause of action against a third party. This is because evidential damage claims would be for pure economic loss, which is generally only recoverable in English tort law where the defendant has voluntarily assumed a responsibility towards the claimant. Added to this is a remoteness objection, since it is highly unlikely that in the two hunters scenario D2 could reasonably have foreseen that his negligence would cause the claimant to suffer from evidential damage, as opposed to personal injury.50 The other way of reformulating the claimant’s damage is to describe it as the increased risk of injury rather than the injury itself. As we have seen, this was what Lord Hoffmann did in Barker to justify apportioning liability between the tortfeasors in cases falling within the Fairchild principle. Clearly this approach is consistent with the outcome in Barker itself. It is not, however, consistent with the outcome in McGhee, where the employer was held liable for the dermatitis, not the increased risk of contracting it brought about by the absence of washing facilities, nor with the outcome in Fairchild, where again liability was imposed in solidum. Still, we must assume that in future the English courts will apply Barker in cases not caught by section 3 of the Compensation Act 2006, and therefore that where liability is imposed in cases of this kind, it will be limited to the degree of additional risk created by the defendant. Unfortunately, however, the reformulation of the damage as the increased risk of injury is no more convincing than the ‘evidential damage’ doctrine. Four problems can be identified. The first is that in Barker, Lord Hoffmann insisted—not surprisingly, bearing in mind the

48 cf Stevens, above n 4, who argues (at 150) that ‘it could not be shown on the balance of probabilities against either defendant individually that he had caused a right not to suffer evidential damage to be infringed’. 49 ibid. 50 ibid. Stevens also argues that the evidential damage argument suffers from a circularity problem. He accepts that D2’s action has obscured the fact that D1 shot the claimant, and so has robbed the claimant of the action he would otherwise have had against D1. However, he argues that if we use this fact to justify imposing liability on both defendants (which we must, since of course we do not know which is which) then it is no longer the case that D2 has caused the claimant loss, since the claimant now has a perfectly valid claim against D1. In his words, ‘if [the argument] works against both [defendants], it works against neither’ (ibid). However, this is not a circularity problem, but an infinite regression, whereby in position x the logic of argument y drives us to position z, but then in position z the logic of argument y no longer applies, so that we are bounced back to position x, where argument y kicks in again, bouncing us back to position z, and so on, ad infinitum. Hence, if we accept the evidential damage argument we are driven to a conclusion which is inconsistent with the argument, but if we abandon that position (and deny recovery), the force of the argument returns. In such cases, we are, as a matter of logic, free to choose between either of the competing positions.

178

Donal Nolan

implications of the alternative—that liability would arise only if the claimant did in fact contract mesothelioma,51 but it is difficult to see why this should be the case if the risk really is the damage.52 The second is that it is difficult to see why the imposition of risk is itself harmful. Of course, we do not want to be subjected to risks of injury, but that is because we do not want to suffer physical injury; if the risk does not cause us to suffer such injury then it is surely harmless,53 and of course in these cases the whole point is that the claimant is unable to prove that the risk created by the defendant has caused injury.54 The third problem is that if we accept that the imposition of risk is itself a harm, and that this provides the justification for the results in McGhee, Fairchild and Barker, we then need a good reason for not imposing liability for the imposition of risk in Wilsher. As we have seen, the single-agent/multiple-agent distinction does not provide such a reason. The final objection is that it is assumed in all these cases that we know for certain that each defendant has materially increased the risk of injury to the claimant, but this assumption is questionable. If we take the scenario in Fairchild, for example, it is possible that a single asbestos fibre caused the claimant’s mesothelioma, and, if this is the case, we have no way of knowing when that particular fibre was ingested. If in fact it was inhaled before the claimant began working for the defendant, then by the time the defendant negligently exposed the claimant to asbestos, the claimant was already doomed, and the negligence of the defendant not only did not, but could not, have made any difference. The argument is even stronger in the two hunters scenario, since here it seems positively unlikely that the hunter who did not in fact shoot the claimant increased the risk of injury to the claimant by negligently firing in his direction. We can test this by asking whether, at the moment the defendant pulled the trigger, there was any possibility of his shot hitting the claimant. Even if we assume a degree of indeterminism in the processes involved, for this to be the case, his gun would still have to be pointing so close to the claimant that in the split second between the defendant pulling the trigger and the bullet reaching the claimant’s position the claimant might

51

Barker [2006] UKHL 20, [2006] 2 AC 572 [48]. See HM Hurd, ‘The Deontology of Negligence’ (1996) 76 Boston University Law Review 249, 263 fn 34 (‘if risk-taking … is not only culpable, but wrongful, then the requirement that a defendant cause harm is morally irrelevant’); and JL Coleman, Risks and Wrongs (New York, Cambridge University Press, 1992) 399. 53 This may not always be the case. A claimant who knows that he has been exposed to the risk of future harm may suffer anxiety as a result, and in severe cases this might result in psychiatric injury, as was alleged to have happened in one of the appeals heard by the House of Lords in Rothwell v Chemical & Insulating Co Ltd [2007] UKHL 39, [2007] 4 All ER 1047. However, where this happens the actionable damage is not the increased risk, but the psychiatric injury it causes. See further on this and related issues, J Steele, Risks and Legal Theory (Oxford, Hart Publishing, 2004) 116–18. 54 See further Hurd, above n 52, at 263, arguing that for an unrealised risk to count as a harm, risks would have to possess a moral ontology separate from any physical or psychological harms that materialise from them, ‘a highly implausible metaphysical claim’. As she goes on to point out, risks are better understood as ‘probabilistic calculations about future events derived inductively from past experience’, in other words as epistemic constructs, not ontological entities. As we shall see below, the same is true of the loss of a statistical ‘chance’ of recovery in the medical negligence context. 52

Causation and the Goals of Tort Law

179

move directly into the line of fire. If not, then by pulling the trigger the defendant imposes no risk whatsoever on the claimant, since there is no way that by doing so he can cause him injury. It could be argued that this is to adopt the wrong perspective.55 From the point of view of the defendant at the time of the breach of duty, a risk was imposed on the claimant (as he could not have known that his bullet would miss), and indeed this must always be the case where the defendant has been negligent towards the claimant, as he has to be in order to be liable.56 But even if we assume that this was the perspective the House of Lords had in mind when they said that in cases falling within the Fairchild principle, the defendant was deemed to have caused the injury when he could be shown to have materially increased the risk of its happening, this switch from objective risk to subjective risk cannot rescue the risk as damage idea. This is because even if we accept the premise that an unrealised risk is a form of harm, this claim can surely be sustainable only if the defendant’s action did in fact create a risk for the claimant; it would be a very strange form of ‘damage’ that depended for its existence on the knowledge available to the defendant at the time of the breach. As we shall see, a mirror image of this objection to the idea of risk as damage can be found in the debate on loss of a chance, where (as we shall see) it has been argued that only if the objective probability of the claimant recovering was greater than zero can the claimant’s loss of a ‘chance’ plausibly be regarded as a form of harm.

III. Loss of a Chance In this section of the chapter, I consider one particular response to the problem of causal indeterminacy that has proven popular in medical negligence cases: the argument that the negligence of the defendant doctor has caused the claimant patient to lose a chance of a positive outcome. The argument is closely related to the argument from risk imposition which we have just looked at, since taking away a chance of a positive outcome and imposing a risk of injury are two sides of the same coin (we usually say that the claimant has lost a chance where the defendant was obliged to take reasonable steps to improve her position, and that the claimant has been subjected to a risk of injury where the defendant was under a duty to take reasonable steps not to worsen her position). It is not surprising, therefore, to find that the loss of a chance argument is subject to similar 55 See eg S Perry, ‘Risk, Harm, Interests, and Rights’ in T Lewens (ed), Risk: Philosophical Perspectives (New York, Routledge, 2007) 197, where it is argued that for many moral purposes ‘the most appropriate epistemic perspective for determining what risk one person has imposed on another will be the perspective either of the actor himself/herself or of some idealized agent, such as a reasonable person, who is imagined to be in the actor’s situation at the time of acting’. 56 See Palsgraf v Long Island Railroad Co, 162 NE 99 (New York CA, 1928); and Bourhill v Young [1943] AC 92 (HL).

180

Donal Nolan

objections. However, before we evaluate the argument, it will be helpful to take a look at the two English cases in which it has been most fully considered, Hotson v East Berkshire Area Health Authority57 and Gregg v Scott.58

A. Hotson and Gregg The claimant in Hotson was a boy who had fallen from a tree and injured his hip. He was taken to the defendant’s hospital, but the injury was not correctly diagnosed, and he was sent home. After five days of serious pain, the claimant returned to the hospital where a proper diagnosis was made, and appropriate emergency treatment given. Unfortunately, the nature of his injury was such that a serious and permanent disability was likely to develop, and this is what eventually happened. The claimant sought damages from the defendant, which admitted that the failure to make a correct diagnosis in the first place had been negligent, but denied that this had been the cause of the disability. At the trial of the action there was a clash of medical evidence on the causation issue, but in the end the judge reached the conclusion that while there was only a one-infour probability that the boy would have avoided developing the condition even if he had been given prompt treatment, the defendant’s negligence had reduced that probability to zero, and that the defendant should therefore pay the claimant damages for the loss of the chance of recovery, those damages being one quarter of the damages which would have been awarded for the disability itself. That decision was affirmed by the Court of Appeal,59 but the House of Lords unanimously allowed the defendant’s appeal. In their Lordships’ opinion, it was not correct to say that the defendant’s breach of duty had deprived the claimant of a one-in-four chance of recovery. On the contrary, the evidence was that the state of the claimant’s leg when he first arrived at the hospital determined one way or the other whether prompt treatment would have brought about a full recovery or whether it would have made no difference. Since the state of the claimant’s leg at that time was a matter of past fact, it was up to the claimant to show on the balance of probabilities that treatment would have made a difference, and since he could not do that, his action failed. In the words of Lord Mackay, [i]n the circumstances of this case the probable effect of delay in treatment was determined by the state of the facts existing when the plaintiff was first presented to the hospital. It is not, in my opinion, correct to say that on arrival at the hospital he had a 25 per cent chance of recovery. If insufficient blood vessels were left intact by the fall he had no prospect of avoiding complete avascular necrosis whereas if sufficient blood vessels

57 58 59

Hotson [1987] AC 750 (HL). Gregg v Scott [2005] UKHL 2, [2005] 2 AC 176 (Gregg). Hotson [1987] 2 WLR 287 (CA).

Causation and the Goals of Tort Law

181

were left intact on the judge’s findings no further damage to the blood supply would have resulted if he had been given immediate treatment, and he would not have suffered the avascular necrosis.60

Similarly, Lord Ackner said that the loss of a chance issue ‘cannot arise where there has been a positive finding that before the duty arose the damage complained of had already been sustained or had become inevitable’.61 Lord Mackay made it clear that he was not ruling out the possibility that a claimant could succeed by proving loss of a chance in a medical negligence case,62 and Lord Bridge also left this question open.63 All that Hotson established, therefore, was that the loss of a chance argument will fail where, on the balance of probabilities, the fate of the claimant was determined at the time of the defendant’s breach of duty. The House of Lords recently revisited the loss of a chance issue in another medical negligence case, Gregg v Scott.64 When the claimant in Gregg had visited the defendant general practitioner about a lump under his arm, the defendant assumed it to be benign, and negligently failed to refer the claimant to a specialist for further investigations. The lump turned out to be cancerous, and the defendant’s negligence meant that the start of treatment was delayed for nine months. During those nine months, the cancer had spread, and it was estimated that as a result the claimant’s chance of being cured (cure being defined as survival for at least 10 years) had been cut from 42 per cent to 25 per cent. The claimant sought damages for the diminution in his prospects attributable to the delay, but both the trial judge and the Court of Appeal dismissed his claim because, on the balance of probabilities, he would not have survived in any case. By a majority (Lord Nicholls and Lord Hope dissenting), the House of Lords agreed. Lord Hoffmann was not convinced by the claimant’s argument that even though he could not prove, on the balance of probabilities, that the defendant’s negligence had deprived him of a cure, the loss of the chance of that outcome ought itself to be recognised as a form of actionable damage. The full logical implication of this argument was that damages should be awarded in all cases in which the defendant might have caused an injury and had increased the likelihood of the injury being suffered. Adhering to this position would require the House to abandon a good deal of authority (including Hotson and Wilsher), and there were ‘no new arguments or change of circumstances which could justify such a radical departure from precedent’.65 Lord Phillips agreed that the appeal should be dismissed. Close analysis of the difficulties surrounding the statistical evidence relied on in the case only served to illustrate that the exercise of assessing the loss of a chance in a clinical negligence case was a difficult one. By contrast, establishing the consequences

60 61 62 63 64 65

Hotson [1987] AC 750 (HL) 785. ibid, at 792. ibid, at 786. ibid, at 782. Gregg [2005] UKHL 2, [2005] 2 AC 176. ibid, at [85].

182

Donal Nolan

of a doctor’s negligence on the balance of probabilities was likely to be very much easier. This practical consideration was a policy reason for rejecting the loss of a chance approach, since a ‘robust test which produces rough justice may be preferable to a test that on occasion will be difficult, if not impossible, to apply with confidence in practice’.66 Baroness Hale arrived at the same conclusion. Whether the loss of a chance argument should be accepted was, in her view, a matter of legal policy, and in the end she concluded that the case for change had not been made out. In particular, she was concerned that it would be unjust if claimants were given a free choice as to whether to formulate the damage as the loss of the chance of a positive outcome, or the loss of the outcome itself, since this would mean that in cases where the balance of probabilities test was satisfied, they would recover full damages, and in cases where it was not they would still recover proportionate damages. And yet if proportionate recovery were introduced across the board, that would ‘surely be a case of two steps forward, three steps back for the great majority of straightforward personal injury cases’.67 Besides, proportionate recovery would add considerably to the complexity of the issues to be resolved, and this would make settlements and trials much more difficult. All in all, the problems to which acceptance of the loss of a chance approach would give rise outweighed any policy advantages it would bring. Lord Nicholls, dissenting, felt that in cases such as Fairchild the courts had demonstrated a willingness to leap an evidentiary gap when overall justice plainly so required. This was not appropriate in cases (such as Hotson) where the patient’s actual condition at the time of the negligence was determinative of the answer to the hypothetical ‘but-for’ test of causation, but it was appropriate in less straightforward cases, where the limitations on scientific and medical knowledge meant that a full identification of the patient’s condition at the time of the breach of duty would not necessarily provide such an answer. And the present case was of the latter kind, since identifying the nature and extent of the claimant’s cancer at the time of the mistaken diagnosis would not provide a simple answer to the question of what the outcome would have been if he had been treated promptly. His Lordship also drew on the ‘empty duty’ argument which had proved so influential in Fairchild: if claims for loss of a chance of recovery were not allowed, then the duty of care of a doctor faced with a patient with a poor prognosis would be ‘hollow’.68 Lord Hope agreed with Lord Nicholls that the appeal should be allowed, and although his reasoning was influenced more by a second argument which the claimant had put forward,69 he made it clear that he also accepted the loss of a chance argument. 66

ibid, at [170]. Gregg [2005] UKHL 2, [2005] 2 AC 176 [225]. 68 ibid, at [4]. 69 This was that the delay in diagnosis had undoubtedly caused some physical injury to the claimant—namely the enlargement of the tumour—and that when it came to quantifying the damages for that injury, an award should be made for the effect it had had on the claimant’s prospects. The 67

Causation and the Goals of Tort Law

183

Before a broader assessment of Hotson and Gregg is made, we should note that Gregg was not a case, like Hotson, where medical negligence had increased the probability of an adverse outcome which had then materialised, since the claimant in Gregg was still alive when the case was heard in the House of Lords. This meant that Lord Phillips was able to leave open the question of whether proportionate recovery should be permitted where an adverse outcome had in fact materialised. It follows that Gregg has not shut the door on claims for loss of a chance, since, although no such reservations were apparent in the opinions of Lord Hoffmann and Baroness Hale, the appeal was dismissed by only a bare majority of their Lordships. If we take Hotson and Gregg together, therefore, all we can say for certain on the English authorities is that a claimant will not be able to recover for a lost chance in a medical negligence case where either, on the balance of probabilities, his condition was such at the time of the breach of duty that the threatened injury was already inevitable, or where the threatened injury has yet to materialise. Where neither of these conditions is satisfied, recovery for loss of a chance remains an open question.

B. Evaluating the Loss of a Chance Argument Before we turn to an evaluation of the loss of a chance theory, there are three red herrings which must be dispensed with. The first is the fact that in contract there is no problem with recovery for lost chances, as is shown by the beauty contest case, Chaplin v Hicks.70 The short answer to this is that breach of contract is actionable per se, and hence the claimant does not have to show (as he does in negligence) that on the balance of probabilities the defendant has caused him damage.71 It follows that in contract cases, recovery for lost chances is a question of quantification, rather than liability, and it is well established that where a hypothetical (as opposed to a past fact) is in issue at the quantification stage, the balance of probabilities test is abandoned, and proportionate recovery permitted. That takes us to the second red herring, which is recovery for lost chances at the quantification stage in personal injury or death cases. Again, the explanation is that this is permitted at this stage where questions of past fact are not in issue.72 The third red herring is claims for loss of a chance of an economic gain, as in Allied Maples Group Ltd v Simmons & Simmons (a firm),73 where a solicitor’s negligence caused a client to miss out on the opportunity of a favourable business transaction. This difficulty with this argument is that it was not obvious that the delayed diagnosis had caused the cancer to spread: it was more likely than not that this would have happened in any case: see NJ McBride and R Bagshaw, Tort Law, 3rd edn (London, Pearson, 2008) 563. 70 Chaplin v Hicks [1911] 2 KB 786 (CA). See also Kitchen v RAF Association [1958] 1 WLR 563 (CA). 71 See M Lunney, ‘What Price a Chance?’ (1995) 15 Legal Studies 1, 4. See also Sellars v Adelaide Petroleum NL (1994) 179 CLR 332 (High Court of Australia) 359 (Brennan J). 72 See eg Davies v Taylor (No 1) [1974] AC 207 (HL). 73 Allied Maples Group Ltd v Simmons & Simmons (a firm) [1995] 1 WLR 1602 (CA).

184

Donal Nolan

time the action is in tort, and the issue goes to liability rather than quantum, but none the less this type of case is of little relevance to the medical negligence context because a chance of economic gain is itself something of economic value, and hence its loss is itself an economic loss. As Tony Weir has written, [l]osing a chance of [a monetary] gain is a loss like the loss of the gain itself, alike in quality, just less in quantity: losing a chance of not losing a leg is not at all the same kind of thing as losing the leg.74

Since the loss of a chance argument is closely related to the idea of risk imposition as damage, it is not surprising to find that it is subject to similar objections. The first is that if the loss of the chance really is the damage—in Jane Stapleton’s terms, the ‘gist’ of the action75—then recovery should be allowed whether or not the threatened injury has actually materialised. However, it seems to be widely accepted, even by proponents of the loss of chance theory,76 that purely speculative claims ought not to be allowed before some physical injury has occurred, and in the end the absence of such an injury was the basis of the rejection of the claim in Gregg. Limiting the application of the loss of chance doctrine to cases where the injury has materialised amounts to a tacit admission that the real damage is the physical injury, rather than the chance of avoiding it, in which case the theory is merely a fiction designed to circumvent the causal indeterminacy problem.77 The second objection is that, if recovery is to be allowed on a loss of a chance basis, then the quid pro quo must surely be that the damages of a claimant who can establish on the balance of probabilities that the defendant caused his injury should be discounted to account for the (less than even) chance that his injury would have happened anyway. And yet it seems impossible to achieve this outcome, since there is nothing to stop the claimant in such a case from saying that the gist of his claim is the injury itself, and that since he can prove causation of that damage, he is entitled to full recovery. Jane Stapleton has argued that under the ‘vicissitudes of life’ principle a discount would in any case be made to reflect the chance that the claimant would have suffered the injury at some time even without the defendant’s breach, and that ‘in many cases, if not all, this approach will give a result indistinguishable from that produced had the claim been framed in

74 T Weir, An Introduction to Tort Law, 2nd edn (Oxford, Clarendon Press, 2006) 80. See also Coote, above n 37, at 772; Sellars v Adelaide Petroleum NL (1994) 179 CLR 332 (High Court of Australia) 355 (Mason CJ, Dawson, Toohey and Gaudron JJ) (‘in a case such as the present, the applicant shows some loss or damage was sustained by demonstrating that the contravening conduct caused the loss of a commercial opportunity which had some value ... the value being ascertained by reference to the degree of probabilities or possibilities’). 75 J Stapleton, ‘The Gist of Negligence, Part II: The Relationship Between “Damage” and Causation’ (1988) 104 Law Quarterly Review 389. 76 See, eg, ibid, at 395, where Stapleton says that it might be possible to restrict the argument to cases where the outcome has occurred or is certain to occur. 77 See T Hill, ‘A Lost Chance for Compensation in the Tort of Negligence by the House of Lords’ (1991) 54 Modern Law Review 511, 517–18; Lunney, above n 71, at 5–6.

Causation and the Goals of Tort Law

185

terms of loss a chance’.78 With respect, however, it is difficult to see the basis for this claim that the two types of discount are likely to produce anything like the same result. Let us suppose, for example, that the probabilities in Hotson were reversed, so that the probability that prompt treatment would have brought about a full recovery was 75 per cent instead of 25 per cent. Full application of the loss of chance idea would require that in these circumstances the boy’s damages be discounted by 25 per cent to reflect the probability that the injury would have come about in any case, but clearly any discount under the vicissitudes of life principle, to reflect the possibility that even after a full recovery the boy would have developed a similar disability in the future, would be far less. The final objection to the loss of chance theory is the mirror image of the final objection to the ‘risk as damage’ idea, namely that in some, if not all, of the so-called loss of a chance cases the claimant has not really lost any chance at all. Where the occurrence or otherwise of the physical injury is causally determined at the time of the defendant’s breach of duty, the claimant has not lost a personal chance, but only a statistical chance.79 This was the basis of the decision in Hotson, where either the injury sustained in the fall ruptured so many blood vessels that necrosis was inevitable, or it did not. In other words, either the injury was treatable, in which case the breach caused the necrosis, or the injury was not treatable, in which case the breach did not cause any damage at all. What the breach did not cause was the loss of a ‘chance’ of avoiding necrosis that the claimant had when he turned up at the hospital. Furthermore, it would be quite misguided to recognise the loss of a mere statistical chance as actionable damage in negligence, since, as Stephen Perry has pointed out, it would be very odd to suppose that ‘the hypothetical situation of 100 people who are assumed to have independently suffered the “same” injury could in some sense be an asset possessed by an actual individual who has in fact incurred such an injury’.80 The same argument against loss of a chance recovery has been put forward by Allan Beever,81 though Beever refers not to statistical and personal chances, but to objective and epistemological probability. Objective probability, he says, is ‘a feature of the world’; an event has an objective probability if ‘there really is a chance that it will or will not happen’.82 By contrast, epistemological probability ‘is not a feature of the world but reflects only our understanding of the world’.83 Suppose, for example, that you negligently 78

Above n 75, at 398. Hill, above n 77, provides an excellent exposition of this argument, as does SR Perry, ‘Protected Interests and Undertakings in the Law of Negligence’ (1992) 42 University of Toronto Law Journal 247 and ‘Risk, Harm, and Responsibility’ in DG Owen (ed), Philosophical Foundations of Tort Law (Oxford, Oxford University Press, 1995) 321. See also S Perry, ‘The Role of Duty of Care in a Rights-Based Negligence Law’ ch 4 of this book, text accompanying n 45ff. 80 Perry, ‘Protected Interests’, above n 79, at 255. See also Hill, above n 77, at 516 (‘a statistical chance has no real value, it is therefore not a loss and consequently not compensatable’). 81 A Beever, ‘Gregg v Scott and Loss of a Chance’ (2005) 24 University of Queensland Law Journal 201. 82 ibid, at 207. 83 ibid. 79

186

Donal Nolan

destroy my lottery ticket, and that I have no record of the number. If this happens before the draw is made, then it is possible (depending on whether we accept a degree of indeterminism in such matters) that there was an objective probability of, say, one in a million, that my ticket would have been the winning ticket, but if the same thing happens after the draw has been made, this is no longer a possibility. Either it was the winning ticket or it was not; we just don’t know. If we now say that there was a one in a million ‘chance’ that it was the winning ticket, this cannot therefore be a reference to an objective chance, but only to an epistemological one. And as Beever says, the loss of an epistemological ‘chance’ is not a real loss: ‘it is simply a form of factual uncertainty’.84 It follows that to establish actionable damage, the claimant in a loss of a chance case would have to show that at the time of the defendant’s breach the later occurrence of his injury was still ‘up in the air’, as opposed to determined one way or the other. Only if this is the case are we in the realms of personal (or objective) chances, as opposed to statistical (or epistemological) ones. This takes us into very deep waters indeed, however, for while in some scenarios—such as Hotson—it will be clear that this is not the case, in others it will be impossible to know either way. On the facts of Gregg, for example, was the claimant’s fate sealed, one way or the other, when he first went to see his doctor? Or could subsequent events have also played a part in the development of the lymphoma?85 If, as Beever argues,86 the law should take a deterministic stance, and assume that in all such cases the outcome was already inevitable, then there is no scope for the loss of a chance theory at all. But even if we do not go this far, the impossibility of knowing whether a subsequent outcome was or was not determined at a particular moment in time makes the distinction unworkable. In any case, since it is up to the claimant to establish that the defendant caused him damage, and since he can do this only by proving the impossible—that the outcome was not determined at the time of the breach—there is nothing to be lost by adopting Beever’s position, and hence rejecting the loss of a chance argument in all cases. One final argument, this time in favour of loss of a chance recovery, should be noted. As we have seen, in contract cases the fact that the claimant does not need to establish actionable damage means that recovery for lost chances is straightforward. It follows that if there is a contractual relationship between the two parties, there is no reason why a patient who has lost a personal chance of recovery as a result of a doctor’s negligence should not be able to recover appropriate damages. Robert Stevens has suggested that the same is true whenever the defendant has voluntarily assumed a responsibility towards the claimant—as will commonly be 84 ibid, at 210. See also at 212 (it cannot be right for the law to compensate for the loss of such a chance, ‘as that would be to compensate for the “loss” of something that does not exist’). 85 In the Court of Appeal in Gregg, Mance LJ took the view that subsequent events might also have played a part, and hence that the case was distinguishable from Hotson: see [2002] EWCA Civ 1471, (2003) 71 BMLR 16. 86 Beever, above n 81. Stephen Perry also seems to incline towards this position: see ‘Risk, Harm, and Responsibility’, above n 79, at 337.

Causation and the Goals of Tort Law

187

the case in the medical negligence context—since in such cases the lost chance is consequent upon the infringement of a right to careful treatment created by the assumption of responsibility.87 Superficially, this is an attractive argument, albeit not one yet considered by the English courts. However, it should not be pushed too far. Stevens seems to think that the argument can be applied both to personal and statistical chances, since he refers to the Hotson case as one where this reasoning might have been used to justify recovery. The contract cases all concern personal chances, however, and the difficulty in a statistical chance case like Hotson is that even if the claim had been made in contract, the court would still have been faced with a question of past fact, requiring resolution on the balance of probabilities. And while Stevens points out that even in a case like Hotson the provision of careful treatment is something for which the claimant would pay,88 it does not follow from the fact that careful treatment has a market value that the failure to provide it causes the claimant loss recoverable in tort.89

IV. Compensation and Deterrence I have argued that it is very difficult, if not impossible, to justify a relaxation of the rules in causal indeterminacy cases by reference to arguments rooted in corrective justice. It should come as no surprise that the same difficulties do not emerge if the justification is sought instead in ‘goals’ of compensation and deterrence. After all, if it were true that the function of tort law is to compensate and to deter, there would be no obvious reason for having a causation requirement at all. A compensation goal looks only to the loss suffered by the claimant, and a deterrence goal only at the defendant’s negligent behaviour. Either way, causation, which links the two, is entirely redundant, and indeed positively counter-productive, since the enforcement of a causation requirement will mean that claimants who have suffered losses will go uncompensated, and that defendants who have acted negligently will go unpunished, with the result that tort law will under-deter.90 It does not necessarily follow that instrumentalist concerns of this kind lie behind the drive to ‘water down’ the requirement of factual causation; the motivation might instead be a mistaken belief that doing so will bring about more just outcomes in individual cases. In fact, however, there is considerable evidence from

87

Stevens, above n 4, at 50. ibid. If there were a contract between the parties, and the claimant had paid for careful treatment, then there would be a stronger case for damages, although calculating them would raise difficulties, and there would certainly be no justification for doing so by reference to a fictional lost chance of recovery. 90 According to two leading lawyer economists, ‘the idea of causation can largely be dispensed with in an economic analysis of torts’: WM Landes and RA Posner, The Economic Structure of Tort Law (Cambridge MA, Harvard University Press, 1987) 229. 88 89

188

Donal Nolan

both the case law and academic commentaries that deterrence and compensation have been significant motivating factors in this connection. This claim is more difficult to substantiate in the case of compensation, but it is important to note that the cases in which the causation rules have been relaxed have all involved employees suing for diseases contracted in the workplace, a context in which there has traditionally been a powerful drive to compensate, even when negligence has not been established. Compensation as a goal may also explain the strong intuition that a claimant who can show that his injury is the result of tortious conduct, but who is unable to pinpoint the tortfeasor, ought to be awarded damages. I have argued that from a corrective justice viewpoint, this factor ought not to operate to the detriment of an individual defendant, but it does seem to give the claimant a particularly powerful societal claim to have his loss spread more widely, and of course the easiest way of doing this is via the defendant’s liability insurance.91 This certainly seems to be the view of Parliament, which responded to the decision in Barker v Corus by legislating for recovery in solidum in mesothelioma cases, thereby making it much more likely that claimants would receive full compensation than under the apportionment approach adopted by the House of Lords. It is much easier to show that moves to resolve causal indeterminacy problems in favour of claimants have been motivated at least to some extent by deterrence concerns. As we have seen, in McGhee two members of the House of Lords were concerned that unless liability were imposed employers would be able to disregard the duties of care they owed their employees with impunity. Similarly, in Fairchild, Lord Bingham, Lord Hoffmann and Lord Rodger all made it apparent that one of their motivations for departing from the general ‘but-for’ rule was the fact that its application would rob the duty to protect employees against mesothelioma of all content. The concern underlying this ‘empty duty’ argument is clearly that in some types of case strict adherence to the usual causation rules would not give employers an incentive to take reasonable care of their employees’ interests. Supporters of the loss of a chance theory have also argued that it would produce more effective incentives than the all-or-nothing approach which is currently used in personal injury cases. For example, John Fleming, an early proponent of ‘probabilistic causation’, wrote that ‘insistence on the traditional criteria of proof … leads to serious under-deterrence of the harmful activity’.92 Furthermore, a deterrence rationale might also serve to explain why the courts have been more willing to modify the causation rules in employers’ negligence cases than in medical negligence cases, since the judiciary may well feel that employers need more 91 See Restatement of the Law, Third, Torts: Liability for Physical Harm, Proposed Final Draft No 1 (Philadelphia PA, American Law Institute, 2005) § 28, comment o, where it is argued in the context of the American ‘market-share doctrine’ that a plaintiff who can demonstrate that the marketing and sale of a product was tortious, and that the product caused their harm, has a strong claim for compensation. 92 Fleming, above n 39, at 662–3. See also Stapleton, above n 75, at 390–91 and 399; and D Rosenberg, ‘The Causal Connection in Mass Exposure Cases: a “Public Law” Vision of the Tort System’ (1984) 97 Harvard Law Review 851, 866.

Causation and the Goals of Tort Law

189

incentives to exercise reasonable care than health-care providers. The ‘empty duty’ argument, for example, carries a lot more weight where the breach of duty is deliberate as opposed to inadvertent, and deliberate breaches are much more likely to happen in a commercial context, where the defendant may save money by not taking the precaution in question. Installing showers in McGhee would have cost money; administering less oxygen in Wilsher would not.93 The wider question of whether compensation and deterrence are best achieved through tort litigation is not my present concern, but it does seem to me to be clear that adapting a mechanism of corrective justice in order to bring about broader social goals of compensation and deterrence will inevitably lead to incoherence. This incoherence is the product of an irreconcilable tension between these two goals and the fundamental nature of the mechanism through which they are sought to be achieved. Both the tension, and the resultant incoherence, are amply demonstrated by the developments in the English law governing factual causation which I have considered in this chapter. The trouble is that the logic of the instrumentalist approach has no obvious stopping point. If we really think, for example, that negligence should be turned into a deterrence mechanism, then not only should we do away with the causation requirement, we should do away with the damage requirement as well. But if we do this, then we find ourselves with something that no longer looks like tort law at all, but much more like the criminal law. Similarly, if we want to turn negligence into a compensation mechanism, all that matters is that the claimant has suffered loss; fault and causation are irrelevant. But again this would leave us with something that looked much more like a social security system than tort law. And because there are no logical stopping points—or rather because the logical stopping points are incompatible with the basic idea of tort law—the courts have had to resort to illogical ones instead. Hence the fact that the single-agent/multiple-agent distinction has been used to prevent the Fairchild exception from becoming the general rule, despite the fact that no rationale can be given for it. And hence also the fact that in Gregg v Scott, Lord Hoffmann ruled out recovery for loss of a chance, but only a year later, in Barker v Corus, his Lordship was prepared to recognise the imposition of risk as a form of actionable damage, when—as Tony Weir has pointed out—‘if contribution to the risk of harm is sufficient, it is hard to see why diminishing the chance of avoiding it should not be’.94 Writing extra-judicially, Lord Hoffmann has explained that the House of Lords in Gregg refused to enlarge the Fairchild exception so as to encompass the loss of a chance cases because ‘one would not know where to stop. Anyone would be able 93 Although the courts have been careful to adopt restrictions (such as the single-agent rule) which are superficially neutral as between industrial disease and medical negligence cases, there are clear suggestions in the case law that in practice a distinction is being drawn between the two contexts: see eg Fairchild [2002] UKHL 22, [2003] 1 AC 32 [69] (Lord Hoffmann). 94 T Weir, ‘All or Nothing?’ (2004) 78 Tulane Law Review 511, 531. See also Lunney, above n 71. Note that since only one agent was in play in Gregg (the cancer), the single-agent rule cannot be used to distinguish the two cases.

190

Donal Nolan

to sue for the possibility that his damage had been caused by a defendant. This would be a radical change in the law of tort’.95 However, the need to preserve the essential nature of tort law by the imposition of such arbitrary limitations would never have arisen had the House of Lords resisted the temptation to use tort law as a mechanism of social policy in the first place. Perhaps, therefore, it would have been better if what his Lordship went on to describe as the ‘general rule that insuperable evidential problems are the claimant’s hard luck’96 had remained not only the general rule, but the universal one.

95 96

LH Hoffmann, ‘Causation’ (2005) 121 Law Quarterly Review 592, 600. ibid, at 601.

8 Looking Outward or Looking Inward? Obligations Scholarship in the Early 21st Century STEVE HEDLEY*

I. Introduction It is a commonplace in the philosophy of law that you can try to understand legal institutions and practices from two points of view: you can look at them from the outside in, or from the inside out. From the outside or ‘external point of view’, you try to make sense of them by looking at the externals: what the various officials and other participants are actually doing and (what is in practice inseparable) what they say they are doing. From the inside or ‘internal point of view’ you try to get into the heads of the lawyers, to grasp what they think they are doing by considering their thoughts and (what is in practice inseparable) their utterances. Neither viewpoint is reducible to the other. Legal systems have real-world effects, which cannot be explained away by pointing out that lawyers don’t always intend those effects or recognise them when they have occurred. And legal systems—I use the expression deliberately—are as much in lawyers’ heads as they are anywhere else. Both viewpoints, then, have something to say and something worth listening to. Both ultimately focus on the same thing—legal language—but they take different routes towards it. This is well-travelled ground in legal theory, which I will not retrace here.1 But to clear the ground, I must make some very basic observations. First, enthusiasm for one point of view is hardly ever meant as a criticism of the other. Hart’s own advocacy of the ‘internal point of view’ was never meant to supplant the external point of view; on the contrary, he took it for granted that much serious thought

* Many thanks to all who have helped with this chapter, including Tony Duggan, John Gava, Shane Kilcommins, William Lucy, Geoff McLay, Jonathan Morgan, Annette Morris, Andrew Robertson, Chaim Saiman, Geoff Samuel, Colin Scott, Stephen Smith, David Wright and Tang Hang Wu. I would be delighted to receive comment on or criticism of this chapter at [email protected]. 1 For Hart’s classic exposition see HLA Hart, The Concept of Law (Oxford, Clarendon Press, 1961) 54–7.

194

Steve Hedley

about legal systems would be from an external perspective (and his own theory was obviously external). Secondly, it is usual to find both views within the same skull. Of course, brief descriptions of the insider/outsider distinction often imply that the insider view is held by (say) judges, whereas the outsider view is held by (say) sociologists from other cultures; but this is an over-hasty slurring over the facts. A sociologist who failed to notice the continual appeals by legal actors to the legal system’s own values would not be a very observant one. And a judge who never appeals to values outside the law—responding to all such appeals with ‘It’s the system, what can you do?’—could only be either a confirmed cynic or seriously in need of Prozac. Both perspectives have their place. Neither excludes the other. What has this to do with obligations? General notions about law always have relevance to obligations, but there is a much closer connection, and that is what I want to consider here. Differences of approach to obligations have become rather stark of late, and a rather sharp division between internal and external perspectives seems to be emerging. On one side, you have writers looking purely inward, creating internal structures and maps for its constituent parts, and being greatly concerned for its overall coherence. These writers are far from unanimous as to which structures are the right ones—some draw on corrective justice, some on Roman law, some on notions of rights. But they are agreed that the solution to the problem they are addressing is to be found within the law and legal thought, not outside in questions of utility or the social effects of the legal system, questions which are seen as essentially political. On the other side, you have those who look externally, asking what purpose and interests legal structures serve, comparing other areas of law that bear on the same problems, and considering how the law should be described and developed against that background. Again, from the externalists there is no single view on any of those questions, no common view of what is the right way for the law of obligations to grow. There is, however, a consensus that these externally-oriented questions must be asked. A mere division of opinion is of course not really cause for comment. We are lawyers and we are academics—difference and debate are what you would expect. But we are in danger of losing track of what we are in fact agreed upon. In particular, there is a developing school of radical internalists who are prepared to deny any validity at all to the external viewpoint, and have been quite explicit that externalists should either be deprived of influence within the law school or even be run out of it (on the argument that externalists are not really ‘doing law’ at all). This view is, I argue, entirely wrong-headed. Each side has something to say, and it is grotesque to exclude proponents of the ‘wrong’ one. I am not unaware that the converse accusation has been levelled at some externalists, including perhaps myself: that our disdain for some structural concerns amounts to refusing any kind of internal view its rightful place in legal thinking. This would be equally bad if proven, but as I hope to show, the accusation is unfounded. Criticism of particular proposed structures, suggestions for alternatives, and discussion of the precise significance to be attached to structures are all very different things from a denial that structure has a role at all. Moreover, much of what the externalists are

Looking Outward or Inward: Obligations

195

doing is merely reminding internalists of what the rest of the legal system—public law and regulatory law especially—is already doing. We are happy to share the law school with the internalists, but expect a similar concession in return. In summary, then, both internal and external points of view play vital roles. Without the internal viewpoint, without at least some attention to structure and process, the legal system is not a ‘system’, and arguably isn’t ‘legal’ either. Whether the same is true of the external point of view is a moot point (it is much-debated, whether a system without externally defensible morality or goals can truly be ‘legal’), but no internalist would (I hope) defend obligations if it were in such a state. Without the internal viewpoint, obligations cannot survive; without the external viewpoint, it does not deserve to. And both viewpoints deserve their place in the legal academy.

II. The Emerging Internalists and Externalists While many of its elements were in place in earlier years, the revival of strongly internalist views is really a phenomenon of the last two decades, at least in the common-law world (excluding for present purposes the United States, which moves to a very different intellectual drum-beat). There were certainly hints of what was to come: Atiyah’s externalist commentaries on contract law2 provoked much comment, as did Fried’s internalist account of contract.3 But the key event was Birks’ Introduction to Restitution,4 which proposed to systematise that subject in a very particular and very precise way, drawing heavily on Romanist models, and pre-supposing that surrounding areas of law could be similarly systematised (a task which Birks seemed to think would be relatively uncontroversial). The ambition was always broader than the banal technicalities of restitution; it was not merely to find restitution on the map, but to ensure that the map itself was properly drawn. The radical internalist project was on the road. The last decade has seen these intellectual tools brought to bear on a wider subject-matter, in most (though not all) cases by those who learned their restitution from Birks. Four main strands are discernable, though perhaps they are not really distinct from one another, and some theorists endorse more than one of them.

2 Much of that strand of his writing was collected into PS Atiyah, Essays on Contract (Oxford, Clarendon Press, new ed 1990). 3 C Fried, Contract as Promise: A Theory of Contractual Obligation (Cambridge MA, Harvard University Press, 1981). 4 P Birks, An Introduction to the Law of Restitution (Oxford, Oxford University Press, 1985). The final statement of this theory was in P Birks, Unjust Enrichment, 2nd edn (Oxford, Oxford University Press, 2005).

196

Steve Hedley

Taxonomists take Birks’ campaign to a wider world, attempting the same sort of analysis of other fields of law. Leading examples are Chambers on property law5 and Pretto-Sakmann on personal property.6 An overall view of English private law under this taxonomy was written by a team assembled by Birks himself,7 though the Birksian taxonomy really only controls the chapter headings, each of the chapters then proceeding on a more individual basis.8 The reluctance of this approach to tackle statutory material, except where it accepts common law assumptions, ensures that the scope for this sort of work is limited.9 Rights Theorists regard the key question in obligations cases as the identification of the plaintiff ’s right, arguing that once this is done, the answer to other questions should fall into place rapidly and uncontroversially. In particular, this minimises reference to policy concerns, which are seen as objectionable. ‘Name the right, define it, and the rest is mere application in the light of the circumstances. More juris, less prudence’.10 Originally proposed in relation to restitution, the theory clearly has broader application, and has recently been urged as the basis both of negligence11 and of tort generally.12 A notable feature of the theory is the reluctance, to date, to definitively list the rights, or describe them in other than vague terms.13 Corrective Justice Theorists were active long before current controversies—many trace the core notions back to a brief and obscure theoretical foray by Aristotle, and indeed some regard later attempts to correct Aristotle’s line of thought as retrograde.14 However, it is obvious that the theory shares many assumptions with 5 R Chambers, ‘Integrating Property and Obligations’ in A Robertson (ed), The Law of Obligations: Connections and Boundaries (London, Routledge Cavendish, 2004) 127; see also the same author’s An Introduction to Property Law in Australia (Sydney, Law Book Co, 2001). 6 A Pretto-Sakmann, Boundaries of Personal Property: Shares and Sub-Shares (Oxford, Hart Publishing, 2005). 7 P Birks (ed), English Private Law (Oxford, Oxford University Press, 2000); see also the 2nd edn, A Burrows (ed), 2007. 8 See book review by N Kasirer (2003) 3 Oxford University Commonwealth Law Journal 249, 255. 9 cf the companion volume, D Feldman (ed), English Public Law (Oxford, Oxford University Press, 2004). 10 D Stevens and J Neyers, ‘What’s Wrong with Restitution?’ (1999) 37 Alberta Law Review 221, 227. 11 A Beever, Rediscovering the Law of Negligence (Oxford, Hart Publishing, 2007). Also relevant to the themes of this chapter is the same author’s ‘Particularism and Prejudice in the Law of Tort’ (2003) 11 Tort Law Review 146. 12 R Stevens, Torts and Rights (Oxford, Oxford University Press, 2007). Both Beever and Stevens also link their claims to corrective justice (see Beever, Rediscovering Negligence, above n 11, at 52–5 and Torts and Rights at 327). 13 Beever does not provide an account of the rights, and thinks it unreasonable that he should be asked to: ‘This book is a theoretical examination of the law of negligence as it operates in terms of the categories of enquiry with which we are familiar … Without becoming entirely unwieldy—and impossibly long—it cannot also become a philosophical investigation of personal and property rights’: Rediscovering Negligence, above n 11, at 62. How happy lawyers will be with such an approach, which apparently requires an ‘unwieldy’ and ‘impossibly long’ philosophical investigation even to answer the most basic questions it raises, must be a matter for speculation. Stevens provides various partial lists of rights (Torts and Rights, above n 12, at 4–17 and 303) though he seems to regard most questions they might raise as still open (at 337–40). 14 J Gordley, Foundations of Private Law (Oxford, Oxford University Press, 2006) 7–32.

Looking Outward or Inward: Obligations

197

the taxonomists, and there has in recent years been regular dialogue between them in the hope of constructing an account of the law acceptable to both, particularly involving leading corrective justice theorists such as Weinrib15 and Gordley.16 Interpretive Theorists seek to discern meaningful patterns in law; interpretive theory ‘is nothing more (nor less) than the attempt to understand legal concepts in terms of their meaning’.17 The theory has roots in Dworkin’s concept of ‘Law as Integrity’,18 though arguably the roots are shallow ones, as Dworkin’s concerns were with individual ‘hard’ cases, not structural ideas per se. Interpretive theory openly draws on the previous three theories, and perhaps is best seen as an attempted synthesis of them rather than as a distinct notion.19 Clearly, these various theorists do not agree on everything (indeed, it is a cause for celebration if we find two corrective justice theorists who can agree on what ‘corrective justice’ is). Equally clearly, they agree on a great deal, and can fairly be regarded as a coherent school or movement within obligations scholarship.20 Common features (all of which indicate the inward focus) are: the concentration on common law at the expense of statute; an insistence that they are describing matters common to all legal systems (which is taken to excuse them from discussing any one system); and a disdain for the purpose of legal institutions. Indeed, some deny that obligations can properly be said to have a purpose at all, unless that purpose is merely to be itself; private law no more has a purpose than Love has a purpose. ‘Love is its own end. My contention is that, in this respect, private law is just like love’.21 Engagements with the external point of view are therefore typically brief and fraught with hostility. Externalist writers by contrast have tended not to cluster together, and are agreed on little except that standards and perspectives external to obligations are nonetheless of relevance within it. Serious engagement with internalist ideas

15

EJ Weinrib, The Idea of Private Law (Cambridge MA, Harvard University Press, 1995). Gordley, above n 14. A Beever and C Rickett, ‘Interpretive Legal Theory and the Academic Lawyer’ (2005) 68 Modern Law Review 320, 328. 18 R Dworkin, Law’s Empire (London, Fontana, 1986) ch 7. 19 A good example is S Smith’s Contract Theory (Oxford, Clarendon Press, 2004) (which, despite what its name suggests, focuses almost exclusively on interpretive theories of its subject-matter). I have written in more detail on this strand of internalist theory in ‘The Shock of the Old: Interpretivism in Obligations’ in C Rickett and R Grantham (eds), Structure and Justification in Private Law—Essays for Peter Birks (Oxford, Hart Publishing, 2008). 20 For a general survey of the area see W Lucy, Philosophy of Private Law (Oxford, Clarendon Press, 2007). 21 Weinrib, above n 15, at 6. This remark is puzzling, even as a comment on love. What Weinrib may mean is that those who are in love will typically show little interest in a rigorous analysis of that phenomenon; perhaps also he has in mind that ‘The mind has a thousand eyes/And the heart but one’ (FW Bourdillon, The Night has a Thousand Eyes). Yet questions about what love is for are routinely asked in literary criticism, social theory, theology, cultural studies, developmental psychology and reproductive biology—and why should they not be? Weinrib’s (purposefully absurd) example of proposing economic efficiency as the explanation of love (at 5) merely shows that some possible answers are wrong, not that it is a mistake to ask the question. For comment and critique see J Gardner, ‘The Purity and Priority of Private Law’ (1996) 46 University of Toronto Law Journal 459. 16 17

198

Steve Hedley

usually involves thorough immersion in them; the diversity of alternative viewpoints militates against the emergence of a unitary ‘externalist’ point of view. Indeed, much ‘externalist’ writing goes little beyond the demand that the role of external preferences when choosing between internal theories should be more openly acknowledged.22 This is, I argue, no bad thing; the dogmatism of the more radical internalists is not a feature to be emulated. New perspectives on obligations have come from public law, from sociology, from political thought, from history and from economics. The best writings in this area have been attempts at synthesis of differing viewpoints to gain a clear perspective on legal theories:23 special mention in relation to torts should go to the writings of Stapleton,24 Cane25 and Keren-Paz;26 in relation to contract, Wightman27 and Collins;28 and in relation to restitution, Dagan.29 A more thorough-going externalism, which would attempt to understand and assess obligations entirely from some perspective outside it, is certainly imaginable. But no such approach has even a toe-hold in common law culture outside the United States. While Christian legal scholarship seems to be growing in volume and sophistication,30 it has had little to say on obligations as yet even in America. The more fundamentalist varieties of law-and-economics are rather better established there, but have yet to make it far into other common law jurisdictions; such lawyer-economists as have prospered in the United Kingdom

22 Most of the ‘anti-Birks’ writings, including those of the present author, are of this type; the alternative theories of restitution urged are for the most part no more externally-oriented than the theories they oppose. 23 Foremost in promoting this strand of scholarship was PS Atiyah, particular with The Rise and Fall of Freedom of Contract (Oxford, Clarendon Press, 1979) and Accidents, Compensation and the Law, 1st edn (London, Weidenfeld & Nicolson, 1970); see also current edition, P Cane, Atiyah’s Accidents, Compensation and the Law, 7th edn (Cambridge, Cambridge University Press, 2006). 24 See especially J Stapleton, Product Liability (London, Butterworths, 1994). 25 See especially P Cane, The Anatomy of Tort Law (Oxford, Hart Publishing, 1997). Cane’s more recent work has concentrated on shoring up the internal perspective, eg ‘Tort Law as Regulation’ (2002) 31 Common Law World Review 305; Responsibility in Law and Morality (Oxford, Hart Publishing, 2002), but not I think to the extent of abandoning the external perspective entirely (unless Peter wants to tell me different!). 26 See especially T Keren-Paz, Torts, Egalitarianism and Distributive Justice (Burlington, Vermont, Ashgate, 2007), and the same author’s ‘Private Law Redistribution, Predictability and Liberty’ (2005) 50 McGill Law Journal 327. 27 See especially J Wightman, Contract: A Critical Commentary (London, Pluto, 1996). 28 See especially H Collins, Regulating Contracts (Oxford, Oxford University Press, 1999), and review of it by D Campbell, ‘Reflexivity and Welfarism in the Modern Law of Contract’ (2000) 20 Oxford Journal of Legal Studies 477. 29 See especially H Dagan, The Law and Ethics of Restitution (Cambridge, Cambridge University Press, 2004); and the same author’s ‘Just and Unjust Enrichments’ ch 17. Also worthy of mention is K Barker’s call for a more pluralist approach: ‘Understanding the Unjust Enrichment Principle’ in J Neyers, M McInnes and S Pitel (eds), Understanding Unjust Enrichment (Oxford, Hart Publishing, 2004) 79. 30 For a general survey see D Skeel, ‘The Unbearable Lightness of Christian Legal Scholarship’ (2008) 57 Emory Law Journal 1471.

Looking Outward or Inward: Obligations

199

have been of the less absolutist sort.31 And if law can be said to have started on the road towards becoming a proper social science, it clearly has a long way to travel.32

III. Why Have We Not Progressed? When it comes to a detailed statement of the law of obligations, it is striking how much of the modern vision of internalists merely repeats the views of leading law teachers circa 1880. Those who look inward see an antique set of rules. Certainly contract has changed very little: a modern law teacher who reads the first editions of Pollock or Anson on contract33 will find little that is strange. The tort books of that time would be a little stranger, no doubt, though mostly on matters of emphasis rather than substance. Restitution is the newest component of obligations, though its writers have rapidly ‘traditionalised’ themselves by the rapid ingestion of large slabs of Romanist thought. Crucially, the underlying political assumptions of the Victorian text writers and the modern internalists would apparently be the same: namely that the topic under discussion is justice between individuals, not wider concerns of social justice; that while there are occasional parliamentary intrusions into the law, there is little point in looking for a pattern in them; that freedom of contract should be assumed to be the norm; and that order and system are to be found in the eternal common law, not in the random forays of here-today-gone-tomorrow politicians and civil servants. (It will be appreciated that I am describing the more ‘traditional’ modern textbooks here; my description does an injustice to the entire range available, but I think identifies a solid strand within modern writing.) And where the internalists appear to have innovated, it is usually simply by completing projects the Victorians themselves started but left incomplete: finding a unitary basis for tort, or finding a more satisfactory description of the materials traditionally labelled ‘quasi-contract’. We are so used to this that it has ceased to strike us as odd. But it is odd. Every other branch of human knowledge—including legal knowledge—has progressed immeasurably over the past century. In any other area, academics would be embarrassed at using much the same theories and attitudes as were advanced a century and a half ago, with nothing to show for the work of the intervening 31 The apparent unexportability of law-and-economics is starting to be an issue in the literature (see eg K Grechenig and M Gelter, ‘The Transatlantic Divergence in Legal Thought: American Law and Economics vs German Doctrinalism’ (2007) 31 Hastings International and Comparative Law Review 295) but there is a way to go in understanding it. 32 G Samuel, ‘Is Law Really a Social Science? A View from Comparative Law’ [2008] Cambridge Law Journal 288. 33 F Pollock, Principles of Contract at Law and in Equity: A Treatise on the General Principles Concerning the Validity of Agreements in the Law of England, 1st edn (London, Stevens, 1876); W Anson, Principles of the English Law of Contract and of Agency in its Relation to Contract, 1st edn (Oxford, Oxford University Press, 1879).

200

Steve Hedley

period except some minor updating. In fact, the similarities between the Victorian text writers in obligations and the modern ‘internalists’ fade when we consider why they wrote as they did. The Victorians took many of the same choices as the modern internalists but for vastly different reasons, which (from our vantage point) seem to be more about their past than their future. Their preference for principle over discretion seems to have been a reaction to the chaos of the medieval common law, and their choice of Roman law as (barely acknowledged) organising tool is unsurprising, given the lack of alternatives;34 anyone adopting those positions today must presumably have rather different reasons. Their lack of concern with individual fairness must be seen in the context of a jury system that could apply the law flexibly to each case—jurists did not need to consider fairness in individual cases, because that could be left to juries. With juries now largely removed from the civil legal system,35 jurists’ neglect of fairness sends a very different message. And of course the Victorian jurists, respectable gentlemen all, had the blindnesses of their generation: drawing a rigid line between private law and public law, and somehow failing to notice the growing mass of governmental regulation law which made a nonsense of the distinction. Terrified at the broadening of the electorate, and the socialist barbarism they feared would quickly follow, they turned their backs on the law as it actually was, and so left us no internal account of the (tortuous but in the event largely peaceful) transformation of law and government which democracy brought about. It is one of the most sobering aspects of modern legal history that the Victorians constructed the basis of the modern administrative state, and yet their best legal brains failed to notice what was going on. The occasional genius did, perhaps. It took a John Stuart Mill to point out that the law of contract always involves a deliberate public choice as to which bargains the state will enforce, and so talk of ‘respecting the will of the parties’ is therefore merely obfuscation;36 and a Frederick Maitland to notice (eventually) how detailed and how thorough regulation of the economy had become, protesting feebly that his colleagues still spoke as if this only involved the occasional use of the royal prerogative.37 In this, as in other matters, the past is a foreign country. The Victorian text writers were not ‘adopting the internal point

34 On the reception of Romanist doctrine in contract see J Gordley, The Philosophical Origins of Modern Contract Doctrine (Oxford, Clarendon Press, 1991) especially chs 6–8. 35 For the (rather sudden) decline of the civil jury in England see M Lobban, ‘The Strange Life of the English Civil Jury, 1837–1914’ in J Cairns and G McLeod (eds), The Dearest Birth Right of the People of England (Oxford, Hart Publishing, 2002). 36 JS Mill, Principles of Political Economy, 7th edn (1871, reprinted, Oxford, Oxford University Press, 1998) 162. This passage (too long to reproduce here) is a remarkable one from a common law point of view. By evaluating private law principles as state action, it anticipates Shelley v Kraemer, 334 US 1 (US Supreme Court, 1948) by three-quarters of a century, and is within spitting distance of the ‘horizontal effect’. 37 F Maitland, The Constitutional History of England (Cambridge, Cambridge University Press, 1913) 417.

Looking Outward or Inward: Obligations

201

of view’; by their standards, they addressed issues broadly and with full regard to relevant policy concerns. They were simply Victorian. And yet it was this generation of scholars that set the tone for almost a century in England and in most of the Commonwealth. The precise history of those who followed them differs from country to country, but the overall stories are much the same: the social sciences separated out and ‘professionalised’ themselves, with the result that each had less to say to the other, if indeed they talked at all; most law schools were dominated by part-time teachers and practitioners, with little interest in what the rest of the university was doing; innovation in legal theory was poorly rewarded by universities and not rewarded at all by legal professional bodies, which tended to assume that new theory with no immediate application in legal practice must be unimportant.38 The effects of this period in the doldrums are still with us, not least because during it many law schools attained ‘cash-cow’ status within their universities—high student numbers, grimly efficient teaching, low research costs—and now find it hard to take a different tack without incurring their superiors’ displeasure. There is much more to say on this, and on why the US legal academy followed such a different path.39 But the low morale and low intellectual productivity of that period in legal scholarship are well known. It was in the 1960s that the tide turned. Increasing government investment in universities, coupled with the obvious importance of law in an increasingly juridified political climate, led to a huge expansion in the number and quality of legal academics. Full-time law staff, on well-defined career tracks where high-quality publications have a good chance of leading to promotion, steadily became the norm. In such a vastly increased university system, there are niches for all sorts of views: the law schools and their occupants are diverse as never before. With an increasingly complex legal system, and with significant numbers of ambitious scholars keen to make their individual marks (for law is still the field of the lone scholar, despite trends to the contrary in other fields), an inevitable consequence has been increased specialisation. The breadth of knowledge achieved by earlier generations of scholars is usually not attainable, and indeed may no longer be desirable, because knowledge so broad must in modern conditions be horribly shallow. The law has essentially been written afresh. Many of the subjects now written on simply did not exist five decades ago. It is a sobering thought that neither administrative law nor labour law were established legal academic subjects in the United Kingdom in 1960, and while family law and constitutional law have an earlier history, those subjects are today unrecognisable from earlier writings. In modern conditions, the law is being constantly re-written both through the

38

Atiyah, The Rise and Fall of Freedom of Contract, above n 23, especially 660–71. For some relevant consequences of which see C Saiman, ‘Restitution in America: Why the US Refuses to Join the Global Restitution Party’ (2008) 28 Oxford Journal of Legal Studies 99; the same author’s ‘Restitution and the Production of Legal Doctrine’ (2008) 65 Washington and Lee Law Review 993; E Sherwin, ‘Legal Positivism and the Taxonomy of Private Law’ in C Rickett and R Grantham, above n 19, at 103. 39

202

Steve Hedley

political process and through emerging academic ideas.40 Change is the only constant. And yet many in obligations still emulate the Victorian textbook style, even though that style, with its blinkered refusal to acknowledge external influences or public concerns, was out-of-date even when originally employed. Viewed externally, the law of obligations is not an island: its rules have social and economic effects like any other area of law, and there is sustained and complex legislative intervention there as elsewhere. What is unique about the law of obligations is a deeply ingrained internalism, which looks for system in the common law only: it still refuses to take seriously the idea that statutory interventions might be purposeful, or might reflect concerns more urgent than those to be found in the old case law. This extreme internalism is not a unanimous view, to be sure. Not everyone treats statutory regulation of contract and tort as some kind of alien intrusion, or prefers Victorian values to modern ones. But there is a persistent strand of thought that the common law safeguards important values of personal responsibility and limited government which legislation necessarily (at least, as currently practiced) disregards. For most of the latter half of the twentieth century, those who thought this way saw obligations as a dying subject, to be fought for while it lasted, but doomed eventually to be submerged by the rising tide of regulation.41 By the 1990s, this fatalistic cast of mind had found a new outlet—promoting unitary theories of obligations which attempted to unite the cases, with theories which concerned themselves only with justice between individuals—and ignored both wider public concerns and most legislation. Particularly influential was the Romanist model proposed by Peter Birks in relation to unjust enrichment, and which is now increasingly being applied to contract and tort as well. And there is no doubt that this vision of obligations suits some very well indeed: those who wish to portray public law as innovative, sophisticated and responsive to modern concerns can do so all the better if they assume that private law is conservative, narrow and wilfully old-fashioned.42 And so long as the question is posed that way—Do we stick with the traditional approach to obligations, or do we adopt a new publicoriented focus?—then we do indeed have to choose, however unpalatable both the alternatives seem to be.

40 For accounts of the history see W Twining, Blackstone’s Tower—The English Law School (London, Sweet and Maxwell, 1994) ch 2; F Cownie, Legal Academics—Culture and Identities (Oxford, Hart Publishing, 2004) ch 2. 41 One of the more eloquent proponents of this view (I do not know if he still holds it) was Nigel Simmonds. See his The Decline of Juridical Reason: Doctrine and Theory in the Legal Order (Manchester, Manchester University Press, 1984) especially chs 2 and 9. 42 M Loughlin, The Idea of Public Law (Oxford, Oxford University Press, 2003); on Loughlin’s view of what private law is, see book review by N Barber (2005) 25 Oxford Journal of Legal Studies 157, 165–6. Contrast D Oliver, Common Values and the Public-Private Divide (London, Butterworths, 1999); and M Moran, ‘The Mutually Constitutive Nature of Public and Private Law’ ch 2 of this book.

Looking Outward or Inward: Obligations

203

IV. The Modern Interweaving of Private and Public Yet this stark choice is an illusion, an ideological holdover from before the fall of the Berlin Wall, an unsophisticated and barely intelligible demand that we chose between State and Market. In fact, we are stuck with both a strong state and an all-pervasive market for the foreseeable future. Governments tempted to reduce the power of the state have won few victories; governments tempted to abolish markets routinely fail to generate wealth by other means, for which they are rapidly punished by their citizens. There are differences over the relative role of state power and of markets, but on all but the fringes of politics they are differences of emphasis only.43 The thesis of unrestrained capitalism has encountered the antithesis of state socialism, and we are now the heirs to the (inelegant and theoretically untidy) synthesis: a blending of public and private concerns, in which market institutions play a major role but are forever subject to public monitoring, review and reform. Where traditional legal rules survive, it is not because they are traditional but because the political system has determined (rightly or wrongly) that they are better than the alternatives on offer. The ideals which inspired an earlier generation of judges to lay them down are no longer to the point. Modern obligations law, therefore, along with most other areas of law, represents an interweaving of public and private concerns. No area of law can be ‘purely private’, neglecting the public good or trusting that the invisible hand of the market will cure all ills. So legislative intervention into ‘private’ areas is commonplace; and the state acts not merely by abolishing those parts of the common law it does not care for, it also modifies and re-moulds the common law to suit its purposes. Law is a flexible tool in the state’s hands, and the state has proved adept at turning old legal institutions to new purposes. But this is not a one-way process: the private influences the public too. ‘The state’ is not really a unitary entity at all, still less a capricious tyrant that can act as it pleases without fear of the consequences. The state could in theory abolish property ownership (or any other legal concept) tomorrow if it wanted to, but would in practice run into irresolvable difficulties if it tried; accordingly, ownership is not a myth.44 The state cannot itself make wealth or improve social conditions—it can only hope to create economic and social circumstances in which others will be able to do achieve this—and the legitimacy and power of those currently at the helm depend heavily on results. The state, then, cannot simply ‘do what it wants’—its powers are limited (not least by the powers of private actors)—and always runs the risk of being held responsible for the results of its actions (including results mediated 43 S Deakin, ‘Private Law, Economic Rationality and the Regulatory State’ in P Birks (ed), The Classification of Obligations (Oxford, Clarendon Press, 1997) 283. 44 Contrast L Murphy and T Nagel, The Myth of Ownership: Taxes and Justice (Oxford, Oxford University Press, 2002), reviewed by F Maultzch (2004) 67 Modern Law Review 508.

204

Steve Hedley

through the reactions of private actors). So the public moulds the private, and the private moulds the public. And an observer who looks at the private alone, as if it were a distinct entity, will miss most of what is going on.45 This process is ongoing throughout the modern law, and is at its most fascinating when it comes to the creation of new forms of private property, usually in trying to locate new technologies within the complex web of modern law. (When are biological entities ‘property’, and what are their ‘owners’ allowed to do with them? Are databases property, and how much proprietary protection does ‘intellectual property’ deserve? And who ‘owns’ the bread-crumb trail of electronic records you leave as you surf the web?) More mundane examples abound, even in the core areas of obligations. Two examples follow. That the law on personal injury compensation could be defined purely ‘internally’ or ‘apolitically’ would have been an impossible position in (say) the middle of the nineteenth century. Industrial safety was a leading concern of the growing trade union movement; and employers were quite open in arguing that generous rights of action for injured employees threatened economic development (because of cost, and because of the loss of managerial flexibility implicit in injury prevention). The steady ‘de-politicisation’ of personal injury law was simply the steady victory of the union side, as increasing democracy made it harder for government to ignore working-class concerns, and as increasing access to legal services ensured that both plaintiffs and defendants had lawyers. The legislative influence is clear enough, with reforms whittling away defences previously available to defendants, increasing the scope of public legal provision, encouraging public liability insurance, and (perhaps most important of all) not providing any other legal outlet for injured plaintiffs.46 The growth of liability was therefore the product of both internal forces (the inner logic of the legal concepts) and some rather forceful influence from outside. What of the brave new world we now live in, where tort is ‘in crisis’, where governments now seek to reduce the scope of liability,47 where much of the public (who are not really into internalism) regards tort law as a bureaucratic intrusion by an over-meddlesome state,48 and where industrial safety 45 For some of the issues see M Zamboni, The Policy of the Law—A Legal Theoretical Framework (Oxford, Hart Publishing, 2007). 46 For the history in one jurisdiction see W Cornish and G de N Clark, Law and Society in England 1750–1950 (London, Sweet and Maxwell, 1989) ch 7. An intriguing counterpoint comes from another common law jurisdiction that took a different path entirely, attempting from the 1970s onwards to replace the common law scheme of liability with a system of payments from public funds. Yet there too the private lawyers have by-and-large refused to engage with the law as it is, falling back on common law principles when the law has (all too obviously) moved on: G McLay, ‘Accident Compensation— What’s the Common Law Got to Do With It?’ [2008] New Zealand Law Review 55. 47 P Cane, ‘Taking Disagreement Seriously: Courts, Legislatures and the Reform of Tort Law’ (2005) 25 Oxford Journal of Legal Studies 393 and also in M Bryan (ed), Private Law in Theory and Practice (London, Cavendish, 2007) 27. 48 Indeed, even some judges see it that way: see JJ Spigelman, ‘Negligence: The Last Outpost of the Welfare State’ (2002) 76 Australian Law Journal 432. For debate on the ‘compensation culture’ see F Furedi, Courting Mistrust (London, Centre for Policy Studies, 1999); E Lee et al (eds), Compensation Crazy: Do We Blame and Claim Too Much? (London, Hodder and Stoughton, 2002); S Thomson,

Looking Outward or Inward: Obligations

205

seems little different from the Victorian scene except in being more overtly racist (as dangerous trades are increasing moved abroad, to developing nations whose rulers say they cannot afford first-world health-and-safety laws—just as firstworld Victorian employers did)? It will be apparent that this area of law cannot be understood without appreciating both its inner logic and how it appears to the community whose interests it is supposed to serve.49 To view it only ‘internally’ is to ignore the one argument which might possibly justify it—namely, that all the alternatives we might put in its place are even worse.50 Equally, the law on the contents of contracts. Here we encounter a peculiarly Victorian blind spot. The Victorian text writers started from the notion that contracts were usually the product of individual agreement (not an unreasonable view, around the middle of the nineteenth century) and reasoned that therefore the individual terms of contracts must also be the product of individual agreement— agreeing to a contract was equated with agreeing to its terms. This is of course a non-sequitur (you can’t deny your agreement simply because you did not set the terms, any more than you can deny your marriage simply because you had no hand in framing the marriage laws51), and the basic contradiction implicit in this has dogged contract theory ever since. In the late Victorian period, it led to the (confused and confusing) distinction between ‘contract’ and ‘status’,52 as well as increasingly desperate attempts to regard all the various sources of terms (custom, industry practice, national legislation, international treaty) as somehow mere aspects of personal agreement. One might have thought that the emergence of the corporate economy, where large economic units simply impose their terms on others without any pretence of ‘agreement’, would have put paid to this talk.53 Yet ‘Harmless Fun Can Kill Someone’ (2002) 1 Entertainment Law 95; L McIlwaine, ‘Tort Reform and the “Compensation Culture”’ [2004] Journal of Personal Injury Law 239; K Williams, ‘State of Fear: Britain’s “Compensation Culture” Reviewed’ (2005) 25 Legal Studies 499; A Morris, ‘Spiralling or Stabilising? The Compensation Culture and Our Propensity to Claim Damages for Personal Injury’ (2007) 70 Modern Law Review 349. 49 For parallel concerns see J Solomon, ‘Judging Plaintiffs’ (2007) 60 Vanderbilt Law Review 1749; R Bagshaw, ‘Tort Law, Concepts and What Really Matters’ ch 10 of this book. 50 For integrated discussions see, eg, P Cane, Atiyah’s Accidents, Compensation and the Law, 6th edn (Cambridge, Cambridge University Press, 2004) chs 18–19; D Howarth, ‘Three Forms of Responsibility’ [2001] Cambridge Law Journal 553. See generally C Robinette, ‘Can There Be a Unified Theory of Torts?’ (2005) 43 Brandeis Law Journal 369. For some excellent ongoing work see the ‘Compensation Culture Project’ (Principal Investigators: S Halliday and C Scott), which employs participant observation in local authority offices to ask (amongst other things) whether the applicable tort regime in fact operates as a system of regulation. 51 Of course, whether marriage is truly a contract has been a talking-point down the ages, though with temporary and local variations depending on what ‘marriage’ is, what ‘contract’ is, and what marriage could possibly be if not a contract. 52 Maine’s famous claim that ‘the movement of the progressive societies has hitherto been a movement from Status to Contract’ (Ancient Law (1861) 170) has been much debated, though as Atiyah notes, it is now almost a cliché to say that this trend has reversed itself, if indeed it was not already in reverse when Maine was writing (Atiyah, above n 23, at 259–60 and 716). See M Rehbinder, ‘Status, Contract, and the Welfare State’ (1971) 23 Stanford Law Review 941. 53 PS Atiyah, The Rise and Fall of Freedom of Contract, above n 23, at 596–601. Note also the late-19th and early 20th century English movement, in many respects successful, to escape orthodox legal views of

206

Steve Hedley

the Victorian habit has stuck with the internalists, who continue to say ‘contract’ and mean their general theory of contract, which stresses individual free choice and ignores the messy battles over which terms shall govern. To study those battles, one must look in the books on individual types of contract, or even on ‘regulation’ and ‘competition law’, where the multitude of collectivities (private and public, national and international) seeking to influence contract terms each seek to have their say. Again, it is not a question of choosing between the internal and external perspectives, as both have some validity. To paraphrase Collins, externalists do not criticise the notion of contract-as-promise in order to dismiss it. We criticise it because unless its vulnerability to criticism is understood and its weaknesses appreciated, the limited role it plays in the modern law of contract cannot be understood. Contract-as-promise is not dead; it is not even dying. But it is only one aspect of a vastly more complex law which regulates the modern institution of contract.54 As these examples illustrate, it is not a question of either ‘internalism’ or ‘externalism’ being untenable. Rather, neither view is enough on its own. The kind of radical internalism I am critiquing here routinely falls into this trap, by carefully limiting its attention to its chosen area (‘common law of obligations’) and proceeding as if it does not matter what is outside it. But the common law is not an isolated bubble, and questions of how judges are to develop the law depend very much on whether they are the best people to do it—which in turn means asking what the functions of other organs of government are and should be, and perhaps re-thinking what the functions of the judiciary should be. The truth is that we are in a period of extreme flux and uncertainty. There was a time when a constitutional law student could answer the question ‘Who makes the law?’ with ‘the Queen in Parliament’, and expect to pass with only a smidgeon of corroborative detail. Today, while the Queen is still on her throne, such a student would be failed outright in any law school worthy of the name. Leaving royalty aside, national parliaments do not make the law to the extent that they did, and many of them complain bitterly at this. Regional and international bodies make more and more of the law, though their remoteness from the people they govern places sharp limits on their legitimacy and their powers. Greater emphasis on rights and proper procedures, as well as a host of other factors, has led to more and more issues being classified as ‘legal’. This juridification has led to a hugely increased judicial sphere of influence: a universally acknowledged fact, if not seen by all as reason to celebrate.55 Against that background, it is unsurprising that judges in the higher courts have in recent years broadened their vision, being what the terms of contracts should be, by means of standard forms, arbitration, and a new Commercial Court which could be relied upon to take a ‘commercial’ view: H Arthurs, Without the Law: Administrative Justice and Legal Pluralism in 19th Century England (Toronto, Toronto University Press, 1985) ch 3. 54

H Collins, Regulating Contracts (Oxford, Oxford University Press, 1999) 9. Indeed, the modern assumption of power by judges is often spoken of almost as a world-wide political coup. See R Hirschl, Towards Juristocracy (Cambridge MA, Harvard University Press, 2004); AS Sweet, ‘The Juridical Coup d’État and the Problem of Authority’ [2007] 8 (10) German Law Journal 915, and other articles in that number, at www.germanlawjournal.com. 55

Looking Outward or Inward: Obligations

207

bolder both in developing the law and in the range of arguments they have been prepared to listen to. Yet in commenting on this we see radical internalism at its worst, praising judicial innovations where they happen to like them (such as in the introduction and progressive refinement of unjust enrichment), but bleating against the judges’ use of ‘policy reasons’ which they do not care for, and insisting that judges should stick to ‘principle’. But a distinction between ‘principle’ and ‘policy’ can only be made to work when those terms are well defined. At this particular juncture, neither is. At the risk of being seen as parochially European, I should perhaps add that these issues of the status and character of private law can all be seen in microcosm in the debates on harmonisation of private law throughout the European Union—debates on whether it should be done, what difference it might make, to what extent the aim should be reform and to what extent merely consolidation, and how it can be justified under the existing treaties.56 One of the most striking features of this is that not merely the arguments but the actual topic of debate is forever changing. Sometimes it is about economics and the facilitation of crossborder trade (But whose view of economics? With what role for differing national policies?57). Sometimes it is about attitudes to the heritage of Roman law (But is this really an appeal to common origins, or merely a bid for power by those who claim to be Roman law’s most faithful guardians? Is the common law really different, or does its serial plagiarism from Roman sources—of which Birks is merely the most recent example—invalidate that argument?58). And sometimes it is about uniformity and the messages uniformity sends (Would harmonisation of laws across Europe be a clear and welcome statement of unity, or an undemocratic usurpation of power which would never have been granted had it been asked for explicitly?59 Is the tenacity of national legal cultures mere lawyerly self-interest, or

56 The literature on this topic is voluminous, repetitive and in a variety of respects tedious beyond all comparison. However, some contributions rise markedly above the usual level, and are particularly relevant to the themes of this chapter. In addition to the pieces referred to in the following footnotes see M Hesselink, The New European Private Law (The Hague, Kluwer, 2002) and C Twigg-Flesner, The Europeanisation of Contract Law (London, Routledge Cavendish, 2008). 57 See, eg, E Carbonara and F Parisi, ‘The Paradox of Legal Harmonization’ [2007] Public Choice 367; F Nicola, ‘Transatlanticisms: Constitutional Asymmetry and Selective Reception of US Law and Economics in the Formation of European Private Law’ (2008) 16 Cardozo Journal of International and Comparative Law 101. 58 See, eg, R Zimmermann, ‘Roman Law and the Harmonisation of Private Law in Europe’ in A Hartkamp et al (eds), Towards a European Civil Code, 3rd edn (The Hague, Kluwer, 2004). More generally on unoriginality and plagiarism in legal development, see AWB Simpson, ‘Innovation in 19th Century Contract Law’ (1975) 91 Law Quarterly Review 247, reprinted in AWB Simpson, Legal Theory and Legal History (London, Hambledon, 1987) ch 10; M Siems, ‘Legal Originality’ (2008) 28 Oxford Journal of Legal Studies 147. 59 See, eg, B Markesinis, ‘Why a Code is Not the Best Way to Advance the Cause of European Legal Unity’ (1997) 5 European Review of Private Law 519; H Collins, ‘The Voice of the Community in Private Law Discourse’ (1997) 3 European Law Journal 407; E McKendrick, ‘Harmonisation of European Contract Law: The State We Are In’ in S Vogenauer and S Weatherill, The Harmonisation of European Contract Law (Oxford, Hart Publishing, 2006) 5–30.

208

Steve Hedley

a legitimate exercise of subsidiarity?).60 Such is the modern legal system: the many and diverse perspectives within it are each entitled to their say. If theorists choose to concentrate on smaller areas, then let them do so; but let them not deny what it is that they are doing, or simply assert that their favourite problem is somehow more important than the problems they leave for others.

V. The Problem: Polarisation and Mutual Disdain The problem is not, therefore, too much internalism/too much externalism, or too much order/too little. Everyone agrees that the law of obligations is more-orless structured, though some are interested in the ‘more’, others in the ‘less’. The problem is how to make each camp pay due respect to the work of the other. Every scholar will chose the sort of work that they find most congenial, or where they think they can best make their mark. What must be avoided is allowing them to suppress or denigrate work against which they have no legitimate complaint, but simply do not care to do themselves. In my view, the writing here is unbalanced. When we go looking for externalists who have denied any validity at all to internal views, we simply do not find them. The need for some sort of structure has never been denied (the dispute has been over the adequacy of the structures on offer, and the need for better ones). Are there any ‘radical externalists’ to be found, who are simply not interested in any kind of legal structure? Kaye has recently protested against the ‘fundamentalists’ in both camps, damning both the rights theorists (internal) and the more enthusiastic law-and-economics theorists (external)61—but that sort of law-andeconomics has no significant uptake outside the United States. Some externalists have been accused of the vile crime of being legal realists (who are assumed to be awful creatures, with no regard for properly structured legal thought), but Dagan’s work has undermined the charge, exposing the idea that legal realists had no respect for law’s internal structure as the nonsense it is.62 Where externalists have been accused of ignoring law’s internal workings, it is usually in the defence of some very particular internalist view—so someone who attacks Birks’ view on how to structure unjust enrichment is told (wrongly) that they are attacking the very idea of structure itself.63 The truth is that externalists understand the 60 See, eg, H Collins, ‘European Private Law and the Cultural Identity of States’ (1995) 3 European Review of Private Law 353; J Smits, ‘A European Law on Unjustified Enrichment? A Critical View of the Law of Restitution in the Draft Common Frame of Reference’ (2008) 19 Stellenbosch Law Review 179–88; S Glanert, ‘Speaking Language to Law: The Case of Europe’ (2008) 28 Legal Studies 161. 61 T Kaye, ‘Rights Gone Wrong: The Failure of Fundamentalist Tort Theory’ (May 2008), available at http://ssrn.com/abstract=1130492. 62 H Dagan, ‘Legal Realism and the Taxonomy of Private Law’ in C Rickett and R Grantham, above n 19, at 147. 63 See eg, the charge that my own writing on restitution is ‘anti-theoretical’ and ‘downplays the importance of principled legal reasoning’: D Sheehan, ‘Implied Contract and the Taxonomy of

Looking Outward or Inward: Obligations

209

strengths of internalism well, and use internalists much as a truffle-hunter uses truffle hounds—if there is order to be found, we can trust them to sniff it out, and we ourselves will do whatever else is required. If internalists feel challenged by criticisms of their structures, then so they should be—and let them respond with reasoned defences of it, not with the nonsensical charge that their critics reject all structural notions.64 A more promising line of internalist attack is one which acknowledges their opponents’ strengths, as for example with Goldberg’s recent charge that those taking up rigid theoretical positions are promoting half-truths about tort law.65 Precisely! Each side has some proportion of the truth—let us politely say that each has a half, until the contrary is shown—and let us then tease out in debate precisely where the whole truth lies. Very different are those writers for whom internalism has become the whole game. For them, it is not a question of balance. Externalist viewpoints must be ignored or their relevance denied, because otherwise legal debate ceases to be distinctively legal: it becomes a free-for-all in which the very meaning of law is lost. While the point has been put in various ways, the essential argument is the same, and has been made many times by leading internalists: that we have a stark choice, between an utterly pure internalism and an utterly pure externalism, and that anyone who chooses the second is simply not a lawyer. The danger is that, as an appeal to policy becomes a more frequent practice, academic lawyers will appeal to policy rather than attempt to refine their understandings of private law so that they are no longer inadequate. If this occurs, then, with respect to the private law, academic lawyering will no longer exist as a discipline … On this view, academic private lawyers will have abandoned the primary task of the academic lawyer, which is to treat the law as a academic discipline.66 [O]ne must either accede to the possibility that law can be understood through itself or deny the possibility that law can be understood at all. Perhaps it is hardly surprising that dissatisfaction with contemporary scholarship has caused exponents of ‘critical legal studies’ to explore this latter skeptical alternative.67 I do not seek for one moment to deny the fascination and significance of jurisprudence, law and economics, law and literature, and the like. I have long been fascinated

Unjust Enrichment’ in P Giliker (ed), Re-examining Contract and Unjust Enrichment: Anglo-Canadian Perspectives (The Hague, Martinus Nijhoff, 2007) 187. This assumes that Birks’ theory is the only theory worthy of the name, and Birks’ principles the only principles. 64

cf P Jaffey, Private Law and Property Claims (Oxford, Hart Publishing, 2007) 30. J Goldberg, ‘Ten Half-Truths About Tort Law’ (2008) 42 Valparaiso University Law Review 1221. Also working towards a more balanced view (though still too ready to damn externalism as destructive, rather than as describing necessary limits to internalism) is E McKendrick, ‘Taxonomy: Does It Matter?’ in D Johnston and R Zimmermann (eds), Unjustified Enrichment—Key Issues in Comparative Perspective (Cambridge, Cambridge University Press, 2002) 627. 66 A Beever and C Rickett, ‘Interpretive Legal Theory and the Academic Lawyer’ (2005) 68 Modern Law Review 320, 335–6. 67 Weinrib, The Idea of Private Law, above n 15, at 18. The choice of the crits as target here is particularly bizarre, given their repeated insistence that law is not simply the product of external political forces, but should be regarded as ‘relatively autonomous’. 65

210

Steve Hedley

by them myself. Without question they have a role to play in a modern law school. But to regard those studies as ‘what proper legal academics should be doing’ seems to me unacceptable. At the end of the twentieth century, the work of the practical legal scholar—and our working relationship with the judiciary—is too important to society for us to sell out to the departments of philosophy, history or economics.68 Waddams is clearly correct to say that map-makers’ claims about the importance of particular rules or decisions are invariably made from the perspective of their general theories of the law. But there is no other way that such claims can be made. There is no external perspective from which claims of importance can be made.69

Various principled reactions to these passages are possible. All the usual warnings about taking isolated statements out of context apply (though in each case, it seems to me that the context confirms that each writer meant every word). It would also be unfair to dismiss views merely because they happen to be overstated or inelegantly expressed (though again, unless we are prepared to dismiss entire articles as mere slips of the pen, the argument seems to miss what is being said). One could also retort that the barbarians at the gate are really public lawyers rather than other sorts of social scientist. ‘Sociology is not law’ is a claim with a certain naïve charm, and may possibly contain some truth; ‘public law is not law’ deserves no such indulgence. Yet this does not address the real problem—the fear of obsolescence in a new century—leading convinced internalists to deny that their externalist colleagues even belong in the law school. This needs to be tackled more directly. Almost by definition, those who concentrate purely on internal legal issues are unlikely to be very good at explaining why those issues matter. A solid argument for the significance of a particular approach involves asking why others should care about it, and what other solutions for the same problem are available—a reluctance to enquire into either will constitute a serious handicap. Building a beautiful theory of law is only half the battle—faced with such a theory, others may assume that its appeal is primarily aesthetic. If internalism is not to become a dead end, its proponents need to pay more attention to why their claims matter. With this in mind, I list eight points to bear in mind when considering this aspect. 1. The worth of a logical system cannot be evaluated from inside that system. What do they know of obligations, who only obligations know? The internal consistency of a legal model is only one feature of it, and is by no means a knockdown argument in favour of its acceptance; many extremely bad theories have been internally consistent.70 The question ‘why?’ will not go away, nor will it answer itself; and attempts to answer it without questioning the system’s definitions will only lead to circular claims. 68 69 70

A Burrows, Understanding the Law of Obligations (Oxford, Hart Publishing, 1998) 113 and 119. S Smith, ‘A Map of the Common Law?’ (2004) 40 Canadian Business Law Journal 364, 380. P Schlag, ‘Law and Phrenology’ (1997) 110 Harvard Law Review 895.

Looking Outward or Inward: Obligations

211

2. World-wide claims require world-wide knowledge. Anyone who claims that a certain theory is part of the common law, not limiting the claim to any particular jurisdiction, would appear to be saying that it is the law in every State in the United States, most of the United Kingdom and Canada, Australia, Singapore, Ireland, New Zealand, and a number of other places. Such a claim to knowledge is not usually credible; it could only be made by an exceptional individual with access to law libraries of superb coverage. If there is a serious claim there, it will usually be a rather lesser one; but if its own author cannot spell it out, it seems unlikely that their readers will be able to do so. 3. If the law is truly in chaos, it is most unlikely that some new theory can instantly restore order. Internalists have tended to assume that ‘order’ and ‘disorder’ in law are purely functions of the concepts the lawyers employ, so that a revolution in those concepts can radically alter the level of certainty in the law. But certainty in law has many roots. The law of murder will always be in conceptual chaos, because decision-makers feel strongly about these things and will not be dictated to by mere legal logic; medical negligence cases will always be harder to resolve than other negligence cases, because of the complexities of causation and the uncertainties of prognosis; juries are entirely uncertain in some respects and entirely predictable in others. A claim to be injecting much-needed certainty into the legal process needs empirical back-up if it is to be taken seriously, and is an unlikely claim if based solely on conceptual considerations. 4. Appeals to the ‘best explanation’ of particular areas of law are a dubious form of argument at best. Rationality is not ensured by a rigorous search for the ‘best explanation’ of the law, because the answer has already largely been determined by the choice of what it is that needs explanation. Framing is all. The ‘best explanation’ of the law on escaped snails may be one thing, the ‘best explanation’ of ginger-beer-consumers’ rights may be another, the ‘best explanation’ of the law of negligent injury may be a third thing; and there is no good reason to suppose that these ‘best explanations’ will cohere with one another. This ‘best explanation’ device, as used in court to resolve hard cases, has solid jurisprudential backing71 (how strong, is not my current concern); but its use in structural arguments does not, and relies heavily on ‘the classic formalist misconception that unsettled law and controversial cases can be resolved solely by conceptual analysis’.72 To paraphrase Llewellyn, a court has the legal issues defined for it by the parties, whereas in academic argument each participant defines the issues as they like, and ‘consequently always rides his straw man down’.73 71

R Dworkin, above n 18, ch 7. P Jaffey, ‘Contract, Unjust Enrichment and Restitution: The Significance of Classification’ in P Giliker (ed), Re-examining Contract and Unjust Enrichment: Anglo-Canadian Perspectives (The Hague, Martinus Nijhoff, 2007) 227. 73 K Llewellyn (unpublished), quoted in W Twining, ‘Some Scepticism about Some Scepticisms’ (1984) 11 Journal of Law and Society 137, 163–4. 72

212

Steve Hedley

5.

Not many people read Aristotle, Justinian or Kant any more. There was a time—not really very long ago—when a classical education was the mark of an educated person, and a philosophical grounding a mark of special distinction. That being so, any would-be legal theorist would be sure to include ample references to the classics or philosophy. But those days are gone, there is too much else to learn, and those traditional disciplines must compete in the market-place of ideas with many others. It may be a mistake, therefore, to treat a proven connection between a favourite theory and that of the ancients as an advantage. If it is suggested that the connection is a useful one, some explanation is called for. And someone who is offended by such questions as ‘Who cares what Immanuel Kant thought?’ may not be an effective advocate for their cause.74 6. No one has a monopoly on ‘rights’ or ‘justice’. To base your theory on either is to invite comment from those with a different view, very probably ‘external’ to your conceptual world. These days, anyone who has time to engage in political argument will have some sort of a view on both, and on the extent to which they should be reflected in law. (My UK audience will know that even The Sun has a view on these matters, though it’s not pretty.) These concepts can no longer be used in a purely technical sense, as if lawyers could define them to suit themselves; there are even those who believe that rights are pre-legal, or may be better off not being protected in law.75 Arguments about rights are not internal legal arguments any more—if they ever were. 7. History is on no one’s side. When intellectual history—legal or otherwise—is done in a hurry, it often results in hasty over-generalisations: The Romans thought ‘x’, or eighteenth century legal theory said ‘y’, or no one ever thought ‘z’ until Lord Denning said it in 1969. This apparent neatness is seen by some internalists as a confirmation of their approach: earlier legal thought can be characterised by these unambiguous general descriptions, so why not continue the habit? But when we have the time to do our legal history properly, this certainty falls apart. At any period, there is argument over how to characterise legal notions. The legal system is always in flux. Even when some legal institutions are uncontroversial, there might be unfathomable differences on why they are uncontroversial. So while historical argument is always a useful resource, properly done it will never yield unambiguous answers. And (contrary to the views apparently held by some) there is no special category of internalist legal history, by which we are entitled to attribute certain views to historical figures to satisfy legal theory, when the facts suggest that their views were otherwise. ‘Actual assertions about the past should be tested and if

74 For a limited defence of the use of Roman law in modern common law contexts, see J Lee, ‘Confusio: Developing English Private Law through Reference to Roman Law’, paper presented at the Fourth Biennial Conference on the Law of Obligations, Singapore, July 2008. 75 See especially M Dembour, Who believes in Human Rights? Reflections on the European Convention (Cambridge, Cambridge University Press, 2006) especially ch 8.

Looking Outward or Inward: Obligations

213

they turn out to be false, should be contradicted’.76 Do legal history properly, or don’t do it at all.77 8. No one does policy well. It follows that no one is excused from doing it merely because they may do it badly. A frequent internalist argument is that judges and legal academics are no good at policy formation, so they should leave it to others. But (leaving aside the point that we have no sure idea where ‘policy’ begins or ends) there is a gap in the argument. No person alone is ever good at policy, because sound policy can only be the product of multiple viewpoints. Whether legal officials are the least-worst people to be trusted with such power will require case-by-case consideration—it is not enough simply to assert that legal resort to policy is an admission of failure.78 There are really two objections here—that lawyers may lack the knowledge to decide these issues, and that they may lack the legitimacy to do so. Neither argument applies in all circumstances (lawyers have both knowledge and legitimacy in many areas). This objection, then, cannot be applied with a broad brush, but needs discrimination.79 Let the political merits of a pure internalism be clearly recognised, so that there is no mistaking its worth. The notion that law is an autonomous science, which yields definitive answers without the need for external intervention, is an extremely valuable one, with deep cultural roots. Much like the myth of Father Christmas, properly employed it can encourage proper behaviour, divert greed into socially useful channels, and promote the impression that virtue is rewarded and its converse punished. It also has a useful educational role: eventually, young people who initially believe the myth will realise for themselves that they have been misled, which will teach them further valuable lessons on the reliability of official pronouncements—lessons which might not be understood if explained in direct language. The difficulty comes when the entire academic establishment is meant to pretend that Father Christmas exists, simply because it would be nice if he did. If ‘internal’ legal theories can only be defended by these means, then they do not deserve to survive in the academy, however they may fare outside it.

VI. The Poverty of Either Approach Alone Neither internalism nor externalism is a viable philosophy on its own. To ignore internalism is to miss much of what makes law ‘legal’; to ignore externalism deprives 76 S Waddams, ‘Private Right and Public Interest’ in Michael Bryan (ed), Private Law in Theory and Practice (Abingdon, Routledge-Cavendish, 2007) 3, 4. 77 Compare R Gordon, ‘Critical Legal Histories’ (1984) 36 Stanford Law Review 57. 78 A Beever, ‘Policy in Private Law: An Admission of Failure’ (2006) 25 University of Queensland Law Journal 287. 79 For an attempt to describe ‘policy’ in narrower and more defensible terms, see A Robertson, ‘Constraints on Policy-Based Reasoning in Private Law’ ch 11 of this book.

214

Steve Hedley

law of any point. And it is all one thing. It’s not that you can’t separate outside from inside, public from private; it’s that nothing makes any sense if you do.80 All legal thinkers need to know this, but some (I have argued) learned it long ago. Every externalist will already be entirely familiar with internalist ways of thought (many law courses consist of little else), and externalist scholarship, whatever its weaknesses, does not suffer from a failure to appreciate internalism’s merits. It is the internalists who have retreated to their bunkers, and have recently dug in deeper. They need to rediscover that, like it or not, they are part of a wider political system. They can only learn about that wider system by talking to public lawyers and even non-legal academics, however unappealing that prospect may now seem. They will learn yet more if they will take press criticism of their discipline seriously, as newspapers are still the main conduit through which public views are conveyed. As ever, Tom Stoppard had it right—‘I’m with you on the free press. It’s the newspapers I can’t stand’—but it is too easy to dismiss genuine public fears on the way law is developing, merely because most journalists do not have law degrees. More generally, the study of the common law should not become a nineteenth century theme park, or an exercise in disgust that the world has moved on from then. In the modern age, private rights are actively made and re-made every day, not merely inherited from the last generation of lawyers—and we all have a responsibility to ensure that they are made as well as they can be. It is absolutely true (as several internalists have insisted) that traditional legal doctrine is very bad regulation, but we must be very careful when we decide which way that cuts. Does it mean that we must abandon the idea that law is regulation (on the basis of the rather snobbish idea that common law is above that)? Or does it mean that considerable improvement is needed in existing legal provision? Neither internalism nor externalism makes sense without the other. They must learn to live together.

80 Compare M Scordato, ‘Reflections on the Nature of Legal Scholarship in the Post-Realist Era’ (2008) 48 Santa Clara Law Review 353.

9 Treating Like Cases Alike: Principle and Classification in Private Law CHARLIE WEBB*

I. Introduction In private law as elsewhere, it is both common and uncontroversial to say that we should treat like cases alike. But what does this mean? Without more, the simple proposition that like cases must be treated alike does not take us very far, for it leaves the crucial question unanswered: when are cases alike and so call for like treatment? Though it is often concealed, this is a question that lies at the heart of recent debates on the classification or taxonomy of private law. Classification involves forming from the mass of legal claims groupings of individual claims on the basis that they share such common features as to make them suitable for collective categorisation. So, to say that a given case or claim ought to be categorised in a particular way is to make a claim about what other cases or claims it resembles and with which it falls to be treated. Disputes over the proper classification of claims are disputes over what cases really are alike, what likenesses and differences matter. As such, classificatory schemes seek to offer—and are valuable precisely because they do offer—an answer to the question of what it is that makes cases materially alike and unalike. In this essay I shall begin by taking a look at recent attempts to develop a classificatory scheme for private law, and, in particular, at the most prominent and widely supported of these—the event-based classification proposed by Peter Birks. Insofar as it focuses attention on the justifiability of the divisions found in law and employed in legal reasoning, classification matters. Like cases must be treated alike, and a basis or structure for identifying likenesses is essential. But, of course, what matters most is not that such questions are addressed but that they are answered correctly. Here, I shall argue, Birks was less successful. An

* Thanks to Robert Stevens, Hanoch Dagan and Stephen Smith for their comments. My attendance at the Obligations IV conference was facilitated by an overseas conference grant from the British Academy.

216

Charlie Webb

examination of the failings of Birks’ scheme, however, provides the opportunity to think again about what we mean when we claim that cases are alike or unalike and hence how a successful classification of private law might take shape. In the final sections, I shall offer a few conclusions on how arguments as to the likeness of cases may be understood and resolved.

II. Legal Classification As a starting point, we may consider what classification seeks to achieve. It is pretty clear that classification is not an end in itself. Rather, the value of legal classification lies in its contribution to the pursuit of other goals we consider valuable. These goals are, or may be, numerous.1 Broadly, however, I think we can see two possible objectives for a classificatory scheme. First, our aim may simply be to present the law as it is in a more easily accessible and digestible way. Without some such arrangement, all we would have is an unordered mass of cases and statutes, and, while all the same information would still be there, it would make getting to grips with it that much harder. For instance, at least one reason why we divide law degree courses into modules such as criminal law, administrative law, tax law and torts is to make the law easier for students to learn and make sense of. Not only do such divisions break down the vast heap of cases and legislation into smaller, more manageable pieces, but each of these pieces also possesses a coherence and unity—though, as these examples show, the nature and bases of that unity can differ—which makes the relevant material more readily understood and integrated. Similarly, some such ordering also helps when it comes to finding out what the law says on a particular issue. So, if a judge wants to know what sum of damages to award for the loss of a limb or if you want to know what permission you need to build an extension onto the back of your house, then having books on damages and on planning law is a lot better than leaping blindly into the law reports and statute books. Secondly, we may want a classificatory scheme to contribute to the development of the law, to shape or influence judicial and legislative decision-making beyond simply providing a neatly ordered account of what the cases and statutes have said up till now. On this approach, classificatory schemes are to be followed, providing

1 For discussion of the possible goals of legal classification, see P Birks, ‘Equity in the Modern Law: An Exercise in Taxonomy’ (1996) 26 University of Western Australia Law Review 1, especially 1–4 and 97–9; P Birks, ‘Definition and Division: A Meditation on Institutes 3.13’ in P Birks (ed), The Classification of Obligations (Oxford, Oxford University Press, 1997); E McKendrick, ‘Taxonomy: Does it Matter?’ in D Johnston and R Zimmermann (eds), Unjustified Enrichment: Key Issues in Comparative Perspective (Cambridge, Cambridge University Press, 2002) especially 632–8; S Smith, ‘Taking Law Seriously’ (2000) 50 University of Toronto Law Journal 241; E Sherwin, ‘Legal Positivism and the Taxonomy of Private Law’ in C Rickett and R Grantham, Structure and Justification in Private Law (Oxford, Hart Publishing, 2008) 115–21.

Treating Like Cases Alike

217

both direction when deciding novel points of law and a standard against which existing law is to be judged. Correct classification, on this view, aims to ensure not only that the law is applied consistently but that it is consistent. Within such a scheme, each category operates in much the same way as a rule,2 stipulating that all cases falling within it are to be dealt with in the same way, by the application of a common set of tests and standards. So, on this basis, to say that a given case or claim is, say, a contract case or an unjust enrichment case is to make a claim about how that case should be dealt with, whether it is to resolved through the application of one set of rules and principles or another. It is important to distinguish these two broad aims of classification because, though not antagonistic,3 they ultimately point in different directions and, as such, support differently arranged classificatory schemes. Borrowing Cane’s terminology, I shall call them, respectively, expository and dispositive classification.4

A. Expository Classification Expository classification aims to help those who want to know what the law is and those who want to communicate this to others. When devising an expository classification, the objective is to find or form groups from the mass of cases and statutes which will make the law more easily conveyed, learned and understood. As such, the qualities we are looking for in such a scheme are clarity, relevance, accuracy and comprehensiveness; in essence whatever qualities make for good and effective communication. It follows that both the choice between competing alternative expository schemes and the decision as to where to place individual cases within that scheme must turn on an evaluation of just those factors. The aim is to make the law more easily understood and articulated and so we want the relevant material to be arranged in such a way as communicates this information as clearly, as accessibly and as accurately as possible. This has a couple of important implications. First, within expository classifications boundary disputes should be comparatively uncommon and largely insignificant. In contrast to dispositive classification, within expository schemes nothing follows from the inclusion of a given case within one category rather than another. Such schemes seek simply to present the law as it is, in a way which makes it easy to learn, understand and articulate; they are not designed to influence development of the law or to act as a guide for judicial decision-making.5 2 Indeed, one might say that the advantages of classification, principally clarity, certainty and consistency, broadly equate to the advantages of rules, or at least rule-based decision-making. As such, it is not at all surprising to see that the strongest advocates of legal classification are also the fiercest critics of judicial discretion. 3 Indeed, we have need for both forms of classification, though it then becomes even more important to keep them distinct. 4 P Cane, The Anatomy of Tort Law (Oxford, Hart Publishing, 1997) 200–202. 5 Save to the extent that both legal development and judicial decision-making take as their starting point an understanding of the law as it stands.

218

Charlie Webb

Accordingly, particular classificatory choices do no more than reflect decisions as to how the relevant material may best be conveyed and understood. As such, debates as to the proper classification of a given case (or of the legal rights and rules it supports) are appropriate only where its inclusion in one category rather than another is likely to lead to it, or the material with which it will be classed, being better understood or more easily located. In many cases, there will be a range of more or less equally reasonable classificatory choices, between which we should have no particular preference.6 Secondly, as we have seen, the aim of expository classification is to make things easier for people who want to find out what the law is and to pass this information on to others. However, people’s needs and interests differ and so what may be helpful for one person may not be helpful for another. This bears on the categories we employ. For example, for those whose concerns do not extend beyond an isolated point of law or setting, a series of contextually defined categories such as business law and consumer law are likely to be of most use. By contrast, people seeking a broader understanding of the law and its aims will gain more if the law is divided up on functional grounds, by reference to the interests it protects and/or the principles it embodies. This leads us to divisions such as those found in the law of obligations, between contract and torts, and indeed the sub-division of the latter into trespass, defamation, nuisance and so on. As such, at different times we have need for different sorts of category and so different divisions. However, this is only problematic if we have to make a choice between categories such as consumer law and categories like torts and contract. For expositional purposes, we do not. As noted already, when our focus is expository classification, nothing follows from the inclusion of a given case within one category rather than another. Expository schemes describe the law as it is, and the same material is described wherever the divisions are drawn. One consequence of this is that, as we have just seen, boundary disputes are largely insignificant. Just as importantly, however, it follows that we have no reason to object to the inclusion of the same material in more than one category. This simply means that the same information is being provided in a variety of ways, some suiting some people and some purposes, some suiting others. Accordingly, we can accommodate a variety of expository classificatory schemes employing different categories. Similarly, there is no objection to the categories employed within a single expository scheme overlapping or intersecting. Once again, the sole objective of expository classification is to see that the law is described effectively. As such, the only issue raised by overlapping

6 So, if we are concerned simply with effectively communicating the law, it makes little difference whether, for example, the rules on rescission are placed into a category called ‘contract’, which covers the law dealing with the making and effects of promises and agreements, or are put into a category headed ‘unjust enrichment’ which looks at the ramifications of non-consensual transfers. Indeed, as we are about to see, within an expository scheme, we should have no objection to including them in both (as indeed mirrors the practice of texts and courses in these areas).

Treating Like Cases Alike

219

categories is that our classificatory scheme will present an account of the law which is in certain respects repetitive. That overlapping categories can be accommodated within an expository scheme should come as no surprise. After all, all undergraduate law courses pull together modules reflecting divisions drawn on a variety of different grounds—sometimes functional, sometimes contextual, occasionally jurisdictional—with ample scope for duplication. So, for instance, the category ‘medical law’ covers, amongst other things, plenty of material found within the category ‘torts’. But this does not mean that law schools must choose between offering one or the other. At most it means that course convenors have to get their heads together to ensure that students do not end up doing the same thing twice.

B. Dispositive Classification The ambition of expository schemes does not extend beyond providing an account of the present state of the law that is clear, accurate and easily navigated. In particular, they make no contribution to judicial decision-making save that, given (or to the extent that) law operates as a system of rules and we embrace a doctrine of precedent, the prior decisions of courts and existing legislation, simply by virtue of their pedigree, provide material relevant to how cases should be decided now and in the future. Accordingly, deciding cases now requires knowledge of what courts and legislatures have said in the past, and expository classification provides this information. Similarly, expository schemes offer no guidance as to the future reform or development of the law beyond the fact that any such proposal must take as its starting point, or point of departure, an understanding of the law as it stands.7 Dispositive classification is different. Dispositive schemes exist to shape or influence the content of the law, beyond simply reminding legal actors of what has gone already. As such, dispositive classification is a tool to be used in deciding cases and in considering future legal developments. A dispositive classificatory scheme is something to turn to for assistance when deciding novel points of law or when resolving disputes where the authorities are muddled or controversial. How can a classificatory scheme do this? The first point to note is that a classificatory scheme can assist in deciding difficult cases only if the classification of individual cases within that scheme is determinative of, or at least relevant to, how those cases should be decided. In other words, within a dispositive scheme, the classification of a case must tell us something about how the law should respond to it: what set of rules and principles we apply to it, what questions we ask and what tests we use. As such, to

7

See too Sherwin, above n 1, at 119.

220

Charlie Webb

claim that a given case be classified one way rather than another is to claim that the case falls to be treated one way rather than another. It follows from this that, for the purposes of dispositive classification, we class particular cases together precisely because they fall to be treated in the same way. If the classification of a case within a particular class determines that the law should respond to it in a particular way, the same must also be true for all other cases lined up next to it in the same category. Accordingly, the effect of grouping cases together within a dispositive scheme is to say that they call for a common response.8 The key question then is how we identify a series of categories which act as guides—good guides—to how cases should be decided. We shall turn to this shortly. Before then, there are a couple of further points worth making about dispositive classification. First, in contrast to expository schemes, when engaged in dispositive classification we are likely to see categories as necessarily mutually exclusive. If the classification of a given case is to determine, at least in part, how that case is to be resolved, and if cases within different classes will fall to be dealt with differently, then it makes no sense to say that a case can fall within or straddle more than one category.9 Secondly, and following on from this, within dispositive schemes boundary disputes are a lot more important. Since cases within different categories are to be treated differently, resolved by different sets of rules and principles, allocating a case to one class rather than another will (or is likely to) lead to a difference in result. So, if, for example, one claims that receipt of trust property falls within a category ‘unjust enrichment’ which provides for strict liability claims in respect of non-intentional enrichments, this suggests that the

8 I should clarify that when I say that under a dispositive scheme all cases classed together call for a common response or fall to be treated in the same way, I do not mean that all cases within a class must lead to exactly the same outcome (eg, A has a claim for £100 from B) or even to a common form or measure of liability. Rather, my claim at this stage is that, if a classification is meaningfully to be described as dispositive, all cases grouped together within a given class must, at the very least, fall to be decided by the application of a common set of rules and principles, and that all cases falling within other categories are to be decided by different sets of rules and principles. How those rules and principles apply to individual cases within a given class may vary. So, to claim that there exists a category of cases (for instance, headed ‘unjust enrichment’) that should be resolved by asking a series of questions (was the defendant enriched? did this enrichment come at the claimant’s expense? was the enrichment unjust? has the defendant changed his position? etc), and that these questions differ from those we should be asking when addressing cases within other categories, is to put forward a dispositive category. 9 In fact the issue is a little more complex since we need to take into account what Birks called ‘alternative analysis’ (see, eg, P Birks, ‘Misnomer’ in WR Cornish et al (eds), Restitution: Past, Present and Future (Oxford, Hart Publishing, 1998) 8–9), whereby different elements of a given set of facts can be emphasised so as to reveal discrete reasons for a legal response. For instance, if you snatch my wallet, then we can identify from those facts both a non-consensual transfer and a deliberate and wrongful infringement of my property interests. Such alternative analyses will provide independent reasons for a claim and often (though not always) different measures of recovery. So, in our example, while either interpretation may be thought to provide a good reason to require you to return the wallet or its value, only when viewed as a wrongful interference with my property do we have a reason for requiring you also to make good any further losses I might suffer. So, taking alternative analysis into account, we may say instead that any one reading or analysis or interpretation of a case must fall exclusively within one dispositive category.

Treating Like Cases Alike

221

liability that attaches to those who receive misapplied trust assets should not be dependent on their being in any way ‘knowing’.

III. Classification and Likeness We are now in a position to see how and to what extent classificatory schemes are concerned with likeness. For the purposes of expository classification, the only ‘likeness’ or connection that the individual cases (rights, rules) within a single category need have is that, by being grouped together, they are more easily found and understood by those wanting to know what the law is, or more easily communicated by those seeking to pass this information on to others. I should stress that I am not claiming that this is the only connection that cases grouped within a particular expository category will have. On the contrary, it is likely that grouping such cases together will make them more easily understood and communicated precisely because they possess some other, independent common feature—though that common feature may take any number of forms (contextual, functional, jurisdictional etc). Nonetheless, the key point is that such likenesses are relevant to the make-up of expository classificatory schemes only where and to the extent that they assist in providing a clear and accessible account of the law. As such, expository classification is not driven to follow and reflect any one conception of likeness or indeed any conception of likeness at all. By contrast, likeness lies at the heart of dispositive classificatory schemes and the divisions drawn within them. As we have just seen, the placement of a given case within a dispositive scheme determines how the law is to respond to that case. When cases are grouped together within a single category, it follows that they all fall to be dealt with in the same way. As such, each category within a dispositive scheme seeks to identify a collection of cases that call for like treatment. Cases in the same category are to be treated alike; cases in different categories are to be treated differently. We can now see that dispositive classification confronts, and seeks to answer, the question we posed at the outset: like cases must be treated alike, but how do we know when cases are alike so as to call for like treatment? Dispositive classification aims to provide us with groups of cases which are all to be treated alike. As such, when constructing a dispositive scheme, what we are looking to identify is groups of cases which are, materially, alike. At this point we may usefully turn to recent contributions on the question of private law classification. Seeing that dispositive classificatory schemes seek to identify groups of like cases, we might expect some insights into how we go about determining what cases are materially alike from the classificatory schemes private law scholars have put forward. However, as I shall argue, the scheme which has received widest support is misconceived and is likely to impede, rather than encourage, fair and principled decision-making. As such, we would do better to put it to one side and start again.

222

Charlie Webb

IV. Classification by Causative Event Recent years have seen a flurry of writing on classification in private law. The principal impetus has come from those working within the fields of restitution and unjust enrichment. Their concern has been to establish the credentials of the subject and to mark out its place within private law. This required ‘restitution’ and ‘unjust enrichment’ to be defined and distinguished from the law’s other organising concepts. This could not be done without some analysis of how and why we should classify and order legal rules and rights. As such, the claims that restitution and/or unjust enrichment be accepted as a discrete and independent branch of the law necessarily entailed a broader claim as to how the law should be arranged and presented. Accordingly, beyond the immediate focus on unjust enrichment, the contributions of unjust enrichment lawyers to the issue of legal classification extend to, and seek to shape our understanding and analysis of, all aspects of private law.10 Of the various classificatory schemes or approaches that have emanated from unjust enrichment lawyers one is now dominant—the event-based classification put forward by Peter Birks. Birks’ claim was that all legal rights arise from particular factual occurrences or events and that we can therefore classify such rights by reference to their causative event, that is, the type of event from which the right arises. Birks identified four general classes of such causative event— manifestations of consent, wrongs, unjust enrichment, and miscellaneous others—and these then provided the four basic categories of his classificatory scheme.11

A. Expository or Dispositive? Given what I have said already about the possible objectives of classification, the first question we should ask is what sort of classificatory scheme was Birks putting forward? The answer to this is not as straightforward as one would imagine. Birks never labelled his scheme as expository or dispositive, and at no point did he draw, or indeed appear to recognise, any such distinction. Moreover, when one turns to the substance of the scheme, Birks himself sent out rather mixed signals. 10 As we can see in the way that classificatory approaches first adopted in the context of restitution/unjust enrichment are increasingly being applied to other branches of private law. See, eg, P Birks, ‘The Concept of a Civil Wrong’ in DG Owen (ed), The Philosophical Foundations of Tort Law (Oxford, Clarendon Press, 1996); P Birks, ‘Equity in the Modern Law’, above n 1; R Chambers, ‘Constructive Trusts in Canada’ (1999) 37 Alberta Law Review 173; J Edelman, Gain-Based Damages: Contract, Tort, Equity and Intellectual Property (Oxford, Hart Publishing, 2002); A Pretto-Sakmann, The Boundaries of Personal Property: Shares and Sub-Shares (Oxford, Hart Publishing, 2005); A Burrows (ed), English Private Law, 2nd edn (Oxford, Oxford University Press, 2008). 11 Accounts of this classificatory scheme can be found throughout Birks’ work. See, eg, Birks, ‘Definition and Division’, above n 1, at 17–23; and P Birks, Unjust Enrichment, 2nd edn (Oxford, Clarendon Press, 2005) 21–8.

Treating Like Cases Alike

223

Certainly, there are times when Birks seemed to deny any dispositive role for his classification. For instance, in relation to the category ‘unjust enrichment’, Birks claimed that whether a particular enrichment was ‘unjust’ was to be determined not by appeals to ‘abstract conceptions of justice’ but simply by looking down to the cases: an unjust enrichment is what the courts have identified as an unjust enrichment.12 Similarly, the category of ‘wrongs’ is determined not by reference to moral notions of wrongdoing but simply by asking whether the law described certain conduct as a breach of duty.13 In light of such statements, it may not be surprising that Sherwin has concluded that Birks’ classificatory scheme should be regarded as ‘formal’,14 meaning that it is merely ‘descriptive’ and ‘without normative implications’.15 This would make Birks’ classification purely expository. Despite this, I think it is pretty clear that Birks intended his scheme to act as a guide to judicial decision-making and future development of the law. This can be seen in the framework Birks set down for deciding claims in unjust enrichment,16 a framework into which existing law could be fitted but which was also to be used to resolve borderline or hard cases in the future.17 It is apparent too in the various changes to the law—in relation to matters such as total failure of consideration, resulting trusts, knowing receipt, restitution for services, change of position—which Birks considered necessary if the law of unjust enrichment, as defined within his scheme, was to achieve coherence and consistency.18 Similarly, Birks’ concern that the categories within a classificatory scheme must not overlap or intersect and must not otherwise be ‘bent’ only makes sense in relation to dispositive classification. As we saw earlier, when classifying purely for expositional 12 P Birks, An Introduction to the Law of Restitution, revised edn (Oxford, Clarendon Press, 1989) 18–19. 13 Hence Birks never addressed the question whether all the things the law described as ‘wrongs’ deserved to be treated as such. As such, his category included examples such as innocent conversion, where the wrong seems to be one of form rather than substance, and which the law appears to have regarded as a tort only by historical accident (see, eg, JH Baker, An Introduction to English Legal History, 4th edn (London, Butterworths, 2002) 399). This refusal to open up the law’s categorisation of such cases as wrongs appears more than a little incongruous given the enthusiasm of Birks and others to prise unjust enrichment claims away from the clasps of contract. See, to similar effect, P Jaffey, The Nature and Scope of Restitution (Oxford, Hart Publishing, 2000) 327. 14 Sherwin, above n 1, at 121. 15 ibid, at 111. cf P Jaffey, ‘Classification and Unjust Enrichment’ (2004) 67 Modern Law Review 1012, 1017–21, where the author notes that Birks’ categories are identified in formalist terms but are then treated as having dispositive effect. 16 Whereby, the availability of a claim in unjust enrichment was to be determined by asking the following questions: Was the defendant enriched? Was it at the expense of the claimant? Was it unjust? Does the claimant have a defence? (see, eg, Birks, Introduction, above n 12, at 21). Latterly Birks inserted a further question: What kind of right did the claimant acquire (ie, personal or proprietary)? (see Birks, Unjust Enrichment, above n 11, at 39). 17 Birks, Unjust Enrichment, above n 11, at 39–40; P Birks, Restitution: The Future (Annandale, NSW, The Federation Press, 1992) x. Moreover, this is just how the House of Lords understood the framework when they adopted it in Banque Financière de la Cité v Parc (Battersea) Ltd [1999] 1 AC 221 (HL). 18 Similarly, Birks, when arguing for the integration of equitable wrongs with common law torts and breach of contract, warned of the dangers of courts reaching decisions which contradict each other: ‘The purpose of good classification is to prevent this happening, by making it possible to apply like rules to like cases’ (Birks, ‘Equity in the Modern Law’, above n 1, at 27).

224

Charlie Webb

purposes, there is no necessary objection to mixed and overlapping categories. Clear and clean divisions are important only if the classification of individual cases affects how they fall to be dealt with. That Birks saw his classificatory scheme as working in this way is plain when one looks at the significance he gave to boundary disputes. A good example of this is provided by Birks’ discussion of claims to recover one’s assets following a successful vindicatio.19 A vindicatio is, in essence, a simple assertion of title to an asset in someone else’s hands. So if you steal my shoes, a vindicatio claim amounts to my asserting that those are indeed my shoes. The causative event of that right— my title to the shoes—will vary, depending on how I acquired the shoes in the first place. So, if, as is likely, I bought them, my title will arise from an event falling within the class ‘consent’. However, for Birks, if I am to go further and claim not only that the shoes you are holding are mine but that you must give them back, I am asserting a further right, a right I necessarily did not have when the shoes were still in my possession. Birks was interested in how that right—the right to have the shoes given back to me—should be classified. His conclusion was that the right clearly did not arise from a manifestation of consent or the commission of a wrong,20 so the choice came down to unjust enrichment or an event falling into the class of miscellaneous ‘others’. Birks’ conclusion was that this right belonged to the ‘others’ category—defining the causative event as receipt of an asset belonging to another—on the basis that, were we to view the event as an unjust enrichment, it would entail that such a right, and hence the vindicatio which it supports, would be subject to the defence of change of position. Birks saw this as undesirable since it would lead to a weakening in the law’s protection of property rights. Now the key point for the purposes of the present discussion is not whether Birks’ analysis of the vindicatio is a good one. Rather it is to see the significance that Birks placed on the classification of this right. Birks’ reason for preferring not to class it within the category of unjust enrichment was precisely because of the impact this would have on the law’s response to such cases. So, irrespective of whether Birks was right or wrong in his view of the scope of the change of position defence, it is plain that he saw the classification of individual cases as having direct implications for the way those cases fell to be treated. Accordingly, there seems to me to be little doubt that Birks saw his scheme as having, and intended it to have, normative effect. If so, his scheme should be regarded, and falls to be evaluated, as a dispositive scheme.21 19 See, eg, Birks, ‘Misnomer’, above n 9, at 21–6; P Birks, ‘Property and Unjust Enrichment: Categorical Truths’ [1997] New Zealand Law Review 623, 656–8. 20 Of course, you have committed a wrong in my example, but vindicatio claims are also available against those who come into possession of another’s assets entirely innocently. Hence we need an account of such rights which does not depend on the defendant’s wrongdoing. 21 Of course, there is nothing to stop Birks’ classification being employed purely as an expository scheme. But then it should play no part in legal reasoning or shaping future legal developments. Certainly the ‘correct’ identification and categorisation of a claim’s causative event become essentially pointless questions, since they would then have no bearing on how such claims fall to be resolved or justified.

Treating Like Cases Alike

225

What should we make of Birks’ classification? While Birks offered plenty of general arguments for the importance of legal classification, his reasons for pushing for a scheme classing rights by causative event are a lot harder to pin down. To the extent that Birks discussed alternative classificatory schemes, his preference for event-based classification appeared to rest more in convention—the established categories of contract and tort being, in Birks’ view, event-based—than on any reasoned conclusion that such a scheme was more useful or offered better guidance.22 If so, this provides rather poor foundations for a dispositive classificatory scheme. As noted already, dispositive classification is concerned with identifying what cases are materially alike and so call for like treatment. However, cases can be alike or unalike in any number of ways, and the key question such schemes must answer is which of these likenesses and differences count? If Birks’ classification is to succeed as a dispositive scheme then it must be on the basis that claims or rights are materially alike where they share the same causative event.23 But the truth of this claim is hardly self-evident and so one would have thought that the onus was on Birks to demonstrate why this conception of likeness is the one we should adopt. For whatever reason he did not, and as such the central premise of Birks’ classification, upon which its success as a dispositive scheme depends, went undefended.

B. The Problems of Constructing an Event-Based Classification An appraisal of Birks’ reasons for adopting an event-based classification is, therefore, made rather difficult given how little we know about what these reasons were. In the end, however, I do not think this matters much, since the deficiencies of Birks’ scheme—indeed of any attempt to classify by causative event—are revealed as soon as we consider how such a scheme would be set up and operated.24 If we are to devise a classificatory scheme which categorises legal rights by reference to the events from which they arise, the first thing we have to do is to identify the causative events of the various rights we want to classify. Now, the facts that need to be in place before a particular right arises will often be many, with each of these facts being necessary for the establishment of the right. Classification by causative event presupposes, therefore, that we can identify from this list of necessary factual elements one, or perhaps some combination, which can be regarded as the causative event, with the other facts to be considered in some sense

22 See, eg, P Birks, ‘The Law of Unjust Enrichment: A Millennial Resolution’ [1999] Singapore Journal of Legal Studies 318, 320–1; Birks, Unjust Enrichment, above n 11, at 20–8; cf Birks, ‘Definition and Division’, above n 1, at 27–32. 23 Though Birks himself never quite made this claim explicitly. Perhaps the nearest he came to this was his assertion that ‘[t]he law of unjust enrichment is the law of all events materially identical to the mistaken payment of a non-existent debt’ (Birks, Unjust Enrichment, above n 11, at 3). 24 This repeats an argument I make in C Webb, ‘What is Unjust Enrichment?’ (2009) 29 Oxford Journal of Legal Studies 215.

226

Charlie Webb

ancillary, or as forming the background to, the ‘true’ causative event. It is far from clear, however, how we are meant to do this. Take for example the right to damages for losses caused by a breach of contract. To make out such a claim the following facts and circumstances would have to have occurred or have been present: each contracting party having achieved majority (so acquiring contractual capacity), the formation of a contract—which can itself be broken down into the distinct elements of offer, acceptance, the provision of consideration, and the satisfaction of any relevant formalities—the defendant’s breach of that contract, and loss to the claimant as a result of that breach. Which of these is the causative event? Birks’ answer was that the causative event is the breach of contract and so he put such claims into the category of wrongs.25 But why? The problem is that they can all be said to be causative events, in that each one must be present if the claimant is to succeed in making out his claim. If any one is absent, then the claimant will have no such right, and, as such, each event is a cause of the right’s existence. Moreover, as each is a necessary, yet none a sufficient, condition of the existence of the right, it cannot be said that any one of these events is more causally important than the others. So, while it is true to say that the defendant is liable because he committed the breach of contract, it is no less true to say that his actions constituted a breach of contract, and hence generated his liability, because he had contracted to render the relevant performance. So, while the project of classification by causative event requires us to pick out one event from the list of necessary factual conditions for any given claim and mark it as the causative event, we are given no indication as to how we are to make this selection. One possible solution would be to say that the causative event is the final event in the series of necessary factual occurrences. For instance, it is only when the breach of contract is committed that the claimant’s cause of action accrues; only then can he demand damages from the defendant.26 On this basis, the breach of contract should be understood as ‘the’ causative event as it triggers the claim. But this soon creates problems. Even in our breach of contract example, we can question whether, taking this approach, it is the breach which should be identified as the causative event. Though breach without more renders the defendant liable to pay nominal damages, substantial compensatory damages will only be recoverable if the breach causes the claimant loss. If the claimant is to make out such a claim, then it is the causing of loss—an event which is distinct from the breach and typically post-dates it—which provides the final necessary factual element.27 Another example is provided by estoppel claims. Such claims are typically understood as requiring both a promise or representation by the defendant and

25 See, eg, Birks, ‘Equity in the Modern Law’, above n 1, at 54; Birks, Unjust Enrichment, above n 11, at 21. 26 Though the doctrine of anticipatory breach may cast doubt on this. 27 This point is even more clear in relation to those torts which are not actionable without proof of loss since here no claim will lie unless and until loss is suffered.

Treating Like Cases Alike

227

some act by the claimant in reliance on this promise such as would cause him detriment if the defendant did not then keep it. What is the causative event? Clearly the promise or representation is an essential element, but it seems equally clear that this alone is not enough and that it is only once some act of detrimental reliance has been performed that any claim can lie. As such, it is this, or perhaps even the defendant’s refusal or inability then to abide by his undertaking, that appears to provide the final trigger for such claims. Yet Birks classed such claims in the consent category.28 This is not plainly wrong, but nor is it plainly right, and it clearly shows that Birks himself did not take this ‘last shot’ approach to identifying causative events. The inappropriateness of such an approach is even more apparent when we consider that, in a number of cases treated by Birks as falling within the consent category, the causative event would instead be the fulfilment of a formality or registration requirement.29 However, the inclusion of a category such as ‘registration’ or ‘signature’, not to mention the separation of such cases from those where apparently similar rights are created or transferred without formality, would rightly appear incongruous in a classificatory scheme which has dispositive ambitions. So we are left with our initial problem: given that the existence of any particular right will depend on the presence of a series of distinct, or at least distinguishable, factual elements, identification of one or more of these as the causative event requires us to make a selection. Yet if this is true, and unless our choice is to appear arbitrary, we need to know how this selection should be made. In other words, we require some standard or set of criteria that can tell us which, from the list of necessary factual elements for a claim’s existence, is the causative event.30 Importantly, the deciding factor cannot be causal significance for the simple reason that, once we have our list of necessary factual conditions for the right to arise, as necessary conditions they will all be equally causally significant.31 As such, whatever these criteria are and wherever they are to be found, it necessarily requires an inquiry beyond simply asking what facts need to be present for a right to arise. In other words, the supposed first step in the construction of an eventbased classification—the identification of the causative events from which rights

28

See, eg, Birks, ‘Equity in the Modern Law’, above n 1, at 63. At least where the formality requirement is fulfilled after all other elements of the relevant transaction have been completed, for example, such as where a transfer of title is perfected only upon registration or by notice to some third party (as is required for the legal assignment of debts under s 136 of the Law of Property Act 1925). 30 And, if sometimes the selection seems obvious or some options appear more intuitively appealing than others, then this is because we are relying on some internalised, unspoken criteria for selection, guiding our views on which of these facts really ‘count’. 31 So, what makes an event ‘causative’ for the purposes of classification by causative event cannot be determined by reference to causation, in the sense of the factual contribution that event makes to the existence of the right. Of course, the language of causation can be, and has been, used in different ways, so it is possible that ‘causative’ events can be marked out as peculiarly ‘causative’ in some other sense. But then we still need to know what makes a right ‘causative’ in this sense, and this necessarily takes our inquiry beyond simply ascertaining the necessary factual conditions for the claim. 29

228

Charlie Webb

arise—must in fact be prefaced by an inquiry into how such causative events can be isolated from the list of necessary factual conditions. This is crucial because it is this prior inquiry and how we answer it that will determine how we identify the causative events that we are seeking to classify. But since how we identify causative events will directly impact on the make-up of our scheme—for instance, identifying the causative event in an estoppel claim as reliance rather than a promise will lead to a difference in how such rights are classified—it is this prior inquiry and not the notion of causative events which determines how the classification will take shape and how the material within it will be distributed. The problems do not end there. Once we identify the causative events of the various rights we are seeking to classify, we are still faced with the task of categorising them. This involves forming from the multiplicity of specific right-creating events groupings of such events on the basis that they are of the same kind, that they can all be regarded as particularised instances of a single generic causative event. But, once again, this is easier said than done. Birks gave us answers—here his four basic categories of consent, wrongs, unjust enrichment and others—but, as before, we are left without any explanation as to why these categories are correct or how they have been chosen. Events in the world do not come neatly packaged into pre-established categories. We have to do this ourselves. As with the initial identification of causative events, the key question is how this is to be done. So, say we want to know whether the receipt of money through the commission of a breach of contract or fiduciary duty is an event materially similar to the receipt of a mistaken payment. In certain respects these two sets of facts resemble one another; notably the receipt of money by the defendant in each case. But there are also some plain factual differences: in the former case the money came (or is likely to have come) from someone other than the claimant; in the latter case the defendant’s participation may have been entirely passive and indeed unknowing. So, these cases, and the claims that do or may arise from them, are factually similar in some ways and different in others. Because of this, it would be perfectly possible to draw up a generic description of an event which embraced them both—say, ‘claims arising from the receipt of money or other benefits’. But the question is not whether we could do this but whether we should. It is equally clear that we could choose to frame our categories so that such claims are kept separate, for instance by having separate classes of ‘wrongful benefits’ and ‘innocent benefits’, or by lumping the former claims in with all others that depend on the commission of a legal wrong and leaving claims of the latter type in a separate category of ‘unjust enrichment’.32 32 Birks famously adopted this last approach, recanting from his earlier ‘perfect quadration’ thesis, which saw all restitutionary claims as arising from the event ‘unjust enrichment’. Birks felt driven to this conclusion given that there was clearly a class of restitutionary claims where the defendant’s liability turned on the fact that he made his gain through the commission of a wrong against the claimant. Given that Birks already had a separate category of ‘wrongs’, this led to overlapping categories: these claims arose from the commission of a wrong, yet they also matched Birks’ initial conception

Treating Like Cases Alike

229

Answers are not to be found in the cases. Courts are not in the habit of identifying and categorising causative events, and, even if they were, one point of dispositive classification is to offer assistance where the law runs out or is unsettled. Dispositive classification also aims to take us outside the cases and to enable us to locate relevant equivalences and differences which the courts have missed. Indeed, this is why unjust enrichment lawyers became concerned with classificatory questions in the first place; they believed that they could identify a collection of like cases which the courts had overlooked. Nor is it enough to look to how the law responds to the cases we want to classify. For instance, it may well be true that the law responds in the same way to both of the cases we looked at in the previous paragraph, at least in so far as it recognises that the defendant is, or may be, liable to give up his gain. But this alone cannot tell us that these two cases ought to be classed together. Dispositive schemes exist to guide legal developments and to assist in the resolution of difficult and borderline cases. This cannot be done if the classification of individual cases is to be determined simply by the law’s response to them. Moreover, to adopt such an approach would effectively be to abandon the project of classification by causative event. Rather than the notion of causative events driving the scheme and directing us as to how cases should be decided, we would be looking at how cases are decided in order to identify their causative events. To take another example, we may ask whether the innocent receipt and use of the claimant’s goods without his consent is materially similar to, and so should be classed with, a deliberate punch on the nose. It may well be that English law sees both cases as involving the commission of a breach of duty, but, again, to treat this as determinative is to revert to classification by (one aspect of) the law’s response to the event. What we need to ask is whether it is right to treat them both as wrongs. Once more, if dispositive classification is to help us determine whether the law is consistent and coherent, the correct classification of any given case cannot follow simply from how the law presently responds to it.

of unjust enrichment, defined as it was to catch all instances of restitutionary liability. Birks’ solution was to cut down his definition of unjust enrichment so as to exclude all restitutionary claims which arise from some other event within his scheme (ie, consent, wrongs or others) (for a full account of Birks’ argument, see Birks, ‘Misnomer’, above n 9). Even if we accept, as for the purposes of dispositive classification I think we should, that overlapping categories must be avoided, the question is: why did Birks choose this solution rather than any other? Birks appeared to consider that the restitution for wrongs cases necessarily had to be put into the category of wrongs since it was plain that the defendant had indeed committed a wrong in such cases and that the claimant was specifically asserting the commission of that wrong in making his claim. But all that means is that the commission of a wrong is indeed one of the necessary factual conditions of such a claim, not that this must be treated as the causative event or determine its classification. That, in such cases, the facts on which the claim is based reveal, and are relied on as revealing, the commission of a wrong no more entails the conclusion that such claims must fall within the category of ‘wrongs’ than does acknowledging that claims for breach of contract require that it be shown that there was indeed a contract between the parties entails that such claims be placed into a category of ‘contract’ or ‘consent’. In truth we could come up with any number of different categorisations which would have avoided the overlaps Birks was so keen to excise, and so what we need is a basis for preferring one such set of categories over the others.

230

Charlie Webb

We can look at the factual elements of such claims as long and hard as we like but they cannot give us the answer. Any two claims are likely to resemble each other in some ways and differ in others. What we need is some basis for determining which resemblances and which differences matter. As before, the only option is to invoke some external set of criteria that can tell us when individual causative events should be viewed as materially similar and so suitable for collective categorisation. Once again, the events-based classification comes up short.33 It requires supplementation before it can tell us what we need to know. And, once again, though it is plain that Birks’ scheme offers answers to these questions, what is missing is any explanation of why his set of categories is preferable to any other. The upshot of all this is that, even once we are agreed on what combination of facts is needed for a given claim to arise, there will usually be no clear answer to the question of which of these is to be identified as the causative event. Nor is it plain how we are to determine which of these individual causative events should be viewed as specific instances of some broader generic causative event, such that they should be brought within a single category. Yet these two questions are key. The very possibility of classification by causative event depends on our being able to answer them, and the answers we give will determine how the scheme takes shape. However, in many, perhaps all, cases there is no one clear answer but instead a range of plausible, alternative answers, each of which will lead, inevitably, to a differently arranged scheme. What this means is that there is not one possible event-based classificatory scheme but many. As such, the immediate challenge to any given event-based classification is: ‘why this one?’

C. Deficiencies of Classification by Causative Event This point is significant. If we were concerned only with expository classification, it would not matter much that we can come up with a variety of alternative classificatory schemes, each of which arranges the law by reference to causative events. However, where our aim is to develop a dispositive classificatory scheme—a scheme which assists in the deciding of cases and the development of the law—this presents a serious problem, since it means that classification by causative event is capable of giving us a number of alternative, mutually inconsistent answers to the question of what cases and claims are alike and so call for like treatment. A scheme which identifies and categorises causative events one

33 This is demonstrated by the fact that other lawyers, who appear to share Birks commitment to classification by causative event, have come up with different categories of causative event: see R Grantham and C Rickett, Enrichment and Restitution in New Zealand (Oxford, Hart Publishing, 2000) 43–50; W Swadling, ‘Property: General Principles’ in A Burrows (ed), English Private Law, 2nd edn (Oxford, Oxford University Press, 2008) 366.

Treating Like Cases Alike

231

way will lead us to decide cases differently to one which identifies and categorises causative events another way. And, of course, they can’t both be right. It follows from this that the only way to ensure that an event-based classificatory scheme does in fact succeed in treating like cases alike is to make sure that we identify and categorise causative events in such a way as to ensure that like cases are indeed grouped together. But this then defeats the purpose of creating a dispositive classification based around causative events. As we have seen, the aim of dispositive classification is to assist in legal reasoning by telling us when cases are materially alike and unalike, or at least to indicate how material likeness is to be determined. As such, we would only choose causative events as the basis for such a classification if we considered that correctly identifying a claim’s causative event determines how that claim falls to be treated. In other words, the premise of such a scheme must be that material likeness is determined by examination of claims’ causative events; that claims are materially alike where and because they arise from the same basic causative event.34 But, once we admit that the only way to ensure that our event-based classification does in fact group like cases together is to reach a prior conclusion as to which cases really are alike, then we are abandoning the idea that it is the notion of causative events that determines material likeness and difference. To reiterate, if we want our scheme to group like cases together, given the variety of ways causative events could be identified and categorised, we must first consider which cases are alike and so should be grouped together and then identify and categorise their causative events accordingly. For instance, we cannot arrive at the conclusion that restitution for wrongs cases should be classed separately from mistaken payments, or that estoppel claims should be classed alongside contract claims simply by examining the lists of factual conditions necessary for such claims to arise. Their respective causative events and the categories between which they are to be allocated can be defined in a variety of ways, each avoiding overlaps and contradiction. Only once we have first decided whether these cases really are alike, and so call for common categorisation and treatment, can we answer the question of how their causative events should be identified and categorised and so present a scheme of legal rights classified by causative event which groups like cases together and keeps unalike cases apart. But what then is gained by such a classification? We may still choose to use the notion of causative events to mediate our presentation of what cases are materially alike, but this can only obscure the fact that it is not causative events that determine the structure of the scheme but a prior decision, founded on unspoken factors, as to which cases should be treated alike. So, while we may still be able to say that cases are materially alike when the rights concerned arise from events that are materially alike, this is not because it follows from the similarity in causative event that the cases call for like treatment. Quite the opposite; it follows from the prior conclusion

34

See above, text accompanying n 23.

232

Charlie Webb

that the cases should be treated alike that the relevant causative events are to be viewed as materially alike. The conclusion, then, is that if our concern is to develop a dispositive classification—a scheme which groups like cases together and so treats like cases alike—the notion of causative events provides no assistance and serves no useful function. We simply cannot answer the questions that need to be answered in order to build an event-based classification without first deciding which cases we regard as deserving like treatment. And, once we have these answers, an eventbased scheme serves no purpose, and indeed will only obscure why it is that we view these cases as materially alike in the first place. We would be better served bringing out into the open the criteria we apply in making these decisions and use these as the basis of our classification.

V. Reasons and Principles Which takes us right back to where we started, in search of some criterion which determines material likeness and so around which a dispositive classification can be based. The way forward is to think again about why we make claims about likeness. Sometimes the principle of treating like cases alike is regarded as bearing only on the administration of the law.35 So understood, the principle requires that legal rules are not applied capriciously or selectively but are instead applied equally to all those who are brought within them. As such, ‘treat like cases alike’ boils down to no more than ‘the law must be applied’ or ‘treat rules as rules’. In this form, then, the principle is rendered otiose. It tells us nothing that the relevant rule does not already tell us, and, in particular, has nothing to say on the proper content of legal rules. More commonly, however, the claim that like cases should be treated alike is seen to be wholly independent of and anterior to the law’s actual treatment of such cases. What we are looking at here is not how rules are applied, but whether the distinctions they draw and the unities they create are sound. On this understanding of the principle, we can question whether a legal system whose rules are clear and consistently applied does in fact treat like cases alike. Necessarily, therefore, the likeness of cases is not (simply) a product of the rules a legal system adopts. Cases can be alike even where the law distinguishes between them. Indeed, the principle that like cases be treated alike is most often invoked when and for the purpose of challenging some distinction presently found in the law. But this then brings us to the problem we identified at the outset: if material likeness is not to be determined by reference to existing legal rules, how is it to be determined? 35 See, eg, HLA Hart, ‘Positivism and the Separation of Law and Morals’ (1958) 71 Harvard Law Review 593, 623–4. cf HLA Hart, The Concept of Law, 2nd edn (Oxford, Clarendon Press, 1994) 159–63, where Hart acknowledges that this is just one possible application of the principle.

Treating Like Cases Alike

233

The first point to stress that likeness is not a matter of factual identity. Indeed, no two cases will ever be factually identical, if only because of differences in the time and location of events or the identities of those involved. It is for this reason that reference is often made, as it has been here, to material likeness. But simply re-framing the inquiry in this way takes us little further. It tells us that cases may be alike while being in certain other (non-material) respects unalike, but gets us no nearer to determining what likenesses and differences are material. The question we need to address is what it means to say that cases are materially alike despite being, in part, factually dissimilar. Any two cases or sets of facts may have any number of similarities and differences, and so may be alike in some ways and for some purposes and not for others. The key, therefore, is to consider what our purposes are when we make claims as to likeness. When we argue that cases are alike or different, we do so to make a claim about how those cases should be decided. If I say that cases A and B are alike, what I am arguing is that these cases, by virtue of their factual similarity and in spite of their factual differences, fall to be treated in the same way. If I claim that, irrespective of whatever factual similarities they may have, they are nonetheless materially different, my claim is that the law should respond to them differently. Like cases must be treated alike precisely because like cases are those that should be treated alike. This may seem entirely unilluminating. However, what it does is to highlight that the sorts of likenesses and differences we are looking at when we make claims about material likeness of cases are those that are relevant to the law’s response to them. Cases are materially alike where the factual differences between them have no bearing on the way those cases should be decided and, conversely, where the similarities between them require that they be decided in the same way. Disputes as to the likeness of cases, therefore, turn on whether the factual differences between them are sufficient to justify their being differently decided. How then do we know whether a factual difference supports a different legal response? This can be answered only by inquiring into the law’s reasons for attaching legal consequences to particular factual occurrences. A factual difference will be immaterial when it does not provide any reason for the law to treat the two cases differently. Cases will be alike where their facts provide the same reason or set of reasons for the law’s intervention. If the same reasons for the law to respond are present in both cases then the law should treat those cases alike.36 36 The argument in the text also suggests, I think, that it is misconceived to understand the principle of treating like cases alike as requiring consistency with past actions and decisions (for examples of the principle being understood in this way, see N Gillespie, ‘On Treating Like Cases Differently’ (1975) 25 Philosophical Quarterly 151; D Lyons, ‘Formal Justice and Judicial Precedent’ (1985) 38 Vanderbilt Law Review 495; cf A Marmor, ‘Should Like Cases be Treated Alike?’ (2005) 11 Legal Theory 27). So conceived, it requires that, once we have decided a particular case one way, we should decide all future like cases similarly. As such, in contrast to the understanding of the principle presented in the text above, in this form it applies only where we are faced with deciding a case which is analogous to a case that has been decided previously. Now, despite its narrower focus, this version of the principle is likewise dependent on an understanding or conception of material likeness, since this is needed to tell us when a later case is properly to be regarded as analogous to, or simply like, the first. In other

234

Charlie Webb

We can make the same point by looking again at dispositive classification. Within a dispositive scheme, the classification of a given case tells us how that case falls to be treated or dealt with. But if the classification of cases is to determine the law’s response to them, it must be that such classification provides or identifies or reflects a reason for the law to respond in a particular way (for instance, by holding the defendant liable to the claimant). Each category must, therefore, provide or reflect a particular reason or set of reasons which justifies attaching a particular legal response to all cases that fall within it. It follows from this that an effective dispositive classificatory scheme must be reason-based.37 Another way to express this is through the language of principles. Our generic reason for assigning a particular legal response to a given set of facts is that the response reflects or gives effect to principles38 we believe should be embodied in or effectuated by the law. Cases are alike where they raise the same questions of principle, where the principles which are or should be reflected in the law apply to them in the same way. Where cases raise the same questions they should receive the same answers. Disputes as to the likeness of cases are therefore disputes as to what are good reasons for legal decisions and/or their application to the facts at hand, or in other words, as to the principles which do or should shape the law and/or as to what the application of those principles to the facts requires. To determine which cases are alike and hence call for common treatment requires an analysis of the principles that shape and dictate the law’s response to them. How then do we identify these principles? Where do we find them? This will depend on the precise claim we are making.

words, the direction to act consistently with past actions requires that we can identify when new cases or decisions are sufficiently or relevantly similar to previous cases or decisions so as to invoke the application of the principle. So, if we come to decision X in case A, the principle of consistency in decision-making requires that we reach decision X in case B only if cases A and B are materially alike. But, if it is true, as I have argued, that material likeness can only be identified by reference to the reasons we have for attaching legal responses to particular combinations of facts, then any successful inquiry into material likeness will tell us not only that the relevant cases are alike (and hence that they should be treated alike) but also what particular treatment is justified in both cases. That is to say, the very criteria for determining likeness also determine just treatment; we can only determine whether two cases are alike by identifying what reasons the law has for responding to them, but, once we have identified those reasons, they will also determine what particular response the law should attach to them. As such, to interpret the principle of treating like cases alike as requiring (only) consistency with past decisions or actions is incoherent, for the very basis upon which likeness is determined tells us not that cases should be decided consistently with earlier decisions but that they should be decided consistently with (and through application of) the principles at stake. For much the same reason I doubt that it can be correct to refer to treating like cases alike as merely a/the principle of formal (as opposed to substantive) justice (cf Lyons, ibid; and N MacCormick, Legal Reasoning and Legal Theory (Oxford, Clarendon Press, 1978) 73). 37

The same conclusion is reached by Sherwin, above n 1, at 116; and Jaffey, above n 15, at 1014. Dworkin famously distinguished principles and policies (see R Dworkin, Taking Rights Seriously (London, Duckworth, 1977) 22–3 and 90–100). The reference to principles in the text is intended to embrace both of these, though I also accept that we can identify sub-divisions within this class of principles as broadly defined. 38

Treating Like Cases Alike

235

First, a claim that two cases are alike may be a claim made entirely independently from and irrespective of the principles which currently find expression in the law. The claim is, therefore, entirely prescriptive. Whatever the law may currently provide, in relation to such cases and elsewhere, we should view such cases as alike and so we should treat them the same way. If, and insofar as, the law does not back this up, this does not affect the validity of the claim, but instead provides an argument for the law to be changed. Such arguments necessarily invoke some ideal set of principles which the claimant believes should inform and shape the law, and, consequently, such arguments can be evaluated and resolved only through consideration and analysis of what we want the law to do, what values we want it to embody and what objectives we want it to pursue. This necessarily takes us into the realms of moral philosophy, economics, public policy, and whatever other fields may plausibly provide the standards and objectives of a legal system. Secondly, and alternatively, one may contend that, given the principles that a given legal system has chosen to embrace, or given the principles which best explain the practices of that legal system, we should view two cases as alike and hence as requiring like treatment.39 Such a claim, in contrast to the first, does require some analysis of existing practices. Any claim that a given set of principles underlies or supports the practices of a particular legal system involves a claim that those principles are given expression, if only implicitly, in those practices. As such, those principles must be shown to fit with the cases and legislation. However, if we are seeking to identify a set of principles which provides a synthesis or best explanation of existing authority, fit alone is not enough. We also need to inquire into the possible principles that a legal system may choose to embody and see which of these provide common threads between these authorities, providing a coherent and persuasive account of (at least certain core or salient features of) current legal practice. As such, even here we cannot avoid addressing the question of what the law is or may be here to do, and these arguments will, as before, require consideration of arguments of justice, utility, public policy and the like. Finally, though I hope this is already clear, it is worth repeating that on neither approach is the material likeness of cases to be determined simply by asking how the law presently responds to such cases. As we have noted, the very point of invoking the maxim that like cases should be treated alike is typically to test or challenge the rules we have. Necessarily, therefore, disputes as to whether cases are alike cannot be resolved by appeals to what cases the law treats as alike. In much the same way, disputes as to what the law should be cannot be settled solely

39 This sort of inquiry now tends to be labelled as ‘interpretive’ (see, eg, S Smith, Contract Theory (Oxford, Clarendon Press, 2004) 4–6; A Beever and C Rickett, ‘Interpretive Legal Theory and the Academic Lawyer’ (2005) 68 Modern Law Review 320). I discuss inquiries of this sort, and the problems they pose, further in C Webb, ‘Property, Unjust Enrichment and Defective Transfers’ in R Chambers, C Mitchell and J Penner (eds), Philosophical Foundations of the Law of Unjust Enrichment (Oxford, Oxford University Press, 2009) 336–41.

236

Charlie Webb

by reference to what the law is.40 What we are seeking to determine is whether the cases that the law treats as alike really are alike, and whether the cases it treats differently really are different.

VI. Implications Despite my rejection of the classificatory scheme introduced by Birks and embraced by many private lawyers, it should be plain that my argument is not motivated by scepticism as to the importance of classification in analysing and explaining the law. I agree with those who have stressed the potential of classification to aid analysis and affect the outcome of cases.41 The divisions and associations we draw in the law matter, since they both reflect and shape our understanding of the nature, basis and incidence of legal rights and claims. Classification requires us to think about the validity of the divisions and associations found in the law and ensures that we remain focused on the need to treat like cases alike. The contribution unjust enrichment lawyers have made in promoting the cause of good classification is to be applauded. Where I differ from such accounts is simply, but fundamentally, on the question of how and where we draw such divisions, on what it is that makes cases materially alike and unalike. The answer we give to the classificatory question is vital, for—just as good classification assists in our understanding of the law and our ability to develop a body of law which is principled and just—defective classificatory schemes create obstacles to such goals. At the very least they fail adequately to explain why we draw the distinctions we draw; at worst, they encourage and entrench distinctions which are unprincipled and unjust. It is a major failing of the approach of many unjust enrichment theorists to classificatory questions that they have consistently downplayed or disregarded the prescriptive aspects of the classifications they propose.42 The tendency has been to present them as (uniquely) latent in the case law, such that no justification is needed, and hence none offered, beyond an account of how the cases can be seen to fit into the proposed scheme. Aside from the inadequacy of such accounts to tell us why we should maintain such an approach to classification, it also opens up its advocates to accusations of inconsistency and hypocrisy, since the case law plainly reveals no such uniform classificatory scheme. Instead, the fit with

40 For present purposes, nothing turns on the answer to the question of where to locate the line between what the law is and what it should be. It is, I think, clear that such a line exists. One may prefer to express the same distinction in terms of what we do (or have been doing) and what we should do. 41 See, eg, A-G v Blake [2001] 1 AC 268 (HL) 290 (Lord Steyn); Foskett v McKeown [2001] 1 AC 102 (HL) 129 (Lord Millett); L Smith, ‘Unjust Enrichment, Property, and the Structure of Trusts’ (2000) 116 Law Quarterly Review 412, 422. 42 See, eg, Birks, An Introduction to the Law of Restitution, above n 12, at 27; Birks, ‘Equity in the Modern Law’, above n 1, at 3.

Treating Like Cases Alike

237

the cases is achieved only by marginalising, distorting or dismissing as wrongly decided those decisions which do not square with the approach of the writer. We are told that we should support the theory because it fits the cases, but we should support the cases only where and to the extent that (and, we must therefore presume, only because) they fit the theory. The cases are relied on where they fit the theory and dismissed where they do not.43 The circularity and deficiency of this position are plain. As we have seen, the likeness of cases cannot be determined by examination of our treatment of such cases. The failing of the accounts offered by the unjust enrichment school is not that they seek to change our understanding of the law or to question the correctness of some of the cases. Rather, it is to think that such conclusions can be justified simply by reference to the law as it stands. A classificatory scheme which seeks to influence future legal developments cannot be valueneutral. Instead, it must address the question of what differences and similarities matter, what makes cases alike or unalike. To answer these questions requires the application of standards which are in a sense external to or independent of the law, in that they are sourced not (only) in the legal rules we may currently have but in some theory of what law exists to do, what its goals and functions are, and, just as importantly, how best to achieve them. This, of course, means that the positions we take on these issues have no means of authoritative resolution, and for that reason will always be contestable. But the fact that the premises on which classificatory schemes are founded will necessarily be open to challenge, and that such challenges can only be addressed (and even then with no guarantee of resolution) through moral or political argument, must not lead us to conclude that such questions are unanswerable or, worse, that it is not the place of lawyers to speculate on them.

43 See S Waddams, Dimensions of Private Law: Categories and Concepts in Anglo-American Legal Reasoning (Cambridge, Cambridge University Press, 2003) 21–22 and 222–5.

10 Tort Law, Concepts and What Really Matters RODERICK BAGSHAW

I. Introduction This chapter explores the relationship between the capacity of tort law to achieve its goals and the nature of the concepts that are incorporated in the law. It seems unlikely that any attractive version of the substantive law of torts could relinquish the use of concepts. By ‘concepts’ I mean those categories of things which are grouped together and referred to with a single term or simple phrase, for example those categories we refer to when we use terms and phrases such as ‘rights’, ‘causal links’, ‘policy arguments’, ‘ultra-hazardous activities’, and ‘acts done in the course of employment’. Clearly the incorporation of concepts permits legal rules to attain a degree of generality that is important in allowing them to be efficiently communicated and applied. But it would be a mistake to assume that the only function of concepts is to enable a generalised presentation of laws which have already been fully specified at a more detailed level. I will argue below that a further function of many concepts is to enable a useful presentation of the law in advance of it being fully specified at a more detailed level, and that the form of the concepts which are incorporated in the law affects their capacity to perform this function. If this argument is successful then it will support the criticism that some tort theorists have unhelpfully restricted the capacity of tort law to achieve its goals by failing to appreciate both the different forms of concepts that can be incorporated into the substantive law and the significance of choices between those forms.

II. Different ‘Forms’ of Concept The simplest form of concepts is often called the ‘classical form’. Concepts that take this form can be defined in terms of some set of conditions (some of which may themselves be concepts). For instance, the concept of ‘widow’ might be

240

Roderick Bagshaw

defined using the concepts of ‘woman’, ‘marriage’, ‘husband’ and ‘death’. One advantage of choosing this form for a concept to be incorporated in the law is that it is easy to describe the process by which a lawyer or judge should determine whether particular facts fall within the conceptual category: the definition provides a checklist of criteria. It seems, however, that some concepts that are used in everyday settings do not exist in this ‘classical form’. For instance, Ludwig Wittgenstein famously discussed the concept ‘game’ and denied that any list of criteria existed that could determine whether each particular activity was properly referred to as a ‘game’.1 Moreover, he argued that the absence of a classical definition did not prevent ‘game’ from being a useful concept: people understand the concept of ‘game’ as a category including a cluster of examples, that are similar to each other in overlapping and criss-crossing ways, and further things which are similar to these examples. This different form of concept is often called ‘the prototype form’. If a prototype-form concept was incorporated in the law then the process by which a lawyer or judge could determine whether particular facts fall within the conceptual category would not involve a checklist but an assessment of similarities between the particular facts and the features of some prototype, or the features of each example within a cluster of related prototypes. Importantly, this process allows different sets of facts to be held to fall within the concept on the basis of strong resemblance to different features of the prototype. For example, both a twowheeled vehicle with a small motor and a three-wheeled pedal-rickshaw might be held to be sufficiently similar to the prototype ‘bicycle’, but in each case the explanation for the conclusion would emphasise the influence of different features of the prototype. Prototype-form concepts may be both useful and stable, so that it is not the case that there will inevitably be pressure to convert any prototypeform concept into one or more classical-form concepts. The category ‘games’ is sufficiently useful in helping us to order and understand the world without the unity of a classical checklist. Perhaps it might be thought that prototype-form concepts are primarily a feature of everyday contexts where precision is not usually important: the sorts of tasks undertaken by everyday speakers mean that they can make do with a prototype-form concept of ‘metal’, which may leave them struggling to classify correctly mercury and the metallic gases of Jupiter,2 whilst the sorts of tasks undertaken by chemists mean that they need a classical-form concept. But even if the claim that everyday speakers do not usually need precision could be empirically validated, this would not negate the possibility that there may be further advantages in not using classical-form concepts. In some circumstances the advantage may be that prototype-form concepts permit everyday speakers 1 GEM Anscombe (trans), L Wittgenstein, Philosophical Investigations, 3rd edn (Oxford, Blackwell Publishers, 2001) ss 65–71. 2 L Sixtrude and R Jeanloz, ‘Fluid Helium at Conditions of Giant Planetary Interiors’ (2008) 105 Proceedings of the National Academy of Sciences 11071.

Tort Law, Concepts and What Really Matters

241

to make use of propositions about categories of things, for instance ‘metals are often valuable’ or ‘metals often conduct electricity’, despite being unaware of the checklist for correct application of the classical-form concept and even if they are unable to access information about the matters on the checklist. Thus everyday speakers may be able to make use of propositions about ‘metals’ without having heard that ‘metals’ are distinguished by the way in which the atoms are bonded, or that hydrogen and helium are metallic under certain conditions, and without having access to direct information about the atomic bonds in any sample they are interested in classifying. To generalise, the suggestion is that ‘prototype-form’ concepts tend to use familiar and accessible features which may make them both easier to learn (teach) and more efficient to use. Studies by psychologists have suggested that even when people are aware of the classical form of a particular concept, they may still use prototypes in everyday reasoning about it.3 This tends to confirm that it may often be quicker and easier to use prototypes than classical forms, and these gains offset any reduction in reliability. Often this advantage in speed will be obvious: in most contexts it will take a far shorter time to make similarity judgements based on lustre, hardness, smoothness, and various other features of metal prototypes than to investigate bonds between atoms.4 If prototype-form concepts allow groupings which are useful despite the lack of any classical unity (eg ‘games’) and allow quicker categorisation even where a classical form might be possible (the everyday speaker’s ‘metals’) then it might be surprising if they were never to be found incorporated in the law. Indeed, I would argue that the prototype form provides a better model than the classical form for many concepts currently used in tort law, for example the concept of ‘employee’, when used in the context of defining a vicarious liability doctrine.5 Do scholars have good reasons to resist the incorporation of prototype-form concepts in tort law? It might be objected that the phrase ‘assessment of similarities’ provides only a thin and opaque description of the process on which the 3 For example, many people are apparently willing to rank even numbers in terms of how good they are as examples of the concept ‘even number’ (8 is commonly ranked as a good example, 30 as a bad one): see, SL Armstrong, LR Gleitman and H Gleitman, ‘What Some Concepts Might Not Be’ (1983) 13 Cognition 263, reprinted in E Margolis and S Laurence, Concepts: Core Readings (Cambridge MA, MIT Press, 1999). 4 Many of the features of prototype ‘metal’ are related to the type of bonding—‘metals are malleable and ductile because the electrons can adjust rapidly to relocation of the metal atom nuclei and there is no directionality in the bonding; they are lustrous because the electrons can respond almost freely to an incident wave of electromagnetic radiation and reflect it’: PW Atkins et al, Shriver and Atkins Inorganic Chemistry, 4th edn (Oxford, Oxford University Press, 2006) 71. I am grateful to Professor Andrew Weller and Dr Robert Jacobs for help with ‘metal’. 5 For instance, on the basis of discussion of the English case law, N McBride and R Bagshaw, Tort Law, 3rd edn (Harlow, Pearson Longman, 2008) 659, proposes ‘a “five indicia” test for determining whether a given contract to do work for another amounts to a contract of service or a contract for services. Under this test, if A has entered into a contract with B under which he undertakes to do work for B, that contract will be a contract of service if most of the … five requirements are met’. Another familiar non-classical concept may be the statutory extension of dependency actions to a person who ‘was living … as the husband or wife or civil partner of the deceased’: Fatal Accidents Act 1976 s 1(3)(b)(iii).

242

Roderick Bagshaw

operation of prototype-form concepts depends in controversial legal contexts. But the fact that the process is difficult to describe does not mean that no such process exists in human reasoning.6 Is it possible to make the process more transparent? We might start by noting that answers to the questions of what features should be considered, how similarity should be gauged, and how assessments of similarity with regard to individual features should be combined into a general assessment, will all reflect ideas as to why a particular selection of things is being thought of as a category and distinguished from other selections of things. So, for example, in judging whether a particular person is ‘similar to’ a prototype ‘employee’ we are unlikely to consider whether his weight is similar, or what volume of airspace each takes up, because the physical dimensions of people have no track record as providing a useful way of organising our thinking about the relationships they can have with someone who pays them for work. Further, the features most pertinent to an assessment of similarity are likely to reflect the prominent features of relevant rival prototype concepts. For instance, the familiar factors which textbooks suggest are useful when seeking to decide if a particular worker is an ‘employee’ are more tailored to distinguishing between ‘employees’ and ‘independent contractors’ than between, for instance, ‘employees’ and ‘volunteers’, or ‘employees’ and ‘family members performing regular chores for agreed rewards’. To make the point more specific, the degree of family relationship between payer and worker might be helpful in allocating a case between ‘contract of service’ and ‘family chores’ (because of the prominence of that feature in the prototype of ‘family chore-doer’) but would probably be judged irrelevant to determining whether a case involved a ‘contract of service’ or a ‘contract for services’. With regard to the combination of degrees of similarity assigned to individual features into an overall judgement, clearly if the facts with regard to every feature plausibly associated with the use of the categorisation point in the same direction then the overall judgement will be straightforward. But, equally, any fact which is judged to be highly unlikely in anticipated variants on the prototype may have decisive weight: for example, if a 15-year-old grandchild of a payer has agreed to wash the car, mow the lawn and vacuum the payer’s house each weekend in exchange for a moderate allowance, our provisional overall judgement may be that the worker is closer to the ‘family chore-doer’ prototype than any ‘basic employee’ prototype; but that judgement might be decisively changed by the further fact that the payer was deducting an amount from the allowance to pay for the child’s membership of a trade union which concentrates on representing part-time workers. To summarise, my claim is that when a lawyer or judge considers whether the paying grandparent is vicariously liable for some negligence of the 15-year-old in the course of washing the car, the lawyer or judge considers a rule which incorporates prototype-form concepts of ‘employee’ and ‘family chore-doer’ rather than 6 Philosophers, psychologists, cognitive scientists and others have investigated the nature of concepts and conceptual reasoning and produced a research literature rich in debate and controversy.

Tort Law, Concepts and What Really Matters

243

one which incorporates a classical-form concept of ‘employee’. In other words, the judge will consider whether the particular facts seem more similar to one of the ‘employee’ prototypes than to the ‘family chore-doer’ prototype; the judge will not act as if there is some checklist of criteria for being an ‘employee’.

III. Competition Between Concepts The idea of competition between rival concepts is familiar from well-known statements like the warning to lawyers not to use the philosophers’ concept of causation.7 But some tort lawyers have a dangerous propensity to treat some familiar concepts as natural rather than as products of design that must compete with rival products. The danger of this propensity is that the influence of the purposes of the designer on the shape of the concept will be overlooked, even when those purposes are in tension with the goals of tort law. Perhaps an illustration of the propensity of lawyers to lose sight of the fact that familiar concepts may have been developed for purposes different from those of the law of torts may be helpful. An example, exaggerated somewhat for the purposes of emphasis, starts from the observation that psychiatrists have developed various diagnostic criteria and labels in order to advance their purposes, such as determining the incidence of various conditions, and developing and sharing information about appropriate treatment. Some tort lawyers then borrowed these concepts to define the line that they required between a compensable injury (‘recognised psychiatric condition’) and an ill-effect (ordinary grief, distress and misery8), which reasonable people would seek to avoid but which could not be used as the foundation for a tort claim without unacceptable waste (and restriction on liberty). Now, such borrowing of concepts may often be both efficient and harmless: a heavy spanner will often do service as a hammer. But borrowing without reflection always carries a risk, and one that may become invisible if the concept later comes to be treated as natural. For instance, when psychiatrists highlighted ‘sudden trauma’ in their diagnostic criteria they were probably intending to draw attention to a potentially significant regularity rather than identify a factor which it would be fair, or efficient, to use as a condition of entitlement. Of course, once it is recognised that many concepts are products of design and that different forms of concept are possible, it becomes worthwhile to consider what makes a particular design for a concept better at serving ‘the goals of 7 See, eg, Yorkshire Dale Steamship Co Ltd v Minister of War Transport [1942] AC 691 (HL) 706 (Lord Wright): ‘Causation is to be understood as the man in the street, and not as either the scientist or the metaphysician, would understand it’. 8 See, eg, Alcock v Chief Constable of South Yorkshire Police [1992] 1 AC 310 (HL) 416 (Lord Oliver): ‘Grief, sorrow, deprivation and the necessity of caring for loved ones who have suffered injury or misfortune must, I think, be considered as ordinary and inevitable incidents of life which, regardless of individual susceptibilities, must be sustained without compensation’.

244

Roderick Bagshaw

the law of torts’. I suggest that there are three measures: (1) the efficiency with which a concept can be applied to distinguish situations, taking into account, in particular, the necessary investment it requires by way of determination of facts; (2) the accuracy with which a concept distinguishes situations which ought to be distinguished; and (3) the extent to which the concept communicates the reasons underlying the distinctions which are drawn using it. The third measure is, it seems, the least important within the common law of torts, in that the effectiveness of legal rules within tort law does not appear to turn to any noteworthy extent on the level of public understanding of why particular distinctions are so significant. Indeed, it seems to be assumed that tort law remains effective despite the fact that many expert practitioners and scholars cannot reach agreement as to the reasons which demand particular distinctions, such as that between ‘economic loss consequential on damage to a claimant’s property’ and ‘pure economic loss’.9 Consequently, in many situations identification of whether one design for a concept is better than another will reflect the outcome of a comparison of the compromises that each design makes between the first and second measures. Debates are common as to whether in a particular context the law should use a bright-line, generally indicating a design which scores highly on the first (efficiency of application) measure, or a more complex device, generally indicating a design for a concept, or combination of concepts, which scores highly on the second (accuracy of normative distinction) measure. Naturally the preferred compromise between the two measures is likely to vary from context to context within the law of torts. For instance, a concept that is easily applicable may be particularly valuable when defining a rule that it is important for people to apply to their own behaviour in the absence of legal advice. By contrast, it may be preferable to use a concept which is less easy to apply but more normatively accurate where it is to be used in the definition of a rule that will chiefly be referred to by expert lawyers advising in the aftermath of a catastrophic event. Choice of concept design in a particular context may also reflect opinions as to the stage of development of the law in a particular context and how it should be modified in the future. The adoption by appellate courts of bright-line concepts, which are efficient to apply, tends to limit the scope for legitimate innovation by inferior courts. Indeed, bright-line concepts are sometimes adopted in conjunction with an express prohibition on judicial development of the law and an invitation to the legislature to intervene.10 By contrast the adoption of highly flexible concepts, which may be expected to score highly for normative accuracy, gives 9 See C Witting, ‘Distinguishing Between Property Damage and Pure Economic Loss in Negligence: A Personality Thesis’ (2001) 21 Legal Studies 481, for both a summary of the inadequacies of previous explanations and a controversial new thesis. 10 See, eg, Frost v Chief Constable of South Yorkshire [1999] 2 AC 455 (HL) 500 (Lord Steyn): ‘In my view the only sensible general strategy for the courts is to say thus far and no further. The only prudent course is to treat the pragmatic categories as reflected in authoritative decisions such as the Alcock case [1992] 1 AC 310 and Page v Smith [1996] AC 155 as settled for the time being but by and large to leave any expansion or development in this corner of the law to Parliament’.

Tort Law, Concepts and What Really Matters

245

first instance courts far more freedom. Conclusions about the stage of development may also be related to opinions about the rectitude of current views as to what compromises will prove optimal: a designer who is not confident about the rectitude of current views is less likely to ‘set them in stone’. Unsurprisingly, there is also an opportunity for tort theorists to choose to use particular forms of concepts in a similarly strategic way. Thus a tort theorist who wishes to restrict opportunities for judicial innovation may choose to promote a structure for tort law composed from relatively rigid classical-form conceptual categories, whilst a tort theorist who is impatient for change may use more flexible components. With regard to other factors which may be relevant when choosing between classical and prototype forms, I noted above that it is straightforward to describe the process by which a lawyer should determine whether particular facts fall within a classical-form concept. Transparency of process is clearly an advantage, but the existence of a checklist of criteria might also seem to promise the more important benefits that determinations will be consistent and predictable, at least so far as fact-finding is consistent and predictable. These combined strengths of classical concepts mean that no legal system is likely to discard that form of concept lightly. But the incorporation into the law of prototype-form concepts may also produce relatively consistent and predictable results, even if the process by which prototype-form concepts assist decision-making lacks transparency. Common experience tells us that people who make similarity judgements based on lustre, hardness, smoothness, and various other features of metal prototypes will classify most samples (at room-temperature and atmospheric-pressure) consistently, predictably and correctly. Moreover, the previous paragraphs have suggested that prototype-form concepts may often allow reliable results to be obtained more quickly than a checklist, assuming that one is available, particularly where items on the checklist are less familiar or more difficult to investigate than the features around which a prototype is built. And, of course, I have suggested that any assumption that a checklist is available to define every concept that has proved its utility in everyday reasoning—for instance ‘games’—is unwise. Standing back, it seems safe to conclude that it would be unwise for any designer of tort law to reject the possibility of incorporating prototype-form concepts, given that in many contexts the efficiency with which they can be applied to distinguish situations—taking into account, in particular, the necessary investment the distinctions may require by way of determination of facts (measure (1))—will make up for any shortfall in accuracy and transparency.

IV. Interpretive Concepts So far I have identified two forms of concepts which may be incorporated into legal rules. But I do not intend to claim that these two forms, and their hybrids,

246

Roderick Bagshaw

exhaust the possibilities. My claim is simply that if a theorist wishes to incorporate concepts into rules so as to allow those rules to attain a useful degree of generality then the incorporated concepts need not take the classical form. As I advertised at the start of this chapter, I intend to go on to establish that some tort theorists have unhelpfully restricted the capacity of tort law to achieve its goals by failing to appreciate the different forms of concepts that can be incorporated. This proposition might be easier to establish if a further form of concepts exists which makes judgements by paying direct attention to purposes or goals, rather than attention to a classical checklist or prototypes. Does such a form of concepts exist? In explaining the concept ‘law’ in doctrinal contexts, such as where someone states that ‘the law in England on a particular point is x’, Ronald Dworkin develops the idea of ‘interpretive concepts’.11 For Dworkin the distinctive characteristic of these ‘interpretive concepts’ is that they are elaborated through assigning value and purpose to the practice in which they are used. Thus, his preferred elaboration of the concept ‘law’ maintains that the proposition ‘the law in England on a particular point is x’ is true if x ‘flows from [the] principles of personal and political morality that provide the best interpretation of the other propositions of law generally treated as true in contemporary legal practice’.12 People who want to know whether a particular example falls within an ‘interpretive concept’ do not have direct access to any checklist of criteria, but instead have an agreed technique: they must construct the best understanding of the scope of that concept in the practice where it is used and determine whether the example falls within this best understanding. Dworkin’s discussion of ‘interpretive concepts’ focuses on highly abstract concepts such as ‘law’, and political values such as ‘justice’, ‘equality’ and ‘liberty’. But his views seem to echo—albeit only with regard to such abstract concepts—the more general position that the structure of concepts ‘consists in their relations to other concepts within more general theories’.13 Could some of the less abstract concepts incorporated in legal rules be ‘interpretive concepts’? The practical consequence of finding that the concept of ‘employee’ in the doctrine of vicarious liability is an ‘interpretive concept’ might be that a lawyer or judge ought to determine whether a particular person is an employee, not by consulting a checklist or assessing similarity to relevant prototypes, but by considering whether ‘the best interpretation of “employee” as it is used in this part of the law extends to cover this person’. One reason for doubting whether concepts such as ‘employee’ can be interpretive is suggested by considering why, if judges and lawyers can determine what ‘flows from [the] principles of personal and political morality that provide the best interpretation of the other propositions of law generally treated as true in 11

R Dworkin, Justice in Robes (Cambridge MA, Belknap Press, 2006) 9–12. ibid, at 14. 13 Margolis and Laurence, above n 3, at 43–51, refer to this as the ‘Theory-Theory of Concepts’, in contrast to the ‘Classical Theory’ and ‘Prototype Theory’. 12

Tort Law, Concepts and What Really Matters

247

contemporary legal practice’, they express legal propositions using intermediate terms such as ‘employee’ at all—why not just announce that employers are vicariously liable when such liability ‘flows from [the] principles of personal and political morality that provide the best interpretation of the other propositions of law generally treated as true in contemporary legal practice’? The most obvious answer to this question seems to be that many of those who need to know the law do not have direct access to a reliable sense of what ‘flows from [the] principles of personal and political morality that provide the best interpretation of the other propositions of law generally treated as true in contemporary legal practice’. By contrast, they may well have a reasonably reliable sense of how to allocate cases by way of similarity between the prototypes of ‘employee’, ‘independent contractor’, ‘family chore-doer’ and ‘volunteer’. But if this answer is correct then it means that treating ‘employee’ as an ‘interpretive concept’ would be self-defeating. If concepts are incorporated into legal rules so as to allow those rules to attain a degree of generality that is important in allowing them to be efficiently communicated and applied, then it would be counter-productive to formulate the concepts in such a way that they could not be efficiently communicated and applied. Such a conclusion, however, does not deny that interpretation, in Dworkin’s sense, might be useful in choosing between or clarifying the prototype- or classical-form concepts which have been incorporated into the law. And it clearly does not deny that abstract concepts, which are more likely to figure in disputes between Herculean theorists than in quotidian legal advice, may best be thought of as interpretive in nature.

V. The Form of Concepts and Interpretive Legal Theory At this stage I will seek to substantiate the charge that some tort theorists have unhelpfully restricted the capacity of tort law to achieve its goals by failing to appreciate that different forms of concepts can be incorporated into legal rules. Thus, I will seek to demonstrate that the previous pages have been more than the preparations for a cavalry charge against a straw man. The primary concern that motivated me to write this chapter was the extent to which the need to choose between different forms of concepts seems to have been ignored in some recent exercises in ‘interpretive legal theory’ as applied to the law of torts.14 So far as the goal of an ‘interpretive legal theorist’ is to ‘reveal’ an ‘intelligible order’ in the present law, there must be a risk that the theorist’s conclusions will be incorrect if he or she assumes that the order can only be

14 A detailed description and defence of ‘interpretive legal theory’ is provided in A Beever and C Rickett, ‘Interpretive Legal Theory and the Academic Lawyer’ (2005) 68 Modern Law Review 320.

248

Roderick Bagshaw

composed of classical-form concepts. But such an assumption is surprisingly prevalent, and not just amongst those who would describe themselves as engaged in ‘interpretive legal theory’. For instance, such an assumption would appear to be made, for instance, by authors who critically discuss the concept of ‘assumption of responsibility’ from the perspective that the concept must be defective if all applications of it do not involve the presence of facts satisfying some other concept, such as ‘promise’ or ‘invited reliance’.15 The same assumption seems to have dogged many discussions of the concept of ‘proximity’.16 No doubt the utility of the latter concept is a matter for debate, but it seems to have often been assumed—I would argue incorrectly assumed—that ‘proximity’ will be unable to play any useful role in legal reasoning unless it is a concept in the classical form. A theorist who cast aside the assumption would be open to the possibility that, for example, ‘assumption of responsibility’ is a concept best understood as a cluster of prototypes and other situations which are sufficiently similar to one or more of these prototypes. I hope that these apparent examples of erroneous assumptions about the form of concepts will be sufficient, since my intention is to concentrate on demonstrating the deleterious effect that such assumptions can have on the capacity of tort law to achieve its goals, not to quantify the prevalence of such assumptions, still less to provide a register of those who have succumbed to the error.17 At this stage then, to make further progress, I must introduce and defend an account of ‘the goals of tort law’.

VI. The Goals of Tort Law Clearly the idea of ‘tort law’ having ‘goals’ is problematic because attributing ‘goals’ to a body of rules and techniques—tort law—involves a form of anthropomorphism that could easily lead to confusion. Is the reference to the goals of makers of tort law, of users of it, of practitioners who must settle disputes about its content and application? I intend to treat the question ‘What are the goals of tort law?’ as an oblique way of asking ‘What are the reasons why we should continue to tolerate the existence of tort law?’ One crucial feature of this reformulation is that it preserves the future-facing idea of ‘goals’. But I will also assume that any 15 See, eg, K Barker, ‘Unreliable Assumptions in the Modern Law of Negligence’ (1993) 109 Law Quarterly Review 461. 16 See, eg, WVH Rogers (ed), Winfield and Jolowicz on Tort, 17th edn (London, Sweet and Maxwell, 2006) 147: ‘According to the Privy Council … [p]roximity means “the whole concept of necessary [ie necessary to produce a duty of care] relationship between the plaintiff and defendant”. Now while this plainly introduces elements other than foreseeability, it is not, it is submitted, the formulation of a test or criterion at all, for it gives no indication of what, in any case, are the other ingredients of the necessary relationship’. It does not seem unfair to conclude from this that the author believes that useful ‘tests’ must consist of checklists of ingredients. 17 I am happy to confess that I have detected the error in my own previously published work.

Tort Law, Concepts and What Really Matters

249

answer must be conditioned by an artificial stipulation. The artificial stipulation flows from the fact that one reason why we ought to tolerate the existence of tort law might be that the immediate cost of eliminating it would be too great relative to the present, discounted, value of the uncertain longer term gains from any alternatives. So perhaps we should tolerate the continuing existence of tort law (or at least parts of tort law) because of the cost of change. But I doubt anyone would want to argue that one of the goals of tort law is to avoid the immediate cost of eliminating it. So I will treat the question ‘What are the goals of tort law?’ as equivalent to ‘What are the reasons why we should tolerate the continuing existence of tort law—on the assumption that there would be no transition cost in eliminating it or substantially altering it?’ Put like that, then, the primary reason—the primary goal of the law of torts— ought surely to be that the law of torts should be tolerated because it makes the world a better place than it would be in the absence of the law of torts. Who would recommend tolerating tort law (in a world where there would be no transition cost in eliminating it or substantially altering it) unless tort law is likely to make the world a better place? No doubt the proposition that the primary goal of the law of torts ought to be to make the world a better place begs a few questions!18 But it is a proposition which involves the immediate rejection, as insufficient, of the rival claim that the law of torts has no purpose other than to be the law of torts. One seemingly formidable response to this rejection might be the claim that the rights which tort law protects are rights which provide persons with equal liberty and are intrinsically valuable. But this response is not sufficient; it may justify the existence of such rights, but it is insufficient to justify public investment in legal enforcement of the rights through tort law.19 Legal enforcement necessarily carries costs, for instance defendants will incur expense in defending themselves against erroneous claims, and costs may be incurred in order to demonstrate the validity of genuine claims against defendants who are mistaken or doubtful about their obligations. The imposition of such costs cannot be sufficiently justified merely by establishing that the underlying rights have intrinsic value. To put the

18 For instance, one might ask how to determine whether an alteration leaves the world a better place where it is advantageous for some but disadvantageous for others, how to deal with uncertainty as to consequences, and how to choose (if this is necessary) between different and incompatible improvements. No doubt the difficulty of answering such questions might appear to enhance the attractiveness of a non-instrumental approach, but the difficulty is not a sufficient justification for such an approach. 19 Robert Stevens, a forceful advocate in favour of the view that ‘it is meaningless to talk of the law of torts having a function or goal at all’ (R Stevens, Torts and Rights (Oxford, Oxford University Press, 2007) 326) concedes that the intrinsic value of a right ‘does not mean that we should pay any price for a legal system which recognized and enforced such a right … [W]e might conclude that the enforcement of such moral rights by law was not worth the candle’ (at 333). It is hard to see why this is not an admission that an appeal to the ‘intrinsic value’ of rights is insufficient to demonstrate that we should tolerate the continuing existence of tort law. Tort law must justify its cost in ‘candles’.

250

Roderick Bagshaw

point a different way, even if tort law is similar to ‘love’,20 the intrinsic value of love is an insufficient (in the sense of incomplete) justification for spending US$200 billion to promote it in the United States each year. (This figure is, of course, one scholar’s assessment of the direct costs of the portion of the US tort system, which deals with personal injury, illness and death.)21

VII. Improving the world and ‘interpretive legal theory’ The claim that tort law’s primary goal is to make the world a better place might not appear to be in conflict with ‘interpretive legal theory’. Surely, an interpreter might conclude that the order of tort law is best explained by pursuit of such a goal? But, on closer scrutiny, there may be a subtle tension between identifying the best understanding of a static snapshot of today’s tort law as an intelligible order and the best understanding of today’s law as a ‘tool’ for continuing to make the world a better place in the future. Another related important contrast might be between an account of tort law based on the premise that the snapshot is of a completed project and an account of tort law based on the premise that it is a work in progress that will have to be further clarified in future in the light of changes in the world and, perhaps more importantly, changes in our knowledge of the effects of our interventions in the world. One distinguishing feature of the second type of account is that it is more likely to incorporate into the law techniques to guide that future clarification and development, and less likely to incorporate forms that assume that a full specification of the law is already available. Moreover, the best account of tort law as a work in progress may present it as currently relatively disordered, in the sense of lacking immediate internal coherence. One reason for such disorder might be that in different practical contexts designers of the law might have reached different conclusions about the appropriate balance between efficiency of application and normative precision, different conclusions about the extent to which the law should be fixed or left free to develop, different conclusions about the suitability of the law in the particular context for judicial development, different conclusions about the practical accessibility of the criteria that might form part of a classical-form concept, different conclusions about 20 Ernest Weinrib famously wrote that: ‘Explaining love in terms of extrinsic ends is necessarily a mistake, because love does not shine in our lives with the borrowed light of an extrinsic end. Love is its own end. My contention is that, in this respect, private law is just like love’. (EJ Weinrib, The Idea of Private Law (Cambridge MA, Harvard University Press, 1995) 6.) More recently, Robert Stevens has claimed that ‘although Professor Weinrib’s claim about private law in general is not only counter-intuitive but implausible, it is true about the law of torts’ (ibid, at 325). 21 KS Abraham, The Liability Century: Insurance and Tort Law from the Progressive Era to 9/11 (Cambridge MA, Harvard University Press, 2008) 3. The figure is derived from the TowersPerrin-Tillinghast estimate: US$247 billion as the cost in 2006.

Tort Law, Concepts and What Really Matters

251

the rectitude of current views as to what compromises will prove optimal, and perhaps different conclusions on these issues at different times. Moreover, these differences may be reasonable in the light of different degrees of knowledge about the effects that different forms of law are likely to have in these different contexts. Perhaps a metaphor will help to capture the difference between the two types of account. The second type of account of tort law will conceive of it as a city, a mixture of ruins, functioning buildings, construction sites, and rules and techniques for guiding its future development, and it will be an account which reflects the city’s primary goal of improving the world in the future. Clearly, such an account of a city may be very different from that which makes most intelligible the greatest proportion of the current city on the assumption that it is static and complete. Moving away from the metaphor, I would argue that it is not realistic to suppose that people in a society such as that in England will have the optimal freedom to choose a pattern of life and flourish provided that private law grants them some security for their persons, possessions and promises. For instance, it seems clear that many patterns of life also depend in a crucial way on interests in the behaviour of third parties, be they breadwinners,22 testators,23 potential employers24 or potential customers.25 Similarly, many patterns of life depend in a crucial way on the behaviour of governmental bodies.26 This said, I suspect that few theorists could be found who would maintain that the courts have identified the best available scope of protection for these important interests. There remains uncertainty both with regard to the value of seeking to protect various similar interests through tort law and the likely consequences of doing so in particular ways. If this is correct, then an account of the law which presents it as static, complete and coherent is likely to be misleading and may restrict progress towards better solutions. Moreover, the most helpful way of presenting the law in a relatively general form which is usable by those who cannot be expected to have access to either perfect information or ideal solutions may be through the identification of prototypes (and anti-prototypes) which can form a basis for ‘similarity’ reasoning.

22 In England the Fatal Accidents Act 1976 extends tort law’s protection to the economic interest of family dependants in the life of a breadwinner. 23 In England White v Jones [1995] 2 AC 207 (HL) extended tort law’s protection to a beneficiary’s economic interest in a testator’s choice not being impeded. By analogy, similar protection now exists where a professional’s negligence impedes a third party’s choice about pension benefits: Gorham v British Telecommunications Plc [2000] 1 WLR 2129 (CA). 24 Spring v Guardian Assurance Plc [1995] 2 AC 296 (HL) arguably imposed a duty to take reasonable care not to harm a claimant’s economic interests by misleading a potential employer about his suitability for employment. 25 A combination of torts protect a claimant’s economic interest in its potential customers not being misled (injurious falsehood), confused (passing off), threatened (intimidation) and unlawfully prevented from dealing (OBG Ltd v Allan [2008] 1 AC 1 (HL)). 26 The scope of tort’s protection of interests which depend on the behaviour of public bodies varies between jurisdictions, but most common law jurisdictions have created some special rights against public bodies. See, eg: Invercargill City Council v Hamlin [1996] AC 624 (PC); Pyrenees Shire Council v Day (1998) 192 CLR 330 (High Court of Australia); JD v East Berkshire NHS Trust [2004] QB 558 (CA).

252

Roderick Bagshaw

VIII. Conceptual Strictures: Rights I have suggested that those who seek a static and complete intelligible order in tort law may tend to neglect the utility of forms of concept other than the classical. In this section I want to draw attention to the extent to which some theorists have constrained the scope of tort law to develop by stringently limiting the form of central concepts. The field of recent tort law scholarship from which my examples are drawn might be described as ‘the debate as to the basic structure of torts’. A significant number of recent contributions to this debate have argued that the content of the law of torts depends on the scope of ‘rights’, and have additionally imposed strict conditions as to what can, and cannot, properly be considered a ‘right’.27 My claim, of course, is that the ‘conceptual strictures’ that these academic theorists have adopted are strictures with a propensity to impede tort law’s primary goal. A first stage in illustrating this propensity is to show how strictures about what can and cannot count as a ‘right’ can influence what a theorist defends as a legitimate part of tort law.28 An example might help. Recently Jason Neyers published an article entitled ‘Rights-Based Justifications for the Tort of Unlawful Interference with Economic Relations’.29 The article asked if this tort could plausibly be described as based on any right enjoyed by potential claimants, taking this to be a necessary first step in an investigation into whether the tort is consistent with any rights-based theory of tort law. In the aftermath of the OBG decision30 the English version of this tort can be presented as presupposing a duty not to cause harm intentionally to a claimant by using unlawful means to interfere with a third party’s liberty to deal with the claimant. Thus, one might think that it was simple to flip this duty and identify the claimant’s claim right behind the tort as being ‘a right not to have others intentionally cause harm to you by using unlawful means to interfere with the liberty of third parties to deal with you’. But Neyers never discusses this possibility. Instead he considers, and knocks down in turn, attempts to base the tort on a claimant’s ‘right to trade’, on some right of the claimant to complain about 27 I doubt it will come as a surprise to anyone to hear that the prominent recent rights-based theories of tort law that I have in mind are those promoted by Ernest Weinrib, Peter Benson, Alan Beever and, to a lesser extent, Robert Stevens. I regard Stevens as a less central target because he is willing to concede that the ‘rights’ protected by tort law can take any shape (‘as many and varied as the wit of human imagination’) provided that they are not created by judges (but by the legislature instead): Stevens, above n 19, at 337–8. 28 For the avoidance of doubt I should emphasise that I am not trying to prove that the particular list of rights that a theorist uses influences what that theorist considers is part of tort law—I take that to be obvious. Clearly a rights-theorist who thinks that people have a right not to be subjected to attention will defend a different law of torts to a rights-theorist who does not think that such a right exists. 29 J Neyers, ‘Rights-Based Justifications for the Tort of Unlawful Interference with Economic Relations’ (2008) 28 Legal Studies 215. 30 OBG Ltd v Allan [2008] 1 AC 1 (HL).

Tort Law, Concepts and What Really Matters

253

non-remote consequences of violations of a third party’s rights, or on a doctrine based around the defendant’s ‘abuse of rights’. The reason why Neyers does not consider a claim right identified by simply flipping the duty is not, of course, because of short-sightedness. Rather he regards it as obvious that a right ‘not to have others intentionally cause harm by using unlawful means to interfere with the liberty of third parties to deal with you’ cannot exist. Rights, properly so-called, cannot take such a form. His ‘design template’ for ‘rights’ will not allow ‘not to have others intentionally cause harm by using unlawful means to interfere with the liberty of third parties to deal with one’ to be recognised as a ‘right’. Why can ‘rights’ not take this form? Neyers does not, at least in this particular article, defend any particular ‘conceptual template’ for ‘rights’. But if we turn to Ernest Weinrib’s primer on corrective justice then he states: ‘As the juridical expressions of correlativity, the rights and duties themselves have to be conceived from a theoretical standpoint that is consistent with correlativity’.31 And as a result ‘considerations of distributive justice are inadmissible. Also inadmissible are considerations whose justificatory force extends only to one party because the fair terms of a bilateral interaction cannot be set on a unilateral basis’.32 How do these strictures preclude a ‘right’ mirroring the OBG duty? The answer seems to be that the claimant’s ‘right’ to a third party’s freedom cannot be founded on the defendant’s intention, since this is a unilateral matter.33 How might a theorist who is not committed to the same strictures explain the OBG tort? Perhaps he or she would recognise the importance of third parties’ freedom to human flourishing in a society where competitive markets are relied on as a major mechanism for allocation, but would note that the omnipresence and near invisibility of economic opportunities will make any duty to avoid damaging them potentially debilitating, and would believe that judges are more suited to identifying and categorising means than motives. Such a theorist might clearly believe that protecting a claimant’s interest in the freedom of third parties against those who intentionally seek to damage that interest and use unlawful means strikes an appropriate balance between facilitating a flourishing market and preserving valuable liberty. But, equally, such a theorist could not rule out the possibility that a better balance might broaden the definition of ‘unlawful means’ whilst insisting that the means must be the ‘instrumentality’ by which the defendant causes loss to the claimant.34 31 EJ Weinrib, ‘Corrective Justice in a Nutshell’ (2002) 52 University of Toronto Law Journal 350, 353. 32 ibid, at 355. 33 See, similarly, P Benson, ‘The Basis for Excluding Liability for Economic Loss in Tort Law’ in DG Owen (ed), Philosophical Foundations of Tort Law (Oxford, Clarendon Press, 1995) 435: ‘If a plaintiff lacks a proprietary or possessory right in something, he has no legal standing to constrain a defendant from intentionally using it as the defendant sees fit, even if this impairs or interferes with the plaintiff ’s interests’. 34 This possibility corresponds to the dissenting opinion of Lord Nicholls in OBG Ltd v Allan [2008] 1 AC 1 (HL) esp at [159] and has subsequently been commended by Lord Walker as useful in defining the appropriate level of protection for the wider range of economic interests that the tort of

254

Roderick Bagshaw

To recap, the first staging post in influencing the propensity of conceptual strictures about what can and cannot be a ‘right’ to impede tort law’s primary goal is to show that such strictures influence35 the views of ‘rights-based theorists’ as to the legitimate content of the law of torts. From this point, the next stage of my argument involves seeking to establish that the influence of these strictures on what is considered a legitimate part of tort law will not always be benign or neutral. Rather, such strictures have the propensity to impede progress towards tort law’s primary goal of making the world a better place. To some extent this part of the case depends upon confession. Those who promote ‘rights-based’ theories of tort law which rely on stringent specifications of what can properly be considered a ‘right’ usually confess that in doing so they intend to reject all consequentialist goals as the primary determinants of the legitimate content of tort law. Thus, if tort law, constrained by their conceptual strictures, succeeded in making the world a better place, this would be fortuitous rather than intentional. But, equally, some further argument is needed to demonstrate that such strictures are likely to impede the primary goal of making the world a better place. Perhaps some ‘rights theorists’ would claim that there is a fortuitous improvement of the world (sufficient to offset the costs) in using tort law to do no more than correct (to the extent that it does) the violations of the ‘rights’ which they recognise. Here I will mention two points to cast doubt on any such claim. First, as I have already argued, it is not realistic to suppose that people in a society such as that in England will have sufficient freedom to choose a pattern of life and flourish if private law limits itself to granting them some security for their persons, possessions and promises. Moreover protecting property ahead of less tangible economic interests tends to perpetuate inequalities with regard to the protection of welfare. Few would deny that the welfare of many people in England is far more closely tied to their future economic prospects than to their present property. Similarly,

conspiracy to use unlawful means protects: see Revenue & Customs Commrs v Total Network SL [2008] UKHL 19, [95]. 35 It may be the case that the word ‘influence’ makes an inappropriate suggestion about the direction of causation. In other words it is possible—perhaps more probable than not—that the views of ‘rightsbased theorists’ as to the legitimate content of the law of torts influence the strictures they apply to their concepts of ‘rights’. Equally, I do not intend to claim that the proponents of ‘rights-based approaches’ to tort law are unaware of the role that conceptual strictures play in their theories. My concern is rather that observers of ‘the debate as to the basic structure of torts’ may be unaware of the purposes which are encoded in the design of one of the key concepts which is deployed. For instance, the Textbook on Tort that I co-author with Nicholas McBride (3rd edn, Harlow, Pearson Longman, 2008) presents each of the torts as involving a breach of a duty owed to potential claimants. Since the algebra of duties and rights assumes that D owing a duty to C is equivalent to C having a right against D, an incautious reader might expect our book to be consistent with some of the recent rights-based visions of tort. But, that is not necessarily the case. The concept of ‘duty’ which we use is relatively non-constraining. If we were forced to explain our ‘strictures’ for determining which ‘verbal formulations’ we are willing to classify as ‘duties’ we might settle for something such as ‘intelligible normative instructions’. As we have seen, some recent rights-theorists use far more stringent ‘strictures’ for their ‘rights’, and would be inclined to exclude from the scope of tort law rules that our non-constraining approach to the concept of ‘duty’ includes.

Tort Law, Concepts and What Really Matters

255

the common insistence that purely economic interests can be legitimately protected in tort law only when an arrangement which mimics a contract has been made is likely to disadvantage, systematically, those whose welfare is more dependent on public services than private markets. The second point flows from the fact that most ‘rights-based’ theorists assume an all-or-nothing approach to protecting particular interests, in that they struggle to accommodate protection for only a portion of an interest or protection of an interest only in a particular context. By contrast, the development of tort law in circumstances where there are uncertainties about practical consequences and ‘best’ designs often requires a cautious approach so that a static snap-shot of the law of torts at any one moment might well reveal partially built structures. Obvious examples of such incomplete protection of interests might include the developing law of privacy and the partial and localised protection of autonomy. Such partially built structures, products of uncertainty and cautious experimentation, can be accommodated more easily by models of the law of torts which impose fewer strictures as to the form of their component parts.

IX. Possible Responses: Rights How might a rights-theorist who uses a restricted concept of ‘rights’ respond to my argument so far? One response might be to reiterate that without tight definition of key concepts such as ‘right’ the law of torts is likely to lack ‘coherence’. There certainly seems little doubt that a minimalist tort law structured around a small number of narrowly defined ‘rights’ is more likely to be ‘coherent’ (and complete) than a version of tort law that may be working towards some protection for a wider range of interests and against a wider range of behaviour. Moreover, few will doubt the importance that recent rights-theorists have attached to ‘coherence’. Ernest Weinrib, for instance, has written that ‘[a]ny sophisticated system of liability aspires to realize the values of fairness and coherence’,36 which does not seem synonymous with ‘aspires to leave the world a better place than it would be without such a sophisticated system of liability’. Should ‘coherence’ be treated as one of the principal goals of tort law? Is ‘coherence’ one of the reasons why we would not abolish tort law if there was no cost in doing so? No doubt ‘coherence’ may be an important measure when it comes to the comparison of rival theories. But it is more difficult to accept it as a primary goal of the product of an ongoing series of practical attempts to make the world

36 Weinrib, above n 31, at 356. The ‘aspirations’ or a ‘system of liability’ are presumably similar to the ‘goals’ of ‘tort law’.

256

Roderick Bagshaw

a better place.37 Do cities strive for coherence? Perhaps it might be thought that I am preaching against a straw man, and nobody really puts coherence ahead of beneficial consequences for human beings. But if so, I must be misunderstanding the purpose of, say Arthur Ripstein, when he says that a duty to attempt easy rescues, although morally attractive, ‘would sit uneasily with the rest of tort doctrine’38 and would be incompatible with his conception of tort law focused on ‘fair terms of interaction’ and ‘risk ownership’. I must also be misunderstanding the purpose of Allan Beever when he attacks White v Jones39 for having ‘stunted our understanding of, and at least arguably also the development of, the law as a whole.’40 And misunderstanding the purpose of Robert Stevens when he castigates Smith v Eric S Bush41 as ‘anomalous’ and proudly proclaims that neither Canadian National Railway Co v Norsk Pacific Steamship Co42 nor Perre v Apand Pty Ltd43 would ‘be followed in England’.44 Such arguments seem to go beyond neutral identification of rules which will reduce ‘the explanatory power’ of existing theories. Indeed, they are arguments that seem to suppose that preserving the explanatory power of a theory is a sufficient reason for disregarding the consequences of these decisions for the welfare of human beings. Certainly the authors have not established that these decisions have had a negative effect on the capacity of ordinary people to choose their own patterns of life and to flourish. A second response that such a ‘rights-based theorist’ might be tempted by would take the form of asserting that, even if my concerns are legitimate, I have no tolerable alternative account of ‘the basic structure of torts’ to offer. Certainly I have not presented a fully-specified alternative account in this chapter, but merely promoted a metaphor about cities and commended wider recognition of the utility of prototype-form concepts. But the second response invites the question how it would be determined whether a rival account was ‘tolerable’.

37 It will be recalled that I am treating the ‘goals of tort law’ as the reasons why we should tolerate the continuing existence of tort law—on the assumption that there would be no transition cost in eliminating it or substantially altering it. Clearly an argument that the primary reason for tolerating tort law is that it is coherent would be unexpected. Moreover, given that contemporary tort law is the product of a lengthy series of developments by different law-makers—and that at each subsequent stage in the series there would have been the potential for a transition cost in eliminating or substantially altering any previous stage—the likelihood of the final product being coherent must be low. 38 A Ripstein, Equality, Responsibility, and the Law (Cambridge, Cambridge University Press, 1999) 92. 39 White v Jones [1995] 2 AC 207 (HL) (holding that a disappointed beneficiary can sue a solicitor for negligently failing to prepare a will reflecting the testator’s instructions). 40 A Beever, Rediscovering the Law of Negligence (Oxford, Hart Publishing, 2007) 269. 41 Smith v Eric S Bush [1991] 1 AC 831 (HL) (holding that a purchaser of a house which proves to be defective may be able to sue a surveyor who did not identify that defect even though the surveyor was valuing the house for a potential lender). 42 Canadian National Railway Co v Norsk Pacific Steamship Co [1992] 1 SCR 1021 (holding a defendant that damaged a railway bridge liable for the losses of the primary user of the bridge). 43 Perre v Apand Pty Ltd (1999) 198 CLR 180 (High Court of Australia) (holding a defendant which was responsible for the introduction of diseased potatoes into a locality responsible for the losses of nearby potato producers even though their crops, machinery and land were not physically damaged). 44 Stevens, above n 19, at 40–43.

Tort Law, Concepts and What Really Matters

257

There seems little doubt that the approaches which insist on all directly protected interests conforming to a strict template of ‘right’ are usually presented as a preferable alternative to rival ‘policy-based approaches’.45 One prominent criticism of ‘policy-based approaches’ is that they require judges to take on constitutionally inappropriate tasks. A second objection is that they require impossible computations with unreliable data, and consequently carry an unacceptably high risk of inconsistency and unpredictability. By contrast, approaches based on a limited number of tightly defined ‘rights’ require very few computations (because they rely on so few ‘concepts’) and the principal source of uncertainty is in the application of these general ‘concepts’ to the detailed facts of real cases. No doubt the concerns behind these criticisms are substantial, but it seems an overreaction to respond by insisting on reversion to a minimalist law of tort chiefly celebrated for the ease with which it can be explained. Such an overreaction misses the fact that there are legal techniques for managing and controlling such concerns, for seeking to steer and control innovation; rules and techniques which aim to ensure that judges only revise or innovate when to do so is likely to lead to tort law which is more suitable for the primary goal of making the world a better place. It is the effectiveness, or lack of it, of these techniques which will determine whether a ‘tolerable’ alternative account can be constructed. I would suggest that one of these techniques is the gradual development of key concepts in the law by way of identification of prototypes and anti-prototypes and assessments of similarity, and that the modest scale of any uncertainty caused by Smith v Eric S Bush,46 White v Jones,47 Canadian National Railway Co v Norsk Pacific Steamship Co48 and Perre v Apand Pty Ltd49 tends to confirm the capacity of this technique to moderate such concerns.

X. Scepticism About Concepts At the start of this chapter I suggested that it was unlikely that any attractive version of the substantive law of torts could relinquish the use of concepts because of their value in allowing legal rules to attain a degree of generality useful for efficient communication and application. But the expression of relatively general rules which incorporate concepts, either in the form of checklists or prototypes, is 45 ‘Policy’ is, of course, not a ‘concept’ in any significant sense, but rather a label for all other concerns. Some courts, notably those in Canada and Singapore (see eg: Cooper v Hobart [2001] 3 SCR 537 at [28] and Ngiam Kong Seng v Lim Chiew Hock [2008] 3 SLR 674 (Singapore Court of Appeal) [41]–[45]), have taken steps towards ‘conceptualising’ policy, by formulating rules which direct judges to use different types of policy in different ways. But there is scope to debate both the success of this ‘conceptualisation’ and the utility of these rules. 46 Smith v Eric S Bush [1991] 1 AC 831 (HL). 47 White v Jones [1995] 2 AC 207 (HL). 48 Canadian National Railway Co v Norsk Pacific Steamship Co [1992] 1 SCR 1021. 49 Perre v Apand Pty Ltd (1999) 198 CLR 180 (High Court of Australia).

258

Roderick Bagshaw

merely one way of making legal outcomes sufficiently predictable. Outcomes will be sufficiently predictable whenever judges reach them by techniques which can be reliably replicated outside the court-room. As is well known, Jane Stapleton has consistently advocated a very different technique, which is exemplified by her preferred approach to identifying whether a duty of care exists in a novel situation.50 This technique involves transparent judgement as to whether ‘taken together and on balance, the pro-liability concerns raised by the facts outweigh factors countervailing to liability’—what might be called ‘the open balancing of competing policy concerns’. The technique’s claim to be reliably replicable outside the court-room depends on: whether judges succeed in articulating an agenda of reasons for and against liability; the effectiveness of the instruction to judges to be cautious before departing more than incrementally from the weight given to these reasons by judges in previous cases where the reasons were relevant; and whether those who need to know the law in advance of judicial resolution will have sufficient access to the information needed to predict how judges are likely to ‘weigh’ these competing concerns.51 Stapleton has coupled promotion of her preferred technique with profound scepticism towards rival approaches based on fact-based categories and opaque assessments of similarity. This scepticism might well stretch sufficiently far to suggest that the law would be better off without prototype-form concepts, not because law should be static and judges forbidden from assessing questions of policy, but because there is little need to seek to control and manage judicial policy choices. I will not seek to cure such scepticism here, but will simply balance it with counter-scepticism as to whether the open balancing of policy concerns can claim to be a reliably replicable technique in a world where most people, including scholars, lawyers and judges, have so little access to information about the consequences of different possible legal interventions.

50 She also commends the open technique in making normative judgments about the scope of liability for consequences, see eg: J Stapleton, ‘Cause-in-Fact and the Scope of Liability for Consequences’ (2003) 119 Law Quarterly Review 388, 412. 51 The judge should ‘identify … which pro-liability and countervailing concerns are raised by the particular fact situation’, should then ‘consider previous case law, if any, in which these concerns were raised and the weight given them by previous courts (being cautious about departing more than incrementally from such earlier weightings)’, and finally the judge should ‘determine what weight, in this judge’s own view, it would be fair, just and reasonable to give to the relevant factors and therefore, on balancing them together, whether a duty of care should be recognized’: J Stapleton, ‘Duty of Care Factors: A Selection from the Judicial Menus’ in P Cane and J Stapleton (eds), The Law of Obligations: Essays in Celebration of John Fleming (Oxford, Clarendon Press, 1998) 91. See also, J Stapleton, ‘Duty of Care: Peripheral Parties and Alternative Opportunities for Deterrence (1995) 111 Law Quarterly Review 301, 301: ‘If the courts desire convincingly to contain liability in the tort of negligence, they must determine and explicitly articulate an agenda of coherent reasons which are “countervailing” to the imposition of liability’. The meaning of the word ‘coherent’ in this proposition may require discussion.

Tort Law, Concepts and What Really Matters

259

XI. Summary and Conclusions In this chapter I have sought to demonstrate that different forms of concepts can be incorporated into legal rules, and that choices between forms may alter tort law’s capacity to meet its primary goal of making the world a better place. I have suggested that some of those who have sought to identify an immediate intelligible order in tort law, treated as a static and complete system, have tended to overlook the valuable role that dynamic concepts might play in a version of tort law treated as constantly developing in the light of new information about the world and the effects of our interventions in it. Moreover, I have argued that prototype concepts developed through assessments of similarity may provide a satisfactory technique for controlling and managing some of the constitutional and practical concerns which seem to have stimulated some accounts of tort law which have particularly restricted the form of key concepts. Underlying all my arguments is a clear preference for a law of torts which can do good, ahead of a law of torts which can be straightforwardly explained.

11 Constraints on Policy-Based Reasoning in Private Law ANDREW ROBERTSON*

I. Introduction The legitimacy of community welfare or policy considerations in deciding cases and shaping private law rules has recently come under sustained attack from scholars advocating a strict corrective justice or rights-based approach.1 A significant element in that attack is the idea that judges who take account of policy considerations in private law decision-making exceed their judicial role and improperly act as legislators.2 The proper role of community welfare considerations in deciding cases and shaping private law rules lies at the heart of questions about the nature of private law. The question whether community welfare considerations have a legitimate role to play in private law underlies the great divide between formalism and instrumentalism in private law scholarship.3 Instrumentalists treat private law as a tool of public policy and focus their attention on the policy implications of private law rules. For the corrective justice purist, on the other hand, private law can have no goals, and policy considerations can have no legitimate role to play in private law adjudication or the development of private law rules. * This chapter represents preliminary work on a project on policy-based reasoning in private law supported by the Australian Research Council’s Discovery Projects funding scheme. I am grateful to participants at the Obligations IV conference as well as Arlen Duke, Anne Orford, Megan Richardson and Christian Witting for helpful suggestions made in discussions of this chapter. I also thank Harold Luntz and Donal Nolan for their comments on a draft. The usual caveat applies. 1 EJ Weinrib, The Idea of Private Law (Cambridge MA, Harvard University Press, 1995); EJ Weinrib, ‘The Disintegration of Duty’ in Madden (ed), Exploring Tort Law (New York, Cambridge University Press, 2005) 143; EJ Weinrib, ‘Does Tort Law Have a Future?’ (2000) 34 Valparaiso University Law Review 561; A Beever, Rediscovering the Law of Negligence (Oxford, Hart Publishing, 2007); R Stevens, Torts and Rights (Oxford, Oxford University Press, 2007). 2 Stevens, above n 1, at 312–14; Weinrib, The Idea of Private Law, above n 1, at 7; and Beever, above n 1, at 170–81. 3 H Dagan, ‘The Limited Autonomy of Private Law’ (2008) 56 American Journal of Comparative Law 809, 809–11 identifies ‘the competition between autonomy and instrumentalism’ and the ‘competition between two accounts of the relationship between private law and social values’ as a perennial debate, which ‘has become particularly acute in recent years’.

262

Andrew Robertson

What the instrumentalist and formalist accounts of private law have in common is the notion that, if policy considerations guide decision-making in private law, they do so in a way that is entirely unconstrained. Both treat the ‘unruly horse’ of public policy as unbridled. The judge who takes account of policy considerations (rightly under the instrumentalist account and wrongly from the formalist point of view) acts as legislator and is completely free to choose between different legislative solutions to a particular problem. The instrumentalist account is premised on an unfettered use of policy, while the formalist account rejects it. A great deal of economic analysis of legal problems proceeds as though the courts were completely unconstrained in their capacity to pursue the policy goal of wealth maximisation. This is consistent with the stark dichotomy drawn by Ernest Weinrib between a private law that is overrun by political choices and one that is entirely above politics and in which community interests have no legitimate role to play.4 Robert Stevens also offers an all-or-nothing choice between a law of torts that is concerned only with the infringement of primary rights and one in which ‘judges take on the role of legislators’ and ‘decide cases based upon an assessment of policy considerations’.5 This chapter represents the beginnings of an attempt to develop an account of the constraints on the use of policy in private law. In doing so, it explores the nature of the via media between unconstrained instrumentalism and the policyfree ‘anti-instrumentalism’ of corrective justice and rights-based theories.6 The chapter provides support for the view that considerations of community welfare have a legitimate role to play in private law decision-making and in shaping private law rules because they are constrained by the framework of judicial decisionmaking in general, and private law decision-making in particular. The next section of this chapter sets up a working definition of policy and identifies the different roles played by policy considerations in private law. The following section of the chapter outlines the principal constraints on the use of policy in private law.

II. The Nature of Policy It is necessary to begin by developing a useful working definition of policy. The ‘policy considerations’ that are the subject of this chapter are justifications (for or against a legal rule or outcome in a particular case) that are concerned with community interests not related to the form of law. That definition must be qualified in four ways, each of which will be explained below. First, policy considerations may be either consequentialist or deontological. Secondly, it is

4

Weinrib, The Idea of Private Law, above n 1, esp at 7–8. Stevens, above n 1, at 312–14. 6 Dagan, above n 3, at 810 describes the ‘neo-Kantian formalism’ of corrective justice as ‘the ultimate anti-instrumental understanding’ of private law. 5

Duty ofConstraints Care in A Rights-Based Theory of Negligence on Policy-Based Reasoning

263

necessary to exclude from discussions of ‘policy’ the formal legal considerations of certainty, consistency and coherence, even though the pursuit of these values might be said to serve community interests. Thirdly, policy considerations relating to the interests of the community must be distinguished from considerations of justice or fairness to and between the parties to a particular dispute. Fourthly, it is important to note that policy considerations do not equate to justifications for legal rules: they do not always operate as justifications for rules and nor do they constitute the only form of justification for rules.

A. Consequentialist and Deontological Considerations The first point to note is that there are two types of community-based justifications for a legal rule that are commonly described as ‘policy’ considerations. Most policy arguments are consequentialist: they justify or oppose a legal rule or outcome because of the consequences that a rule or outcome of that type is expected to produce. Economic analysis of law is the most common type of consequentialist policy analysis in the legal literature: it is concerned with the effect of legal rules on people’s behaviour. Some policy arguments are less concerned with the consequences of legal rules or outcomes than with the idea that a particular legal rule or outcome would be right or wrong in itself. An example is the ‘value of human life’ argument in the wrongful birth cases, which will be discussed below. Deontological policy considerations are moral arguments, which, according to Weinrib’s analysis, must be considered external to private law because they do not reflect the unity of the private law relationship.7 Weinrib claims that private law is not in the service of any external purpose, even the purpose of supporting non-instrumental moral obligations. Deontological policy considerations are external to private law on Weinrib’s view because they are not relational; even if they identify a moral obligation, they involve moral reasoning that does not justify the existence of a correlative right.8 At an even broader level, deontological policy considerations such as the ‘value of human life’ argument are not concerned with the parties’ moral responsibilities to one another. They are simply concerned with the moral rightness of a legal rule. It is not possible to draw a sharp distinction between consequentialist and deontological policy considerations. Many policy considerations which appear to be concerned with intrinsic notions of right and wrong can also be justified on the basis of a concern to produce socially desirable outcomes. The ‘value of human life’ and ‘blessing’ considerations in the wrongful birth cases are sometimes said to be founded on notions of right and wrong. In McFarlane v Tayside Health 7 Weinrib, The Idea of Private Law, above n 1, at 50. The distinction implicitly drawn by Weinrib between a non-instrumental account of morality and a non-instrumental account of private law is discussed by S Perry, ‘The Role of Duty of Care in a Rights-Based Theory of Negligence Law’ ch 4 of this book. 8 See Weinrib, The Idea of Private Law, above n 1, at 48–52.

264

Andrew Robertson

Board, for example, Lord Millett supported the denial of damages for the costs of raising a healthy child by reference to the assertion that: ‘[i]t is morally offensive to regard a normal, healthy baby as more trouble and expense than it is worth’.9 In Cattenach v Melchior, however, McHugh and Gummow JJ put this as a question of community welfare. They asked whether ‘the underlying values respecting the importance of human life, the stability of the family unit and the nurture of infant children until their legal majority are an essential aspect of the corporate welfare of the community’.10 Similarly, the notion that a plaintiff ’s participation in a joint criminal enterprise with the defendant might prevent a duty of care arising has been justified on the basis that ‘the civil law cannot condone breaches of the criminal law’.11 This justification seems to be based on the deontological notion that it would be wrong for the civil law to condone criminal conduct. Denial of a duty of care in these circumstances could, however, equally be justified on the consequentialist basis that the civil law should discourage criminal conduct. Moreover, at a deeper philosophical level, consequentialist and deontological policy considerations collapse into one another. It is impossible to distinguish sharply between deontological and consequential policy considerations because an enhancement of community welfare or social utility (based on the consequences of a rule) can itself ground a claim that a particular rule is morally ‘right’. In other words, deontological claims can be made on the basis of consequentialist arguments. Although consequentialist and deontological considerations are thus entangled, it is useful to note the two different strands of thinking in the idea of policy.

B. Considerations of Form Secondly, it is necessary to distinguish broader social or economic policy considerations from those that relate to the form of law itself, ie, the need for certainty, coherence, consistency with other legal rules and adherence to precedent. One could observe that these formal legal considerations serve broader social ends and should therefore be treated as one type of policy consideration, amongst others. Although we can properly describe formal values (certainty, consistency and coherence) and other community interests collectively as ‘legal concerns’,12 it is necessary to distinguish between these different types of ‘legal concerns’ in order to understand the relationship between them. We need to distinguish formal legal considerations from broader social and economic policy because, as we will see, considerations of form operate as a constraint on the use of broader social or economic policy justifications. Judges can take account of broader social and economic considerations only to the extent that they can be reconciled with 9

McFarlane v Tayside Health Board [2000] 2 AC 59 (HL) 114. Cattenach v Melchior (2003) 215 CLR 1 (High Court of Australia) 35. Gala v Preston (1991) 172 CLR 243 (High Court of Australia). 12 J Stapleton, ‘The Golden Thread at the Heart of Tort Law: Protection of the Vulnerable’ (2003) 24 Australian Bar Review 135, 137. 10 11

Duty ofConstraints Care in A Rights-Based Theory of Negligence on Policy-Based Reasoning

265

the need to maintain certainty, coherence and consistency. Thus, although formal legal virtues can be seen as a type of policy argument, formal legal values operate to constrain the operation of non-formal policy arguments.

C. Considerations of Fairness and Justice Between the Parties Thirdly, it is necessary to distinguish considerations of justice between the parties to a particular dispute from those that relate to broader community interests. The ‘policy’ label is sometimes given to broad considerations of fairness or justice as between the parties before the court, such as the need to ensure a fair allocation of risks and responsibilities. In Kienzle v Stringer,13 for example, Zuber JA observed that ‘policy’ is relevant to a consideration of what is ‘reasonably foreseeable’ or ‘within the reasonable contemplation of the parties’ in the application of the remoteness rule in tort and contract. He said: The governing term is reasonable and what is reasonably foreseeable or reasonably contemplated is a matter to be determined by a court. These terms necessarily include more policy than fact as courts attempt to find some fair measure of compensation to be paid to those who suffer damages by those who cause them.14

It is clear that by ‘policy’ Zuber JA meant considerations of fairness and justice between the parties. Policy considerations, in the sense of broader community interests, are rarely even mentioned in remoteness cases in contract.15 In the causation context, community interests have also sometimes been conflated with considerations of justice between the parties under the heading of ‘policy’. In March v E & MH Stramare Pty Ltd, McHugh J observed that ‘the use of commonsense notions of causation’ and expressions such as ‘real cause’ or ‘proximate cause’ ‘require the application of a policy choice, for they allow the tribunal of fact to determine legal liability on broad grounds of moral responsibility for the damage which has occurred’.16 The notion of ‘policy’ here seems to incorporate both the defendant’s ‘moral responsibility’ for the damage in question and the community interest in limiting ‘legal responsibility for damage resulting from acts or omissions of the kind in question’.17 Considerations of policy were also brought

13

Kienzle v Stringer (1981) 35 OR (2d) 85 (Ontario CA) 90. ibid, at 90, citing K Swinton, ‘Foreseeability: Where Should the Award of Contract Damages Cease’ in BJ Reiter and J Swan (eds), Studies in Contract Law (Toronto, Butterworths, 1980) 61. In Kienzle v Stringer (1981) 35 OR (2d) 85 (Ontario CA) 92, Wilson JA also referred to the ‘public policy component inherent in the remoteness concept’. 15 See A Robertson, ‘The Basis of the Remoteness Rule in Contract’ (2008) 28 Legal Studies 172, 188–95 (explaining the role of considerations of fairness between the parties). An exceptional case in which broader community welfare considerations were discussed (but not ultimately taken into account) is Unity Insurance Brokers Pty Ltd v Rocco Pezzano Pty Ltd (1998) 192 CLR 603 (High Court of Australia). 16 March v E&MH Stramare Pty Ltd (1991) 171 CLR 506 (High Court of Australia) 531. 17 ibid, at 532. 14

266

Andrew Robertson

to bear on the causation issue in Fairchild v Glenhaven Funeral Services Ltd, where Lord Bingham said: It can properly be said to be unjust to impose liability on a party who has not been shown, even on a balance of probabilities, to have caused the damage complained of. On the other hand, there is a strong policy argument in favour of compensating those who have suffered grave harm, at the expense of their employers who owed them a duty to protect them against that very harm and failed to do so, when the harm can only have been caused by breach of that duty and when science does not permit the victim accurately to attribute, as between several employers, the precise responsibility for the harm he has suffered. I am of opinion that such injustice as may be involved in imposing liability on a duty-breaking employer in these circumstances is heavily outweighed by the injustice of denying redress to a victim.18

Although, as Donal Nolan observes,19 there were hints of deterrence in the reasoning in Fairchild, Lord Bingham seems to be using the word policy here to mean fairness and justice between the parties. The ‘strong policy argument’ seemed to be that it was fair and just, as between the parties, for the employer to be held liable to compensate the employee for the harm in question. As with formal legal values, it is necessary to distinguish the concern to do justice to each of the parties from broader social or economic reasons for or against a particular legal rule or outcome. As will be explained below, the need to do justice to the parties operates as a constraint on the circumstances and the ways in which the courts can take account of broader community interests in shaping and applying common law rules. For considerations of justice to the parties to operate as a constraint on the use of policy considerations requires a notion of justice that can be distinguished from community interests. If ‘justice to each of the parties’ for this purpose means corrective justice as articulated by Weinrib, and the courts can only give effect to community interests that are consistent with corrective justice, then community goals are not really being pursued at all. The governing criterion is then corrective justice, which just happens in particular instances to accord with broader community interests. The pursuit of policy goals can therefore be meaningfully constrained by the need to do justice to the parties only if this notion of justice is something that is distinguishable from community interests, but is broader and more flexible than corrective justice. As we will see, the need to do justice to each of the parties does operate as a constraint in private law decision-making, but operates more strongly in favour of the defendant than it does in favour of the plaintiff. In other words, the courts are more concerned to ensure that the defendant is not unjustly burdened by liability than to ensure that the plaintiff is not unjustly denied enforceable rights. The key constraining factor is that liability will not be imposed on a defendant on the basis

18 19

Fairchild v Glenhaven Funeral Services Ltd [2003] 1 AC 32 (HL) 67. D Nolan, ‘Causation and the Goals of Tort Law’ ch 7 of this book.

Duty ofConstraints Care in A Rights-Based Theory of Negligence on Policy-Based Reasoning

267

that it is in the community interest to do so unless that liability is justifiable by reference to the defendant’s responsibility to the plaintiff.

D. The Roles Played by Policy Considerations One view of the proper role of policy is that legal principles are applied to decide cases, while policy considerations can only be used to inform the development of legal principles.20 The direct application of policy considerations is said to be more uncertain than the application of rules derived from the policy considerations. In Caltex Oil (Australia) Pty Ltd v The Dredge ‘Willemstad’, for example, Stephen J suggested that [t]o apply generalized policy considerations directly, in each case, instead of formulating principles from policy and applying those principles, derived from policy, to the case in hand, is, in my view, to invite uncertainty and judicial diversity.21

On this view, policy considerations should only be used to assist judges in determining what the relevant rules ought or ought not to be.22 This view is consistent with Emily Sherwin’s analysis of the relationship between rules and rationales. Sherwin argues that rationales can properly shape the formulation of rules, but ‘once a rule is in place their work is done’.23 In the cases, however, the role of policy considerations is not confined to informing the development of legal principles. Policy considerations influence private law decision-making in a number of different ways. First, as already noted, policy considerations influence the development of legal rules. Community welfare considerations therefore underlie and are embedded in many private law doctrines, even though relevant cases can be and are decided without any reference to those underlying rationales. Secondly, some legal rules, such as the restraint of trade doctrine in contract, require the direct consideration of community welfare in the determination of each case that falls to be decided under the rule. Thirdly, policy considerations can be taken into account in the application of broad, open-textured standards that form part of a particular legal rule.24 In the application of these standards, policy considerations may be taken into account

20 A Mason, ‘Policy Considerations’ in Blackshield et al (eds), The Oxford Companion to the High Court of Australia (Melbourne, Oxford University Press, 2001) 535. 21 Caltex Oil (Australia) Pty Ltd v The Dredge ‘Willemstad’ (1976) 136 CLR 529 (High Court of Australia) 567. 22 P Cane, ‘The Anatomy of Private Law Theory: A 25th Anniversary Essay’ (2005) 25 Oxford Journal of Legal Studies 203. 23 E Sherwin, ‘The Rules of Obligations’ ch 18 of this book. cf Frederick Schauer, ‘Formalism’ (1988) 97 Yale Law Journal 509, 534, discussed below n 24. 24 This is consistent with the view that the integrity of a rule as a rule is not undermined if resort is had to its underlying rationale in order to choose between multiple applications of the rule, all of which are consistent with its language. Schauer, ibid, at 534, argues that the distinction between a rule and a reason collapses, and the rule is undermined, only when the purpose of a rule is used to trump its language.

268

Andrew Robertson

but need not be. An example of this is the ‘fair, just and reasonable’ part of the Caparo Industries plc v Dickman25 test for identifying the existence of a duty of care. A fourth method of utilising policy is to formulate a legal rule by reference to a particular policy rationale, and then direct judges to apply the rule in light of the relevant policy goals. This approach is exemplified by Bazley v Curry, where the Supreme Court of Canada instructed trial judges to apply the relevant test ‘with a sensitive view to the policy considerations underlying it’: In summary, the test for vicarious liability for an employee’s sexual abuse of a client should focus on whether the employer’s enterprise and empowerment of the employee materially increased the risk of the sexual assault and hence the harm. The test must not be applied mechanically, but with a sensitive view to the policy considerations that justify the imposition of vicarious liability—fair and efficient compensation for wrong and deterrence. This requires trial judges to investigate the employee’s specific duties and determine whether they gave rise to special opportunities for wrongdoing.26

III. Constraints on the Use of Policy A. Institutional Constraints The methodological constraints on the use of policy to be discussed below need to be considered against the background of the broader institutional constraints on judges, which have been thoroughly explored in the legal literature. It is well accepted that judges are significantly constrained by the need to explain their reasons for judgment.27 Judges address their reasons to the parties and also, by stating them publicly, to other judges and lawyers, the legal academy and the broader community.28 The prospect of peer scrutiny and a desire to maintain one’s reputation both within and outside the profession significantly limit the extent to which judges can rely on unorthodox justifications.29 This constraint is reinforced by the fact that judges in appellate courts can only develop legal principles by acting collectively. As Windeyer J observed in Mount Isa Mines Ltd v Pusey, judges do not determine issues of policy based on an individual ‘view of what social interests dictate’.30 They must act ‘in company and not alone’ and 25

Caparo Industries plc v Dickman [1990] 2 AC 605 (HL). Bazley v Curry [1999] 2 SCR 534 [46]. 27 Eg, M Kirby, ‘Judicial Activism? A Riposte to the Counter-Reformation’ (2004) 24 Australian Bar Review 219. 28 M Eisenberg, The Nature of the Common Law (Cambridge MA, Harvard University Press, 1988). 29 See, eg, RA Posner, How Judges Think (Cambridge MA, Harvard University Press, 2008) 140–57; EW Thomas, The Judicial Process (Cambridge, Cambridge University Press, 2005) ch 10; I Englard, ‘Law and Economics in American Tort Cases: A Critical Assessment of the Theory’s Impact on Courts’ (1991) 41 University of Toronto Law Journal 359, 366–9. 30 Mount Isa Mines Ltd v Pusey (1970) 125 CLR 383 (High Court of Australia) 396. 26

Duty ofConstraints Care in A Rights-Based Theory of Negligence on Policy-Based Reasoning

269

are guided by ‘[a]nalogies in other courts, and persuasive precedents as well as authoritative pronouncements’.31 The legal profession and the legal academy also provide a significant institutional constraint by policing (in textbooks and scholarly literature) consistency, coherence and doctrinal stability, and by scrutinising and criticising assumptions made by judges about the potential social and economic consequences of particular legal rules.32

B. Common Law Method and Convention The institutional constraints outlined above reinforce the doctrinal stability that is produced by the conventional techniques of common law reasoning. The doctrine of precedent is the strongest formal constraint on common law courts, but the practical strictures it imposes are quite limited. There are three principal reasons for this. First, although it is not without controversy,33 there is strong support for the view that the doctrine of precedent requires courts to follow only the outcomes of previous cases, and not the reasoning of the judges deciding those cases.34 Insofar as it requires only consistency of outcomes, the doctrine of precedent operates as a weak constraint on the use that may be made of policy considerations in deciding cases and developing legal rules. On the precedentas-outcomes view, articulated rules are a by-product of the common law system, rather than the productive machinery.35 The requirement that like cases be treated alike does not significantly constrain the kinds of considerations that can be taken into account or the ways in which they can be utilised.36 Secondly, even courts at first instance routinely deal with issues that they are not strictly bound, by the principle of stare decisis, to resolve one way or another. Thirdly, ultimate appellate courts that have the power to overrule their own decisions are not of course in any strict sense bound by the doctrine of stare decisis. 31

ibid. See Eisenberg, above n 28. 33 See, eg, R Cross and JW Harris, Precedent in English Law, 4th edn (Oxford, Clarendon Press, 1991), esp ch 2; Victoria v Commonwealth (1995) 187 CLR 416 (High Court of Australia) 484–5 (Brennan CJ, Toohey, Gaudron, McHugh and Gummow JJ): ‘It is to seek to distort the principles of stare decisis and of ratio decidendi to contend that a decision lacks authority because it might have been reached upon a different path of legal reasoning to that which was actually followed. That would be to replace what was decided by that which might have been decided’. 34 See, eg, N Duxbury, The Nature and Authority of Precedent (Cambridge, Cambridge University Press, 2008) ch 3; J Kraus, ‘From Langdell to Law and Economics: Two Conceptions of Stare Decisis in Contract Law and Theory’ (2008) 94 Virginia Law Review 157. Even the rule-based theory of precedent defended by L Alexander, ‘Constrained by Precedent’ (1989) 63 Southern California Law Review 1, does not require adherence to the rule announced by the precedent court: at 18–19. 35 S Perry, ‘Judicial Obligation, Precedent and the Common Law’ (1987) 7 Oxford Journal of Legal Studies 215, 252: ‘The body of settled common law is, therefore, not to be regarded as the productive machinery of the common law process, which is how it is conceived on the positivist model, but as a by-product; it is never more than a weighted but nevertheless provisional approximation to those requirements of morality the determination and application of which constitute the ultimate obligation of the common law judge’. 36 K Greenawalt, ‘Policy, Rights and Judicial Discretion’ (1977) 11 Georgia Law Review 991. 32

270

Andrew Robertson

The real constraint on the use of policy considerations in judicial reasoning does not arise from what judges are strictly bound to do, but from convention, adherence to which is required by the institutional factors outlined above. As a matter of strongly-established convention the courts must, in resolving disputes, give priority to values that have found expression in legal documentary form.37 Peter Cane has suggested that this requirement is the foundation of the common law’s stability and conservatism and is essential to the distinctiveness of legal reasoning, which prevents it from becoming ‘all things considered reasoning’.38 This constraint leads courts to ground their decisions in the application or adaptation of existing (documented) principles or rules, rather than the direct application of (undocumented) policy considerations.39 Where policy considerations are employed, the priority given to legal documentary forms ensures some consistency, through time and across jurisdictions, in the types of policy considerations that are taken into account, and the circumstances in which they are employed. As a matter of judicial practice, the current orthodoxy in Commonwealth jurisdictions lies much closer to the precedents-as-reasons than the precedents-asoutcomes position. Legal reasoning begins with, and gives priority to, ‘announced rules’.40 In the quotidian practice of judicial reasoning, announced rules have a very strong influence, whether or not they form part of the ratio decidendi of the decisions in which they have been articulated. Judges routinely follow rules laid down in earlier decisions without stopping to ask whether they are bound to do so. Even in matters of controversy, judges are often more concerned to identify and apply the rules that have been articulated by an ultimate appellate court than to consider whether they are strictly bound to follow those rules. The weight that is given to rules announced or laid down in previous decisions therefore operates as a powerful stabilising influence, despite the weakness of the formal constraint imposed by the principle of stare decisis. Even at the ultimate appellate level, the power to depart from previous lines of authority is exercised sparingly. The courts routinely express a concern to rationalise or explain existing case law, rather than to change the law in a way that would be inconsistent with previous decisions.41 Adherence to these conventions means that when the courts do take account of policy considerations in the determination of cases and the development of legal rules, this does not involve the exercise of anything like a simple legislative choice between different policy options.

37 P Cane, ‘Taking Disagreement Seriously: Courts, Legislatures and the Reform of Tort Law (2005) 25 Oxford Journal of Legal Studies 393, 407–8. 38 ibid, at 408. 39 Thus, in Benning v Wong (1969) 122 CLR 249 (High Court of Australia) 305, Windeyer J observed that ‘it is as sound a maxim for law as for war that operations should be from a firm base, that an advance must be from a position which has been securely established’. 40 Eisenberg, above n 28, at 63. 41 See, eg, Bazley v Curry [1999] 2 SCR 534 [15], where McLachlin J (for the Supreme Court of Canada) said that her aim was to ‘articulate a rule consistent with both the existing cases and the policy reasons for vicarious liability’.

Duty ofConstraints Care in A Rights-Based Theory of Negligence on Policy-Based Reasoning

271

C. Consistency and Coherence Related to the doctrine of precedent is the notion that the courts are constrained by the formal legal goal of ensuring consistency between related principles and related bodies of law. There is also a strong value attached to ensuring that the common law remains consistent over time, which is manifested in the judges’ concern with maintaining doctrinal stability.42 Kirby J observed in Cattanach v Melchior that judges are bound to refine the law and apply it to new circumstances ‘in ways that are logically reasoned and shown to be a consistent development of past decisional law’ and ‘have no authority to adopt arbitrary departures from basic doctrine’.43 Consistency over time is particularly highly valued in private law, and is therefore a particularly strong constraining influence, because of the perceived need for certainty in the rules governing commercial activity. Whether private law can accommodate policy considerations at all depends on the nature of the coherence that is required or, more precisely, what particular features of private law are required to cohere with each other. Weinrib argues that the need for coherence precludes policy-based reasoning altogether.44 This is based on a particular notion of coherence. Weinrib’s stipulation that the justifications for a defendant’s liability must correlate with the justifications for the plaintiff ’s right is founded on his insistence on coherence between the features of private law relationships. For Weinrib, coherence is the ‘intrinsic unity of the features that cohere’45 and this requires that ‘the consideration that justifies any feature of the relationship must cohere with the considerations that justify every other feature of it’.46 Weinrib’s notion of coherence requires symmetrical justifications for the plaintiff ’s right and the defendant’s liability. Since policy considerations are always one-sided, the use of policy considerations always produces incoherence in the sense of asymmetry between the justifications underlying the plaintiff ’s right on the one hand and the defendant’s obligation or liability on the other. Most judges and private law scholars do not demand coherence between the considerations justifying the features of private law relationships, but insist only on a much broader notion of coherence within categories of obligation and between different private law doctrines. In McFarlane v Tayside Health Board, for example, Lord Millett observed that, in taking account of policy considerations in assessing a novel head of damages, the court is ‘concerned to maintain the coherence of the law’.47 This seemed to mean something similar to Neil MacCormick’s understanding of coherence as the notion that ‘the multitudinous rules of a

42

Eisenberg, above n 28. Cattanach v Melchior (2003) 215 CLR 1 (High Court of Australia) 53. 44 Eg, Weinrib, The Idea of Private Law, above n 1, esp at 8–14, 48–52, 72–5, 210–14 and ‘The Disintegration of Duty’, above n 1. 45 Weinrib, The Idea of Private Law, above n 1, at 35. 46 ibid, at 32. 47 McFarlane v Tayside Health Board [2000] 2 AC 59 (HL) 108. 43

272

Andrew Robertson

developed legal system should “make sense” when taken together’.48 A significant way in which legal rules can cohere with each other is through the promotion of common justifications. Coherence here means unity of the justifications and rationales that underlie legal rules.49 Coherence between related rules requires unity between the principles and policies that are embedded in those rules.50 A set of legal rules is coherent in this sense if the rules are united by a common set of justifications. Normative coherence with other rules in the given set does not, on its own, justify the adoption of a particular legal rule, but serves only to establish a kind of baseline eligibility for the adoption of the rule.51 On this view, one way in which legal rules can cohere with each other is through the promotion of common policies or through the utilisation of policy considerations in ways that are common to the different sets of rules.

D. Bipolarity and the Need for Justice to Both Parties The fourth and final constraint on the use of community welfare considerations is the need in each case to do justice to the particular parties before the court. Correlativity between the plaintiff ’s right and the defendant’s obligation is a defining feature of private law. In private law litigation, this means that the plaintiff ’s entitlement necessarily corresponds with the defendant’s obligation or liability: a remedy for the plaintiff can only be achieved through liability for the defendant, and relieving the defendant from liability means denying the plaintiff. The bipolar structure of private law is undoubtedly a constraint on the pursuit of community interests in private law. If, in private law litigation, a judge is to give effect to social or economic policy goals, the judge is at least to some extent constrained in doing so by the need to do justice to both the plaintiff and the defendant.52 The crucial question, though, is the nature and strength of the bipolarity constraint in private law: How are judges constrained in private law decision-making by the correlativity between right and obligation, and the consequent need to reconcile the pursuit of public goals with the need to do justice to individual plaintiffs and defendants? From a purely instrumentalist perspective, private law rules are or should be framed by reference to the goal of producing socially desirable outcomes, such as encouraging efficiency-enhancing behaviour. The need to do justice to the parties 48 N MacCormick, Legal Reasoning and Legal Theory (Oxford, Clarendon Press, 1978) 152. See also N MacCormick, Rhetoric and the Rule of Law: A Theory of Legal Reasoning (Oxford, Oxford University Press, 2005) ch 10. 49 BB Levenbook, ‘The Role of Coherence in Legal Reasoning’ (1984) 3 Law and Philosophy 355, 360. 50 ibid. 51 N MacCormick, ‘Coherence in Legal Justification’ in A Peczenik et al (eds), Theory of Legal Science (Dordrecht, D Reidel Publishing, 1984) 244. 52 In Mustapha v Culligan of Canada Ltd [2008] SCC 27 [16], for example, the Supreme Court of Canada observed that ‘[t]he law of negligence seeks to impose a result that is fair to both plaintiffs and defendants, and that is socially useful’.

Duty ofConstraints Care in A Rights-Based Theory of Negligence on Policy-Based Reasoning

273

to particular disputes is simply not part of the equation. Whether or not legal rules are ever framed without regard to considerations of justice between the parties, a great deal of scholarly analysis proceeds as though they can be. The notion of a penalty default rule in contract law provides a good example of this. A penalty default rule is a default rule (ie, one that can be modified or avoided by express contractual provision), which is deliberately set to produce an outcome that one or both of the parties would not want, in order to provide them with an incentive to disclose information to each other or to the courts.53 Parties who fail to respond to the incentive provided by the rule are penalised, either through the imposition of liability or through the denial of a right that would be recognised under a default rule that gave effect to conventional assumptions or expectations. Whether the rule operates justly or unjustly in its application to parties who fail to respond to the incentive has no bearing on the formulation of the rule. A good illustration of this kind of thinking is provided by Avery Katz’s analysis of the role that promissory estoppel should play in pre-contractual negotiations.54 Katz argues that in pre-contractual negotiations there is an optimal time for the parties to begin investing in the transaction. The risk of pre-contractual reliance should be imposed on the party who is best able to identify the optimal time for reliance and either take steps to ensure that the parties become bound at that time or take the risk of the reliance loss if the transaction does not proceed. Since this proposed rule is formulated by reference to the goal of encouraging efficient behaviour, rather than giving legal effect to a conventional understanding, Katz acknowledges that less experienced negotiators might have trouble adjusting. He suggests, however, that ‘a few costly mistakes might help speed up the learning process’.55 The interests of individual parties are to be sacrificed in pursuit of the community interest in encouraging efficient behaviour. The purely instrumental perspective takes no account of the need to do justice as between individual plaintiffs and defendants. At the opposite extreme, Weinrib argues that the reasons for the plaintiff ’s right and the defendant’s liability must be found in the interactions between the parties, and the reasons for the defendant’s obligation must correlate with the reasons for the plaintiff ’s entitlement. But even for those who view Weinrib’s notion of correlativity as excessively demanding, there is no doubt that the bipolar structure of private law constrains the pursuit of public goals. As Hanoch Dagan suggests, there is a middle ground between Weinrib’s notion of correlativity and unconstrained instrumentalism.56 Dagan argues, against unconstrained instrumentalism, that it is not enough that it is socially desirable for a class of people that includes the defendant to bear a responsibility to a class of people that 53 I Ayers and R Gertner, ‘Filling Gaps in Incomplete Contracts: An Economic Theory of Default Rules (1989) 99 Yale Law Journal 87, 91. 54 A Katz, ‘When Should an Offer Stick? The Economics of Promissory Estoppel in Preliminary Negotiations’ (1996) 105 Yale Law Journal 1249. 55 ibid, at 1303. 56 Dagan, above n 3, and ‘Just and Unjust Enrichments’ ch 17 of this book.

274

Andrew Robertson

includes the plaintiff. The constraint that is imposed by the bipolar structure of private law is not correlativity of justification, but the requirement of reconciling the promotion of community interests with the need to do justice as between individual parties. To begin to identify the nature of the bipolarity constraint, it is useful to consider three questions. First, can an obligation be imposed on a defendant purely on the basis of community welfare concerns? Secondly, can the extent of the defendant’s secondary or remedial obligation be determined exclusively by community welfare concerns? Thirdly, can a right that might otherwise be recognised be denied solely on the basis of community welfare concerns?

(i) Policy Supporting the Imposition of an Obligation There are some instances in which courts articulate ‘policy’ reasons in favour of liability.57 For example, the courts sometimes refer to the ‘policy’ of ensuring that victims of harm receive adequate compensation. There may be community welfare benefits in ensuring that those who are seriously injured are adequately compensated by those who have harmed them. This may, for example, help to ensure that the cost of care does not fall on the community, or that the severely injured are not left destitute. References to the ‘policy’ of ensuring that victims of harm receive adequate compensation often seem not to point to these community interests, however, but rather to considerations of fairness and justice between the parties. Policy, in the sense of considerations of community welfare, tends to weigh more heavily in favour of denial of a right than the imposition of an obligation. In her survey of ‘moral and policy concerns underlying the resolution of novel claims for a duty of care in negligence’, Jane Stapleton found relatively few convincing factors in the judgments in favour of the recognition of novel duties.58 The three pro-liability community welfare factors she identified (compensation, deterrence, and loss spreading) were said to be unconvincing because they are ‘present and would also weigh in favour of a duty in all cases, including the plethora of recent cases in which a duty has been denied’.59 Although these considerations are sometimes mentioned in judgments as weighing in favour of liability, it seems that, perhaps for the reasons Stapleton gives, they are insufficient on their own to justify liability. The defendant’s obligation must ultimately be justified by reference to the defendant’s responsibility. In Bazley v Curry, for example, the Supreme Court of Canada reformulated the principles relating to vicarious liability by reference to two ‘social goals’ weighing in favour of liability: first, providing effective compensation to victims of harm

57 See S Waddams, Dimensions of Private Law: Categories and Concepts in Anglo-American Legal Reasoning (Cambridge, Cambridge University Press, 2003) 202–3. 58 J Stapleton, ‘Duty of Care Factors: A Selection from the Judicial Menus’ in P Cane and J Stapleton (eds), The Law of Obligations: Essays in Celebration of John Fleming (Oxford, Clarendon Press, 1998) 72. 59 ibid, at 64.

Duty ofConstraints Care in A Rights-Based Theory of Negligence on Policy-Based Reasoning

275

and, secondly, reducing the risk of future harm by encouraging employers to engage in ‘imaginative and efficient administration and supervision’ of activities which create risks.60 The court was concerned to formulate and apply the rule in a way that ensured that the interests of the defendant were not sacrificed in pursuit of the relevant policy goals. The goal of compensation could be pursued only to the extent that this was ‘fair’ to the employer, which required that it was justifiable in light of the employer’s responsibility for the harm, as a result of introducing the risk into the community.61 The ‘enterprise risk’ test formulated by the Supreme Court in Bazley was ‘animated’ by the identified policy goals, but the pursuit of those policy goals was constrained by the need to reconcile them with the bipolar nature of private law rights and the bipolar structure of private law litigation.

(ii) Policy Influencing the Extent of the Remedy A more controversial question is whether the nature and extent of a plaintiff ’s right must be justifiable in terms that relate specifically to the plaintiff. Is the defendant also entitled to ask for a non-public justification as to why this particular plaintiff should be entitled to the remedy in question?62 This issue arises in relation to punitive and gain-based damages awards, which are commonly opposed on the basis that they provide a windfall to the plaintiff.63 The conversation in relation to these awards is a familiar one. The awards are justified on the (public) basis that they punish wrongdoers and deter wrongful conduct. The fact that the justification for the defendant’s liability is grounded in the community interest raises the question of why the plaintiff should receive the award. Another public justification is routinely advanced: paying these awards to the plaintiff provides the plaintiff with an incentive to sue, and thus reinforces the deterrent effect of the rule. The remedy then has an entirely public justification, founded on the community interest in deterrence, rather than an explanation for the entitlement of the particular plaintiff. As Weinrib says, the plaintiff ’s right of action then serves ‘as a mechanism for bribing someone to vindicate the collective interest in deterring the defendant’s inefficient behaviour’.64 The widespread concern among scholars about the windfall this is said to provide to the plaintiff may indicate that there is something in the structure of private law that requires an explanation for a private law remedy that is based on the relationship between the plaintiff and the defendant. If such an explanation is needed, then in the case of punitive damages

60

Bazley v Curry [1999] 2 SCR 534 [29]–[34]. The conclusion that it was fair for the employer to bear the risk was said to be ‘buttressed’ by the policy consideration that ‘the employer is often in a position to spread the losses through mechanisms like insurance and higher prices, thus minimizing the dislocative effect of the tort within society’: ibid, at [31]. 62 Dagan, above n 3, at 816 answers this question in the affirmative: ‘the plaintiff … needs to justify her entitlement to the specific measure of recovery she seeks to impose on the defendant’. 63 See, eg, R Cunnington, ‘Should Punitive Damages Be Part of the Judicial Arsenal in Contract Cases?’ (2006) 26 Legal Studies 369, 382. 64 Weinrib, The Idea of Private Law, above n 1, at 11. 61

276

Andrew Robertson

awards, it may be found in the notion that punitive awards redress moral injury65 or respond to dignitary loss66 resulting from the wrongdoer’s exploitation of, and assertion of superiority over, the victim. Whether it may be possible to reconcile exemplary damages with the principle of corrective justice on the basis of the ‘moral injury’ theory67 is beside the point for present purposes. Our present concern is not with the question whether particular outcomes can be reconciled with corrective justice principles or, more broadly, rationalised on grounds that avoid policy or community welfare considerations. We are concerned here with what motivates judicial decisions, and whether particular outcomes are motivated by community welfare considerations unconstrained by the need to provide a ‘non-public’ justification for the outcome. The orthodoxy of appellate court judgments holds that exemplary damages awards are motivated by community welfare considerations. In considering whether exemplary damages awards could be made in response to inadvertently negligent conduct in A v Bottrill, for example, Lord Nicholls of Birkenhead observed that ‘[t]heir primary function is to punish. They also serve as an emphatic vindication of the plaintiff ’s rights and as a deterrent’.68 In Gray v Motor Accident Commission the High Court of Australia held that exemplary damages may not be awarded in cases where substantial punishment has been inflicted through the criminal law, because ‘the purposes for the awarding of exemplary damages have been wholly met if substantial punishment is exacted by the criminal law’.69 Those purposes had earlier been identified as being ‘to punish the wrongdoer and deter others from like conduct, not to compensate the party that was wronged’.70 An earlier decision that exemplary damages may be awarded even where the wrongdoer is insured under a scheme of compulsory insurance was motivated in part by the idea that exemplary damages awards serve a secondary purpose of assuaging the victim’s urge for revenge and discouraging any temptation to ‘endanger the peace’ by retaliating.71 These, too, are obviously community welfare goals. If the purposes of exemplary damages are punishment and deterrence, and these purposes are ‘wholly met’ by substantial punishment of the wrongdoer through the criminal law, then it would seem that in cases where exemplary damages are awarded, the extent of the remedy is governed, entirely in the case of Gray

65 J Hampton, ‘Correcting Harms Versus Righting Wrongs’ (1992) 39 UCLA Law Review 1659; M Galanter and D Luban, ‘Poetic Justice: Punitive Damages and Legal Pluralism’ (1993) 42 American University Law Review 1393; PW Lee, ‘Contract Damages, Corrective Justice and Punishment’ (2007) 70 Modern Law Review 887. 66 A Ripstein, Equality, Responsibility and the Law (Cambridge, Cambridge University Press, 1999) 151. Ripstein argues that the essential feature of punitive damages cases is that ‘the wrongdoer acts in pursuit of his or her own ends by intentionally exposing others to injuries or unreasonable risks’. 67 As argued by Lee, above n 65. 68 A v Bottrill [2003] 1 AC 449 (PC) 457. The dissentients in A v Bottrill, Lords Hutton and Millett, also took the view that ‘the primary purpose of exemplary damages is to punish’ (at 465). 69 Gray v Motor Accident Commission (1998) 196 CLR 1 (High Court of Australia) 14. 70 ibid, at 12. 71 Lamb v Cotogno (1987) 164 CLR 1 (High Court of Australia) 9–10.

Duty ofConstraints Care in A Rights-Based Theory of Negligence on Policy-Based Reasoning

277

and almost entirely in the case of Bottrill, by community welfare considerations. There is no real attempt to justify the extent of the plaintiff ’s right in terms that relate specifically to the plaintiff. The bipolar structure of the private law relationship does not, therefore, appear to constrain the pursuit of community welfare goals in the making of exemplary damages awards. In these cases it seems that community interests trump the desirability of ensuring that the plaintiff does not obtain a windfall.

(iii) Policy Requiring the Denial of Rights The other context in which community interests can operate as a trump is in relation to the denial of rights that a person would otherwise have. These are situations in which rights that would otherwise exist are denied on the basis of an overwhelming community interest, without any concern for reconciling the outcome with the interests of justice between the parties. We see this, for example, when a contractual provision or a trust is struck down on the basis of a ‘public policy’ consideration. In the restraint of trade cases, for instance, contractual rights are denied on the basis of the overriding public interest in each person being able to carry on his or her trade or profession.72 We also see the denial of rights in the interests of community welfare when the courts refuse to enforce a contractual or equitable obligation of confidence on the grounds of an overriding public interest in the disclosure of the information in question.73 In AttorneyGeneral v Observer Ltd, Lord Goff of Chieveley said that although the basis of the law’s protection of confidence is that there is a public interest that confidences should be preserved and protected by the law, nevertheless that public interest may be outweighed by some other countervailing public interest which favours disclosure … It is this limiting principle which may require a court to carry out a balancing operation, weighing the public interest in maintaining confidence against a countervailing public interest favouring disclosure.74

In Lion Laboratories Ltd v Evans,75 for example, the public interest in the disclosure of confidential information about the reliability of a breath-testing unit used by police was held to override the obligation of confidence that would otherwise have been owed in relation to that information. In these cases, there is no attempt to justify the outcome in terms of justice or fairness between the parties. The concern to ensure justice between the parties is overwhelmed by the community

72 The clearest example is the restraint of trade doctrine; see, eg, Nordenfelt v Maxim Nordenfelt Guns & Ammunition Co Ltd [1894] AC 535 (HL); Esso Petroleum Co Ltd v Harper’s Garage (Stourport) Ltd [1968] AC 269 (HL); Peters (WA) Ltd v Petersville Ltd (2001) 205 CLR 126 (High Court of Australia) and MJ Trebilcock, The Common Law of Restraint of Trade: A Legal & Economic Analysis (Toronto, Carswell, 1986) esp at 106–19, 321–3 and 346–54. 73 See, eg, Minister for Immigration and Citizenship v Kumar [2009] HCA 10 [25]–[27]. 74 Attorney-General v Guardian Newspapers Ltd (No 2) [1990] 1 AC 109 (HL) 282. 75 Lion Laboratories Ltd v Evans [1985] QB 526 (CA). See further Y Cripps, The Legal Implications of Disclosure in the Public Interest, 2nd edn (London, Sweet & Maxwell, 1994).

278

Andrew Robertson

interest, which alone justifies depriving a party of rights he or she would otherwise have.76 As with exemplary damages, it does not matter for present purposes if a justification for some of these doctrines can be given that does not involve community welfare concerns. Stephen Smith has, for example, offered an autonomy-based justification for what he calls ‘substantive limitations on the enforceability of agreements’, which broadly cover the ground of illegality and public policy unenforceability doctrines.77 The relevant justification Smith offers is that there can be no moral obligation, and therefore no legal obligation, to engage in an objectionable activity, and this includes agreements that impose undue restrictions on personal autonomy. That explanation is ‘internal to contract law’ in the sense that, rather than being created by the parties and struck down in the public interest, a binding obligation does not ever arise for reasons grounded in notions of interpersonal morality. But Smith does not rationalise the current law on restraint of trade on the basis of autonomy; indeed he questions whether the courts are too ready to intervene in such cases.78 Even if some version of the restraint of trade doctrine could be rationalised on the basis of a concern with individual autonomy, the important point is that this is not the way in which restraint of trade cases are currently decided. The protection of the community interest is the explicit reason the plaintiff is denied the rights that he or she would otherwise have. The point for present purposes is that, whether or not other explanations might be given for the outcomes of some of the cases in question, community interests do operate, on their own, as justifications for the denial of legal rights. In the relevant cases the courts make no attempt to justify the outcome on any other basis. The recognition that parties are routinely deprived of private law rights they would otherwise have because of overwhelming community welfare considerations may have implications for the debate over the use of policy considerations in the law of negligence. Ernest Weinrib and Allan Beever have expressed concern about the situations in which a duty of care which would otherwise be held to arise is denied solely on the basis of a community interest, such as a concern that recognition of a duty of care in the given situation would create indeterminate liability or lead police or other public authorities to adopt defensive practices.79 Weinrib opposes the notion that policy considerations can play a decisive role in the denial of a duty of care: A plaintiff can therefore be denied compensation on the basis of policy considerations that, while one-sidedly pertinent to the defendant or to persons carrying on a similar activity, have no normative bearing on the position of the plaintiff as the sufferer of an injustice. From the plaintiff ’s point of view, the denial of recovery, operating (as the

76

Other examples are noted by Waddams, above n 57, at 203–5. SA Smith, Contract Theory (Oxford, Clarendon Press, 2004) 249–63. ibid, at 258–60. 79 See, eg Mitchell v Glasgow City Council [2009] UKHL 11 and E Chamberlain, ‘Negligent Investigation: Tort Law as Police Ombudsman’ ch 12 of this book. 77 78

Duty ofConstraints Care in A Rights-Based Theory of Negligence on Policy-Based Reasoning

279

court says) extrinsically to simple justice, amounts to the judicial confiscation of what is rightly due to the plaintiff in order to subsidize policy objectives unilaterally favourable to the defendant and those similarly situated.80

If the denial of rights that would otherwise exist on the basis of community interest does amount to ‘judicial confiscation’ or treating the plaintiff ‘merely as a means to others’ ends’,81 a broader view reveals that this is something that happens throughout private law. We can see in the law of contract, the law of trusts and the law of confidential information that the notion of ‘private right’ giving way to considerations of ‘public interest’82 underlies long-established private law doctrines. If private law exists as a distinctive category of law then we should expect it to be coherent as a justificatory system. The fact that rights (that would otherwise exist) are routinely denied on the basis of overwhelming public interest in one area of private law lends legitimacy to a similar justificatory pattern in another area of private law. As noted above, normative coherence does not provide a complete justification for any particular rule denying liability in the public interest, but it does establish a baseline eligibility or legitimacy for private law rules that have that effect.83

IV. Conclusions This chapter has attempted to sketch the different ways in which common law judges are constrained in the use they can make of policy considerations in framing legal principles and deciding cases. Taken together, the constraints discussed above significantly limit the extent to which, and the ways in which decisionmaking in private law can be guided by policy considerations. This is not to underestimate the importance of community welfare considerations in private law. Policy considerations have played and continue to play a very significant role in shaping private law rules; they therefore underlie and are embedded in many private law doctrines and are given effect in many cases where they are not explicitly mentioned. Community welfare considerations also have a direct role to play in the application of important private law doctrines. The constraints identified in this chapter show, however, that when judges take account of policy considerations they do so in a constrained judicial capacity and not in a relatively unconstrained legislative capacity. This has implications for both formalist and instrumentalist approaches to private law. The existence of these constraints undermines the formalist claim that judges cannot take account of community welfare considerations without taking on an illegitimate legislative role. On the 80 81 82 83

Weinrib, ‘Does Tort Law Have a Future?’, above n 1, at 566. Beever, above n 1, at 176–7. To borrow the language of Waddams, above n 57, ch 10. See above nn 48–51 and accompanying text.

280

Andrew Robertson

other hand, a purely instrumentalist analysis of private law doctrines, which treats private law as a system of regulation governed exclusively by community welfare, also offers an incomplete account. The utility of economic analysis in the development of private law doctrines is limited if it fails to take account of the bipolar structure of private law, and the need to balance community welfare with the need to do justice to parties to particular disputes.

12 Negligent Investigation: Tort Law as Police Ombudsman ERIKA CHAMBERLAIN*

I. Introduction Justice Allen Linden has argued that tort law should serve an ‘ombudsman’ function as it moves into the future.1 Noting that the compensation function of tort law has waned in importance in the modern welfare state, and that damages may not always be an effective deterrent, Linden has called for tort law to be used ‘as an instrument of social pressure upon centres of governmental, financial and intellectual power’.2 More recently, Linden has argued that the ombudsman function of tort law provides psychological and political empowerment to victims, allowing them to expose shortfalls in government, obtain an acknowledgment of poor treatment, and stimulate changes in policies or practice.3 Indeed, tort law’s most influential role in the new millennium may be in drawing publicity toward the conduct of certain defendants, thereby forcing them to justify their actions not only to the court, but to the public at large. A tort claim may also encourage defendants to modify their conduct to avoid similar suits or to convince the public that they have improved their standards. Among other areas, Justice Linden has argued that tort law can fill an ombudsman role in supervising the actions of police. Tort actions for battery, false imprisonment, and malicious prosecution have helped to regulate police conduct during the process of arrest, detention and prosecution.4 However, the focus * The author would like to thank the attendees of the Obligations IV conference for their helpful comments, and Daniel Daniele for his research assistance. 1 AM Linden, ‘Tort Law as Ombudsman’ (1973) 51 Canadian Bar Review 155; AM Linden, ‘Reconsidering Tort Law as Ombudsman’ in F Steel and S Rogers-Magnet (eds), Issues in Tort Law (Toronto, Carswell, 1983) 1; and AM Linden, ‘Torts Tomorrow—Empowering the Injured’ in NJ Mullany and AM Linden (eds), Torts Tomorrow: A Tribute to John Fleming (Sydney, Law Book Company Information Services, 1998) 321. 2 Linden, ‘Tort Law as Ombudsman’, ibid, at 156. 3 See generally Linden, ‘Torts Tomorrow—Empowering the Injured’, above n 1. 4 See eg, Christie v Leachinsky [1947] AC 573 (HL); and Koechlin v Waugh and Hamilton (1957) 11 DLR (2d) 447 (Ontario CA), where the respective courts allowed tort claims for wrongful arrest because the arrestees were not informed of the true charges against them.

284

Erika Chamberlain

of this chapter is on the more contentious tort sometimes known as ‘negligent investigation’. This tort applies to two main sets of plaintiffs: victims of criminal conduct who claim that police ought to have apprehended the criminal sooner or offered protection to the victim; and suspects who are wrongly charged, tried or convicted of crimes as a result of flawed police investigations. Both types of claim have received a relatively cool reception in the common law courts. Nevertheless, the recent Supreme Court of Canada decision in Hill v Hamilton-Wentworth Regional Police Services Board,5 which affirmed that police owe a duty of care to criminal suspects, has renewed the debate about the role of negligence law in overseeing the conduct of police investigations. This chapter argues that negligent investigation is an appropriate use of the ombudsman function of tort law. The policy reasons typically used to deny a duty of care—interference with police discretion and the unwanted potential for defensive policing—are unsubstantiated and, at any rate, hardly unique to police. As long as the courts insist on close proximity and maintain a lenient standard of care, there is no reason to believe that the potential for negligence liability will hinder police investigations. Instead, the recognition of a duty of care should encourage greater care by police and an improvement in standard investigatory practices. Perhaps more importantly, negligence liability will provide vindication to those who have suffered losses as a result of negligent police conduct. The existing literature suggests that police complaints procedures do not provide satisfaction and, if anything, affirm the complainants’ negative perceptions of the police; in particular, complainants are often sceptical about the quality and objectivity of internal investigations.6 By contrast, a tort claim can bring public attention to allegations of police negligence, and can reassure the victim that her complaints have been heard. The tort of negligent investigation can, therefore, provide a voice to historically disadvantaged groups. This chapter focuses on two such groups: victims of domestic violence, who complain that police disregard their requests for protection; and minority groups, who complain that they are unfairly targeted by criminal investigations. It concludes that the tort of negligence can serve an important political and social function in monitoring police conduct and exposing the treatment of these historically under-represented groups.

5

Hill v Hamilton-Wentworth Regional Police Services Board [2007] SCC 41, 285 DLR (4th) 620. The literature is vast. See eg, AJ Goldsmith, ‘What’s Wrong with Complaint Investigations? Dealing with Difference Differently in Complaints Against Police’ (1996) 15 Criminal Justice Ethics 36; T Landau, ‘When Police Investigate Police: A View from Complainants’ (1996) Canadian Journal of Criminology 291; K Strudwick, ‘Is Independence the Only Answer to Complainants’ Satisfaction of the Police Complaints Process? A Perspective from the United Kingdom’ (2003) 4 Police Practice and Research 35; and PJ Lesage, Report on the Police Complaints System in Ontario (Toronto, Ontario Ministry of the Attorney General, 2005) (Lesage Report). 6

Tort Law as Police Ombudsman

285

II. Negligence Claims by Victims of Crime A. The General Commonwealth Position Throughout most of the Commonwealth, negligence claims by victims of crime have been dismissed at the duty stage. While there are differences in the tests for the duty of care in various jurisdictions, the analysis generally involves some examination of foreseeability, proximity, and policy.7 The seminal English case, Hill v Chief Constable of West Yorkshire,8 focused on the latter two elements. The claimant in Hill was the mother of the last victim of Peter Sutcliffe, the ‘Yorkshire Ripper’. She alleged that, had police not been negligent in conducting their investigation, they would have arrested Sutcliffe before he had a chance to kill her daughter. Thus, the claim effectively sought to hold police responsible for the criminal acts perpetrated by a third party. Lord Keith, who wrote the leading opinion denying a duty of care, stressed that such claims require a special degree of proximity. For instance, in Home Office v Dorset Yacht,9 the plaintiffs had been successful because their yacht was moored in close proximity to an open borstal institution. They therefore had a close and direct relationship to the defendant officers, whose negligent supervision allowed some boys to escape and damage their yacht. By contrast, Lord Keith argued, Miss Hill was one of ‘a vast number of the female general public who might be at risk from [Sutcliffe’s] activities but was at no special distinctive risk in relation to them’.10 Lord Keith explained: Miss Hill cannot for this purpose be regarded as a person at special risk simply because she was young and female. Where the class of potential victims of a particular habitual criminal is a large one the precise size of it cannot in principle affect the issue. All householders are potential victims of an habitual burglar, and all females those of an habitual rapist. The conclusion must be that although there existed reasonable foreseeability of likely harm to such as Miss Hill if Sutcliffe were not identified and apprehended, there is absent from the case any such ingredient or characteristic as led to the liability of the Home Office in the Dorset Yacht case.11

Thus, the plaintiff ’s claim failed for insufficient proximity. Nevertheless, Lord Keith went on to explain that the claim also failed on the grounds of public policy. His policy concerns can be summarised as follows: —

Tort liability would not appreciably improve the standards of police behaviour, since they are already motivated by a ‘general sense of public duty’.

7 The governing case in England is Caparo Industries plc v Dickman [1990] 2 AC 605 (HL), which asks whether the loss was reasonably foreseeable, whether there was a proximate relationship between the parties, and whether it is fair, just and reasonable for the court to impose a duty of care. See also Pyrenees Shire Council v Day (1998) 192 CLR 330 (High Court of Australia); and Cooper v Hobart (2001) 206 DLR (4th) 193 (Supreme Court of Canada). 8 Hill v Chief Constable of West Yorkshire [1989] 1 AC 53 (HL). 9 Home Office v Dorset Yacht [1970] AC 1004 (HL). 10 Hill [1989] 1 AC 53 (HL) 62. 11 ibid.

286 — —

—

Erika Chamberlain The threat of tort liability might cause police to carry out their activities in a ‘detrimentally defensive frame of mind’. The conduct of police investigations is primarily discretionary, involving the balancing of various interests and the appropriate allocation of resources. Such considerations are largely non-justiciable. Police will be subject to a flood of claims, and will have to divert financial resources and manpower to the defence of those claims. This will detract attention from their primary function of suppressing crime.

As a result of these concerns, Lord Keith believed that the best solution was to make police immune from claims of the sort raised in Hill. This immunity was soon tested in the Court of Appeal’s decision in Osman v Ferguson,12 where a school teacher became obsessed with the plaintiff, his student.13 The plaintiff complained to the school and police, and the headmaster referred the teacher for psychological assessment. The teacher harassed the plaintiff over the course of several months, committing numerous acts of vandalism, all of which were reported to police. The police decided to arrest the teacher for criminal damage, but took no effective action to make the arrest. The teacher ultimately shot and wounded the plaintiff, and killed the plaintiff ’s father. The plaintiff sued the police for failing to arrest or thoroughly investigate the teacher, or to link him to the stolen shotgun with which the crimes were committed. Osman seemed a much stronger case than Hill, because there was a greater degree of proximity between the plaintiff and the police: Osman was a known target of a known criminal, and there was plenty of warning that the violence would escalate. However, relying on Hill, the court rejected the claim for policy reasons, thereby affirming police immunity in negligence for their conduct of a criminal investigation. This decision was, of course, criticised by the European Court of Human Rights for denying the plaintiff the right of access to a court.14 The Commission decided that the public policy factors should not have acted as a watertight defence, and that the plaintiff should have had the opportunity to have his case adjudicated on its merits. While the European Court’s decision may have caused the English courts to avoid using language suggesting blanket immunity,15

12 Osman v Ferguson [1993] 4 All ER 344 (CA). See also Alexandrou v Oxford [1993] 4 All ER 382 (CA), where the plaintiff claimed that the police were negligent in responding to a burglar alarm at his clothing shop. The Court of Appeal dismissed the claim on the authority of Hill. 13 For a more complete summary of the facts, see LCH Hoyano, ‘Policing Flawed Investigations: Unravelling the Blanket’ (1999) 62 Modern Law Review 912. 14 Osman v UK (1998) 29 EHRR 245. This was contrary to Art 6(1) of the European Convention on Human Rights. 15 See Brooks v Commissioner of Police for the Metropolis [2005] UKHL 24 [27] (Brooks). See also Swinney v Chief Constable of the Northumbria Police [1996] 3 All ER 449 (CA), where the Court of Appeal declined to refer to Hill as conferring ‘blanket immunity’, and instead called it a ‘principle of general application’.

Tort Law as Police Ombudsman

287

its effects have been minimal and police still enjoy relative immunity in negligence for their conduct during an investigation.16 The continued vigour of Hill is illustrated in the recent House of Lords decisions in Brooks v Commissioner of Police for the Metropolis17 and Smith v Chief Constable of Sussex Police.18 The plaintiff in Brooks was a friend of Stephen Lawrence and a co-victim of the vicious racial attack in which Lawrence was killed. He sued police for his treatment at the scene of the crime, where he was stereotyped as an agitated young black man who could not assist in capturing the perpetrators. No one showed concern for his condition or tried to calm him down. In addition, he suffered frustration at the failed prosecution of the perpetrators, which was partly the result of ‘fundamental errors’ in the investigation.19 In the final report of the inquiry into Lawrence’s death, Sir William Macpherson summarised: ‘The investigation was marred by a combination of professional incompetence, institutional racism and a failure of leadership by senior officers’.20 In ultimately dismissing Brooks’s claim against the police, Lord Steyn conceded that there were a ‘litany of derelictions of duty and failures in the police investigation’.21 He also took a ‘more sceptical approach’ to examinations of police conduct than had the courts in Hill.22 Nevertheless, he concluded that the ‘core principle of Hill has remained unchallenged’.23 Lord Steyn wrote: A retreat from the principle in Hill would have detrimental effects for law enforcement. Whilst focusing on investigating crime, and the arrest of suspects, police officers would in practice be required to ensure that in every contact with a potential witness or a potential victim time and resources were deployed to avoid the risk of causing harm or offence. Such legal duties would tend to inhibit a robust approach in assessing a person as a possible suspect, witness or victim. By placing general duties of care on the police to victims and witnesses the police’s ability to perform their public functions in the interests of the community, fearlessly and with despatch, would be impeded. It would, as was recognised in Hill, be bound to lead to an unduly defensive approach in combating crime.24

16 The effects of Osman v UK (1998) 29 EHRR 245 were muted by the European Court’s subsequent decision in Z v UK (2002) 34 EHRR 2 (reviewing the House of Lords decision in X (Minors) v Bedfordshire County Council [1995] 2 AC 633). The Court found that the dismissal of a claim against child welfare authorities for lack of duty was not based on immunity or an exclusionary rule, but on the principles of English law. Thus, it was not a denial of Convention rights to dismiss the case at the duty stage. 17 Brooks [2005] UKHL 24. 18 Smith v Chief Constable of Sussex Police [2008] UKHL 50. 19 Secretary of State for the Home Department, ‘The Stephen Lawrence Inquiry’ (Cm 4262-I, 1999) para [46.1]. 20 ibid. 21 Brooks [2005] UKHL 24 [8]. 22 ibid, at [28]. 23 ibid, at [30]. 24 ibid.

288

Erika Chamberlain

Similar reference to the community interest served to undercut the alleged duty of care in Smith.25 The plaintiff had been assaulted by his estranged partner, who was taken into custody but never prosecuted. The plaintiff subsequently received numerous threats from his partner and reported these to police, explaining that he feared for his life. The police did not take notes about Smith’s complaints or complete a crime form. Eventually, Smith’s partner came to his home and attacked him with a claw hammer, leaving him with a fractured skull and brain damage. Although Lord Hope conceded that there was a ‘highly regrettable failure to react to a prolonged campaign’ of threatened violence,26 he affirmed that, ‘in the interests of the whole community’, a duty of care should not be imposed.27 Thus, the policy considerations from Hill can serve to negate any duty of care, even in cases where there is a close and direct relationship with the plaintiff, no threat of indeterminate liability, and no obvious question of discretion or resource allocation, and where the relevant police conduct is roundly criticised. The duty of care is declined due to rather vague and unsubstantiated claims about the public interest. Respectfully, it is difficult to understand how it is in the public interest for police to treat victims like Brooks and Smith in such an insensitive, stereotypical way, with no regard for their safety. Moreover, it is remarkable that the House of Lords has continued to defend the exercise of police discretion in cases where that discretion is exercised in a racist, sexist or homophobic fashion. The Australian courts have largely accepted the authority of Hill,28 although most of the leading appellate decisions deal with duties owed to suspects under investigation, and are accordingly discussed in Section III of this article. Elsewhere in the Commonwealth, the principles of Hill have been substantially, though not universally,29 accepted with respect to claims by crime victims.30 It is often accepted without argument that a duty of care would cripple police in their task of investigating and suppressing crime. While the courts have acknowledged that this might leave apparently deserving plaintiffs without remedy, they conclude that this is the price to be paid for the greater public good of an efficient police service.31 Police are given a de facto immunity for negligence in the conduct of an investigation, a privilege that a diminishing number of professions can claim.

25

Smith [2008] UKHL 50. ibid, at [72]. ibid, at [75]. 28 See especially Sullivan v Moody (2001) 207 CLR 562 (High Court of Australia) [56], where Hill was described as ‘instructive’, even though ‘not determinative’. See also Peat v Lin [2004] QSC 219 (Supreme Court of Queensland). 29 See especially Carmichele v Minister of Safety and Security (2001) 10 B Const LR 995, where the Constitutional Court of South Africa concluded that foreseeability and proximity were sufficient to control the scope of liability, and that police immunity in negligence was unnecessary. 30 See I Fraser, ‘Police Negligence in the Pacific: Building a Case around Hill’ (2006) 32 Commonwealth Law Bulletin 299, for a review of police negligence cases in the Pacific where public policy issues were considered. 31 See for example, Brooks [2005] UKHL 24 [31]. 26 27

Tort Law as Police Ombudsman

289

(i) Jane Doe v Metropolitan Toronto (Municipality) Commissioners of Police32 The decision of MacFarland J in Jane Doe is perhaps the best-known counterpoint to Hill v Chief Constable of West Yorkshire in a case alleging negligent investigation. The plaintiff was attacked and raped in her Toronto apartment by a man who came to be known as the ‘balcony rapist’. All of his victims lived in second- or third-floor apartments, to which he gained entry through the balcony door, and all lived within the Church/Wellesley area of Toronto. Police suspected that the rapist might live in the area, but did not want to issue a warning to residents for fear that women would become hysterical, the attacker would flee, and the investigation would be compromised.33 When the plaintiff was interviewed by police after the attack, she expressed shock that women in the area had not been warned. She later claimed, and MacFarland J agreed, that women in the area were being used as ‘bait’ for the rapist.34 They were left vulnerable to attack and, due to their ignorance of attacks in the area, were denied the opportunity to take precautions for their own safety. The duty of care was examined by reference to the test in Anns v Merton London Borough Council,35 as then accepted by the Supreme Court of Canada in Kamloops (City) v Nielsen.36 MacFarland J accepted the reasons of Moldaver J from an earlier motion in the Divisional Court.37 Since the defendants knew that there was a serial rapist at large, it was ‘eminently foreseeable that he would strike again’ and harm another victim.38 The element of proximity was satisfied because Jane Doe was a member of a ‘narrow and distinct group of potential victims’: all were single, white females; all lived in the Church-Wellesley area; and all lived in second- or third-floor apartments. Further, while police might make a legitimate policy decision not to warn women on the grounds that it would cause unnecessary panic and lead to greater harm, Moldaver J concluded that this would not excuse police from their duty to protect those at risk. In other words, police had a duty to either warn the women (so that they could take their own protective measures) or take steps to protect them from attack. In failing to do either, police ‘failed utterly in the duty of care they owed Ms. Doe’.39 MacFarland J ultimately awarded Jane Doe approximately CAN$220,000 in damages.40 32 Jane Doe v Metropolitan Toronto (Municipality) Commissioners of Police (1998) 160 DLR (4th) 697 (Ontario Court of Justice, General Division) (Jane Doe). 33 The police believed that another serial rapist, known locally as the Annex rapist, had previously fled the area because of intensive media coverage and police presence. 34 Jane Doe (1998) 160 DLR (4th) 697 (Ontario Court of Justice, General Division) [118]. 35 Anns v Merton London Borough Council [1978] AC 728 (HL). 36 Kamloops (City) v Nielsen [1984] 2 SCR 2. 37 Jane Doe v Metropolitan Toronto (Municipality) Commissioners of Police (1990) 74 OR (2d) 225 (Div Ct). 38 ibid, at 230. 39 Jane Doe v Metropolitan Toronto (Municipality) Commissioners of Police (1990) 74 OR (2d) 225 (Div Ct) [174]. 40 This included the costs of psychiatric and chiropractic treatment, prescription medication, and night-time cab fares (since Ms Doe was no longer able to take public transportation at night).

290

Erika Chamberlain

It has been argued that the decisions in Jane Doe gave scant consideration to policy under the second stage of Anns,41 and it is true that the courts did not consider the various factors that served to negative the duty in Hill v Chief Constable of West Yorkshire. Nevertheless, it can hardly be said that policy considerations were absent in Jane Doe. MacFarland J wrote at length about police attitudes toward sexual assault victims and investigations, explaining that, in spite of claims of reform, problems at the police force remained. Officers were inadequately trained, treated victims insensitively, and persisted in believing rape myths.42 The evidence suggested that police had not taken the reports of previous rapes seriously, did not follow up in a timely fashion, and believed that some of the reports were fictitious. They even considered laying a public mischief charge against one of the complainants. It is reasonable to assume that the rapist might have been apprehended sooner, if police had assigned greater priority to the investigation. Moreover, MacFarland J found that the defendants’ conduct of the investigation, especially the failure to warn, was ‘motivated and informed by the adherence to rape myths as well as sexist stereotypical reasoning about rape, about women and about women who are raped’.43 Thus, it is inaccurate to suggest that MacFarland J did not consider policy factors in reaching her decision, which was quite clearly based on her assessment of police attitudes toward sexual assault and the effects of those attitudes on the investigation. She made extensive reference to reports on police treatment of sexual assault,44 and concluded that they were not adequately educating officers and enforcing their own protocols. MacFarland J evidently believed that the public interest weighed in favour of imposing a duty of care in the situation, as it might help to improve accountability and force the police to make more effective reforms. Indeed, given her findings regarding the stereotypes that motivated the police conduct, it would have been surprising if MacFarland J had negated the duty of care based on the policy factors outlined in Hill. It would have been unreasonable for the police to avoid liability on the basis that their functions are discretionary, when the evidence in the case suggested that they exercised their discretion on stereotypical, discriminatory grounds. Further, the police in Jane Doe had not acted out of a sense of public duty: they showed little concern for the safety of women in the area and downplayed the seriousness of sexual assault. Finally, since police had not taken effective action on the numerous reports about the poor conduct of

41 See C Moroz, ‘Jane Doe and Police Liability for Failure to Apprehend: The Role of the Anns Public Policy Principle in Canada and England’ (1995) 17 Advocates’ Quarterly 261. 42 Jane Doe v Metropolitan Toronto (Municipality) Commissioners of Police (1990) 74 OR (2d) 225 (Div Ct) [154]–[159]. 43 ibid, at [162]. MacFarland J therefore found that the plaintiff ’s equality rights under s 15 of the Charter of Rights and Freedoms had been violated. 44 For instance, the Report of the Police Committee on Rape (Metropolitan Toronto Police Force, 1975); and the Report of the Public Task Force on Violence Against Women and Children (1984) (the Godfrey Task Force). McFarland J also referred to numerous internal police documents and memoranda.

Tort Law as Police Ombudsman

291

sexual assault investigations in the past, it was apparent that additional sanctions were necessary to encourage reform. In fact, this is exactly what Jane Doe wanted.45 She brought her claim so that police tactics would be brought to public attention. As she stated, ‘[i]t was a political action for me, and that was accomplished. I won the day the trial started’.46 In the words of Justice Linden, the tort claim empowered the victim both psychologically and politically. Toronto City Council subsequently apologised to Doe and to all the women of Toronto.47 Jane Doe also received an apology from the Chief of Police, who undertook to make improvements in the conduct of sexual assault investigations.48 Finally, while not expressly linked to the outcome of the Jane Doe litigation, it seems that police changed their tactics during the so-called Scarborough rapes of 1999, where they issued widespread public warnings to potential victims.49 Thus, the action in Jane Doe was promising for the ombudsman function of tort law: it brought attention to discriminatory practices, garnered an apology from the actors, and inspired institutional change. In response to Jane Doe, the City Auditor conducted a review of sexual assault investigations and made several recommendations, including ongoing training for officers and better communications with victims and the public.50 The Auditor followed up with a report in 2004 and established a Sexual Assault Audit Steering Committee.51 Although the Committee was recently disbanded,52 and there are lingering questions about the extent to which sexual assault investigations have been substantively improved,53 it is significant that, in the 10 years since Jane Doe, the conduct of sexual assault investigations by Toronto police has remained a public issue. Jane Doe herself remains a prominent figure in discussions about the police treatment of sexual

45 Early on, she had offered to settle her case with police for CAN$50,000, an apology, and a promise of change in sexual assault procedures: see M Randall, ‘Sex Discrimination, Accountability of Public Authorities and the Public/Private Divide in Tort Law: An Analysis of Doe v Metropolitan Toronto (Municipality) Commissioners of Police’ (2001) 26 Queen’s Law Journal 451, 463. 46 Linden, above n 3, at 333, quoting from K Honey, ‘Rape Victim Wronged by Police, Judge Rules’ The Globe and Mail (4 July 1998) A1. 47 P Gombu, ‘Doe Gets Apology’, The Toronto Star (10 July 1998) A1. 48 C Freeze, ‘Jane Doe Gets Apology from Chief of Police’, The Globe and Mail (17 July 1998) A3. 49 Randall, above n 45, at 494. But see S Childs and P Ceyssens, ‘Doe v Metropolitan Toronto Board of Commissioners of Police and the Status of Public Oversight of the Police in Canada’ (1998) 36 Alberta Law Review 1000, 1013, where the authors are critical of Jane Doe for providing little guidance to police in discharging their duty to warn. 50 Review of the Investigation of Sexual Assaults—Toronto Police Service (Toronto, Audit Services, 1999). 51 The Auditor General’s Follow-up Review on the October 1999 Report Entitled: ‘Review of the Investigation of Sexual Assaults—Toronto Police Service’ (Toronto, Audit Services, 2004). 52 Toronto Police Services Board, Minutes of Public Meeting (21 May 2008) at www.tpsb.ca/FS/ Docs/Minutes/2008, accessed on 1 November 2008. 53 See A Zerbisias, ‘Police Report on Rape Fails All of Our Jane Does’, The Toronto Star (21 May 2008), which criticises the city for disbanding the Steering Committee. The article quotes Jane Doe as saying: ‘Things are worse than they have ever been, worse because we know better now’.

292

Erika Chamberlain

assault,54 and it is doubtful that the ongoing reviews of police procedures would exist but for her pivotal litigation and continued advocacy.

B. The Ombudsman Function As Jane Doe illustrates, tort law can be an effective means to expose and potentially reform the conduct of police, especially in an area where other attempts at reform have been inadequate.55 There has been little motivation from within police forces to improve the treatment of sexual assault victims and increase the chances of conviction. While there had been various internal and external reviews prior to Jane Doe,56 almost universally critical of the system, the evidence suggests that the investigation of sexual assaults was still dire. If nothing else, Jane Doe brought renewed attention to the situation and showed the public that the problem had not been solved. Moreover, the threat of a substantial damages award may prove to be better motivation to improve police practices than any government inquiry or policy paper. It is unfortunate that the claim in Hill v Chief Constable v West Yorkshire was struck out, as it might have exposed some of the systemic problems, including poor organisation and communication, which prevented the more timely apprehension of the Yorkshire Ripper.57 For instance, while Sutcliffe was interviewed nine times by various officers before his arrest, four of the interviewers believed they were questioning him for the first time.58 What is perhaps more troubling is how the investigation was skewed by the police perception that the Ripper targeted prostitutes and other women of ‘loose morals’. This led them to exclude certain victims from the list of crimes attributed to him (as they did not meet their profile), to erroneously include some victims in that list on account of their dubious lifestyles, and to act on the basis of letters that turned out to be hoaxes. While some of this information came to light during Sutcliffe’s criminal trial, it never became the focus of attention in the civil suit, which was struck out at a preliminary stage. This meant that police attitudes and procedures never received the judicial or public scrutiny that might have prompted reform. Moreover, because Hill granted police near immunity for negligent investigation, there was no financial incentive to make changes to the system. Hill was, therefore, a missed opportunity to use negligent investigation as a means of inspiring institutional change. The ombudsman function is not so much concerned with compensating victims as with providing an incentive for police to improve their standards. This

54 Perhaps ironically, Jane Doe remains the ‘face’ of sexual assault in Toronto. See eg, Zerbisias, ibid; and M Henry, ‘ “Nothing’s Going to Change”: Jane Doe’ The Toronto Star (22 May 2008). 55 See generally J Gregory and S Lees, Policing Sexual Assault (London, Routledge, 1999). 56 See above n 44. 57 For a more detailed analysis, see J Smith, Misogynies (London, Faber and Faber, 1989) 163 ff. 58 ibid, at 164–5.

Tort Law as Police Ombudsman

293

was recognised in 1968 by the very progressive dissenting opinion of Keating J in the major American case of Riss v New York.59 The plaintiff had repeatedly complained to police that she was being harassed by a rejected suitor, who threatened: ‘If I can’t have you, no one else will have you, and when I get through with you, no one else will want you’. The plaintiff ’s pleas for protection went largely unheeded by police. Ultimately, the man hired a thug to throw lye (caustic soda) in the plaintiff ’s face, leaving her blinded and permanently scarred. Her tort claim against the police was rejected by a majority of the New York Court of Appeals, on the basis that there was no tortious duty to protect. The majority relied on familiar policy concerns, such as maintaining police discretion for the allocation of resources and avoiding the potential for indeterminate liability. Justice Keating analysed each of the policy concerns and found them unpersuasive. However, his opinion is most interesting as an illustration of the ombudsman function of tort law. Keating J found that the facts in Riss were a wake-up call regarding the lack or use of police resources in the city. He wrote, If the police force of the City of New York is so understaffed that it is unable to cope with the everyday problem posed by the relatively few cases where single, known individuals threaten the lives of other persons, then indeed we have reached the danger line and the lives of all of us are in peril. If the police department is in such a deplorable state that the city, because of insufficient manpower, is truly unable to protect persons in Linda Riss’ position, then liability not only should, but must be imposed. It will act as an effective inducement for public officials to provide at least a minimally adequate number of police. If local officials are not willing to meet even such a low standard, I see no reason for the courts to abet such irresponsibility.60

It was not only in the plaintiff ’s interest, but the public interest, to impose liability. The financial threat of tort liability could force the city to assign greater resources to police operations. Police liability for negligent investigation is, at bottom, a question of accountability. Police are given a great deal of power over citizens in modern society. In return, the public expects a minimum degree of protection. If there is no effective accountability through complaints or disciplinary processes, then perhaps financial accountability through tort actions is the only way to stimulate reform. Indeed, Keating J suggested that the tort action should also have the normative function of encouraging certain standards of practice. He noted that the imposition of liability on other government authorities had had ‘healthy side effects’.61 Tort actions do not simply compensate a victim or punish a wrongdoer; they encourage the wrongdoer to amend its conduct and avoid future liability. But, Keating J observed, as long as police are immune from liability in tort, they would be completely untouched by this ‘salutary control’.62 In other words, tort liability 59 60 61 62

Riss v New York 22 NY 2d 579 (New York CA). ibid, at 587. ibid, at 590. Riss v New York 22 NY 2d 579 (New York CA) 590.

294

Erika Chamberlain

could serve as a sort of quality-control mechanism through which problems in police practices could be identified and corrected. Without civil claims, some of those problems might remain hidden from scrutiny.

C. Potential Use by Victims of Domestic Violence Claims for negligent investigation may help to provide a voice for victims of domestic abuse. Evidence from the last several decades indicates that police do not take domestic violence as seriously as other violent crimes: they have been slow and often ineffective in their response to complaints; they arrest perpetrators less frequently than for non-domestic assaults; and they even blame victims for their own misfortune.63 While these problems have been exposed for some time, substantive reform has been sluggish. It apparently remains the case that police are hesitant to ‘interfere’ in domestic assault situations, leaving victims to fend for themselves. This hesitance even occurs in situations where women have obtained restraining or protective orders against their abusers.64 In the United States, some women have been successful in suits against the police for failing to provide equal protection of the law.65 These suits, brought under Title 42 § 1983 of the United States Civil Code, require the claimants to show either that police treat women victims of violence differently from men, or that they treat domestic assault differently from other crimes. Such equal protection claims, while instructive, do not have a direct equivalent in most Commonwealth jurisdictions, and their relevance for negligence claims is minimal. However, it is notable that, in addition to her negligence claim, Jane Doe brought a successful claim against the Toronto Police for violating her rights under section 15 of the Canadian Charter of Rights and Freedoms,66 which similarly guarantees equal protection and benefit of the law.67 Such resort to constitutional provisions, nevertheless, remains rare, and it is more likely that victims of domestic violence in the Commonwealth will resort to tort law if police are negligent in investigating complaints and apprehending

63 See generally, BS Blackwell and MS Vaughn, ‘Police Civil Liability for Inappropriate Response to Domestic Assault Victims’ (2003) 31 Journal of Criminal Justice 129. See also DG Saunders, ‘The Tendency to Arrest Victims of Domestic Violence: A Preliminary Analysis of Officer Characteristics’ (1995) 10 Journal of Interpersonal Violence 147. 64 See GS Rigakos, ‘Situational Determinants of Police Responses to Civil and Criminal Injunctions for Battered Women’ (1997) 3 Violence against Women 204; and RJ Kane, ‘Police Responses to Restraining Orders in Domestic Violence Incidents: Identifying the Custody-Threshold Thesis’ (2000) 27 Criminal Justice & Behavior 561. See also DL Lordi, ‘Police Liability under State Tort Law for Failure to Enforce Protection Orders: The Last Demand for Accountability’ (2006) 85 Oregon Law Review 325. 65 See the summary of such suits in Blackwell and Vaughn, above n 63. 66 Pt I of the Constitution Act, 1982, being Sch B to the Canada Act 1982 (UK) (the Charter). 67 Jane Doe did not receive any additional damages for this violation, as they were seen to duplicate her damages in negligence.

Tort Law as Police Ombudsman

295

perpetrators.68 These claims seem to be ideal candidates for finding a duty of care: there is often a pattern of violence, so harm is reasonably foreseeable; and both the victim and perpetrator are identifiable, so there is no threat of indeterminate liability. Further, if a protective order is in place, police have a pre-existing duty to take action. There is little room for discretion, and no obvious conflict with other police duties. There is some authority, admittedly limited, to suggest that police might owe a duty of care to victims of domestic violence. In Batchelor v Tasmania,69 the plaintiff sued police after his father killed his mother and committed suicide. On the day of the killings, the plaintiff ’s mother complained to police that her husband had seriously assaulted her, intended to kill her, and owned firearms. She requested a restraining order and police took steps to help her obtain one. While she was at the police station, the father arrived. Police told him that they were going to the family home to seize his firearms, and they escorted the mother there to get her belongings. When they arrived at the residence, they noticed that one of the father’s firearms was missing, but they left the mother alone inside. The father approached from outside and shot the mother dead through a window. The father then killed himself. The plaintiff sued police for the loss of his parents and his psychiatric harm. He claimed that the police were negligent in not arresting his father (contrary to the pro-arrest policy then in force), in telling his father that they were going to the family residence to seize his firearms, and in failing to evacuate the home when they realised that one of the guns was missing. Blow J was cautious in analysing the duty of care, noting that it was rare to impose a duty to protect someone from the criminal conduct of a third party.70 Nevertheless, given the conflicting authorities, including Hill and Jane Doe, Blow J refused to strike out the plaintiff ’s cause of action for the death of his mother.71 Blow J summarised the various indicia which might give rise to a duty to protect the plaintiff from criminal acts. First, there may be a duty where the criminal actions are predictable and foreseeable, as when they have been threatened or have occurred in the past.72 This is often the case with domestic violence. Secondly, there may be a duty when the plaintiff is especially vulnerable. Again, victims of domestic violence are often vulnerable and powerless to change their situation, apart from through appeals for police protection. Finally, Blow J found that there may be a duty if the defendants have assumed responsibility for the situation. In Batchelor, the police had taken control of the situation and, arguably, made it 68 Notably, the House of Lords rejected a claim against the police under the Human Rights Act 1998 in Chief Constable of Hertfordshire Police v Van Colle [2008] UKHL 50, where the plaintiffs’ son was murdered after assisting police with a criminal investigation into theft. 69 Batchelor v Tasmania [2005] TASSC 11 (Supreme Court of Tasmania). 70 ibid, at [12]. 71 The plaintiff ’s claim for the death of his father was struck out, as there was no duty of care to prevent the father from committing suicide. 72 Batchelor v Tasmania [2005] TASSC 11 (Supreme Court of Tasmania) [13], citing Gleeson CJ in Modbury Triangle Shopping Centre v Anzil (2000) 205 CLR 254 (High Court of Australia).

296

Erika Chamberlain

worse by informing the father that they were travelling to his residence to seize his firearms. Blow J therefore concluded that the police might have owed the mother a duty of care. Batchelor is useful for illustrating that a duty to take reasonable care in investigating crime and apprehending offenders need not raise the spectre of indeterminate liability, nor challenge the allocation of police resources and the prioritisation of tasks. The victim and perpetrator were clearly identifiable, and there was evidence of a pattern of violence. Police had already made the decision to assist the mother. However, they were arguably negligent in the way they carried out the investigation, particularly by putting the father on notice and leaving the mother alone in the home. Batchelor is, more or less, a straightforward case of asking what a reasonable officer would do when faced with a known violent man and a vulnerable woman who seeks protection. It is not a delicate balancing of interests, resources or priorities, and there is, accordingly, no justifiable reason for police to be immune if their actions caused the plaintiff ’s loss. Of course, this assumes that a plaintiff in such a case can prove causation, which may be a daunting task, as illustrated by BM v British Columbia (Attorney General).73 The plaintiff, Bonnie Mooney, was in the process of separating from Roland Kruska. Kruska had a history of violence, including criminal convictions for assault causing bodily harm, sexual assault, manslaughter and unlawful confinement, and had been convicted previously for brutally assaulting Mooney with a cane. On one occasion, Kruska pursued Mooney in his vehicle. She complained to police, but the officer assigned to her case said that he could do nothing for her. He advised her to apply for a restraining order and to stay in public places. These actions were contrary to the proactive arrest policy then applicable. About six weeks later, Kruska broke into Mooney’s home, shot and killed her friend, and injured her daughter. After making a formal complaint about the officer’s negligence in her case (which was upheld after an internal investigation), Mooney sued the police for the psychological damage that she suffered. The trial judge (Collver J) and at least one judge at the Court of Appeal found that police owed Mooney a duty of care.74 Donald JA distinguished BM from Hill v Chief Constable of West Yorkshire, since Mooney had ‘direct engagement’ with police and was not a member of an indeterminate class.75 Moreover, the proactive arrest policy meant that there was little room for police discretion in their handling of domestic assault. The policy explicitly recognised that domestic assault victims are at risk and should be given priority.76 Nevertheless, Collver J 73

BM v British Columbia (Attorney General) (2004) 31 BCLR (4th) 61 (British Columbia CA) (BM). Donald JA was the only appellate judge to consider the duty of care in any detail. Hall JA avoided the issue and found that the case was more appropriately determined on the issue of causation. 75 BM (2004) 31 BCLR (4th) 61 (British Columbia CA) [46]. 76 ibid, at [50] (Donald JA). Donald JA dissented on the issue of causation, and would have granted recovery to the plaintiff. 74

Tort Law as Police Ombudsman

297

and a majority of the Court of Appeal dismissed Mooney’s case for lack of causation. Given that criminal behaviour is unpredictable, and that Kruska seemed undeterred by past convictions, the plaintiff had not sufficiently established that more effective police action would have prevented Kruska’s fatal attack.77 As the trial judge summarised, ‘police are guardians, not guarantors, of public wellbeing’.78 The result in BM suggests that claims for negligent investigation in domestic assault cases may be difficult to establish under the current causation framework.79 However, BM is critical for recognising that police might owe a duty of care to victims of domestic violence who claim to be threatened and seek assistance. Especially in situations involving an established pattern of violence, the harm is foreseeable and there is proximity between police and the victim. There is no risk of indeterminate liability, because the victim is distinct and identifiable. Moreover, BM suggests that policy concerns regarding police discretion will not be sufficient to negate the duty of care, particularly if official policies impose a de facto duty to take action on complaints of domestic violence. This last point is important if tort claims are to serve an ombudsman function regarding police treatment of domestic violence. Many of the pro-arrest policies currently in force were a response to the historical tendency of police to downplay the severity of domestic violence. Essentially, such policies remove the discretion that allowed police to turn a blind eye to domestic abuse. Thus, it would be incongruous if police were able to avoid a duty of care on the basis that they require discretion in their prioritisation of complaints. BM helps to reaffirm that police must take action in cases of domestic assault, and cannot use their discretion to justify their historical preference for non-interference. Indeed, BM complements proactive arrest policies by providing a sort of external enforcement mechanism. Such policies, while laudable, often lack effective consequences for their breach. While Mooney received an apology and a promise of ‘corrective action’ as a result of her complaint to police, this was perhaps little consolation for the death of her friend, injury to her daughter, and lifelong psychological consequences.80 Moreover, the facts in BM attest that internal policies, alone, are not sufficient to change behaviour. As one author has written, ‘Policy

77 The majority refused to adopt a modified standard of causation that might have assisted the plaintiff in proving her case. For a more detailed analysis of the causation issue, see MI Hall, ‘Duty, Causation and Third-Party Perpetrators: The Bonnie Mooney Case’ (2005) 50 McGill Law Journal 597. 78 BM [2001] BCSC 419 (British Columbia Supreme Court) [64] (Collver J). 79 It has been argued that Mooney’s claim might have been more successful if framed in terms of the equality rights under s 15 of the Charter, as Jane Doe had been: see EA Sheehy, ‘Causation, Common Sense and the Common Law: Replacing Unexamined Assumptions with What We Know about Male Violence against Women or from Jane Doe to Bonnie Mooney’ (2005) 17 Canadian Journal of Women & the Law 87. In particular, it might have been argued that the officer’s failure to act was motivated by erroneous and stereotypical attitudes regarding abused women. 80 The officer involved was not apparently disciplined for his inaction, but received ‘operational guidance’: see Sheehy, ibid, at 103.

298

Erika Chamberlain

guides; liability deters’.81 The apparent lack of improvement in police attitudes suggests that more severe and more public reprimands are necessary. Donald JA remarked in dissent that ‘the right to police protection in these circumstances is so strong and the need for teeth in the domestic violence policy so great’ that recovery is warranted.82 BM puts police on notice that their internal policies may have consequences beyond professional discipline, and gives them a financial incentive to see that their policies are followed.83 Of course, recovery in tort does not adequately compensate for the sometimes lethal consequences of police inaction in these cases. However, it does provide victims with an additional avenue of redress, and one that will attract public scrutiny in a way that internal investigations and private letters of apology do not. If nothing else, negligence claims against police in domestic violence cases attract public attention, stressing that the problem has not disappeared. They may help to shake the sense of complacency and the illusion of progress that internal policies might engender. Finally, in Justice Linden’s terminology, tort actions provide psychological empowerment to victims, to tell how they were not only abused by a spouse, but neglected by police. Unfortunately, given the House of Lords’ recent reaffirmation of Hill in Smith, it is unlikely that negligence claims will be available for victims of domestic violence in the United Kingdom. The majority in Smith indicated that the risk of defensive policing outweighed any claims for compensation by victims. More troubling, however, is the language used by several of the Lords in describing domestic abuse. For instance, Lord Hope commented, ‘Not every complaint [of domestic abuse] is genuine’, such that police needed freedom to determine which complaints to pursue.84 Further, Lord Carswell explained: One must recognise that police officers may quite properly be slow to engage themselves too closely in such domestic type matters, where they may suspect from experience the existence of a degree of hysteria or exaggeration on the part of either or both persons involved.85

Such comments suggest that the House of Lords shares the police’s preference for non-intervention in cases of domestic abuse, and their patently stereotypical attitudes about complainants. Thus, there is little hope that tort law will be used in the United Kingdom to compensate or empower those abused by partners and ignored by police.

81

Hall, above n 77, at 615. BM (2004) 31 BCLR (4th) 61 (British Columbia CA) [12]. 83 The author is more optimistic about the potential for claims against the police than some other writers. For example, Sheehy, above n 79, believes that BM has closed the door on such claims, and has undone any progress that was made by Jane Doe. 84 Smith [2008] UKHL 50 [76]. 85 ibid, at [107]. 82

Tort Law as Police Ombudsman

299

III. Negligence Claims by the Wrongly Accused A. The General Commonwealth Position The Commonwealth jurisprudence on the duty of care owed by police to a suspect under investigation has drawn heavily on the policy considerations discussed in Hill v Chief Constable of West Yorkshire.86 The courts tend to focus on the discretionary nature of police investigations and the threat of over-deterrence. Additionally, there is concern that a duty of care would hinder police in their vigorous pursuit of all lines of investigation. These concerns are typically seen as sufficient to negative any prima facie duty of care that is owed toward suspects, and are usually accepted without supporting evidence. For instance, in ElguzouliDaf v Commissioner of Police of the Metropolis,87 Steyn LJ seemed to accept without question that it would be detrimental for the Crown Prosecution Service (CPS) to owe a duty toward suspects: In my view such a duty of care would tend to have an inhibiting effect on the discharge by the CPS of its central function of prosecuting crime. It would in some cases lead to a defensive approach by prosecutors to their multifarious duties. It would introduce a risk that prosecutors would act so as to protect themselves from claims of negligence. The CPS would have to spend valuable time and use scarce resources in order to prevent law suits in negligence against the CPS. It would generate a great deal of paper to guard against the risks of law suits. The time and energy of CPS lawyers would be diverted from concentrating on their prime function of prosecuting offenders. That would be likely to happen not only during the prosecution process but also when the CPS is sued in negligence by aggrieved defendants. The CPS would be constantly enmeshed in an avalanche of interlocutory civil proceedings and civil trials.88

While there may well be good reason not to impose a duty of care on the CPS, one would expect that such de facto immunity should not be granted without more conclusive evidence of the alleged consequences. As discussed in Section III.B, it is a credit to the Supreme Court of Canada that they refused to accept police policy arguments that were speculative and unsubstantiated. Their Commonwealth counterparts, by contrast, have largely accepted police assertions as fact. For instance, in Tame v New South Wales,89 the various members of the High Court of Australia concluded straightforwardly that police do not owe a duty of care to those under investigation, while using language suggesting that their conclusions were based on mere speculation.90 Gaudron J asserted,

86 87 88 89 90

Hill v Chief Constable of West Yorkshire [1989] 1 AC 53 (HL). Elguzouli-Daf v Commissioner of Police of the Metropolis [1995] QB 335 (CA). ibid, at 349. Tame v New South Wales (2002) 211 CLR 540 (Tame). See also Wilson v New South Wales [2001] NSWSC 869 (Supreme Court of New South Wales).

300

Erika Chamberlain

It would be incongruous and, perhaps, give rise to incompatible duties if a person charged with the investigation of a possible offence were to owe a duty of care to the person whose conduct is the subject of that investigation.91 (emphasis added)

Similarly, Gummow and Kirby JJ supposed, without further analysis: It is unlikely that an investigating police officer owes a duty of care to a person whose conduct is under investigation. Such a duty would appear to be inconsistent with the police officer’s duty, ultimately based in the statutory framework and anterior common law by which the relevant police service is established and maintained, fully to investigate the conduct in question.92 (emphasis added)

While Mrs Tame’s claim was admittedly weak,93 it is noteworthy that nearly blanket immunity was granted to police on the basis of no evidence and minimal analysis. Similarly, the 2004 decision of the New South Wales Court of Appeal in Cran v New South Wales94 suggests that the policy principles outlined in Hill v Chief Constable of West Yorkshire remain highly persuasive in the Australian courts. The plaintiff was in custody and could have been released if the authorities had properly filled out forms to expedite an analysis of a substance they thought to be LSD, but which turned out to be paper stickers. He was clearly vulnerable to police negligence, and the timely completion of the forms presented no obvious conflict of duties for police. Yet, Santow JA reasoned: Even ministerial tasks involve decisions as to priorities in the deployment of resources … If the police were for fear of civil action to employ additional resources to check that forms were properly filled out … that may well involve an allocation of resources away from active police enforcement to administration.95

Thus, police have been able to achieve immunity for their negligent conduct during investigations based on the discretion inherent in resource allocation. As discussed below, it is strange that police have been so successful with this argument, when comparable professions have not. For example, it would be shocking if medical professionals were granted immunity when their negligent completion of paperwork resulted in injury to a patient. Yet, medical professionals seem to have an equally plausible argument that paperwork detracts resources from front-line patient care. It is puzzling that police should maintain such widespread preferential treatment for so long.96 91

Tame (2002) 211 CLR 540 [57]. See also [292] (Hayne J); and [126] (McHugh J). ibid, at [231]. 93 Mrs Tame suffered psychiatric injury when she learned, several years after a motor vehicle crash for which she successfully sued the other driver, that an officer on the scene had erroneously recorded her blood alcohol concentration as 0.14%, when it was in fact 0.00%. The error had been corrected almost immediately and had not affected her in any way, but she became obsessed with the possibility that it might have been communicated to others and damaged her reputation. 94 Cran v New South Wales (2004) 62 NSWLR 95. 95 ibid, at [48]. 96 There are occasional detractors. See eg, Eichelbaum CJ in Whithair v Attorney General [1996] 2 NZLR 45 (New Zealand High Court). 92

Tort Law as Police Ombudsman

301

B. Hill v Hamilton-Wentworth Regional Police Services Board97 In Hill, the Supreme Court of Canada became the highest court in the Commonwealth to recognise that police owe a duty of care to suspects under investigation. While the court ultimately dismissed the plaintiff ’s claim because police had met the standard of care, Hill is important for its evaluation of the various policy factors that are normally used to negative any prima facie duty. McLachlin CJ, for the majority, was careful to stress that the duty should not be negated on policy grounds that were speculative, unsubstantiated, and did not apply in the particular case. Her decision is laudable because it cuts through the ‘thin blue line’ and serves notice that police will be held accountable in negligence when their carelessness leads to wrongful charge or conviction. The plaintiff was tried and convicted of robbery in the mid-1990s. The evidence against him consisted primarily of eye-witness identification. Although he was eventually acquitted on appeal due to errors made by the trial judge, he had spent 20 months in jail. He sued, inter alia, the police for their alleged negligence in carrying out the investigation. The allegations of negligence included: that police had published a photo identifying Hill as a suspect early in the investigation, thereby tainting subsequent eye-witness identifications; and that police had failed to thoroughly investigate information suggesting that other persons might have committed the robberies. Perhaps most egregiously, police had conducted a photo line-up using Hill (an Aboriginal Canadian) and 11 Caucasian foils. It is noteworthy that both eye-witness identification procedures and police ‘tunnel vision’ have been identified as primary causes of wrongful conviction.98 Thus, Hill was a textbook case of how negligent police practices can lead to wrongful conviction, and it was not surprising that the court was sensitive to the political importance of its decision. This importance was heightened by the racial undertones of the case: Aboriginal Canadians are incarcerated at a disproportionately high rate,99 and there is literature suggesting that they are treated differently by police than Caucasian suspects.100 Hill thus had the potential to raise awareness about police investigatory procedures generally, and also about any systemic prejudices toward Aboriginals.

97

Hill v Hamilton-Wentworth Regional Police Services Board [2007] SCC 41, 285 DLR (4th) 620. See B Macfarlane, ‘Convicting the Innocent: A Triple Failure of the Justice System’ (2006) 31 Manitoba Law Journal 40, for a summary of international inquiries into wrongful conviction over the past century. The discussion of eye-witness misidentification can be found starting at para 147. 99 See Aboriginal Peoples and the Criminal Justice System (Ottawa, Canadian Criminal Justice Association, 2000); and J-A Brzozowski, A Taylor-Butts and S Johnson, ‘Victimization and Offending among the Aboriginal Population in Canada’ (2006) 26(3) Juristat (Canadian Centre for Justice Statistics). 100 A summary of the research can be found in JV Roberts and AN Doob, ‘Race, Ethnicity, and Criminal Justice in Canada’ (1997) 21 Crime & Justice 469. 98

302

Erika Chamberlain

Hill being a novel case, the duty of care was analysed in terms of the Anns test, as restated by the Supreme Court in Cooper v Hobart101 and Childs v Desormeaux.102 While the majority’s analysis of the ‘proximity’ requirement was rather disappointing,103 its analysis of so-called ‘residual’ policy considerations under stage two of the Anns/Cooper test was not. McLachlin CJ proceeded through the various arguments raised by the police and found that each was either unsubstantiated and/or unpersuasive. Inter alia, the police had argued that the duty of care should be negated because the conduct of investigations was discretionary, and because a duty of care would lead to defensive policing and a flood of litigation.

(i) The Discretionary Nature of Police Investigations While McLachlin CJ acknowledged that criminal investigations require the exercise of discretion, she found that this did not distinguish police from other professionals.104 For example, medicine also involves the exercise of ‘discretion, intuition and occasionally hunch’, but this does not render doctors immune in negligence.105 Also, given the demise of barristers’ immunity throughout much of the Commonwealth,106 it would be difficult to justify the continued immunity of police, alone, among the professions. McLachlin CJ further reasoned that the discretion exercised by police was hardly unrestrained: there are various statutory, constitutional and common law obligations that direct how police can exercise their discretion in a democratic society. To hold them to a standard of reasonableness is merely to affirm their professional standards. Indeed, if it had been found that police exercised their discretion in a discriminatory manner on account of Hill’s ethnicity, few would argue that this had been a valid exercise of their discretion. McLachlin CJ rightly concluded that the discretionary nature of police functions should not serve to exclude them entirely from scrutiny in negligence. Any concerns police had about the exercise of discretion, the majority found, could be addressed by reference to the standard of care. The standard adopted in Hill v Hamilton-Wentworth, that of a reasonable police officer in the circumstances, was highly deferential and took into account the discretion inherent in many police functions.107 It recognised that police sometimes have to make decisions based on less-than-conclusive information, hearsay or hunch. At the same time, the standard of care must account for the potentially serious harms that might flow from the negligent conduct of investigations: the arrest and imprisonment of 101

Cooper v Hobart (2001) 206 DLR (4th) 193 (Supreme Court of Canada). Childs v Desormeaux (2006) 266 DLR (4th) 267 (Supreme Court of Canada). 103 See E Chamberlain, ‘Negligent Investigation: The End of Malicious Prosecution in Canada?’ (2008) 124 Law Quarterly Review 205. 104 Hill v Hamilton-Wentworth [2007] SCC 41, 285 DLR (4th) 620 [51]–[54]. 105 ibid, at [53]. 106 JS Hall and Co v Simons [2003] 3 WLR 543 (HL); Chamberlains v Lai [2006] NZSC 70 (Supreme Court of New Zealand). Canadian barristers have been subject to negligence claims for considerably longer: Leslie v Ball (1863) 22 UCQB 522; Wade v Ball (1870) 20 UCCP 302 (Upper Canada Court of Common Pleas); and Demarco v Ungaro (1979) 95 DLR (3d) 385 (Ontario High Court). 107 Hill v Hamilton-Wentworth [2007] SCC 41, 285 DLR (4th) 620 [73]. 102

Tort Law as Police Ombudsman

303

innocent persons.108 While police should be given some leeway to pursue investigations as they see fit, the potential consequences for the suspect demand that the police act reasonably in the circumstances. In Hill, the police were absolved of liability based on this relatively generous standard of care. The majority prefaced its analysis by noting that police practices have improved since 1995, and that the defendants’ conduct should be judged by the standards prevailing at that time.109 With respect to the photo line-up, the majority accepted the trial judge’s finding that there were no rules regarding line-ups in 1995. While a modern-day police officer would likely use foils of the same race as the suspect, the defendants’ conduct was not unreasonable in its day.110 This conclusion will sit uneasily with some observers: it seems a matter of common sense that the foils in a line-up should be the same race as the suspect. Nevertheless, some of the foils had similar skin tones and facial features to Hill, so his race did not make him obviously stand out. On the facts, the trial judge and the majorities of the Court of Appeal and Supreme Court all concluded that the racial composition of the line-up did not result in unfairness. The other main allegation of negligence was that police down-played evidence that another perpetrator might have been responsible for the crimes. They arrested and charged Hill even though they had tips implicating other men, and then proceeded with the case against him even though the robberies continued while he was in custody. Although another suspect was identified and charged with some of the robberies, the lead detective did not seek to delay Hill’s trial to permit further investigation. McLachlin CJ conceded: Had Detective Loft conducted further investigation, it is likely that the case against Hill would have collapsed. Had he re-interviewed the eyewitnesses, for example, and shown them [the other suspect’s] photo, it is probable that matters would have turned out otherwise; when the witnesses were eventually shown the photo of [the other suspect], they recanted their identification of Hill as the robber.111

In spite of this observation, McLachlin CJ found that the detective’s conduct fell within the acceptable range of police discretion.112 She stressed that, in 1995, ‘awareness of the danger of wrongful convictions was less acute than it is today’.113 Thus, the majority concluded that the detective met the standard of a reasonable officer in the circumstances. The Supreme Court’s analysis of this issue indicates that other Commonwealth courts have been overly cautious in granting police near blanket immunity for their conduct of investigations. There is nothing about the discretionary nature of police activities that suggests they should receive greater protection than lawyers, 108 109 110 111 112 113

ibid, at [72]. ibid, at [77]. ibid, at [80]. ibid, at [83]. ibid, at [88]. ibid.

304

Erika Chamberlain

doctors, or other professionals. As long as the courts maintain the rather deferential standard adopted in Hill, there is no reason for police to fear that their professional judgement will be second-guessed by the courts. Hill suggests that police will only be found liable in the most obvious cases of negligence.

(ii) The Threat of Defensive Policing The defendants in Hill also raised the argument, often successful in Commonwealth courts, that imposing a duty of care on police would force them to carry out their investigations in a defensive manner. Also known as the ‘chilling effect’, this argument insinuates that police will be fearful of vigorously pursuing all lines of investigation, lest the suspects bring civil claims after the fact. Charron J’s dissenting opinion suggested that a duty toward suspects would discourage police from laying charges ‘except in cases where the evidence is overwhelming’.114 However, the majority found that this argument was speculative, and that the record did not support the assertion that potential tort liability would alter police behaviour. Instead, the majority cited literature indicating the tort liability had no chilling effect.115 Further, the majority reiterated that the standard was only that of reasonableness. While McLachlin CJ acknowledged that ‘police might become more careful in conducting investigations if a duty of care in tort is recognized’, she concluded that ‘this is not necessarily a bad thing’.116 It is hard to argue with the outcome of discouraging unreasonable conduct.

(iii) Floodgates Finally, the police in Hill resorted to the argument that recognising a duty of care toward suspects would result in a flood of litigation. Even worse, they argued, law suits could be brought by factually guilty parties who managed to avoid conviction on a technicality.117 Unlike many Commonwealth courts, which have accepted this argument without supporting evidence, the majority of the Supreme Court concluded that the threat of floodgates was not credibly substantiated. McLachlin CJ noted that both Ontario and Quebec had already recognised claims for negligent investigation, and neither province had experienced a glut of litigation. She also seemed critical of the defendants for raising this argument in the abstract, without supporting data. Pointing to ‘a relatively small number of lawsuits, the cost of which are unknown, with effects on the police that have not been measured’, was not sufficient to negate the prima facie duty of care.118 Regarding the possibility of recovery by factually guilty parties, McLachlin CJ concluded that the chance of such ‘injustice’ was not any greater than in other 114

ibid, at [152]. ibid, at [57]. 116 ibid, at [56]. 117 The dissenting judges (Bastarache, Charron and Rothstein JJ) found this argument persuasive, noting that it would be difficult to determine which plaintiffs should rightfully recover without introducing a verdict of ‘factually innocent’ into Canada’s criminal justice system. 118 Hill v Hamilton-Wentworth [2007] SCC 41, 285 DLR (4th) 620 [61]. 115

Tort Law as Police Ombudsman

305

tort actions. The tort system is not perfect. For example, a ‘person who recovers against her doctor for medical malpractice may, despite having proved illness in court, have in fact been malingering’.119 No one would suggest, however, that such a possibility should lead to blanket immunity for medical practitioners. Further, McLachlin CJ felt that the possibility of erroneous awards would be minimised by the other safeguards in the tort process, noting especially the difficulties of proving causation in negligent investigation cases (ie, that the suspect would not have been charged at all but for the police negligence). It is heartening that the Supreme Court challenged the police’s rather bald argument regarding the floodgates of litigation. The courts have too easily accepted this argument in the past, on the assumption that any suspect who is exonerated will subsequently bring a civil claim against the police. This assumption ignores that most suspected criminals lack the financial resources to pursue lengthy litigation against the state, and that most are probably relieved to be rid of the justice system once found not guilty. Given the deferential standard of care adopted in Hill, it seems that only the most obvious cases of negligence will succeed in any event.

C. The Ombudsman Function As with claims for negligent investigation brought by crime victims, claims by criminal suspects may serve an ombudsman function by bringing faulty police practices to public attention. In Hill, for example, there was renewed analysis of police identification procedures and tunnel-vision, as well as some indirect criticism of police treatment of racial minorities. Some of these practices have already been disparaged by various official inquiries into wrongful conviction;120 thus, the need for scrutiny in the civil courts may not be as great as in the case of domestic assault. Nevertheless, it is reasonable to assume that a greater amount of scrutiny will place additional pressure on police to improve practices. Moreover, inquiries and reviews of conviction are rare, and typically only occur in what are considered grave miscarriages of justice. In Canada, for instance, the ability to review convictions is strictly limited.121 The official inquiries involve notorious cases of wrongful conviction, with accused persons who have spent many years in prison. There have only been a handful of such cases in recent decades. Conversely, where someone has been arrested and charged, but ultimately acquitted, there is little opportunity to bring public attention to any negligent 119

ibid, at [63]. For instance, P Cory, The Inquiry Regarding Thomas Sophonow: The Investigation, Prosecution and Consideration of Entitlement to Compensation (Winnipeg, Manitoba Justice, 2001); and F Kaufman, The Commission on Proceedings Involving Guy Paul Morin: Report (Toronto, Ontario Ministry of the Attorney General, 1998). 121 Section 696.1 of the Criminal Code, RSC 1985, c C-46, allows convicted offenders to petition for ministerial review on the grounds of miscarriage of justice. Such applications are not intended as further appeals but are ‘extraordinary’ remedies. 120

306

Erika Chamberlain

police practices. The suspect is (rightly) released, but is not compensated for the anxiety, stress, damage to reputation, financial loss, and family disruption that he has suffered. Unless the police or prosecutors acted with malice, the exonerated suspect has no claim for damages under the traditional law of the Commonwealth. The majority in Hill accurately observed that, without a claim for negligent investigation, ‘an important category of police conduct will go unremedied’.122 The negligence claim is necessary ‘to complete the arsenal of already existing common law and statutory remedies’.123 This argument is critically important to the ombudsman function of tort law. Obviously, police conduct that is malicious should not go unremedied; however, there is a much broader category of negligent conduct that should not be resistant to scrutiny. The literature indicates that most wrongful convictions do not arise out of deliberate police wrongdoing, but out of substandard practices, poor identification techniques, and inattention to alternative lines of inquiry.124 While it is useful to criticise these practices in official inquiries, experience suggests that more direct consequences are necessary to achieve substantive reform. Perhaps the best evidence of this, at least in Canada, is the exclusion of evidence that is obtained through breach of the accused’s Charter rights.125 The potential that evidence will be excluded and, in many cases, that the accused will be acquitted, has motivated police to improve the way that they search and question suspects, provide them with access to counsel, and inform them of their rights.126 It is doubtful whether these improvements would have been made without the prospect of losing convictions. However, not all substandard investigatory practices are breaches of Charter rights; they are simply negligent practices that may lead to the wrongful charge or conviction of the suspect. In these cases, some other mechanism is required to motivate police to improve their practices. Through the action for negligent investigation, the wrongfully accused may bring public attention to negligent investigatory techniques and provide a financial incentive to make improvements. One of the arguments raised against negligent investigation, in Hill and elsewhere, is that it will allow the courts to second-guess the exercise of discretion by police in the course of an investigation. As discussed above, this fear seems exaggerated. The standard of care adopted in Hill still allows police a great deal of flexibility in the way that they conduct investigations. Hill suggests that the courts will be relatively deferential to police expertise, and it is unlikely that courts will criticise the way that police pursue various lines of inquiry. Instead, the claim for negligent investigation will likely be most useful in cases where police have 122

Hill v Hamilton-Wentworth [2007] SCC 41, 285 DLR (4th) 620 [35]. ibid. 124 See the rather comprehensive review article by Macfarlane, above n 98. 125 Section 24(2) of the Charter allows a court to exclude evidence gained through breach of the accused’s Charter rights, if its admission would bring the administration of justice into disrepute. 126 The jurisprudence on these issues is vast, but see especially R v Collins [1987] 1 SCR 265; and R v Manninen [1987] 1 SCR 1233. 123

Tort Law as Police Ombudsman

307

adopted practices that have become discredited or have been shown to contribute to wrongful convictions. Police might also be found liable for negligent actions that are clearly ‘operational’ in character, for instance, failing to keep adequate notes, misplacing evidence, or poor communication among officers. Such actions can hardly be described as ‘discretionary’, and the courts are entirely competent to conclude that they are negligent. Moreover, such actions could lead to liability in a variety of other professions that do not enjoy immunity in negligence; the liability of medical professionals for poor record-keeping is a prime example.127 It is highly disingenuous to suggest that a police officer who misplaces a credible tip or overlooks a key piece of evidence should be immune from liability due to the inherently discretionary nature of his functions. As the majority in Hill opined, it ‘is not necessarily a bad thing’ if the imposition of liability encourages police to be more careful in conducting investigations.128 The financial threat of negligence liability has had a salutary effect on other professions and industries, leading to, inter alia, better warning labels on products,129 better disclosure of the risks of medical procedures,130 and the more responsible service of alcohol.131 On the whole, society is better off for the greater care that negligence liability has encouraged in these areas. It seems curious that we would not demand similar care from police officers, who hold substantial power in society and are in a position to affect our liberty.

D. Potential Use by Ethnic Minorities Beyond its potential to improve police practices generally, the tort of negligent investigation may help to provide a voice for racial and ethnic minorities, who are typically incarcerated at a disproportionately high rate.132 There is a widespread perception among these minorities that police hold stereotypical attitudes toward them and unfairly target them in their investigations.133 As with domestic violence, there may well be systemic factors which lead police to treat visible minority suspects differently than Caucasian suspects. Further, Hill illustrates that the conduct of certain identification procedures may unfairly single out visible minority suspects. 127

Braun Estate v Vaughan [2000] 3 WWR 465 (Manitoba CA). Hill v Hamilton-Wentworth [2007] SCC 41, 285 DLR (4th) 620 [56]. 129 Hollis v Dow Corning Corp [1995] 4 SCR 634. 130 Reibl v Hughes [1980] 2 SCR 880. 131 Jordan House Ltd v Menow [1974] SCR 239. 132 Obviously, disproportionate incarceration rates are not solely or even largely attributable to police practices. There are numerous social factors that lead to increased crime and crime reporting rates in Aboriginal communities: see Roberts and Doob, above n 100, at 481–2. However, there is also evidence of discrimination within the justice system. 133 See HC Jain, P Singh and C Agocs, ‘Recruitment, Selection and Promotion of Visible-Minority and Aboriginal Police Officers in Selected Canadian Police Services’ (2000) 42 Canadian Public Administration 46. 128

308

Erika Chamberlain

In Canada, the problem may be particularly acute for Aboriginals.134 One of the most notorious miscarriages of justice in Canada was the wrongful conviction of Donald Marshall, Jr, a Mi’kmaq, for the murder of a black youth in Nova Scotia. Marshall spent 11 years in prison before his name was cleared, and even then the Court of Appeal suggested that he was responsible for his own wrongful conviction.135 This suggestion was forcefully repudiated by the Royal Commission appointed to inquire into the case, which concluded that the ‘criminal justice system failed Donald Marshall, Jr at virtually every turn’.136 The Commission found that the police role in the miscarriage of justice was ‘central throughout’.137 They wrote: If it had not been for the inadequate and unacceptable investigation by the Sydney City Police Department in 1971, for example, Marshall would never even have been charged. If it had not been for the incompetent reinvestigation of the original incident by the RCMP in 1971, Marshall would have been freed within a few months of his conviction.138

The Commission also concluded that Marshall’s race played a role in his treatment by the justice system. The police, courts and corrections officials involved held some stereotypical beliefs about Aboriginal people. For example, an admissions document from Corrections Canada described Marshall as a ‘typical Indian lad’ who ‘enjoys a good fight while intoxicated’.139 However, the Commission revealed a more insidious problem regarding the conduct of criminal investigations involving Aboriginals: ‘We believe that certain persons within the system would have been more rigorous in their duties, more careful, or more conscious of fairness if Marshall had been White’.140 This suggests that, even in circumstances where police do not take overtly racist actions, the degree of diligence exercised when investigating Aboriginal suspects is not as great as when investigating white suspects. Aboriginals seem more likely to be the victims of police negligence, and accordingly, more likely to be potential plaintiffs in claims for negligent investigation. Of course, the Donald Marshall investigation occurred in the early 1970s, and there has been progress in the police treatment of Aboriginal and other minority suspects in the interim.141 Nevertheless, both Aboriginal and black Canadians still feel that they are treated unfairly by police,142 and this view is even shared by many

134 A similar problem seems to exist in Australia: see J Stapleton, ‘The Golden Thread at the Heart of Tort Law: Protection of the Vulnerable’ (2003) 24 Australian Bar Review 135. 135 R v Marshall (1983) 57 NSR (2d) 286 (Nova Scotia Supreme Court, Appeal Division). 136 Nova Scotia, Royal Commission on the Donald Marshall Jr Proscution, Commissioners’ Report, vol 1 (Halifax, 1989) (Chair: Hickman CJNS) 15. 137 ibid, at 249. 138 ibid. 139 ibid, at 148. 140 ibid, at 162. 141 See ibid, at 148–9. The Charter of Rights and Freedoms, entrenched in 1982, the various provincial human rights commissions, and countless advocacy groups have helped to raise awareness about racism in Canada generally and particularly among police. It is impossible to provide a comprehensive survey of these developments here. 142 See Jain, Singh and Agocs, above n 133, at 48, 61–2.

Tort Law as Police Ombudsman

309

white Canadians.143 In addition, ethnic minorities are sceptical about the objectivity and efficacy of police complaints procedures.144 And while there have been several public inquiries about the wrongful convictions or deaths of Aboriginals at the hands of police,145 such inquiries are typically reserved for the most serious cases. They are of little use to those who have been wrongfully charged with lesser offences or are acquitted at an earlier stage. Such persons may, however, have suffered considerable loss, and the tort of negligent investigation may provide a new avenue to police accountability. Further, to the extent that claims for negligent investigation expose substandard practices or indirect prejudice, they will help ethnic minorities to keep police/race relations in the public spotlight. Finally, the threat of substantial damages awards may be the necessary incentive to achieve meaningful reform.

IV. Conclusion: The Role of Tort Law in Promoting Police Accountability The tort of negligent investigation can play a meaningful role in promoting police accountability in modern society. It is a useful supplement to current police complaints procedures, which are often seen as biased or incomplete. It allows plaintiffs, whether crime victims or the wrongly accused, to bring greater public attention to police negligence. It provides more public recognition of the wrongs they have suffered and, through its financial consequences, can provide the necessary ‘teeth’ to motivate institutional change. This may be particularly important in cases involving disadvantaged groups, like domestic violence victims, Aboriginals, and other racial minorities, who have had difficulty achieving substantive reform through more traditional legal means. Committee reports and policy guidelines ring hollow if they are not supported by action, and experience suggests that there is little action in these areas unless there are practical consequences. As with the liability of manufacturers and medical practitioners for failure to warn, the potential for tort liability may provide the incentive to change that policies, regulations and disciplinary procedures could never do.

143 See DA Northrup, ‘Public Perceptions of Police Treatment of Minority Groups and Disadvantaged in Metropolitan Toronto’ (1996) 11(1) Institute for Social Research (York University) Newsletter, where the author found that 50% of Caucasian respondents to a Toronto survey thought that police treat black people worse than whites. 144 See Paying the Price: The Human Cost of Racial Profiling (Toronto, Ontario Human Rights Commission, 2003) 26–7. See also the Lesage Report, above n 6, at 24–5. 145 See Royal Commission on the Donald Marshall, Jr Prosecution, above n 136 (wrongful conviction); Manitoba, Report of the Aboriginal Justice Inquiry of Manitoba: The Death of John Joseph Harper (Winnipeg, Aboriginal Justice Implementation Commission, 1999) (shoddy investigation into police shooting of Aboriginal man); SB Linden, The Ipperwash Inquiry, Commissioner’s Report (Toronto, Ministry of the Attorney General, 2007) (police shooting of an Aboriginal protester).

310

Erika Chamberlain

Of course, the action for negligent investigation is no panacea. It will not, by itself, eradicate wrongful conviction or stereotypical beliefs about women and minorities. Practically speaking, negligence claims will be available in only the more egregious cases, and only to plaintiffs with some financial support and the will to engage in extended litigation. For many, further contact with the justice system will be the last thing they desire. However, for those plaintiffs who can point to clear police negligence, the tort of negligent investigation may be the best ombudsman available. It empowers them psychologically, politically and financially, and it serves the normative function of raising the standards of investigation. As discussed by the majority in Hill v Hamilton-Wentworth Regional Police Services Board, there is no established reason to maintain police immunity in negligence. Their investigatory powers, while discretionary, are not unchecked. Like doctors, lawyers, and accountants, they should and do aspire to high professional standards. And like those other professionals, they should be subject to judicial scrutiny when their conduct falls below those standards, particularly when we remember that citizens’ physical security or liberty may be at stake. Negligence liability will help to maintain the standards of professional police conduct, and to ensure that they exercise the skill and care that society expects. If it makes them more diligent in conducting their investigations, then, as McLachlin CJ quipped, ‘this is not necessarily a bad thing’.146

146

Hill v Hamilton-Wentworth [2007] SCC 41, 285 DLR (4th) 620 [56].

13 Deterrence in Private Law YOCK LIN TAN

I. Introduction Contemporary jurisprudence on the idea of private law continues to be strongly occupied with taxonomies with their predicate of immanence or internal intelligibility and self-sufficiency.1 There are, as might be expected, different explanations of this immanence but whatever the explanations, these taxonomies are uniformly presented as, or assumed to be, self-establishing.2 They need no further explanations in themselves and are dependent on no external and contingent causes for their existence. For instance, each category of private legal right (contract, tort, restitution, property)3 responds to a different ‘causative event’ but, in common with each other, rests on an identical, self-sufficient, non-reductive, autonomous premise. In the opposite camp, taxonomies are regarded as didactic and helpful to a point but ultimately are futile edifices on which to build a conclusion;4 taxonomies are really fungible expressions of a diversity of considerations, reflected in the former approach to private law, but in principle unrestricted to the private interests of the parties in dispute.5 Under the former approach, if by the goals of private law we mean a resultoriented, functional notion, there should be no real goals to speak of; or perhaps we can say there is only one goal to speak of, namely to ensure that the idea of law is faithfully expressed and implemented.6 Contrary to the current fashion of 1 See, eg, EJ Weinrib, The Idea of Private Law (Cambridge MA, Harvard University Press, 1995); P Benson, ‘The Unity of Contract Law’ in P Benson (ed), The Theory of Contract Law: New Essays (Cambridge, Cambridge University Press, 2001); P Birks, English Private Law (Oxford, Oxford University Press, 2000) ‘Introduction’; and P Birks, ‘Equity in Modern law: An Exercise in Taxonomy’ (1996) 26 Western Australian Law Review 1. 2 Weinrib, ibid, at 16 refers to ‘a self-understanding enterprise’. 3 Weinrib leaves out property from his enumeration of private law rights. cf H. Dagan, ‘The Distributive Foundation of Corrective Justice’ (1999) 98 Michigan Law Review 138. 4 They may be dangerous if they foster dogmatism: G Samuel, Law of Obligations and Legal Remedies, 2nd edn (London, Cavendish Press, 2001) 65. 5 See, eg, S Waddams, Dimensions of Private Law: Categories and Concepts in Anglo-American Legal Reasoning (Cambridge, Cambridge University Press, 2003); HW Tang, ‘Storytelling in the Law of Unjust Enrichment’ ch 19 of this book. 6 cf S Smith ‘The Rights of Private Law’ ch 5 of this book.

312

Yock Lin Tan

invoking policy whenever explanations of liability fail, policy in particular will have no place as a primary determinant of right. The role and notion of goals expands more meaningfully if our view of rights requires us to consider interests and policies going beyond the parties in dispute. So the notion of goals makes more sense to legal scholars of the opposite persuasion. Legal scholars invariably tacitly assume one or the other posture, or explicitly argue one or the other, when they consider justifications based on rationale, principles, and policies.7 There has been no small number of views, going both ways, but noticeably, in this burgeoning theoretical and jurisprudential literature, the role of deterrence in private law is obscure or neglected.8 In the civil law, there are only pockets of law in which deterrence is a rationale or ground of liability or remedial principle.9 What this chapter represents is a conviction that the question whether deterrence has a role to play in private law is meaningful, important and cannot be neglected in serious studies of private law goals. I first sketch out the competing theories of private law. Then I express a view on where deterrence in theory fits in.10 Finally, I review the cases on a selective basis to see whether professed views of deterrence in fact provide the closest fit in these areas. In doing all this, I do not forget our discipline’s taste for the particular and for concrete manifestations but I am comforted that there is room for testing its particular features against bigger theoretical models. If there is nothing else to gain from this exercise, we shall at least benefit from understanding the law’s unstated orientation.

II. Is Private Law An Immanent Idea? Theories of private law may be grouped under three headings. There is the Aristotelian-derived category which opposes corrective or commutative justice and distributive justice11 and assigns the former as the unique and distinctive and essential characteristic of private law.12 Commutative justice can be conceived of 7 S Hedley, ‘Looking Outward or Looking Inward? Obligations Scholarship in the Early 21st Century’ ch 8 of this book, calls for a dialogue between the above two camps. 8 J Edelman, Gain-Based Damages: Contract, Tort, Equity and Intellectual Property (Oxford, Hart Publishing, 2002) and the literature on exemplary damages are notable exceptions. For a sample of the latter, see NJ McBride, ‘A Case for Awarding Punitive Damages in Response to Deliberate Breaches of Contract’ (1995) 24 Anglo-American Law Review 369; A Phang and PW Lee, ‘Restitutionary and Exemplary Damages Revisited’ (2003) 19 Journal of Contract Law 1; R Cunnington, ‘Should Punitive Damages be Part of the Judicial Arsenal in Contract Cases?’ (2006) 26 Legal Studies 369. 9 For an account of the significance of deterrence reasoning in tort law scholarship, see GT Schwartz, ‘Mixed Theories of Tort Law: Affirming Both Deterrence and Corrective Justice’ (1997) 75 Texas Law Review 1801. In the criminal law, deterrence as a sentencing principle has become unfashionable in most places: see A Ashworth, ‘Criminal Justice and Deserved Sentences’ in N Lacey (ed), A Reader on Criminal Justice (Oxford, Oxford University Press, 1994) 206–20. 10 This attempt has some important precedents. 11 In mutually exclusive terms: Weinrib, above n 1, at 61–2. 12 See Weinrib, above n 1; EJ Weinrib, ‘The Gains and Losses of Corrective Justice’ (1994) 44 Duke Law Journal 277; EJ Weinrib, ‘Restitutionary Damages as Corrective Justice’ (2001) 1 Theoretical

Duty of Care Deterrence in A Rights-Based Theory in Private Law of Negligence

313

as the universal virtue of rendering to everyone what belongs to him. This virtue is demanded when there is an inequality, ie when one person gains in holdings by another’s loss. The proper notion of equality—and of gain (benefit) and loss (harm) in holdings—is critical, of course, and is obtained by abstracting from personal virtues and circumstances so as to arrive at an irreducible state where we compare the state of the two persons, one supposedly harmed and the other benefited, to their counterfactual states had the harmful action not occurred. When that is done, we have transcended the particular contingencies of the relations between these two persons and are left with the pure notion of equality of justice. Among those who adhere to the immanent idea of law, some attempt is typically made to explain and clarify the process of abstraction which makes precise what constitutes an equality of holdings. Weinrib, for instance, derives the element of equality from the Kantian practical reason. An individual’s autonomy or capacity for self-determination is necessarily subject to the demands of practical reason because it must be consistent with the autonomy and purposive capacity of another. In other words, a legal right is the necessary and equal expression of the purposive capacity.13 Distributive justice on the other hand answers a distinctly different question of proportionality and desert. Premised on a scarcity of resources or funds, theories of distributive justice aim to provide justifications for apportioning and sharing the common pie. Distributive justice is thus that universal virtue of rendering or distributing to everyone that reward or punishment according to his merits, ‘observing a just proportion by comparing one person or fact with another so that neither equal persons have unequal things, nor unequal persons things equal’.14 A necessary condition is that the merits of distribution should be instrumentally necessary to resolve the scarcity and not merely contingent, absolute in scope and not relative, universal in application and not particular. No or little abstraction is needed because the notion premises that scarcity produces inequality and that intrinsic value cannot be attributed to equality alone. Egalitarian considerations are important of course (and are preeminent in strict egalitarianism)15 but it is necessary to isolate, identify and prescribe a differentia of merit which should be equalised or prioritised. Examples include resource equality, welfare equality, opportunity equality that raises the absolute position of the least advantaged in society, desert equality, or welfare maximisation. It follows that inequality is not wrong in itself. Whichever distributional merit is considered appropriate in any

Inquiries in Law 1; EJ Weinrib, ‘Punishment and Disgorgement as Contract Remedies’ (2003) 78 Chicago-Kent Law Review 55. 13 J Rawls, A Theory of Justice (Oxford, Oxford University Press, 1971) derives it from an original covenantal position. Noticeably, those who subscribe to the idea of law as corrective justice focus on right and very few on duty. 14 It is easy, given the Cartesian nature of the opposition between corrective and distributive justice, to identify the proponents of this school, even when they do not adopt overtly the terminology of the jurisprudence. 15 I would put Jean Hampton’s ‘The Retributive Idea’ in JG Murphy and J Hampton (eds) Forgiveness and Mercy (Cambridge, Cambridge University Press, 1988) here.

314

Yock Lin Tan

category of activity or state of affairs, inequality which satisfies or is consistent with the accepted differentia of merit is justifiable. The second broad category of theories of private law predicating that all justice is distributive naturally denies the immanence of right. These theories have again, as might be expected, numerous manifestations. A legal right to some is an expression or prioritisation of policy. A legal right in the view of others is merely a conclusion according to a balance of interests. Waddams’s Dimensions of Private Law16 may be cited as an example; a strong notion of distributive justice underlies his numerous demonstrations showing that decisions of the common law courts are reached not so much in terms of juridical categories of right but in terms of broader notions of exchange interest, unjust enrichment, security of property and so on. Most if not all economic theories of private law are erected on the same premises coupled with some prioritisation. This is because even when they seem focused entirely on what is rational and efficient individualism, they do so in order to attain global efficiency. Utilitarian theories of private law also belong here. An act or policy, utilitarians exhort, is right if and only if it produces at least as much utility as any other alternative. A wider version of such theories formulates this precept indirectly in terms of the rules to be adopted or pursued. A rule is right if and only if it produces at least as much utility as any other rule. Whether in its narrower or wider form, utilitarian theories of private law are simply distributional theories. Another broad category of theories of private law, a mixed category, may be identified.17 Instead of saying that all private law is corrective or that it is distributive, one could work out a role for each, without excluding the other or requiring exclusive dominance for one or the other.18 One’s starting point may be different. Beginning with the view that private law is primarily corrective justice, one could advocate distributional principles as remedial or institutional measures which ensure that private law’s primary function will not break down or be frustrated.19 Kronman seems to be arguing along these lines.20 Private law rights exist or do not exist in accordance with principles of corrective justice and cannot primarily be derived from principles of distributive justice. Nevertheless, distributive justice is unavoidable in relation to contract law. Seeing contract as an exchange which orders wealth holdings, he demonstrates that it is consistent with libertarian views to have regard to distributive justice and that this is preferable to taxation in many respects.21

16

Above n 5. See S Smith, ‘The Rights of Private Law’ ch 5 of this book. See Schwartz, above n 9; L Smith, ‘Restitution: The Heart of Corrective Justice’ (2001) 79 Texas Law Review 2115; cf Dagan, above n 3. EJ Weinrib, ‘Deterrence and Corrective Justice’ (2002) UCLA Law Review 621 sees no difficulty in according deterrence a role in the operation of legal norms as an ensemble, although it should have no influence on the contents of those norms. 19 Implicit in this statement is the view that in public law, distributive justice is primary and corrective justice serves as a secondary check against anomalies and inappropriateness. 20 AT Kronman, ‘Contract Law and Distributive Justice’ (1980) Yale Law Journal 472. 21 I do not overlook that Kronman’s position is a stronger one. He is not saying that distributive justice plays an auxiliary and facilitative role but quite the reverse, he suggests that distributive justice is an essential element of private law because conflict is inevitable. ‘The implementation of any patterned conception of justice is bound to require interference in people’s lives’: ibid, at 511. 17 18

Duty of Care Deterrence in A Rights-Based Theory in Private Law of Negligence

315

III. Where Does Deterrence Fit In? As a concept, deterrence does not easily fit in with either rights-based accounts of private law or distributive and communitarian accounts. Superficially, any rule of liability offers a potential deterrent to a person who would act in contravention of it;22 deterrence in this sense is afforded by the potential or actual effect or consequence of rule violation and is greater or smaller according to the person’s contingent circumstances and risk proclivities. This superficial notion of deterrence must be rejected as being of little if any material value to the rights-view of private law,23 even when we introduce more objective elements such as the harm occasioned to third parties if the person is not deterred. Weinrib points out, for instance, that if deterrence of harmful conduct ‘is justified at all, it is justified even for injuries that cannot be deterred’.24 Similarly, I would add, if deterrence were the ground of liability in negligence, every person harmed in fact must be entitled to recover not only his own loss but also the loss of others. ‘[N]othing about the justificatory consideration expressed by the goal of deterrence limits its operation to acts that cause compensable injury’ to any one victim.25 Put more generally, neither the infinitely variable external effects and consequences of rule violation nor the individual motivations of actors can constitute a necessary and essential notion of equality of corrective justice.26 They must be abstracted from in order to reach that essential notion of equality which measures the need for corrective justice. When that is done, only the act of the actor that has harmed the other is relevant and only the harm that is caused by the act of the actor is an intrinsic disturbance of the equality of purposive capacity between them. Any notion that an actor has increased in purposive capacity by the indirect harm that his act has encouraged another to inflict is simply contrary to what corrective justice demands. The superficial notion of deterrence as making the actor responsible for all the effects and consequences of his act equally fails to satisfy as a measure of distributive justice.27 We are for the moment constrained to accept that each distributional approach produces a different outcome28 and that no one has to date convincingly demonstrated that we must follow one path to the exclusion of all

22

Especially to one who is risk-averse. Every rule of liability would be a rule of liability if we thought of deterrence like this. 24 Weinrib, above n 1, at 38. See also Weinrib, above n 18. 25 Weinrib, above n 1, at 38. 26 Save that exemplary damages awarded when the defendant’s ‘malevolence … injures the plaintiff ’s feelings of dignity and self-worth’, they are compensatory and not punitive and are then consistent with principles of corrective justice: see Weinrib, above n 1, at 91. 27 This notion of deterrence underlies the argument that the prospect of liability ex ante should be determined by deterrence reasoning and that it is necessary to induce an appropriate deterrence response for the prevention of tortious harm: see Schwartz, above n 9. 28 For instance, if we think of harm caused by negligent conduct, a principle that equalises welfare leads to indemnification of the victim but if resources are to be equalised, the loss should be spread among the persons with greatest capacity to bear it. Utilitarian welfare maximization principles will produce another difference in directing that the loss should be borne by the person who could have prevented it most economically. 23

316

Yock Lin Tan

others. Proof may well be impossible, since there appears to be no invariance in the choice of resource, welfare etc as differentia because the premises of material scarcity are not invariant. The diversity in differentiae results from the different ways in which we may define the material premises and we are unable to choose one of these differentiae to the exclusion of all others because we are unable to establish a true premise to the exclusion of all others. We can only require that while there can be variation between differentiae, there ought to be no variation in what makes up a valid differentia. Whatever these attributes may be, the attribute of instrumental necessity and sufficiency must be one of them. A valid differentia is instrumentally necessary and sufficient, as opposed to contingent, if it furnishes a fully rational and necessary nexus to the harm to be distributed, without being either under-inclusive or over-inclusive. For instance, if we define harm to resources as a deprivation to the victim of what he had before the act of the actor, which he would have maintained had the act not occurred, then a proposal to distribute that harm in accordance with equalisation of resource will be necessary and sufficient. It will be neither under-inclusive nor over-inclusive as long as harm is defined in that counterfactual sense. It would only be under-inclusive or overinclusive if we defined harm in terms of damage to welfare and then purported to apply the principle of equalisation of resources. If the premises are that welfare is the virtue in scarce supply, only equalisation or maximisation of welfare will satisfy the requirement of instrumental necessity and sufficiency.

A. Problem of Class Identity To satisfy the requirements of an instrumentally necessary and sufficient differentia of merit, a differentia of distributive justice must be an a priori allocation capable of determination by a thought experiment. The problem is that the effects and consequences of rule violation in the deterrent sense are incapable of a priori determination. The harm and loss to an identified or identifiable person is indeed capable of a priori determination (for instance, we can think of harm as making a person worse off than she would have been at the material time if the act had not occurred) and we can sensibly imagine a system of justice which distributes the negative goods (ie the harm arising from some conduct or activity). But the effects and consequences of rule violation in the deterrent sense are not confined to an identified or identifiable person. They are, as it were, comprehended by a notion of global gain and loss of a fluctuating body, which is not definable because the identity of members of the class comprising both victims and other actors will alter according to the context and circumstances. In some contexts, the act committed will encourage similar conduct on the part of other actors and some members of the class of victims are in fact future persons and unidentifiable. Present members who are harmed by the actor’s conduct are identifiable but even then the identity or composition of this class will vary according to time and other circumstances of the occurrence of the act as well as the efficacy of past deterrence.

Duty of Care Deterrence in A Rights-Based Theory in Private Law of Negligence

317

In the case of harm to an identified person, it never matters whether the act occurs later or earlier in the sense that there will never be any loss of identity in-between. But in deterrence cases, loss of identity is inescapable because, ex hypothesi, the harm is larger or smaller according to the efficacy of past deterrence. This relativity of harm is even more pronounced if there are future actors and future persons harmed by the acts of future actors. Put another way, deterrence in the sense of burdening the actor with responsibility for difficulties of practical enforcement of a rule of liability immediately creates discordance between the differentia of notional harm defined in terms of such responsibility and the material premises of resource or welfare or opportunity. Instrumental necessity and sufficiency is lost for a number of reasons. First, preceding actors and not the actor in question have contributed to these difficulties whereas the actor’s contribution is to add to the difficulties which the next victim will encounter. So when notional harm encompasses these matters, there is a departure from the material premises which involve the resource or welfare or opportunity of an identified person. Further, the actor’s contribution to difficulties of practical enforcement varies according to the efficacy of past and present deterrence of herself and other actors. Again, there is a departure from the resource or welfare or opportunity of an identified person. The upshot is that the identity of the class of persons who are notionally harmed cannot be established and the notional harm is indeterminate. These criticisms are equally applicable to deterrence in the strongest sense which saddles the actor with other persons’ harm. In a phrase, deterrence seeks the recovery of notional harm which cannot be exactly limited and located in an identifiable person or class of persons, making it impossible to satisfy a resource-based, welfare-based principle of distributive justice that the differentia for allocating harm must be instrumentally necessary and sufficient and not contingent.

B. Redundancy of Deterrence Even if we can ensure that the class which will suffer from the actor’s act will retain the same identity throughout all material times, deterrence will be hard to explain in terms of a number of accepted principles of distributive justice. In the first place, we can question whether the distribution of loss according to deterrence is a distributive question to be settled according to some notion of welfare maximisation. Under welfare-based principles, we should indemnify the victim who loses most from the act of the actor. Conversely, we should maintain an adverse position against an actor who benefits most from his acts. Many of our actions have indirect effects on the conditions of life of others, and actions that harm others are no different. For instance, a breach of contract may indirectly cause the innocent party to fail to pay his creditors who make him bankrupt and may in turn contribute to his wife committing suicide and his children being left destitute. Consistently with welfare-based principles, nothing less than complete

318

Yock Lin Tan

indemnification will do and the indemnification of the person harmed alone should sufficiently explain why we should take resources from the actor and transfer them to the person harmed, regardless of the actor’s fault. Deterrence is actually irrelevant under such welfare-based principles. It adds nothing to the case for indemnification of the victim. For instance, with reference to a popularly conceived and simplistic principle of utilitarianism, it is not wrong to make one person suffer so that there is a net gain for another but this can be done whether or not it will deter the actor from acting.29 Under resource-based principles or resource egalitarianism, it is simply not clear why or how deterrence restores equality to endowments and resources as between parties who are implicated, directly or indirectly. Resource-based principles seek equality of resources but allow the resource-holder to express his ambition and to abide by the consequences of his choice of use of these resources. Inequalities in use are not unjust per se, provided that to reflect this qualification of ambition sensitivity, we require that circumstances over which a person has no control should not adversely affect his choice of use of resources. Resourcebased principles thus prescribe that we should help the worse off by equalising resources and endowments, and this leads us to loss-spreading, not deterrence. The imposition of tortious liability according to the capacity to spread the loss or the capacity for insurance protection is a good example of resource-based distributive principles. Again, deterrence adds nothing to this system of distributive justice. Under Rawls’s difference principle as formulated for the distribution of positive goods, justifiable social and economic inequality must satisfy two conditions. They must be attached to positions or offices open to all under conditions of fair equality of opportunity and they must be to the greatest benefit of the least advantaged. So as to apply this principle to negative goods (ie to the just distribution of harm), let us posit that an unequal distribution of the burden of harm is justifiable if it equalises opportunities for harm avoidance and produces the greatest benefit to the least advantaged. Is deterrence a difference principle? Assuming that opportunities are equal, will deterrence raise and maximise the absolute position of the least advantaged in society? It is hard to see that the person harmed by an activity is necessarily the least advantaged in society. On the other hand, deterrence directs attention away from the position of the person harmed to the actor. It cannot matter that the person harmed can suffer the harm well and is not among the least advantaged, because it is the actor who must be deterred and she will not be deterred if she is to be leniently treated when the person harmed is not among the least advantaged.

29 We can of course refine welfare-based applications by introducing a comparison between the harm done by the actor and the counterfactual worst harm to another person, thus giving an incentive to the actor to pursue the least harmful course of conduct. It is only then that we can say that the actor should be deterred from causing harm which is not the least that can be inflicted.

Duty of Care Deterrence in A Rights-Based Theory in Private Law of Negligence

319

C. Constructing a Notion of Notional Harm As already said, a common difficulty in fitting deterrence among the distributive principles just described is that it relies on a relative notion of harm (dependent on time and context and especially class identity and on the actor’s rational decision to be deterred) whereas these principles presuppose a notion that is invariant or absolute. Naturally, one possible response to the relative harm objection is to re-formulate the notion of gain and loss or benefit and harm so as to avoid comparative elements. One could suggest that a person is harmed by another if, as a result of the other’s actions, the person falls below a normatively-defined threshold state. That defined state encompasses both the potential and actual harm which may be or is caused by the actor to all victims similarly situated and circumstanced, actual or putative. There is a further refinement. Two reference points are possible in respect of the actor. One is that the actor herself is referenced and the other is that others like the actor are included. If so, the defined threshold state could be refined to encompass the potential and actual harm which may be or is caused by the actor or other similar actors to all persons situated similarly to the victim. Under either formulation, the variable identity or lack of identity of people harmed will not preclude a conclusion that harm has been caused to the victim which is a fit matter for deterrence. This assumed notional other persons’ harm seems to underlie the recommendations of those legal scholars and courts who seriously advocate specific deterrence as a critical decisional consideration in selective legally significant transactions. Becker’s economic theory of deterrence in crime and punishment30 has found lodging in some circles, as providing an analogous answer of private law. The theory, simply put, is that intentional harm is deterred by setting sanction equal to gain from harmful conduct. Building on this model, some have argued that damages for fraudulent misrepresentation would insufficiently deter fraudulent misrepresentation unless there are rules which strongly favour the representee, such as more generous remoteness rules, rules for the recovery of profit and rules which vary the date of loss assessment.31 Deterrence theories of vicarious liability lean in the same direction. They advocate that where direct liability is too costly to enforce, whether because of high costs of detection or other practical difficulties, indirect liability is justifiable if it is the only practical way to maintain the applicable principles of distributive justice.32 They therefore shift the burden of liability from the actor to another with control over the actor in some sense.

30 GS Becker, ‘Crime and Punishment: An Economic Approach’ (1968) 76 Journal of Political Economy 169. 31 See Q Zhou, ‘A Deterrence Perspective on Damages for Fraudulent Misrepresentation’ (2007) Journal of Interdisciplinary Economics 83. 32 See DA Strauss et al, ‘Brief of Amici Curiae (Kenneth Arrow et al) in Support of Petitioners in the Case of Metro-Goldwyn-Mayer Studios Inc v Grokster Ltd’, at www.copyright.gov/docs/mgm/kennethArrow-et-al.pdf, accessed on 9 December 2008.

320

Yock Lin Tan

They focus on creating efficient mechanisms for law enforcement by deterrence of inefficient acts. Making suitable adjustments for notional harm,33 we can develop the argument from the views just rehearsed that the notional harm arising from the difficulties of policing and enforcing fraudulent misrepresentation should be distributed according to desert, meaning efforts or contributions to enhancement, containment, or avoidance of this notional harm. So far as deterrence-based vicarious liability goes, the argument for desertbased distribution is more sophisticated. It is typically advanced as a ‘parasitic’ argument because it predicates that the primary rule of direct liability cannot be enforced adequately or effectively.34 An additional, parasitic or auxiliary layer of liability is interposed on the indirectly involved party who is in a position to deter the harmful conduct and which is aimed at making the right of the plaintiff efficacious against ineffectual self-deterrence on the part of the primary actor. Typically when we consider the vicarious liability of employers, we are asked to accept the high costs of detection and enforcement of direct liability against an employee (arising from difficulties in identifying and suing the employee). These difficulties, with respect, are often exaggerated. There are numerous cases where they are no more distinguished in scope and extent than in non-employment cases. In any event, if deterrence is pertinent, the employer is to be deterred not so much from taking advantage of the difficulties of detection as from creating opportunities for an employee to take advantage of them. In other words, we are concerned less with difficulties of detection and enforcement than difficulties of obtaining satisfaction from possibly men of straw. Even in this sense, the employer is to be deterred not from taking advantage of difficulties of obtaining satisfaction of harm from an employee of straw but from creating unequal opportunities for

33 This is necessary because the actor’s conduct (making the fraudulent misrepresentation) will not have added to existing difficulties of detection and enforcement (evidenced by prevalence of the relevant activity). These difficulties ex hypothesi confronting the victim are the contributions of preceding actors. On the other hand, the doing of the act may add to the difficulties and externalities confronting a future victim of another’s act. If so, it is the future social product which is impeded. Again, it must be asked: in what sense can it be said that making a present defendant pay for the historical losses is a valid desert-based response consistent with theories of responsibility and agency? To surmount the charge of inconsistency, we shall need to refine the defined normative threshold state of harm proposed earlier so that the actor has harmed someone if she causes this person to be worse off than another person would have been harmed had the actor acted differently. So if the actor had acted differently and another person was harmed, that harm would not include the costs of detection and enforcement engendered by past activities and actors. However, because the actor did not act differently but acted similarly to acts to be deterred, the victim has suffered in addition the costs of detection and enforcement. Under this version, a weaker theory of responsibility is adopted under which an actor can harm someone when she does not choose a better option which would avoid the harm. So the harm occasioned by difficulties of detection and enforcement is attributed to the present actor because she does not choose a better option which would avoid those difficulties for the victim. On what basis of responsibility should an actor choose the better option in the sense of the least harmful option when inflicting harm? 34 cf GT Schwartz, ‘The Hidden and Fundamental Issue of Employer Vicarious Liability’ (1996) 69 Southern California Law Review 1739.

Duty of Care Deterrence in A Rights-Based Theory in Private Law of Negligence

321

obtaining satisfaction of harm. If, hypothetically, the employer had conducted herself in the way the employee has and satisfaction would have been obtained from her, then by substituting an employee for herself the employer would be creating unequal opportunities for obtaining satisfaction for the harm inflicted by the employee. We are to deter her not from causing harm which is easily covered up or not easily proved but from not contributing to the creation of good and happiness. If this is correct, then the argument for deterrence, far from being parasitic, should be framed as a primary argument.

D. Is Deterrence Desert-Based? Is deterrence in either case under examination truly a valid desert-based principle of distributive justice? Broadly, such principles prescribe punishment or reward according to one’s contribution to socially unproductive or socially productive work. Activities which raise the social product are rewarded and those which diminish it are punished. But with respect to the case for deterring fraudulent misrepresentation, the first difficulty is that it suggests that the defendant has taken advantage of the difficulties of detection and therefore deserves a high penalty, but it ignores facts to the contrary if they exist. Some, if not many, will take advantage of these difficulties; but not all will. Of those who are prone to take advantage (because they have the requisite knowledge) there are circumstances in which they nevertheless act ethically and not selfishly. More important perhaps is the question: How does a deterrence formula square with the desert of the ignorant actor? Knowledge of the rule of liability is crucial to desert-based decisional thinking since the actor who does not know of the rule does not take deterrence into consideration when she chooses to be a rule-breaker. She simply is not deterred. Ignorance blunts deterrent effects and this rule-breaker is only deterred the second time when it may be too late. So as a principle of desert-based distributive justice, deterrence only makes sense if there is prior knowledge of the need for deterrence. This severely limits the role of deterrence in private law, which in numerous instances treats knowledge of the existence of a rule of liability as inconsequential. To overcome the contingency that some actors know of the need for deterrence but others do not, desert-based principles which are person-oriented must be restricted to the initial design of liability. They must focus not on individual cases but on the rules governing transactions to be deterred. This means that we shall not only need to rely on a threshold notion of harm, but we shall also need to rely on a threshold notion of action. We shall need to propose a notion of unconscious deterrence which would make knowledge of the existence of a rule of liability inconsequential. Such a proposal is possible if what we mean by deterrence is ruledeterrence and not act-deterrence. Let me base the argument on the hypothetical liability to compensate for harm occasioned by fraudulent misrepresentation. In devising a scheme that will be even-footed and will not be derailed by practical difficulties of enforcement of liability, the judge may suppose that if one rule is open to abuse, it should be augmented or supplemented by another rule that

322

Yock Lin Tan

permits her to add the notional harm or to order disgorgement of the notional gains of the actor. She may reason that the augmentation or supplementation will be effective because if it is known it will deter the actor. The efficacy of the augmentation is determined in terms of unconscious deterrence. It does not matter to the argument that an actor, through ignorance of the rule, breaks the rule. Rule-deterrence will make the scheme efficient. Thereafter, however, the judge when applying the rule no longer troubles herself with deterrence. In this sense, the initial design is based on unconscious deterrence. There is another sense of rule-deterrence which may be explored. In relation to the initial design of a scheme of liabilities, the judge may consider that the scheme will be more efficient if it is incumbent on all who fall within the terms of the scheme to take steps to acquire knowledge of what is expected of them, in particular of those rules which serve deterrence purposes. Whether the actor then knows or does not know that a rule she has broken serves deterrence purposes is immaterial because she should have known. The judge will presume knowledge and the initial design of the scheme of liabilities will be erected on the basis of presumed knowledge of deterrence considerations. Unconscious deterrence here refers to presumed knowledge that deterrence is part of the scheme. However, even if we can in this way overcome the problem of ignorance in desert by conceiving a role for unconscious deterrence in the initial design, we shall need to overcome another impediment. This is that the application of desertbased principles in devising the initial design requires specification of those activities that will count as socially productive and those which will count as socially unproductive and the computation of the net balance. A comparative analysis of desert is necessary. We shall need to examine the relative or comparative deserts of both actor and victim in order to decide whether, as a matter of the initial design or disposition of liability, the need for deterrence of the act arises. In the case of vicarious liability, the argument that the liability is a species of desert-based deterrent justice fares even more poorly because this additional consideration seriously weakens the argument for deterrence. So far from impeding the social product, the corporate organisation invested with separate legal identity must employ agents to do its acts which, ex hypothesi, raise the social product. The extent to which the social product is enhanced would, under a desert-based approach, be relevant in determining whether there is an overall reduction or diminution of social product as a result either of an increase in the difficulties of enforcement (through the multiplication of power to harm in the hands of diverse agents) and detection, or of a curtailment of the opportunities for obtaining satisfaction. However, those who have proposed deterrence as a just solution have never considered such scrutiny of the social product to be relevant.35 35 Further, it may be unnecessary and superfluous to refer to deterrence as a type of desert-based justice when other principles such as the difference principle already provide the justification for imposing vicarious liability on the employer. To the extent that the employer creates opportunities which disturb the equal distribution of opportunities of satisfaction for loss caused by the employee, the difference

Duty of Care Deterrence in A Rights-Based Theory in Private Law of Negligence

323

Ultimately, desert-based principles of distributive justice assume first that individual decisions, rational in isolation, can produce enhanced as well as disastrous collective consequences which promote or impair the individual ability to make future personal decisions and secondly that there is a meaningful sense of the community to be continued and propagated. This sense is measured by a defined normative threshold of harm and of action. To my mind, this process of imputing a notional harm to an actor is not distributive reasoning because the class associated with the notional harm lacks identity. If there is no clear and unequivocal class identity, it is insensible to think of distributing benefits and harms. I have also shown that we shall need to refine the notional harm so that it is changed from a counterfactual harm to a notion of the least best available harm. This is not possible within the limits of reasonable conjecture because it would require a threshold notion of similar acts. If deterrence must mean deterrence from pursuing an act when similar acts would incur less harm, how do we define a notion of similar acts?

IV. Deterrence is a Tax While I have shown that deterrence cannot be a principle of distributive justice, I do not say that there is no room for deterrence in private law. I argue quite the reverse that there is room for deterrence, but that deterrence is a tax and that its role in private law ought to reflect the tax principle. The tax principle is either criticised as a robbing principle (unjustifiable theft) or presented as an economic principle of an auxiliary nature (to meet social costs). It has hardly been broached as a principle of justice. Kronman, however, takes a more enlightened legal view of the contrast between taxation and contractual regulation in accordance with distributive principles of justice. He makes a number of interesting points which show, although he does not fully elucidate it, that tax is also a matter of justice. In outcome, there seems little to distinguish contractual regulation and tax imposition. Both have the effect of re-ordering private wealth, if we understand the notion of wealth in the widest sense. Both are similar in end (and share the same boundaries in that sense) but are set apart in the means. They are different methods of wealth re-distribution. The way Kronman explains it, a tax is the price people pay for being left free to transact as they please. The initial assignment of shares of society’s resources will over time distort and tax is the means to correct and restore the original principle of desert. ‘In this way, the original pattern of holdings can be re-established, and distributive

principle already states that the conditions for unequal re-distribution are met. First, the actor has created unequal opportunities and attached them to a position or office and secondly, the unequal re-distribution is to the greatest benefit of the least advantaged. Vicarious liability follows without any need to consider the merits of deterrence as a principle of distributive justice.

324

Yock Lin Tan

justice preserved’.36 The taxation principle is a dynamic of background justice in agreement with the way Rawls sees tax as an aspect of institution building, and of background justice. It protects and preserves the basic structure, the framework of entitlements within which individuals remain free to pursue their ends through voluntary transactions with others. It helps immensely that in Kronman’s discussion, certain misconceptions of the taxation principle are perspicuously avoided. Tax is not necessarily personoriented; for instance, the imposition of a Goods and Services Tax is transaction or activity-oriented, not person-oriented. Thus, as Kronman argues, the difference between contractual regulation for reasons of distributive justice and tax is not in transaction specificity. If so, the idea of deterrence as being a tax payable for pursuing a private activity is unobjectionable on the ground of any distinction between activity and personal orientation. Nor is there a real difference between the supposed neutrality of taxation and supposed non neutrality of contractual regulation. ‘Any workable scheme for re-distributing wealth is bound to discriminate unfairly in favour of some individuals or activities’, and it would be a mistake to suppose that neutrality is intrinsic to taxation; as between income tax and selective income tax, for example, it is obvious that there is variation in neutrality.37 Again, Kronman correctly observes that the supposed difference that tax is compulsory whereas contractual regulation is not, does not exist. Tax is only compulsory in a more or less practical sense. Even if only in theory, those subject to tax remain free to avoid it by reorganising their resources or activities. However, certain misconceptions remain to be cleared up. The notion that a tax is forced transfer or re-distribution to the state is taken for granted. There are ample indications of this in such passages as: ‘a tax scheme … requires those subject to the tax to share with the state the fruits of their voluntary transaction with other individuals’.38 And again, when he writes: ‘Usury laws, minimum wage laws, rent control laws’ are examples of such contractual regulation but ‘[i]n each case, the same end could be achieved through an appropriate combination of taxes and subsidies’.39 In my view, the assumption that a tax, unlike contractual regulation, is a transfer to the state is dubious. A tax has the effect of a tariff but the identity of the recipient or transferee is not essential to the nature of a tax. What is central is that it expresses a threshold condition for participating in an activity (including the accumulation of resources) which reflects the price exacted for participation in it; the intended actor being free to avoid it by a reorganisation of her resources or activities. Whether its incidence is defined in terms of the personal possession of resources or involvement in a taxable activity, identity is not critical to the tax principle. In contrast, identity is critical to a principle of distributive justice, since no such principle can be defended successfully if it does not predicate that an 36 37 38 39

Kronman, above n 20, at 499. Kronman, above n 20, at 503. ibid, at 504. ibid, at 507.

Duty of Care Deterrence in A Rights-Based Theory in Private Law of Negligence

325

instrumentally necessary, absolute and universal differentia of merit exists, which is in turn impossible without objective certainty of identity. Where objective certainty of identity is lacking, the principle that is just, is a rationing principle that assigns negative goods to those who will pay the price for participation in the activity proscribed and positive goods to those who will not. With respect to transactions in which objective certainty of identity is impossible, there is a threshold condition to satisfy which determines the pertinent level of notional harm or other person’s harm. That threshold condition is a function of liberty and solidarity. It expresses the price exacted in the name of solidarity; it is not capable of measurement in terms of any index of equality, and is unashamedly instrumental, contextual and always particular. If so, the examples of contractual regulation which Kronman gives are in truth examples of a tax. Though he calls them contractual regulations based on principles of distributive justice, it seems to me that his explanation runs into difficulties. Usury laws and minimum wage laws, to take some of his most basic examples of contractual regulation in the name of distributive justice, are not really distributional principles but tax principles. Moreover, as I believe that principles of corrective justice are a correct starting point for any meaningful discourse about private law, the principles of distributive justice relevant in private law must be of secondary character, not primarily intended to re-order wealth but to secure consistency with distributive principles of justice as established by public law. A court’s pure role in redistribution is therefore the modest one of discovering what principles of justice established by public law are relevant and have recognisable impact on principles of corrective justice. Significantly, the principles of corrective justice and distributive justice do not change although an activity is prohibited in the sense that a tax is imposed. A just tax, however, is one that produces a proportionate response between individual liberty and solidarity.

V. Fiduciary Obligations Cases on fiduciary obligations provide excellent material for selectively exploring the implications of deterrent justice because this is where deterrence makes a powerful impression or claim to be the singular rationale for strictly enforcing fiduciary duties.40 If the foregoing thesis is correct, this deterrence-based understanding of fiduciary obligations is a matter of regret. It has marred, if not retarded, the development of ‘no conflict’ duties, which are of course numbered importantly among fiduciary duties. There is no shortage of pronouncements of the pre-eminent role of deterrence in shaping fiduciary duties. Sometimes, these deterrence-based explanations go 40 See AJ McLean, ‘The Theoretical Basis of the Trustee’s Duty of Loyalty’ (1969) 7 Alberta Law Review 218; C Harpum, ‘Fiduciary Obligations and Fiduciary Powers: Where Are We Going?’ in P Birks (ed), Privacy and Loyalty in the Law of Obligations (New York, Oxford University Press, 1997).

326

Yock Lin Tan

along more procedural lines; for instance, that deterrence is required to offset the greater capacity to do evil arising from the fiduciary’s control of information and knowledge flowing or available to the beneficiary, or to remove the fiduciary’s evidential advantage in concealing her wrongdoing, or to relieve the beneficiary from exhaustively proving impropriety.41 More frequently, we are directed to desert-based deterrence explanations which proclaim the need to deter the fiduciary from being swayed by considerations of personal interest or more positively, the need to protect the integrity of the fiduciary relationship by encouraging loyalty. Long passage of time apparently has not blunted the sharpness of these pronouncements, including those which combine the need for substantive and procedural deterrence. As example the passage in Murad v Al-Saraj may be cited where Arden LJ said: It may be asked why equity imposes stringent liability of this nature. The passage just cited from the judgment of Lord Eldon LC makes it clear that equity imposes stringent liability on a fiduciary as a deterrent—pour encourager les autres. Trust law recognises what in company law is now sometimes called the ‘agency’ problem. There is a separation of beneficial ownership and control and the shareholders (who may be numerous and only have small numbers of shares) or beneficial owners cannot easily monitor the actions of those who manage their business or property on a day to day basis. Therefore, in the interests of efficiency and to provide an incentive to fiduciaries to resist the temptation to misconduct themselves, the law imposes exacting standards on fiduciaries and an extensive liability to account.42

A. Distinction Between Positive and Performance Duties As we confront these fiduciary liabilities, it is necessary to resolve some initial confusion of terminology. The epithet ‘fiduciary’ is not so much a description of the contents of the positive duty or discretion which she has undertaken or accepted, as of the manner in which that positive duty or discretion is to be performed or

41 In Parker v McKenna (1874) 10 Ch App 96 at 125 James LJ said: ‘the safety of mankind requires that no [fiduciary] shall put his [beneficiary] to the danger of such an inquiry’ into motives and fault. A Duggan, ‘Gain-Based Remedies and the Place of Deterrence in the Law of Fiduciary Obligations’ ch 15 at n 98 suggests that these high agency costs which make the defendant’s wrongdoing hard to detect are addressed in part by relaxing the burden of proof on the plaintiff. 42 Murad v Al-Saraj [2005] EWCA 959 [74]. For an older statement that an account of profits for breach of fiduciary duty is imposed ‘not only for the sake of justice in the individual case, but for the protection of the public generally’ see Benson v Heathorn (1842) 1 Y & CCC 326, 342–3 (Knight-Bruce VC). Academic writings do not question the deterrence rationale. For Canadian authorities, see Soulos v Korkontzilas [1997] 2 SCR 217; 3469420 Canada Inc v Strother (2007) 281 DLR (4th) 640 (Supreme Court of Canada). For an Australian account, see Harris v Digital Pulse Pty Ltd (2003) 56 NSWLR 298 (New South Wales CA) [414]–[415]. See Duggan, ibid. Deterrence explanations are typically also linked to efficiency arguments. Duggan argues that ‘by making the defendant more trustworthy, it increases the value of his services to the plaintiff and presumably also the amount the plaintiff is willing to pay for them’. For him, the question whether a proprietary remedy should be given instead of a personal remedy depends on deterrence benefits exceed bankruptcy costs.

Duty of Care Deterrence in A Rights-Based Theory in Private Law of Negligence

327

exercised.43 Thus, the duty to distribute trust property is designated as being a fiduciary one not because there is anything distinctive about the nature of a distribution or the nature of the property to be distributed. The fiduciary designation is rather a statement about the distinctive way the positive duty should be performed: first, that the positive duty is non-delegable and must be performed by the trustee herself; secondly, that the positive duty is assumed or undertaken for the exclusive and sole benefit of the beneficiaries and therefore must be performed for their benefit; thirdly, that as the positive duty is undertaken for another’s benefit, it is to be discharged not only according to tenor but subject to the overriding considerations or requirements of loyalty and fidelity;44 and fourthly, that no excuse is tolerated for non-discharge of the positive duty, since a fiduciary under a positive duty or obligation must do the thing she is obligated to do unless in exceptional circumstances she has a lawful ground for relief.45 Put more generally, the epithet ‘fiduciary’ indicates that the decisions and actions taken by a fiduciary when discharging her positive duty or exercising her discretion will implicate fiduciary standards of appraisal. Doubtless then, fiduciary duties are occupied solely with the fiduciary’s performance of her positive duties or discretions and are stricter than contractual duties as a result of concern over how they are performed; but how can it be said that they are imposed for the sake of deterrence? Although they are performanceduties parasitic, as it were, on positive duties undertaken by the fiduciary, they arise because they are assumed or not at all.46 According to deterrence-based rationales, the fiduciary must be compelled to perform her positive duties to the principal by removing all distractions of personal interest, especially those in the form of opportunities for gain. The problem is that the notion of an imposition of duty for the purposes of deterrence flatly contradicts the consensual nature of fiduciary duties. The reason that the fiduciary’s performance of duty is strictly

43 GMB Bean, Fiduciary Obligations and Joint Ventures: the Collaborative Fiduciary Relationship (Oxford, Clarendon Press, 1995) acknowledges that the ‘duty’ is not used strictly but refers to the fiduciary’s duty to act for the beneficiary: see n 34 at 215. In the text at the same place he states: ‘It is imperative that in acting the fiduciary should have no competing interest, either personal or to a third party, which would divert it from its purpose’. [footnotes omitted] With respect, L Smith, ‘The Motive, Not the Deed’ in J Getzler (ed), Rationalizing Property, Equity and Trusts: Essays in Honour of Edward Burn (London, LexisnexisUK, 2003) 55 provides a better explanation in terms of positive duties and the motive with which they are discharged. cf P Birks, ‘The Content of Fiduciary Obligation’ (2000) 34 Israel Law Review 3; A Burrows ‘We Do This at Common Law but That in Equity’ (2002) 22 Oxford Journal of Legal Studies 1. 44 In this sense, some have approached fiduciary obligations as disabilities because the performance of the positive duty or power is regulated by a scheme of disabilities. 45 If the fiduciary does not execute her obligation as and when she should, the aggrieved party has the right to obtain a sum of money equivalent to the advantage which the effective and exact execution of the obligation would have given her and which indemnifies her for the prejudice caused to her by the obligor’s non-execution. However, unique to the trust, the aggrieved party in a trust case is entitled to compel the timely and proper performance of that duty; the court may enforce it and may do so by removing those trustees who refuse to join in. 46 Smith, above n 43, at 61 writes: ‘The fiduciary relationship is best understood as premised on a voluntary undertaking to assume an office to which the law attaches fiduciary obligations, or outside of that, on a voluntary undertaking to put another’s interests ahead of one’s own’. (footnote omitted)

328

Yock Lin Tan

enjoined is simply that she has undertaken to act for the benefit of the beneficiary. She has undertaken not only the duty to achieve a result or carry out a task but also to do so without distraction. Her consent to do so is sufficient to ground her liability for failure to act as she has consented.

B. Systemic Deterrence as an Explanation of Fiduciary Duties If the deterrence-based explanation is right, consistency with the consensual nature of fiduciary duties requires us to restrict it to ‘knowledge-based’ deterrence. We shall have to suggest that a fiduciary who assumes a fiduciary duty is deterred by knowledge that performance of the duty is demanded in accordance with fiduciary standards. This notion of knowledge-based deterrence is problematic, since those who have assumed fiduciary duties without that knowledge should not be visited with stern application of fiduciary standards and there is no indication in the case law that knowledge of the existence of the rule of stern application makes any difference to fiduciary liability. Alternatively, we shall need to suggest that the Australian High Court is correct to say that the need for deterrence is an expression of ‘the policy of the law in holding fiduciaries to their duty’.47 Edelman argues, in similar vein, that the high degree of institutional protection afforded to fiduciary relationships can be justified as necessary to place the fiduciary on constant alert even to the possibility of innocent breach. Such vigil is necessary in relationships characterised by vulnerability and susceptibility to abuse.48

Let us understand these remarks to be advocating building in deterrence effects in the initial design of fiduciary duties so that what is permitted is the consensual assumption of such duties as have been mapped out. Deterrence arguments then are systemic arguments, they are arguments for evaluating the rules as a whole, not for evaluating individual fiduciary transactions. Deterrence might have been thus offered as a systemic argument when fiduciary duties were of less concern than fiduciary relationships. It is conceivable, if the existence of fiduciary relationships is of primary interest, that we are to attach a fixed regimen of pre-mapped fiduciary duties on consensual fiduciary relationships once these have come into being. If, however, this conception of fiduciary duties was ever held at any time, that time has long passed. No one today thinks of fiduciary duties other than as context-specific and contingent.49 For the sake of argument, let us however suppose that fiduciary duties are imposed for the sake of deterrence and ask whether there is any sensible

47

Maguire v Makaronis (1996) 188 CLR 449, 468. Edelman, above n 8, at 83–5. Lord Wilberforce in New Zealand Netherlands Society ‘Oranje’ Inc v Kuys [1973] 2 All ER 1222 (PC) said that ‘the precise scope of [the obligation not to profit] must be moulded according to the nature of the relationship’: at 1130. 48 49

Duty of Care Deterrence in A Rights-Based Theory in Private Law of Negligence

329

improvement to our understanding of how they arise and are operative. To Nolan, closer scrutiny of the body of cognate duties broadly comprehended by the doctrine of conflict of interests50 reveals not a monolithic structure but a bifurcated one.51 Some of these conflict duties (broadly designated as ‘conflict duties’) are restitutionary (such as the no self-dealing duty) and some redress wrongdoing, ie are wrongs-based.52 However, somewhat surprisingly, Nolan did not conclude thereby that the deterrence rationale was an excessive argument and should be confined to conflict duties which are wrongs-based, such as the ‘no profit’ duty and the ‘no conflict’ duty. He tacitly accepted that even the fiduciary rules which reverse unjust enrichment by subtraction are also prophylactic and proscriptive.

C. Redundancy of Deterrence Generally Taking my point of departure here, I would like to show that even among conflict duties which are (let us say) wrongs-based, such as the no profit duty and the no conflict duty, the deterrence rationale is redundant and unconvincing.53 What seems profoundly interesting is that we cannot prove that a loss occasioned by a breach of a conflict duty is recoverable for the sake of deterrence. Statements to this effect (that a fiduciary is liable for loss occasioned by breach of a conflict duty in order to deter her from being swayed by personal interest) might draw support from the cases that decide that a fiduciary must not compete with her beneficiary. Clauson J in Re Thompson54 certainly thought that the ‘no competition’ duty was part of the doctrine of conflict of interest and duty. But is the recovery of losses occasioned through breach of such duty bottomed on deterrence of wrongful conduct? Superficially, the fiduciary ought to bear the loss if she must be deterred from being swayed by interest inconsistent with duty. But it seems that, invariably, where there is loss, that loss will be directly attributable, if at all, to failure to achieve the result undertaken (such as to preserve property or to obtain the contract) or to engage in conduct for the sake of the beneficiary. Conflict of interests, if any, merely explains the provenance of the breach or supplies the motivation for 50 E Simpson, ‘Conflicts’ in P Birks and A Pretto (eds), Breach of Trust (Oxford, Hart Publishing, 2002) describes the rules preventing conflict of duty and interest as informed by an overarching conflict principle. 51 RC Nolan, ‘Conflict of Interest, Unjust Enrichment, and Wrongdoing’ in WR Cornish et al (eds), Restitution—Past, Present and Future: Essays in Honour of Gareth Jones (Oxford, Hart Publishing, 1998) ch 7. 52 cf G Jones, ‘Unjust Enrichment and the Fiduciary’s Duty of Loyalty’ (1968) 84 Law Quarterly Review 472. 53 cf M Conaglen, ‘The Nature and Function of Fiduciary Loyalty’ (2005) 121 Law Quarterly Review 452, 460, who argues that the function of fiduciary loyalty is the prophylactic protection of non-fiduciary duties. In precluding the duty of good faith and the duty to act for proper purposes from the duty of loyalty, Conaglen would appear to have left out the exercise of fiduciary powers from the duty of loyalty. 54 [1930] 1 Ch 203 (Ch) 215–16. See also LS Sealey, ‘Some Principles of Fiduciary Obligation’ [1963] Cambridge Law Journal 119, 125, fn 37.

330

Yock Lin Tan

it. If loss is caused by breach of a positive duty, it would be stretching the notion of causation to take the influence of conflicting interest as the cause of the breach. In these circumstances, the conflicting interest merely presented the occasion or provided the state of things under which the fiduciary chose to breach a positive duty owing to her beneficiary. A fiduciary who puts herself in a position of conflict does not necessarily have to submit to personal interest, and to say that the act of placing herself in that position caused the loss to the principal is a highly artificial application of the notion of causation.55 There is no addition to our understanding of conflict duties in saying this. Suppose a fiduciary competes with her beneficiary in a tender for a building contract where the fiduciary is charged to obtain a successful tender. There is no need to speak of conflict of interest and duty because any loss suffered by the beneficiary is simply recoverable on the basis of breach of the duty to obtain a successful tender.56 More generally, suppose a fiduciary is charged to obtain a designated investment for her beneficiary. If the fiduciary diverts the same investment to herself and chooses some inferior investment in which to place the fiduciary fund, so that the fund incurs an income loss, she would be liable for the loss because she has breached his duty to accomplish the required result.57 There is no need to invent a new liability based on conflict between the fiduciary’s interest and her duty to advance the interests of the principal when the loss is already recoverable through breach of her duty to accomplish the specified result. If the positive duties the fiduciary has undertaken are conduct-duties (ie she has obliged herself to engage in an endeavour, to expend certain efforts, or to conduct herself in a matter without guaranteeing a result, including exercising a power), there will be an independent sphere for performance-duties to operate, since performance-duties represent the standards of appraisal of her conduct. One could explain conflict duties as standards for identifying and gauging the materiality and strength of those interests which excuse and those which do not excuse failure of endeavour. So when a fiduciary undertakes to try and obtain a contract for her beneficiary, we will judge the quality of his efforts in accordance with the performance measures prescribed by the doctrine of conflict of interests. Unlike the case of result-duties, the deterrence rationale may appear to be no small help for judging the extent or relevance of the state of actual distraction in which the fiduciary has performed her conduct-duties. Even so, this is merely an 55 Unless what is meant is that the loss is suffered as a result of failure to disclose the existence of a conflict of interest and duty. See Huff v Price (1990) 51 BCLR (2d) 282 (British Columbia CA); Standard Investments Ltd v CIBC (1985) 52 OR (2d) 473 (Ontario CA) leave to appeal refused [1986] 1 SCR vi; Hodgkinson v Simms [1994] 9 WWR 609 (Supreme Court of Canada). cf Daly v Sydney Stock Exchange Ltd (1986) 160 CLR 371 (High Court of Australia) where stockbrokers who failed to disclose their precarious financial condition when advising a client to deposit money in their stockbroker’s account for later investment were held liable for breach of the no conflict duty. 56 As J Mowbray et al (eds), Lewin on Trusts, 17th edn (London, Sweet & Maxwell, 2000) does at 456–7. 57 That loss is not occasioned by any breach of the duty of care but by failure to accomplish the result.

Duty of Care Deterrence in A Rights-Based Theory in Private Law of Negligence

331

illusion. In all but one case, fiduciary obligations are intrinsically exclusive and considerations of extraneous interests are immaterial, making it quite irrelevant whether the fiduciary must be deterred, when performing her conduct-duty, from being swayed by this or that extraneous interest. If for instance the fiduciary should act honestly in carrying out her duty to sell the beneficiary’s property but makes false representations in doing so, resulting in loss to her beneficiary, she must bear responsibility for the loss. It matters not at all that her making the false representations was the consequence of being swayed by personal interest and that the personal interest is not such interest as will afford her an excuse for not acting honestly. The fact is that she must act honestly and that is that. Among the conflict duties, there are those which inherently express the subordination of all personal interests actually in conflict with the underlying conductduty. The duty not to compete is one of them. Compliance with these conflict duties is impossible without the fiduciary suppressing all actually conflicting interests when performing her conduct-duty. We do not need deterrence to tell us this. It is already implicit in the formulation and conception of the performanceduty which we designate the duty not to compete. If for instance the fiduciary should not compete with her beneficiary when she is undertaking a conduct-duty but she does so and this produces loss to the beneficiary, there must be recovery for the loss. The reason is simply breach of performance-duty, not deterrence from breach of duty. By definition, the fiduciary competes with her beneficiary when she fails to suppress her actually conflicting interest, and is liable for loss occasioned by breach of performance-duty. It simply does not matter whether she knew or did not know that there was an actually conflicting interest to be suppressed, so that by her failure to do so she has no excuse for not complying with his performance-duty. It results accordingly that just as loss occasioned by failure to achieve a promised result is recoverable per se, regardless of whether there is any need for deterrence, so loss occasioned by failure to carry out the endeavour in accordance with the conflict duties will be recoverable, regardless of the need for deterrence.58 Neither is deterrence necessary to protect future performance of a result-duty or conduct-duty, since the beneficiary is already entitled to an injunction to protect the future exercise of positive duty where there is some cogent evidence of a pending breach or threat to breach.59 In the exceptional instance of the no conflict duty, strictly so-called, we might more plausibly suggest that the obligation to avoid putative or potential real and sensible conflict between duty and interest rests on the need for deterrence from breach of duty. We might imagine that the conflicting duty is the performance-duty

58 The fact that it is the result of or motivated by possibly distracting interests is neither here nor there unless the notion of competition is widened to include related matters. This however would be contrary to the specificity with which fiduciary duties are formulated. 59 See Glassington v Thwaites (1823) 1 Sim & Stu 124, 133; Re Thomson [1930] 1 Ch 203 (Ch). The jurisdiction rests upon injury to property actual or prospective. See Emperor of Austria v Day (1861) 3 De GF & J 217, 253–4. See also RCA Corp v Pollard [1983] Ch 135 (CA).

332

Yock Lin Tan

(and not the positive duty, whether result-duty or conduct-duty)60 and that the no conflict duty requires that the performance-duty relevant to the underlying conduct-duty must be safeguarded in turn from putative or potential extraneous distractions. For instance, the performance-duty that we designate as the duty not to compete must be safeguarded and we shall therefore not allow the discharge of this duty to be distracted by putative or potentially conflicting extraneous interests. However, this explanatory framework also fails to be persuasive. If there is breach of the performance-duty resulting in loss, there will be liability for the loss regardless of whether the breach was facilitated by the fiduciary creating and putatively falling prey to a distraction.61 The intrinsic stringency of the performance-duty already ensures that we have no interest in potential distractions. Creating a potential distraction no more adds to liability than not doing so excuses liability. If, however, there is no breach of performance-duty, we can only speak of loss in the notional sense. This is an unintelligible sense in which we impute a loss, not in fact suffered, to the beneficiary on the basis that the fiduciary had in fact been swayed by extraneous interests and that loss had in fact occurred.62 The only intelligible sense of loss if there is no present breach of performance-duty is the future loss which will occur if the putative conflict materialises and the extraneous interest causes breach of the performance-duty in the future. However, although we might imagine the no conflict rule as prohibiting a conflict between interest and a future performance of a performance-duty, this is also unhelpful in reality. Any loss that is occasioned when at some future date the performance-duty is breached will be recoverable on the ground of breach of duty alone. If it is anticipated that that duty will be breached, injunctive relief will be available and it is not necessary to have a no conflict rule for preventive purposes.

D. Focusing on the No Conflict Duty There is equally no need to invent a new deterrence-based liability for recovery or entrapment of any gains that are made in breach of positive duties or performanceduties. Edelman has argued that the stripping of a fiduciary of profits made from a wrong serves a deterrence purpose, even when the breach of duty is not deliberate 60 It does not make much sense to suppose that the duty envisaged is the positive conduct-duty since that duty is discharged simply by engaging in it and there is never any real and sensible conflict between engaging it and a distracting interest. It is in the manner of engaging in the conduct that conflict can arise. 61 If there is non-performance of conduct altogether, there is breach of positive conduct-duty. 62 GMB Bean, above n 43, seems to be relying on notional loss when he argues (at 220): ‘If the information related to the purpose for which the fiduciary was employed then use of that information by the fiduciary could potentially damage the beneficiary, and would be a rationale for recognizing a fiduciary duty … It is no answer to say that the lack of damage is irrelevant. The no conflict rule applies without proof of damage because (1) it only applies in situations where the potential for damage exists, and (2) proof of damage is not required because the fiduciary is considered to be in a position to manipulate the facts and, hence, the beneficiary is less able to prove that the conflict caused the damage’. (footnote omitted)

Duty of Care Deterrence in A Rights-Based Theory in Private Law of Negligence

333

or reckless or calculated for gain.63 In the cases considered previously, the loss and gain could be correlative in the sense that the gain corresponds to, and hence is a measure of, the loss suffered. If the fiduciary competes with her beneficiary and causes her loss, that loss very often in practice correlates to the fiduciary’s gains. There are two other different possible outcomes, namely where the gains exceed the loss, because the fiduciary is (say) more endowed in resources and able to reap economies of scale not enjoyed by the beneficiary, and where there is no loss but only gains. However, even in these circumstances, in which all agree that the beneficiary can elect to take the profits made in breach of duty,64 the need for profitstripping for the sake of deterrence is not obvious. Rather, the fact that it would be unconscionable for a fiduciary in breach to profit from her breach sufficiently explains the fiduciary’s liability to account for these gains made from breach of performance-duty. If the fiduciary diverts the funds into investments which turn out to be highly lucrative, she is liable for the gains, for the simple reason that she has made them out of property in her charge or under her control or out of her position as fiduciary and it would be unconscionable to assert beneficial ownership of the gains.65 A fortiori if the fiduciary makes gains out of her breach of duty. Deterrence would only appear to make an independent contribution to the no conflict liability and have a non-trivial significance if it is directed at appropriating the gains made by the fiduciary to the principal when there is no breach of pre-existing performance-duty or unconscionability in the use of the principal’s property or one’s position as fiduciary.66 Edelman has argued that fiduciaries are liable to disgorge any profits made in breach of their duties, however innocently, because of the need for deterrence from the possibility of innocent breach.67 [T]he high degree of institutional protection afforded to fiduciary relationships can be justified as necessary to place the fiduciary on constant alert even to the possibility of innocent breach. Such vigil is necessary in relationships characterised by vulnerability and susceptibility to abuse.68

So soon as the need for vigil is posited, the nomenclature of the no conflict rule for entrapping gains proves to be misleading. Performance-duties other than the no conflict duty already ensure that breach which results in gains for the fiduciary will be redressed, without admitting any excuse in terms of personal interest, and I have in the foregoing paragraphs shown that the no conflict duty is furthermore not a meta-duty designed to ensure compliance with performance-duties.69 There is no pre-existing performance-duty or potential unconscionability that must be

63

Edelman, above n 8, at 83. See Dean v MacDowell (1878) 8 Ch D 345 (CA) 351. 65 This is the no profit rule which is discussed below. 66 ‘The liability arises from the mere fact of a profit having, in the stated circumstances, been made’: Regal (Hastings) Ltd v Gulliver [1967] 2 AC 134 (HL) 144. 67 Edelman, above n 8, at 85. 68 Edelman, above n 8, at 85 ibid. 69 cf Smith, above n 43; and Conaglen, above n 53. 64

334

Yock Lin Tan

brought into potential or putative conflict with an interest before the no conflict duty is attracted. The duty said to be conflicted is actually an inherent part of the no conflict duty. It is empty of meaning and the phrase ‘real and sensible conflict of duty and interest’ merely signifies a new duty having the effect of redirecting the gains made by the fiduciary to her principal. The interests which count are those which putatively occasion the redirection, the beneficiary, according to some, being relieved from having to prove that the profits were indeed made by preferring interest over duty. A very significant criticism of the no conflict duty is unavoidable and, as will be shown shortly, it removes the ground on which the deterrence rationale stands in this one area in which it appears to be meaningful. This is that the expression ‘real and sensible conflict’ is also devoid of meaning because there is nothing by which to measure the degree or extent of conflict. There is no base reference performance-duty which is possibly or potentially conflicted and against which we may assess the extent of the conflict, and the expression of real and sensible conflict is merely a vague allusion to a personal interest of the fiduciary which has surfaced and should be suppressed. The question is whether as a matter of initial design of fiduciary liabilities (or disabilities), rule-deterrence makes sense; but how is rule-deterrence a sensible notion in the absence of a reference point which can direct us to a best or most desirable and available state of affairs or option? As proved in the first section of this chapter, without a reference point for comparison, we cannot define the other persons’ harm, and any decision made that there is real and sensible conflict between duty and interest is pure semantics. It will be a conclusion without real grounds, derived from a subjective impression that the fiduciary’s gain producing activity is sufficiently similar to the general conduct or responsibilities she has undertaken for the beneficiary to justify redirecting the gains to the beneficiary. If the no conflict rule was truly bottomed on the need to deter (in the rule-deterrence sense) the fiduciary from anything less than utmost vigilance in complying with her duties for the sake of the fiduciary institution, we should be identifying not only the interest which should be suppressed but also ensuring that the fiduciary was fully aware that that interest had surfaced (although in the rule-deterrence sense, she need not know that that interest should be suppressed). We should not be presuming conclusively that the gain is made by not suppressing that interest. An actor is only specifically deterred by punishment or disincentives which she can choose to avoid. If the gain is presumed conclusively to be made once the relevant interest has been found, the punishment is not truly one that the fiduciary can entirely choose to avoid if the interest found to exist was purely coincidental and one of which she was unaware. Secondly, if we follow the suggestion for desert-based deterrence, we should be evaluating the comparative effect on the global social product, and not concentrating on the fiduciary’s lack of vigilance when we make her disgorge her gains to the beneficiary. We should, for instance, be inquiring as to whether the fiduciary has obtained something which she could put to superior use; but that is something the law will not do, though equity may sometimes temper the wind to the shorn lamb.

Duty of Care Deterrence in A Rights-Based Theory in Private Law of Negligence

335

A slight digression may help to make concrete these criticisms of a deterrencebased account of the no conflict duty. If the criticisms are sound, it follows that there is an intrinsic dissimilarity between the no profit and no conflict rules, and the former constitutes no proof that the latter exists for the sake of prevention and deterrence. Contrary to popular opinion, they do not actually overlap; or if they do, the overlap is entirely fortuitous and the result of misunderstanding the no profit rule. The assimilation of the two is typical of many pronouncements which fail to distinguish between the appropriation of property which is a consequence of a breach of duty (which is the no profit rule) and the appropriation of property which is deemed to be a breach of duty. The no conflict rule, as has been argued, determines by a kind of equivalence of conduct that the gains made out of similar or equivalent acts should be redirected, and deems the acquisition to be a breach of duty. Regretfully, this point has not been fully recognised in the case law. This is because although the statement of principle in Chan v Zacharia70 was very clear about the separation between the two rules, it did not go far enough. It explained that the no profit rule serves to disgorge personal gains secured by actual use of property or position, whereas the no conflict rule serves to deter the fiduciary from being swayed by considerations of personal interest.71 But then it went on to say that the two rules, ‘while overlapping, are distinct. Neither … fully comprehends the other’.72 These remarks could be misunderstood to mean that the no profit rule responds to actual conflict while the no conflict rule to likely conflict. Thus, Lewin on Trusts,73 dealing with the relationship in some detail, writes that what is crucial to the application of the profit rule is not that the profit is the fruit of the trust property or trusteeship, but that it is made in circumstances where the conflict exists and falls within the scope and ambit of the conflicting duty. Normally in a case where the conflict rule applies, the trustee will also have taken advantage of the trust property or trusteeship. But that will not always be so. For example, where a trustee starts up a business in competition with a business owned by the trust, but is not enabled to do so either by reason of the trust property or trusteeship, he is still accountable for the profits of his competing business.74

In other words, we are instructed that the no profit rule is engaged only when advantage is actually taken of one’s position as trustee, whereas the no conflict rule is engaged as long as there may be a conflict between interest and duty.75 These explanations very quickly suggest that the no conflict rule is an extension 70 Chan v Zacharia (1984) 154 CLR 178 (High Court of Australia). See also Warman International Ltd v Dwyer (1995) 69 ALJR 362 (High Court of Australia) 368; Don King Productions Inc v Warren [1999] 3 WLR 276 (CA). 71 See also A J McLean, ‘The Theoretical Basis of the Trustee’s Duty of Loyalty’ (1969) 7 Alberta Law Review 218, 219. 72 Chan v Zacharia (1984) 154 CLR 178 (High Court of Australia). 73 Above n 56. 74 ibid, at 448–9. 75 ibid, at 456–7.

336

Yock Lin Tan

of the no profit rule for preventive and deterrent purposes; the fiduciary must be prevented from making gains out of a real and sensible conflict to instill vigilance against, and to prevent her from, sliding into the act of making gains out of actual conflicts.76 The illusion of overlap is amplified by judicial moves which have expanded the notion of property to cover information capable of immediate exploitation77 and to include the position occupied as fiduciary.78 One thing should at least be made clear, namely that the notion of profiting from the position occupied as fiduciary must be a narrow one. It only contributes to confusion to give the notion of profiting from one’s position a wide and vague meaning, going far beyond the notion of profiting from the principal’s property, to include doing anything at all whilst also happening to occupy the position of a fiduciary. Properly confined, the notion implies that a fiduciary does not make a profit from her position as such unless her position as such is not just the occasion but the raison d’etre for profit making.79 Thus, it is not sufficient that profits are made from information acquired during the currency of a fiduciary relationship; rather the information must also have been obtained for the sake of, and for purposes within the scope of, the fiduciary relationship.80 (On the other hand, if as in Reading v AttorneyGeneral,81 the principal is not able to benefit from using its property in the way that the fiduciary did, this will be inconsequential because the property is in fact used for gain.) From this, it follows that if the principal is not able to benefit directly or indirectly from using that information, the fiduciary cannot be said to have profited from her position.82 76 Thus JC Shepherd, The Law of Fiduciaries (Toronto, Carswell, 1981) 150 repeats that the no conflict rule and the no profit rule are ‘precisely the same thing when taken to their logical conclusion: a fiduciary is liable for choosing against the interests of his beneficiary’. JRM Gautreau, ‘Demystifying the Fiduciary Mystique’ (1989) 68 Canadian Bar Review 1, 21 sees that the same need to dissuade a fiduciary from being swayed by personal interests underscores the no profit rule. If the fiduciary cannot profit, she will not be swayed. 77 See Ultraframe (UK) Ltd v Fielding [2006] FSR 17 (Ch); Foster Bryant Surveying Ltd v Bryant [2007] 2 BCLC 239 (CA). 78 See Docker v Somes (1834) 2 My & K 655; Aberdeen Town Council v Aberdeen University (1877) LR 2 App Cas 544 (HL); Nordisk Insulinlaboratorium v Gorgate Products [1953] 1 All ER 986 (CA). Bean, above n 43, at 249 takes the view that the misuse of beneficiary’s property rule is distinct from the no profit and no conflict rules. 79 See Deane J in Chan v Zacharia (1984) 154 CLR 178 (High Court of Australia) 198 who refers to ‘by use or reason of ’ of the fiduciary position. I would argue that the phrase ‘misuse or abuse of position’ is actually referring to breach of the performance-duties arising out of the fiduciary’s position. 80 See Aas v Benham [1891] 2 Ch 244 (CA). See also Trimble v Goldberg [1906] AC 494 (PC). 81 Reading v A-G [1951] AC 507 (HL). 82 Suppose the fiduciary diverts the principal’s funds to an inferior investment but places her own funds in that investment she would otherwise have selected for the principal. She can be said to have profited from her position, though she has not used the principal’s funds to make a profit for her own purposes. The principal could have benefited from the investment and the fiduciary was to put the principal’s money in the investment. If, however, the principal is barred from making the investment and has not otherwise determined to keep the information from others, it cannot be said that the fiduciary has used her position to make profits by using the information. cf Diamond v Oreamuno 301 NYS 2d 78 (New York CA, 1969) 80–81: ‘It is well established, as a general proposition, that a person who acquires special knowledge or information by virtue of a confidential or fiduciary relationship with

Duty of Care Deterrence in A Rights-Based Theory in Private Law of Negligence

337

The above submissions are derived from appreciating that the key to the no profit rule lies in the concept of unconscionable conduct. A fiduciary who makes a profit out of the principal’s property or her position holds it on constructive trust, regardless of what she intended,83 because it would be objectively unconscionable for her to assert a beneficial title to those profits; the objective appraisal making an inquiry into motives and competing considerations irrelevant. In contrast, the no conflict rule predicates a fiduciary who places her own funds in a venture in circumstances in which she cannot be said to have breached any specific performance-duty owed to her principal. On this view, the no profit rule is completely dissimilar to the no conflict rule84 and is not about preventing the fiduciary from slipping into an actual conflict of duty and interest. The no conflict rule, properly understood, is of very limited scope and must be singled out from all the other rules about the performance of a positive duty of a fiduciary, including the no profit rule, which is based on unconscionable conduct in relation to property acquisition. It is not a meta-performance-duty designed to ensure the undistracted performance of fiduciary duties such as the no profit duty. Neither is it an extension of the no profit duty, since we cannot properly speak of extending a rule about unconscionably using the principal’s property to a rule which is not about the principal’s property. Rather, it is a decision that profits made from acts similar to those undertaken must be redirected to the beneficiary. If the no conflict rule, as narrowed down, is prophylactic, then it is the undertaking of acts similar to that already undertaken (and not the sliding temptation) which must be deterred. There are formidable difficulties against such a deterrence-based explanation of the no conflict rule. I have already shown that the irrelevance of whether the fiduciary suppressed the potential interest negatives the suggestion that the fiduciary must be deterred from being swayed by conflicting interests in performing her duty, whatever that may be. A desertbased deterrence fails to explain the no conflict rule. In sketching the differences another is not free to exploit that knowledge or information for his own personal benefit but must account to his principal for any profits derived therefrom… The primary concern, in a case such as this, is not to determine whether the corporation has been damaged but to decide, as between the corporation and the defendants, who has the higher claim to the proceeds derived from the exploitation of the information. In our opinion, there can be no justification for permitting officers and directors, such as the defendants, to retain for themselves profits which, it is alleged, they derived solely from exploiting information gained by virtue of their inside position as corporate officials’. 83 See Gordon v Gonda [1955] 2 All ER 762 (CA). See also Helmore v Smith (No 1) (1887) 35 Ch D 436 (CA). 84 This clarification is also different from that offered by Nolan—‘[f]irst, a fiduciary cannot on his own account, take any benefit which his undertaking requires or authorises him to obtain or pursue in his representative capacity. Secondly, a fiduciary, even though acting outside the scope of his undertaking, cannot take a benefit made through actual misuse of his representative position … the predominant aim of the first rule is to obviate any conflict of duty and interest which might tempt a fiduciary away from the loyalty to his principal, while the primary aim of the second rule is to counter the risk that a fiduciary will be disloyal to his principal if he can make some incidental profit out of his activities as a fiduciary’: see Nolan, above n 51, at 100.

338

Yock Lin Tan

between the no profit rule and the no conflict rule, I was concerned to show that the no conflict rule is not about deterring the fiduciary from sliding into actual conflict and thereby causing loss to the beneficiary. Even if it were, it would be about building in deterrent considerations in an initial framework to augment or supplement or amend other rules within the framework. That is hardly true deterrence but is more an argument for efficiency. If we must deter a fiduciary from undertaking similar acts to that already undertaken regardless of whether she is under any pre-existing duty, we must be able to point to some other harm to a class of unidentifiable persons which must be protected by levying a deterrence tax. We are, however, unable to identify any other group that must be protected from harm. Although the existence of similar beneficiaries could constitute a class whose interests could justify imposing a deterrence tax, every similar beneficiary is already protected by existing performance-duties and the warrant for further protection is absent. If the beneficiary is made up of existing persons and future (unborn) persons, it may be possible to argue that a deterrence tax is justifiable for the protection of unborn beneficiaries whose interests will not be fully looked after by existing beneficiaries. But this possibility has never been adverted to in the case law and would in any case only be true of certain trustees. In fact, deterrence-based explanations are as uneasy as they are unnecessary. In my view, the no conflict rule is really a way of regulating the future employment of the fiduciary. Unless she is a trustee, the fiduciary cannot be prevented from resigning her position and undertaking acts similar to those she has previously undertaken for the beneficiary, even though such resignation may be an inconvenience and more importantly deprive the beneficiary of the fiduciary’s valuable contributions in future. The no conflict duty with its prospect of redirection of the gains made from similar acts strikes a compromise between the future needs of the fiduciary and beneficiary. It serves to preclude sudden resignations and leaving the beneficiaries in the lurch after an amount of spade-work has been done. If there is any idea of deterrence at all behind the no conflict rule, it is that the no conflict duty is a restraint or disincentive to sudden resignations in order to pursue similar acts or an incentive to the fiduciary to continue to serve her principal for as long as this is consistent with her personal future interests. But then these matters hardly belong to the realm of deterrence from other persons’ harm which would warrant a tax imposition and should be resolved on a balance of fairness, of future costs and benefits. From the point of view of rationale, a rigid rule approach such as is typical of a tax and exemplified in Boardman v Phipps85 is misconceived. The no conflict duty must express a balance of fairness through an open ended inquiry into the fairness of appropriating an opportunity for gain undertaken by a fiduciary at her own expense and risk to the principal.86 85

Boardman v Phipps [1967] 2 AC 46 (HL). See also Canadian Aero Services Ltd v O’Malley (1973) 40 DLR (3d) 371 (Supreme Court of Canada) 391, where it was said: ‘The general standard of loyalty, good faith and avoidance of a conflict of duty and self-interest to which the conduct of a director or senior officer must conform, must be 86

Duty of Care Deterrence in A Rights-Based Theory in Private Law of Negligence

339

VI. Conclusion I would argue that deterrence arguments are in reality arguments for a tax or transfer of wealth as a notional measure of ‘other person’ harm. This is an ambitious argument and if I am taxed with an excessive flight of imagination, I would plead that it is not without some benefits if the result is that we begin balancing the so-called real and sensible conflict of duty and interest. The no conflict duty I would defend is not the rigid and inflexible one implied by the real and sensible conflict test but one determined by a more open-ended test that gives each relevant factor its due weight when striking the balance of fairness. The reason is that deterrence is here immaterial.

tested in each case by many factors which it would be reckless to enumerate exhaustively. Among them are the factor of position or office held, the nature of the corporate opportunity, its ripeness, its specificness and the director’s relation to it, the amount of knowledge possessed, the circumstances in which it was obtained, whether it was special or, indeed, even private, the factor of time in the continuation of the fiduciary duty after termination of the relationship with the company, and the circumstances under which the relationship was terminated, that is whether by retirement or resignation or discharge’.

14 Justifying Fiduciary Allowances MATTHEW HARDING*

I. Introduction Where a defendant who is a defaulting fiduciary has profited in breach of fiduciary obligation, she will be made to disgorge that profit, typically by way of an order to pay to the claimant a sum of money following the taking of an account in order to work out what profit was made.1 Although, strictly speaking, the taking of an account is conceptually distinct from the order to pay by which disgorgement is achieved once the account is taken, the account-taking process and the order to pay are usually described together as the equitable remedy of ‘account of profits’. In some disgorgement cases where the remedy of account of profits has been awarded, courts in England and elsewhere have made what is typically described as an ‘equitable’ or ‘just’ ‘allowance’ in favour of a defendant who, in breach of fiduciary obligation, has expended time and skilled labour in generating the profit that stands to be disgorged.2 This allowance has the effect of reducing the sum of money that the defendant is ordered to disgorge to the claimant after the account has been taken. It therefore benefits the defendant. In this essay, I seek to understand what might justify making an allowance, thereby benefiting the defendant, notwithstanding that, in a case where an allowance is made, the defendant has necessarily acted in breach of fiduciary obligation. My thesis is that the justification of allowances is to be found in the defendant’s desert. The essay is laid out in five parts. In Section II, I shall argue

* My thanks go to Elise Bant, Michael Bryan and Andrew Robertson for helpful comments and suggestions. 1 Disgorgement may also be achieved by the declaration and enforcement of a constructive trust of specific assets. For convenience, I refer in the text only to the account of profits although, as below n 25 shows, allowances may be made in disgorgement cases where a constructive trust is declared. 2 Re Macadam [1946] 1 Ch 73 (Ch); Re Jarvis (decd) [1958] 1 WLR 815 (Ch); Phipps v Boardman [1964] 1 WLR 993 (Ch); DPC Estates Pty Ltd v Grey and Consul Developments Pty Ltd [1974] 1 NSWLR 443 (New South Wales CA); O’Sullivan v Management Agency and Music Ltd [1985] 1 QB 428 (CA); Estate Realties Ltd v Wignall [1992] 2 NZLR 615 (New Zealand High Court); Warman International Ltd v Dwyer (1995) 182 CLR 544 (High Court of Australia); Murad v Al-Saraj [2005] EWCA Civ 959; Chirnside v Fay [2006] NZSC 68.

342

Matthew Harding

that the making of allowances demands justification in light of the aim of the account of profits in disgorgement cases, which is to meet the goal of deterrence. In Section III, I shall consider two possible explanations of the role of allowances in disgorgement cases: one that ties allowances to an inquiry into causation; and another that ascribes to allowances the function of a rule of remoteness. In Section IV, I shall argue for a justification of allowances based on a goal that is properly pursued by the law in disgorgement cases notwithstanding the goal of deterrence. That goal is the goal of satisfying desert, or, to put it another way, the goal of giving to people what they deserve. Finally, in Section V, I shall argue that the concept of desert assists in explaining why an allowance must be withheld from a dishonest defendant, and I shall consider when and on what basis allowances might be limited or withheld even where the defendant has been honest.

II. Allowances Demand Justification A. The Aim of the Account of Profits That the making of allowances demands justification may be seen in light of the aim of the account of profits in disgorgement cases. The identity of that aim has been the topic of lively recent debate, and in particular it has been suggested that the aim of the account of profits might be to compensate loss. I do not wish to pursue the arguments for and against that counter-intuitive claim.3 Rather, I wish to proceed on what I take to be the mainstream assumption that the aim of the account of profits in disgorgement cases is deterrence.4 In making this assumption, I accept the proposition that the award of an account of profits in a particular case aims to deter the defendant from breaching her fiduciary obligations in the future and aims to deter other fiduciaries from breaching their fiduciary obligations as well. I also accept the related proposition that deterrence is an appropriate

3 To gain a sense of the debate, see C Rotherham, ‘ “Wrotham Park Damages” and Accounts of Profits: Compensation or Restitution?’ [2008] Lloyd’s Maritime and Commercial Law Quarterly 25. 4 J Edelman, Gain-Based Damages: Contract, Tort, Equity and Intellectual Property (Oxford, Hart Publishing, 2002) 83–5, 212; M McInnes, ‘Account of Profits for Common Law Wrongs’ in S Degeling and J Edelman (eds), Equity in Commercial Law (Sydney, Law Book Company, 2005) 428; M Conaglen, ‘The Nature and Function of Fiduciary Loyalty’ (2005) 121 Law Quarterly Review 452, 463; J Edelman, ‘Gain-Based Damages and Compensation’ in A Burrows and A Rodger (eds), Mapping the Law: Essays in Memory of Peter Birks (Oxford, Oxford University Press, 2006) 147–50; J Edelman and E Bant, Unjust Enrichment in Australia (Melbourne, Oxford University Press, 2006) 45–8; S Worthington, Equity, 2nd edn (Oxford, Oxford University Press, 2006) 131, 134; A Duggan, ‘Solicitors’ Conflict of Interest and the Wider Fiduciary Question’ (2007) 45 Canadian Business Law Journal 414, 421–2; A Duggan, ‘Gains-Based Remedies and the Place of Deterrence in the Law of Fiduciary Obligations’ ch 15 of this book. See also Harris v Digital Pulse Pty Ltd (2003) 56 NSWLR 298 (New South Wales CA) [161] (Mason P), [306], [414] (Heydon JA).

Justifying Fiduciary Allowances

343

goal for the law to pursue in disgorgement cases. In other words, I accept that the goal of deterrence not only explains but also justifies the award of an account of profits.5 As James Edelman points out, deterrence usually justifies the award of a disgorgement remedy, like an account of profits, where what is being deterred are ‘cynical’ breaches of obligation. However, Professor Edelman argues that, in cases of breach of fiduciary obligation, deterrence may justify disgorgement even where what are being deterred are ‘non-cynical’ breaches: ‘[disgorgement] is justified in these cases because the trust and vulnerability inherent in the fiduciary relationship requires greater protection of the rights of the principal by a higher degree of deterrence from breach’.6 The aim of the account of profits being deterrence, it follows that the more a defendant is required to disgorge upon the taking of an account, the more fully the aim of the account of profits is achieved. The goal of deterrence depends on a defendant who has been made to disgorge profit in a past case taking that fact to be a reason against acting in breach of fiduciary obligation in the future. It also depends on other persons taking that fact to be a reason against acting in breach of fiduciary obligation in analogous cases. Like all reasons, the fact that a defendant has been made to disgorge profit in a past case may have a greater or lesser degree of weight in the practical reasoning of a person—whether it be that defendant or someone else—who takes it to be a reason for action.7 So, the weightier the fact of past disgorgement grows as a reason against acting in breach in an analogous case, the more fully is the goal of deterrence met, in the sense that the goal is more likely to be achieved in a larger number of cases. A person who responds properly to reasons will, all else being equal, treat the fact that a defendant was ordered to disgorge 100 per cent of her profit as a weightier reason against acting in breach in an analogous case than she would treat the fact that the defendant was ordered to disgorge 50 per cent or 80 per cent of her profit. She is thus more likely to refrain from acting in breach; she will be deterred from so acting to a greater extent than she otherwise would have been.8

5 In this, I reject the view that private law remedies are justified only to the extent that they are informed by the requirements of corrective justice. For the best-known exposition of this view, see EJ Weinrib, The Idea of Private Law (Cambridge MA, Harvard University Press, 1995). See also L Smith, ‘The Motive, Not the Deed’ in J Getzler (ed), Rationalizing Property, Equity and Trusts: Essays in Honour of Edward Burn (London, LexisNexis, 2003) 53. 6 Edelman, ‘Gain-Based Damages and Compensation’, above n 4, at 149–50. 7 For the classic account of how reasons have the dimension of weight or strength, see J Raz, Practical Reasons and Norms (Oxford, Oxford University Press, 1999) ch 1. 8 Note that meeting the goal of deterrence does not necessarily require that a person actually refrain from acting in breach; the goal is met once a person is discouraged from acting in breach in the sense that she is given additional or weightier reasons against so acting than she otherwise would have had. The fact of past disgorgement will cause a person to refrain from acting in breach only where it outweighs the reasons for so acting. This may not occur if the risk of getting caught is low and the profit to be made is substantial. Deterrence is but one strategy that may be deployed in pursuant of the ultimate goal of preventing breaches of fiduciary obligation from occurring: others may include encouraging a culture of professional ethics, supervising fiduciaries, and imposing formalities to make breaches difficult to achieve.

344

Matthew Harding

B. Justifying Allowances Given what is entailed in meeting the goal of deterrence, it may be seen why allowances demand justification. In a case where an allowance is made in favour of the defendant, the defendant will be made to disgorge less profit than she would have had to disgorge had the allowance not been made. The fact of the defendant’s disgorgement will therefore carry less weight as a reason against acting in breach of fiduciary obligation in future cases than it would have carried had the allowance not been made. To this extent, the goal of deterrence is compromised. In Guinness plc v Saunders, Lord Goff suggested that allowances might be justified to the extent that they do not conflict with the goal of deterrence: [A]s I see it, a conflict … will only be avoided if the exercise of the jurisdiction [to make allowances] is restricted to those cases where it cannot have the effect of encouraging trustees in any way to put themselves in a position where their interests conflict with their duties as trustees.9

I hope to have said enough to show that, given what is entailed in meeting the goal of deterrence, the making of allowances always has the effect to which Lord Goff alluded and therefore always conflicts with the goal of deterrence.10 It is simply not possible to argue that allowances are justified to the extent that they are consistent with the goal of deterrence unless one is prepared to conclude that allowances are never justified,11 a conclusion that Lord Goff did not endorse in Guinness plc v Saunders.12 Notwithstanding this, it is possible to argue that allowances are justified despite the fact that they always conflict with the goal of deterrence. Such an argument may point to a goal, other than deterrence, that it is appropriate for the law to pursue in disgorgement cases and that, in some cases, may override the goal of deterrence. In Section IV of this essay, I attempt such an argument, pointing to the goal of satisfying desert by way of justification. However, before attempting that argument, I wish to consider two accounts of allowances that seek their justification not in any goal that the law pursues in disgorgement cases, but instead in widely-accepted limitations on liability in private law generally. These limitations are established: first, by the requirement that there be a causal connection between a breach of obligation and the consequences in respect of which liability for breach is to be imposed; and secondly, by the rule according to which liability ought not to be imposed in respect of consequences that are deemed too remote from a breach of obligation, even if they are causally connected to that breach. 9

Guinness plc v Saunders [1990] 2 AC 663 (HL) 700 (Lord Goff). A point also made by L Aitken, ‘Reconciling “Irreconcilable Principles”—A Revisionist View of the Defaulting Fiduciary’s “Generous Equitable Allowance” ’ (1993) 5 Bond Law Review 49, 61; J Palmer, ‘The Availability of Allowances in Equity: Rewarding the Bad Guy’ (2004) 21 New Zealand Universities Law Review 145, 157. 11 This is indeed the conclusion that Ms Palmer reaches: ibid, at 172. 12 His Lordship described the jurisdiction to make allowances as ‘restricted’, not ‘non-existent’: Guinness plc v Saunders [1990] 2 AC 663 (HL) 700 (Lord Goff). 10

Justifying Fiduciary Allowances

345

III. Allowances, Causation and Remoteness A. Causation On one view, making an allowance is a strategy deployed by the court in order to identify profit which the defendant has made since her breach of fiduciary obligation but which is insufficiently causally connected to the breach to require disgorgement. Another such strategy is to apportion the profit made by the defendant since her breach based on what the claimant and the defendant had previously agreed to in an antecedent profit-sharing arrangement.13 The strategy of apportionment depends for its justification on the idea that the claimant, in light of the profit-sharing arrangement, would have consented to the defendant having some of the profit if the defendant had sought the claimant’s consent to making it. To that extent, ‘but for’ causation, linking the breach and the profit, is not established. According to those who would place allowances alongside the strategy of apportionment, allowances are also a means by which the court identifies profit that is not sufficiently causally connected to the defendant’s breach of fiduciary obligation. However, the best interpretation of this view of allowances is that it is not concerned with whether ‘but for’ causation is established. Rather, it is concerned with whether, to the extent that a defendant has made profit through expenditure of time and skilled labour, it is possible to describe that defendant’s breach of fiduciary obligation as a cause of her profit. This is the best interpretation of this view of allowances because, except where there has been an antecedent profit-sharing arrangement, ‘but for’ causation is not applied in disgorgement cases. Instead, the causation inquiry seeks to identify a basic factual connection between the defendant’s breach and the profit.14 The High Court of Australia appears to have viewed allowances as tied to a causation inquiry in Warman International Ltd v Dwyer, especially in the following passage, which is worth setting out in full. [I]t may be appropriate to allow the fiduciary a proportion of the profits, depending on the particular circumstances. That may well be the case when it appears that a significant proportion of an increase in profits has been generated by the skill, efforts, property and resources of the fiduciary, the capital which he has introduced and the risks he has taken, so long as they are not risks to which the principal’s property has been exposed. Then it may be said that the relevant proportion of the increased profits is

13 Warman International Ltd v Dwyer (1995) 182 CLR 544 (High Court of Australia) 562; Murad v Al-Saraj [2005] EWCA Civ 959 [160] (Clarke LJ); Chirnside v Fay [2006] NZSC 68 [36] (Elias CJ). 14 Regal (Hastings) Ltd v Gulliver [1967] 2 AC 134 (HL) 144–5 (Lord Russell of Killowen), 153 (Lord Macmillan); Chan v Zacharia (1984) 154 CLR 178 (High Court of Australia) 199 (Deane J); Murad v Al-Saraj [2005] EWCA Civ 959 [72]–[79] (Arden LJ). See also the excellent discussion in C Mitchell, ‘Causation, Remoteness, and Fiduciary Gains’ (2006) 17 King’s College Law Journal 325, 331–6, and the criticism of the current law in R Cunnington, ‘The Assessment of Gain-Based Damages for Breach of Contract’ (2008) 71 Modern Law Review 559, 582–4.

346

Matthew Harding

not the product or consequence of the plaintiff ’s property but the product of the fiduciary’s skill, efforts, property and resources.15 (emphasis added)

The same thinking is evident in the work of James Edelman and Graham Virgo.16 Professor Edelman suggests that allowances ensure that ‘only the profits flowing from the wrong are disgorged’.17 And Professor Virgo argues that the equitable allowance seeks to apportion profits so that the principal only recovers those profits which derive from the breach of fiduciary duty and the fiduciary is allowed to retain those profits which can be considered to derive from his or her work and skill.18

Elsewhere, Professor Virgo states that ‘the existence of the allowance is … preferably analysed as being grounded on principles of causation’.19 The view of allowances that ties them to a causation inquiry in disgorgement cases, although it enjoys the endorsement of the High Court of Australia and distinguished scholars, appears to suffer from two shortcomings. First, it appears to be inconsistent with the practices of courts when making allowances. Secondly, and more seriously, it is based on a misunderstanding of what constitutes a breach of fiduciary obligation in a disgorgement case. I will consider each of these shortcomings in turn. The case law reveals clearly that allowances are typically made when taking the account that will underpin the court’s disgorgement order.20 The purpose of taking this account is to ‘expose the defendant’s financial affairs insofar as they pertain to the claimant’, thus providing an evidentiary basis for the disgorgement order.21 Specifically, courts have factored allowances into this preliminary account-taking exercise in at least two ways, both in cases of breach of fiduciary obligation and in older cases, often cited in fiduciary cases, where a partner has had to account to the estate of his deceased partner for profit derived from the partnership business after that deceased partner’s death. First, courts have permitted, and sometimes required, the entry of specific items as disbursements on the account. These items include expenditures of money or other capital,22 as well as expenditures of time and

15

Warman International Ltd v Dwyer (1995) 182 CLR 544 (High Court of Australia) 561. Edelman, Gain-Based Damages, above n 4, at 104–5 and 171–2; G Virgo, The Principles of the Law of Restitution, 2nd edn (Oxford, Oxford University Press, 2006) 516–19; G Virgo, ‘Restitutionary Remedies for Wrongs: Causation and Remoteness’ in C Rickett (ed), Justifying Private Law Remedies (Oxford, Hart Publishing, 2008) 301, 309–10 and 325–6. 17 Edelman, Gain-Based Damages, above n 4, at 172. 18 Virgo, The Principles of the Law of Restitution, above n 16, at 518. 19 Virgo, ‘Restitutionary Remedies for Wrongs’, above n 16, at 310. 20 On the nature and function of the account, see R Grantham and C Rickett, Enrichment and Restitution in New Zealand (Oxford, Hart Publishing, 2000) ch 20; P Millett, ‘Proprietary Restitution’ in S Degeling and J Edelman (eds), above n 4; McInnes, above n 4. 21 McInnes, above n 4, at 407. 22 Phipps v Boardman [1964] 1 WLR 993 (Ch) 1018; DPC Estates Pty Ltd v Grey and Consul Developments Pty Ltd [1974] 1 NSWLR 443 (New South Wales CA) 473 (Hutley JA); Murad v Al-Saraj [2005] EWCA Civ 959 [89] (Arden LJ); Chirnside v Fay [2006] NZSC 68 [21] (Elias CJ). 16

Justifying Fiduciary Allowances

347

skilled labour by the defendant,23 perhaps quantified with reference to an hourly rate of pay.24 Secondly, courts have, when taking the account, credited the defendant with an allowance that is calculated, not with reference to specific items of expenditure, but rather in an ‘all things considered’ way.25 An allowance that is calculated in this second way may capture more than items of expenditure, permitting the defendant to take a further share of profit as well.26 What is important to note about these two ways in which courts have factored allowances into the account-taking process is that courts have not viewed them as ways of working out what proportion of profit made is sufficiently causally connected to the breach of fiduciary obligation to require disgorgement. Rather, courts have viewed them as ways of working out what, for the purposes of disgorgement, counts as profit in the first place. Another way of putting this point is to say that courts have required that the defendant must disgorge, not a proportion of the gross profit, but all of the net profit that is revealed upon the taking of the account.27 And because the practice of courts is, on the taking of the account, to deduct allowances from the gross profit that has been derived as a consequence of the breach, the role of allowances has been in helping to ascertain what the net profit to be disgorged is.28 Given this, it cannot be said that courts, in making allowances, have found a causal link between breach and profit to be lacking; instead the proper conclusion is that courts have found that, to the extent of the allowance, profit that has been made as a consequence of breach of fiduciary obligation does not count for the purposes of disgorgement. In addition to being inconsistent with the practices of courts, the view of allowances that treats them as the product of an inquiry into causation suffers from a more serious shortcoming: it is based on a misunderstanding of what constitutes a breach of fiduciary obligation in a disgorgement case. Consider a typical disgorgement case in which the defendant, presented with a business opportunity that 23 In some of the early partnership cases, allowances were permitted for expenditures of money but not for expenditures of time and skilled labour: Burden v Burden (1813) 1 V & B 170; Stocken v Dawson (1848) 9 Beav 239. However, these cases were exceptional and have not been relied on in breach of fiduciary obligation cases. Cases in which courts appear to have countenanced entering expenditures of time and skilled labour as disbursements on the account include: Brown v Litton (1711) 1 P Wms 140; Yates v Finn (1880) 13 Ch D 839 (Ch); Chirnside v Fay [2006] NZSC 68 [153] (Tipping J). 24 Chirnside v Fay [2006] NZSC 68 [153] (Tipping J). 25 Brown v De Tastet (1821) Jac 284; Featherstonhaugh v Turner (1858) 25 Beav 382; Lord Provost of Edinburgh v Lord Advocate (1879) 4 App Cas 823 (HL); Phipps v Boardman [1964] 1 WLR 993 (Ch); O’Sullivan v Management Agency & Music Ltd [1985] 1 QB 428 (CA). Such an ‘all things considered’ allowance appears also to have been made or at least contemplated in cases where the disgorgement remedy in view was not an account of profits but rather the declaration and enforcement of a constructive trust of specific assets: Re Jarvis (decd) [1958] 1 WLR 815 (Ch); Paul A Davies (Australia) Pty Ltd v Davies [1983] 1 NSWLR 440 (New South Wales CA); United States Surgical Corporation v Hospital Products Ltd [1983] 2 NSWLR 157 (New South Wales CA). 26 O’Sullivan v Management Agency & Music Ltd [1985] 1 QB 428 (CA). 27 ‘[T]he concept of profit here refers not [to] the gross value of receipts, but to the margin of receipts over costs’: Grantham and Rickett, above n 20, at 466. 28 ‘The Court would have to ascertain the amount of the net profits made by the concern (after making a liberal allowance)’: Featherstonhaugh v Turner (1858) 25 Beav 382, 389.

348

Matthew Harding

he would not have been presented with had he not been in a fiduciary position, diverts the opportunity from his principal to himself and develops it profitably over some months or even years. When considering this case, it is unnecessary and inconsistent with principle to divide the defendant’s actions over the course of time into the categories of ‘breach of fiduciary obligation’ and ‘no breach of fiduciary obligation’, and then to assert that profit flowing from actions in the latter category—including actions that take the form of expenditure of time and skilled labour—may be retained by the defendant because such profit has not been caused by a breach of fiduciary obligation. To divide up the defendant’s actions in this way is unnecessary and inconsistent with principle because of what constitutes a breach of fiduciary obligation in a disgorgement case. In our hypothetical case, according to the ‘no profit’ rule, all profit that the defendant makes that may be attributed to his initial misuse of the fiduciary position—diverting the business opportunity to himself—must be disgorged.29 Implied by the fact that the ‘no profit’ rule requires disgorgement in these circumstances is the assumption that the initial misuse of fiduciary position itself suffices as a breach of fiduciary obligation.30 The defendant’s conduct since that misuse of position is irrelevant to establishing the causal connection between the misuse of position and the profit that the defendant has subsequently made. If that conduct were to be relevant to the question of causation, it could only be relevant according to some doctrine of intervening cause, in which case it might break the chain of causation between the misuse of position and the profit. However, it has never been suggested that such a doctrine has any application in disgorgement cases.

B. Remoteness Another strategy that courts deploy in disgorgement cases is to fix a date on the calendar after which any profit is deemed to be too remote from the defendant’s breach of fiduciary obligation and therefore does not require disgorgement.31 The justification of this strategy appears to be that profit need not be disgorged, 29 Regal (Hastings) Ltd v Gulliver [1967] 2 AC 134 (HL) 144–5 (Lord Russell of Killowen), 153 (Lord Macmillan); Chan v Zacharia (1984) 154 CLR 178 (High Court of Australia) 199 (Deane J); Murad v Al-Saraj [2005] EWCA Civ 959 [72]–[79] (Arden LJ); Cunnington, above n 14, at 580 (discussing apportionment) and 581 (discussing allowances). Note also that, according to the ‘no conflict’ rule, a breach of fiduciary obligation actually occurs at the first moment when there is a real possibility that the defendant’s personal interest in diverting the business opportunity will come into conflict with one or more of the contractual (and perhaps other) obligations that he owes to his principal: Boardman v Phipps [1967] 2 AC 46 (HL). This moment is likely to precede the misuse of the fiduciary position. However, it is unnecessary for present purposes to pursue this point. 30 If one prefers not to use the language of ‘breach of obligation’ to describe a fiduciary’s accountability under the ‘no profit’ rule, then this idea might be expressed as follows: a fiduciary must account for all profit made qua fiduciary, and all profit that is causally connected with the initial misuse of position is made qua fiduciary. For a discussion of the view that the ‘no profit’ rule establishes a liability to account and not an obligation breach of which constitutes a civil wrong, see Mitchell, above n 14, at 329–30. 31 Warman International Ltd v Dwyer (1995) 182 CLR 544 (High Court of Australia) 561.

Justifying Fiduciary Allowances

349

notwithstanding that it is causally connected to the breach, where, according to a rule of remoteness, it is deemed to be too remote from that breach. Some scholars, including Andrew Burrows, Ross Grantham and Charles Rickett, take the view that the making of allowances, like the fixing of a date after which profit is deemed to be too remote, reflects the operation of such a rule of remoteness.32 These scholars accept that there is a causal connection between profit that is captured by an allowance and the breach of fiduciary obligation as a consequence of which that profit was made.33 However, they point out that if only a causal connection between breach and profit were required before disgorgement could be ordered, disgorgement would be ‘potentially unlimited’, and might bring about hardship for the defendant:34 hence the need for a rule of remoteness, whose function is performed—in part—by the making of allowances.35 The view of allowances put forward by Professors Burrows, Grantham and Rickett avoids the shortcomings of the view that ties allowances to an inquiry into causation. First, the view of these scholars is consistent with the practices of courts when making allowances. One may argue that the function of allowances is to reflect the operation of a rule of remoteness while recognising that, in practice, the making of allowances goes towards the calculation of net profit to be disgorged upon the taking of the account. By contrast, as I argued above, the view that allowances are tied to a causation inquiry is inconsistent with that practice. Secondly, pointing to a rule of remoteness as the functional explanation of allowances may be reconciled with a correct understanding of what constitutes a breach of fiduciary obligation in a disgorgement case. Applying a rule of remoteness does not entail dividing up a defendant’s actions into the categories of ‘breach of fiduciary obligation’ and ‘no breach of fiduciary obligation’ so that the profit causally connected to actions in each category may be identified. A rule of remoteness is properly applied after the causal connection between the breach—in the form of the initial misuse of fiduciary position—and profit is established; it is not applied in order to establish that connection. However, while the idea that allowances perform the function of a rule of remoteness avoids certain problems, it actually tells us nothing meaningful about how allowances may be justified. Even if we accept that the function of allowances is to identify profit that is deemed to be too remote from the defendant’s breach of fiduciary obligation to require disgorgement, we are confronted immediately by the question: On what basis will profit be deemed to be too remote? The answer to that question—which in fact may be posed without reference to the concept 32 A Burrows, The Law of Restitution, 2nd edn (London, Butterworths, 2002) 500–501; Grantham and Rickett, above n 20, at 486–7. See also P Birks, An Introduction to the Law of Restitution (Oxford, Clarendon Press, 1985) 351–5; Mitchell, above n 14, at 336–9. 33 Burrows, ibid, at 500; Grantham and Rickett, above n 20, at 486. 34 Grantham and Rickett, above n 20, at 487. 35 I say ‘in part’ because, as I point out in the text, the function of a rule of remoteness may be performed by strategies other than the making of allowances, such as the fixing of a date on the calendar after which profit need not be disgorged.

350

Matthew Harding

of remoteness—is what promises to supply a justification of allowances. If the language of remoteness must be deployed, it ought not to divert attention from the fact that what is required is concentration on what Charles Mitchell has called the ‘substantive normative reasons’ for applying a rule of remoteness in disgorgement cases.36 And it will not suffice, in identifying those reasons, simply to point to vague concepts. For example, to say that profit will be deemed to be too remote from the defendant’s breach to the extent that disgorgement would cause ‘hardship’37 to the defendant is not to supply a justification of allowances; it is a way of stating the conclusion that an allowance is justified. And to assert that ‘the liability of the fiduciary should not be transformed into a vehicle for the unjust enrichment of the plaintiff ’,38 and that to that extent allowances are justified, is also to state a conclusion and not an argument.

IV. Allowances and Desert In summary, then, the aim of the account of profits is deterrence. In light of that aim, the making of allowances demands justification, because to the extent that an allowance is made in a disgorgement case, the goal of deterrence is compromised. Any attempt to justify allowances by tying them to an inquiry into causation is likely to fail, both for being inconsistent with the practices of courts when making allowances and for being at odds with the correct understanding of what amounts to a breach of fiduciary obligation in a disgorgement case. A more promising line of argument maintains that allowances perform the function of a rule of remoteness in disgorgement cases. However, to assert that allowances perform this function is only to pose the question of justification in new language: the basis on which profit may properly be identified as too remote to require disgorgement must still be worked out. Appeals to vague concepts like hardship will not do. Some more clearly articulated basis for the making of allowances is needed. Whatever that basis is, it must be sensitive to one fact about allowances. A defendant has no right to an allowance in a disgorgement case. Whether or not a defendant is entitled to her allowance as of right is a question that has troubled courts and commentators alike.39 However, the correct view is that, in the words of Tipping J of the High Court of New Zealand (as he then was), the making of an

36

Mitchell, above n 14, at 336. Grantham and Rickett, above n 20, at 487. 38 Warman International Ltd v Dwyer (1995) 182 CLR 544 (High Court of Australia) 561. Of course, it is clear that in this passage the High Court of Australia used the phrase ‘unjust enrichment’ in a broad sense and not to convey the specific meaning that the phrase has come to have in private law. 39 P Finn, Fiduciary Obligations (Sydney, Law Book Company, 1977) 121; Guinness v Saunders [1990] 2 AC 663 (HL) 700 (Lord Goff); Chirnside v Fay [2006] NZSC 68 [36]–[37], [46] (Elias CJ). 37

Justifying Fiduciary Allowances

351

allowance is ‘a matter of discretion not of right’.40 Indeed, the making of an allowance must be a matter of discretion. It is a long-standing principle of fiduciary law that a fiduciary is not permitted any remuneration qua fiduciary, except where this has been authorised by the principal.41 To make an allowance to a defendant for expenditures of time and skilled labour in generating gross profit that has been caused by a breach of fiduciary obligation is to permit remuneration to the defendant qua fiduciary when working out the net profit to be disgorged. This is precisely what the rule against unauthorised remuneration prohibits. So to say that a defendant has a right to an allowance in a disgorgement case is to say that the defendant has a right to have something that, at the same time, she is under an obligation not to have. This is morally—indeed logically—impossible. By contrast, to say that a defendant is under an obligation not to have unauthorised remuneration but that the court may, in its discretion, make an allowance in her favour does not entail any such impossibility. Rather, it entails an acceptance that the court may exercise its discretion to override the rule against unauthorised remuneration, and then invites a consideration of the reasons for and against overriding that rule.42 Accepting that a court may exercise discretion to override a rule in this way should present no difficulties except to those who would not tolerate any role for discretion in the award of equitable remedies. But the debate about discretionary remedialism is for another day.43 While a justification of allowances must be sensitive to the fact that defendants have no right to them, it is an error to conclude from the fact that defendants never have rights to allowances that allowances ought never to be made, just as it is an error to conclude from the fact that making allowances compromises the goal of deterrence that allowances ought never to be made. Earlier, I pointed out that allowances may be justified despite the fact that they always conflict with the goal of deterrence and that their justification may point to a goal, other than deterrence, that it is appropriate for the law to pursue in disgorgement cases. The pursuit of that goal need not entail recognising that defendants have rights to allowances. It simply may entail recognising that there are reasons to make allowances in favour of certain defendants in disgorgement cases, reasons that may be understood in light of the goal in question. In what follows, I argue that the goal of satisfying desert yields reasons for making allowances in some disgorgement cases. To that extent, desert supplies at least a prima facie justification of allowances.

40 Estate Realties Ltd v Wignall [1992] 2 NZLR 615 (New Zealand High Court) 627, 631. See also Harris v Digital Pulse Pty Ltd (2003) 56 NSWLR 298 (New South Wales CA) [311]–[336] (Heydon JA); Virgo, ‘Restitutionary Remedies for Wrongs’, above n 16, at 309–10. 41 This is best viewed as an outworking of the ‘no profit’ rule, but is sometimes spoken of as a separate requirement. 42 Such an exercise of discretion ought to be responsive to these and all other relevant reasons, which means that the reasons in question limit the choices that properly may be made by the court. See generally A Robertson, ‘Constraints on Policy-Based Reasoning in Private Law’ ch 11 of this book. 43 See generally P Birks, ‘Rights, Wrongs and Remedies’ (2000) 20 Oxford Journal of Legal Studies 1; S Evans, ‘Defending Discretionary Remedialism’ (2001) 23 Sydney Law Review 463.

352

Matthew Harding

A. Desert ‘To say that a person deserves something is to say that there is a certain sort of propriety in his having it’.44 Joel Feinberg’s statement captures the essence of desert as a moral concept, an essence that may be found in the relation of ‘propriety’ that Professor Feinberg points to. The essence of desert being propriety, desert may be distinguished as a moral concept from entitlement. When one is entitled to something, it may be said that one has a right to have that thing. However, when one deserves something, although one’s desert is a reason to have that thing it does not form the basis of a right.45 Because the essence of desert is propriety and not right, desert is a promising candidate for the justification of allowances in disgorgement cases. As I pointed out earlier, any such justification must be sensitive to the fact that defendants have no right to allowances, and desert is sensitive—indeed it depends for its conceptual identity—on just that fact. However, to say that desert entails a relation of propriety is only to begin an inquiry into its nature. This inquiry leads to a series of questions. On what basis does a person deserve something? What does a person deserve? And why does desert matter? For two reasons, I address each of these questions before turning to how desert may figure in the justification of allowances in disgorgement cases: first, so that potential general objections to desert-based arguments may be headed off; and secondly, so that a clearer picture emerges of precisely what I think desert does—and does not—justify. First, then, on what basis does a person deserve something? On this question, the philosophers who have written about desert are all agreed: the basis of desert is always some fact about a person’s qualities or actions, what she has been or what she has done.46 It is ‘in virtue of ’ such qualities or actions of a person that her desert may be apprehended.47 Thus are we able to make sense of statements like ‘the best chess player deserves to win the prize’ or ‘that kind of bravery deserves a medal’, as opposed to a statement like ‘every third candidate deserves to receive an honours grade’. At this point, I do not wish to explore the basis of desert any further—I shall turn shortly to some specific facts that may form the basis of desert in disgorgement cases—except to make two observations. First, the basis of desert is always backward-looking.48 And secondly, the basis of desert is always personal. Thus, for example, it makes no sense to say that Clare deserves to be punished

44 J Feinberg, ‘Justice and Personal Desert’ in his Doing and Deserving: Essays in the Theory of Responsibility (Princeton NJ, Princeton University Press, 1970) 56. 45 ibid, at 86; SR Munzer, A Theory of Property (Cambridge, Cambridge University Press, 1990) 260. 46 Feinberg, above n 44, at 58; J Kleinig, ‘The Concept of Desert’ (1971) 8 American Philosophical Quarterly 71, 73; D Miller, Social Justice (Oxford, Clarendon Press, 1976) 85; L Becker, Property Rights: Philosophic Foundations (London, Routledge, 1977) 50; Munzer, ibid, at 257; JW Harris, Property and Justice (Oxford, Oxford University Press, 1996) 205. 47 Feinberg, above n 44, at 58. 48 A point emphasised by Kleinig, above n 46, at 73.

Justifying Fiduciary Allowances

353

because she might commit a crime in the future; and it is equally bizarre to say that Luke deserves a reward because Rachel rescued a child drowning in the sea. These aspects of the basis of desert—that it is always backward-looking, and that it is always personal—are significant when thinking about the justification of allowances in disgorgement cases. They show that if desert does justify allowances, it supplies a justification very different in type from that which justifies disgorgement remedies like the account of profits, because far from being backwardlooking and personal, the goal of deterrence is forward-looking and impersonal. Because they supply very different types of justification, one would expect not only to find the goals of deterrence and desert in conflict in disgorgement cases; one would expect also to find that the conflict cannot be resolved by understanding more clearly the proper relationship between the two goals. Rather, one would expect to find that, when understood in their proper relationship, the two goals are incommensurable. The second question that I posed earlier is: What does a person deserve in light of what she has been or what she has done? At a fundamental level, Professor Feinberg points out that a person deserves to have expressed to her, and possibly to others as well, a ‘responsive attitude’ that is appropriate given the basis of her desert.49 At this level, the form that the expression of the appropriate attitude takes is neither here nor there. However, as Jim Harris has argued, against the backdrop of human interactions and institutions, desert is ‘hostage to convention’.50 In other words, against that backdrop, what a person deserves is what is conventionally given to those whose qualities or actions are the same as those that form the basis of that person’s desert; only in this conventional form may the appropriate attitude be expressed satisfactorily to the deserving person and, possibly, to others as well. Thus, according to Professor Harris, [i]f workers exert themselves to such effect that their work deserves recognition, and if convention dictates that appropriate recognition in such a context consists in more pay, it is no answer to suggest that we should give them a clap instead. In such a context pay, not applause, is the just reward for meritorious labour.51

Similarly, if a crime deserves punishment, and punishment in our community always takes the form of imprisonment, then what criminal Nick deserves in our community is imprisonment and not, say, a fine. If desert justifies the making of allowances in disgorgement cases, the fact that what a person deserves is determined by convention is significant, because it means that convention ought to play a role in quantifying the allowances to be awarded. In Chirnside v Fay, one of the questions that the Supreme Court of New Zealand had to determine was whether to calculate an allowance, to be awarded to a defendant whose breach of fiduciary obligation brought to fruition

49 50 51

Feinberg, above n 44, at 82. Harris, above n 46, at 207. See also Feinberg, above n 44, at 82; Miller, above n 46, at 91. Harris, above n 46, at 208.

354

Matthew Harding

a property development project, by reference to an hourly rate of pay or by reference to the value of the development project.52 To choose the latter path would have been in keeping with the conventional way of remunerating property developers. However, the Court chose the former path.53 It is arguable that what the defendant deserved in Chirnside v Fay—assuming now that desert justified an allowance in that case even though, for reasons that I shall discuss shortly, it almost certainly did not—was an allowance calculated in accordance with the conventional way of remunerating property developers. If that argument may be made, there is a reason to think that the Court chose the wrong path when quantifying the allowance. Finally, it must be asked why desert matters. This question, which presents a general challenge to desert-based arguments, may be asked in another way: Why should the fact that a person has had a particular quality or has performed a particular action mean that there is propriety in that person having something? Why should such a fact have any particular moral significance? This general challenge may be met by asserting that desert is a necessary and basic moral concept, so that it makes no sense from a moral point of view to ask why it matters. It is necessary because, as Professor Becker argues, without it there could be no understanding of moral responsibility: once we deny the intelligibility of saying that a person may deserve reward or punishment for her actions, we abandon the ‘whole enterprise of making moral judgments’.54 It follows that only the moral sceptic can do without desert. Moreover, Professor Becker argues that desert is a basic concept in the sense that it is pointless to try to ‘go behind it’ in order to understand its moral force.55 In this regard, asking why desert matters is rather like asking why entitlement matters, or why need matters. For those who are prepared to engage in moral argument at all, the answer must be: It just does. Because desert is a necessary and basic moral concept, where it supports an individual’s claim to something, the claim demands to be taken seriously by any person—including a judge—charged with its determination. In particular, it will be no answer to a desert claim to argue that its basis is impermissible or irrelevant because of the existence of facts other than those which form the basis of the desert. It follows that, if desert justifies the making of an allowance in a disgorgement case, it will not do to refuse to make the allowance, or to limit it, on the basis that allowances must be limited or withheld in all such cases because, for example, the goal of deterrence demands this. Such an argument will not do because it sidelines altogether a necessary and basic moral concept in a case where it has at least a prima facie justificatory role. If desert justifies an allowance in a disgorgement case, any argument against that allowance will have to explain not only why

52 53 54 55

Chirnside v Fay [2006] NZSC 68 [117] (Tipping J). ibid, at [153] (Tipping J). Becker, above n 46, at 49–50. Becker, above n 46, at 45.

Justifying Fiduciary Allowances

355

the allowance must be limited or withheld, but also why the allowance must be limited or withheld despite the fact that the defendant deserves it. The three questions about desert that I just posed led me to three thoughts, with respect to disgorgement cases in which allowances are made, that require further consideration. First, the conflict in such cases between the goal of deterrence and the goal of satisfying desert appears to be a conflict between incommensurables. Secondly, the quantum of a deserved allowance ought to be dictated by convention. And thirdly, if there are grounds on which a deserved allowance may be limited or withheld, they will need to explain not only why an allowance must be limited or withheld in a given case, but also why the allowance must be limited or withheld despite the fact that the defendant deserves it. I take up these thoughts in Section V. However, before doing that, I need to explain how desert may, in light of specific facts in certain disgorgement cases, justify making an allowance in favour of the defendant. That is my aim in the next sub-section.

B. Desert as a Justification of Allowances The obvious starting point when thinking about why a defendant might deserve an allowance in a disgorgement case is to consider whether the defendant’s expenditure of time and skilled labour—what for convenience, I shall simply call her labour—might be the basis of her desert. One type of moral claim, famously discussed by John Locke, bases itself on the fact that the claimant has mixed her labour with some thing.56 This claim based on labour-mixing, which is typically a claim to property, suffers from well-known shortcomings; primary among these is that the claim depends on the idea of self-ownership, an idea that entails what Professor Harris calls a ‘spectacular non sequitur’.57 However, a claim based on labour-mixing is not the same as a desert claim based on labour. A desert claim based on labour does not depend on the idea of self-ownership, nor does it depend on some metaphysical process by which a thing is ‘mixed’ with the labour that has been expended on it. Instead, a desert claim based on labour asserts that, in certain circumstances, the expenditure of labour itself—whether it results in the physical transformation of a thing, the exploitation of an opportunity, or something else—may be the basis of desert of a benefit. The benefit may be property, but it may also be something else; unlike the labour-mixing argument, a desert claim based on labour is typically a claim to a benefit other than ownership rights over the product of the labour.58 56 See generally J Locke and P Laslett (ed), Two Treatises of Government (Cambridge, Cambridge University Press, 1988) II.V. 57 The non sequitur is in the proposition that, because I am not a slave, I own myself. Professor Harris points out that it is consistent with my not being a slave that no one owns me: Harris, above n 46, at 196–7. For further shortcomings, see the celebrated critique in R Nozick, Anarchy, State and Utopia (New York, Basic Books, 1974) 174–8, and J Waldron, The Right to Private Property (Oxford, Clarendon Press, 1988) ch 6. 58 Although a desert claim based on labour may, in certain circumstances, justify property: Becker, above n 46, at 50–53.

356

Matthew Harding

Labour may be the basis of desert of a benefit in at least two ways. First, as George Sher points out, in many cases where a person expends labour in the pursuit of a worthy goal, that person deserves to succeed.59 Desert of success in achieving goals that one has laboured to achieve is not relevant when considering the justification of allowances, and I shall consider it no further here. The second way in which labour may be the basis of desert of a benefit, which is relevant when considering the justification of allowances, is that, in certain circumstances, where a person expends labour, she deserves a tangible benefit—like remuneration or an honour of some description—the precise identity and quantum of which will vary depending on what convention dictates.60 For convenience, I shall refer to this desert as desert of a reward. The phrase ‘in certain circumstances’ is designed to draw attention to the fact that labour, taken on its own, is not a sufficient condition of desert of a reward. For example, no one would suggest that Ian deserves a trophy simply because he has laboured for years at his hobby of constructing miniature ships inside bottles. Labour is only the basis of desert of a reward when it is accompanied by some other fact or facts.61 One fact that, when it accompanies the fact of labour, may support a desert claim to a reward is that the labour in question has benefited people other than the labourer. Professor Becker argues that labour may be the basis of desert where it ‘adds value’ to the lives of others.62 Professor Harris states the same argument as follows: ‘If a person by his labour confers benefit on others, he deserves to be rewarded’.63 The fact that, by one’s labour, one has benefited other people may be relevant to a desert claim in a macro sense and also in a micro sense.64 In the macro sense, the expenditure of labour that has brought about benefit to the community or to an organisation deserves reward whose source is that community or organisation. In this way, desert of remuneration or honours for public servants or the employees of a company may be understood. In the micro sense, where one’s labour has benefited another person, one may make a desert claim to a reward that is directed against the other person. If James washes Cynthia’s car, his claim to be rewarded based on what he deserves is best directed at Cynthia. Turning then to disgorgement cases, if desert claims based on labour figure in the justification of the making of allowances, then allowances ought to be made in those cases where the defendant, in breach of fiduciary obligation, has benefited the claimant by her expenditure of time and skilled labour. Allowances ought to 59

G Sher, Desert (Princeton NJ, Princeton University Press, 1987) 61–2. For the role of convention in working out what a person deserves, see above nn 50–51 and accompanying text. 61 Becker, above n 46, at 52 and 55. 62 Becker, above n 46, at 50–51. Indeed, there is a hint of this line of argument in Locke, in the passage where he says, ‘[t]o which let me add, that he who appropriates land to himself by his labour, does not lessen but increase the common stock of mankind’: Locke, above n 56, at II.V 37. 63 Harris, above n 46, at 204. See also H Dagan, ‘Comment: Property and the Public Domain’ (2006) 18 Yale Journal of Law and the Humanities 84, 90–92. 64 Harris, above n 46, at 210–12, argues that desert claims based on labour work better in a micro sense. 60

Justifying Fiduciary Allowances

357

be made in such cases because the fact of labour and the fact of consequent benefit to the claimant combine to provide sufficient support for a desert claim, in a micro sense, directed at the claimant. And indeed, in the case law the fact that the defendant’s labour benefited the claimant appears to have been treated as a significant factor in decisions in which allowances have been made in the defendant’s favour.65 This suggests that, even if courts have not made allowances for reasons that refer explicitly to desert, they have been sensitive, when making allowances, to precisely those facts that may form the basis of desert claims.66 One excellent example is to be found in the judgment of Wilberforce J in Phipps v Boardman, the well-known case in which a fiduciary acquired a controlling interest in an ailing company and turned its fortunes around, all in breach of fiduciary obligation.67 In Phipps v Boardman, the claimant benefited substantially as a result of the defendant’s activities. Having ordered that an account of profits be taken and having stated that the defendant’s expenditures of money should be entered on the account as disbursements, Wilberforce J continued: It seems to me that this transaction … was one of a special character calling for the exercise of a particular kind of professional skill. If Boardman [the fiduciary] had not assumed the role of seeing it through, the beneficiaries would have had to employ (and would, had they been well advised, have employed) an expert to do it for them. If the trustees had come to the court asking for liberty to employ such a person, they would in all probability have been authorised to do so, and to remunerate the person in question. It seems to me that it would be inequitable now for the beneficiaries to step in and take the profit without paying for the skill and labour which has produced it.68 (emphasis added)

Consequently, an allowance on a ‘liberal scale’ was required.69 According to this passage, the defendant’s labour, plus the fact that that labour benefited the claimant, would have made it ‘inequitable’ for an allowance to be refused. That those facts supported a desert claim to a reward enables one to see why it would have been inequitable to refuse an allowance. A further example of a court referring to labour benefiting the claimant when justifying making an allowance is to be found in O’Sullivan v Management Agency and Music Ltd.70 There, an account of profits was ordered against fiduciaries 65 See, eg, Brown v Litton (1711) 1 P Wms 140; Brown v de Tastet (1821) Jac 284; Featherstonhaugh v Turner (1858) 25 Beav 382; Yates v Finn (1880) 13 Ch D 839 (Ch); Re Macadam [1946] 1 Ch 73 (Ch); Re Jarvis (decd) [1958] 1 WLR 815 (Ch); Phipps v Boardman [1964] 1 WLR 993 (Ch); O’Sullivan v Management Agency and Music Ltd [1985] 1 QB 428 (CA); Murad v Al-Saraj [2005] EWCA Civ 959. 66 Such facts also appear to have been treated as significant when courts have considered the question of what proportion of the value of a thing a defendant will be permitted to retain, where the defendant has increased the value of the thing by his labour: see Docker v Somes (1834) 2 My & K 655, 668 (Lord Brougham); Chirnside v Fay [2006] NZSC 68 [125]–[126] (Tipping J). 67 Phipps v Boardman [1964] 1 WLR 993 (Ch). 68 ibid, at 1018. This passage was referred to with approval by Denning MR and accepted by Pearson LJ in the Court of Appeal: Phipps v Boardman [1965] 1 Ch 992 (CA) 1020 and 1030 respectively; and it was referred to with approval by Lords Cohen and Hodson in the House of Lords: Boardman v Phipps [1967] 2 AC 46 (HL) 104 and 112 respectively. 69 Phipps v Boardman [1964] 1 WLR 993 (Ch) 1018. 70 O’Sullivan v Management Agency & Music Ltd [1985] 1 QB 428 (CA).

358

Matthew Harding

whose work as promoters and managers had played a significant role in the success of the claimant as a writer and performer of popular music.71 The members of the Court of Appeal were unanimous in the view that an allowance should be made in favour of the defendants. Dunn LJ referred to the fact that the claimant enjoyed no success prior to his involvement with the defendants and was equally unsuccessful after his relationship with the defendants ceased. However, during the years of his collaboration with the defendants, the claimant’s success was, according to Dunn LJ, ‘phenomenal’.72 His Lordship referred to these facts as ‘significant’.73 Fox LJ agreed, saying: I am not satisfied that it would be proper to exclude [the defendants] from all reward for their efforts. I find it impossible to believe that they did not make a significant contribution to Mr O’Sullivan’s success. It would be unjust to deny them a recompense for that.74

The substantial benefit that the defendants’ labour delivered to the claimant was clearly the principal reason for the making of an allowance in O’Sullivan v Management Agency and Music Ltd. Fox LJ stated that it would be ‘unjust’ to refuse to make an allowance in the circumstances. As with Wilberforce J’s description of refusing an allowance as ‘inequitable’, Fox LJ’s description of such a refusal as ‘unjust’ calls for further explanation. That explanation is provided once it is seen that, in light of the defendants’ labour and the benefit it brought to the claimant, the defendants had a desert claim to a reward in the form of the allowance that the Court of Appeal made in their favour.75 What about cases in which the defendant’s labour has not benefited the claimant? In some such cases, an allowance has been withheld from the defendant.76 However, in other cases, the court has been prepared to make an allowance in favour of the defendant notwithstanding that the defendant’s activities did not benefit the claimant in any way. For example, in Fraser Edmiston Pty Ltd v AGT

71 Strictly, the account of profits was unnecessary and contrary to principle, because the case was one in which rescission was ordered based on undue influence: see D O’Sullivan, S Elliott and R Zakrzewski, The Law of Rescission (Oxford, Oxford University Press, 2008) [2.21]–[2.23]. 72 O’Sullivan v Management Agency & Music Ltd [1985] 1 QB 428 (CA) 458 (Dunn LJ). 73 ibid. 74 ibid, at 468 (Fox LJ). See also Waller LJ at 473. 75 It might be thought that my analysis, focused as it is on the fact that the defendant benefited the claimant, would best be presented with reference to the concept of (counter-) restitution of unjust enrichment. (For an exploration of that concept in a different setting, see E Bant, ‘Restiutio in Integrum and the Change of Position Defence: Lessons from Rescission’ [2007] Restitution Law Review 13.) However, for several reasons, I do not think that this is the case. First, an unjust enrichment analysis has difficulty drawing what to my mind is the plausible distinction between the honest and the dishonest defendant (see Part V below). Secondly, an unjust enrichment analysis does not sit easily with the fact that an allowance is within the discretion of the court (see Part IV above). Finally, to my mind the idea of (counter-) restitution of unjust enrichment simply does not capture the essence of what a court seeks to achieve by awarding an allowance and is therefore intuitively unsatisfying when applied to disgorgement cases. 76 For example: A-G v Blake [2001] 1 AC 268 (HL) (where an account of profits was ordered in response to a breach of contract); perhaps also Guinness plc v Saunders [1990] 2 AC 663 (HL).

Justifying Fiduciary Allowances

359

(Qld) Pty Ltd, the defendants profited by taking for themselves the benefit of a lease that they ought to have shared with the claimant. The defendants were permitted an allowance, although not a generous one.77 In Chirnside v Fay, an allowance was made in favour of a defendant who excluded the claimant from the benefit of a lucrative property development that the two had initially embarked upon in joint venture.78 And in Warman International Ltd v Dwyer, an allowance was made in favour of a defendant whose activities not only failed to benefit the claimant, but also caused the claimant to suffer a substantial financial loss.79 How may allowances be justified in such cases? The short answer is that if allowances are justified in such cases, it is unlikely to be on the basis of desert. Where a defendant’s labour has not benefited the claimant in any way, there appears to be no basis in desert for making an allowance in favour of that defendant when taking the account of profits. The defendant may deserve a reward because her labour may have benefited people other than the claimant. However, that desert will usually be insufficient to justify making an allowance in her favour against the claimant; it will found either a desert claim in a macro sense, or desert claims in a micro sense directed against the people other than the claimant who have benefited. For example, in Warman International Ltd v Dwyer the defendant’s activities brought to fruition a successful business importing, assembling and selling gear-boxes for motor cars which no doubt benefited many people, including stakeholders and customers of the business.80 However, it is far from clear why the defendant deserved a reward as against the claimant, given that the claimant did not benefit from the defendant’s activities. In a case like Warman International Ltd v Dwyer, if an allowance is to be made it must be justified on a basis other than desert; one contender for this basis might be the aim of supplying entrepreneurs with sufficient incentives to develop business opportunities. But it is also possible that there is simply no justification for an allowance in a case where the defendant’s labour has not benefited the claimant. In some disgorgement cases, judges have identified the extent to which the defendant was exposed to risk in his profit-making activities as a fact that might be relevant to the making of an allowance.81 Exposure to risk may be relevant to the making of an allowance based on what the defendant deserves. However, exposure to risk is not relevant as the basis of the defendant’s desert: the basis of desert is always what someone has been or what he has done, and exposure to risk

77 Fraser Edmiston Pty Ltd v AGT (Qld) Pty Ltd [1988] 2 Qd R 1 (Supreme Court of Queensland) 12–13. 78 Chirnside v Fay [2006] NZSC 68 [153] (Tipping J). 79 Warman International Ltd v Dwyer (1995) 182 CLR 544 (High Court of Australia) 553, 568. 80 ibid, at 551–2. 81 Phipps v Boardman [1965] 1 Ch 992 (CA) 1022, 1031 (Pearson LJ), 1032 (Russell LJ); Estate Realties Ltd v Wignall [1992] 2 NZLR 615 (New Zealand High Court) 630; Warman International Ltd v Dwyer (1995) 182 CLR 544 (High Court of Australia) 561; Chirnside v Fay [2006] NZSC 68 [27] (Tipping J), [110] (Elias CJ) referring to the findings of the trial judge.

360

Matthew Harding

matches neither description.82 Rather, exposure to risk may be relevant in determining the quantum of the allowance that is to be made in favour of a deserving defendant.83 As I pointed out earlier, the quantum of a deserved allowance ought to be determined according to convention.84 If the convention is to remunerate risk-takers more highly than non-risk-takers—perhaps to encourage entrepreneurial activity—then a risk-taking defendant who deserves an allowance because his labour has benefited the claimant may deserve more than a non-risk-taking defendant would deserve. It is strongly arguable that such a convention does exist in communities whose economies are organised on a market basis. To the extent that it does, courts in those communities are appropriately sensitive to risk-taking when considering the quantum of allowances in disgorgement cases.

V. Limiting Allowances Although it is usually acknowledged that allowances are justified in at least some disgorgement cases, there appears to be disagreement about when and on what basis courts ought to limit or even withhold allowances. For instance, in Guinness plc v Saunders, members of the House of Lords expressed the view that allowances ought to be made only in exceptional circumstances.85 But Warman International Ltd v Dwyer and Chirnside v Fay suggest that some judges are prepared to make allowances in a wide range of cases.86 In this section of the essay, I consider the question of when and on what basis allowances may properly be limited or withheld in disgorgement cases: first, in cases where the defendant has been dishonest; and secondly, in cases where the defendant has not been dishonest.

A. Dishonesty A review of the case law reveals that courts have been uncertain about what effect the dishonesty of the defendant ought to have on the making of allowances in disgorgement cases. Some judges appear to have viewed dishonesty as a sufficient

82 A person may expose himself to risk, and such an action may form the basis of a desert claim. However, the fact that the action has brought about exposure to risk ought to be irrelevant to any desert claim based on the action. If that fact were relevant to desert claims, counter-intuitive results would follow. The person who staked everything on the roulette wheel would be more deserving of a reward than the careful investor. 83 It may also be relevant when considering whether a claimant is barred from pursuing a claim due to delay: Clegg v Edmondson (1857) 8 De GM & G 787, 814. 84 See above nn 50–51 and accompanying text. 85 Guinness plc v Saunders [1990] 2 AC 663 (HL) 693–4 (Lord Templeman), 700–701 (Lord Goff). See also Chirnside v Fay [2006] NZSC 68 [38] (Elias CJ). 86 Warman International Ltd v Dwyer (1995) 182 CLR 544 (High Court of Australia); Chirnside v Fay [2006] NZSC 68 [137]–[139] (Tipping J). See also Estate Realties Ltd v Wignall [1992] 2 NZLR 615 (New Zealand High Court).

Justifying Fiduciary Allowances

361

reason for withholding an allowance altogether;87 others appear to have regarded dishonesty as no impediment to the making of an allowance.88 Still others have suggested that dishonesty may be relevant but have not explained precisely how.89 Moreover, in most cases where dishonesty is mentioned, it is mentioned in passing and there is no sustained analysis of its effect. In the one case and the few contributions to the literature in which the impact of dishonesty has been considered, the discussion has concentrated on whether withholding an allowance from a dishonest defendant has a punitive aim, and on whether such an aim is properly pursued in equity. On the one hand, Professor Edelman has asserted that the refusal to make an allowance in favour of a dishonest defendant is ‘in effect, a punitive measure’,90 and Mason P of the New South Wales Court of Appeal has recognised and endorsed a punitive aim in disgorgement cases where the defendant has been dishonest.91 On the other hand, Jessica Palmer has argued against withholding allowances in order to punish a defendant,92 and Heydon JA (as he then was) of the New South Wales Court of Appeal has put the view that the refusal to make an allowance to a dishonest defendant is not punitive in aim and that punishing the defendant ought never to be an aim of equity.93 In putting this view, Heydon JA drew on the well-known remarks of Sir WM James LJ in Vyse v Foster to the effect that the Court of Chancery was not a court of penal jurisdiction.94 Finally, Lee Aitken has addressed the question of dishonesty assuming that it falls to be considered with reference to the permissibility of punishment in private law, but has refrained from taking sides in the debate.95 Given that the making of an allowance is a matter within the discretion of the court, and that a defendant never has a right to an allowance, it is difficult to see how withholding an allowance from a dishonest defendant—indeed, from any defendant—could be punitive in aim. In what sense could a court aim to punish a person by not giving her that to which she otherwise has no right? As Heydon JA pointed out in Harris v Digital Pulse Pty Ltd, the proper view is that the court does not withhold an allowance in order to punish a dishonest defendant, but rather chooses to make an allowance in favour of a defendant whose honesty is a 87 Phipps v Boardman [1965] 1 Ch 992 (CA) 1020 (Lord Denning MR); United States Surgical Corporation v Hospital Products Ltd [1983] 2 NSWLR 157 (New South Wales CA) 243. 88 O’Sullivan v Management Agency and Music Ltd [1985] 1 QB 428 (CA) 468 (Fox LJ); Estate Realties Ltd v Wignall [1992] 2 NZLR 615 (New Zealand High Court) 631; Warman International Ltd v Dwyer (1995) 182 CLR 544 (High Court of Australia) 560, 568. 89 O’Sullivan v Management Agency and Music Ltd [1985] 1 QB 428 (CA) 456 (Dunn LJ). See also Finn, above n 39, at 242. 90 Edelman, Gain-Based Damages, above n 4, at 105 and 172 (the quotation is on 105). 91 Harris v Digital Pulse Pty Ltd (2003) 56 NSWLR 298 (New South Wales CA) [165]–[166] (Mason P). It is not clear whether Professor Edelman endorses such a punitive aim; his phrase ‘in effect, a punitive measure’ equivocates between aim and effect. 92 Palmer, above n 10, at 163–5. 93 Harris v Digital Pulse Pty Ltd (2003) 56 NSWLR 298 (New South Wales CA) [311]–[336] and passim (Heydon JA). 94 Vyse v Foster (1872) LR 8 Ch App 309 (Ch) 333. 95 Aitken, above n 10, at 64.

362

Matthew Harding

‘passport to an indulgence’.96 Whether the effect of withholding an allowance is punitive—whether a dishonest defendant who is refused an allowance thinks she has been punished—is neither here nor there. Indeed, as Heydon JA’s analysis in Harris v Digital Pulse Pty Ltd reveals, whether equity ought to be punishing defendants is utterly irrelevant to the question of when, if ever, a court should refuse to make an allowance in favour of a dishonest defendant. The real impact of dishonesty may be seen more clearly in light of a desertbased analysis. Put shortly, in a case where the defendant has been dishonest, the defendant does not deserve to have an allowance made in her favour. It follows that allowances ought never to be made in favour of dishonest defendants in disgorgement cases, not because dishonest defendants must be punished, but because the justification of allowances—desert—is absent where the defendant has been dishonest. To see how the argument works, consider a hypothetical example. Imagine that Peta benefits Adam by cooking and feeding him a nutritious dinner. Now imagine that Peta kidnapped Adam prior to the meal and that he has eaten his nutritious dinner under duress. The facts of Peta’s labour and Adam’s benefit, in the form of the nutrition his body has derived from the meal and the money he has saved by dining at Peta’s expense, are in this example insufficient as a basis for any desert claim that Peta might otherwise have had. For instance, it would be absurd to suggest that, given these facts, Peta deserves praise or thanks for providing the meal, let alone that Adam should be the one to praise or thank her. To the contrary, Peta ought never to have kidnapped and fed Adam in the first place. This example shows that where a person labours and consequently benefits another person, the facts of her labour and that other person’s benefit will not suffice as the basis of a desert claim where the expenditure of the labour is in violation of some moral duty.97 In a disgorgement case where the defendant has been dishonest, the defendant will typically have acted in ways that violate moral duties that she owes to the claimant. For example, the defendant may have lied to the claimant in order to conceal her activities, or perhaps the defendant has deliberately used the claimant in some other way as a means to the defendant’s own ends. Such actions, evidence of which will typically justify a finding of dishonesty, reveal violations of moral duty of a type famously discussed by Kant,98 although one need not be a Kantian to appreciate how moral duties of this type may exist.99 Having acted in violation of moral duty, the defendant, who might otherwise have deserved an allowance based on the fact that her labour benefited the claimant, deserves no reward, whether in the form of an allowance or in any

96

Harris v Digital Pulse Pty Ltd (2003) 56 NSWLR 298 (New South Wales CA) [335] (Heydon JA). Becker, above n 46 at 51–2; Sher, above n 59, at 67–8. 98 I Kant and M Gregor (trans & ed), Groundwork of the Metaphysics of Morals (Cambridge, Cambridge University Press, 1997) 38: ‘So act that you use humanity, whether in your own person or in the person of any other, always as an end, never merely as a means’. 99 B Williams, Truth and Truthfulness (Princeton NJ, Princeton University Press, 2002) 122. 97

Justifying Fiduciary Allowances

363

other form.100 Where the defendant has been dishonest, an allowance ought to be withheld altogether. But this ought to happen not because dishonesty calls for the limitation of a deserved allowance; it ought to happen because the dishonest defendant does not deserve an allowance in the first place.

B. Limiting Deserved Allowances That leaves cases in which an honest defendant has, by her labour, benefited the claimant. It is well established that an honest defendant who has benefited the claimant may still be liable to disgorge profit according to the ‘no profit’ rule.101 In such cases, the defendant’s desert claim to an allowance has a sufficient basis, and an allowance may not be limited or withheld for the reason that the defendant does not deserve it. However, there are two ways in which an allowance in such a case might possibly be limited or withheld justifiably. First, convention may dictate that the quantum of the allowance be limited; this might occur in a case where a defendant seeks an allowance calculated in an ‘all things considered’ way but the court quantifies a smaller allowance with reference to an hourly rate of pay or some other standard method of working out remuneration for labour of the type performed by the defendant.102 And secondly, an allowance may be limited or withheld, despite the fact that the defendant deserves it, to the extent that a goal of the law other than satisfying desert supplies overriding reasons for limiting or withholding the allowance in question. The obvious contender for such a goal is deterrence. However, as I pointed out earlier, where the goal of deterrence is in conflict with the goal of satisfying desert in a disgorgement case, this is most likely a conflict between incommensurables, because the goal of deterrence is forwardlooking and impersonal and the goal of satisfying desert is backward-looking and personal. Consequently, it is not clear when and why the goal of deterrence might override the goal of satisfying desert in a disgorgement case.103 One way forward might be to isolate the goal of deterrence and ask whether that goal is appropriately pursued in cases where an honest defendant, by her labour, has benefited the claimant. If the goal of deterrence were found to be an 100 If the dishonest defendant has expended labour in ways that do not benefit the claimant, then the defendant deserves no allowance not only because of her dishonesty but also because her labour has not benefited the claimant. 101 Regal (Hastings) Ltd v Gulliver [1967] 2 AC 134 (HL) 144–5 (Lord Russell of Killowen): ‘The rule of equity which insists on those, who by use of a fiduciary position, make a profit, being liable to account for that profit, in no way depends on fraud, or absence of bona fides; or upon such questions or considerations as whether the profit would or should otherwise have gone to the plaintiff, or whether the profiteer was under a duty to obtain the source of the profit for the plaintiff; or whether he took a risk or acted as he did for the benefit of the plaintiff, or whether the plaintiff has in fact been damaged or benefited by his action. The liability arises from the mere fact of a profit having, in the stated circumstances, been made’. 102 See above nn 22–26 and accompanying text. 103 It might be thought that one way of breaking this deadlock is to take the position that deterrence ought to be pursued by withholding from defendants that which they deserve. No doubt deterrence could be achieved thus, but the very permissibility of taking that position is what is in question.

364

Matthew Harding

inappropriate goal for the law to pursue in such cases, then to that extent the goal of satisfying desert would cease to be in conflict with the goal of deterrence and there would be no deterrence-related impediment to pursuing the goal of satisfying desert. However, if the goal of deterrence were found to be an inappropriate goal for the law to pursue in certain disgorgement cases, then more would be at stake than the justification of allowances—there would be no justification in those cases for making defendants liable to disgorge their profit in the first place. There would be no need even to consider making allowances in favour of such defendants. There are signs that the view that an honest defendant ought not to be required to disgorge profit to a claimant who has benefited by the defendant’s activities might be gaining traction in some quarters.104 If that view were to prevail in the law, the honest defendant whose activities have benefited the claimant would not need to resort to a desert claim to an allowance in a disgorgement case; she would have a right to all of her profit because, according to fiduciary principles, that profit would have been lawfully made.

104 J Langbein, ‘Questioning the Trust Law Duty of Loyalty: Best Interest or Sole Interest?’ (2005) 114 Yale Law Journal 929; Murad v Al-Saraj [2005] EWCA Civ 959 [82] (Arden LJ), [122]–[123] (Jonathan Parker LJ), [158] (Clarke LJ). See also the precursors to this view: G Jones, ‘Unjust Enrichment and the Fiduciary’s Duty of Loyalty’ (1968) 84 Law Quarterly Review 472, 497 and the enigmatic remarks of Deane J in Chan v Zacharia (1984) 154 CLR 178 (High Court of Australia) 204–5.

15 Gains-Based Remedies and the Place of Deterrence in the Law of Fiduciary Obligations ANTHONY DUGGAN*

I. Introduction Loss-based remedies compensate the plaintiff for harm caused by the defendant’s wrongdoing, and the most common example is damages. Gains-based remedies deprive the defendant of profits earned by wrongdoing. The most common examples are the constructive trust, which is a proprietary remedy, and the account of profits, which is a money claim. In Pettkus v Becker,1 the Supreme Court of Canada held that a constructive trust may be awarded to prevent unjust enrichment, subject to proof of correspondence between the defendant’s gain and the plaintiff ’s loss, and in LAC Minerals Ltd v International Corona Resources Ltd,2 the court apparently assumed that in the wake of Pettkus v Becker, the prevention of unjust enrichment was now the only basis for constructive trust relief. A note on terminology may be helpful before proceeding further. In other jurisdictions, restitution scholarship recognises two categories of unjust enrichment: unjust enrichment by subtraction and unjust enrichment by wrongs. Unjust enrichment by subtraction occurs where the defendant’s gain equals the plaintiff ’s loss. Unjust enrichment by wrongs occurs where the defendant profits from wrongful conduct, such as breach of fiduciary obligation, whether or not the plaintiff suffers a corresponding loss.3 The prevention of unjust enrichment by subtraction is commonly understood to be grounded in considerations of corrective justice, whereas the prevention of unjust enrichment by wrongs reflects * I developed some of the ideas in this paper over a period of years teaching a course on Commercial Equity with Michael Bryan in the LLM program at the University of Melbourne. I am in Professor Bryan’s debt for the many stimulating discussions we have had, inside and outside the classroom, about the issues this paper canvasses. My thanks also to Stephen Waddams, Michael Trebilcock, Ed Iacobucci, Craig Rotherham and James Edelman for their helpful comments. All errors are mine. 1 [1980] 2 SCR 834. 2 [1989] 2 SCR 574. 3 See, eg, P Birks, An Introduction to the Law Of Restitution (Clarendon Press, Oxford, 1989) 6.

366

Anthony Duggan

deterrence considerations.4 In Pettkus v Becker and later cases, the court speaks of the constructive trust to prevent unjust enrichment without distinguishing between unjust enrichment by subtraction and unjust enrichment by wrongs. However, given the court’s insistence in Pettkus v Becker on a correspondence between the defendant’s gain and the plaintiff ’s loss, unjust enrichment by subtraction is clearly what it had in mind.5 To restate, Pettkus v Becker suggests that constructive trust relief is limited to cases of unjust enrichment by subtraction. On the other hand, in Soulos v Korkontzilas6 the court held that a constructive trust may be imposed either for restitutionary reasons (to prevent unjust enrichment by subtraction) or for deterrence reasons (to prevent unjust enrichment by wrongs) and where the purpose is deterrence, it makes no difference whether the plaintiff has suffered any loss. 3469420 Canada Inc v Strother7 is a parallel decision holding that an account of profits may be awarded either for restitutionary reasons or to deter wrongful conduct and that, where the purpose is deterrence, the plaintiff ’s loss does not matter. In summary, Soulos and Strother seem to support the use of gains-based remedies for deterrence purposes and without reference to considerations of corrective justice. However, the majority judgment in Soulos is not unequivocal, while both Soulos and Strother were subject to strong dissents. The upshot is that the question remains at least to some extent unsettled. Part II, below discusses the constructive trust cases—Pettkus v Becker, LAC Minerals and Soulos—with particular reference to their ambivalence regarding the deterrence objective. Part III analyses Strother with emphasis on observations in the minority judgment about the nature and function of the account of profits remedy. Part IV identifies some implications the cases may have for the law of restitution at large. Part V explores the underlying policy considerations, arguing that the unsettled state of the case law is at least in part attributable to an unresolved tension between deterrence and corrective justice objectives compounded, in the case of the constructive trust, by bankruptcy concerns.

II. The Constructive Trust A. Pettkus v Becker Pettkus v Becker8 is the source in Canada of the constructive trust to prevent unjust enrichment. The case involved a dispute over the division of property between a 4 Though some argue that unjust enrichment by wrongs can also be accounted for in corrective justice terms: see below n 101. 5 See further, Part II A, below. 6 [1997] 2 SCR 217. 7 (2007) 281 DLR (4th) 640 (Supreme Court of Canada). 8 [1980] 2 SCR 834.

Gains-Based Remedies

367

man and woman who had lived together in a common law relationship for nearly 20 years. The property was in the man’s name, but the woman had made regular and substantial household contributions over the years. Prior to Pettkus v Becker, the courts had relied on resulting trusts principles to resolve cases of this kind.9 However, the resulting trusts approach depended on proof of a common intention that a partner’s contributions entitled her to a pro tanto share of the disputed property and, in the absence of an express statement of intention, the courts had no choice but to fall back on implication or, perhaps more accurately, imputation: their task was to ‘glean “phantom intent” from the conduct of the parties’.10 In Pettkus v Becker, Dickson J, writing for the majority, excoriated the artificiality of the resulting trusts approach and rejected it in favour of an unjust enrichment analysis. Dickson J went on to articulate three key requirements for a finding of unjust enrichment. He said there must be: (1) an enrichment accruing to the defendant; (2) a corresponding deprivation of the plaintiff; and (3) the absence of a juristic reason for the enrichment.11 In the present case, the defendant was enriched by the plaintiff ’s contributions to their joint household over the years, the plaintiff was correspondingly deprived by the defendant’s refusal to share the property and there was an absence of juristic reason for the enrichment because at the time of making her contributions the plaintiff reasonably expected a share of the disputed assets and the defendant should have realised she had this expectation. To reverse the unjust enrichment, the court awarded the plaintiff a constructive trust for a share of the disputed property proportional to her contributions.12 Pettkus v Becker was a family property case, but the constructive trust to prevent unjust enrichment clearly has a wide range of other applications as well, including commercial disputes. In Peter v Beblow, Cory J suggested that the courts may need to limit the use of the remedy in commercial cases,13 but McLachlin J in the same case doubted ‘the wisdom of dividing unjust enrichment cases into two categories— commercial and family—for the purpose of determining whether a constructive trust lies’.14 No doubt as a consequence of this statement, claims for constructive trust relief have become quite common in commercial litigation.15 9

Murdoch v Murdoch [1975] 1 SCR 423; Rathwell v Rathwell [1978] 2 SCR 436. Pettkus v Becker [1980] 2 SCR 834 at 843 (Dickson J). 11 Ibid, at 848. 12 Query whether the inquiry into the parties’ reasonable expectations is any less elusive than the ‘judicial quest for [a] fugitive common intention’ which the court rejected in Pettkus v Becker, ibid, at 842. Pettkus v Becker has a tragic postscript. The defendant played hard ball and the plaintiff was unable to collect from him. She ended up committing suicide. In a suicide note, she protested against the legal system which gave her a remedy but failed to help her enforce it: AH Oosterhof, Oosterhof on Trusts: Text, Commentary and Materials (6th edn, Toronto, Thomson Carswell, 2004) 613. 13 [1993] 1 SCR 980 at 1022. 14 Ibid, at 996. 15 Though with mixed success. Two relatively recent examples are Baltman v Melnitzer (1996) 43 CBR (3d) 33 (Ontario Court of Justice) and Ellingsen (Trustee of ) v Hallmark Ford Sales Ltd (2000) 190 DLR (4th) 47 (British Columba CA). For a discussion of earlier cases, see DM Paciocco, ‘The Remedial Constructive Trust: A Principled Basis for Priorities over Creditors’ (1989) 68 Canadian Bar Review 315. 10

368

Anthony Duggan

Pettkus v Becker decided that the prevention of unjust enrichment by subtraction is a justification for constructive trust relief. A predictable question that arose in its wake was whether the prevention of unjust enrichment by subtraction was now the only basis for the remedy. If this was the case, then how was Pettkus v Becker to be reconciled with the long line of authority, dating back to Keech v Sandford,16 in support of constructive trust orders to deter breach of fiduciary obligations?17 There are at least two possible approaches the court might have taken in response to this difficulty. The first would be to reconceptualise the older case law in subtractive unjust enrichment terms, while the second would be to concede that the prevention of unjust enrichment by subtraction is not the constructive trust’s only function and that the prevention of wrongful conduct is an alternative and equally legitimate objective. As it happens, the court has done both. It took the first approach in LAC Minerals Ltd v International Corona Resources Ltd18 and it took the second approach in Soulos v Korkontzilas.19

B. LAC Minerals In LAC Minerals, Corona was a junior mining company. It discovered that a property belonging to Mrs Williams sat on top of a valuable gold deposit. Corona did not have the resources to mine the deposit by itself and so it had discussions with LAC, a senior mining company, about the possibility of a joint venture or the like. In the course of negotiations, Corona disclosed to LAC the information about Mrs Williams’ property. LAC went behind Corona’s back and purchased the property for itself and it spent a large amount of money on the property in the early stages of development. Corona sued LAC for breach of confidence and breach of fiduciary duty. By way of relief it asked for compensation or alternatively a constructive trust over the property. The case had a mixed history on its way up through the courts, but in the Supreme Court it was held unanimously that LAC was liable for breach for confidence, by a majority that it was also in breach of fiduciary obligation and by a differently constituted majority that Corona was entitled to constructive trust relief. La Forest J, who delivered the leading judgment on the remedy question, proceeded on the assumption that, following Pettkus v Becker, there must be a finding of subtractive unjust enrichment before the court can award a constructive trust. The problem in the present case was that, while the Williams property was clearly LAC’s gain, there seemed to be no corresponding deprivation suffered by Corona because the property was not Corona’s to begin with. La Forest J got round the problem by holding, in effect, that what LAC deprived Corona of was not the 16

(1726) Sel Cas Ch 61, 25 ER 223 (Ch). See also Regal (Hastings) Ltd v Gulliver [1967] 2 AC 134 (HL); Boardman v Phipps [1967] 2 AC 46 (HL); Canadian Aero Service Ltd v O’Malley [1974] SCR 592. 18 [1989] 2 SCR 574. 19 [1997] 2 SCR 217. 17

Gains-Based Remedies

369

property itself, but the opportunity to acquire the property and the resulting profits. However, this response is itself problematic because without LAC’s interception, Corona probably would not have acquired the Williams’ property by itself. It would have needed help from LAC to purchase and develop the property. Therefore, the most likely outcome is that the parties would have purchased the property jointly and shared the development costs. It is true that Corona had to pay LAC for the cost of the improvements as a condition of the remedy, but Corona ended up getting all the surplus.20 In other words, the constructive trust gave Corona more than it had lost and, while a supra-compensatory remedy might be justifiable on deterrence grounds, it is inconsistent with the corrective justice agenda mapped out in Pettkus v Becker. In this connection, it is worth noting La Forest J’s remarks about the importance of preserving ‘the institution of good faith bargaining or relationships of trust and confidence’ and the need for a remedy which ‘acts as a deterrent to the breach of duty and strengthens the social fabric those duties are imposed to protect’.21 These statements signal a switch of tack from corrective justice to deterrence considerations. If the objective really is deterrence, the function of the remedy is to strip away the defendant’s gain and so it should be irrelevant whether there has been a corresponding deprivation to the plaintiff or not. The plaintiff ’s deprivation only matters if corrective justice is the objective. On the other hand if Pettkus v Becker requires proof of subtractive unjust enrichment, as La Forest J seemed to think, the case is hard to square with the deterrence agenda he was clearly flirting with. La Forest J’s reasoning on the unjust enrichment point calls to mind Roy Goode’s rationalisation of the Keech v Sandford line of cases in subtractive unjust enrichment terms.22 These cases all involve a fiduciary’s misuse of position to make a profit for himself. As in LAC, it is not immediately obvious where the unjust enrichment lies because the profits were not the plaintiff ’s to begin with. However, according to Goode, the problem is resolved by shifting the focus from the actual profits to the opportunity of acquiring them. Goode used the expression ‘deemed agency gains’ to describe the defendant’s profits: deemed agency gains are ‘gains which derive not from appropriation of property previously held by P but from activity undertaken by D for his own benefit which he was under an equitable duty, if he undertook it all, to pursue for P, so that D in effect acted as P’s constructive agent and the resulting gains will be treated in equity as if they 20 G Hammond, ‘Equity and Abortive Commercial Transactions’ (1990) 106 LQR 207 at 210–12. In addition to the point made in the text, La Forest J’s judgment is also open to the charge that his analysis begs the question. To fit the case into the subtractive unjust enrichment mould, La Forest J is forced to say, in effect, that the defendant’s gains belonged in equity to the plaintiff, even though the plaintiff may not have earned those gains in the absence of the defendant’s wrongdoing. However, if the defendant’s gains belong in equity to the plaintiff, the reason must be because the defendant holds them on trust for the plaintiff and the existence of the trust is the point in issue. 21 [1989] 2 SCR 574 at 673. 22 R Goode, ‘Property and Unjust Enrichment’ in A Burrows (ed) Essays on the Law of Restitution (Oxford, Clarendon Press, 1991), ch 9.

370

Anthony Duggan

had in effect been procured for P’.23 To avoid over-compensating the plaintiff, Goode argued that the constructive trust in deemed agency cases should always be conditional on the plaintiff reimbursing the defendant for her time, money and other inputs into the profit-making venture.24 Goode’s primary concern was with the bankruptcy implications of the constructive trust. If the plaintiff suffers no deprivation, a constructive trust over the defendant’s gains would remove property from the defendant’s bankruptcy estate, thus benefiting the plaintiff at the expense of the unsecured creditors. On the other hand, if the constructive trust is limited to cases of subtractive unjust enrichment, the remedy leaves the defendant’s estate no worse off than it would have been but for the restitutioninducing event.25

C. Soulos v Korkontzilas In Soulos v Korkontzilas, the plaintiff employed the defendant real estate agent to purchase a property on his behalf. The defendant purchased the property for himself instead and concealed his action by telling the plaintiff that the vendor had changed his mind about selling the property. The plaintiff sued the defendant for breach of fiduciary duty, claiming a constructive trust over the property. Specifically, the plaintiff asked for the property to be transferred to him for the price the defendant had paid, subject to certain adjustments. The market price of the property had fallen since the date of the defendant’s purchase, but both parties claimed to want the property because it held special significance for them as members of the Greek community. The defendant argued that there must be subtractive unjust enrichment before the court can award a constructive trust and, in the present case, there was no gain to the defendant because the market value of the property had fallen. Likewise, the plaintiff was not deprived. On the contrary, the defendant’s actions saved him from a bad bargain. In the Supreme Court McLachlin J, writing for the majority, rejected the defendant’s argument on two grounds. First, she said, the constructive trust remedy is not limited to the reversal of unjust enrichment. It is also available to deter breach of fiduciary obligation and other wrongful conduct. In any event, the plaintiff had established subtractive unjust enrichment, given his desire to own the property for non-monetary reasons.26 The conditions for constructive trust relief in wrongful conduct cases are that: (1) the defendant was under an equitable obligation; (2) the assets in the defendant’s hands resulted from deemed or actual 23

Ibid, at 219. Ibid, at 226. For a different view, see M Harding, ‘Justifying Fiduciary Allowances’ ch 14. 25 Goode, above n 22, at 226. Goode’s analysis is subject to the same concern as La Forest J’s judgment in LAC Minerals, ie, it is essentially question-begging: see above n 20. 26 [1997] 2 SCR 217 at 242–3. Sopinka and Iacobucci JJ dissented, holding that: (1) the constructive trust remedy depends on proof of unjust enrichment; (2) the trial judge had rejected the plaintiff ’s contention that the property held special value for him and there was no basis for disturbing this finding; and (3) there was no unjust enrichment because the plaintiff had suffered no loss. 24

Gains-Based Remedies

371

agency gains of the defendant; (3) the plaintiff has a legitimate reason for seeking a proprietary remedy; and (4) there are no other factors that would make it unjust to award a constructive trust. McLachlin J cited Goode’s paper, discussed in Part (b), above, as the source of these four conditions.27 It will be apparent from the above account of Goode’s analysis that the second condition is inappropriate if the purpose of the remedy is, as this part of the judgment presupposes, to deter wrongful conduct. The only reason for requiring proof of deemed or actual agency gains is to rationalise the remedy in unjust enrichment terms. However, the need for this exercise disappears once deterrence is recognised as an alternative basis for the remedy. Needless to say, McLachlin J’s deemed agency gains limitation has led to quite a bit of confusion in later, lower court decisions.28 Moreover, it could be read as suggesting ambivalence about the legitimacy of the constructive trust’s deterrence function and this impression is reinforced by her more recent statements, discussed in Part III, below. The second ground of the majority’s judgment is open to criticism because it is imprecise as to the nature of the enrichment. The plaintiff ’s deprivation was the non-monetary value he placed on the property. But what was the defendant’s enrichment? McLachlin J did not address this question, but presumably it was the non-monetary value the defendant placed on the property. This can’t be the end of the story, though, because unjust enrichment by subtraction depends on a correspondence between the defendant’s enrichment and the plaintiff ’s deprivation and there appears to have been no basis for assuming that the parties valued the property equivalently. Goode’s main concern was with the bankruptcy implications of the constructive trust and the court’s decision may, perhaps, be justifiable from this perspective. If the defendant had become bankrupt, the constructive trust would probably not have prejudiced the unsecured creditors. This is because they, unlike the defendant himself, presumably had no special attachment to the property and so removal of the property from the estate would not diminish the pool of assets available for distribution. On the contrary, given the plaintiff ’s undertaking to pay an amount equivalent to the price the defendant paid for the property, the property’s removal may actually have increased the asset pool by ridding the estate of a bad bargain.29 On the other hand, there is no indication in the judgment that the majority was thinking along these lines. The deemed agency gains requirement in Soulos suggests some ambivalence about the policy justification for constructive trust relief in wrongful conduct

27

Ibid, at 241. A Duggan, ‘Constructive Trusts and the Deemed Agency Limitation’ (2004) 83 Canadian Bar Review 151. 29 See C Rotherham, Proprietary Remedies in Context: A Study in the Judicial Redistribution of Property Rights (Oxford, Hart Publishing, 2002) 184–5, making a similar point about Keech v Sandford. 28

372

Anthony Duggan

cases and this creates uncertainty at the doctrinal level.30 For example, given that the deemed agency gains requirement reflects Goode’s conception of the constructive trust, it means that in a deemed agency gains case, that is to say, a case in the Keech v Sandford mould, the court must give the defendant an allowance.31 By contrast, the conventional wisdom is that the court has a discretion and it may deny the allowance if the defendant acted dishonestly.32 A discretionary allowance makes sense from a deterrence perspective, but it cannot be rationalised in unjust enrichment terms. The reverse is true of a non-discretionary allowance. It is also unclear, in the wake of Soulos, what should happen in a wrongful conduct case where there are no actual or deemed agency gains.33 Bribe-taking is the paradigm case. In Lister v Stubbs,34 the plaintiff employed the defendant as a purchasing officer. The defendant took a bribe from one of the plaintiff ’s suppliers and invested the bribe money. The plaintiff claimed a constructive trust over the defendant’s investments. The court denied the constructive trust and limited the plaintiff to an account of profits for the amount of the bribe money itself. According to Goode, the refusal of a constructive trust was ‘clearly correct’, because there was no subtractive unjust enrichment.35 More precisely, there were no deemed agency gains: the bribe was not a form of benefit which it was D’s duty to obtain for P; it resulted from conduct in which D ought not to have engaged at all. The distinction is between a benefit which P would have obtained if D had fulfilled his duty and a benefit which would never have been conferred at all if D had observed that duty.36

However, in the more recent case of A-G for Hong Kong v Reid,37 the Privy Council declined to follow Lister v Stubbs, asserting that ‘a fiduciary must not be allowed to benefit from his own breach of duty’.38 Reid’s case clearly supports the deterrence theory of constructive trust liability and by implication it rejects Goode’s subtractive unjust enrichment theory. Where does the law in Canada stand, post-Soulos? On the one hand, the Soulos majority’s apparent endorsement of the deterrence theory implies support for Reid’s case. On the other hand, the deemed agency gains restriction implies support for Goode’s subtractive unjust enrichment theory and Lister v Stubbs.

30 See L Smith, ‘Constructive Trusts—Unjust Enrichment—Breach of Fiduciary Obligation— Soulos v. Korkontzilas’ (1997) 76 Canadian Bar Review 539. 31 See text accompanying n 24, above. 32 Boardman v Phipps [1967] 2 AC 46 (HL). See also Harris v Digital Pulse Pty Ltd (2003) 56 NSWLR 298 (New South Wales CA) at 371–84 and cases cited there. 33 See Smith, above n 30. 34 (1890) 45 Ch D 1 (CA). 35 Goode, above n 22, at 231. 36 Ibid. Cf Rotherham, above n 29, at 184. 37 [1994] 1 AC 324 (PC). 38 Ibid, at 336.

Gains-Based Remedies

373

III. The Account of Profits A. Introduction 3469420 Canada Inc v Strother39 was a case involving a lawyer’s fiduciary obligations to his client. The client had engaged Strother to provide advice about film production tax shelter investments. There was a change in the tax laws aimed at defeating this kind of tax shelter and Strother advised the client there was no way round the new rules. Some time later, Strother discovered a loop-hole but, instead of telling the client, he exploited it himself. The client sued Strother and his firm alleging breach of fiduciary duties and claiming an account of profits. The trial judge dismissed the claim,40 but this decision was reversed by the British Columbia Court of Appeal.41 A further appeal to the Supreme Court resulted in a 5–4 split in the plaintiff ’s favour. The main point of disagreement was the scope of the defendant’s retainer; the majority held that the retainer obliged the defendant to keep the plaintiff informed of new developments, while the minority held the opposite.42 However, there was also disagreement over the remedy. The trial judge found that the plaintiff had suffered no loss because, by the time of Strother’s alleged wrongful conduct, it had already wound down its film production tax shelter business and was not in a position to revive it. Consequently, it could not have taken advantage of the new developments even if Strother had disclosed the information to it. The orthodox teaching is that an account of profits may be awarded for reasons of corrective justice (to prevent unjust enrichment by subtraction) or deterrence reasons (to prevent wrongful conduct) and where deterrence is the objective, the plaintiff ’s loss is neither here nor there.43 The majority judgment in Strother reflects this orthodoxy: [w]here, as here, disgorgement is imposed to serve a prophylactic purpose, the relevant causation is the breach of a fiduciary duty and the defendant’s gain (not the plaintiff ’s loss). Denying Strother profit generated by the financial interest that constituted his conflict teaches faithless fiduciaries that conflicts of interest do not pay. The prophylactic purpose thereby advances the policy of equity, even at the expense of a windfall to the wronged beneficiary.44

On the other hand, the minority, led by McLachlin CJC, went out of its way to question the legitimacy of using the account of profits remedy for solely deterrence

39

(2007) 281 DLR (4th) 640 (Supreme Court of Canada). 3469420 Canada Inc v Strother (2002) 26 BLR (3d) 235 (British Columbia Supreme Court). 3469420 Canada Inc v Strother (2005) 38 BCLR (4th) (Supp) 159, reasons 8 BLR (4th) 1, additional reasons 256 DLR (4th) 319. 42 For a fuller account of this aspect of the case, see A Duggan, ‘Solicitors’ Conflicts of Interest and the Wider Fiduciary Question’ (2007) 45 Canadian Business Law Journal 16. 43 Chan v Zacharia (1984) 154 CLR 178 (High Court of Australia) at 198 (Deane J), quoted in Strother (2007) 281 DLR (4th) 640 at [75] (Binnie J). 44 Ibid. 40 41

374

Anthony Duggan

purposes. Without deciding the issue, McLachlin CJC made the following points: (1) it is open to question whether the account of profits remedy is available in a case where the breach of fiduciary obligations relates not to the management of the plaintiff’s assets, but to an opportunity that arises to the defendant in his fiduciary capacity; (2) the account of profits remedy should be sparingly applied because ‘the effect may be to give the plaintiff priority over other creditors’, should the defendant be insolvent;45 (3) it may be anomalous to award the plaintiff an account of profits for breach of fiduciary duty in cases where, as in Strother itself, ‘the fiduciary duty arises from the same facts as a concomitant contractual duty’;46 and (4) there is a ‘tension between the need to deter fiduciaries from abusing their trust on the one hand, and the goal of achieving a remedy that is fair to all those affected, on the other’.47 It is true that these remarks are obiter, but they are still noteworthy because they contrast so sharply with McLachlin J’s48 take on the remedies question in Soulos.

B. The Account of Profits Remedy for Diversion of Commercial Opportunities McLachlin CJC’s suggestion that an account of profits may not be the appropriate remedy in cases where the defendant diverts to himself a business opportunity encountered in his fiduciary capacity is contrary to the Keech v Sandford line of cases. She bases her suggestion on the following statement by David Hayton in an Australian book of equity essays: Where the trustee’s profit is not made out of the trust property but out of an opportunity that arises to him in his office as trustee, … many have questioned whether it is equitable or fair that the trust beneficiaries should have a proprietary interest in that profit (rather than a mere personal claim).49

However, the reference in this passage is to proprietary remedies, such as the constructive trust, whereas the account of profits is a personal remedy. In any event, Hayton goes on to say, in a passage McLachlin CJC does not refer to:50 There is no justification to distinguish between misuse of property and misuse of office (whether honestly or dishonestly), while distinguishing them may well be capricious. In the vernacular, ‘beneficiaries are entitled to expect not to be ripped off by their trustee, or if he tries to rip them off, then they are entitled to expect that his profits become part of the trust fund’. 45

Ibid, at [155]. Ibid, at [156]. 47 Ibid. She also suggested that ‘where extra deterrence is required, it [may] be better achieved by remedies such as exemplary damages, which, unlike account, can be tailored to the particular situation’: ibid. For discussion, see Duggan, above n 42, at 27–8 and 30–3. 48 As she then was. 49 D Hayton, ‘Unique Rules for the Unique Institution, The Trust’ in S Degeling and J Edelman (eds), Equity in Commercial Law (Sydney, Law Book Company, 2005) 284, quoted in Strother (2007) 281 DLR (4th) 640 at [155]. 50 Ibid, at 285 (emphasis added). 46

Gains-Based Remedies

375

In a work which McLachlin CJC does not cite, James Edelman demonstrates that gains-based money remedies fall into two categories which he calls ‘restitutionary damages’ and ‘disgorgement damages’.51 Restitutionary damages is a remedy ‘which operates to reverse wrongful transfers of value from a claimant to a defendant’.52 The clearest example is the right to restitution that arises in mistaken payment cases. ‘Disgorgement damages operate to strip a defendant of profit made by wrongful conduct’.53 Examples in the fiduciary context include Boardman v Phipps54 and Canadian Aero Service v O’Malley.55 The rationale is deterrence.56 The main difference between restitutionary damages and disgorgement damages is that, in the case of disgorgement damages: whether a transfer of value occurs or not is irrelevant. There might be no transfer of value. The value might be generated from a third party or through the defendant’s own skill and initiative. For example, consider the case where a defendant trustee wrongfully uses information obtained while acting in a fiduciary capacity and makes a personal profit. The amount of the profit might be primarily the result of the defendant’s skill and experience. The defendant’s actions might exploit an opportunity which the claimant could never have acquired or even be to the benefit of the claimant, enhancing the claimant’s wealth as well as the defendant’s. But when disgorgement damages are available this gain must be given up.57

This statement is at odds with McLachlin CJC’s suggestion in Strother that the account of profits remedy may be inappropriate ‘in a case where the breach did not arise from the management of property, where it did not cause the plaintiff any loss and where, viewing the same case through the lens of contract law, the plaintiff would have recovered nothing’.58 McLachlin CJC’s error was to elide the two categories of gains-based damages Edelman identifies. By contrast, the majority in Strother noted that the account of profits remedy ‘may be directed to either or both of two equitable purposes’, the one ‘prophylactic’ and the other ‘restitutionary’59 and these statements square with Edelman’s analysis.

C. The Insolvency Implications of the Account Remedy McLachlin CJC suggests that the account of profits remedy should be sparingly applied because ‘the effect may be to give the plaintiff priority over other creditors’, should the defendant be insolvent. Again, this statement confuses proprietary

51 J Edelman, Gains-Based Damages: Contract, Tort, Equity and Intellectual Property (Oxford, Hart Publishing, 2002). 52 Ibid, at 1. 53 Ibid. 54 [1967] 2 AC 46 (HL), cited in Edelman, above n 51, at 72. 55 (1973) 40 DLR (3d) 371 (Supreme Court of Canada) cited in Edelman, above n 51 at 72. 56 Edelman, above n 51, at 82. 57 Ibid, at 72. 58 Strother (2007) 281 DLR (4th) 640 at [154]. 59 Ibid, at [74]–[75].

376

Anthony Duggan

and personal remedies. While proprietary remedies such as the constructive trust may have the effect McLachlin CJC contends for, the account of profits does not because it is only a money claim and so it gives the plaintiff no stake in any of the defendant’s assets. If the defendant becomes bankrupt, the plaintiff ’s only recourse will be to lodge a proof of claim for the amount of the judgment and to share pro rata with the other unsecured creditors in any dividend distribution. The personal nature of the account of profits remedy lies at the heart of the debate over the relative merits of Lister v Stubbs60 and Attorney-General for Hong Kong v Reid.61 In Lister v Stubbs, the court held that the defendant was in breach of fiduciary duty for accepting bribes. However, it refused to order a constructive trust over investments the defendant had purchased with the bribe money, confining the plaintiff instead to an account of profits equivalent to the value of the bribes. The court relied on bankruptcy considerations in support of its conclusion and, by implication, concluded that the need to preserve the integrity of the bankruptcy distribution system was more important than the deterrence of defaulting fiduciaries. By contrast, in Reid’s case, where the facts were comparable, the Privy Council awarded the plaintiff a constructive trust over the second-generation profits, citing deterrence considerations in support of its conclusion and downplaying the bankruptcy implications. If the account of profits was a proprietary remedy, as McLachlin CJC seems to think, the court in Lister v. Stubbs would have achieved nothing on the bankruptcy front by denying constructive trust relief: either way, the plaintiff would have had priority over the defendant’s unsecured creditors.62 In Reid’s case, the defendant had invested the bribe money in real estate. The reason the plaintiff asked for constructive trust relief was because it wanted to 60

(1890) 45 Ch D 1 (CA). [1994] 1 AC 324 (PC). 62 McLachlin CJC seems to have got the idea that the account of profits is a proprietary remedy from Sarah Worthington’s book, Equity (Oxford, Oxford University Press, 2003). According to Worthington: 61

the fiduciary holds the profits of his breach of trust for his principal from the moment they are obtained. The principal then has preferred status on the fiduciary’s insolvency, and this is inevitably controversial. Nevertheless, the analysis supporting this proprietary status is both simple and historically compelling. Disgorgement requires the fiduciary to pay over the profits of his disloyalty to the principal. [A] proprietary reaction in these circumstances is consistent with Equity’s treatment of other Equitable obligations that demand payment over to the claimant. If the defendant has the asset in question in his hands, Equity always regards the obligation as specific, and as entitling the claimant to treat the asset as already her own in Equity; it never regards the obligation as merely a personal obligation to pay over the value in money’ (at 124). McLachlin CJC’s citation is to the first edition of Worthington’s book, which has no footnotes. A second edition was published in 2006 (Equity (2nd edn, Oxford, Oxford University Press, 2006)), and this does include footnotes so it is now possible to identify the sources on which Worthington relied for the statement quoted above. The authority she cites in support of the first sentence is Reid’s case. However, as indicated above, the remedy the plaintiff sought and obtained in that case was a constructive trust, not an account of profits. Moreover, Worthington’s claim that her analysis is ‘both simple and historically compelling’ is now qualified by a footnote stating that ‘other views have held sway’ and citing Lister v Stubbs (at 135, n 20, emphasis added). The new footnote says it all.

Gains-Based Remedies

377

lodge a caveat and it needed a proprietary basis for doing so. Suppose that instead the plaintiff had asked for an account of profits representing the value of the land. Would the court have had jurisdiction to grant the remedy? Lister v Stubbs suggests that the answer may be ‘no’: the plaintiff is entitled to an account of profits only for the amount of the bribe money itself and the second generation profits are too remote. In Reid’s case, Lord Templeman reached the same conclusion relying on some extra-judicial statements by Sir Peter Millett.63 However, it makes no sense to say, in the name of deterrence, that the plaintiff is entitled to a constructive trust over the second-generation profits but that he cannot claim the money equivalent by way of a personal remedy. The better view is that, at least in cases of deliberate wrongdoing, the plaintiff should have the choice of remedy, subject to the court’s discretion to disallow a proprietary claim if the circumstances demand.64 This conclusion is inconsistent with Worthington’s analysis which submerges the personal remedy in the proprietary claim.65 However, it is consistent with the Pettkus v Becker66 line of cases in Canada which stresses the remedial character of the constructive trust and the court’s discretion to award a money remedy instead. McLachlin CJC’s reservations in Strother about the appropriateness of gainsbased remedies for breach of fiduciary duty suggest that she may be having second thoughts about the correctness of her leading judgment in Soulos, though she did not refer to Soulos in her Strother judgment. Soulos, like Strother, involved a breach of fiduciary duty which caused the plaintiff no identifiable loss and, in both cases, the plaintiff sought a gains-based remedy (an account of profits in Strother and a constructive trust in Soulos). Assuming, for the sake of argument, that the account of profits is a proprietary remedy, then the concerns McLachlin CJC expressed in Strother about the bankruptcy implications of the account of profits remedy apply with equal force to the constructive trust. The correctness of the decision in Soulos depends on two separate questions: (1) whether gains-based remedies are appropriate at all if the plaintiff has suffered no loss; and (2) whether proprietary remedies are appropriate, having regard to the potential bankruptcy implications. The answer to the first question turns on whether deterrence (preventing wrongful conduct), as distinct from corrective justice (preventing unjust enrichment by subtraction), is a legitimate function of gains-based remedies and the answer to the second question depends on a trade-off between deterrence and bankruptcy considerations.67 The point for present purposes is that there are two separate questions in play and the problem with McLachlin CJC’s analysis in Strother is that she runs them together.

63

P Millett, ‘Bribes and Secret Commissions’ [1993] Restitution Law Review 7. See Edelman, above n 51, at 108–11 and sources there cited. 65 See above n 62. 66 [1980] 2 SCR 834. 67 As to which, see, eg, A Duggan, ‘Constructive Trusts from a Law and Economics Perspective’ (2005) 55 University of Toronto Law Journal 217 and Paciocco, above n 15. 64

378

Anthony Duggan

D. The Contract-Fiduciary Overlap In Strother, McLachlin CJC suggested that the account of profits remedy may be inappropriate in cases, like the present one, where the plaintiff has overlapping causes of action for breach of contract and breach of fiduciary obligation. Why, she asks, should the plaintiff be entitled to an account of profits if he sues in equity when, if he had sued for breach of contract he would have recovered nothing?68 Citing Canson Enterprises Ltd v Boughton & Co,69 she raises the question ‘whether the remedy of account, like equitable compensation, should be harmonized with common law remedies where the facts support concurrent equitable and legal claims’.70 This last statement assumes that damages is the only money remedy for breach of contract. On the contrary, in A-G v Blake,71 the House of Lords held that in appropriate circumstances the plaintiff in a contract action may claim an account of profits.72 Of course, it would be open to the Supreme Court to hold that Blake’s case should not be followed in Canada, but McLachlin CJC does not mention Blake and there is no hint in her judgment that disapproving the case is what she had in mind.73 In Harris v Digital Pulse Pty Ltd,74 the New South Wales Court of Appeal gave a similar reason for holding that punitive damages should not be available for breach of fiduciary obligation that was also a breach of contract. In Australia, punitive damages, or exemplary damages as they are called there, cannot be awarded for breach of contract and a majority of the court in Digital Pulse held that it would be anomalous if the plaintiff could succeed in obtaining punitive damages by suing for breach of fiduciary obligation rather than for breach of contract. This view is open to question. In Digital Pulse, the plaintiff ’s punitive damages claim depended on the fiduciary relationship, not the contract. Allowing punitive damages for a breach of fiduciary obligation which also happens to be a breach of contract is not the same as allowing punitive damages for breach of contract at large. Furthermore, allowing punitive damages for breach of fiduciary obligation except where there also happens to be a breach of contract is itself anomalous; the case for punitive damages rests on the need to deter fiduciary wrongdoing and that imperative does not disappear just because the parties are in a contractual relationship.75 The same goes for Strother. In cases like Strother, 68

Strother (2007) 281 DLR (4th) 640 at [153]. [1991] 3 SCR 534. 70 Strother (2007) 281 DLR (4th) 640 at [158]. 71 [2001] 1 AC 268 (HL). 72 For discussion, see M McInnes, ‘Account of Profits for Common Law Wrongs’ in Degeling and Edelman, above n 49, at 409. 73 Cf America Canada v Mutual Trust Co [2002] SCR 601 at [30–3], where the court appeared to acknowledge the availability of gains-based remedies for breach of contract; however, the discussion is brief and there is no mention of Blake’s case. 74 (2003) 56 NSWLR 298 (New South Wales CA). 75 See A Duggan, ‘Exemplary Damages in Equity: A Law and Economics Perspective’ (2006) 26 OJLS 303 at 324. 69

Gains-Based Remedies

379

the account of profits remedy serves the same deterrence function as punitive damages and the fact that the fiduciary relationship happens to be grounded in a contract between the parties should make no difference.

E. Deterrence and the Prevention of Unjust Enrichment by Subtraction In Strother, McLachlin CJC referred to a debate taking place ‘throughout the Commonwealth’ about the appropriate remedies for breach of fiduciary duties and that, ‘underlying this debate is the tension between the need to deter fiduciaries from abusing their trust on the one hand, and the goal of achieving a remedy that is fair to all those affected, on the other’.76 In other words, is deterrence a legitimate goal for gains-based remedies in the absence of subtractive unjust enrichment? This is the same as the question the court had earlier addressed in Soulos. There the court apparently concluded that the answer was ‘yes’, although the suggestion that the constructive trust depends on proof of actual or deemed agency gains cuts in the other direction. McLachlin CJC’s reservations in Strother about the deterrence function of the account of profits remedy contrast sharply with her appeal to deterrence considerations as a justification for the constructive trust in Soulos. On the other hand, they are perfectly reconcilable with her support for the deemed agency gains requirement. In summary, McLachlin CJC’s judgment in Soulos is ambiguous and the Strother minority judgment reinforces the ambiguity of her position. Bribe-taking involves fiduciary wrong-doing, but without actual or deemed agency gains. Soulos leaves the status of Reid’s case uncertain in Canada and Strother does not make matters any clearer. The Strother majority judgment has no direct bearing on the question. While the majority held that an account of profits may be justifiable on deterrence grounds, even without proof of subtractive unjust enrichment, the same is not necessarily true for the constructive trust having regard to the proprietary character of the remedy and the bankruptcy implications which flow from this.77 The minority judgment in Strother is inconsistent with Reid’s case. Furthermore, it could be read as suggesting, contrary to Lister v Stubbs, that the plaintiff should not even be entitled to a money remedy given the absence of a corresponding loss to the plaintiff. McLachlin CJC did not refer to the bribe cases in Strother and the likelihood is that the implications of her remarks did not occur to her. Nevertheless, when and if the Supreme Court is confronted with a bribe case, she will have to nail her colours to the mast. Support for Reid’s case will require reaffirmation of her commitment to the deterrence objective of gains-based remedies for breach of fiduciary duty. On the other hand,

76 77

Strother (2007) 281 DLR (4th) 640 at [156]. See text accompanying n 25.

380

Anthony Duggan

a conversion to the corrective justice school of thought would force her to the conclusion that Reid’s case was wrongly decided.

IV. The Perils of Top-Down Reasoning The importance of Pettkus v Becker extends beyond the constructive trust. Pettkus v Becker is one of three seminal cases in the Canadian law of restitution. The other two are Deglman v Guaranty Trust of Canada78 and Garland v Consumers’ Gas Co.79 In Deglman, the Supreme Court for the first time acknowledged unjust enrichment as a source of obligations in the common law. Pettkus v Becker’s contribution was to provide a framework for analysing unjust enrichment claims, while in Garland, the court confirmed that the Pettkus v Becker approach and, in particular, the inquiry into ‘juristic reason’, applies not just in constructive trust cases but to unjust enrichment claims at large.80 Pettkus v Becker has received a somewhat negative press in Australia, including one early assessment which unkindly dismissed it as ‘fog on stilts,’ in reference to the apparent indeterminacy of the unjust enrichment concept.81 In Roxborough v Rothmans of Pall Mall Australia Ltd,82 a case involving a claim for restitution of unlawfully imposed taxes, Gummow J, the co-author of the ‘fog on stilts’ quip and now on the High Court, launched a spirited attack on the restitution movement, saying that unjust enrichment was an example of ‘top-down reasoning’, ‘whereby a theory about an area of law is invented or adopted and then applied to existing decisions to make them conform to the theory and to dictate the outcome in new cases’.83 Top-down reasoning puts the theory first so that the theory drives the cases. This is antithetical to the common law tradition where the influence is the other way round.84 Furthermore, he said, if unjust enrichment is transformed into a definitive legal principle, substance and dynamism may be restricted by dogma. In turn, the dogma will tend to generate new fictions in order to retain support for its thesis. It may also distort well settled principles in other fields, including those respecting equitable doctrines and remedies, so that they answer the newly mandated order of things. Then various theories will compete, each to deny the others. There is support in Australasian legal scholarship for

78

[1954] SCR 725. [2004] 1 SCR 629. 80 L Smith, ‘Demystifing Juristic Reasons’ (2007) 45 Canadian Business Law Journal 281 at 281–2. 81 JD Heydon, WMC Gummow and RP Austin, Cases and Materials on Equity and Trusts (4th edn, Sydney, Butterworths, 1993) at [3111]. 82 (2001) 208 CLR 516 (High Court of Australia). 83 Ibid, at [73], citing RA Posner, ‘Legal Reasoning From the Top Down and From the Bottom Up: The Question of Unremunerated Constitutional Rights’ (1992) 59 University of Chicago Law Review 433. 84 Ibid, at [72]. 79

Gains-Based Remedies

381

considerable skepticism respecting any all-embracing theory in this field, with the treatment of the disparate as no more than a species of the one newly discovered genus.85

In Farah Constructions Pty Ltd v Say-Dee Pty Ltd,86 a case concerning recipient liability for breach of fiduciary duty, the High Court in a unanimous judgment quoted this statement with approval.87 It also said that ‘[u]njust enrichment is not a “definitive legal principle according to its own terms” ’,88 ‘the restitution basis is unhistorical’89 and it ‘reflects a mentality in which considerations of ideal taxonomy prevail over a pragmatic approach to legal development’.90 The unsettled state of the Canadian case law discussed in Part II, above underscores the force of these points. Pre-Pettkus v Becker, gains-based remedies were the province of equity, not the law of restitution. It was well-established in equity that both the account of profits and the constructive trust had not only a restitutionary function, but a deterrence—or prophylactic—function as well. Pettkus v Becker obscured this point by articulating a new basis for constructive trust relief which took no account of the remedy’s deterrence function. The court has subsequently struggled to address the problem in two different ways, first by attempting to reconceptualise the older cases in unjust enrichment terms and secondly, by conceding that deterrence is a separate and legitimate basis for the remedy. These responses are inconsistent with one another and the tension between them remains unresolved. The problem could have been avoided had the court in Pettkus v Becker resisted the temptation to invent a new theory of constructive trust relief. The account of profits remedy has fared somewhat better. It would have been tempting to conclude, post-Pettkus v Becker, that, by analogy with the constructive trust, an account of profits depends on proof of unjust enrichment by subtraction. Fortunately, the majority in Strother resisted the temptation and clung to orthodoxy. On the other hand, the minority took the bait and, by doing so, demonstrated the power of dogma to distort settled principles ‘so that they answer the newly mandated order of things’.91 Of course, this is a minority position and it might be no more than a curiosity, were it not for the fact that the Chief Justice wrote the judgment and it was subscribed to by three other members of the court. For these reasons it would be premature to celebrate orthodoxy’s triumph over the new order. Incidentally, it is probably fair to say that the Roxborough case and Farah Constructions have stopped the emerging Australian restitution movement in its

85

Ibid, at [74]. [2007] HCA 22. 87 Ibid, at [151]. 88 Ibid, quoting David Securities Pty Ltd v Commonwealth Bank of Australia (1992) 175 CLR 353 (High Court of Australia) at 378–9 (Mason CJ, Deane, Toohey, Gaudron and McHugh JJ). 89 Farah Constructions [2007] HCA 22 at [154]. 90 Ibid. For similar views about legal taxonomy, see SM Waddams. Dimensions of Private Law— Categories and Concepts in Anglo-American Legal Reasoning (Cambridge, Cambridge University Press, 2003). 91 See text accompanying n 85 above. 86

382

Anthony Duggan

tracks, at least temporarily. The status of the law of restitution is also dubious in the United States but, interestingly, for quite different reasons. According to Chaim Saiman, the restitution movement in the United States has fallen victim to legal realism because realism shifts the focus from doctrinal exposition to extradoctrinal justifications for legal outcomes.92 From a realist perspective, a case like Pettkus v Becker, for example, is unexciting because the shift from common intention to unjust enrichment analysis makes absolutely no difference to the sub-text of the decision and the court’s real motivations. In Saiman’s words, ‘[the realist’s] skepticism is a subset of a more general suspicion of the explanatory power of inductive and deductive modes of legal reasoning and argument’.93 By contrast, the Australian High Court’s position is firmly rooted in doctrinalism. From the realist’s perspective, the restitution movement is simply irrelevant. However, from the High Court’s doctrinal perspective it is positively harmful because of its potentially distorting effect on settled doctrines and also because it is antithetical to the established common law tradition of reasoning by induction. The Australian position is at least in part a function of history. There is a strong New South Wales contingent on the High Court at present, as there has often been in the past. New South Wales has the particular distinction of being one of the last common law jurisdictions to enact Judicature Act reforms and it maintained separate common law and equity bars until the 1970s. The consequence was to perpetuate a keen sense of difference between common law and equity, grounded in their separate origins, and a commitment to maintaining the separation in the interests of doctrinal purity. To suggest a merger of common law and equitable doctrines, for example by allowing a common law defence to an equitable cause of action, is to commit the ‘fusion fallacy’—a crime hardly less reprehensible than killing or stealing.94 For unreconstructed traditionalists, the restitution movement is the fusion fallacy writ large. Adoption of unjust enrichment as the unifying theme cuts across established legal boundaries and it is inconsistent with maintaining the separation between common law and equity that the old-style equity lawyer is committed to. In other words, the two agendas are diametrically opposed. With these considerations in mind, it becomes easier to understand the High Court’s uncharacteristically heated reaction to the emerging restitution movement in Australia. Saiman characterises law and economics as the right wing branch of legal realism in the United States.95 This may be an over-simplification but, in any event, the opposition to the restitution movement in Australia is a manifestation, not of right-wing attitudes as such, but of conservatism in its literal sense.

92 C Saiman, ‘Restitution in America: Why the US Refuses to Join the Global Restitution Party’ (2008) 28 OJLS 99, pt 1. 93 Ibid, at 107. 94 RP Meagher, JD Heydon, and MJ Leeming, Meagher, Gummow and Lehane’s Equity: Doctrines and Remedies (4th edn, Sydney, Butterworths, 2002) ch 2. 95 See above n 92, pt V.

Gains-Based Remedies

383

V. The Competing Policy Concerns For the realist, of course, policy matters more than orthodoxy. In Canada, as matters presently stand post-Strother, an account of profits can be awarded for deterrence reasons alone and without proof of a corresponding loss to the plaintiff. Is this outcome justifiable in policy terms? Gains-based remedies in tort are potentially inefficient because they give the defendant an incentive to spend more on precautions to avoid liability than the plaintiff ’s expected loss.96 Likewise, gains-based remedies in contract are generally speaking uneconomical because they discourage efficient breach.97 On the other hand, different considerations apply in the fiduciary context. Deterrence of fiduciary misconduct is important to preserve the integrity of fiduciary relationships and a loss-based remedy will be insufficient deterrence if the defendant’s gain exceeds the plaintiff ’s loss. To see why, consider the following example: D obtains for himself a contract with T which he was under a fiduciary obligation to obtain for P. P values the contract at $4 and D values it at $7. Assuming a probability of 1 that P will discover D’s wrongdoing and successfully sue for recovery, D’s expected gain is $7. His expected loss is $4 if P’s remedy is a loss-based one, but it rises to $7 if P’s remedy is a gains-based one.98

Allowing for P’s successful suit, D will still be ahead if P is limited to a loss-based remedy, but not if P can sue for a gains-based remedy instead. Note that a gains-based remedy does no more than prevent unauthorised takings: it does not preclude the parties from trading entitlements and, in fact, it encourages bargaining. In the above example, a gains-based remedy makes it unprofitable for D to misappropriate the contract, but it would still be profitable for D to buy the contract from P at any price up to $7 while P, for her part, should be willing to sell for any price over $4. Fiduciary law permits transactions like this provided P freely consents and the terms are fair.99 In summary, ‘just as 96 AM Polinsky and S Shavell, ‘Punitive Damages: An Economic Analysis’ (1998) 111 Harvard Law Review 870 at 879–80. 97 Bank of America Canada v Mutual Trust Co [2002] SCR 601 at [31]. Cf FH Easterbrook and DR Fischel, ‘Contract and Fiduciary Duty’ (1993) 36 Journal of Legal Studies 425 at 441–6. 98 See Duggan, above n 75, at 311. Fiduciary relationships typically involve high agency costs which make the defendant’s wrongdoing hard to detect. Fiduciary law addresses this problem in part by relaxing the burden of proof on the plaintiff. This increases the probability of detection and successful litigation and in most cases makes additional measures, such as punitive damages, unnecessary: RJ Cooter and BJ Freeman, ‘The Fiduciary Relationship: Its Economic Character and Legal Consequences’ (1991) 66 New York University Law Review 1045. 99 There is a risk that a reviewing court may wrongly conclude that P did not properly consent or that the terms are unfair. However, this does not make the gains-based remedy inefficient. To see why, consider the following variation on the example in the text:

P values the contract at $4, whereas D values it at $7. D is prepared to pay P $4 if P lets D take the contract. D’s total costs of negotiating with P and T are $1. There is a ¼ ex ante probability that P will end up suing D and that the court will wrongly reject D’s justification of the deal. On these figures, D’s expected loss is 25 cents ($1 × ¼). His expected gain is $2.25 ($7-$4) × (1-1/4). Since D’s expected gain exceeds his expected loss, he is likely to go ahead with the transaction: see Duggan, above n 75, at 312.

384

Anthony Duggan

restitution plus an additional penalty induces the would-be thief to enter into market transactions instead, the profits remedy induces the parties to contract explicitly. It is a contract-inducing, not a contract-frustrating, approach’.100 Of course, this is a law and economics perspective and other theoretical perspectives may lead to different conclusions. For example, it is much harder to justify gains-based remedies from a corrective justice perspective in cases where there is no corresponding loss to the plaintiff.101 These different ways of thinking perhaps underlie the disagreement between the majority and the minority in Strother over the remedies question, though neither judgment is cast in explicitly theoretical terms. Soulos is ambiguous on the availability of constructive trust relief for deterrence reasons and without proof of unjust enrichment by subtraction. In this respect, the decision reflects the tension in the earlier case law, including Lister v Stubbs and Reid’s case. As it happens, the question involves a difficult trade-off between deterrence and bankruptcy considerations and the indecision in the cases reflects the difficulty of the choice. Bribe-taking is the paradigm case of fiduciary wrongdoing without subtractive unjust enrichment. In the bribe cases, the plaintiff ’s loss bears no necessary relation to the defendant’s gain and, as in Reid’s case itself, it may be unquantifiable.102 This means that a loss-based remedy is likely to result in systematic under-compensation. Given the risk of under-compensation, the plaintiff ’s likely ex ante preference will be for a remedy to deter the defendant’s wrongdoing so the compensation issue does not arise. For effective deterrence, the remedy must capture all the defendant’s gains. An account of profits limited to the amount of the bribe money, as in Lister v Stubbs, falls short in this regard. An account of profits that extends to the defendant’s second-generation profits would be an improvement, but even this might result in under-deterrence if the defendant has a special attachment to the disputed property or there is a chance that the property will increase in value after the judgment date.103 The only sure-fire way of extracting all the defendant’s gains is to impose a constructive trust on the bribe money and its traceable proceeds. Ex ante, the availability of the constructive trust remedy makes both the plaintiff and the defendant better off because, by making the defendant more trustworthy, it increases the value of his services to the plaintiff and presumably also the amount the plaintiff is willing to pay for them.

100

Easterbrook and Fischel, above n 7, at 444. But see, eg, EJ Weinrib, ‘Restitutionary Damages as Corrective Justice’ (2000) 1 Theoretical Inquiries in Law 1 (gains-based remedies in tort); L Smith, ‘The Motive, Not the Deed’ in Joshua Getzler (ed), Rationalizing Property, Equity and Trusts (London, LexisNexis, 2003) 53 (gains-based remedies for breach of fiduciary obligation). For an overview and critique of the arguments, see K Barnett, Corrective Justice and Gains-Based Remedies (paper presented at ‘The Goals of Private Law’ Obligations IV Conference, National University of Singapore, July 23–25, 2008). 102 [1994] 1 AC 324 (PC) at 331 (Lord Templeman). 103 See Rotherham, above n 29, at 178–82 where this point is addressed at greater length. 101

Gains-Based Remedies

385

It is sometimes argued that the cost of the constructive trust is the costs the unsecured creditors bear if the defendant becomes bankrupt.104 However, this is an ex post perspective. Ex ante, the creditors may be better off because, by making the defendant more trustworthy, the constructive trust improves his employment prospects and, presumably, his income-earning potential and the capacity to repay his debts.105 The real concern with the constructive trust remedy is arguably that it may increase the cost of bankruptcy proceedings by encouraging strategic claims which may be costly to process and administer and which may lead to the outlay of scarce estate resources on expensive litigation. These considerations all point towards a default rule, but the question is what the default rule should be.106 Lister v Stubbs supports a default rule denying the constructive trust remedy but leaving the parties free to opt in if they want to. Reid’s case supports a default rule allowing the remedy but leaving the parties free to opt out. In choosing between the two approaches, the aim should be to minimise the aggregate costs of transacting around the default rule. This in turn depends on the difficult question of whether parties at large are more likely than not to want the constructive trust remedy and, given the empirical nature of the inquiry, it is not surprising that the courts may differ. Contrast the corrective justice perspective on constructive trusts with the economic approach described above. From an economic perspective, since the primary goal of all remedies is deterrence rather than compensation or restitution, the presence or absence of a corresponding loss to the plaintiff makes no difference: either way, the question is whether the deterrence benefits of the remedy exceed the bankruptcy costs. On the other hand, from a corrective justice perspective, the presence or absence of a corresponding loss to the plaintiff makes all the difference. Emily Sherwin argues that a constructive trust may be imposed even if the defendant is bankrupt, but only for the purpose of restitution.107 In the bankruptcy context, the question is whether the defendant’s unsecured creditors, rather than the defendant herself, have been unjustly enriched at the plaintiff ’s expense. This question depends on whether the plaintiff was an involuntary creditor: in other words, did the plaintiff voluntarily accept the risk of the defendant’s bankruptcy? If not, then the plaintiff should have priority.108 104 See, eg, R Goode, ‘Proprietary Restitutionary Claims’ in W Cornish et al (eds), Restitution, Past, Present and Future: Essays in Honour of Gareth Jones (Oxford, Hart Publishing, 1998); EL Sherwin, ‘Constructive Trusts in Bankruptcy’ [1989] University of Illinois Law Review 297. 105 See Duggan, above n 67, at 230. 106 For a summary and critical analysis of the literature on default rules, see I Ayres, ‘Default Rules for Incomplete Contracts’ in P Newman (ed), New Palgrave Dictionary of Economics and the Law (London, MacMillan, 1998) vol 1 at 585. 107 EL Sherwin, above n 104. 108 There is a substantial element of indeterminacy here. For example, if the defendant tricks the plaintiff into a transfer of the disputed asset, it could be argued that the plaintiff is an involuntary creditor because the defendant’s duplicity subverted the plaintiff ’s choice. On the other hand, it could be argued that the risk of fraud is an incident of contracting and, while it justifies a remedy, it does not justify giving the plaintiff priority in the defendant’s bankruptcy. According to Sherwin, the question turns on a number of variables including the nature of the fraud and the plaintiff ’s commercial

386

Anthony Duggan

On the other hand, Sherwin goes on to say, a constructive trust should never be awarded against a bankrupt defendant if the purpose is deterrence. The reason is that if the defendant is bankrupt, the remedy impacts on the defendant’s creditors, rather than the defendant herself and this is pointless from a deterrence perspective.109 Sherwin accepts the case for constructive trust relief on deterrence grounds outside bankruptcy and her analysis implies that the award should be defeasible if the defendant becomes bankrupt subsequently. The difficulty is that changing creditors’ relative entitlements in bankruptcy encourages opportunism: a creditor may put the debtor into bankruptcy not for the purpose of maximising returns to the creditors at large, but simply as a mechanism for avoiding the constructive trust. Sherwin concedes this point, but she argues, contentiously, that corrective justice considerations are more important.110 Perhaps with these difficulties in mind, Goode maintains that the courts should never grant constructive trust relief without proof of unjust enrichment, even outside the defendant’s bankruptcy.111 The trouble is that this leaves the constructive trust’s deterrence potential untapped even in cases where, as events turn out, there are no bankruptcy implications.112 To summarise, law and economics scholars and corrective justice theorists are sharply divided on the proper function of remedies and the ambivalence in the case law at least partly tracks these different ways of thinking. Furthermore, even within each of the two schools of thought, there is room for disagreement over questions such as how much weight to give bankruptcy considerations in determining the appropriateness of constructive trust relief. Given these differences, the unsettled state of the case law is hardly surprising.

VI. Conclusion The Canadian cases discussed in this paper reveal a deep-seated ambivalence about the appropriateness of gains-based remedies for breach of fiduciary duty and other wrongs. This uncertainty contrasts sharply with the state of the law in sophistication and the ‘constructive trust claimant’s position as an involuntary creditor is a question of degree’: ibid, at 352. It may be difficult to predict in advance how a court might rule on such questions. Furthermore, some might say that preferred creditor status is a matter best left to the legislature because there may be other types of involuntary creditor who are more deserving than constructive trust claimants in the type of case before the court. 109

Ibid, at 337–9. Ibid, at 361–4. 111 Above n 22. 112 Of course, the criminal law is a fallback, but not all actionable wrongs are criminal offences and offences are not always prosecuted. The literature on the relative merits of public and private law enforcement is substantial. Contributors include Hobbes, Rousseau, Weber Becker and Stigler. For a useful survey, see K Roach and MJ Trebilcock, ‘Private Enforcement of Competition Laws’ (1996) 34 Osgoode Hall Law Journal 461. 110

Gains-Based Remedies

387

other jurisdictions and, for that matter, with the state of the law in Canada prePettkus v Becker. In part, the unsettled state of the case law in Canada is a function of the court’s particular method of reasoning in Pettkus v Becker, namely the use of induction to develop a theory of constructive trust relief which happened to fit the facts of the case itself but overlooked other categories of dispute. And in part, the unsettled state of the law is a function of competing policy objectives, namely deterrence and corrective justice, compounded in the case of the constructive trust by bankruptcy concerns. This policy tension is not confined to Canada, but is evident in other jurisdictions as well, and it is likely to remain unresolved until the courts squarely address it.113

113

See Rotherham, above n 29, at 196 expressing more or less the same point as follows:

The wisdom that the ‘progeny’ of equity ‘must be legitimate—by precedent out of principle’ leads to efforts to rationalize outcomes as the consequence of exercises in deduction from formal premises found in the existing law. The danger is that, in being hidden from view, the normative concerns that motivate the court are not fully examined.

16 The Normative Foundations of Restitution for Wrongs: Justifying Gain-based Relief for Nuisance CRAIG ROTHERHAM

I. Introduction The common law has long allowed certain claimants relief based on the benefit derived by the defendant as a result of a wrong.1 Such relief is not, however, universally available and there is little agreement as to the principles governing when claimants should be allowed recourse to this alternative to compensatory relief. Broadly speaking, there are two principal types of approach to the issue. On the one hand, a methodology that can be characterised as ‘formalistic’ or, perhaps less perjoratively, ‘conceptualistic’2 assumes that the norms regulating entitlement to gain-based relief are internal to the private law3 and proceeds by identifying analytical features of particular torts that indicate whether or not such relief should be available. On the other hand, approaches that can be characterised as ‘pragmatic’4 identify normative objectives that would justify granting gain-based relief and determine whether these would be served in a particular context without being constrained by notions of the internal logic of particular legal doctrines. This essay explores the general issue of the justification of gain-based relief through the lens of the tort of nuisance. An examination of the case law on the 1 For overviews of the different ways in which the common law has provided for this outcome, see Strand Electric & Engineering Co Ltd v Brisfords Entertainments Ltd [1952] 2 QB 246 (CA) (Denning LJ) and A-G v Blake [2001] 1 AC 268 (HL) (Lord Nicholls). This interpretation of the relief in question is not universally accepted. Some courts and commentators treat such relief as providing compensation for loss. For a critique of this view, see C Rotherham, ‘ “Wrotham Park Damages” and Accounts of Profits: Compensation or Restitution?’ [2008] Lloyd’s Maritime and Commercial Law Quarterly 25. 2 See, eg, C Saiman, ‘Restating Restitution: A Case of Contemporary Common Law Conceptualism’ (2007) 52 Villanova Law Review 487. 3 See eg EJ Weinrib, The Idea of Private Law (Cambridge MA, Harvard University Press, 1995). For his account of restitution for wrongs, see EJ Weinrib, ‘Restitutionary Damages as Corrective Justice’ (2000) Theoretical Inquiries in Law 1. 4 For an account of pragmatism in legal reasoning, see R Posner, The Problems of Jurisprudence (Cambridge MA, Harvard University Press, 1990).

390

Craig Rotherham

subject in England and the United States reveals that gain-based relief for nuisance has attracted judicial attention only occasionally and has yet to be satisfactorily resolved. This essay then considers two formalist approaches to the availability of gain-based relief generally, and evaluates their application to the tort of nuisance in particular. Perhaps the most categorical example of such an exercise in fleshing out the immanent doctrinal logic of particular torts was provided by Peter Birks’ distinction between ‘anti-harm’ wrongs and ‘anti-enrichment wrongs’.5 A second and more enduring example of this form of reasoning can be seen in analyses that link the right to gain-based relief to whether or not the right breached is proprietary in nature. On closer scrutiny, neither of these accounts proves satisfactory—gain-based relief cannot be justified by reference to conceptual distinctions alone. This essay seeks to move beyond formal explanations for the availability of gain-based relief and to provide an analysis that directly addresses the difficult but ultimately inescapable normative questions of individual moral desert and/or the collective good. Focusing on the objectives of punishment and deterrence that are generally regarded as central to the justification of gain-based relief, this essay asks to what extent these purposes would be furthered by awarding such relief in the context of nuisance. It becomes clear that such an approach to the matter is unlikely to result in the endorsement of the view that gain-based relief should be automatically available for any tort that enriches the defendant. A pragmatic approach is, instead, likely to draw more detailed distinctions based on factual specificities relating to the importance of different interests, the moral quality of the defendant’s conduct and the likelihood that gain-based relief would deter breaches of the sort in question. This is apparent in the context of nuisance. The uncertain process of balancing competing rights to land use means that gainbased relief would often be inappropriate. Such relief can, nonetheless, be readily justified for blatant transgressions. It is useful to examine the approach favoured in the proposed Restatement (Third) of Restitution and Unjust Enrichment, which is broadly in sympathy with such a nuanced approach to restitution for wrongs generally and for nuisance in particular.

II. Gain-Based Relief for Nuisance Even taking a wide view of its importance in this context, it is unlikely that gainbased relief has a particularly prominent role to play in the law of nuisance. Most nuisances will not enrich the tortfeasor. Nonetheless, there are several ways in which a claimant might argue that a defendant has benefited from causing a nuisance, and as many ways in which gain-based relief can be measured.

5

P Birks, An Introduction to the Law of Restitution, rev edn (Oxford, Clarendon Press, 1989) 328.

Justifying Gain-based Relief for Nuisance

391

First, a defendant may profit by skimping on measures needed to prevent or abate a nuisance.6 In these circumstances, the defendant’s enrichment might be quantified by calculating how much the defendant saved as a result of not adopting the necessary measures. Secondly, a defendant might be said to be enriched as a result of causing a nuisance by not paying the price that the claimant might have demanded for granting an easement allowing the defendant to interfere with the claimant’s enjoyment of his or her land. In these terms, the measure of a defendant’s enrichment would be linked to what would be a reasonable fee for licensing the defendant’s actions.7 Finally, a defendant might be said to be enriched as a result of deriving a profit from the activity that caused a nuisance. On this basis, relief might be measured in terms of some or all of the net profit in question. Such a measure is, however, liable to give rise to serious questions of causation and remoteness and difficult issues of apportionment.

III. Treatment of the Issue in Case Law A. English Case Law The issue has attracted attention in English law only relatively recently and the resulting authority is limited and controversial. Carr-Saunders v Dick McNeil Associates Ltd8 is often treated as authority for allowing gain-based relief for nuisance. However, the extent to which the decision supports this proposition is debatable. The defendant committed a nuisance by adding two storeys to its premises and obstructing natural light to which the claimant enjoyed a prescriptive right. Following Wrotham Park Estate v Parkside Homes,9 Millett J reasoned that general damages could be calculated by determining what would have amounted to a reasonable fee for the right to build in the manner in question. Millett J treated evidence of the claimant’s loss of amenity resulting from the interference as a factor to be taken into account in determining a hypothetical fee and remarked that, in calculating an appropriate figure, he would also have had regard to evidence concerning the defendant’s profit from the development had this been made available.10 The value of the decision as authority for providing gain-based relief for nuisance is undermined by Millett J’s suggestion that the award in question was one 6 Often the nuisance will arise because of the defendant’s negligence in taking inadequate precautions. Indeed, such fault is a condition for liability if the defendant’s use of the land is considered ‘natural’: see Cambridge Water v Eastern Counties plc [1994] 2 AC 264 (HL). In these circumstances, the issues arising as to whether gain-based relief should be available for nuisance are the largely the same as those which would have to be considered in relation to the provision of such relief for negligence more generally. 7 See, eg, Carr-Saunders v Dick McNeil Associates Ltd [1986] 1 WLR 922 (Ch). 8 Carr-Saunders [1986] 1 WLR 922 (Ch). 9 Wrotham Park Estate v Parkside Homes [1974] 1 WLR 798 (Ch). 10 Carr-Saunders [1986] 1 WLR 922 (Ch) 931.

392

Craig Rotherham

for loss. If this appears to be difficult to reconcile with his Lordship’s reliance on Wrotham Park, some light may be thrown on the matter by his subsequent analysis of that decision in Jaggard v Sawyer.11 There, Millett LJ remarked that it was ‘plain’ that, in Wrotham Park, ‘Brightman J’s approach was compensatory’ on the basis that ‘[h]e sought to measure the damages by reference to what the plaintiff had lost’ by fixing upon ‘the amount which he judged the plaintiff might have obtained as the price of giving its consent’.12 What is more, in applying Wrotham Park in Carr-Saunders, Millett J placed some significance on the fact that damages had been sought in lieu of an injunction.13 Subsequently, in Jaggard v Sawyer, Millett LJ explained Wrotham Park on this basis and defended the Court of Appeal’s refusal to grant gain-based relief for a breach of contractual covenant in Surrey County Council v Bredero Homes Ltd14 on the ground that the claimant in that case was not in a position to apply for an injunction.15 This suggests that gain-based relief might be available only where a claimant’s proprietary right has effectively been expropriated in that a monetary award is being granted in return for the defendant’s being permitted to continue the activity in question. The subsequent Court of Appeal decision in Stoke-on-Trent City Council v W & J Wass Ltd16 is, in contrast, sometimes regarded as authority for the proposition that gain-based relief is not available for nuisance.17 The reality is that the effect of the decision is largely explained by and limited to the particular facts of the case. The defendant ran a market in breach of the claimant council’s monopoly to operate such an enterprise in the area. The defendant’s activity amounted to a nuisance as a result of an irrebuttable evidential presumption that a market conducted within a particular radius of the claimant’s market caused the claimant loss.18 As a result, the claimant was entitled to an injunction. The defendant, however, initially failed to comply with enforcement notices and then, following a conviction, simply shifted its market to a nearby location that was still within the zone covered by the Council’s monopoly. As a result, the Council was anxious to pursue further legal measures against the defendant and brought a claim for monetary relief. In the High Court, in making an inquiry as to damages, Peter Gibson J found that, in fact, the defendant’s breach of the claimant’s monopoly caused no harm to the claimant’s market. Nonetheless, he held that the claimant was entitled to relief based on what would have been a reasonable fee to have paid 11

Jaggard v Sawyer [1995] 1 WLR 269 (CA). ibid, at 291. 13 Carr-Saunders [1986] 1 WLR 922 (Ch) 931. 14 Surrey County Council v Bredero Homes Ltd [1993] 1 WLR 1361 (CA). 15 Jaggard v Sawyer [1995] 1 WLR 269 (CA) 290. Similarly, in Wass v Stoke-on-Trent CC [1988] 1 WLR 1406 (CA) 1420, Nicholls LJ distinguished Wrotham Park on the basis that it involved an award of damages in lieu of an injunction. This issue was also raised in Forsyth-Grant v Allen [2008] EWCA Civ 505. 16 Stoke-on-Trent City Council v W & J Wass Ltd [1988] 1 WLR 1406 (CA). 17 See, eg, G Virgo, Principles of the Law of Restitution, 2nd edn (Oxford, Oxford University Press, 2006) 472; A Burrows, The Law of Restitution, 2nd edn (London, Butterworths, 2002) 475. 18 See Sevenoaks District Council v Pattullo & Vinson Ltd [1984] Ch 211 (CA). 12

Justifying Gain-based Relief for Nuisance

393

for the privilege to operate a market. On appeal, however, the Court of Appeal found for the defendant and limited the claimant to nominal damages. It is, however, easy to exaggerate the significance of the decision. Nourse LJ concluded that, with the exception of cases in which the award was made in lieu of an injunction, relief in the form of a hypothetical reasonable fee was available only where the claimant had suffered loss.19 The reality is that this interpretation would mean that gain-based relief would almost invariably be available in cases of nuisance. The facts of Wass were anomalous in that a nuisance was established in that case only as the result of an evidential presumption of loss. Generally, however, a nuisance cannot exist unless there is harm in the form of interference with claimants’ enjoyment of their land.20 The most recent decision in this context is interesting for its discussion of the possibility of a claim for consequential profits rather than a reasonable licence fee. In Forsyth-Grant v Allen,21 the nuisance in question was, once again, in the form of an interference with a right to light.22 In the course of developing two houses next to the claimant’s hotel, the defendant’s architect offered to compensate the claimant for any interference with the claimant’s right to light that would result from the development. The claimant, however, declined to reply to any communication on the subject. Evidence suggested that the diminution in the value of the claimant’s hotel as a result of the loss of light was £1,848 and that the defendant’s development was worth £6,767 more than it would have been had it been built respecting the claimant’s right to light. While, the judge at first instance was not inclined to grant gain-based relief, he concluded that an award of a reasonable fee based on a fair percentage of the defendant’s profit would have resulted in a figure that would have been less than the compensatory damages to which the claimant was entitled.23 The claimant appealed, arguing that she was entitled to the entire profit resulting from the breach. The Court of Appeal was not persuaded. Patten J, with whom Mummery LJ concurred, concluded that, at the very least, such relief was reserved for exceptional cases.24 Moreover, his Lordship suggested that, if necessary, he would have gone further and concluded that an account of profits is not available for nuisance.25 Delivering the other opinion, Toulson LJ also 19

Stoke-on-Trent City Council v W & J Wass Ltd [1988] 1 WLR 1406 (CA) at 1414. Nonetheless, in Forsyth-Grant v Allen [2008] EWCA Civ 505 [19], Patten J remarked that, ‘[i]n relation to claimants in nuisance, it is, I think, significant that the possibility of an award of damages on a restitutionary basis has been limited by … [Wass].’ 21 Forsyth-Grant v Allen [2008] EWCA Civ 505. 22 See also Tamares Ltd v Fairpoint Properties Ltd [2007] EWHC 212 for another recent example arising in this context. 23 One might doubt that this represents a sound conclusion. After all, it is difficult to imagine that anyone in the claimant’s position would have accepted a fee that was less than the actual loss that would result from the actions in question. Common sense suggests that a reasonable fee should be drawn from in somewhere in the range between the claimant’s loss and the defendant’s profit from the wrong. 24 Forsyth-Grant v Allen [2008] EWCA Civ 505 [26] (Patten J) (following the approach taken in A-G v Blake [2001] 1 AC 268 (HL) 284 (Lord Nicholls)). 25 ibid, at [32]–[33]. 20

394

Craig Rotherham

concluded that the claimant was not entitled to an account of profits. Nonetheless, his Lordship was less definitive on the question as to whether the remedy could ever be available in a claim for nuisance. Thus, Toulson LJ stated: I would reject the argument that this remedy is, or should be, generally available as a matter of course in cases of nuisance … I would [however] leave open the question whether … [such an] award may be available on the facts of a particular case of nuisance, given the developing state of the law, because it seems to me better than any such questions should be addressed in the light of particular features said to justify that relief in the particular case.26

There is something to be said for Toulson LJ’s cautious approach. For one thing, if the claimant in Forsyth-Grant v Allen had not acted so unreasonably, it is quite possible that the Court would have looked more sympathetically upon her request to strip the defendant of a profit that was directly attributable to a knowing breach of the claimant’s right to light. More generally, it may be that the quite different interests protected in the widely divergent circumstances in which a nuisance may occur make sweeping judicial generalisations on the appropriateness of gainbased relief unwise.

B. US Case Law The law in the United States on this question is not much clearer. Some courts have concluded that gain-based claims should not be available for nuisance. For example, in Marmo v Tyson,27 the Court of Appeals of the 8th Circuit dismissed an appeal of the District Court of Nebraska’s decision to refuse to allow the plaintiff to amend its pleadings to include an unjust enrichment claim based on the defendant’s saving from failing to install appropriate pollution control equipment. The Court of Appeals justified its conclusion that the claim was ‘futile’ on the basis that there was no compelling authority for a gain-based claim in nuisance in Nebraskan law and that there was no indication that the Nebraskan Supreme Court would entertain such a claim. The majority, however, offered neither any state appellate authority conclusively denying such a claim nor any reason of principle as to why such a claim should not be recognised. Such a narrow approach hardly appears compelling in light of the fact that, as the majority recognised, not only had courts in other States allowed gain-based claims for nuisance,28 but also at least one Nebraskan district court had recently refused a defendant’s motion to dismiss such a claim.29 As a result, Judge Arnold’s dissenting view that the appeal

26

ibid at [47]. Marmo v Tyson 457 F 3d 748 (2006). 28 The court noted the following decisions in Branch v Mobil Oil, 778 F Supp 35 (US District Court, Oklahoma, Western District, 1991) 36; NC Corff Partnership Ltd v OXY USA Inc, 929 P 2d 288 (Oklahoma CA, 1996) 295; Evans v City of Johnstown, 96 Misc 2d 755, 410 NYS 2d 199 (Supreme Court of New York, 1978). Both are discussed below. 29 Schwan v CNH America LLC, 2006 WL 1215395 (US District Court, Nebraska). 27

Justifying Gain-based Relief for Nuisance

395

should be allowed, on the basis that fundamental principle appeared to support the claim, has much to recommend it. Courts that have provided more substantive reasons for denying gain-based relief for nuisance have often relied on unconvincing grounds. Thus, in Abatiello v Monsanto,30 the New York District Court displayed a rather fundamental misunderstanding of the law in dismissing such a claim on the basis that an unjust enrichment claim required a contractual or quasi-contractual relationship between the parties. Another common misconception in American restitution law is apparent in the decision of the District Court of Oklahoma in Quarles v Little River Energy Co.31 Judge Frizzell concluded that, because unjust enrichment actions are equitable in nature, they should be allowed only where legal remedies are inadequate and that this condition cannot not be satisfied where a compensatory claim is available. Both the major and minor premises of the syllogism at the heart of this explanation are flawed. The notion that equitable actions lie only where legal remedies are inadequate is at best dubious and at the very least demands some account of the notion of the adequacy of remedies before it can be applied.32 In any event, there is no reason to characterise the relief sought as equitable: such a claim would represent an extension of the common law of ‘quasi-contract’. There is, however, some American authority for unjust enrichment claims for nuisance. For example, the theoretical availability of such relief is well recognised in Oklahoman law. In Branch v Mobil Oil,33 the District Court of Oklahoma concluded that a claim in unjust enrichment for a gain generated by improperly disposing of pollutants stated a valid cause of action under State law. Subsequently, in NC Corff Partnership Ltd v OXY USA Inc,34 the Oklahoma Court of Appeals upheld an appeal against a decision to grant summary judgment dismissing claims based on unjust enrichment. The action involved a claim by the owners of an industrial park against the former operator of nearby oil and gas wells. The defendant had failed to deal with contamination from its operations with the result that the groundwater of the park was polluted. The plaintiff alleged that the defendant was enriched as a result of saving money in not financing the cleanup for which it was legally responsible. The Court took the view that an action for unjust enrichment would lie provided that the plaintiff could establish that the defendant was responsible for the nuisance, that it could not be abated,35 and 30 Abatiello v Monsanto 522 F Supp 2d 524 (US District Court, New York, Southern District, 2007). 31 Quarles v Little River Energy Co 2008 WL 185715 (US District Court, Oklahoma, Northern District). 32 See, generally, D Laycock, The Death of the Irreparable Injury Rule (Oxford, Oxford University Press, 1991). 33 Branch v Mobil Oil 778 F Supp 35 (US District Court, Oklahoma, Western District, 1991) 36. 34 NC Corff Partnership Ltd v OXY USA Inc, 929 P 2d 288 (Oklahoma CA, 1996) at 295. 35 The role of the requirement that the nuisance should not be capable of being abated is unclear. However, distinctions based on the duration of the nuisance are relatively common: see below text accompanying nn 65–67 and n 126.

396

Craig Rotherham

that the defendant benefited economically as a result. In its apparently categorical analysis of the requirements for a claim for unjust enrichment in the context of nuisance, the Court appeared to go further than simply concluding that the plaintiff ’s pleadings disclosed an arguable cause of action—which is all that would have been strictly necessary for the purposes of determining that the plaintiff ’s cause of action should not have been summarily struck out. Moreover, in sharp distinction to the approach taken by the District Court in Quarles,36 the Court in Corff rejected the notion that a gain-based claim was unavailable where the plaintiff was in a position to bring a claim for compensation for loss. Subsequently, the United States Court of Appeals for the 10th Circuit in Burlington Northern & Santa Fe Ry Co v Grant37 followed Corff in concluding that Oklahoman law provided for an unjust enrichment claim in this context. Courts in other States have, in preliminary proceedings, concluded that pleadings citing unjust enrichment arising out of a nuisance stated tenable causes of actions. Thus, in Evans v City of Johnstown,38 Justice Shea of the New York Supreme Court accepted that the plaintiffs’ statement of claim disclosed an arguable cause of action in alleging that the local municipality gained by not incurring expenditure to maintain a proper sewerage system, with the result that the plaintiffs’ properties were polluted with effluent. Justice Shea drew an analogy between the facts of the case before him and trespass cases where defendants had been required to pay a reasonable fee for using plaintiffs’ property,39 particularly those in which the defendant’s use involved the dumping of waste.40 Similarly, in Russell-Stanley Corp v Plant Industries Inc,41 a New Jersey court refused a motion to dismiss a claim for unjust enrichment relating to the defendant’s failure to take steps to abate contamination resulting from an ‘abnormally dangerous’ activity carried out on the premises in question while they were managed by the defendant. Nothing in the case law as it stands offers a thoroughgoing analysis of the issue. The decisions all took place in the context of preliminary proceedings where the question was whether the claim was an arguable one that deserved to be fully heard instead of being summarily dismissed. In all of the cases, gainbased and loss-based relief were sought in the alternative, and in none of the cases is it apparent that there would have been any substantive or procedural advantage to the plaintiff in claiming restitution rather than compensation. Nonetheless, the cases cited suggest that, as a matter of principle, a gain-based 36

See above n 31 and accompanying text. Burlington Northern and Santa Fe Ry Co v Grant 505 F 3d 1013 (US CA, 10th Cir, 2007). 38 Evans v City of Johnstown 96 Misc 2d 755, 410 NYS 2d 199 (Supreme Court of New York, 1978). 39 Citing the celebrated decision in De Camp v Bullard, 159 NY 450, 54 NE 26 (New York CA, 1899). 40 Evans v City of Johnstown, 96 Misc 2d 755, 410 NYS 2d 199 (Supreme Court of New York, 1978) 766, 205, citing Bomptin Realty Co v City of New York, 276 App Div 1094, 96 NYS 2d 414 (Supreme Court of New York, Appellate Division, 1950). For an analogous case in English law, see Whitwham v Westminster Brymbo Coal and Coke Co [1896] 2 Ch 538 (CA). 41 Russell-Stanley Corp v Plant Industries Inc, 595 A 2d 534 (Superior Court of New Jersey, 1991). 37

Justifying Gain-based Relief for Nuisance

397

claim should be available. In addition, there is some academic support for this position.42 The reasoning offered in the case law thus far suggests that all that a plaintiff would be required to demonstrate was that the defendant had been enriched as a result of an actionable nuisance. There is no suggestion, for example, that a plaintiff would need to establish that the nuisance was wilfully committed.43

IV. Formal Approaches to the Availability of Gain-Based Relief A. Anti-Enrichment Wrongs and Anti-Harm Wrongs Peter Birks developed the thesis that the availability of restitution for wrongs could be explained by three considerations. He identified the wilfulness of the defendant’s conduct as an important consideration and additionally justified strict gain-based liability for breaches of fiduciary duties by the need to deter any wrongdoing in that context. Beyond that, he explained the availability of gainbased relief on the basis of a distinction between anti-harm and anti-enrichment wrongs.44 According to this analysis, where the purpose of the wrong was the prevention of harm as opposed to the prevention of enrichment, recovery would be restricted to compensation for loss. While this appears to offer an elegant method for determining entitlement to gain-based relief, it cannot withstand close analysis. For one thing, it is hard to see what wrong cannot be said to be ‘anti-harm’. After all, it is difficult to think of wrongs for which compensatory damages are not available. Moreover, we might be suspicious of the premise that it is possible to determine in the abstract whether a wrong is of one kind rather than the other without reference to remedies that are made available. Wrongs, along with the rights of which they represent breaches, do not exist entirely free from the relief that is granted to remedy them. We can say that a defendant owes a duty or a claimant has a right because the law responds by providing certain remedies when a defendant acts in a particular way. The availability of some form of relief is essential to the very existence of rights and duties and the range of remedies provided will dictate the content of those rights and duties. If one is sure that a certain wrong is an anti-enrichment wrong, it is probably in large part because it has already been established that gain-based relief is available for such a wrong. To be useful, Birks’ distinction must be able to explain why 42 See A Kanner, ‘Unjust Enrichment in Environmental Litigation’ (2005) 20 Journal of Environmental Law and Litigation 111. 43 cf the position taken in the proposed new Restatement: see below text accompanying nn 137–9. 44 Birks, above n 5, at 328.

398

Craig Rotherham

the relief in question is made available or refused in the first place and guide us in novel cases where the courts have yet to determine whether gain-based relief should be available. In their treatise on the Australian law of restitution, Mason and Carter, noting this problem, observed that ‘the fact that a particular wrong has not been the subject of precedents recognising restitutionary damages is not conclusive. What matters is whether the wrong is in fact of an anti-enrichment type’.45 This, however, begs the question as to whether there is a ‘fact’ of the matter in play here. It might be objected that the notion that rights and duties are simply a reflection of remedies is too narrowly positivistic. In hard cases, when considering whether a certain type of relief should be made available, it is possible to think about the interests in question more holistically as institutions with a purpose that can point to the scope of the right and the relief that should be made available in the event of a breach. This approach, however, requires us to go beyond conceptual distinctions and to reflect on the broader considerations of policy that motivated the recognition of the interest. Birks initially regarded nuisance as a paradigmatic anti-harm wrong, thereby excluding the possibility of gain-based relief. Subsequently, however, he concluded that Millett J’s analysis in Carr-Saunders confounded this interpretation.46 Birks ultimately abandoned the anti-harm/anti-enrichment distinction, conceding that, in explaining the availability of gain-based relief, ‘there is no hope of certainty in the pursuit of a purely conceptual line’.47 Birks’ distinction initially attracted a largely sceptical response from other commentators48 and now attracts little academic interest.49 Nonetheless, in recent years, some Canadian courts have found the categorical nature of Birks’ analysis seductive. Thus, in Reid v Ford Motor Co,50 the Supreme Court of British Columbia heard an application by the plaintiff to amend its pleadings to include a ‘claim for waiver of tort’ in an action brought by her on behalf of members of a class of litigants against a defendant alleged to have provided cars with faulty ignition switches. The plaintiff argued that the defendant had saved on the costs of recalling the vehicles, replacing defective systems and redesigning a new system to eliminate the defect and, moreover, had been positively enriched as a result of sales of replacement ignition units.

45

K Mason and JW Carter, Restitution Law in Australia (Sydney, Butterworths, 1995) 619. See Birks, above n 5 at 471. For an analysis of that case, see text accompanying n 8 above. 47 P Birks, ‘Civil Wrongs: A New World’, Butterworths Lectures 1990–91 (London, Butterworths, 1992) 55. 48 See, eg, R Goff and G Jones (ed), The Law of Restitution, 6th edn (London, Sweet & Maxwell, 2002) 783; Burrows, above n 17, at 479. 49 After previously offering a critique of the concept, the distinction is not discussed in the most recent edition of Goff and Jones: see R Goff and G Jones (ed), The Law of Restitution, 7th edn (London, Butterworths, 2007) para [36-006]. Nor is it treated in some of the other leading texts: see eg Virgo, above n 17. 50 Reid v Ford Motor Co (2006) BCSC 712. 46

Justifying Gain-based Relief for Nuisance

399

The argument is hardly an implausible one. Nonetheless, Gerow J was emphatic in refusing to amend the statement of claim in the manner requested. She reasoned that [r]estitutionary claims are not made in negligence and nuisance because they are in the main ‘anti-harm wrongs’ in relation to which it is impossible, even if they lead to an enrichment of the wrongdoer, to elevate the prevention of enrichment to the level of a primary purpose.51

This is hardly convincing. All wrongs are primarily concerned with the prevention of harm. Indeed, it is arguable that the prevention of enrichment is always a ‘secondary purpose’: gain-based relief is principally designed to punish and deter potentially harmful conduct and serves to protect the legal institutions at issue from being undermined.52 Birks’ analysis was discussed subsequently in Serhan Estate v Johnson & Johnson53 in relation to a claim for negligence, fraudulent misrepresentation and conspiracy in the context of the defendant’s provision of allegedly defective medical devices. The Ontario Divisional Court heard an appeal from the decision of Cullity J in the Ontario Superior Court refusing to strike out a claim for disgorgement of profits as failing to disclose a cause of action.54 While noting Gerow J’s treatment of the question in Reid, Epstein J distinguished the case on the basis that fraud was identified by Gerow J as a tort that could be waived, and took the view that the additional issues that the case raised as to whether a gain-based claim for deceit or negligence could be maintained in the absence of proof of loss raised questions of policy that should be resolved after full argument.55 The most recent consideration of the issue was offered by the Ontario Superior Court in Heward v Eli Lilly & Co56 in the context of a preliminary application for the certification of a class action for negligence in respect of an anti-psychotic drug that allegedly had serious side effects that the defendant had failed to disclose. As an alternative to claims for compensatory and punitive damages, the plaintiffs sought an order for the disgorgement of the defendant’s profits. Again, the defendant argued that the plaintiffs’ gain-based claim failed to disclose a tenable cause of action. Cullity J convincingly explained his reasons for rejecting the defendant’s application in remarking that I am … far from satisfied that … I should conclude that the proposed classification is sound. My principal difficulty is that it is by no means clear to me how one is to determine whether the primary purpose of a tort that has resulted in the enrichment of a

51

ibid, at 29. IM Jackman, ‘Restitution for Wrongs’ (1989) 48 Cambridge Law Journal 302. 53 Serhan Estate v Johnson & Johnson [2006] OJ No 2421. 54 Serhan Estate v Johnson & Johnson (2004) 72 OR (3d) 296. 55 Serhan Estate v Johnson & Johnson [2006] OJ No 2421 [68]. 56 Heward v Eli Lilly & Co [2007] CanLII 2651. Leave to appeal Cullity J’s judgment on the relevant point of law was subsequently denied by Lederman J in the Ontario Divisional Court: Heward v Eli Lilly & Co [2007] CanLII 26607. 52

400

Craig Rotherham

defendant requires it to be placed within one category or the other. If circularity is to be avoided—and the terms ‘anti-enrichment’ and ‘anti-harm’ are not merely labels that indicate that a disgorgement remedy will, or will not be available—there must be some good reason of principle or policy that will enable torts to be classified one way or the other.57

B. Restitution for Wrongs Limited to Proprietary Rights (i) Property Rights and Gain-Based Relief There is a persistent suggestion that the availability of gain-based relief is largely dependent upon whether or not the right breached is proprietary in nature. The purported analytical link between property and gain-based relief could be understood in various ways. Framed in terms of necessary and/or sufficient conditions for the availability of gain-based relief, a breach of a property right could be any one of the following: (a) a necessary and sufficient condition; (b) a necessary but not sufficient condition; or (c) a sufficient but not necessary condition. Alternatively, the matter could be framed less categorically in terms of the existence of a breach of a property right providing a reason for relief that might be buttressed or overridden by other reasons. The notion that a breach of a property right is a necessary and sufficient condition for gain-based relief sometimes surfaces in judicial discourse. Thus, in the Court of Appeal’s controversial decision in Surrey County Council v Bredero Homes Ltd,58 Steyn LJ denied a gain-based claim for the breach of a contractual covenant on the ground that such relief was available only for infringements of property rights. It was on this basis that his Lordship distinguished the celebrated decision in Wrotham Park v Parkside Homes,59 reasoning that gain-based relief had been rightly awarded in that case because the restrictive covenant breached amounted to a proprietary interest.60 In a similar vein, the Court of Appeal recently held, in denying profit-based relief for the tort of breach of statutory duty in Devenish Nutrition Ltd v Sanofi-Aventis SA, that such relief was limited to torts that involve the breach of a proprietary right.61 In addition, the availability of accounts of profit for breaches of intellectual property rights is sometimes 57 ibid, at [39]. Cullity J had already declined to respect the apparently categorical distinction between ‘anti-harm’ and ‘anti-enrichment’ wrongs in certifying the claimants’ class action in Serhan v Johnson & Johnson (2004) 73 OR (3d) 296 (Ontario Superior Court of Justice) against the defendant’s objections that it failed to disclose a cause of action. See above n 54 and accompanying text. 58 Surrey County Council v Bredero Homes Ltd [1993] 1 WLR 1361 (CA). 59 Wrotham Park v Parkside Homes [1974] 1 WLR 798 (Ch). 60 Surrey County Council v Bredero Homes Ltd [1993] 1 WLR 1361 (CA) 1369–70. For a criticism of this view, see L Smith, ‘Disgorgement of the Profits of Breach of Contract: Property, Contract and “Efficient Breach” ’ (1995) 24 Canadian Business Law Journal 121; A-G v Blake [2001] 1 AC 268 (HL) 283 (Lord Nicholls). 61 Devenish Nutrition Ltd v Sanofi-Aventis SA [2008] EWCA Civ 1086 [75] (Arden LJ), [155] (Tuckey LJ). In reaching this conclusion the Court claimed to be following Stoke-on-Trent CC v W & J Wass Ltd [1988] 1 WLR 1406 (CA). It is, however, difficult to see how that decision is consistent with such an interpretation.

Justifying Gain-based Relief for Nuisance

401

explained by reference to notions of property, thereby suggesting that a breach of a proprietary interest is a sufficient condition for such relief.62 Many commentators have also explained the availability of gain-based relief by reference to the proprietary status of the right breached.63 Daniel Friedmann provided one of the most influential accounts of this kind in suggesting that the defendant’s ‘appropriation’ of a claimant’s property right provides a sufficient condition for gain-based relief.64 In his view, however, a breach of this nature is not a necessary condition for such relief: gain-based relief might also be justified for breaches of other rights where it is thought necessary in order to deter conduct of the type in question. Finally, Friedmann favoured an intermediate category of ‘quasi-property’ rights where gain-based relief is available provided the defendant’s breach is deliberate. Friedmann argued that any nuisance that results in injury that is ‘continuous or systematic’ amounts to an appropriation of the claimant’s property rights in the form of an involuntary easement.65 He argued that, in contrast, in the case of more transitory nuisances, ‘there is no appropriation of the injured interest and hence no basis for a restitutionary claim’.66 While seductive, the temporal distinction relied upon here is not entirely convincing. There is no reason why a defendant should not equally be benefited by a transitory breach—should it matter, for example, whether a quantity of a pollutant seeped gradually onto the claimant’s property in a fashion that could be described as ‘systematic’ or whether it was allowed to escape in a single discharge? Other commentators have eschewed any temporal requirement, simply reasoning that it follows from the fact that a nuisance involves a breach of a claimant’s proprietary rights that the defendant should be liable to account for any resulting gain.67

(ii) The Premises of the Notion that Property Entails the Right to Gain-Based Relief The notion that property logically entails a right to gain-based relief appears to be based on a number of interrelated assumptions. First, it presupposes that 62

See, eg, Potton Ltd v Yorkclose Ltd [1990] FSR 11 (Ch) 18 (Millett J) (breach of copyright). See, eg, P Watts, ‘Restitution—A Property Principle and a Services Principle’ [1995] Restitution Law Review 49, 54–6; Weinrib, above n 3; P Jaffey, The Nature and Scope of Restitution (Oxford, Hart Publishing, 2000) 136; D Howarth, ‘Torts and Miscellaneous Other Wrongs’ in S Hedley and M Halliwell (eds), The Law of Restitution (London, Butterworths, 2002); Virgo, above n 17, at 475; R Grantham and C Rickett, Enrichment and Restitution in New Zealand (Oxford, Hart Publishing, 2000) 484–5. 64 D Friedmann; ‘Restitution of Benefits Obtained Through the Appropriation of Property or the Commission of a Wrong’ (1980) 80 Columbia Law Review 504. 65 One weakness in Friedmann’s analysis lies in the notion that a defendant ‘appropriates’ a claimant’s proprietary rights by committing a nuisance. The notion is demonstrably false: claimants retain the right to the quiet enjoyment of their property, for it is this that allows them to sue the defendant: see C Rotherham, ‘The Conceptual Structure of Restitution for Wrongs’ (2007) 66 Cambridge Law Journal 172. The defendant’s ‘appropriation’, if it can be so described, lies in interfering with the claimant’s rights in a fashion that, if granted voluntarily, would amount to a licence or an easement. 66 Friedmann, above n 64, at 531. 67 See eg Virgo, above n 17, at 473. 63

402

Craig Rotherham

property performs a fundamental normative role in legal argument, as illustrated by the claim that the rights at issue follow from the fact that the claimant’s proprietary rights were infringed. Secondly, it assumes that we have some unitary and stable conception of property in our legal culture from which we can deduce such outcomes. Thirdly, it presumes that this conception of property is absolutist enough to demand more than compensation for loss. Finally, the understanding in question is premised on the notion that this absolutist conception of property does not merely dictate that property rights are highly protected in some way but rather logically entails the precise form of relief in question. The following sections examine each of these premises in turn, arguing that none of them withstand close analysis.

(iii) Deriving Legal Rules from Abstract Notions of Property One type of argument explaining the availability of gain-based relief by reference to the fact that the defendant has breached the claimant’s proprietary rights posits that there is an analytic link between the right breached and the remedy in question. Such an assumption is apparent in Friedmann’s statement that ‘[t]he essential attribute of property has traditionally been the existence of a right of exclusive enjoyment, so exploitation of property by another gives rise to an enrichment that is necessarily at the expense of the owner’s right’.68 Arguments postulating an analytic link between a right and a remedy have been proffered in other contexts to justify contested private law norms. Thus, Richard Epstein argued that liability in tort should be strict rather than fault-based, on the ground that it analytically follows from our notion of property that those who damage another’s assets are liable whether or not their actions were morally culpable.69 This, of course, does not represent the state of our law. Moreover, Epstein’s analysis has rightly been criticised as inconsistent with the role that property plays in shaping legal norms.70 The same criticism may be made of arguments that attempt to deduce the right to gain-based relief from a priori notions of property. Accounts premised upon the ‘essential’ attributes of property are misconceived. Property is an ‘essentially contested’ concept,71 the contours of which are always liable to be a matter of controversy, and will vary both from one legal system to another and within a single legal system over time. As Coleman and Kraus have suggested, property can be understood as a ‘conceptual place-holder’

68

ibid, at 507. For the most explicit expression of this view, see R Epstein, Takings: Private Property and the Power of Eminent Domain (Cambridge MA, Harvard University Press, 1985) 97–8. Generally, this perspective underpins his account of strict liability. See R Epstein, A Theory of Strict Liability: Toward a Reformulation of Tort Law (San Francisco, Cato Institute, 1988). 70 JL Coleman, ‘Property, Wrongfulness and The Duty to Compensate’ (1987) 63 Chicago-Kent Law Review 451, 454; NE Simmonds, ‘Epstein’s Theory of Strict Liability’ (1992) 51 Cambridge Law Journal 113. 71 WB Gallie, ‘Essentially Contested Concepts’ (1955–56) 56 Proceedings of the Aristotelian Society 167; J Waldron, The Right to Private Property (Oxford, Clarendon Press, 1988) 51. 69

Justifying Gain-based Relief for Nuisance

403

requiring a normative vision to give it content.72 While any legal culture is liable to understand the allocation of entitlements to resources in terms of notions of property, norms determining who can enjoy these rights, which resources can be owned and the precise content of those rights are likely to vary somewhat.73 There are two plausible interpretations of the role that abstract notions of property play in determining more concrete norms, such as those determining the availability of benefit-based relief.74 One understanding is that such concepts of property have no role to play whatever in the formation of legal norms. According to this plausible vision, the common law develops in an incremental and piecemeal fashion as the result of ad hoc policy responses as novel questions arise without any reference to broad notions of property. Understood in this way, property is largely an organising concept that we utilise to analyse matters after the event. From this positivistic perspective, even if we thought it were true that rights of property are absolute in English law, we might be wary of suggesting that this explains why benefit-focused relief for breaches of property rights is so widely available. For this is liable to confuse cause and effect. Instead of concluding that the rights in question arise because an owner’s rights to property are absolute, a more positivistic view would be that it is only because such rights are recognised that an owner’s rights are absolute. Alternatively, we might seek to understand efforts to deduce specific measures of relief from notions of property as appealing not to some immutable pre-legal concept of property but rather as reflecting a particular, historically-situated conception of property.75 Such a conception could be viewed, in a Dworkinian vein,76 as encapsulating principles that are widely recognised in our law and which might support a particular outcome even if there is no precedent directly on the point. On this view, the conception of property as absolute could be understood as reflecting our legal culture’s commitment to individual autonomy above other values. Such an account could be valid, however, only if the conception in question could really be shown to be part of our legal culture.

(iv) A Unitary Conception of Property? The premise that the right to gain-based relief can be deduced from some absolutist understanding of property is undermined by the absence in our law of any unitary conception of property in the sense of a single indivisible bundle of rights that invariably characterises any legal relationship that we would call proprietary. Our concepts of property developed in a relatively ad hoc fashion in 72 JL Coleman and J Kraus, ‘Rethinking the Theory of Legal Rights’ (1986) 95 Yale Law Journal 1335, 1340. 73 C Rotherham, ‘Property and Justice’ in M Kramer (ed), Rights, Wrongs and Responsibilities (London, Palgrave, 2001). 74 C Rotherham, ‘Conceptions of Property in Common Law Discourse’ [1998] Legal Studies 41. 75 On the distinction between the concept of property and particular conceptions of property, see Waldron, above n 71, at 31. 76 See, eg, R Dworkin, Taking Rights Seriously (London, Duckworth, 1977) 14.

404

Craig Rotherham

the casuistic environment provided by the writ system and the forms of action.77 Proprietary concepts such as the fee simple, the lease and the trust all developed incrementally through the recognition of new remedies and their extension to novel situations.78 The common law protects different assets in diverse ways. Thus, while it readily provides specific remedies for real property, it has generally left those who complain of interferences to personal property to claim damages. In addition, there are substantive differences between equitable and legal property rights. For one thing, equity provides a number of specific remedies unavailable at common law. For another, while the common law has little regard for the interests of innocent purchasers, equitable property rights are defeated by the bona fide purchase of a legal interest. This led some to doubt that the rights recognised in equity could properly be labelled ‘property’79 and others, more plausibly, to revise such absolutist assumptions.80 Similarly, the protection provided for intellectual ‘property’ inevitably varies considerably from that afforded to material assets, thereby provoking a rather arid debate as to whether such rights are truly proprietary. Theorists writing on property have long recognised the absence of any fixed set of entitlements. Thus Hohfeld emphasised that the rights that we characterise as property consist of a bundle of discrete legal entitlements that are separable so that the presence of one entitlement does not necessarily indicate that of any others.81 Similarly, Tony Honoré reasoned that it is not possible to formulate a definition of ownership in terms of a fixed set of necessary and sufficient conditions.82 Instead, drawing on Wittgenstein, Honoré proposed that what legal relationships habitually described as proprietary in nature share is a ‘family resemblance’. He argued that, while there are certain incidents that are typical of ownership, one or more of them might be absent in a legal relationship that we might nonetheless characterise as proprietary.83 77 D Seipp, ‘The Concept of Property in the Early Common Law’ (1994) 12 Law and History Review 29. 78 See AWB Simpson, A History of the Land Law (Oxford: Clarendon Press, 1986) 47–65 (fee simple); 71–7 (leases); 181–206; and JH Baker, An Introduction to English Legal History, 3rd edn (London, Butterworths, 1990) 286 and 350–51 (trusts). 79 See eg FW Maitland, Equity: A Course of Lectures, 2nd edn (Cambridge, Cambridge University Press, 1936) 47; H Stone, ‘The Nature of the Rights of the Cestui Que Trust’ (1917) 17 Columbia Law Review 467. 80 See eg T Honoré, ‘Rights of Exclusion and Immunities Against Divesting’ (1960) 34 Tulane Law Review 453, 466. 81 WN Hohfeld, ‘Some Fundamental Legal Conceptions as Applied in Judicial Reasoning’ (1913) 23 Yale Law Journal 16; and (1917) 26 Yale Law Journal 710. For an exploration of the consequences of Hohfeld’s analysis in American legal thought, see M Horwitz, The Transformation of American Law 1870–1960: The Crisis of Legal Orthodoxy (New York, Oxford University Press, 1992) ch 5. 82 T Honoré, ‘Ownership’ in AG Guest (ed), Oxford Essays in Jurisprudence (Oxford, Oxford University Press, 1961) 107. 83 Consequently, it is an error for commentators to rely on Honoré to argue that a legal outcome is inevitable on the basis that the right to profits is an essential element of a proprietary relationship. Cf S Worthington, ‘Justifying Claims to Secondary Profits’ in EJH Schrage (ed), Unjust Enrichment and the Law of Contract (The Hague, Kluwer, 2001) 463.

Justifying Gain-based Relief for Nuisance

405

It is perhaps tempting to treat certain entitlements as entailed by the very notion of property in that we can point to some core set of rights that any legal system is likely to confer upon owners of resources. In these terms, the protection of our existing relationships with particular resources by some degree of rights of exclusion and some degree of freedom from expropriation can be seen as central to any conception of property. In contrast, the right to require the defendant to account for any profit made from infringing a proprietary right is liable to be seen as less elementary and the extent to which it is recognised is likely to vary among different legal cultures. The fact that there is no consensus in our legal culture that there is a right to claim a gain made from a breach of property is demonstrated by the tendency of courts to insist that relief is normally available only where the claimant has suffered loss.84

(v) Does Our Law Treat Property as Absolute? The varying scope of protection afforded to different proprietary interests indicates that it is implausible to conceive of property as absolute in English law. The reality is that our judge-made law makes numerous compromises whereby property rights are not protected as completely as they might be. A pertinent illustration of the limits of proprietary rights is provided by judicial practice in the exercise of the jurisdiction to award damages for future loss of amenity in lieu of issuing an injunction. The courts have sometimes declined to exercise this jurisdiction on the grounds that the refusal of an injunction would amount to an expropriation of property rights—a view that supports the notion that property is absolute.85 Nonetheless, the courts routinely exercise the jurisdiction in the context of mandatory injunctions86 and, occasionally, for prohibitory injunctions.87 In limiting claimants to damages, the courts allow a breach of property rights to continue, albeit compensated. This was so in Jaggard v Sawyers,88 where the Court of Appeal refused to award an injunction for a breach of a restrictive covenant and, instead, required the defendant to pay an award reflecting what would amount to a reasonable sum for the relaxation of the right in question. Millett LJ remarked that ‘references to the “expropriation” of the plaintiff ’s property are somewhat overdone’.89 In his view, the refusal of an injunction could not be properly characterised as an expropriation

84 See eg Stoke-on-Trent City Council v W & J Wass Ltd [1988] 3 All ER 394 (CA) 397 (Nourse LJ); Halifax Building Society v Thomas [1996] Ch 217 (CA) 227–8 (Gibson LJ). 85 See eg Shelfer v City of London Electric Lighting Co [1895] 1 Ch 287 (CA) 316 (Lindley LJ); Anchor Brewhouse Developments Ltd v Berkley House (Docklands Developments) Ltd (1987) 38 Build LR 87 (Ch) 101 (Scott J). 86 Isenberg v East India House Estate Co Ltd (1863) 3 De G J & S 263, 46 ER 637; Wrotham Park Estate Co Ltd v Parkside Homes Ltd [1974] 1 WLR 798 (Ch). Cf Regan v Paul Properties DPF No 1 Ltd [2007] Ch 135 (CA). 87 See eg Bracewell v Appleby [1975] Ch 408 (Ch). 88 Jaggard v Sawyers [1995] 1 WLR 269 (CA). 89 ibid, at 287.

406

Craig Rotherham

of the claimant’s rights in any strong sense because the decision to issue an injunction lay within the court’s discretion. Thus, it seems that a proprietary interest does not dictate an absolute right to injunctive relief. Instead property rights are defeasible in that the courts are effectively entitled to sanction the compulsory purchase of the claimant’s right by the defendant. The judicial willingness to award damages in lieu of mandatory injunctions demonstrates the error of attempting to derive concrete norms from an absolutist notion of property. It would be similarly mistaken to endeavour to deduce a right to gainbased relief in this way.

(vi) Difficulties in Deriving Precise Norms for Gain-based Relief from an Absolutist Conception of Property A further problem with the resort to proprietary explanations in this context is that it is difficult to see that some notion of property can dictate either the choices between different measures of gain-based relief or the choice between personal and proprietary relief. Some commentators explain the availability of ‘reasonable fee’ awards by reference to notions of property but suggest that further justifications are needed to explain the availability of accounts of consequential profits.90 Yet, there is equally a tendency to explain the right to consequential profits by an appeal to property.91 It is unclear, however, whether some abstract notion of property can usefully inform our choice as to whether defendants should be charged what would be a reasonable fee for the right to have acted as they did or whether they should account for the consequential benefits obtained from exploiting the claimant’s property. Equally, if we accept the notion that a conception of property entails a right to gain-based relief, it is difficult to see why the claimant should be restricted to personal rights. If property logically entails a claim to any benefit derived from an owner’s property, why should it not invest the claimant with a proprietary interest in any profit so derived? Yet, if this were so, instead of a claim for restitution, the common law should provide for a claim for conversion for any profits derived from the claimant’s property. Similarly, equity would make the constructive trust more widely available for proprietary wrongs. For example, a claimant would be

90 See, eg, J Edelman, Gain-Based Damages: Contract, Tort, Equity and Intellectual Property (Oxford, Hart Publishing, 2002); and Jaffey, above n 63. 91 Such reasoning is particularly apparent in the context of intellectual property, where the account of profits is widely used. See, eg, A-G v Guardian Newspapers (No 2) [1990] 1 AC 109 (Ch) 161 (Scott J) (breach of confidence) and Potton Ltd v Yorkclose Ltd [1990] FSR 11 (Ch) at 18 (Millett J) (breach of copyright). Intellectual property scholars, in contrast, tend to focus on a complex of different normative justifications and institutional considerations in choosing between different measures of relief, see eg, R Blair and T Cotter, Intellectual Property: Economic and Legal Dimensions of Rights and Remedies (Cambridge, Cambridge University Press, 2005).

Justifying Gain-based Relief for Nuisance

407

invariably entitled to a constructive trust in the context of breaches of intellectual property rights.92 The restrictions placed on the recovery of consequential profits and the apparent reluctance to invest claimants with proprietary rights in such profits demonstrate the futility of explaining gain-based relief in terms of some apparently absolute and immutable conception of property. Whether and to what extent owners can claim consequential profits made from the breach of their proprietary rights will generally depend, not on abstract notions of property, but rather on judicial judgements of the importance of the rights in question, the moral quality of defendants’ behaviour and the deterrent effect of such relief.93 Similarly, the limited availability of proprietary relief is not the result of the immanent logic of some conception of property; it is rather the reflection of concern for the interests of defendants’ general creditors.94

(vii) Taking Property Rights Less Seriously We would expect most legal systems to provide some degree of gain-based recovery where one person benefits from breaching another’s rights. Such relief punishes those committing breaches, thereby exacting due retribution and symbolically signalling our law’s commitment to those rights; equally, it deters breaches in the future. Nothing just said, however, would indicate that we should draw distinctions between breaches of property rights and infringements of other rights. Indeed, it could be argued that the same considerations come into play in deciding when and to what extent gain-based relief should be available for breaches of contract.95 While such relief might be thought to be merited for most breaches of proprietary rights, it would be no real surprise to see some exceptions being made on the basis of a mixture of considerations regarding, for example, the quality of the defendant’s conduct and the importance of the rights in question.96 Whether the right is proprietary or not, the issue is whether such relief is desirable for the better protection of the institutions in question.

92 For recent discussions of the issue of proprietary relief for intellectual property rights, in the context of breach of confidence, see HW Tang, ‘Confidence and the Constructive Trust’ (2003) Legal Studies 136; M Conaglen, ‘Thinking about Proprietary Remedies for Breach of Confidence’ (2008) Intellectual Property Quarterly 82. 93 H Dagan, Unjust Enrichment: A Study of Private Law and Public Values (Cambridge, Cambridge University Press, 1997). 94 On the considerations of policy regarding the imposition of constructive trusts, see E Sherwin, ‘Constructive Trusts in Bankruptcy’ [1989] University of Illinois Law Review 297; C Rotherham, Proprietary Remedies in Context: A Study in the Judicial Redistribution of Property Rights (Oxford, Hart Publishing, 2002); A Duggan, ‘Gain-Based Remedies and the Place of Deterrence in the Law of Fiduciary Obligations’ ch 15 of this book. 95 See A-G v Blake [2001] 1 AC 268 (HL) 283 (Lord Nicholls). 96 Rotherham, above n 65, at 194–5.

408

Craig Rotherham

V. Weighing the Case for Gain-Based Relief for Nuisance A. The Notion that Gain-Based Relief Should be Available for All Benefits Gained from Any Wrong Some commentators have been content with the notion that any wrong from which a defendant derives a profit should give rise to a gain-based claim.97 Thus, Goff and Jones argue that English courts should adopt the principle that, ‘if it can be demonstrated that the tortfeasor has gained a benefit and that benefit would not have been gained but for the tort, he should be required to make restitution’.98 They note, however, that this does not represent the current state of the law. The idea that gain-based relief might be available for every nuisance that enriches the defendant merits serious consideration. This position could be defended on the basis that it punishes culpable defendants99 and/or deters future wrongdoing.100 In addition, in some instances, a gain-based claim might be favoured as a substitute for loss-based relief where the claimant’s loss is liable to be too subjective to satisfy conventional methods of quantification.101 On the other hand, it is difficult to see that there is any other compelling reason for awarding the defendant’s gain to the claimant. Generally, the enrichment in question is treated as a windfall of which the defendant should be stripped rather than something to which the claimant has a compelling claim.102 Given that the gain was made by breaching the claimant’s rights, it would be preferable that the windfall should be enjoyed by the claimant rather than the defendant. If, however, there were another way of disgorging the benefit, for instance, by its seizure by the state, the claimant could not really object.103 97 See eg PD Maddaugh and JD McCamus, The Law of Restitution (Aurora, Ontario, Canada Law Book, 2005) 24–9. 98 Goff and Jones, above n 49. 99 Thus some commentators take the view that punishment plays at least some role in this context. See eg Jaffey, above n 63, at 136; S Hedley; Restitution: Its Division and Order (London, Sweet & Maxwell, 2001) 85; Birks, above n 47, at 86; Burrows, above n 17, at 47. Similarly, the Law Commission of England and Wales took the view that ‘deliberate and outrageous disregard of the claimant’s rights’ was sufficient grounds for ‘restitutionary damages’ in the case of tortious and equitable breaches of rights: see Report on Aggravated, Exemplary and Restitutionary Damages (Law Com No 247, 1997) 43. 100 For commentators for whom deterrence plays at least as important a role in this context, see, eg, Edelman, above n 90; Birks, above n 47, at 96; Dagan, above n 93; Virgo, above n 17, at 434; Burrows, above n 17, at 480. 101 A consideration emphasised by Lord Nicholls in A-G v Blake [2001] 1 AC 268 (HL) 278–84. 102 See eg GE Palmer, Law of Restitution, (Boston, Little, Brown & Co, 1978) vol 1, § 2.11 at 141. 103 The Court of Appeal’s controversial decision in Halifax Building Society v Thomas [1996] Ch 217 (CA) can be defended on this basis. The property that the defendant purchased with a mortgage obtained by defrauding the claimant rose considerably in value. After the defendant defaulted on the loan, the claimant exercised its rights as mortgagee to sell the property and recoup the sums outstanding under the loan. However, a considerable surplus remained. The claimant brought an action arguing that the defendant should account to it for this profit but it found itself in competition with

Justifying Gain-based Relief for Nuisance

409

There are, on the other hand, a number of arguments against awarding gain-based relief for nuisance or at least for restricting the circumstances in which it is made available and/or limiting the quantum awarded. For one thing, there is a considerable degree of reluctance among commentators and courts to acknowledge that punishment is a legitimate objective for gain-based relief.104 It is thought that, if it were to pursue retributive objectives, the civil law would impinge on the province of the criminal law without being curbed by the same safeguards that constrain the latter branch of the law. Certainly, if gain-based relief were thought to be justified for punitive reasons, it is likely to be reserved for instances where the defendant’s conduct is especially worthy of reproach. Thus, if justified on this basis, gain-based relief would in all likelihood be restricted to deliberate wrongdoing. It is unlikely that negligence would be regarded as providing a level of fault sufficient to justify a punitive response from the civil courts. Where, in contrast, gain-based relief is justified on the basis of deterrence, it is important to understand that the existence of a post facto gain from a breach may be a matter of chance rather than an indication that disgorgement is needed to ensure that such breaches are discouraged. Optimal deterrence needs to be considered from the perspective of incentives weighing upon a potential infringer ex ante any breach. The extent to which deterrence is valued is likely to depend upon whether such wrongs are liable to result in financial harm and/or they transgress a right that is regarded as particularly important for non-economic reasons. Whether gain-based relief is needed to provide adequate deterrence will depend on a number of considerations including the strength of other sanctions for the conduct and the risk of under-enforcement of the right. Courts also need to ensure that the benefits brought by deterrence outweigh any associated costs. There is a risk, for example, that supra-compensatory relief might lead parties to make investments to safeguard against breaches that exceed the potential costs of such breaches or to avoid certain socially useful conduct altogether. In addition, supra-compensatory remedies can potentially create moral hazards by not providing right-holders with adequate incentive to prevent breaches and/or to intervene quickly when a breach has taken place.105 A concern for adequate deterrence may justify a presumption in favour of the availability of gain-based relief for deliberate wrongs. Those who have wilfully infringed are more likely to have weighed the benefits and risks of their conduct and are therefore more likely to be responsive to the incentives that supra-compensatory relief brings to bear. On the other hand, gain-based relief the Crown Prosecution Service, which was seeking to claim the profit as the proceeds of a criminal act. The Court found in favour of the Crown Prosecution Service. See Burrows, above n 17, at 476. 104 For academic perspectives, see Edelman, above n 90; and H Dagan, The Law and Ethics of Restitution (Cambridge, Cambridge University Press, 2004). For judicial expressions denying punishment a role in this context, see eg My Kinda Town v Soll [1983] RPC 15 (Ch) 55 (Slade J); and Celanese International Corporation v BP Chemicals Ltd [1999] RPC 203 (Ch) 248 (Laddie J). 105 G Calabresi and D Melamed, ‘Property Rules, Liability Rules and Inalienability: One View of the Cathedral’ (1972) 85 Harvard Law Review 1089.

410

Craig Rotherham

is less likely to deter breaches that are merely negligent. Whether such relief can discourage carelessness is likely to depend on a number of considerations regarding the nature of the norms in question and the extent to which actors in a particular context are likely to be responsive to legal incentives.

B. Gain-based Relief in the Context of Nuisance (i) The Vagueness of Nuisance Norms There is nothing necessarily inconsistent in the law’s making gain-based relief available for some wrongs but not others. Because legal interests can be quite diverse in nature it is understandable that they are protected to different degrees and in disparate ways.106 Thus, the state of the law suggests that, while it might well be appropriate that gain-based relief should be available for breaches of some interests even if the wrong in question was not wilful, in other instances it might be appropriate to make conscious wrongdoing a pre-condition for such relief, to limit the quantum of any award or even to deny this form of relief altogether. One way of approaching the issue is to compare the arguments for granting gain-based relief for nuisance with those for awarding such relief for trespass. The comparison is instructive, first, because the provision of gain-based relief is well established for trespass and, secondly, because trespass is often regarded as the tort most closely analogous to nuisance. At first glance, the two torts have much in common in that both involve an interference with claimants’ rights to the exclusive enjoyment of their land. While a trespass is marked by a defendant’s physically invading the claimant’s land and thereby breaching the claimant’s privilege of exclusive possession, nuisance involves defendants’ using their land in a way that interferes with claimants’ quiet enjoyment of their own land. While the abstract similarities of the two torts are clear, it is equally true that they differ in important ways. The rules that determine the existence of a trespass are more clear-cut than those governing nuisance.107 A trespass is committed where the defendant transgresses a geographical limit that is, or at least could be, clearly delineated. Because the existence of a boundary crossing is generally obvious, the advantages of favouring bright-line rules for determining liability and allocating remedies are readily apparent. Thus, almost any boundary crossing will be a trespass per se, with no requirement to prove loss and with little available in the way of defences or excuses. Moreover, the claimant will almost invariably be entitled to an injunction to prevent on-going transgressions rather than being limited to monetary relief. While such a regime might be justified as the best means of promoting individual

106 A theme that is most fully explored in Dagan, above n 93. See also, D Friedmann, ‘Restitution for Wrongs: The Measure of Recovery’ (2001) 79 Texas Law Review 1879. 107 For a discussion of the choice between determinate rules and indeterminate standards, see E Sherwin, ‘The Rules of Obligations’ ch 18 of this book.

Justifying Gain-based Relief for Nuisance

411

autonomy, it is also, in all likelihood, the most efficient regime available.108 By providing a clear rule, the norm has several advantages. Ex post an alleged breach, the rule leaves little doubt as to whether an infringement has taken place. Ex ante, it allows individuals to organise their behaviour so as to avoid transgressing the norm and provides the parties with a clear starting point for negotiations should they wish to alter their rights. It is not difficult to see how gain-based relief can be viewed as having a role to play in reinforcing the framework of incentives provided by the trespass norms. By raising the costs associated with being sued for an infringement, the provision of gain-based relief can serve to discourage trespasses being committed in the first place and to encourage bargaining instead. The possibility that the defendant might escape being required to pay a reasonable fee or to account for consequential profits would undermine the incentives to bargain that are provided by the tort’s strict standard of liability and the automatic availability of injunctive relief. Equally, the provision of gain-based relief can be justified on the basis of fairness, as the defendant’s transgression of a clear boundary will generally be deliberate or at the very least grossly careless. Nuisance is, in contrast, very different. While there will, of course, be examples of clear breaches, there will equally be many instances where it is less apparent what the law requires. For, rather than a single bright-line rule, the law of nuisance is inevitably more nuanced, incorporating relatively vague standards. English nuisance law, for example, provides for a mixture of rigid rules and more flexible standards. Thus, it favours bright-line rules in providing that physical damage to a claimant’s property is actionable as of right109 and that prohibitive injunctions are available as of right without regard to considerations of the public good.110 Equally, however, English nuisance law incorporates standards that necessitate the balancing of competing interests. Thus, in contrast to the rule prevailing for physical damage to property, disturbances that interfere with claimants’ less tangible enjoyment of their land are actionable only if they are unreasonable having regard to the nature of the locality in question.111 In addition, courts have held that temporary building works that create a level of disturbance that would ordinary be characterised as a nuisance will not be actionable if carried out with reasonable skill.112 American law goes even further in regulating competing land uses through standards rather than rules.113 The current orthodoxy is that the

108 See T Merrill, ‘Trespass, Nuisance and the Costs of Determining Property Rights’ (1985) 15 Journal of Legal Studies 13. 109 St Helens v Tipping (1865) 11 HL Cas 642, 11 ER 1483. 110 A-G v Birmingham Corporation (1859) 4 K & J 528, 70 ER 220. 111 Sturges v Bridgman (1879) 11 Ch D 852 (CA). 112 Andreae v Selfridge & Co Ltd [1938] 1 Ch 1 (CA). 113 Although at least commentator has argued that, in contrast to the academic orthodoxy, in practice, US courts favour bright-line rules in much the same way as their English counterparts: see HE Smith, ‘Property and Exclusionary Rules in Nuisance’ (2004) 90 Virginia Law Review 965.

412

Craig Rotherham

finding of a nuisance is automatic only in cases of ‘severe’ harm, while liability in other cases is dependent upon a utilitarian balancing test.114 In addition, there is a greater readiness to refuse to grant injunctions and to limit claimants to damages where specific relief would be unjust or inefficient.115 The vagueness of the prevailing norms makes a blanket provision of gain-based relief more difficult to justify in the context of nuisance. For one thing, it throws into doubt the extent to which defendants really merit punishment in this context. As law and economics scholars have emphasised, nuisances can be more usefully understood as involving clashes of incompatible uses of land rather than one party’s inflicting harm upon the other.116 While there will obviously be instances of blatant infringements, there will equally be cases where the finding of a nuisance is a close thing. Moreover, human psychology is such that defendants will tend to be self-serving in their interpretation as to whether their actions are reasonable,117 thereby increasing the likelihood of defendants’ committing nuisances while honestly believing that their activity was within legal bounds. The provision of gain-based relief in such cases would show too little regard for defendants whose actions merit little or no reproach. From the point of view of deterrence, it is more difficult to justify extracompensatory sanctions for the breach of vague standards than it is for bright-line rules. The role of deterrence is to provide incentives for future behaviour; but, in the absence of clear rules, it is difficult for parties to calculate the costs of their actions. Making supra-compensatory relief available in respect of marginal transgressions of vague standards is liable to result in excessive precaution leading to over-investment in inefficient pollution mitigating measures or parties’ eschewing altogether investment in socially useful activity.

(ii) Choosing Remedies in the Light of Market Failures As mentioned, one justification for deterring breaches of rights by providing for gain-based relief is in order to encourage parties to bargain for a licence to act in the manner contemplated.118 However, this rationale loses some of its force where obstacles make it difficult to contract. This is liable to be the case where an activity is likely to interfere with the rights of a significant number of parties so that transaction costs are prohibitive and the danger of strategic bargaining high.119 While trespass typically involves an interference with a single 114 Restatement of the Law, Second, Torts (St Paul MN, American Law Institute Publishers, 1979) §§ 826 and 829. 115 See eg Boomer v Atlantic Cement Co, 257 NE 2d 870 (New York CA, 1970). 116 See R Coase, ‘The Problem of Social Cost’ (1960) 3 Journal of Law & Economics 1, 2; Calabresi and Melamed, above n 105, at 1102. 117 See P Valdesolo and D Desteno, ‘The Duality of Virtue: Deconstructing the Moral Hypocrite’ (2008) 44 Journal of Experimental Social Psychology 1334. 118 J Gordley, Foundations of Private Law: Property, Tort, Contract, Unjust Enrichment (Oxford, Oxford University Press, 2006) 451–2. 119 See Calabresi and Melamed, above n 105; Merrill, above n 108.

Justifying Gain-based Relief for Nuisance

413

landowner’s rights, nuisance very often involves more diffuse disturbances that affect numerous owners’ enjoyment of their land. Thus, in the context of any interference that affects enough parties to be characterised as a public nuisance, it would make little sense to go beyond compensatory relief and favour injunctions and gain-based relief if the intention is to encourage defendants to bargain for the right to conduct the activity in question.120 The provision of injunctive relief creates the potential for strategic bargaining whereby, in return for the grant of a licence to pollute, a single landowner is in a position to exact a fee out of all proportion to the real value of the right in question.121 A right to gain-based relief potentially gives rise to the similar hold-out problems. Thus, particularly in circumstances in which the courts would be likely to refuse to grant an injunction, the possibility of being able to claim any gain generated by a nuisance would encourage those potentially affected by a nuisance to endeavour to exact an extortionate fee for consenting to the activity. This danger could be mitigated by courts’ refusing to grant accounts of profits and limiting claimants to reasonable fee awards that reflect the market price that would be paid by a willing seller rather than a sum that might be demanded by a claimant benefiting from a bilateral monopoly.122

(iii) Nuisance as a Means of Internalising Externalities It is unclear that the common law can provide effective incentives by awarding gain-based relief in respect of nuisances that were not deliberately caused. The determination of responsibility to compensate does not necessarily amount to a finding of moral culpability. Instead, the tort of nuisance can be understood as ensuring that, in response to the problems arising from incompatible uses of land, certain parties are required to bear the costs of their activities without regard to issues of blameworthiness. This is perhaps most obvious in cases of uses of land that fall outside the protection of the ‘reasonable user’ principle. Defendants engaging in such uses of land will be liable for any resulting harm regardless of fault.123 In this context, gain-based relief would be inappropriate, for the law does not wish to deter such behaviour. Rather, it seeks only to ensure that those undertaking it take responsibility for the risks involved by bearing all the costs of their 120 Merrill, above n 108. Of course, an approach that favours the award of an injunction might be justified nonetheless as providing a fairer or more workable standard than any other alternative. See, eg, JE Penner, ‘Nuisance and the Character of the Neighbourhood’ (1993) 5 Journal of Environmental Law 1; Smith, above n 113. 121 See Calabresi and Melamed, above n 105. 122 This is consistent with the relief granted at common law in English courts. See eg Wrotham Park Estate Co Ltd v Parkside Homes Ltd [1974] 1 WLR 798 (Ch); Jaggard v Sawyers [1995] 1 WLR 269 (CA); Bracewell v Appleby [1975] Ch 408 (Ch). 123 Cambridge Water v Eastern Counties plc [1994] 2 AC 264 (HL) 299 (Lord Goff). This applies a fortiori to that category of uses characterised as ‘non-natural’ in Rylands v Fletcher (1866) LR 1 Ex 265 (Ex Chamber). While such uses have been generally treated as being subject to a separate tort, they are probably better understood, as Lord Goff argued in Cambridge Water, as subject to ordinary principles of nuisance.

414

Craig Rotherham

activities rather than allowing some of them to spill over onto other members in the community.124 Forcing compliance with specific environmental standards and/or requiring actors to adopt specific emission-limiting measures can be better achieved through government regulation. It may be that generally the best that the law of nuisance can do is to ensure that negative externalities are internalised. As a result, gain-based relief might be best reserved for nuisances that are deliberately caused or, at least, for breaches where the defendant’s behaviour was clearly socially detrimental and demanding of punishment and/or deterrence.

(iv) Proportionality Where a claimant has profited from an activity that has generated a nuisance, the court might award, at one extreme, all profits causally linked to the wrong. Alternatively, the court might award relief according to a hypothetical fee that is linked to a share of the defendant’s profits. The approach to fixing such awards is highly flexible: the figure chosen may reflect not only any burden endured by the claimant but might also vary according to the strength of the parties’ relative bargaining positions. It may be that the claimant effectively benefited from a bilateral monopoly in that only the claimant could have enabled the defendant to exploit his land in the desired manner. In these circumstances, the court might conclude that appropriate fee is anywhere between, on the one hand, the value that would be attached to such an easement in circumstances where there is no question of a bilateral monopoly and, on the other hand, the maximum price that an avaricious landowner might have extracted in light of her position of power. The price chosen is liable to vary according to the court’s sympathy for the defendant in light of the circumstances of the breach.125 Courts ought, however, to have regard to considerations of proportionality in relation to nuisances that, while causally necessary for the profit in question, were relatively minor in nature. Particularly in circumstances in which the activity has ceased or been enjoined, the court might conclude that it should limit relief to compensation for loss or at least to a rather minimal reasonable fee.126

(v) The Significance of the Permanent Deprivation of Rights An important distinction can be drawn between gain resulting from past nuisances that the defendant has ceased or has been enjoined by the court from 124 See, eg, Orchard View Farms Inc v Martin Marietta Aluminium Inc, 500 F Supp 984 (US District Court, Oregon, 1980). 125 In Jaggard v Sawyers [1995] 1 WLR 269 (CA) 282–3, Millett LJ suggested that ‘[i]n situations of this kind a plaintiff should not be treated as eager to sell, which he very probably is not. But the court will not value the right at the ransom price which a very reluctant plaintiff might put on it’. Other decisions, however, indicate the extent of discretion available to the courts: see eg Ministry of Defence v Ashman [1993] 2 EGLR 102 (CA) (subjective devaluation allowed); Inverugie Investments Ltd v Hackett [1995] 3 All ER 841 (PC) (defendant who had wrongly ejected a hotelier and operated the business for 15 years at a loss required to pay a reasonable fee derived from the prevailing market rental for shortterm occupancy rather than rate discounted for long-term occupancy). 126 Friedmann, above n 106.

Justifying Gain-based Relief for Nuisance

415

continuing, and gain that will result where a defendant is permitted to continue the activity in the future. The availability of an injunction provides significant protection for the claimant’s subjective preferences. Where, in contrast, a court declines to enjoin the activity in question and limits the claimant to monetary relief, it is likely to recognise that this represents a greater interference with the claimant’s rights. Thus, where relief is awarded in exchange for allowing the defendant to continue the activity in question and thereby serves, in some sense, as the price given in exchange for a compulsory purchase of the claimant’s property rights, the courts are more likely to award a reasonable fee and the quantum of any such fee is likely to be higher than in cases where relief is sought for past interferences with the claimant’s rights. A distinction of this sort appears to play a role in the use of gain-based relief for nuisance in American law. Thus, at least some courts have indicated that it is a precondition for such relief that the nuisance in question must be incapable of abatement.127 A similar distinction may be in play in English law, where considerable emphasis has been placed on whether or not a court is exercising the jurisdiction to award damages in lieu of an injunction.128 However, at present, the distinctions drawn appear to be too formalistic and arbitrary to correspond to any salient normative considerations.

VI. Pragmatism in Practice: Gain-Based Relief for Nuisance in The Restatements of Restitution § 3 of the first Restatement of Restitution proffered the disarmingly simple principle that, ‘[a] person is not permitted to profit by his own wrong at the expense of another’.129 Nonetheless, the accompanying comments of Warren Seavey, the reporter for that part of the Restatement devoted to ‘quasi-contracts’, betrayed a real degree of ambivalence as to the merits of such relief. Thus, Seavey remarked, that [t]he desirability of permitting restitution in such cases is ordinarily not so obvious as in the cases where there has been no tort since the tortfeasor is always subject to liability in an action for damages and … the right to maintain an action for restitution in such cases is largely the product of imperfections in the tort remedies, some of which imperfections have now been removed.130

In reality, the first Restatement’s treatment of restitution for wrongs represents a strange mixture of a very broad proposition of general principle with a very 127

Above n 35 and accompanying text. See above nn 13–15 and accompanying text. Restatement of the Law of Restitution, Quasi Contracts and Constructive Trusts (St Paul MN, American Law Institute Publishers, 1937) § 3. 130 ibid, at § 3 comment. 128 129

416

Craig Rotherham

limited recognition of specific wrongs for which gain-based recovery was actually acknowledged as being available. In particular, based on his reading of the state of the prevailing case law, Seavey took the view that the right to gain-based relief in respect of trespass to land was very limited indeed.131 The Restatement’s discussion of restitution for wrongs largely ignored instances of mere infringements of property rights and was, instead, framed in terms of defendants’ physically misappropriating assets, drawing its examples largely from the law of conversion. Given this, it is perhaps not surprisingly that there is no discussion of the tort of nuisance. Efforts to provide a revised account of the law of restitution for the second series of Restatements came to nothing and the only draft for discussion that came out of the project did not deal in any depth with restitution for wrongs.132 Consequently, an account of the law drafted by Andrew Kull as part of the third series of Restatements will provide the first comprehensive interpretation of the American law regarding gain-based relief for over 70 years. Chapter five of the tentative draft of the Restatement (Third) of the Law of Restitution and Unjust Enrichment is devoted to the subject of restitution for wrongs. Rather like the first Restatement, the introductory note to the proposed revision presents the strikingly simple general principle that, ‘[g]ains realized in violations of another’s legally protected rights must be given up to the person whose rights have been violated’.133 The following four sections of the proposed new Restatement provide a series of rules illustrating the application of this principle. These sections are devoted in turn to: trespass and conversion; the misappropriation of financial assets;134 intellectual property rights; and fiduciary or confidential relationships. However, chapter five closes with a catchall provision, § 44, which provides that defendants may be accountable for benefits obtained as a result of an ‘interference with other protected interests’ in cases not covered by the previous four sections.135 § 44 is replete with qualifications and provisos that allow for a considerable degree of balancing of interests, an approach with which American legal culture appears more comfortable than its English counterpart.136 Interestingly, the threshold for gain-based liability for miscellaneous legal interests in § 44 is set higher than for those interests that are specifically mentioned

131 ibid, at § 129. The position taken in the Restatement reflected the state of the law at the time the final version was concluded in May 1936. This was a month before the groundbreaking decision in Edwards v Lee’s Administrator, 96 SW 2d 1028 (Kentucky CA, 1936), which in turn was delivered before the publication of the Restatement and therefore without the benefit of its view on the area. 132 Although the tentative draft in question reiterated the general principle providing liability for gains derived from wrongs. See Restatement of the Law, Second, Restitution (Tentative Draft No 1) (American Law Institute, 1983) § 4. 133 Restatement of the Law, Third, Law of Restitution and Unjust Enrichment (Tentative Draft No 4) (American Law Institute, 2005) 37. 134 It might be objected that such breaches will generally involve restitution by subtraction and therefore have little to do with the controversies with which this area is generally concerned. 135 Restatement of the Law, Third, Law of Restitution and Unjust Enrichment, above n 133, at § 44. 136 See Horwitz, above n 81.

Justifying Gain-based Relief for Nuisance

417

in the preceding sections. A defendant will be liable to make restitution only for a ‘conscious interference with another’s legally protected interests’.137 This may be contrasted with the proposed new Restatement’s position that, ‘[a]s a general rule, even an innocent and inadvertent interference with the claimant’s property leads to a liability in restitution for the value of what has been taken or the use made of it’.138 Generally, the presence of knowledge is relevant only in determining whether the defendant is liable for consequential profits as opposed to what would have been a reasonable fee for the claimant to have charged in return for the relaxation of the right in question.139 In addition, § 44 indicates that conscious breaches of miscellaneous interests will not entitle a claimant to gain-based relief if ‘competing legal objectives make such liability inappropriate’.140 More specifically, the section lists more detailed reasons for which restitution will be denied.141 First, restitutionary relief will not be provided ‘if the court would refuse to enjoin the wrongful acts of the defendant, assuming timely application and an absence of procedural or administrative obstacles’.142 This provision suggests that, in circumstances where the courts would refuse to award an injunction, similar considerations would mean that it would be equally inappropriate to award gain-based relief. However, it is difficult to see that this would invariably be the case. In all likelihood, the normal rule that deliberate wrongdoers are liable for consequential profits should not apply. For, such relief would be liable to have the same effect as imposing an injunction— effectively allowing the claimant to hold the defendant to ransom.143 However, it is less clear why the award of a reasonable fee would necessarily be inappropriate. Such a relief would seem to be the least the court should award if the defendant’s breach was deliberate. Moreover, given that the denial of an injunction has the effect of allowing the defendant to compulsorily acquire a proprietary interest from the claimant, the courts should be prepared to exact a reasonable fee even where a defendant’s conduct was not reprehensible. Thus, in English law, reasonable fee awards are widely used in cases where an injunction has been denied, regardless of the quality of the defendant’s conduct.144 Equally, a discussion of the provision in the proposed new Restatement tentatively suggests that, where an injunction is denied, it might be appropriate to grant relief based on what would be a reasonable fee for a licence to continue the conduct in question.145 137

Restatement of the Law, Third, Law of Restitution and Unjust Enrichment, above n 133, at § 44(1). ibid, introductory note to ch 5. 139 See Restatement (Third) of the Law of Restitution and Unjust Enrichment (Tentative Draft No 5) (American Law Institute, 2007) ch 7. 140 Restatement of the Law, Third, Law of Restitution and Unjust Enrichment, above n 133, at § 44(1). 141 ibid, at § 44(3). 142 ibid, at § 44(3)(a). 143 See above n 121 and accompanying text. 144 See eg Wrotham Park Estate Co Ltd v Parkside Homes Ltd [1974] 1 WLR 798 (Ch); Jaggard v Sawyers [1995] 1 WLR 269 (CA). 145 See Restatement of the Law, Third, Law of Restitution and Unjust Enrichment, above n 133, at § 44(d) illustration 14. 138

418

Craig Rotherham

Secondly, gain-based relief will be denied, ‘to the extent that disgorgement would result in an inappropriate windfall to the claimant, or would otherwise be inequitable in a particular case’.146 This is something of a catchall category relying on rather open-ended standards. It would appear, however, to provide scope for mediating concerns about proportionality. In particular, the reference to restitution being denied ‘to the extent’ it would be ‘inappropriate’ or ‘inequitable’ suggests that the courts might distinguish between consequential profits and the reasonable fee measure in applying the section. Just as it is more difficult to justify the availability of gain-based relief for the miscellaneous wrongs covered by § 44 (thereby warranting limiting such relief to conscious wrongdoers), so it is harder to justify higher measures of gain-based relief that would entitle the claimant to share in all or at least a portion of the profit made from the defendant’s activity. While the general rule is that conscious wrongdoers are liable for consequential profits,147 it is not clear that this should be readily applied to the miscellaneous rights covered by § 44. On the other hand, reasonable fee awards can probably be tailored to meet objections of this sort. Thirdly, restitution will be denied, ‘if the benefit derived from the defendant’s interference with the claimant’s interests cannot be adequately measured’.148 This limitation is understandable. Arguments in favour of gain-based relief are generally relatively finely balanced and, where the assessment of the defendant’s enrichment would be highly speculative, it would be arbitrary to make use of that measure. Indeed, one of the justifications for utilising gain-based relief is that often the claimant’s loss cannot be accurately measured.149 This justification loses its force when the defendant’s gain is equally difficult to gauge. Finally, it is said that gain-based relief will be denied, ‘if allowance of the claim would conflict with a remedial scheme established by other law, defining the claimant’s remedies or the defendant’s liability for the wrong’.150 This rationale might be employed where, for example, a statutory scheme sets a ceiling on claims for compensation and the resort to a restitutionary claim might be thought to be inconsistent with the legislative intention in question.151 § 44 is explicitly said to govern miscellaneous instances of misappropriation of property not covered in the previous sections, some instances of fraud, deceit, interference with contractual relationships, unfair competition, breaches of rights of privacy, and illegal conduct causing injury to the claimant. While the reporter’s principal comments do not specifically address the tort of nuisance, the provision of gain-based relief for that tort might be thought to be consistent 146

ibid, at § 44(3)(b). See above n 139. 148 ibid, at § 44(3)(c). 149 A-G v Blake [2001] 1 AC 268 (HL) at 288 (Lord Nicholls). 150 Restatement of the Law, Third, Law of Restitution and Unjust Enrichment, above n 133, at § 44(3)(d). 151 See eg The Franciscan No 1 v The Qualicum Producer (2001) 198 DLR (4th) 267 (British Columbia CA). 147

Justifying Gain-based Relief for Nuisance

419

both with the general statement of principle in § 41 and with the approach taken to miscellaneous rights in § 44. Moreover, an example of nuisance derived from the celebrated case of Boomer v Atlantic Cement152 features in a discussion of the qualifications to the availability of relief, illustrating the proposition that gainbased relief will be denied where the imposition of gain-based relief would be inequitable.153 The suggestion that restitution should be denied in these particular circumstances implies that it should otherwise be available. The approach taken by Andrew Kull in drafting the new Restatement of Restitution and Unjust Enrichment has much to recommend it. The balance struck in § 44 in relation to breaches of miscellaneous rights would appear well suited to the regulation of nuisance even if there is little discussion of that tort.

VII. Conclusion While Goff and Jones lament that ‘no English court has sought to weave any sophisticated gold thread to unite the cases on so-called waiver of tort’,154 perhaps the real problem lies in the suggestion that coherence can be brought to our law by an elegant filament of principle. The flaw in such an assumption is well illustrated by attempts to ground the determination of the availability of gain-based relief generally and for nuisance in particular on formal conceptions. The notion that we can distinguish between anti-enrichment and anti-harms wrongs has proved surprisingly seductive but hardly stands up to close analysis. The claim that gainbased relief should be automatically available for breaches of property rights, while easier to grasp, should also be resisted as being analytically misconceived and for failing to engage in the key issues of policy that arise in this context. There is no knock-out argument for or against gain-based relief for nuisance. However, the case for awarding such relief is probably weaker than is often supposed and the case becomes harder to make the higher the measure of relief that is sought. The proposed Restatement of Restitution and Unjust Enrichment takes a nuanced approach to the issue of gain-based relief. Its willingness to accommodate a diversity of treatment in determining the availability of such relief for different wrongs suggests a sensitivity to the difficult balance that needs to be struck in this context.

152

Boomer v Atlantic Cement, 257 NE 2d 870 (New York CA, 1970). See Restatement of the Law, Third, Law of Restitution and Unjust Enrichment, above n 133, at § 44(d) illustration 14. 154 Goff and Jones, above n 49, at para [36-006]. 153

17 Just and Unjust Enrichments HANOCH DAGAN

I. Introduction Notwithstanding (or perhaps due to) the impressive blossoming of the law of restitution and unjust enrichment, its most fundamental question—what distinguishes just from unjust enrichments?—is still being hotly debated. This essay is an attempt to contribute to this question. My entry into this inquiry is idiosyncratic. I start with two observations. The first is that, at least in modern law, the law of restitution does not address most of the enrichments we deem to be unjust. Tax and redistribution schemes actually deal with most of these unjust enrichments, when ‘unjust’ stands for a sequence of enrichments yielding a distribution we find unacceptable. The second observation is that, beyond this realm of correcting distributive injustices, most enrichments are not perceived as problematic. Indeed, although the title of this essay piggybacks on Michael Walzer’s celebrated Just and Unjust Wars, a stark distinction prevails between the way we treat wars as a priori unjust (so that good justifications are needed to render a war into a just war) and enrichments, which we justifiably perceive (once we set aside distributive concerns) as a priori just. This privileged status of enrichments is clear not only from the, admittedly awkward, comparison to wars. The significantly differential treatment of harms and benefits in private law brings home this very same point. I argue that unpacking these two rather obvious and uncontroversial observations can help to identify both the types of reasons that cannot and, even more significantly, the types of reasons that can justify restitution. In other words, it can help us identify the proper goals of the law of restitution.

II. A Universe of Enrichments Restitution lawyers are preoccupied with paradigmatic private law cases of mistaken payments, self-interested conferrals of benefits, wrongful enrichments, and the like. The academic attention focused on these cases is of course commendable, and largely responsible for our enhanced understanding of the law of restitution. But the inward-looking and in-depth research of these types of cases has also

424

Hanoch Dagan

had its costs. It may have obscured a wide range of enrichments that are either reversed outside the law of restitution or happily accepted as no cause for legal intervention. My first task in this essay is to situate the types of cases that prompt liability in restitution within this much wider universe of enrichments.

A. Taxable Enrichments Consider first the truism—albeit one hardly mentioned as such—that most of the legal regulation of enrichments occurs beyond the boundaries of the law of restitution, however these boundaries are defined. Both in terms of intensity and in terms of magnitude, tax law rather than restitution law is the primary legal mechanism for handling enrichments. To be sure, not only enrichments are taxed; tax law, at least in some jurisdictions, also targets consumption (as in Value Added Tax) as well as the simple holding of valuable assets (as in property tax). Still, the taxation of enrichments is quite a significant subset of contemporary tax law. Some of the enrichment events that prompt or are said to justify prompting taxation are windfalls; Henry George’s famous scheme of taxing land improvements is a prime example.1 But most of the enrichments subject to taxation are not the result of a transfer deemed to be in any way dubious. In many jurisdictions, bequests as well as gifts are taxed even though no one doubts the validity of these gratuitous transfers. Furthermore, the primary tax in modern times, the taxation of income, targets enrichments resulting from people’s labour. Tax law is complex, and the justifications of taxation are an important corpus within political philosophy. For the purposes of this essay, however, a few straightforward propositions will suffice. Some taxation of people’s enrichments can be justified by reference to the quid pro quo they receive by way of certain public goods, such as security;2 and some may be justified by the fact that their production of income relies on and utilizes existing societal goods for which it is just to make them pay.3 And yet, as libertarian authors constantly remind us, the reality of modern-day taxation in liberal democracies can hardly be justified solely in these terms.4 Currently, tax law is very much in the business of addressing distributive concerns. No wonder, then, that along with efficiency, distributive justice (or equity) is canonically presented as a prime goal of tax law and policy.5

1 H George, Progress and Poverty (London, Dent, 1911). See also, A Bell and G Parchomovsky, ‘Givings’ (2001) 111 Yale Law Journal 547. 2 See RA Epstein, Bargaining with the State (Princeton, New Jersey, Princeton University Press, 1993) 145. 3 See RS Avi-Yonah, ‘Why Tax The Rich? Efficiency, Equity, and Progressive Taxation’ (2002) 111 Yale Law Journal 1391, 1404. 4 See, eg, RA Epstein, Takings: Private Property and the Power of Eminent Domain (Cambridge, Massachusetts, Harvard University Press, 1985) 295–303. 5 See, eg, A Smith and E Camman (ed), An Inquiry into the Nature and Causes of the Wealth of Nations 4th edn (London, Methuen & Co., 1925) vol V, ch 2, 310–11. The text skirts over quite divergent views on the matter, ranging from notions of equal sacrifice to various ideas of redistribution. But these differences, notwithstanding their importance for tax theory, are unimportant for the modest purposes of this section.

Just and Unjust Enrichments

425

Let us consider the meaning of these simple propositions. The bulk of our law serving to regulate enrichments is justified not by reference to any deficiencies in the transfer of resources from one individual to another, but rather by the injustice of the resulting distribution of resources in society. This justification, as Robert Nozick forcefully demonstrated, is analytically incompatible with the absolutist conception of property. If we assume, with Nozick, that property rights are necessarily absolute, any distribution we deem just will quickly become unstable due to unobjectionable voluntary exchanges.6 This means that most of our legal regulation of enrichments is justified only after we are reminded that just distributions do not, in all likelihood, follow Blackstone’s infamous model of ‘sole and despotic dominion’,7 but include much more refined definitions of ownership.8 Realising that the public side of the law of enrichment (tax law) actually rejects Blackstone’s conception of ownership is significant for our purpose because this conception is frequently said to supply the private side of this law (the law of restitution) with the baseline purportedly separating just from unjust enrichments.9 To be sure, ruling out the absolutist understanding of ownership from public law does not necessarily dismiss the possibility of a division of labour between a private law that relies on a strictly libertarian conception of ownership and a public law that remedies the entailed deficiencies in terms of distributive justice.10 And yet, this lesson of the public law of enrichment is nonetheless valuable. It releases us from the essentialist claim concerning the allegedly absolutist meaning of ownership, which still prevails despite the devastating critique it has endured.11 It thus leaves private law libertarians, and particularly those who hold that the absolutist conception of property can neatly separate just from unjust enrichments, with a rather insurmountable burden. They must offer support for their claim that private law doctrines should be founded on this conception of property, by showing that public law can and does supplement private law with rules that adequately 6

See R Nozick, Anarchy, State and Utopia (New York, Basic Books, 1974) 160–64. W Blackstone, 2 Commentaries 2 (London, Butterworth & Son 1760). 8 See W Kymlicka, Contemporary Political Philosophy (Oxford, Oxford University Press, 1990) 102–3 (Rawlsians are likely to endorse a property regime that anticipates the need ‘to mitigate the effects of undeserved natural advantages after the initial distribution’). See also, H Dagan, ‘The Social Responsibility of Ownership’ (2007) 92 Cornell Law Review 1255. 9 See, eg, RB Grantham and CEF Rickett, Enrichment and Restitution in New Zealand (Oxford, Hart Publishing, 2000) 18–20, 45–6, 53 and 58–60; NJ McBride and P McGrath, ‘The Nature of Restitution’ (1995) 15 Oxford Journal of Legal Studies 33, 37–9 and 43. cf A Kull, ‘Rationalizing Restitution’ (1995) 83 California Law Review 1191, 1196–7, 1200, 1209 and fn 54; J Carter, ‘Restitution and Contract Risk’ in M McInnes (ed), Restitution: Developments in Unjust Enrichment (Sydney, Law Book Company, 1996) 137. 10 This legal architecture is forcefully suggested in EJ Weinrib, ‘Poverty and Property in Kant’s System of Rights’ (2003) 78 Notre Dame Law Review 795. In addition to the critique mentioned in the text below, I also find Weinrib’s account of why property rights at the stage of unilateral acquisition must be absolute and need not be tempered by something like the Lockean proviso quite perplexing. A thorough discussion of Weinrib’s account of property is beyond the scope of this essay. 11 See, eg, LS Underkuffler, The Idea of Property: Its Meaning and Power (New York, Oxford University Press, 2003). A competing account of property—which is, I argue, both normatively and descriptively more appealing—perceives property as an umbrella of a set of institutions (I call them ‘property institutions’) with some family resemblance to each other: see H Dagan, ‘The Craft of Property’ (2003) 91 California Law Review 1517, 1558–65. 7

426

Hanoch Dagan

remedy the injustices of the libertarian private law they advocate, both in terms of distribution and in terms of interpersonal dependence.12

B. Benevolent Enrichments My second point of departure is another truism, also usually unnoticed. Injurers who affect other people’s interests without their consent are commonly required to compensate their victims. While there are of course exceptions to this rule, the law of torts does recognise a general duty of compensation for harm. By contrast, people who benefit from other people’s interests are required to make restitution only in certain carefully delineated categories; in other words, there is no general duty of restitution.13 Why does the law of restitution take this position? Why does it perceive enrichments as a priori unproblematic,14 maybe even desirable?15 This question is most clearly urgent for those common law jurisdictions that adhere to the traditional method of ‘unjust factors’, which highlights the need for providing reasons for restitution.16 But it is also relevant within the civilian model, in which liability for restitution is triggered if the defendant’s grounds for retaining the disputed enrichment are denied. Even civilians do not find enrichments as such repugnant (contrast again with the a priori evil of harms), and only require unjustified enrichments to be reversed.17 This split between private law’s treatment of harms and its treatment of enrichments or benefits is in fact not that puzzling, at least not for those who appreciate the admittedly complex ways in which private law relies on our collective normative commitments.18 More specifically, good reasons—rooted in some of the most fundamental commitments of liberal societies to autonomy, utility and community—render harms a priori objectionable, while rendering many enrichments, even at the expense of another, innocuous or even laudable.19 (This 12 This seems to be an insurmountable burden both due to the realities of interest group politics in the promulgation of tax legislation and due to the unique role of private law in the construction of our conceptions of a person and of a community. See below, text accompanying n 52. 13 S Levmore, ‘Explaining Restitution’ (1985) 71 Virginia Law Review 65 (supporting this difference); and A Porat, ‘Private Production of Public Goods: Liability for Unrequested Benefits’ (2009) 108 Michigan Law Review (forthcoming) (challenging it). 14 See K Barker, ‘Responsibility for Gain: Unjust Factors or Absence of Legal Grounds? Starting Points in Unjust Enrichment Law’ in C Rickett and R Grantham (eds), Structure and Justification in Private Law: Essays for Peter Birks (Oxford, Hart Publishing, 2008) 56 (‘gains require no moral or legal explanation’). 15 See P Jaffey, The Nature and Scope of Restitution: Vitiated Transfers, Imputed Contracts and Disgorgement (Oxford, Hart Publishing, 2000) 17 (gains are ‘a matter of celebration, not regret’). 16 cf Barker, above n 14, at 57–63 and 73. 17 On the distinction between the two systems’ orientations, see, eg, D Klimchuk, ‘Restitution and Realism’ (2007) 20 Canadian Journal of Law & Jurisprudence 225, 230–32. 18 See generally, H Dagan, ‘The Limited Autonomy of Private Law’ (2008) 56 American Journal of Comparative Law 809. 19 This proposition seems to underlie Kit Barker’s recent attempt to reconceptualise every instance of responsibility for gain in terms of responsibility for harm: see K Barker, ‘The Nature of Responsibility

Just and Unjust Enrichments

427

proposition is true only after we set aside concerns of distributive justice largely addressed by the public side of enrichment law, namely: tax law, and not by the law of restitution.20) While harms derogate from people’s welfare and often undermine their autonomy, benefits add to their welfare and expand the various possible ways for them to write the story of their lives.21 Barring reasons to the contrary, then, as is the case in the various categories of the law of restitution, the sheer receipt of benefits is conducive rather than detrimental to these two fundamental liberal values. On its face, these observations as to the distinction between benefits and harms ignore the person at whose expense the recipient has been enriched. But this discrepancy is only troublesome when the recipient is responsible for some harm endured by that person, again indicating that what is problematic is the plaintiff ’s harm rather than the defendant’s benefit.22 Indeed, the a priori innocence of the receipt of benefits that concerns me here becomes clear precisely when no harm is inflicted, or when the defendant’s benefit exceeds the plaintiff ’s harm. In such cases, if the plaintiff ’s insistence on restitution is backed merely by her disapproval of the defendant’s enrichment, it ends up as a fanciful disguise for envy. Some, indeed, argue that in the larger context of the public law of enrichments, envy—in the sense of preferences regarding relative wealth or other primary goods—carries normative significance.23 But even if we take this view as a given, it does not follow that ‘sentiments of envy or spite’24 should inform private law and guide the resolution of disputes between a particular claimant and a particular defendant.25 What about the difficulty of free-riding on other people’s efforts and investments? Were free-riding a socially repugnant behaviour, its prohibition might indeed carry over to our question about the moral status of enrichments. But a blanket disapproval of free-riding is indefensible. As Wendy Gordon explains, ‘if deliberate uses of others’ efforts always triggered an obligation of payment, it would cause paralysis’. No culture could exist ‘if all free riding were prohibited within it’ because culture ‘is interdependence, and requiring each act of deliberate for Gain: Gain, Harm, and Keeping the Lid on Pandora’s Box’ in R Chambers, C Mitchell and J Penner (eds), Philosophical Foundations of Unjust Enrichment (Oxford, Oxford University Press, 2009) 146. 20 As the text implies, there is no inherent reason for excluding distributive justice from private law. See generally, D Lewinsohn-Zamir, ‘In Defense of Redistribution through Private Law’ (2006) 91 Minnesota Law Review 326. 21 See Barker, above n 14, at 56. 22 As the text implies, in these cases we should focus directly on the defendant’s responsibility for the claimant’s predicament, which is not always obvious. In many contexts, such responsibility would be inappropriate. Often, a profit we deem to be legitimate is, by definition, obtained at the expense of others, and that is indeed the very essence of competitive markets. 23 For the competing views on this question, see MJ Roe, ‘Backlash’ (1998) 98 Columbia Law Review 217, 232 and fn 28. 24 CT Wonnell, ‘Replacing the Unitary Principle of Unjust Enrichment’ (1996) 45 Emory Law Journal 153, 181. 25 Contra M Gergen, ‘What Renders Enrichment Unjust’ (2001) 79 Texas Law Review 1927 at 1927, 1929, 1945 and 1953–4 (discussing JP Dawson, Unjust Enrichment: A Comparative Analysis (Boston MA, Little Brown & Company, 1951)).

428

Hanoch Dagan

dependency to render an accounting would destroy the synergy on which cultural life rests’. Furthermore, no community could exist because community is defined as interdependence: ‘persons learn from each other, sell complementary products, build on a common heritage’.26 This value of interdependence, which is the defining feature of each human community, explains why law does not interfere in many cases where one person’s activity generates positive externalities others enjoy.27 Thus, three core humanistic values—autonomy, utility, and community— explain law’s a priori approval of enrichments. We may therefore plausibly expect that this presumption would be upset when at least one of these core values is actually undermined by allowing the recipient to retain her enrichment. My brief survey of the wide universe of enrichments that are not regulated by the law of restitution ends here with two main lessons that will guide me in the remainder of this essay, one negative and one positive. The first is the implausibility of relying on Blackstone’s absolutist conception of property as the baseline for deciding which enrichment is just and which is not. The second is the determinative role of the three core liberal values of autonomy, utility and community in this formidable task.

III. From Normative Goals to Private Law Rules The task of justifying restitutionary rules requires setting up baselines that determine the justice (or injustice) of the enrichments at stake. These baselines cannot rely on a plaintiff ’s reference to the benefit at stake as ‘mine’ or to the defendant’s enrichment as being at her expense. These colloquial statements beg the question that needs to be decided, namely, whether to allocate the entitlement to the specific benefit at hand to one party or the other.28 There are no shortcuts in 26 WJ Gordon, ‘Of Harms and Benefits: Torts, Restitution, and Intellectual Property’ (1992) 21 Journal of Legal Studies 449, 468; WJ Gordon, ‘On Owning Information: Intellectual Property and the Restitutionary Impulse’ (1992) 78 Virginia Law Review 149, 167–8. 27 Moreover, even in terms of the more individualistic desert for labour theory of property, the entitlement claimed by labourers vis-à-vis free riders is often exaggerated. Labourers, under this theory, merit a reward because by engaging in value-creating activities they contribute to the betterment of the human predicament. But this means that their prerogatives, even from the very perspective of desert theory, must be subject to three and rather demanding limitations: ‘Laborers deserve rewards if and only if they engage in the right kind of activity (useful and purposeful, rather than destructive, inconsequential, or simply inadvertent). They can make a claim to be rewarded if and only if they have actually added value. And, finally, the deserved reward must be proportional to the added value they indeed generated’ (H Dagan, ‘Property and the Public Domain’ (2006) 17 Yale Journal of Law and Humanities 84 (Supp), 90). 28 Recall that, on the one hand, we have no reason to presuppose that any gains derived from resources properly described as the plaintiff ’s property are necessarily within her entitlement and that, on the other, the entitlement to profit is at times an element of rights that we do not usually classify as proprietary: see H Dagan, The Law and Ethics of Restitution (Cambridge, Cambridge University Press, 2004) 21.

Just and Unjust Enrichments

429

answering this question, which requires us to employ the guidelines of autonomy, utility, and community in the various contexts traditionally covered by the law of restitution. At its core, then, restitution theory—the theory that determines which enrichments are just and which unjust, and furthermore prescribes the appropriate remedial response for unjust enrichment—is distributive. Unlike the public side of the law of enrichments, this distribution is not between the better off and the worse off in society. And yet, the theory underlying the law of restitution still clearly constitutes a society-wide distribution of burdens and benefits regarding a diverse set of resources and of activities.29 Identifying the types of normative considerations that restitution theory invokes along these lines is likely to raise two important concerns. Some may worry about the predictability of the resulting doctrine. Others may inquire whether my analysis does not collapse restitution law into a garden variety form of regulation, thus obscuring its characteristic features as an important part of private law usually analysed in terms of corrective rather than distributive justice. Addressing these concerns is important both for their own sake and to refine the ways in which autonomy, utility and community should legitimately inform the law of restitution.

A. Vague Values and Predictable Rules My initial concern, predictability, takes two forms. One is that allowing judges to consult with the values of autonomy, utility and community is likely to threaten the important benefits of ‘the rule of rules’ in the authoritative settling of disputes that secures the moral benefits of co-ordination, expertise, and efficiency.30 Rules, as Emily Sherwin explains, ‘are designed to translate the implications of normative values into concrete prescriptions for action or decision’. To do so effectively, rules ‘must be general, in the sense that they prescribe outcomes for classes of cases’, and they must be ‘sufficiently determinate that rule-appliers can understand what the rules prescribe without first resolving the very normative questions the rules are designed to settle’. In other words: for rules to function as rules they must be followed by those applying them ‘in all cases, without considering whether the local outcome of the rule conforms to the values the rule is supposed to advance’.31

29 For a more detailed account of the distributive dimension of one restitutionary doctrine, governing cases of wrongful enrichments, see H Dagan, Unjust Enrichment: A Study of Private Law and Public Values (Cambridge, Cambridge University Press, 1997) 22–3. See also P Cane, Responsibility in Law and Morality (Oxford, Hart Publishing, 2002) 186–90 and 282–3. cf K Barnett, Corrective Justice and Gains-Based Remedies (paper presented at ‘The Goals of Private Law’ Obligations IV Conference, National University of Singapore, 23–25 July 2008). 30 See L Alexander and E Sherwin, The Rule of Rules: Morality, Rules and the Dilemma of Law (Durham NC, Duke University Press, 2001) 12–15. See also F Schauer, ‘The Generality of Law’ (2004) 107 West Virginia Law Review 217, 224–34. 31 E Sherwin, ‘Rule-Oriented Realism’ (2005) 103 Michigan Law Review 1578, 1589–91.

430

Hanoch Dagan

Restitution theory should thus distance itself from the dubious nominalistic approach of case by case adjudication. Identifying autonomy, utility, and community as the underlying values of the law of restitution should not lead us to the unfortunate strategy of inviting judges to routinely make ad-hoc judgments based on these values.32 By the same token, it should not imply the problematic approach of rule-sensitive particularism, allowing judges to depart from rules whenever the outcome of the particular case at hand so requires, while taking into account both substantive values and the value of preserving the rule’s integrity.33 Instead, it should advise some legal actors—notably, judges of appellate courts— to occasionally use new cases as triggers for an ongoing refinement of rules, as opportunities to revisit the normative viability of existing rules qua rules, and to re-examine the adequacy of the legal categorisation that organises these rules.34 But even if we limit the role of autonomy, utility and community to a circumscribed number of cases, the worry about unacceptable legal indeterminacy—the second version of the predictability concern—may still linger. Would not allowing restitutionary doctrines to rest on these three values without a predetermined formula for measuring and balancing them entail unbridled judicial discretion?35 I do not think so, even when disregarding the fact that classical legal formalism, the main alternative type of legal reasoning, is hopelessly malleable and thus indeterminate.36 To see why, we must first realise that a jurisprudence based on value pluralism, such as the one I propose here for the law of restitution, is very different from one that endorses the (philosophically suspect) positions of relativism, scepticism, or nihilism. These positions undermine any possibility of

32 Similarly, pointing out the distributive dimension of the law of restitution should not be understood as a prescription to evaluate individual cases distributionally, but rather as an observation on the doctrine’s nature. See AT Kronman, ‘Contract Law and Distributive Justice’ (1980) 89 Yale Law Journal 472, 501. 33 See Sherwin, above n 31, at 1591–4. See also, A Goymour, ‘Book Review’ (2006) 65 Cambridge Law Journal 456, 458. 34 Sherwin correctly interprets my position as aimed at limiting ‘the role of autonomy, utility, and community to judgments about the overall effects of rules’. She finds this position ‘logically impeccable’, but argues that ‘the fine distinction it calls for, between particular errors and net overall errors, may be difficult for human beings to achieve’. The reason for this is that ‘once a judge has consulted the rule’s rationale to answer the problem of overall justification, it will be exceedingly difficult for the judge to screen out the rule’s rationale at the point of application … therefore, the temptation will be strong to overrule rules that are beneficial overall’: see E Sherwin, ‘The Rules of Obligations’ ch 18 of this book. Were the authority to overrule been ascribed to trial court judges, this might have indeed been an overwhelming concern, because these judges decide ‘live disputes’ and have unmediated contact with ‘the effect of the rule on parties who have narrated their stories before the court’: Sherwin, ‘The Rules of Obligations’, at n 49. Fortunately, only appellate courts are authorised to overrule. Hopefully, while these courts are still sufficiently attentive to the real life context of the dispute, they are not affected too much by sympathies to the particular parties. They can follow, in other words, the maxim that hard cases make bad law. 35 See J Edelman, ‘Book Review’ (2006) 69 Modern Law Review 131, 134. See also, P Saprai, ‘Book Review’ (2005) 121 Law Quarterly Review 521, 523. 36 The main and inescapable reason for this doctrinal indeterminacy is the availability of multiple and potentially applicable doctrinal sources, rather than the indeterminacy of discrete rules: see H Dagan, ‘The Realist Conception of Law’ (2007) 57 University of Toronto Law Journal 607, 613–7.

Just and Unjust Enrichments

431

moral justification, evaluation, or for that matter criticism, and thus the idea of law itself.37 By contrast, value-pluralist jurisprudence recognises a broad menu of incommensurable human alternatives, but acknowledges a minimal core of moral truths:38 in our case, the moral significance of autonomy, utility and community. The heterogeneity prescribed by value pluralism—meaning the idea that different values or different balances of the same values should guide different areas of law (and life)—should not be equated with an invitation to apply subjective preferences. Context matters. Our lives are divided into economically and socially differentiated segments.39 Each ‘transaction of life’ has some features of sufficient normative importance, features that gain significance from the perspective of our general normative commitments and justify distinct legal treatment.40 Indeed, good lawyers cautious to look at each one of the (rather narrow) categories of restitution, benefit from an ‘alert sense of actuality [that] checks our reveries in theory’, as well as from ‘the illumination which only immediacy affords and the judiciousness which reality alone can induce’.41 They thus realise that different types of human interactions, and therefore different categories of restitutionary doctrines, call for different balances of these broad normative commitments. Indeed, here, as elsewhere, the requirement to explicitly apply judgement, which needs to be normatively and contextually justified, is a real constraint.42 In some categories of cases this contextual normative inquiry might lead to a standoff, with reason unable to adjudicate between two (or more) competing accounts. But the relevant question is not whether such cases are possible, and the sheer existence of hard cases scarcely undermines the determinacy of the law. Rather, the question is whether these cases are prevalent enough so that they threaten a restitution theory premised on these guidelines. In many cases, a sufficiently robust contextual normative account can have quite sharp doctrinal 37

See SJ Burton, Judging in Good Faith (Cambridge, Cambridge University Press, 1992) 19–21. See, eg, J Kekes, The Morality of Pluralism (Princeton NJ, Princeton University Press, 1993) 31–4 and 118–38; M Walzer, Thick and Thin: Moral Argument at Home and Abroad (Notre Dame, Indiana, University of Notre Dame Press, 1994) 1–19; I Berlin, ‘The Pursuit of the Ideal’ in H Hardy (ed), The Crooked Timber of Humanity: Chapters in the History of Ideas (Princeton NJ, University of Princeton Press, 1991) 11; WA Galston, ‘The Legal and Political Implications of Value Pluralism’ (1998) 57 Maryland Law Review 236, 241. 39 See P Berger et al, The Homeless Mind: Modernization and Consciousness (Harmondsworth, Middlesex, Penguin, 1973) 63–82. 40 See, eg, M Walzer, Spheres of Justice: A Defense of Pluralism and Equality (Oxford, Martin Robertson, 1983); E Anderson, Value in Ethics and Economics (Cambridge MA, Harvard University Press, 1993); E Anderson, ‘Pragmatism, Science, and Moral Inquiry’ in RW Fox and RB Westbrook (eds), In the Face of Facts (Washington DC, Woodrow Wilson Center Press, 1998) 17. 41 H Oliphant, ‘A Return to Stare Decisis’ (1928) 14 American Bar Association Journal 71, 73–4 and 159. See also, WW Fisher III, ‘The Development of American Legal Theory and the Judicial Interpretation of the Bill of Rights’ in MJ Lacey and K Haakonssen (eds), A Culture of Rights (Cambridge, Cambridge University Press, 1992) 272–3 and 275. Joseph Raz’s analysis of the distinguishing cases phenomenon brings home a similar point: see J Raz, The Authority of Law: Essays on Law & Morality (Oxford, Clarendon Press, 1979) 183–97. 42 See JW Singer, ‘Property Norms: Reflections on the Externalities of Ownership’ in GS Alexander and E Penalver (eds), Property and Community (New York, Oxford University Press, forthcoming) §§ 2.2.2–2.2.3. 38

432

Hanoch Dagan

teeth, and I attempt to demonstrate this in the last section of this essay. Although some of my accounts may obviously be controversial, they can hardly be challenged by the sheer difficulty of measuring or balancing autonomy, utility and community, or by the possibility that there are other pertinent values.43 To challenge my approach to restitution, a detailed demonstration of the superiority of a competing account is needed, rather than a blanket claim that a better account will always be available.44

B. Private Law and Public Values Restitution lawyers may nonetheless be reluctant to endorse an approach relying on a foundational distributive scheme informed by the trio of autonomy, utility and community. Their position may be explicitly or implicitly informed by the orthodox association between restitution law, as part of private law, and corrective justice. More precisely, with Ernest Weinrib, they may worry that my approach fails to provide adequate justification for restitutionary liabilities because it may not explain ‘why the plaintiff ’s unilateral transfer of the benefit creates an obligation in the defendant’.45 Reasons such as autonomy, utility or community often refer to ‘the normative position of each of the parties separately rather than [to] the relationship between them’.46 Herein lies a fatal failure, says Weinrib: ‘Basing liability on such considerations exemplifies the fallacy that what is unilaterally applicable to the plaintiff can have bilateral significance for the relationship of plaintiff and defendant’.47 Indeed, in Weinrib’s view, no social purpose or social value, even if ostensibly desirable otherwise, can legitimately inform private law. Private law is an autonomous realm with its own inner intelligibility, isolated from the social, economic, cultural, and political realms. For Weinrib, this isolation derives from the bilateral logic of private law adjudication, understood as a unique forum for the vindication of infringed rights. As such, private law must comply with the injunction of correlativity, requiring both that the reasons underlying the plaintiff ’s right be the same as the reasons that justify the defendant’s duty, and that these very reasons also explain the specific remedy inflicted on the defendant. The commands of correlativity, in this view, are so robust that they leave no space for any other (social) value.48

43 cf J Gordley, ‘The State’s Private Law and Legal Academia’ (2008) 56 American Journal of Comparative Law 639. 44 cf R Dworkin, Law’s Empire (Cambridge MA, Harvard University Press, 1986) 76–86 (responding to the challenge of ‘external critique’). 45 EJ Weinrib, ‘Restoring Restitution’ (2005) 91 Virginia Law Review 861, 868. 46 ibid, at 869. 47 ibid, at 870. 48 See, eg, EJ Weinrib, The Idea of Private Law (Cambridge MA, Harvard University Press, 1995) 3–14, 212 and 214; EJ Weinrib, ‘Restitutionary Damages as Corrective Justice’ (1999) 1 Theoretical Inquiries in Law 1, 3–5 and 37. See also, D Klimchuk, above n 17.

Just and Unjust Enrichments

433

I agree that private law is not just a means for normative regulation, meaning that intrinsic features of private law constrain the types of rules it can legitimately promulgate. Brute instrumentalism, which perceives civil suits as one ‘mechanism whereby the state authorizes private parties to enforce the law’,49 is misguided because, like any other aspect of state coercion implied by law, the bipolar structure of private law litigation also entails some justificatory burden.50 Weinrib is thus correct in insisting that private law should be able to justify to defendants both the identity of the beneficiary of any liability imposed on them, and the exact type and degree (or magnitude) of that liability. Insofar as the correlativity requirement insists that the defendant’s liability and remedy correspond to the plaintiff ’s entitlement, it is indeed essential to the integrity of private law and thus to any credible restitution theory. But the correlativity between plaintiff ’s entitlement and defendant’s liability should not obscure the rich social fabric that serves as the inevitable context for the parties’ relationship. Pace Weinrib, correlativity need not mean that private law has, or should have, an inner intelligibility decipherable without recourse to social values.51 Quite the contrary: the pivotal role of private law in defining our mutual legitimate claims and expectations in our daily interactions52 undermines the legitimacy of a private law regime that ignores these values.53 For this reason, the parties’ ex ante entitlements, from which this correlativity must be measured, are best analysed by reference to our social values.54 Refusing to isolate private law from social values does not render correlativity vacuous; it does not, in other words, collapse private law into just another form of regulation. After all, as Weinrib correctly insists, taking correlativity seriously 49 EL Rubin, ‘Punitive Damages: Reconceptualizing the Runcible Remedies of Common Law’ [1998] Wisconsin Law Review 131, 154. The distinction between private law adjudication and public law regulation is particularly blurred in some law and economics accounts of the common law: see, eg, RA Posner, Economic Analysis of Law, 6th edn (New York, Aspen Publishers, 2003) 383–5. 50 On the dialectical relation between law’s coercion and its nature as a justificatory practice, see Dagan, above n 36, at 622–37. 51 Recall that by ‘social values’ I do not mean only collective values, although I do include these values (utility, community) as well. Quite the contrary: one of the most important values served by the law of restitution in my account is autonomy. 52 See H Dagan, ‘Just Compensation, Incentives, and Social Meanings’ (2000) 99 Michigan Law Review 134, 148–52. 53 On its face, there is an important institutional counterargument to my claim, at least insofar as we are dealing with a private law regime which is developed by the judiciary, as opposed to the legislature. But upon closer examination an institutional analysis supports, rather than undermines, my claim. Such an analysis can take two forms: it can be based on concerns of either accountability or expertise. But excepting in the most formal sense of the word, legislators do not, in fact, enjoy any advantage over judges in terms of accountability regarding private law issues; and, as I explain above (text accompanying nn 41–42), in fact fare better than legislators in terms of expertise in these matters. See generally, Dagan, above n 36. Finally, as the text implies, there is no escape from judicial responsibility for enacting social values because the precedential effect of private law cases means that judges are inevitably in the business of shaping our social life. 54 Cf J Gordley, Foundations of Private Law: Property, Tort, Contract, Unjust Enrichment (Oxford, Oxford University Press, 2006) 11; P Cane, ‘Distributive Justice and Tort Law’ [2001] New Zealand Law Review 401.

434

Hanoch Dagan

precludes assertions that the plaintiff ’s right is simply the ‘analytic reflex’ of the defendant’s duty (or vice versa).55 Yet, understanding that correlativity is situated within a thick layer of social values does require a more modest formulation of the correlativity injunction. In this formulation, correlativity prescribes that it is not enough for a plaintiff to demonstrate the desirability of the state of affairs that would result if the type of complaint she raises were to generate the remedy sought. Rather, a private law plaintiff has an additional justificatory burden: to give reasons why people in her predicament should be entitled to extract from people in the defendant’s category the kind of remedy she now requires. One can easily overcome this additional hurdle in some cases, as in the paradigmatic case of an injured plaintiff seeking remedy from a defendant who negligently caused her harm. But some cases are more challenging as, for instance, when the defendant can plausibly ask ‘why me?’ (why should she be forced to be the agent of remedying the plaintiff ’s unjustified harsh predicament), or ‘why you?’ (why should the plaintiff be allowed to invoke the state’s machinery to remove an unjust privilege that the defendant currently holds). And even if the plaintiff has good answers to both questions, she still needs to justify her entitlement to the specific measure of recovery she seeks to impose on the defendant. Weinrib goes even further by requiring that the reasons for the parties’ entitlements (and not only the entitlements as such) should be correlative, namely, that the reasons themselves be wholly internal to the parties’ relationship. ‘Because the justifications that ground liability are constitutive of the normative relationship between the parties’, he says, ‘those justifications must themselves have a relational structure’.56 But this additional requirement of relational reasons is excessively demanding and unwarranted. The bipolar form of private law is only one object, albeit important, of the justificatory burden prompted by the application of state coercion invoked by private law adjudication. Therefore, it should not be entitled to exclusivity in determining the types of normative considerations we must take into account; it should not, in other words, be allowed to overwhelm our justificatory inquiry. Cases in which the reasons for the parties’ entitlements are correlative may well be easier insofar as the integrity of private law is concerned. But this does not and should not imply that they are the only cases, or that we should a priori assume that even if the reasons are not correlative, their implications with respect to the parties’ entitlements will not be sufficiently convergent. Quite the contrary, the sheer fact that the reasons supporting the plaintiff ’s entitlement are not the same as those supporting the defendant’s liability should not necessarily prevent the complaint from proceeding if the convergence between these two sides of the coin is sufficiently wide, and the justificatory burden of showing the desirability of allowing it in its private law form is, on balance, properly met. Not only should private law allow cases of substantial convergence between prescriptions entailed 55 56

Weinrib, above n 48, at 124. Weinrib, above n 45, at 874.

Just and Unjust Enrichments

435

by reasons pertaining to the plaintiff and the defendant as a matter of course. There may even be categories of cases wherein the ‘justificatory deficit’ of considering reasons relating to the interests of third parties is actually outweighed by the significant normative superiority of allowing such cases in private law.57 Each type of case thus requires a careful account of the reasons for and against recognising the plaintiff ’s entitlement vis-à-vis the defendant, reasons of the kind that Weinrib indeed tries to exclude from private law. These reasons relate to whether law’s endorsement of such claims supports or distorts the ideal construction of the type of relationship under consideration, and whether using private law in this way is necessary, and overall conducive, to the public purpose at hand.58

IV. Examples Thus far, I have presented several abstract arguments on the substance and form of restitution theory, which separates just from unjust enrichments. I argued that the concept of property cannot serve as the baseline for this cardinal question, which should instead refer to the foundational liberal values of autonomy, utility and community. I further claimed that this orientation need not generate legal indeterminacy or strip the law of restitution from its constitutive characteristics as part of our private law, but can yield a happy doctrine composed of sharp rules instead of vague standards, and responsive to the refined prescription of correlativity as identified above. The time has come to see these broad propositions in action, examining their validity and power in the specific contexts governed by the law of restitution. This is an important component of my task in this essay because, like any theory of the law, part of the assessment of my theory must lie in the appeal of its specific results as well as in its ability to perform across a range of questions, to help in solving problems, and to confront possible criticisms.59 A detailed account of even 57 cf SA Smith, ‘The Rights of Private Law’ ch 5 of this book (offering reasons for ‘external pluralism’). 58 Note again that this does not collapse my view into brute instrumentalism because, unlike most instrumentalists, I take seriously the requirement that the plaintiff give reasons for why someone in her predicament should be entitled to extract from someone like the defendant the kind of remedy she now requires. Indeed, the practical implications of this additional justificatory burden are limited, but by no means trivial. In some cases, allowing a private law claim of the sort required would be unjust, even when likely to bring about a desirable state of affairs. This is why, for instance, I agree with Weinrib that, in most cases, the law of restitution should not allow recovery in excess to both the plaintiff ’s harm and the defendant’s net profits: see H Dagan, ‘Restitution’s Realism’ in Chambers, Mitchell and Penner (eds), Philosophical Foundations, above n 19, at 67–8. 59 In other words, as a lawyer and a pragmatist, I am committed to pay attention to both the particular contexts of specific doctrinal questions and to the motivating principles and policies underlying their resolution. cf D Herzog, Without Foundations: Justification in Political Theory (Ithaca NY, Cornell University Press, 1985) 223, 232–3. The top-down structure of this essay is purely expositional, and any valuable legal analysis needs to include both the ‘top’ and the ‘down’ sides.

436

Hanoch Dagan

one paradigm category in the law of restitution exceeds the scope of this essay. Fortunately, it is also unnecessary. For our purposes, brief sketches of two case studies—which I discuss at some length elsewhere60—can (I hope) do the job.

A. Mistakes The mistakes that typically concern the law of restitution, touching on the unilateral conferral of benefits that bring into play the recipient party’s detrimental reliance, or at least frustrated expectations, invoke the values of autonomy and of utility.61 Autonomy is relevant from the perspective of the mistaken party because, as involuntary transfers, mistakes ‘invite the law’s corrective measures to reinstate the commands of a mistaken party’s will, thus expanding her freedom of action and securing her integrity of self ’. In such cases, however, the recipient’s autonomy is at stake also because, ‘[i]f recipients were required to hold themselves always ready to give back any benefits they received, the security and stability of their affairs would be severely threatened’. Therefore, an autonomy-based law of mistakes must assign entitlements and liabilities based on a careful attempt to ‘reconcile our liberty to make mistakes with our aspiration for security and stability’.62 Utility is also relevant to the way law should (ex ante!) assign the entitlements to parties who find themselves in these types of cases. The law of mistakes is likely to affect (at least at the margins) the efforts of both transferors and transferees of benefits in trying to avoid mistakes. Given that both these efforts and the administration of the pertinent legal rule are costly (and borne by the parties), the humanistic injunction of utility calls for a careful application of a cost-benefit analysis of the possible legal rules.63 Putting these two perspectives to work in the rather prevalent context of mistaken payments can help to demonstrate the doctrinal teeth of these seemingly abstract propositions. It can also vindicate my claim as to the potential determinacy of restitutionary rules informed by my suggested normative infrastructure, and support my insistence as to the type of correlativity vital to the integrity of private law. Both autonomy and utility point to two considerations that must be taken into account when assigning the risk of mistaken payments between payors and

60

See Dagan, above n 28, chs 3 and 5. One may argue that community is also relevant. Minimal notions of social responsibility require that owners (or owners-to-be), who can avoid legal accidents such as mistakes at relatively low cost, do so in order to prevent the severe harm another person is likely to incur if the accident does take place. In other words, the prescription embedded in the comparative avoidance ability test, discussed below, may be restated in terms of our moral obligations to other members of our community. cf H Dagan, ‘Codification, Coherence, and Priority Conflicts’ in K Siehr and R Zimmermann (eds), The Draft Civil Code for Israel in Comparative Perspective (Tuebingen, Mohr Siebeck, 2008) 149, 163–4. 62 Dagan, above n 28, at 38–9 and 45–6. 63 ibid, at 52–63. 61

Just and Unjust Enrichments

437

recipients. More specifically, these values require that we compare the parties’ mistake-avoidance capacities, as well as their ability to insure or self-insure against this unfortunate contingency.64 A detailed analysis along these lines shows that the best rule often leaves both sides to the legal drama with some risk of harm. Particularly in the traditional cases of what I will refer to as ‘private contexts’, where the parties are more or less equal in terms of their avoidance capabilities and their ability to spread the risks of mistakes, the law of mistakes should allocate the recipient’s detrimental reliance according to a fault-based test. Two points bear emphasis here. First, in these types of cases, both values at hand are likely to caution us against imposing ‘superfluous burdens on potentially mistaken parties by unnecessarily limiting their liberty and by inefficiently inducing redundant precautions’. For this reason, ‘both analyses recommend looking carefully at whether a mistaken payment actually generates an irrevocable harm before any liability is assigned to the mistaken party—that is, before limiting her entitlement to restitution’.65 Secondly, a careful analysis of the injunctions of both autonomy and utility in our context recommends that we follow in the footsteps of tort law, which basically faces a similar challenge: rather than using the outdated test of relative fault adopted by the first Restatement of Restitution,66 we should look at the parties’ comparative negligence. Hence, although my analysis largely confirms the desirability of the existing rules as regards to this type of cases, it also helps to refine the current state of (American) law. Other types of cases call for very different conclusions. There are (increasingly important) categories of cases that fall under the rubric of what I call ‘institutional contexts’—mistaken payments by banks, brokers, or insurance companies, or improper tax payments. In these cases, the significant disparity in the parties’ avoidance capacities and abilities to spread the costs of mistakes among the beneficiaries of the activity that generated those mistakes (eg, the banking system), justifies a unilateral allocation of that risk. Insofar as this analysis applies to improper tax payments, and at least insofar as American law is concerned, it implies quite a radical reform: not only are there no good reasons for the prevailing restrictive approach to the restitution claims of mistaken taxpayers compared to other cases of mistaken payers, but this category’s unique characteristics indeed require the adoption of a rule more liberal than the one governing other cases of mistaken payments, which grants unlimited restitution for improper tax payments. The case of other institutional contexts (mistaken payments by banks, brokers, or insurance companies) is more nuanced. On its face, my suggestion is quite radical there as well. But a careful reading of some American cases shows that current law already reflects, albeit implicitly and imperfectly, the need to 64

This paragraph, and the two that follow, summarise Dagan, ibid, at 49–65 and 77–80. ibid, at 65. 66 See Restatement of the Law of Restitution, Quasi Contracts and Constructive Trusts (St Paul MN, American Law Institute Publishers, 1937) §§ 69(2) and 142(2). 65

438

Hanoch Dagan

apply a different legal regime to mistaken payments in these institutional contexts.67 A new reading of the existing defence of good faith creditor jurisprudence shows that it oftentimes serves (in an admittedly tormented fashion) as the doctrinal home for a rule denying restitution of mistaken payments in institutional contexts where, by and large, the mistaken party is both the cheapest cost-avoider and superior risk bearer. Be it as it may, this brief summary of my analysis of mistaken payments shows that, although the values of autonomy and utility may at first seem too abstract to be of any assistance to the particular doctrinal questions at hand, they are actually quite informative once situated in the robust context that the law of restitution typically faces. Indeed, their analysis can yield quite clear and determinate rules. Furthermore, as noted, we should not a priori assume that reasons supporting a plaintiff ’s entitlement will not sufficiently converge with other reasons supporting the defendant’s liability. The category of mistaken payments provides a good example of such convergence. Because the rules governing such payments affect every payer and every recipient, the ex ante entitlements for these cases must be set up so that they properly address the autonomy and welfare concerns of both.68 These considerations, as noted, by and large coincide.69 Thus, for instance, in private contexts, the plaintiff ’s autonomy justifications for restitution are strongest when the defendant’s autonomy justifications for retention are weakest—that is, when there is no detrimental reliance—and vice versa. Therefore, although the parties’ autonomy concerns are indeed different, a private law doctrine can promote both simultaneously by imposing a rather trivial burden (to make restitution) on a defendant who has not relied, and limiting the restitution of a payer who negligently caused detrimental reliance. A legal system adopting such a doctrine should have no real difficulty explaining to the recipient why she, of all people, should be responsible (in the cases and insofar as she is) for remedying the payer’s harsh predicament,70 and why the payer, of all people, is the one she should reimburse for the payment she mistakenly received.

67 And, in fact, some Commonwealth cases as well. See HW Tang, ‘The Role of Negligence and NonFinancial Detriment in the Law of Unjust Enrichment’ [2006] Restitution Law Review 55. 68 As Dennis Klimchuk puts it, ‘liability vindicates … not (or at least … not only) the particular interest this particular plaintiff has in her autonomy. Instead (or as well) it vindicates the value of autonomy more generally, a value in which the defendant can be understood to have an interest, no less than the plaintiff ’: D Klimchuk, ‘The Normative Foundations of Unjust Enrichment’ in Chambers, Mitchell and Penner (eds), Philosophical Foundations, above n 19, at 97. Klimchuk suggests that this raises the worry that the defendant’s duty to make restitution is inappropriately contingent because it ‘rests on the adoption by the legal order of particular values, interpreted in particular ways’: (at 98). To this worry he correctly replies by saying that ‘we would not protect the abstract entitlements the corrective justice account identifies in the plaintiff unless they served a value in which we have an interest’: (at 99). There is another response: as I explained above, the parties’ ex ante entitlements are not even intelligible without resorting to such underlying normative commitments. 69 Although there is some divergence. See Dagan, above n 28, at 64–5. 70 Her unique role as the proper defendant derives from the fact that, in this type of cases, her autonomy would be least jeopardised by carrying the (quite nominal) burden of remedying the plaintiff ’s justified grievance.

Just and Unjust Enrichments

439

B. Self-Regarding Conferral of Benefits I turn now to my second example: the law governing the self-regarding conferral of benefits. Like the law of mistakes, the analysis of this branch of the law of restitution will help demonstrate the viability of my proposed approach and will show that it generates predictable rules complying with the private law injunction of correlativity. Unlike the law of mistakes, however, it will also help us realise the limits of this last injunction and lead us to discuss a category of cases where taking third-party effects into account is actually justifiable. Or, at least, this is what I will argue. Conventional analysis suggests that restitutionary liability in this category arises when the defendant has been unjustly enriched, unless the plaintiff ’s conferral of the unrequested benefits is deemed officious.71 The problem, of course, is to set the boundaries between these conclusionary epithets; to offer a principled distinction between unrequested benefits that justify granting restitution, when we would consider it unjust for the defendant to retain the benefit, and other unrequested benefits that would not be thus privileged, when we would label the benefactor a volunteer (and the payment officious) and refuse to grant restitution. Autonomy and utility can again be recruited for the task.72 The key point here is that restitution can help overcome free-riding problems that may hinder jointly beneficial actions. Restitution, more specifically, is necessary in categories of cases where the parties’ interests are sufficiently locked in, but their cooperation in jointly serving these interests is unlikely. Typical situations of this category would be, for instance, cases wherein the parties are co-owners, or are both subject to a common liability, or cases of class actions, or of the maritime doctrine of general average. Although the promotion of the parties’ self-interest requires co-operation in these cases, their jointly beneficial collective action might be frustrated absent legal intervention because the individual interest of each of them might supersede their collective good.73 Under such circumstances, restitution becomes an attractive tool for forcing the parties to pay their proportionate share of the collective good, thus overcoming these free-rider difficulties. Solving free-rider problems in the provision of collective goods serves not only the parties’ utility interest. If carefully fine-tuned, it is also an important injunction of our commitment to autonomy. Autonomy stands for self-determination,

71 See, eg, P Birks, An Introduction to the Law of Restitution, rev edn (Oxford, Oxford University Press, 1989) 95–6; A Burrows, The Law of Restitution, 2nd edn (Oxford, Oxford University Press, 2002) 113; DB Dobbs, Law of Remedies, 2nd edn (St Paul MN, West Publishing, 1993) vol 1, § 4.3(4) at 604–5; C Mitchell, The Law of Subrogation (Oxford, Clarendon Press, 1994) 4 and 9; GE Palmer, The Law of Restitution (New York, Aspen Law & Business, 1978 & supp 1998) § 1.5(b) at 23–4. 72 This paragraph, and the two that follow, summarise Dagan, above n 28, at 130–36 and 139–40. 73 This contingency is the expected outcome if no single member of the group is likely to derive sufficient benefits from that good to justify paying the entire cost of supplying it alone, and no coalition of members can feasibly divide the costs among members.

440

Hanoch Dagan

and negative liberty (or ‘personal gatekeeping’) is but a means, albeit frequently quite an important one, for this ultimate value.74 Therefore, where law’s abstention (rather than its intervention) is likely to frustrate goals and aspirations that require collective action, overriding the explicit disinterest of restitution defendants in participating and paying for collective action may indeed be justified from an autonomy perspective as well.75 If the parties’ autonomy is to sanction law’s intervention, however, identifying categories of cases in which law’s intervention is necessary for the solution of collective action problems is not enough. It is not enough that the defendants’ proportionate benefit from law’s necessary intervention objectively exceeds their proportionate share in the costs of supplying the collective benefit. If autonomy is the name of the game, restitutionary obligation should only apply where potential restitution defendants are also subjectively better off receiving the collective benefits and paying than doing without the benefits entirely. This second condition points to the familiar doctrine of subjective devaluation, which essentially ensures that the divergence of the defendant’s explicit preferences (objection) from what we assume to be her self-interest (participation) derives from the pay-off structure locking her and the other participants in a collective action and making defection the dominant strategy, rather than from her genuine subjective preferences. Appreciating this justification for restitution underscores the need for regulating with precise rules, rather than vague standards, the restitutionary liability for the self-regarding conferral of benefits. As Gordon explains, individualised inquiries in this context tend to be particularly difficult. A litigated restitution case may imply that the free-riding at issue is not detrimental since, in these cases, a restitution claimant has ‘already engaged in the valuable activity’. As long as a restitution court is dealing with an isolated fact pattern, it is hard for the court to determine ‘if there exists a substantial class of people, like the claimant, who have not yet engaged in the valuable activity but would do so if restitution were assured’.76 This uncertainty justifies a refusal of restitution in such isolated factspecific cases, making intervention justified only where a court identifies classes of situations where free-riding is systemically detrimental and where subjective devaluation seems unlikely.77 Not only does the restitutionary category under consideration seem particularly hospitable to rules, but the specific rules it calls for are in fact derivative of our specific normative commitments.78 This is most clearly evident when the recommendations of autonomy and utility differ in the specific rules they endorse for delineating the scope of subjective devaluation. Thus, a utility-enhancing 74 See HLA Hart, ‘Between Utility and Rights’ in his Essays in Jurisprudence and Philosophy (Oxford, Clarendon Press, 1983) 198, 206–7; and W Kymlicka, above n 8, at 120 and 123–5. 75 cf HW Tang, ‘Storytelling in the Law of Unjust Enrichment’ ch 19 of this book. 76 Gordon, above n 26, at 474–5. 77 ibid; W Gordon, ‘On the Economics of Copyright, Restitution, and “Fair Use”: Systemic Versus Case-by-Case Response to Market Failure’ (1997) 8 Journal of Law and Information Science 7, 11–12. 78 This paragraph summarises Dagan, above n 28, at 139–45 and the sources therein.

Just and Unjust Enrichments

441

perspective recommends denying restitution when the utility loss to the defendant, had she been forced to make restitution, outweighs the gain to the plaintiff from the action that restitution could have facilitated. The injunction of autonomy is quite sharper. Even when we are relatively certain that overall the action is indeed jointly beneficial, awarding restitution in cases of potentially subjective devaluation offends the liberal commitment to individual free choice and selfdetermination. This normative difference generates a doctrinal one. Utility would support a restrictive interpretation of subjective devaluation, which potential realisability in money should suffice to defeat because, even if the recipient lacks appreciation of the conferred benefit, the market’s appreciation suffices to ensure that restitution does not generate a utility loss.79 By contrast, autonomy is more demanding. Insisting on people’s right to order their own priorities means that a benefit is deemed incontrovertible only if it has been actually converted into money or if ‘it is inevitable that the defendant will [in fact] exercise the benefit’.80 Indeed, the choice between these two alternative tests, offered respectively by the two great authorities of English law—Goff and Jones and the late Peter Birks81— requires a normative choice between the injunctions of utility and autonomy.82 Clarifying this foundational normative question turns out to be a prerequisite rather than an enemy of doctrinal clarity. Contemporary law largely follows Birks’ view, as illustrated by the orthodox position that denies restitutionary liability for improvements of the defendant’s existing interest, as opposed to its preservation. Although the line between improvements and preservation is often murky, a defendant’s objection to investing in an improvement is more likely to express her genuine valuation rather than be a strategic holdout.83 While champions of utility may challenge this prevailing normative position,84 neither existing law nor its more utilitarian alternative is indeterminate or arbitrary. Both positions make quite precise prescriptions. Moreover, given that both sets of prescriptions are morally acceptable (unlike many other possible legal regimes), a good faith choice between them by the legislature or the judiciary 79 There may be hard cases such as, for example, a landowner subjectively estimating the value of her house as far higher than its market value, with or without the cooperative improvement (to which she attributes little or no value). Imposing an immediate restitutionary liability on such a defendant may cause a utility loss if it requires her to liquidate her property at its (lower) market value. But even in these (rare?) cases, potential realisation seems to ensure the efficiency of a restitutionary liability as long as the plaintiff ’s remedy takes the form of an equitable lien to be realised at the moment of realisation of the property, rather than immediate monetary payment. 80 G Virgo, The Principles of the Law of Restitution (Oxford, Oxford University Press, 1999) 79. 81 Compare R Goff and G Jones (ed), The Law of Restitution, 6th edn (London, Sweet & Maxwell, 2002) 25, with Birks, above n 71, at 121–4. 82 My own take on this matter is that the autonomy analysis should prevail for individuals, but where the recipient is a profit-making enterprise, utility concerns should trump. 83 See H Dagan and M A Heller, ‘The Liberal Commons’ (2001) 110 Yale Law Journal 549, 586–8; D Friedmann, ‘Unjust Enrichment, Pursuance of Self Interest, and the Limits of Free Riding’ (2003) 36 Loyola of Los Angeles Law Review 831, 850. 84 See Porat, above n 13.

442

Hanoch Dagan

represents a legitimate application of authority rather than a whimsical usurpation of power. Furthermore, nothing in the ‘publicness’ of the values that inform the parties’ entitlements infringes the injunction of correlativity, properly interpreted. To begin with, autonomy and utility (in whichever balance is ultimately decided) are not introduced in any particular dispute, but rather ex ante. These values, then, set the limits for the right of potential benefactors to confer and seek restitution and the correlative obligation of potential beneficiaries to pay for such unsolicited benefits. Moreover, as is true of mistakes, although the reasons for a particular beneficiary’s liability in this account are not wholly internal to her relationship with the particular benefactor who seeks restitution, correlativity is not infringed. The pertinent rules affect both potential benefactors and potential beneficiaries. And the pertinent autonomy and utility of both of these groups that, from an ex ante perspective, by and large converge in any event, are largely convergent. Finally, at least up to this point, no interest of a third party has been invoked. Indeed, the garden variety restitution cases in our category that we have considered so far are quite easy. But this category also does (or, in my view, should) include a class of cases in which the interest of third parties is part of the justificatory premise for the parties’ entitlements. As noted,85 allowing such external reasons admittedly creates a justificatory deficit. But because the injunction to comply with the prescription of correlativity is only one, albeit important, component of private law’s justificatory burden, it may at times be outweighed by stronger competing normative concerns. The most important instance of this possibility in contemporary restitution law is indeed the expansion of the doctrine governing self-regarding conferral of benefits. In its wake, the doctrine is made to include classes of cases in which, even though the parties’ interests are not as clearly locked in, the interest of thirdparties affected by the benefactor’s decision concerning the benefit strongly support recognition of a restitutionary claim. This is the case when, for instance, one defendant settles with the plaintiff in a torts case and seeks subrogation from the other defendants after attaining full exoneration. Although, as between a payor and a tortfeasor, cases of payments to victims by a non-liable original defendant may not comply with the prescription of correlativity, the law justifiably allows subrogation in the name of the victims’ interest to be compensated promptly.86 A similar analysis applies to insurance payments for a loss that is generally within the basic scope of the policy coverage, but is beyond the insurer’s liability.87 This expansion of insurance subrogation is justified because, were insurers not 85

See text accompanying n 58. See Re Air Crash Disaster, 86 F 3d 498 (US CA, 6th Cir, 1996) 511–13, 549–50 and 552. See also, Restatement of the Law, Third, Restitution and Unjust Enrichment, Tentative Draft No 2 (Philadelphia PA, American Law Institute, 2002) § 26 comment c. 87 See State Farm Fire & Casualty Co v East Bay Municipal Utility District, 62 Cal Rptr 2d 72 (California CA) 75 and 77. See also Continental Insurance Co v Federal Insurance Co, 266 SE 2d 351 (Georgia CA, 1980) 352. 86

Just and Unjust Enrichments

443

allowed to recover colourable claims that were paid even if these claims were not ultimately covered, they would be more hesitant to accept claims altogether. An expansive approach to insurance subrogation helpfully encourages them ‘to err on the side of caution when rejecting claims’.88 Finally, as I show in some detail elsewhere, a similar analysis applies to one of the most high-profile patterns of litigation in recent times, in which governments sue industries (the tobacco industry is a prominent example) for costs they incurred in ameliorating injuries suffered by their citizens and arguably inflicted by the defendants, and for the costs of preventing or reducing the risk of such injuries in the future.89

V. Concluding Remarks My main claim in this essay is quite straightforward. The goals of the law of restitution reflect our more general normative commitments, at least in liberal settings, to autonomy, utility and community. This correspondence should not be surprising for a private law doctrine. Private law, more than any other part of our law, structures our daily interactions as individuals. No wonder, then, that the accumulated wisdom of its doctrines encapsulates the social vision that underlies these interactions.

88 MS Quinn, ‘Subrogation, Restitution, and Indemnity’ (1996) 74 Texas Law Review 1361, 1380. See also State Farm Fire & Casualty Co v East Bay Municipal Utility District, 62 Cal Rptr 2d 72 (California CA) 76–7. 89 H Dagan and JJ White, ‘Governments, Citizens, and Injurious Industries’ (2000) 75 New York University Law Review 354, 358–405.

18 The Rules of Obligations EMILY SHERWIN*

I. Introduction The choice between determinate, authoritative rules and indeterminate standards or rules of thumb is a recurring theme in the law of obligations.1 In the days of separate courts of law and equity, law tended toward rules, with equity serving as a meliorating force. Today, the role once assigned to equity is sometimes associated with restitution and the principle of unjust enrichment. In this essay, I begin with some general observations about rules, ‘equitable’ exceptions to rules, and unjust enrichment. In my view, using restitution and unjust enrichment to avoid the consequences of rules distorts the field of restitution and undermines its utility within the law of obligations. Next, I address the role Dean Hanoch Dagan has assigned to rules in his contribution to this collection.2 Dagan, to his credit, does not treat the principle of unjust enrichment as an invitation to judges to do equity in particular cases. Instead, he attempts to strike a balance between the benefits of rules and the realisation of certain broad social values he associates with restitution. Although I concur in much of what he says, I am not convinced that he has resolved the dilemma of rules.

*

Thanks to Hanoch Dagan for helpful comments. For present purposes, I accept Peter Birks’ description of the law of obligations. Birks proposed that private law is divisible into the law of persons, the law of rights, and the law of actions. Rights are further divisible into proprietary rights, which operate in rem against an indefinite number of persons, and obligations, which operate between particular parties (or their representatives). The law of obligations comprises rights based on contract, rights arising from wrongs, rights based on unjust enrichment, and rights derived from certain ‘other events’: see P Birks, English Private Law (Oxford, Oxford University Press, 2004) vol 1, xxxvl–xliii. See also TW Merrill and HE Smith, ‘The Property/Contract Interface’ (2001) 101 Columbia Law Review 773; and TW Merrill and HE Smith, ‘What Happened to Property in Law and Economics?’ (2001) 111 Yale Law Journal 357, 359, 385–7 and 394–5. The distinction between property rights and what Birks calls obligations can be difficult: see Merrill and Smith, ‘The Property/Contract Interface’. However, the problem I am concerned with in this essay is common to both, therefore nothing of significance turns on a clean division between the two. 2 See H Dagan, ‘Just and Unjust Enrichments’ ch 17 of this book. 1

446

Emily Sherwin

II. Rules, Equity and Restitution A rule, for my purposes, prescribes an outcome in terms that are readily and consistently understood by those who must apply the rule.3 Determinacy is an essential feature of rules: if the meaning of a prescription is inaccessible or controversial, it is indeterminate and cannot function as a ‘rule’ in the sense I have in mind.4 Most often, indeterminacy results from references to moral or evaluative criteria, as when actors are instructed to do what is fair, just, or reasonable. Moral or evaluative terms, however, do not necessarily produce indeterminacy, if their meaning is uncontroversial within the relevant community. At the same time, objective, non-evaluative terms may be indeterminate if they are overly technical or complex.5 A fully-effective rule must be authoritative as well as determinate. A rule is authoritative if it is intended to apply, and is in fact applied, to all cases that fall within the understood meaning of its terms. If decision-makers treat the rule as a non-binding guide or presumption, to be set aside when the outcome it prescribes fails to conform to the underlying reasons for the rule, it is not an authoritative rule but merely a rule of thumb.6 The primary function of determinate, authoritative rules is to settle controversy.7 Actors who agree on basic moral standards may nevertheless disagree about the application of those standards to concrete problems.8 Determinate rules provide answers to these questions of application, within the range of particular cases to which they apply. Authoritative settlement, in turn, allows actors to

3 For a comprehensive analysis of the variety of rules, see F Schauer, Playing By the Rules: A Philosophical Examination of Rule-Based Decision-Making in Life and Law (Oxford, Clarendon Press, 1991) 1–16. Larry Alexander and I have written extensively on the subject of rules elsewhere. See generally L Alexander and E Sherwin, The Rule of Rules: Morality, Rules, and the Dilemmas of Law (Durham NC, Duke University Press, 2001) 26–36 and 53–95. 4 On the importance of determinacy, see Alexander and Sherwin, ibid, at 30–31; Schauer, ibid, at 53–4. I assume—and have argued elsewhere—that language is capable of conveying determinate meaning: see, L Alexander and E Sherwin, Demystifying Legal Reasoning (Cambridge, Cambridge University Press, 2008) 18–24; K Greenawalt, Law and Objectivity (New York, Oxford University Press, 1992) 34–89; Schauer, ibid, at 53–68; HLA Hart, The Concept of Law (Oxford, Clarendon Press, 1961) 132–44; JL Coleman and B Leiter, ‘Determinacy, Objectivity, and Authority’ (1992) 142 University of Pennsylvania Law Review 549. 5 See Alexander and Sherwin, above n 3, at 32–4. 6 See Schauer, above n 3, at 4–5 and 104–5; J Rawls, ‘Two Concepts of Rules’ (1955) 64 Philosophical Review 3. 7 On the settlement function of law, see, eg, Alexander and Sherwin, above n 3, at 11–15; J Raz, Ethics in the Public Domain (Oxford, Clarendon Press, 1994) 187–92; MA Eisenberg, The Nature of the Common Law (Cambridge MA, Harvard University Press, 1988) 4–7. 8 The need for determinate, authoritative rules stems from disagreement and error in the application of more general standards, not from individual repudiation of the standards prevailing in a society. If everyone agrees on the meaning of a prevailing standard but some defect, the solution is sanctions against defectors; determinate rules add nothing to what is already understood: see Alexander and Sherwin, above n 3, at 11–15.

The Rules of Obligations

447

co-ordinate their actions with the actions of others.9 When rules are well designed, they also reduce the decisional errors that result from imperfect reasoning on the part of decision-makers. Determinate rules are not always the ideal choice for governance. In some contexts, settlement and co-ordination may not be predominant values.10 When this is the case, indeterminate standards or rules of thumb that allow for individual judgement may be appropriate. Indeterminate standards or rules of thumb may also be the best choice when rule-makers lack the information or wisdom necessary to design sound rules. A determinate, authoritative rule is superior to unconstrained decision-making only when compliance with the rule will result in a lesser sum of error (including the type of error that results from lack of co-ordination) over the range of cases to which the rule applies.11 Rules that do not meet this standard of net error reduction are not justified rules. The problem of unjustified rules should not be confused with the problem of sound rules that yield erroneous outcomes in particular cases. Errors of overinclusion are an inevitable, or nearly inevitable, by-product of binding rules that state their prescriptions in blunt, determinate terms.12 Decision-makers may be able to avoid errors of this kind by deviating from rules when they believe the results they call for are wrong. Over the long run, however, the same uncertainties and mistakes of judgement that made the rules necessary in the first place will lead decision-makers to err in the process of second-guessing rules. Moreover, when decision-makers fail to treat a rule as authoritative, it will no longer be reliable as a focal point for co-ordination. Thus, if the rule is sound in the sense that universal application of the rule will result in a net reduction of error, including errors of judgement and errors due to lack of co-ordination, attempts to avoid the infelicitous outcomes of rules are misguided. At least from the point of view of the rulemaking authority, the best course for decision-makers is to follow rules according to their terms, without further reflection on the merits of the outcome.13

9 On the value of coordination, see, eg, HM Hurd, Moral Combat (Cambridge, Cambridge University Press 1999) 214–21; TD Campbell, The Legal Theory of Ethical Positivism (Aldershot, Dartmouth Press, 1996) 6, 50, 53 and 58; J Raz, The Morality of Freedom (Oxford, Clarendon Press, 1986) 49–50; GJ Postema, ‘Coordination and Convention at the Foundation of Law’ (1982) 11 Journal of Legal Studies 165, 172–86. 10 One possible example is legal regulation of fraud in the context of sexual encounters. See E Sherwin, ‘Infelicitous Sex’ (1996) 2 Legal Theory 209, 224–8. For analysis of the comparative benefits of rules and standards, see generally CR Sunstein, Legal Reasoning and Political Conflict (New York, Oxford University Press, 1996) 27–8; L Kaplow, ‘Rules versus Standards: An Economic Analysis’ (1992) 42 Duke Law Journal 557, 560–72; I Erlich and RA Posner, ‘An Economic Analysis of Legal Rulemaking’ (1974) 3 Journal of Legal Studies 257, 261–71. 11 See Raz, above n 9, at 70–80 (discussing the ‘normal justification’ of rules). 12 See Schauer, above n 3, at 47–52. Some rules select between equal or neutral options for action, in order to facilitate coordination, and accordingly produce no errors. The rule ‘drive on the right’ is an example. 13 Larry Alexander and I have discussed at length the ‘gap’ in perspective that arises between rulemaking authorities, who rationally prefer that decision-makers faithfully apply rules in all cases, and decision-makers acting under rules, who may rightly conclude that the decision the rule requires them

448

Emily Sherwin

Nevertheless, there is a strong temptation for decision-makers to depart from rules not only when they believe the rules are unjustified as rules but also when they believe the rules are justified overall but produce the wrong result in a particular case. Indeed, a decision-maker who follows a rule when he or she believes it yields the wrong result is not behaving rationally.14 At the same time, decision-makers often appreciate the advantages of determinate, authoritative rules. Accordingly, in law as in other areas of life, decision-makers may search for ways to simultaneously respect rules and make exceptions to the requirements of rules. In English and American law, the first avenue of escape from the constraints of determinate and apparently binding legal rules was the court of equity. In the time of separate common-law and equity courts, equity provided relief when changed conditions or novel problems made legal rules obsolete15 and also when otherwise acceptable rules appeared to yield harsh results.16 These equitable interventions, however, were not viewed as repudiations of the governing rules. In theory, the chancellors were not adjudicating legal rights and titles but only ordering litigants to do what was right: technically, therefore, the legal rules remained in effect.17 Both common-law judges and Chancellors shared this view of the matter. As a

to make is mistaken: Alexander and Sherwin, above n 3, at 553–95. See, in addition, L Alexander, ‘The Gap’ (1991) 14 Harvard Journal of Law & Public Policy 695; Schauer, above n 3, at 128–34. 14 It is rational to follow a rule when the decision-maker believes that, although the result of the rule is incorrect, departing from the rule will undermine the rule’s capacity to coordinate conduct and reduce error, and that potential harm to the rule outweighs the potential harm from the rule’s erroneous result: see Schauer, above n 3, at 94–100 (discussing ‘rule-sensitive particularism’). It is also rational to follow the rule if the decision-maker understands that his or her own judgment may be mistaken and believes that the rule-maker is a better judge of the case at hand than the decision-maker himself: see Hurd, above n 9, at 63–5. But if, after taking account of both systemic harm from noncompliance and the rule-maker’s expertise, the decision-maker nevertheless concludes that expected harm from compliance is greater than expected harm from non-compliance, it is irrational to follow the rule: see Hurd, above n 9, at 62–94; see also GS Kavka, ‘The Toxin Puzzle’ (1983) 43 Analysis 33. For suggestions that rule-following can be rational, see, eg, SJ Shapiro, ‘The Difference that Rules Make’ in B Bix (ed), Analyzing Law (Oxford, Clarendon Press 1998) 45–54; Raz, above n 9, at 88–99; D Gauthier, ‘Commitment and Choice: An Essay on the Rationality of Plans’ in F Farina, F Hahn and S Vanncucci (eds), Ethics, Rationality, and Economic Behavior (Oxford, Clarendon Press, 1996) 217; EF McClennon, ‘Pragmatic Rationality and Rules’ (1997) 26 Philosophy & Public Affairs 210; MC Murphy, ‘Surrender of Judgement and the Consent Theory of Political Authority’ (1997) 16 Law & Philosophy 115. 15 One example is Tulk v Moxhay (1848) 41 ER 1143, in which the English Lord Chancellor developed the doctrine of equitable servitudes, enforcing restrictive covenants by injunction in circumstances that did not satisfy the antiquated privity requirements for relief at law. 16 This is the Aristotelian conception of equity: see Aristotle and T Irwin (trans), Nicomachean Ethics, (Indianapolis IN, Hackett, 1985) 1138b25. In Anglo-American law, the doctrine of constructive fraud and the highly malleable constructive trust remedy were common tools for equity of this kind. An vivid example, decided after the merger of law and equity in New York, is Sharp v Kosmalski, 351 NE 2d 721 (New York CA, 1976), in which the court granted a constructive trust in favour of a farmer who transferred his farm and other assets to a woman he hoped to marry, who then evicted him from the farm. 17 See, eg, JR v MP, YB (1459) 37 Henry VI, fol 13, pl 3 (translated and condensed in ED Rae and JR Rae, Remedies, 5th edn (New York, Foundation Press, 2000).

The Rules of Obligations

449

result, the legal system was able to maintain both an official body of rules and a less official, less publicly understood, channel for exceptions to rules. The political distinctions between equity and law eroded over time, and ultimately their functions were combined in procedurally merged courts.18 To some extent, merged courts continued to invoke equity as a source of exceptions to rules.19 It became increasingly difficult, however, to characterise equitable exceptions as independent from, and therefore not destructive of, categorical legal rules. In the modern law of obligations, the law of restitution and unjust enrichment is viewed by some as a successor to equity: a body of law that allows courts to avoid undesired consequences of rules without directly confronting the rules themselves. Restitution, on this view, is a special repository of moral principles that stand behind the formal rules of obligations and step in to provide relief when the rules produce unjust or unpalatable results. As in the case of equity, courts might use restitution to avoid obsolete legal rules,20 or they might use it to escape the requirements of normally desirable rules in recalcitrant cases.21 This approach to restitution and unjust enrichment is by no means universal. Oxford’s late Regius Professor, Peter Birks, for example, insisted that courts should develop and apply the principle of unjust enrichment by reference to established law, rather than treating it as a licence to do what they believe is just.22 In discussions of the new draft Restatement of Restitution and Unjust Enrichment, opinions have divided over the role of unjust enrichment in law. Some of those involved maintain that unjust enrichment is a principle of extraordinary justice that authorises courts to depart from rules when the rules produce unfortunate results: any case in which a rule allocates value to the wrong party is, potentially, a case of unjust enrichment. For others, however, unjust enrichment is simply shorthand for a doctrinal category composed of reasonably determinate legal rules for recovery of gains.23 18

See Federal Rules of Civil Procedure (US) r 2 (recognising ‘one form of action’). An example is Hardy v Burroughs, 232 NW 200 (1930) in which the Michigan Supreme Court cited a prior decision ‘at law’ that appeared to rule out relief to mistaken improvers of property but proceeded to grant relief ‘in equity’ in the case at hand, which presented more engaging facts. 20 In what is generally viewed as the first instance of quasi-contractual restitution at law, Lord Mansfield invoked the principle of unjust enrichment to avoid outdated procedural rules that stood as a barrier to remedies for fraud: Moses v Macferlan (1760) 97 ER 676. 21 This is exemplified by Riggs v Palmer, 22 NE 188 (New York CA, 1889) in which the court relied on the principle of unjust enrichment and the restitutionary remedy of constructive trust to make an exception to otherwise applicable rules governing wills when the designated beneficiary had murdered the decedent. 22 See P Birks, An Introduction to the Law of Unjust Enrichment (Oxford, Oxford University Press, 1985). 23 See, eg, remarks of Peter Linzer in Proceedings of the American Law Institute, 2000 (17 May 2000) 237. In general, though not in all specifics, the draft Restatement has so far adopted the second of these approaches: it does not treat unjust enrichment as a principle of justice that authorises courts to reach decisions that have no support in, or that contradict, established doctrinal rules. See Restatement of the Law, Restitution and Unjust Enrichment, Discussion Draft, 31 March 2000 (Philadelphia, American Law Institute, 2000) § 1 comment b. 19

450

Emily Sherwin

I am firmly in the second camp. In my view, unjust enrichment is an unfortunately open-ended term that should play, at most, a descriptive role in law; it should not be used to avoid the consequences of rules.24 The law of restitution and unjust enrichment responds to a variety of legal problems involving unjustifiable gains, most often due to transfers of value that were not fully voluntary. There is no reason to think that the imperatives of particularised justice are stronger when gains rather than losses are at stake. Nor are the benefits of determinate rules, including incentives for conduct and co-ordination of actions, less important in transactions involving gains than they are in transactions that cause harm or result in a loss of expectancy.25 Thus, when the principle of unjust enrichment is invoked to support departures from rules in particular cases, the result is an unexplained inconsistency in the treatment of obligations. It follows that the law of restitution and unjust enrichment should have the same formal properties as any other sector of the law of obligations. Typically, solutions to problems of wrongful gain should take the form of rules developed in the usual way by judicial and legislative rule-makers. In some areas, less determinate standards may be appropriate, but judges should not feel free to use the notion of ‘unjust enrichment’ as a tool for correcting any and all transactional outcomes that appear unjust. Understood this way, as a doctrinal category within the law of obligations, restitution is entirely distinct from, and provides no answers to, the problem of determinate rules. In all areas of law, including restitution, determinate rules are necessary to settle controversy, control error, and support co-ordination. In all areas of law, determinate rules may be poorly designed or may outlive their utility, and even when well-designed and current, may misfire in particular cases. Neither restitution, nor any other legal device, can solve this problem in a reliable way.26 When rules are justified—that is to say, when they prevent more error than they cause—they should be treated as authoritative and followed in all cases.27 Rules that may not be justified, either because they are poorly designed or because they

24

See E Sherwin, ‘Restitution and Equity’ (2001) 79 Texas Law Review 2083. See ibid, at 2095–101. 26 In theory, a legal system might be able to capture the benefits or rules and also do justice in individual cases, by limiting public awareness of the results of adjudication. In other words, law-making authorities might issue a set of determinate rules governing conduct, while judges applied a less determinate, more flexible, set of standards to decide individual cases: see GJ Postema, Bentham and the Common Law Tradition (Oxford, Clarendon Press, 1986) 403–8 and 448–52 (discussing Bentham’s proposal to separate law from adjudication); M Dan-Cohen, ‘Decision Rules and Conduct Rules: On Acoustic Separation in Criminal Law’ (1984) 97 Harvard Law Review 625, 630–36 (examining the possibility of ‘selective transmission’). To some extent, the relatively arcane defences that apply only to equitable remedies may result in a separation of this kind between rules and outcomes: see E Sherwin, ‘Law and Equity in Contract Enforcement’ (1991) 50 Maryland Law Review 253, 263–4. 27 Whether rational agents can do this is a difficult question. As noted earlier, it is not rational to follow a rule in a case in which the rule-follower believes the rule requires an erroneous result: see above n 14 and accompanying text. At the same time, we often do manage to follow rules without reflection. 25

The Rules of Obligations

451

are obsolete, should be confronted, evaluated, and overruled if they fail the test of net error reduction over the run of cases to which they apply.

III. Dagan’s Compromise In his contribution to this volume, Hanoch Dagan presents a broad outline of the nature, goals, and methods of restitution.28 Although Dagan and I agree on a number of points, I believe his description of the role of determinate rules in the law of restitution needs some further refinement. Dagan associates restitution with three fundamental values that are widely accepted in liberal societies: autonomy, utility and community.29 These three values serve as guideposts that mark off when an enrichment should be deemed unjust. In general, human interactions that yield benefits (as opposed to harms) advance the values of autonomy, utility and community. Enrichment, therefore, is presumptively just: only when ‘at least one of these core values is … undermined’ by allowing the recipient to retain her enrichment should the law step in to reallocate gains.30 Although the values Dagan invokes to distinguish between just and unjust enrichment are broad and imprecise, Dagan does not suggest that judges should treat restitution and unjust enrichment as an antidote to determinate legal rules. To the contrary, Dagan recognises the benefits of rules and contemplates that in most cases, the values that shape restitution should be implemented through rules. Judges ordinarily should refer to established doctrine to decide cases, rather than engage in ‘ad-hoc judgments’ based on autonomy, utility and community.31 So far, there is no inconsistency in Dagan’s position. Rules seldom function exclusively to co-ordinate conduct. Most rules are also designed to advance some background principle or value by instantiating the principle or value in a more concrete prescription for action.32 The objective is that, on average, compliance with the rule will increase compliance with the principle or value it reflects. Thus, rules can certainly co-exist with abstract values of the kind Dagan cites. The difficulty lies in defining the proper relationship between rules of restitution and the reasons that support them. When a rule is treated as authoritative, its underlying rationale—the principle, value, or combination of values that stands behind it—ceases to operate as a ground for decision.33 Rationales motivate the promulgation of rules and shape their terms, but once a rule is in place their 28

See Dagan, above n 2. See ibid. 30 ibid. 31 ibid. 32 See Schauer, above n 3, at 54 (describing rules as concrete ‘instantiations’ of background values). 33 See J Raz, The Authority of Law (Oxford, Clarendon Press, 1979) 16–19, 22–3 and 30–33 (discussing rules as ‘exclusionary reasons’ for action). 29

452

Emily Sherwin

work is done. Decision-makers are expected to follow the rule without inquiring into the fit between particular outcomes of the rule and its background rationale. Of course, the meaning of a rule may be ambiguous in some cases, even when the rule is stated in substantially determinate terms. In cases of doubt, meaning depends on the intentions of the authority that posited the rule,34 which may be a legislative body or a precedent court.35 In either case, however, the only intentions that count in applying an authoritative rule are intentions about the import of the words used in the rules. Intentions about what should happen in cases not covered by the rule, or what should happen when the outcome of the rule is contrary to the rule’s underlying rationale, are not relevant to application of the rule. Thus, when a decision-maker refers to the rationale of a rule to determine what the rule-maker would have intended in a case to which the rule does not apply, the decision-maker is effectively positing a new rule. When the decision-maker refers to the rationale of the rule to determine that the rule-maker would not intend a result required by the terms of the rule, the decision-maker is effectively overruling the rule.36 In either of these situations, the rule itself does not constrain the decision. Returning to Dagan, although Dagan sees the virtues of rules and endorses rule-guided decision-making in the field of restitution, he is not willing to treat established doctrinal rules as fully and inevitably authoritative. Instead, he proposes that, while judges should not ‘routinely’ make direct reference to autonomy, utility and community, ‘some legal actors—notably judges of appellate courts— [should] occasionally use new cases as triggers for an ongoing refinement of rules, as opportunities to revisit the normative viability of existing rules qua rules’.37 Dagan’s position on the relationship between rules and the rationales that motivate and shape them can be understood in two ways. First, he may have in mind that judges should engage in what Frederick Schauer has called ‘presumptive positivism’.38 Secondly, he may have in mind that judges should overrule rules when the rules no longer meet the test of net error reduction over the range of cases to which they apply, but otherwise should follow the rules without exception. Presumptive positivism is an intermediate approach to rules. To clarify just what presumptive positivism entails, Schauer contrasts it first with ‘particularism’, 34 This is a contested point. Larry Alexander and I have defended our intention-based view of interpretation in other writings: see Alexander and Sherwin, above n 4, at 98–101 and 131–45. 35 There are, of course, problems in conceptualising, let alone divining, the intentions of a multiperson rule-making body. For present purposes, I set these aside and assume that in at least some cases, intentions can be identified: see Alexander and Sherwin, above n 3, at 98–122. When the intentions of the rule-maker are not ascertainable from the words and context of the rule, the rule has no effect. In this situation, what purports to be ‘interpretation’ of the rule is, in effect, an exercise of independent rule-making authority: see Alexander and Sherwin, above n 3, at 119–22; Alexander and Sherwin, above n 4, at 170–71. 36 Alexander and Sherwin, above n 4, at 68–70 (maintaining that outcomes of a rule may be infelicitous and nevertheless be part of the intended meaning of the rule). 37 Dagan, above n 2. 38 See Schauer, above n 3, at 202–5; F Schauer, ‘Rules and the Rule of Law’ (1991) 14 Harvard Journal of Law and Public Policy 645.

The Rules of Obligations

453

and then with ‘rule-sensitive particularism’.39 Simple particularism gives no independent weight to rules: the judge decides in accordance with the rule if, but only if, the rationale for the rule (or the judge’s own best moral reasoning) leads to the same result. Rule-sensitive particularism takes into account the advantages of determinate rules and the expected loss of rule-based advantage if the judge defects from the rule. A decision that departs from the rule will undermine, to some extent, the future reliability of the rule and, accordingly, will decrease the rule’s capacity to settle controversy, co-ordinate conduct and constrain erroneous decision-making. Call this systemic harm. On the other hand, because rules are blunt instruments, a decision that conforms to the rule may undermine the values that support the rule (in Dagan’s case, autonomy, utility and community). Call this harm to values. When systemic harm outweighs harm to values, a rulesensitive particularist will follow the rule; when harm to values predominates, the rule-sensitive particularist will depart from the rule and decide on the basis of its underlying rationale. Both Schauer and Dagan reject rule-sensitive particularism.40 The difficulty with rule-sensitive particularism is that the effects of defections from rules are cumulative. With each defection, the rule is somewhat less reliable; therefore additional defections by will cause less systemic harm, and harm to values will more easily prevail. Ultimately, nothing is left of the rule.41 Presumptive positivism, which Schauer endorses, purports to bolster the authority of rules. According to this view, judges ordinarily should apply determinate legal rules according to their terms, without evaluating the fit between outcome and underlying rationale.42 However, judges can and should ‘peek’ at rationales to check for egregious conflicts.43 If the outcome of the rule appears significantly at odds with the rule’s rationale, the judge can then engage in a full rulesensitive-particularist analysis of the harms and benefits of following the rule. This is an appealing approach, which appears to produce a compromise between rules and the values they are supposed to promote. Presumptive positivism, however, is not a stable solution to the dilemma of rules. As long as there are some cases in which the conflict between outcomes and rationales is obvious and harm to underlying values overcomes systemic harm—as the theory of presumptive positivism assumes there must be—defections will occur and systemic harms will carry less weight.44 Perhaps there is some equilibrium state in which obvious conflicts arise so seldom that systemic harm will not be a significant concern. But there is no assurance that the standard that defines the point at which the presumption in favour of rules is overcome—egregious conflict, obvious conflict, 39

See Schauer, above n 3, at 94–100. See ibid; and Dagan, above n 2. 41 See Alexander and Sherwin, above n 3, at 61–8. The effects of defection from rules can be telescoped, such that the rule is loses all power to constrain from the outset. 42 See Schauer, above n 3, at 202–5; Schauer, above n 38. 43 Schauer, above n 38, at 677. 44 See Alexander and Sherwin, above n 3, at 68–73. 40

454

Emily Sherwin

substantial conflict, or some other tipping point—will capture that equilibrium state. Moreover, decision-makers may err in applying the standard. Therefore, equilibrium seems unlikely, and if equilibrium is not reached, the rule imposes no effective constraint.45 Alternatively, Dagan may simply be suggesting that courts must have power to overrule established rules, based on the broad societal values that inform the law of restitution. The power to overrule is an important feature of any legal system that recognises authoritative rules. Without it, there would be no escape from unjustified rules, including rules that were poorly designed from the outset or rules that have lost their justification as conditions have changed. Legislatures have inherent power to overrule past enactments, as a consequence of their general authority to enact rules. Because legislative intervention is not always practical, most systems also afford similar powers to courts both to establish binding precedent rules and to overrule precedent rules.46 When courts overrule precedents, they must then rely on their own best judgement, informed by applicable values, either to decide the single case before them or to establish a new rule. The power to overrule, however, differs substantially from presumptive positivism. Presumptive positivism purports to leave established rules in place but permits courts to make exceptions to the rules. Overruling eliminates existing rules.47 Overruling is appropriate when, and only when, the rule in question is not a justified rule. That is to say, overruling is appropriate when the rule, judged by reference to the values that stand behind it, will cause more errors because of obsolescence or poor design than it will prevent by settling controversy, coordinating conduct and pre-empting faulty reasoning. Overruling is not an appropriate strategy when rules that are justified over the full range of cases to which they apply produce bad results in some of those cases. Dagan’s reference to rules ‘qua rules’ suggests that he does in fact intend to limit the role of autonomy, utility and community to judgements about the overall effects of rules. In other words, he is endorsing the practice of overruling unjustified rules, but not the practice of presumptive positivism. Autonomy, on this interpretation, does not support departure from a rule unless the rule, on average, produces a net reduction in autonomy. If this is Dagan’s intent, his approach is

45 A further difficulty is that, just as it is irrational for a decision-maker to follow a rule when the decision-maker believes the outcome of the rule is wrong, so too it is irrational to presume in favour of a rule (and, accordingly, follow the rule) when the decision-maker suspects the outcome of the rule is wrong. 46 See Eisenberg, above n 7, at 104–5 (suggesting that the power to overrule is a complement to the power to develop law). 47 The practice of distinguishing rules is a form of overruling. Courts sometimes hold that, although a precedent rule applies by its terms to a case currently before the court, the rule is distinguishable because some fact that was present in the precedent case that established the rule is not present in the current case. This form of analysis creates the impression that the precedent rule is still in place. However, because the facts of current cases will always differ in some respects from the facts of precedent cases, a ‘distinction’ of this kind will always be possible. It follows that the precedent rule has no constraining effect on the current decision.

The Rules of Obligations

455

logically impeccable. Yet, the fine distinction it calls for, between particular errors and net error overall, may be difficult for human decision-makers to achieve. Legislatures naturally assess rules as rules, because their role as legislatures is to enact rules prospectively, having in mind the full range of cases to which the rule applies. Exceptions may occur: legislators may be moved to act by a particular case or problem that has captured public attention, and as a result may overlook less salient consequences of the rule they select. Typically, however, legislatures act with reference to an indefinite set of future cases. In contrast, the primary role of judges is to apply rules to discrete events that have already materialised. The first aspect of a rule the judge confronts is the rule’s outcome in the particular case before the court. In a rule-governed system, rules are authoritative; therefore the judge is expected to decide the case as the rule dictates without reflection on the underlying values that motivate the rule. When judges also have power to establish and overrule rules, their relationship to rules becomes quite complex. To determine whether a rule is justified, the court must compare the rule’s performance to its rationale.48 If—judged in terms of whatever set of values lies behind the rule—errors of over-inclusion and misdesign resulting from application of the rule outweigh the errors of judgement and co-ordination errors that will occur in the absence of a rule, the rule lacks justification and should be overruled. If errors of over-inclusion and mis-design are outweighed by errors of judgement and co-ordination errors, the rule is justified. In that case, the judge should not overrule and should proceed to apply the rule according to its terms. Yet, once a judge has consulted the rule’s rationale to answer the question of overall justification, it will be exceedingly difficult for the judge to screen out the rule’s rationale at the point of application. Discrepancies between the reasons that support the rule and the rule’s outcome in the case before the court will be evident, and consequently hard to tolerate; therefore the temptation will be strong to overrule rules that are beneficial overall.49 Thus, if what Dagan has in mind is not presumptive positivism but cautious overruling of unjustified rules, his approach is sound in theory but problematic in practice. In theory, judges should refer to established rules to define just and unjust enrichments when the rules are justified, but should overrule rules that, judged as rules, run contrary to the values of autonomy, utility and community. In practice, this is a very difficult idea to implement. One possible solution is to adopt a presumption against overruling: judges should overrule only those rules that generate significantly more error than they 48 Ultimately, the rationale of any rule (as well as the proper basis of any decision) is to do what is best, all things considered. Due to the practical limitations of human reasoning, both legislatures and courts are likely to focus on a more limited set of principles or values both in promulgating rules and in repealing or overruling rules. 49 The situation is exacerbated by the common cognitive bias in favour of salient, readily ‘available’ facts. For judges deciding live disputes, the effect of the rule on parties who have narrated their stories before the court is likely to overshadow the benefits of the rule in other case, which remain comparatively abstract and remote.

456

Emily Sherwin

prevent.50 A presumption of this kind would help to counteract the natural tendency of judges to overrule prematurely, when justified rules produce unfortunate results. Yet, like all compromises in regard to rules, a presumption against overruling is unreliable in application. There is no way to draw the line between mildly unjustified rules, which judges should leave in place, and seriously unjustified rules, which judges should excise from the body of law. Moreover, wherever the dividing line may lie, it is not rational for judges to retain and obey rules that, in their view, do more harm than good.

IV. Conclusion The dilemma of rules arises because justified rules—rules that, judged accordingly to their underlying rationales, prevent more error than they cause—will produce bad outcomes in some of the cases to which they apply. This dilemma is especially troublesome for judges, who confront rules from two angles. As adjudicators, judges apply rules to particular facts; as common law rule-makers, they must determine whether rules are justified as rules, taking into account the full range of cases encompassed by the rules’ predicate terms. Accordingly, the dilemma of rules is most pronounced in the area of obligations, in which rules typically are crafted by judges. There is no easy solution to the problem. Unconstrained decision-making not only gives too much play to bias and faulty judgement, but also deprives private actors of the benefits of co-ordination. Blind obedience to rules entrenches rules that are not, or are no longer, justified. Compromises in which judges follow rules most of the time but depart from rules when they are seriously offended by the rules’ results are ultimately unstable. A judicial power to overrule, limited to rules that are not justified as rules, is an analytically sound response, but elusive in practice. Of course, judges have been managing the dilemma of rules for centuries. The law of obligations has not remained static. At the same time, most would agree that it has evolved in the form of recognisable doctrine that imposes at least some constraint on the outcomes of adjudication. For a time, equity provided the courts with a means of updating rules without appearing to undermine their authority. Modern courts have been more willing to overrule rules, either directly or under the guise of distinguishing cases or reasoning by analogy. In one form or another, judicial resistance to rules is an inevitable and probably healthy feature of a legal system. The law of restitution, however, serves independent purposes and should not be conscripted for the purpose of evading either unjustified rules or unpalatable outcomes of justified rules.

50

See Alexander and Sherwin, above n 4, ch 2, § C.

19 Storytelling in the Law of Unjust Enrichment TANG HANG WU*

I. Introduction The development of the law of unjust enrichment is unique in several respects. First, the law has developed at an amazing pace considering that the House of Lords only officially recognised this cause of action less than 20 years ago in Lipkin Gorman v Karpnale.1 The second remarkable feature of this branch of law is the great influence of academic thinking in the advancement of the law.2 Among all the legal scholars who have contributed to the development of the law, Professor Peter Birks must stand out as one of the central figures.3 Together with Robert Goff and Gareth Jones, Birks was one of the modern architects who gave the subject a formal structure through his writings and the legendary graduate seminars at Oxford.4 In reading Birks’ work, I am always struck by how prevalently stories feature in his scholarship. For example, Birks starts An Introduction to the Law of Restitution with the following story: ‘Children quickly learn that if they give their toys away, they cannot expect to get them back. Anxious about spoiling, parents will not intervene unless there is more to it than a mere change of mind’.5 Birks used this illustration as a prelude before launching into a discussion of the competing tensions * I am grateful to my wife, Choog Ing, for proofreading an earlier draft. Thanks are also due to Andrew Robertson, Matthew Harding, Tham Chee Ho, Birke Häcker, Mindy Chen-Wishart and Graham Virgo for their helpful comments. This chapter is dedicated to the memory of my father, Tang Eng Teik, who told me many bedtime stories when I was a child. 1 Lipkin Gorman v Karpnale [1991] 2 AC 548 (HL). 2 Three Lexis-Nexis searches conducted on 24 June 2008 using the key words ‘Birks w/5 restitution’, ‘Burrows w/5 restitution’ and ‘Virgo w/5 restitution’ in the Commonwealth and Irish cases database of judgments revealed 179, 95 and 28 decisions respectively containing these key words. These key words refer to Peter Birks, Andrew Burrows and Graham Virgo who are the leading academics on the law of unjust enrichment. 3 See Lord Walker’s tribute to Peter Birks in Deutsche Morgan Grenfell Group plc v Inland Revenue Commissioners [2007] 1 AC 558 (HL) 611. 4 For a recent overview on the growth of the subject in the Commonwealth, see C Saiman, ‘Restitution in America: Why the US Refuses to Join the Global Restitution Party’ (2008) 28 Oxford Journal of Legal Studies 99. 5 P Birks, An Introduction to the Law of Restitution (Oxford, Clarendon Press, 1989) 9.

458

Tang Hang Wu

in the law in this area, ie the impulse to relieve a person from the consequences of a mistaken transfer of wealth versus the security of receipt. This is vintage Birks; he uses everyday stories to clarify or reinforce a particular legal position. Although Birks does not explicitly say so, he is, whether consciously or unconsciously, employing a form of storytelling in the law.6 Birks’ technique of storytelling is to use narratives as support for conventions and constructs7 of many of the doctrines in the law of unjust enrichment. Kim Lane Scheppelle describes this technique in a legal storytelling symposium as follows: ‘Stories carry power because they have the ability to convey truths even if the stories themselves are not the only ways of seeing the world. Stories re-present experience, and can introduce imagination and new points of view’.8 The purpose of this chapter is to study these stories, which animate much of the important doctrines in unjust enrichment law, and, at times, to challenge them and present counter-narratives. One might ask: Is it worth devoting an entire chapter to Birks’ stories? And, what has this got to do with the theme of this book, ie The Goals of Private Law? Taking the first question, I believe, on closer scrutiny, many of Birks’ stories, which greatly inform his scholarship, are subject to challenge. For example, I must confess I have always been quite puzzled by Birks’ tale about children, toys and their parents, especially the air of confidence accompanying this particular narration.9 How is Birks so sure of how parents would react in this situation? His stories are often told so confidently that they impart an air of inevitability about the legal conclusion that follows. In this chapter, I suggest many of Birks’ stories are controversial in that either they do not reflect the typical context in which unjust enrichment claims arise or they are based on certain controvertable assumptions. Hence, the legal conclusions that follow might not be as inevitable as they appear. Thus, it is important to re-examine many of Birks’ stories because these narratives inform and shape many of the doctrines found in the law of unjust enrichment. Otherwise, there is the danger that Birks’ narratives could become the ‘stock stories’10 which form the foundation of the dominant discourse in the law of unjust enrichment. The answer to the second question is closely related to the first point: there is an important relationship between Birks’ stories and goals of unjust enrichment law. Many of the themes in unjust enrichment law such as: (a) the furtherance of autonomy of the claimant; (b) the protection of the defendant’s interest in security of receipt through the operation of the change of position defence; (c) the promotion 6 There are many strands of legal storytelling. See JB Baron, ‘Law, Literature and the Problem of Interdisciplinarity’ (1999) 108 Yale Law Journal 1059; P Brooks and P Gerwirtz (eds), Law’s Stories: Narrative and Rhetoric in the Law (New Haven CT, Yale University Press, 1996). 7 For a paper on stories as conventions and constructs in the law see JB Baron and J Epstein, ‘Is Law Narrative?’ (1997) 45 Buffalo Law Review 141. 8 KL Scheppele, ‘Foreword: Telling Stories’ (1989) 87 Michigan Law Review 2073, 2075. 9 In fact, one of his former students pointed out to Birks that this statement was wrong as a matter of law: see G Dannemann, ‘In Memoriam Peter Birks’ (2004) Oxford University Comparative Law Forum 2, at ouclf.iuscomp.org/articles/ In_Memoriam_Peter_Birks.shtml, accessed on 1 November 2008. 10 On stock stories, see G López, ‘Lay Lawyering’ (1984) 32 UCLA Law Review 1.

Storytelling in the Law of Unjust Enrichment

459

of rationality in the law through the development of a coherent system of legal taxonomy; and (d) the use of Roman and comparative law in unjust enrichment scholarship to advance understanding, are fortified and reinforced by the stories that Birks tells. While these goals are certainly laudable, the interesting questions which arise from this study of Birks’ stories are these: Who is the autonomous legal person? What interest is contained in the concept of security of receipt? Is the defendant’s interest really protected by the change of position defence? What are the implications of the Birksian legal taxonomy? Why is Roman law so heavily used in the Western world? My aim in writing this chapter is not to undertake the onerous task of presenting a wholly new construction of the structure of unjust enrichment law; rather the aspiration of this chapter is more modest. In re-evaluating Birks’ stories, we are forced to re-assess our understanding of some of the fundamental assumptions underpinning the goals of unjust enrichment law and also the legal methodology used to develop this branch of the law. It is my argument that some of these assumptions are false, and therefore the doctrinal rules in some of these areas and the jurisprudential method employed ought to be reconsidered.

II. The Story of the Autonomous Legal Person One of the most important stories implicit in unjust enrichment law is the story of the autonomous legal person. The autonomous legal person is as crucial in unjust enrichment law as the man on the Clapham omnibus or Bondi tram is in tort law. Although the legal person in unjust enrichment law is not usually explicitly mentioned, I believe that he is the central figure who acts as the normative yardstick for much of unjust enrichment law. Consider Birks’ original articulation of the grounds of restitution, ie that an ‘unjust factor’ must be present in order for an unjust enrichment claim to be allowed. In An Introduction to the Law of Restitution,11 Birks says that in order for the plaintiff to establish a claim to an enrichment received by the defendant, the plaintiff has to demonstrate the proposition—‘I did not mean you to have it’. According to Birks, the plaintiff is effectively saying in these cases: ‘My judgment was not properly exercised (or, was not exercised at all) in the matter of your getting it (the enrichment)’.12 Together with Robert Chambers, Birks has stated that: ‘the unjust factor can be represented in nursery terms as: 1) I didn’t mean it; 2) it was bad of you to receive it; and 3) Mother says give it back anyway’.13 Category 1 includes circumstances where the transferor lacked the intent to make the transfer, ie she was ignorant or mistaken about the transaction. It could also be that the transfer was made on a particular 11

Birks, above n 5, at 140. ibid. cf H Dagan, ‘Just and Unjust Enrichments’ ch 17, who argues that the law of unjust enrichment is motivated by the values of autonomy, community and utility. 13 P Birks and R Chambers [1997] Restitution Law Review (Supplement) 2. 12

460

Tang Hang Wu

basis which subsequently failed. This analysis is sometimes called the ‘unjust factor’ approach because the claimant has to identify a ground of restitution, eg ignorance, mistake, failure of consideration, etc. With regard to category 2, the unconscientiousness of the receipt is the explanatory force for restitution. The final category is where there is a compelling policy reason why the law mandates the defendant to make restitution to the claimant. The interesting thing about this ‘unjust factor’ approach is that although this approach has been accepted by some judges14 and many leading academics working in this field,15 there is a fundamental jurisprudential question that remains largely unexplored: Why should the plaintiff ’s lack of intent be a source of a right against the defendant?16 To my knowledge, Birks has never developed this inquiry fully save for saying that the nature of these transactions ‘cripples autonomy’.17 Therefore, it seems that at the heart of the ‘unjust factor’ approach is this story: an unintended transfer of wealth must be reversed save where the defendant has changed his position because to let the transaction stand would be to compromise the claimant’s autonomy. When we explore this line of thought, a whole host of interesting questions arises. Who is the autonomous legal person?18 Why is his autonomy affected by an unintended transfer of wealth? While much valuable work has been done in explicating the legal person in other areas like tort19 and criminal law, the concept of the legal person in the law of unjust enrichment has never really been developed. Yet his unspoken presence looms large because it is his values, his compromised autonomy, that dictate whether a claim in restitution should be allowed. And so it is vital for unjust enrichment scholars to unpack the concept of the autonomous legal person. In the dominant account of English restitution law, the concept of the autonomous legal person is equated with the idea of the claimant’s pure subjective intent. It is as if the notion of the self is merged with the person’s intention. This is illustrated by Andrew Burrows’ example: If a person who loves dogs and hates cats pays £1,000 to a charity for the care of cats mistakenly believing that it is a charity for the care of dogs, that is surely a serious mistake that should ground restitution … [although] the reasonable person will not be told that the payor loves dogs but hates cats.

14 See eg Deutsche Morgan Grenfell Group plc v Inland Revenue Commissioners [2007] 1 AC 558 (HC); and R (Rowe) v Vale of White Horse District Council [2003] 1 Lloyd’s Rep 418 (QB). 15 See, eg, A Burrows, The Law of Restitution (London, Butterworths, 2002). G Virgo, Principles of the Law of Restitution (Oxford, Oxford University Press, 2006). cf H Dagan, The Law and Ethics of Restitution (Cambridge, Cambridge University Press, 2004) 42–4; Dagan, ‘Just and Unjust Enrichments’ ch 17, who contends that the normative goals of unjust enrichment are autonomy, utility and community. 16 For recent attempts to solve this problem see JM Nadler, ‘What Right Does Unjust Enrichment Protect’ (2008) 28 Oxford Journal of Legal Studies 245. 17 P Birks, Unjust Enrichment, 2nd edn (Oxford, Clarendon Press, 2005) 106. 18 See, eg, N Naffine, ‘Who Are Law’s Persons? From Cheshire Cats to Responsible Subjects’ (2003) 66 Modern Law Review 346 for an attempt to flesh out the legal person in law. 19 See, eg, M Moran, Rethinking the Reasonable Person (Oxford, Oxford University Press, 2003).

Storytelling in the Law of Unjust Enrichment

461

In this construction of the autonomous legal person, once the plaintiff ’s subjective intent with regard to a transfer of wealth is compromised, so too is his autonomy. As Hanoch Dagan explains, [v]alidating a wealth transfer based on such vitiated judgement would offend the liberal commitment to individual free choice; it would violate the maxim that the exercise of (subjective) free will should be the prerequisite to any legitimate transfer of, or interference with, resources.20

If there are no countervailing concerns like the defendant’s change of position, then the law of unjust enrichment must step in and reinstate the claimant’s autonomy. Once this conception of the autonomous legal person is articulated, one can see a glaring artificiality in the construction of the autonomous legal person. It is as if the legal person exists in the abstract without any connection to anyone else. He is essentially an atomistic individual who values the furtherance of his subjective intent above everything else. It is unclear whether this so-called autonomous legal person forms relationships with other human beings; if so, we are not told how to balance the value of these relationships with the legal person’s obsession with fulfilling his subjective intent. The lack of contextual relationships surrounding such an autonomous legal person has led some scholars like Naffine to call him the ‘antisocial man in the state of nature’.21 Mindy Chen-Wishart reaches a similar conclusion and christens him the ‘Super Detached Man’.22 The autonomous legal person seems to be someone who cares principally about maximising his subjective intent—a person who mostly wants to be left alone, does not usually form relationships and is never a parent, a spouse or a friend. As Chen-Wishart points out, ‘the paradigm autonomous person is the rugged individualist’, someone who is ‘self-sufficient, self-knowing, and rationally self interested’.23 It would seem that this model of the autonomous legal person is a form of a separation thesis about what it means to be an autonomous human being.24 A sense of separateness from others is the key to achieving autonomy. As Robin West explains, the separation thesis assumes: ‘I am one human being and you are another, and that distinction between you and me is central to the meaning of the phrase “human being”. Individuals are … distinct and not essentially connected with one another’.25 The problem with this idea of the autonomous legal person is that this model breaks down when it is used to analyse a situation where there is a close relationship 20 H Dagan, The Law and Ethics of Restitution, above n 15, at 42. However, it must be noted that Dagan’s account of unjust enrichment law is vastly different from most English scholars. Dagan argues for pluralistic norms in unjust enrichment law, ie utility and community in addition to autonomy. 21 N Naffine, ‘In Praise of Legal Feminism’ (2002) 22 Legal Studies 71, 83. Craig Rotherham has suggested to me that a more accurate term would an ‘asocial man’ instead of ‘antisocial man’. 22 M Chen-Wishart, ‘Undue Influence: Vindicating Relationships of Influence’ (2006) 59 Current Legal Problems 230, 240. 23 ibid, at 240. 24 R West, ‘Jurisprudence and Gender’ (1988) 55 University of Chicago Law Review 1. 25 ibid, at 1.

462

Tang Hang Wu

between the claimant and the defendant. By ignoring the contextual relationship between the parties and by giving precedence to the claimant’s subjective intent above all other considerations save for the defendant’s monetary change of position, the dominant school of thought in unjust enrichment scholarship paints an unflattering picture of relationships and sometimes reaches counter-intuitive conclusions. Two situations where this problem is particularly stark are: (a) undue influence; and (b) the law on mistaken gifts. The atomistic view of the autonomous legal person in unjust enrichment law affects the conceptualisation of the jurisprudential basis of the law of undue influence and, more importantly, how the law depicts the nature of close human relationships. Birks and Chin, in an influential article on undue influence, explain the rationale of the law of undue influence as follows: the relevant weakness of the plaintiff is that, within the relationship, by reason of excessive dependence on the other person, he or she lacks the capacity for self-management which the law attributes to the generality of adults … The judgmental capacity of the party seeking relief must have been markedly sub-standard.26 (emphasis added)

Thus, Birks and Chin justify relief for undue influence ‘on the ground of impaired judgmental capacity usually arising from morbid dependence on another’ (emphasis added).27 In analysing the iconic case of Allcard v Skinner,28 which involved a former novice nun giving up most of her wealth to a convent, Birks and Chin say that the claim succeeded because the novice nun was ‘excessively spell-bound … her autonomy was impaired to an exceptional degree’.29 Is this depiction of relationships and autonomy accurate? Is it true that when a person trusts and loves someone else completely, the person is morbidly dependent on another so much so that an individual’s personal autonomy is threatened? And that the law must guard against someone becoming excessively spell-bound with his or her religion? It is suggested that these characterisations of relationships are unflattering and inaccurate. They are erroneous because Birks and Chin proceed with the wrong starting point in their construction of the autonomous legal person, which leads them to paint such an unattractive and, ultimately, inaccurate picture of human beings and relationships. The description of excessive or morbid dependence on another person as the basis of undue influence is objectionable; it depicts a person’s dependence on another human being as unnatural and undesirable. A more accurate depiction of human autonomy and a new conception of the autonomous legal person which is sensitive to the relationship between the parties is required to analyse the doctrine of undue influence. The starting point in constructing a more realistic view of an autonomous person is this: not all relationships are necessarily an anathema to personal 26 P Birks and NY Chin, ‘On the Nature of Undue Influence’ in J Beatson and D Friedman (eds), Good Faith and Fault in Contract Law (Oxford, Oxford University Press, 1995) 66. 27 ibid, at 85. 28 Allcard v Skinner (1887) 36 Ch D 145 (CA). 29 Birks and Chin, above n 26, at 68.

Storytelling in the Law of Unjust Enrichment

463

autonomy. Some relationships are positive and enabling. In order ‘to be differentiated human persons, unique individuals, we must have relationships with the social and natural word’.30 As Jennifer Nedelsky argues, ‘[i]f we ask ourselves what actually enables people to be autonomous, the answer is not in isolation, but relationships—with parents, teachers, friends, loved ones—that provide the support and guidance necessary for the development and experience of autonomy’.31 In making this argument I am not arguing that all relationships have a positive effect on personal autonomy. Of course, some relationships are destructive and abusive and these relationships pose a definite threat to an individual’s autonomy. The difficult challenge is to differentiate relationships which are enabling and those that threaten an individual’s autonomy. It is only the latter which should be caught by the doctrine of undue influence. Birks and Chin’s strategy of looking for some form of morbid dependence is simply an over-inclusive view of analysing human relationships. Chen-Wishart, in a recent article on undue influence, rightly disputes Birks and Chin’s characterisation of autonomy and relationships; she argues that ‘relationships have obvious instrumental values’32 and that the Birks-Chin model of undue influence ‘is unrealistic and insulting as a general description of undue influence. It denigrates the values of attachment’.33 Instead, Chen-Wishart re-imagines a model of the law of undue influence using the concept of relational autonomy. At the risk of over-simplification, Chen-Wishart’s idea of relational autonomy and undue influence involves: (a) setting relational norms about how people in close relationships ought to behave; and (b) vitiating transactions that seriously jeopardise the claimant’s chances of leading an autonomous life. Chen-Wishart’s thesis is certainly more attractive than the Birks-Chin’s model and presents a more realistic and compelling depiction of close personal relationships. Apart from undue influence, the story of the autonomous legal person also affects the legal doctrine of when a person may claim restitution for an alleged mistaken gift. Since there is no direct case law on this issue, influential academic literature provides us with a clue as to how the law might develop. The dominant discourse in unjust enrichment law is that a donor ought to be able to get restitution so long as the donor can show that ‘but for’ a mistake, the donor would not have made the gift.34 Consider Burrows’s analysis of an uncle giving a birthday present of £1,000 to his niece not knowing that she married X, a man the uncle hates. Burrows argues that, provided the ‘payor can establish that, had he known that the niece had married X, he would not have made the payment, 30

MJ Radin, ‘Market-Inalienability’ (1987) 100 Harvard Law Review 1849, 1904. J Nedelsky, ‘Reconceiving Autonomy: Sources Thoughts and Possibilities’ (1989) 1 Yale Journal of Law & Feminism 7, 9. 32 Chen-Wishart, above n 22, at 242. 33 ibid, at 243. 34 See, eg, Birks, above n 17, at 148–50. Cf Lord Scott’s preliminary observation in Deutsche Morgan Grenfell Group plc v Inland Revenue Commissioners [2007] 1 AC 558 (HL) 812 that the causative mistake approach may be inapplicable to gifts. 31

464

Tang Hang Wu

there seems no good reason to deny restitution’.35 I have argued elsewhere that this analysis should be rejected for a number of reasons.36 What is relevant here is that Burrows’ analysis relies on a flawed construct of the autonomous legal person and ignores the fundamental nature of gifts and relationships. Most gifts are by nature concerned with relationships and a legal framework that overlooks its social context is intrinsically deficient. A new conception of autonomy, which reconciles the value of self-determination and ‘the centrality of relationships in constituting the self ’37 is necessary in dealing with this issue. I would argue that in this case the uncle’s autonomy is not compromised if he was ignorant of the niece’s spouse. The other motivations for the gift such as love and affection for his niece and his desire to forge an emotional connection remain intact at the time the birthday gift was made. It is highly artificial to say that the uncle’s autonomy is compromised in this story. It is interesting that Birks tells another variation of the story in Unjust Enrichment: If an uncle gives £1,000 each to all his nephews and nieces and then finds that one of them is gay, the fact that he is a notorious homophobe will not in itself show that he made a mistake which caused that gift. At the very least the uncle would have to show that his belief as to sexual orientation was actively in his mind when the gift was made.38

Birks’ story is another vivid example of a version of autonomy which equates the concept of subjective self intent with the idea of the autonomous legal person. In this case, it can be argued restitution is not justified even if the uncle’s belief about the donee’s sexual orientation was actively in his mind. Why should the uncle’s subjective bias and prejudices be deemed to be more important than the relationship meant to be forged by the gift? Why is it that the uncle’s unspoken assumptions take precedence over the nephew/niece’s status as a rational and freewilled agent? I would argue that by taking a more relational view of autonomy, the conclusion would be that restitution of the gift ought not be allowed. From the examples in undue influence and mistaken gifts, it is tempting to caricature the autonomous legal person in Birks’ world as a purely rugged individual who does not care for other individuals. However, I believe the position is more complex and nuanced than this. Recall another of Birks’ famous stories which has spawned many law review articles. It is originally told as follows: Suppose that I see a window-cleaner beginning to clean the windows of my house. I know that he will expect to be paid. So I hang back unseen till he has finished the job; then I emerge and maintain that I will not pay for work which I never ordered. It is too late, I have freely accepted the service. I had my opportunity to send him away. I chose instead to let him go on. I must pay the reasonable value of his work.39 35 36 37 38 39

See Burrows, above n 15, at 144–5. HW Tang, ‘Restitution for Mistaken Gifts’ (2004) 20 Journal of Contract Law 1. Nedelsky, above n 31, at 12. Birks, above n 17, at 149–50. Birks, above n 5, at 265.

Storytelling in the Law of Unjust Enrichment

465

Birks narrates this story as support for the doctrine of ‘free acceptance’. The idea of free acceptance is that an unjust enrichment claim is justified where the recipient accepts a benefit knowing that the benefit was conferred by the claimant with a non-gratuitous intent and in circumstances where the recipient had the opportunity to reject the benefit. As such, free acceptance acts both as a ground of restitution and as a method of establishing the defendant’s enrichment. The tale of the window cleaner is used as an archetypical situation where it is unconscientiousness for the house owner to claim that: (a) the enrichment was not unjust; and (b) there was no enrichment. Birks said this story was a ‘clear and simple example’ of free acceptance. What is fascinating about the window cleaner story is that Birks’ initial intuition about this illustration is that the house owner had a duty to speak up to save the window cleaner from the unnecessary labour before he embarked on the cleaning job. Thus, the legal person in Birks’ world (at least in this context) does embody an ethic of care which necessitates him to intervene to save another from incurring unnecessary labour and expense. According to Birks, free acceptance ‘works because of the unconscientiousness of the recipient in not availing himself of the opportunity to save the intervener from the risk’.40 This narration of the window cleaner story has come under trenchant attack from numerous influential commentators. Authors like Burrows,41 Mead42 and Simester43 have all criticised Birks’ window-cleaning example. These scholars make two principal arguments against imposing liability on the house owner. First, a risk-taker should not be protected by the law. Burrows says that it is not clear why the law should allow a restitution claim since the window cleaner is a risk-taker, arguing that ‘there would be no injustice in … not paying a risk taker … Why should we now want to protect him against the very risk that he undertook?’44 Secondly, Mead and Simester point out that the construction of free acceptance is inconsistent with the general philosophy of the common law, ie that there is no general duty to rescue a third party. It is interesting that Birks, in his reply to Mead, said that the fact the law does not impose a duty to rescue is something which ‘does not engender pride’.45 Simester, in defence of Mead’s point, argues that the common law values an individual’s freedom to ‘live their own lives and pursue their own goals’ and that ‘this is why the systematic imposition of (criminal or civil) liability for failures to act is to be resisted’.46 The resultant position after these strong criticisms is the abandonment of the window cleaner example as a situation where an unjust enrichment claim is 40 P Birks, ‘In Defence of Free Acceptance’ in A Burrows (ed), Essays on the Law of Restitution (Oxford, Clarendon Press, 1991) 111. 41 A Burrows, ‘Free Acceptance and the Law of Restitution’ (1988) 104 Law Quarterly Review 576. 42 GH Mead, ‘Free Acceptance: Some Further Considerations’ (1989) 105 Law Quarterly Review 460. 43 A Simester, ‘Unjust Free Acceptance’ [1997] Lloyd’s Maritime & Commercial Law Quarterly 101, 118–20. 44 Burrows, above n 41, at 578. 45 Birks, above n 40, at 125. 46 Simester, above n 43, at 119.

466

Tang Hang Wu

justified. Birks now says that it was an error to present this story as ‘a clear and simple example’.47 He says it is not clear whether the window cleaner would have desisted in cleaning the window if the house owner spoke out; the window cleaner could be likened to a street busker at the Edinburgh festival who performs for the public. A member of the public could ostensibly listen to the performance and refuse to pay with a clear conscience. In his last book, Unjust Enrichment, Birks’ window cleaner story is now conspicuously absent and he says that free acceptance is a ‘result of muddled thinking’.48 Birks has now famously shifted from the ‘unjust factor approach’ to the absence of basis analysis.49 A major sub-theme of his analysis is that a claim ought to be denied if the claimant had taken the risk that the desired outcome would not be achieved.50 Birks gives the following as examples of risk-taking where restitution ought to be denied: (a) a busker who makes music available to the public;51 (b) a builder speculating that a contract is to come into existence begins to do work on site;52 and (c) a local authority who had not disclosed its intention to charge for sewerage services for many years and then makes a claim for the service. Birks has now shifted from his initial flirtation with imbuing his legal person with some degree of an ethic of care to one where the legal person must be sensitive to the risk involved when conferring a benefit on someone else. What then do we make of the legal person in the dominant discourse in unjust enrichment law? He is primarily a rugged individual who values his freedom. It is not entirely clear whether he forms relationships with others and how these relationships feature in the construction of his autonomy. The argument in this section is that a new conception of the autonomous legal person is required in analysing some areas of law, like undue influence and gifts. Another interesting aspect which arises from this study of the autonomous legal person in unjust enrichment law is the complex picture which emerges. It is erroneous to say, even in the dominant discourse of unjust enrichment, the autonomous legal person is entirely without concern for others. As Birks’ initial concern for the hapless window cleaner demonstrates, the legal person does display some traces of an ethic of care. This observation vindicates Robin West’s thesis that it is impossible to caricature a human being as either one who values or is wary of connectedness with other people. Human beings, as West argues, live in a fundamental contradiction between our fear of invasion and our need for intimacy.53 We are complex creatures who at various times ‘value intimacy, fear separation, dread invasion, and crave individuation’.54

47

Birks, above n 17, at 42. ibid. cf G Virgo, ‘Demolishing the Pyramid—the Presence of Basis and Risk-Taking in the Law of Unjust Enrichment’ ch 19 of this book. 50 Birks, above n 17, at 130 and 160. 51 ibid, at 130 52 ibid, at 144. 53 R West, ‘Jurisprudence and Gender’ (1988) 55 University of Chicago Law Review 1, 54–8. 54 ibid, at 61. 48 49

Storytelling in the Law of Unjust Enrichment

467

III. The Story of the Back Pocket and a One-Sided View of Autonomy In the first chapter of Unjust Enrichment entitled ‘A Core Case’ Birks narrates the following tale: You go shopping with a friend. As you are leaving a department store an assistant comes running up to tell you that he has accidentally given you change for £50 when you had in fact paid with a £20 note. He gave you £30 more than he owed.55

According to Birks, the shopkeeper’s carelessness is irrelevant; there is liability to make restitution: ‘so long as you still have the mistaken money, there is no answer to the shop’s demand to have it back’.56 As a narrative for purposes of teaching beginners, this story demonstrates Birks’ innate gift in presenting complex ideas with disarming clarity.57 This simple and elegant story provides a strong instinctive appeal for an unjust enrichment claim. If the shopper still had the extra £30, why should the shopper not give it back to the shopkeeper? There appears to be no injustice in ordering the shopper to make restitution. The intuitive power of this story seems to compel us to formulate a rule which allows a claim. Immediately after telling this story, Birks says ‘[i]t is evident … the strict liability from which we recoil is actually the only acceptable demand’ (emphasis added). However, it is suggested that this narrative is unrepresentative of the archetypical situation where a restitution claim is sought. This story is unique; central to this story are a few implicit features which may not be present in other claims for overpayment of money. First, the mistake is discovered almost immediately. In the shopper story, the mistake comes to light when the shop assistant runs after the shopper almost immediately. As such, the overpayment would not have a significant impact in the shopper’s life choices. Secondly, the overpayment involves a relatively trivial sum. It follows that the obligation to make restitution would not cause much hardship to the shopper. After all, the money is still in the shopper’s purse or wallet. And finally, the recipient’s change of position, if any, is always quantifiable in monetary terms. In the example, the change of position was whether the shopper would, in reliance on the receipt of the money, spend £30 on lunch rather than £5 on a snack. Steve Hedley very perceptively describes such a narrative as committing ‘the fallacy of the back pocket’, ie where it is assumed that the mistake comes to light almost immediately and the money is still in the payee’s back pocket.58

55

Birks, above n 17, at 6–7. ibid. 57 See A Pretto-Sakmann, ‘ “You Never Can Tell with Bees”: Good Advice from Pooh for Students of the Lex Aquilia’ in A Burrows and A Rodger (eds), Mapping the Law: Essays in Memory of Peter Birks (Oxford, Oxford University Press 2006) 475–7 for a short account of Birks’ teaching style. 58 S Hedley, Restitution: Its Division and Ordering (London, Sweet & Maxwell, 2001) 27–8. 56

468

Tang Hang Wu

Due to the elegant simplicity of the shopper story, this tale has the power to shape and reinforce the main structure of the law of unjust enrichment; the persuasive nature of this story seems to compel us to formulate liability in the following terms: once it is demonstrated that the payer was mistaken, the payee must make restitution unless there is a change of position. Carelessness by the payer is irrelevant. This formulation is no doubt influenced by the unique features of the so-called core case. Since the shopper suffered no harm (the money is still with her), it is only right that the shopper must return the money. But is the shopper story really a typical situation whereby an unjust enrichment claim might operate? Far from being typical, I would venture to say that the shopper scenario is an atypical situation, at least where litigation is pursued. Case law59 and newspaper reports60 tell us that a more common example might involve an institutional payer (usually a bank or pension fund) making an overpayment to a consumer. The mistake only comes to light after a few years and the payer asks for the money back. In this situation, the obligation to make restitution poses numerous challenges to the recipient which are not present in the shopper story. First, an unjust enrichment claim in this context must balance the interest of restoring the autonomy of the payer with that of not unduly impinging on the autonomy of the payee. Since a longer period has passed between the overpayment and discovery of the mistake, the money would, in all likelihood, have affected the payee’s life choices. To reverse the overpayment would frustrate the payee’s expectations and affect her autonomy. The autonomy aspect of the payee has hitherto not been fully developed by most restitution scholars who have concentrated on monetary change of position.61 Secondly, there is the question of difficulty of proof with regard to change of position, especially when the overpayment was made a few years ago. And finally, the overpayment can affect the payee in terms of legal costs and undue distress. These concerns, which are very real in the messy world of litigation, are absent in the shopper narrative and may mandate a rethink of the elegant simplicity of couching liability in terms of strict liability subject to monetary change of position. Each of these points will be developed below. Birks’ story obscures an important aspect of an unjust enrichment claim, namely that a person might have placed certain expectations on her financial position and embarked on certain choices based on this belief. These life decisions could be relatively minor (having a better meal, buying new items of clothing, redecorating) or major (having a baby, getting married, retiring). It should be

59

See, eg, Scottish Equitable plc v Derby [2001] 3 All ER 818 (CA). See, eg, R Colbey, ‘Please Give Us All That Cash Back: Lloyds TSB is Telling Scottish Widows Policyholders They Got Too Much’, The Guardian (London), Jobs and Money Pages, 16 June 2001 at 4; Gray, ‘Bank Recovers £1 m Bond Error’, The Scotsman, 21 June 2001, at 3; R Bennett, ‘Blair Apologises to Tax Credit Fiasco Families’, The Times (London), Home News, 25 June 2005, at 10; T Levene, ‘Their Mistake, But Who Pays the Price?’, The Guardian (London), Guardian Money Pages, 3 May 2008, at 8. 61 See, eg, J Palmer, ‘Chasing a Will-O’-The Wisp? Making Sense of Bad Faith and Wrongdoers in Change of Position’ [2005] Restitution Law Review 53; R Grantham and CEF Rickett, ‘A Normative Account of Defences to Restitutionary Liability’ (2008) 67 Cambridge Law Journal 92. 60

Storytelling in the Law of Unjust Enrichment

469

noted that the defendant’s assumption of her financial wealth would be generated by the mistaken payment by the payor and that the defendant is usually a passive recipient who did nothing in encouraging this mistaken payment. By allowing a restitution claim, the defendant’s past, present and future life choices might be disrupted. This disruption is not trivial and costless as portrayed in Birks’ narrative. Daily minor increases in the defendant’s expenditure can add up to quite a considerable figure. A counter-narrative to Birks’ story can be told to illustrate this point. Suppose a person of humble means is overpaid £5,000 by her pension fund. Six years later, the pension company says to her, ‘Oops, we made a mistake. Can we have our money back?’ Imagine the devastating disruption to the payee’s life if the claim is allowed. The payee in this case would have led her life for the past six years based on the expectation of her current wealth. Dagan correctly casts the payee’s interest in this respect as a form of autonomy interests arguing ‘[p]eople’s legitimate expectations regarding the availability of wealth in future inform their life plans … frustrated expectations are disruptive to individual autonomy’.62 This same point was made by Abbot CJ in Skyring v Greenwood,63 who said, ‘it is important that a man or woman should know the state of their account and the state of their finances, and a wise man or a wise woman looks at that statement before they spend money’. Birks’ narrative is flawed because it presents a onesided view of autonomy, ie concern only for the claimant’s autonomy while ignoring the defendant’s autonomy. Instead, Birks frames the defendant’s autonomy concerns as a form of ‘non-disenriching’ change of position.64 He says that, if a payee could establish ‘[t]his mistaken £100,000 persuaded us that we could at last start a family’ and the payee’s wife was already pregnant, this might serve as a defence.65 However, he concluded that valid examples of such non-disenriching change of position would be rare. It is suggested that Dagan’s characterisation of the problem as an issue of autonomy is more accurate than Birks’ analysis of the situation as a form of non-disenriching change of position. An unjust enrichment claim which has the potential to affect a defendant’s past, present and future life choices is surely more accurately described as a disruption to the defendant’s autonomy. The story about the careless shop owner has blinded us to the fact that concern for individual autonomy in an unjust enrichment claim is a two-sided inquiry (for the claimant and the defendant). Another objection to the shopper story is the implicit assumption that the ensuing monetary change of position by the defendant is easy to prove. In the story, the only change of position in reliance on the receipt of the money is that the shopper 62

Dagan, The Law and Ethics of Restitution, above n 15, at 48. Skyring v Greenwood (1825) 4 B & C 281. 64 Birks seemed to have thought quite hard about the point of ‘non disenriching’ change of position: see A Burrows, ‘Address given at the Memorial Service for Peter Birks on 20 November 2004 in the University Church of St Mary the Virgin Oxford’ in A Burrows and A Rodger (eds), Mapping the Law: Essays in Memory of Peter Birks (Oxford, Oxford University Press 2006) viii–ix. 65 Birks, above n 17, at 260–1. 63

470

Tang Hang Wu

decided to spend £30 on lunch instead of £5 on a snack. But in reality, monetary change of position is very hard to prove, especially after a long lapse of time. In Avon County Council v Howlett,66 the overpayment involved a sum of £1,007 made over a period of five years. Sheldon J found that the more the defendants had in their pocket, ‘the more they were able to spend—a better cut of meat, maybe, from time to time, or something extra from the grocer’.67 An unjust enrichment claim in such cases might operate unfairly against the defendant because the defendant would have to prove a change of position which may be extremely subtle and may happen over a long period of time. The final adverse impact on the defendant, not captured in Birks’ shopper narrative, is that an unjust enrichment claim may affect the defendant in terms of having to spend unnecessary legal costs in defending the claim and suffering undue stress and anxiety. While legal costs and stress are part and parcel of litigation, the point here is this: if the claimant had exercised a little more care in not making an overpayment, legal proceedings (and the stress and anxiety accompanying the law suit) would have been unnecessary. To contrast with Birks’ shopper story, I offer another counter-narrative which demonstrates how an unjust enrichment claim can have a devastating impact on the defendant. Imagine a 65-year-old pensioner who cashes in his pension fund. The pensioner was overpaid £170,000. He did not know of this overpayment. The pensioner uses over £40,000 to discharge his mortgage and close to £10,000 on modest improvements to his lifestyle. More than a year later, the pension fund asks for its money back. The pensioner fights the claim and is brought to near financial ruin because of the legal costs of close to £100,000.68 As a result of the legal battle, the pensioner’s health suffers. These facts are loosely based on the facts of Scottish Equitable plc v Derby.69 In this case, Simon Brown LJ said that while the court felt considerable sympathy for the defendant, an unjust enrichment claim could only be denied on legal principle. This observation is very interesting because it demonstrates that the dominant thinking in unjust enrichment law has influenced the law so much that the adverse impact of an unjust enrichment claim on the defendant is characterised as irrelevant or unprincipled. But why are these concerns unprincipled? If it is true that the defendant’s health suffered due to the anxiety caused by an unjust enrichment claim, why should this not be taken into account?70 What about the defendant’s expenditure of time and legal costs in unearthing bills and detecting changes in his spending pattern in order to 66

Avon County Council v Howlett [1981] IRLR 447 (QB). ibid, at 449–50. 68 According to a newspaper article, legal costs in Scottish Equitable plc v Derby [2001] 3 All ER 818 (CA) were estimated to be £100,000: see D McMillan, ‘£200,000 Error Leaves Man, 70, On the Brink of Ruin’, The Evening Standard (16 March 2001). Cf Some newspaper articles indicated that Derby was legally aided: C Hall, ‘The Cruel Cost of Pension Blunders: Fair Play Campaign New Errors Come to Light After Scandal of Man Brought to Edge of Poverty’, The Daily Mail (28 November 1999). 69 Scottish Equitable plc v Derby [2001] 3 All ER 818 (CA). 70 Cf Mitchell interprets this decision as not completely ruling out the relevancy of the defendant’s ill health if it can be proved this was a result of the overpayment. See C Mitchell, ‘Change of Position: The Developing Law’ [2005] Lloyd’s Maritime and Commercial Law Quarterly 169. 67

Storytelling in the Law of Unjust Enrichment

471

defend an unjust enrichment claim? If the defendant has to pay all the legal costs, the net result would be that the defendant is in a poorer financial situation than before the overpayment was made. All this trouble could have been avoided if the pension fund had exercised a little more care in not making the extra payment. Surely, it is not too much to ask this of large institutions making payments to their customers. I have argued elsewhere,71 that all these factors prompt a rethink of the mantra found in case law and leading academic literature that carelessness on the payor’s part is always irrelevant. In the context of a modest payment from an institutional payor to a consumer involving a mistake which is discovered only after a long period of time, I argue that a restitutionary claim ought to be denied, as the consumer should be presumed to have changed her position.

IV. Mapping the Law: Legal Taxonomy and Comparative Law Birks uses the mapping metaphor heavily in his work.72 In Unjust Enrichment, there is actually a chapter called ‘Three Maps’.73 The reference to mapping is closely linked to Birks’ work on legal classification and his frequent reference to Roman and German law.

A. Legal Classification and Animal Taxonomy There are a couple of recurrent themes related to the mapping metaphor in Birks’ scholarship, namely, the importance of legal taxonomy, and the use of comparative law (especially Roman and German law) in developing the law. It could be argued that these are simply techniques used in analytical jurisprudence and not really a form of legal narrative. However, I would argue that on closer scrutiny there are several implicit stories embedded here. Consider Birks’ work on legal taxonomy. Taxonomy, according to Birks, ‘changes nothing, but it promotes understanding. Without it there is only a chaos of unsorted information … [ie] a heap of good learning’.74 While it is hard to disagree with this stated goal, what 71 HW Tang, ‘The Role of Negligence and Non-Financial Detriment in the Law of Unjust Enrichment’ [2006] Restitution Law Review 55. 72 The book of essays in his memory also refers to the cartography metaphor: Burrows and Rodger, above n 57. 73 Birks, above n 17, ch 2. Cf H Dagan, ‘Legal Realism and the Taxonomy of Private Law’ in C Rickett and R Grandtham (eds), Structure and Justification in Private Law: Essays for Peter Birks (Oxford, Hart Publishing, 2008) 154–5, who disputes that the law can be mapped. Dagan argues that the law is dynamic in its changing nature and the cartography metaphor is fallacious because it assumes stability in the content of the law. 74 P Birks, ‘Equity in the Modern Law: An Exercise in Taxonomy’ [1996] 26 Western Australian Law Review 2, 3–4. Cf C Webb, ‘Treating like Cases Alike: Principle and Classification in Private Law’ ch 9 of this book.

472

Tang Hang Wu

is controversial is the constant invocation of animal taxonomy in this exercise of legal classification. For example, Birks writes in Unjust Enrichment:75 To require a good map of the law is by metaphor to insist on sound classification. Classifications answer questions. Where do animals live? Some are aquatic, some terrestrial, some amphibious, others (like tapeworms) live in other habitats. What do animals eat? Some are omnivorous, some carnivorous, some herbivorous, and others eat other things. When different classifications are combined, a complex hierarchy results, each level dividing according to its own criterion. The hierarchical combination can be undone. That is to say, one can revert to a series of single-question classifications.76

The aim of this taxonomical project, trying to create order out of chaos, is certainly laudable and cannot be faulted.77 However, I would suggest that there is an implicit, and, ultimately, controversial story in the manner in which Birks engages in legal taxonomy. The story is this: legal classification is a dispassionate and value-neutral exercise where the legal scholar, like an animal taxonomist, observes in a detached manner the basic features of the law and sorts it into the most accurate and coherent of categories. Just as vertebrates can be grouped into categories such as mammals, birds, reptiles, amphibians and others, the entire private law, according to Birks, can be classified under the following headings: consent, wrongs, unjust enrichment and others.78 In Birks’ schema, these categories are causative events which trigger certain legal responses.79 So contracts are slotted under consent and torts under the umbrella of wrongs. The story of the dispassionate legal taxonomist quietly working in the library and sorting out neat categories out of chaos is subject to challenge. First, Birks’ classification scheme does not appear to tolerate overlaps. This is no doubt influenced by the analogy with animal taxonomy. It is hard to imagine an animal being, let’s say, both an amphibian and a bird. However, is such a rigid classification scheme, which does not tolerate overlapping categories helpful? I would suggest that an inflexible system of classification might prove to be an inaccurate representation of the law. Dagan correctly argues that the difficulty in classification is that the law is messy and may raise overlapping concerns.80 Cross-fertilisation occurs in many areas of private law and a classificatory system which insists on pristine categories might end up hampering the development of the law. Dagan also points out the difficulties with certain doctrines such as product liability law, which may be hard to pigeon-hole as either contract or tort.

75 See, eg, P Birks, ‘Annual Miegunyah Lecture: Equity, Conscience, and Unjust Enrichment (1999) 23 Melbourne University Law Review 1, 10; and Birks, ibid, at 6. 76 Birks, above n 17, at 20. A similar point made in his ‘Property, Unjust Enrichment, And Tracing’ (2001) 54 Current Legal Problems 231, 238. 77 See M McInnes, ‘Taxonomic Lessons for the Supreme Court of Canada’ in Rickett and Grantham, above n 73, at 77 for a defence of the utility of Birks’ taxonomy. 78 See, eg, P Birks, ‘The Law of Unjust Enrichment: A Millennial Resolution’ [1999] Singapore Journal of Legal Studies 318. 79 Birks, above n 17, ch 2. 80 See Dagan, The Law and Ethics of Restitution, above n 15, at 34; and Dagan, ‘Just and Unjust Enrichments’ ch 17.

Storytelling in the Law of Unjust Enrichment

473

Secondly, as Geoffrey Samuel and Steve Hedley81 have argued, there is a huge difference between the study of law and zoology. As Samuel asks rhetorically, ‘[c]an legal science … realistically and profitably be put side-by-side with natural sciences or is it like trying to compare, say, astrology with astrophysics?’82 With regard to zoology, we are dealing with ‘specifically identifiable objects which can be said to exist independently of the science’.83 These specimens in question can be subject to objective empirical observation and inquiry. Take the animal that Birks’ frequently mentions in his work—the platypus.84 It is an empirical fact that: (a) such an animal exists; (b) it has body hair; (c) it has a duck-bill; and (d) it is warm-blooded. The question then becomes: How do we go about classifying this animal? The classification scheme and the object to be classified are distinct. But can we say the same thing about the law? I suggest that we cannot. Legal concepts are never static and there is a relationship between any classification scheme of the law and the development of the law. As Samuel argues, the ‘fundamental danger is to fail to appreciate that taxonomy is about the construction of systems and systems have their own function’.85 Samuel warns that such a taxonomical scheme may go ‘beyond mere description’ and be ‘capable of creating its own reference points’.86 Therefore, it is important to always keep in mind that any taxonomical scheme is not immutable. Samuel is correct to argue that Birks’ taxonomy could end up setting the agenda and the reference points. This could then lead to a situation where arguments outside these reference points are deemed to be wrong. In this sense, although the classification is presented as scientific and neutral, it is not value-neutral. For example, by placing contracts under the umbrella of consent, Birks is making an implicit preference for the ‘will theory’ of contract law. Such an approach might have a doctrinal impact in the development of the law of contract and unjust enrichment.87 While the ‘will theory’ is one of the defensible analytical theories of the law of contracts, the point is this: there are other plausible contract theories. By presenting the categories of consent, wrongs, and unjust enrichment as somehow incontestable, other contract theories are necessarily excluded without further debate. The ‘will theory’ is then implicitly held out as the ‘meta theory’ that trumps all else. As such, the taxonomical scheme, far from being value-neutral, ends up setting the agenda and framework for discussion and analysis.

81 S Hedley, ‘The Taxonomic Approach to Restitution’ in A Hudson (ed), New Perspectives on Property Law, Obligations and Restitution (London, Cavendish, 2003). 82 G Samuel, ‘Can Gaius Really Be Compared to Darwin?’ (2000) 49 International Comparative Law Quarterly 297, 297. 83 ibid. cf D Sheehan, ‘Implied Contract and the Taxonomy of Unjust Enrichment’ in P Giliker (ed), Re-examining Contract and Unjust Enrichment (Dordrecht, Martinus Nijhoff Publishers, 2007) 185 for a defence of Birks’ taxonomy. 84 See, eg, Birks, ‘Annual Miegunyah Lecture’, above n 75, at 23. 85 Samuel, above n 82, at 319. 86 ibid. This point is also made by Dagan, above n 73, at 155–6. 87 One example is how the law deals with settlement contracts as a defence to an unjust enrichment claim: see HW Tang, ‘A Settlement Agreement as a Defence to An Unjust Enrichment Claim’ [2008] Restitution Law Review 63.

474

Tang Hang Wu

Another illustration of the influence of an inflexible classification scheme on the way we view and analyse the law can be found in the debate between Birks and Grantham/Rickett/Virgo on the proper place of property within private law. In an essay tellingly entitled, Property and Unjust Enrichment: Categorical Truths, Birks says ‘the law of property is concerned with proprietary rights and must look for the events that bring those rights into being. It is a category formed upon a species of response or effect’.88 As such, Birks argues it is a categorical error to argue that restitution may respond to a vindication of a property right. Interestingly, Birks presents this proposition as an incontestable truth because he proceeds on a taxonomical scheme of private law being divided into events (consent, wrongs, unjust enrichment and other events) and responses.89 This classification method inevitably leads Birks to the conclusion that the concept of property is inert and needs to be given ‘teeth’ by the law of obligations. But the fallacy of this analysis is that it implicitly assumes that the taxonomical scheme is somehow pre-ordained. Virgo90 and Grantham and Rickett91 have persuasively pointed out that a competing rationalisation of the law may be advanced, namely, an interference with a pre-existing property right may be characterised as a legally relevant event which triggers a restitution claim. Birks’ analysis, by implication, downplays the significance of the law of property as compared to the law of obligations. In summary, it is important to realise that an exercise in classification of the law might be important for purposes of teaching and organisation; however, we must guard against buying into the implicit story that the taxonomy of the law, like classification of animals, does not change the subject-matter under observation. As demonstrated above, legal taxonomy can influence the content and perspective of the law in a significant way.

B. Roman Law and Comparative Law Anyone with a brief acquaintance with Peter Birks’ scholarship would be struck by the frequent reference to Roman and German law. Gaius, Birks proclaims, is the Darwin of the law.92 In fact, when Birks re-aligned the law of unjust enrichment from the ‘unjust factors’ approach to a single ground, ie absence of basis, he gave two reasons for this significant change namely: (a) the force of precedent;93 and 88 P Birks, ‘Property and Unjust Enrichment: Categorical Truths’ [1997] New Zealand Law Review 623, 631. 89 Explained in Birks, above n 17, ch 2. 90 See, eg, Virgo, above n 15, at 11–7. 91 See, eg, RB Grantham and CEF Rickett, ‘Property and Unjust Enrichment: Categorical Truths or Unnecessary Complexity’ [1997] New Zealand Law Review 668; RB Grantham and CEF Rickett, ‘Property Rights as a Legally Significant Event’ (2003) 62 Cambridge Law Journal 717. 92 P Birks, ‘Definition and Division: A Meditation on Institutes 3.13’ in P Birks (ed), The Classification of Obligations (Oxford, Clarendon Press, 1997) 1–2. 93 cf R Stevens, ‘Absence of Basis Accepted: Unjust Factors Rejected (chs 5–6)’ [2004] Restitution Law Review 270, who criticises Birks’ approach as being inconsistent with the House of Lords decision in Kleinwort Benson v Lincoln City Council [1999] 2 AC 349 (HL).

Storytelling in the Law of Unjust Enrichment

475

(b) Sonja Meier’s (a German jurist) arguments from a comparative perspective.94 Therefore, it is interesting to delve into Birks’ methodology on comparative and Roman law, which obviously had such a profound influence on his thinking. Birks is quite careful to say that his goal in using comparative law is comparison and not legal transplantation. Comparative law, he says, is used in this context to deepen analysis and accelerate understanding.95 What does it mean that comparative and Roman law is used merely to deepen analysis and accelerate understanding? It seems Birks is invoking a story of the survival of the fittest and applying it in the legal context. The argument goes as follows: the Roman legal taxonomy has survived all these years because it is ‘better’ and tested by a long period of use. Consider the riposte Birks gives to the critics of his taxonomy; he charges that sceptics must argue against ‘the entire experience of the Western legal tradition as it has developed over the last two thousand years’.96 Later on, he says: ‘[t]hose who propose the new taxonomy will have to demonstrate exactly how the results of that long experience are defective and how the new map will eliminate those defects’.97 In the conclusion section of this essay, Birks again says that sceptics ‘have to show in the teeth of the experience of the entire Western legal tradition there is a better way of organizing the law which dispenses with unjust enrichment’.98 Birks’ insistence on adopting the classification based on Roman law is premised on the assumption that such a scheme has been tried and tested by the long passage of time. As such, unless shown otherwise, the Roman law taxonomy is the ‘best’ available manner of classifying the law. This thinking can be gleaned from Birks’ argument that ‘[i]t is in the nature of competing classifications, like all competing hypotheses, to fight to the death’. Is Birks’ story of the ‘survival of the fittest’ an apt analogy in this context? I suggest that it is not. First, with the nature of classification, it is very difficult to demonstrate definitively that a legal taxonomy is false unless the scheme is fundamentally flawed. As argued above, the construct of the law is not an empirical reality which can be tested objectively against a taxonomical structure. Secondly, the story that Roman law thinking survives in the Western legal tradition because it is ‘better’ is a controversial hypothesis. Alan Watson, who has studied legal transplants, effectively debunks this myth. According to Watson, there are a few reasons why Roman law is extensively borrowed in the Western legal world. He argues: Law is treated [by lawyers] as existing in its own right; it is being in conformity with lawness that makes law law. Hence, first, the means of creating law, the sources of law, come to be regarded as a given, almost as something sacrosanct … Secondly, law has to 94 See, eg S Meier, ‘Unjust Factors and Legal Grounds’ in D Johnston and R Zimmermann (eds), Unjustified Enrichment: Key Issues in Comparative Perspective, (Cambridge, Cambridge University Press, 2002); and Birks, above n 17, at 44–5. 95 P Birks, ‘Comparative Unjust Enrichment’ in P Birks and A Pretto (eds), Themes in Comparative Law: In Honour of Bernard Rudden (Oxford, Oxford University Press 2002) 151. 96 ibid, at 144. 97 ibid. 98 ibid, at 151.

476

Tang Hang Wu

be justified in its own terms; hence authority has to be sought and found. That authority (in some form, which may be perverted) must already exist; hence law is typically backward-looking. These two features make law inherently conservative.99

Roman law concepts thrive not because they are ‘better’ legal constructs. Lawyers constantly borrow from Roman law because lawyers are generally ‘creatures of habit’ and ‘in altering the law they seek … either to play down the extent of the change, or to borrow a rule from some foreign legal system with great prestige and authority’.100 Watson’s insight is valuable because it demonstrates a few important propositions. First, Birks’ ‘survival of the fittest’ assumption is misplaced. Legal concepts do not ‘battle to the death’ for dominance. Secondly, there is a cultural reason for the prevalence of Roman law concepts in the Western legal world, ie the inherent nature of how lawyers work and think. While this cultural factor provides an explanation on a descriptive level for the prevalence of Roman law concepts in Europe, it does not supply a normative justification for the adoption of Roman law taxonomy in English law.

V. Conclusion Birks is such an important figure in the law of unjust enrichment that the stories that he weaves into his work assume considerable importance. I have tried to show that his stories are not mere narratives used as a form of exposition of the law; rather, these narratives are utilised to advocate a certain position in the law. Since Birks is extremely influential and his work has been extensively relied on by the judiciary and other academics, it is vital that we critically examine these stories. Otherwise, there is the danger of the law of unjust enrichment becoming entirely self-referential. In this chapter, I have evaluated: (a) the story of the autonomous legal person; (b) the core story of an unjust enrichment claim; (c) the story of legal taxonomy being analogous to animal taxonomy; and (d) the story of the ‘survival of the fittest’ in law. No doubt there are many more stories in the dominant discourse of unjust enrichment, which remain unexplored. In this chapter, I contend that the stories that I have examined affect: (i) the way we think of close human relationships; (ii) certain doctrines, like restitution for mistaken gifts and the ‘change of position’ defence; and (iii) jurisprudential methods. The argument in this chapter is that we must guard against making these the ‘stock stories’ that form the foundation of many of the conventions and constructs of the doctrines of the law of unjust enrichment. Otherwise, equally valid concerns will be excluded as being either unprincipled or irrelevant. As demonstrated in this chapter, some of these narratives are controversial and equally valid counter-narratives could be told. 99 A Watson, The Evolution of Law (Oxford, Blackwell Press, 1985) 119. Watson’s transplant thesis is lucidly summarised in W Ewald, ‘Comparative Jurisprudence II: The Logic of Legal Transplants’ (1995) 43 American Journal of Comparative Law 489. 100 Ewald, ibid, at 489 and 499.

20 Demolishing the Pyramid—the Presence of Basis and Risk-Taking in the Law of Unjust Enrichment GRAHAM VIRGO

I. Introduction The orthodox analysis of the unjust enrichment principle adopts a linear formula as follows: enrichment + proof that this was at the claimant’s expense + a recognised ground of restitution + no relevant defence = award of a restitutionary remedy.1

According to this orthodox analysis, specific grounds of restitution need to be positively identified, such as mistake, duress or total failure of consideration.2 These grounds of restitution are recognised3 because they establish that the claimant did not intend the defendant to receive a benefit in the circumstances, either because the claimant never had an intent to benefit the defendant or the intent was vitiated or qualified in some way. Other grounds of restitution are triggered by the defendant’s conduct4 or simply on policy grounds.5 Birks advocated a radical re-interpretation of these grounds of restitution in his book Unjust Enrichment.6 He argued that the orthodox analysis which required the identification of a specific ground of restitution from a recognised list was

1 Banque Financière de la Cité v Parc (Battersea) Ltd [1999] 1 AC 221 (HL) 227 (Lord Steyn) and 234 (Lord Hoffmann), who added that a further question was whether there were any reasons of policy for denying a remedy. 2 As expounded by Lord Mansfield in Moses v Macferlan (1760) 2 Burr 1005, 1012. 3 See P Birks, ‘No Consideration: Restitution after Void Contracts’ (1993) 23 University of Western Australia Law Review 195, 206. 4 Such as the grounds of restitution founded on the defendant’s exploitation of the claimant. 5 Such as the ground of restitution which enables overpayments to be recovered from a public authority (Woolwich Equitable Building Society v Inland Revenue Commrs [1993] AC 70 (HL)), or the ground of necessitous intervention. 6 P Birks, Unjust Enrichment (Oxford, Oxford University Press, 2003); now in its second edition (2005).

478

Graham Virgo

unreliable, cumbersome and could produce the wrong results.7 He concluded that it should be rejected in favour of a civilian approach8 whereby there should be one single ground of restitution, which is that there was no explanatory basis for the defendant’s receipt of an enrichment.9 If there was no such basis, the receipt should be considered to be unjust. There would be a basis for receipt if, for example, the enrichment was transferred to discharge a contractual or other obligation or was a gift.10 Whether this absence of basis approach should be adopted in English law remains a matter of controversy. That it does not yet reflect the state of English law is clear, even though Birks considered it to have been accepted implicitly by cases which had recognised that money paid pursuant to a void ‘interest rate swap’ transaction could be recovered simply because the transaction was void.11 In Deutsche Morgan Grenfell Group plc v Inland Revenue Commissioners12 all the judges in the House of Lords analysed the unjust enrichment principle with reference to the need to establish specific grounds of restitution, and none adopted the absence of basis analysis in respect of a claim relating to premature payment of tax in breach of EC law. However, Lord Walker did acknowledge the persuasiveness of Birks’ approach and recognised that it might be appropriate to consider its reception into English law at a future date.13 The legitimacy of adopting absence of basis in English law consequently remains a live issue.

7

ibid, at 160. Building on the work of Sonja Meier: ‘Restitution after Executed Void Contracts’ in P Birks and F Rose (eds), Lessons from the Swaps Litigation (London, Mansfield Press, 2000); ‘Unjust Factors and Legal Grounds’ in D Johnston and R Zimmermann (eds), Unjustified Enrichment: Key Issues in Comparative Perspective (Cambridge, Cambridge University Press, 2002). See also R Zimmermann, ‘Unjustified Enrichment: the Modern Civilian Approach’ (1995) 15 Oxford Journal of Legal Studies 403; S Meier and R Zimmermann, ‘Judicial Development of the Law, Error Iuris and the Law of Unjust Enrichment’ (1999) 115 Law Quarterly Review 556; J du Plessis, ‘Toward a Rational Structure of Liability for Unjustified Enrichment: Thoughts From Two Mixed Jurisdictions’ (2005) 122 South African Law Journal 142; F Giglio, ‘A Systematic Approach to “Unjust” and “Unjustified” Enrichment’ (2003) 23 Oxford Journal of Legal Studies 455; HLE Verhagen, ‘Absence of Basis and Unjustified Enrichment in Dutch Law’ [2004] Restitution Law Review 132. 9 Birks, above n 6, chs 5 and 6. This approach had been specifically rejected by Lord Goff in Woolwich Equitable Building Society v IRC [1993] AC 70 (HL) 172. cf Canada where ‘no juristic basis’ has been recognised by the Supreme Court: Pettkus v Becker [1980] 2 SCR 834; Peel v Canada (1993) 98 DLR (4th) 140. See now Garland v Consumers’ Gas Distributors Inc [2004] 1 SCR 629; Pacific National Investments v Victoria [2004] SCC 75. 10 Birks, above n 6, at 104. 11 See especially Guinness, Mahon and Co Ltd v Kensington and Chelsea Royal London Borough Council [1999] QB 215 (CA). 12 Deutsche Morgan Grenfell Group plc v Inland Revenue Commissioners [2006] UKHL 49, [2007] 1 AC 558. 13 Deutsche Morgan Grenfell Group plc v IRC [2006] UKHL 49 [158] (Lord Walker). Lord Hoffmann also made reference to this absence of basis analysis: at [28]. See also Lord Hope in Sempra Metals Ltd v IRC [2007] UKHL 34, [2008] 1 AC 561 [25], who did refer to ‘no legal ground’ as the ground of restitution, although he had earlier accepted that the ground was mistake of law. Lord Mance left the point open: at [192]. 8

Demolishing the Pyramid—Unjust Enrichment

479

Clearly the law of unjust enrichment is at a crossroads:14 should we stick with the traditional analysis of specific grounds of restitution or move to a single over-arching ground? This chapter will engage with that question but, in doing so, it will be argued that the choice of routes is not as significant as might first be thought. Although Birks asserted that it was not possible to mix the two methods,15 the orthodox analysis of unjust enrichment already incorporates absence of basis reasoning. The real weakness of the unjust enrichment principle, at least as expounded by commentators including myself, is the failure to identify clearly the bars to unjust enrichment claims.16 But those bars already exist within the cases. If they are expounded clearly it will be seen that they do much of the work envisaged by Birks’ absence of basis principle and they do that work more coherently and effectively. Two key bars to unjust enrichment claims need to be identified expressly. The first is where there is a legally-effective basis for the receipt of an enrichment received by the defendant, which can usefully be called the ‘presence of basis’ bar. The second is where, even though there is no basis for the receipt, the claimant can be considered to have taken the risk that there was or would be no basis. Both bars need to be formally recognised because, where there is either a basis for the receipt of an enrichment or the claimant took the risk that there was no basis, it is not appropriate for the defendant to be required to make restitution.

II. What is a Basis? The fundamental feature of Birks’ new approach is the determination of whether or not there is a basis for the receipt of an enrichment. It is consequently important to be clear as to what he meant by a basis for these purposes. In Germany the test is whether the defendant acquires something at the claimant’s expense ‘without legal ground’.17 Birks said of the civilian approach: [E]nrichments are received with the purpose of discharging an obligation or, if without an obligation, to achieve some other objective as for instance the making of a gift, the satisfaction of a condition, or coming into being of a contract. Those outcomes succeeding, the enrichment is sufficiently explained. An enrichment which turns out to have no such explanation is inexplicable and cannot be retained. The recipient is not entitled to it.18

14 As Krebs had previously recognised: T Krebs, Restitution at the Crossroads: A Comparative Study (London, Cavendish, 2001). 15 Birks, above n 6, at 113. cf Burrows who considers that the absence of basis principle is a useful cross-check on the operation of the unjust factors to determine in novel cases whether the enrichment was unjust: A Burrows and A Rodgers (eds), Mapping the Law: Essays in Memory of Peter Birks (Oxford, Oxford University Press, 2006) 34. 16 As opposed to defences where a great deal of work has been done in exploring their boundaries. 17 Section 812(1) German Civil Code. 18 Birks, above n 6, at 102.

480

Graham Virgo

A. No Explanatory Basis Birks explored the meaning of ‘basis’ in some detail in Unjust Enrichment. He drew a key distinction19 between situations where the claimant participated in the transfer of the enrichment and where he failed to participate in the sense that the claimant had no control over the transfer. Where the claimant did participate, two distinct explanatory bases for the transfer could be identified. First, where the claimant transferred the enrichment to discharge an obligation, such as a contractual or other liability, including a court judgment.20 If the obligation was valid, an explanatory basis could be identified. If there was no valid obligation, there was no explanatory basis for the enrichment. Secondly, where the claimant voluntarily transferred the enrichment to achieve a purpose, such as to make a contract, constitute a trust or as a gift. If this purpose were not fulfilled, there would be an absence of basis for the transfer. So, for example, if the claimant transferred a benefit with a view to making a contract, but no contract was forthcoming, there will have been no explanatory basis for the transfer.21 However, it is in this context in particular that the claimant may well be considered to have taken the risk that the purpose will not be achieved, and this may be sufficient to defeat a claim for restitution.22 Alternatively, where the enrichment has been transferred involuntarily,23 such as where it has been stolen or the claimant is powerless to prevent the transfer, there will have been no explanatory basis for the transfer, save where there was legal authority for acquisition by the defendant.

B. Subjective Assessment of Basis Birks’ analysis of basis in the context of participatory transfers focuses on a subjective interpretation of the basis, in the sense that it is dependent on the claimant’s perception of whether the enrichment was transferred for the purpose of discharging an obligation or achieving some other objective.24 This is especially well illustrated by Birks’25 analysis of the difficult decision in Kerrison v Glyn, Mills, Currie and Co,26 where the House of Lords allowed a claimant to recover money that had been paid in anticipation of a liability arising in the future, when the liability never materialised. The House of Lords concluded that the claimant had paid the money by mistake, even though there was no existing mistake but only a mis-prediction as to what might happen in the future. However, Birks justified the award of the restitutionary remedy on the ground that the claimant’s 19 20 21 22 23 24 25 26

ibid, at 129. ibid, at 140. Birks considered that the vast majority of cases would be of this type: ibid, at 130. See, eg, Chillingworth v Esche [1924] 1 Ch 97 (CA). Attorney-General of Hong Kong v Humphreys Estate (Queen’s Gardens) Ltd [1987] AC 114 (PC). Birks, above n 6, at 154. ibid, at 104. ibid, at 153. Kerrison v Glyn, Mills, Currie & Co (1911) 81 LJKB 465 (HL).

Demolishing the Pyramid—Unjust Enrichment

481

purpose had been to discharge a future liability and, if that liability never arose, it followed that there was a failure of basis. In such a case the identification of the basis depends on the claimant’s perception as to what might happen in the future, although Birks’ analysis does raise concerns about the appropriateness of awarding a restitutionary remedy where the claimant can be considered to have taken the risk of the basis failing.27 In particular, his recognition of the relevance of the payer’s belief that a payment was conditional on there being counter-performance would mean that there was a failure of basis whenever the condition failed, even if there was no enforceable right to receive the counter-performance. This makes the notion of basis a factual test which becomes much more difficult to define and which could potentially make all gifts fail, depending on the nature of the condition which the claimant has attached to the payment.

C. Partial Failure of Basis Birks was adamant that there is no such thing as a partial failure of basis.28 It follows that the basis either fails completely or not at all. However, his analysis of the absence of basis principle does suggest that he was willing to contemplate a partial failure. For example, where a claimant has paid the defendant £50 to discharge a £20 liability then, assuming that the additional payment was not intended as a gift, presumably there was a failure of basis as regards the £30 excess.29 The full payment cannot be recovered, since there was no failure of basis as regards £20, but there should be severance as regards the remaining £30 which could be recovered. Birks’ own analysis of the difficult decision of the High Court of Australia in Roxborough v Rothmans of Pall Mall Australia Ltd30 is consistent with this analysis. In that case the claimants had purchased cigarettes from the defendant, the price for which included a licence fee which the defendant was to pay to the state tobacco authority. Before the licence fee was paid to the state, the statute which authorised it was declared invalid. The claimants recovered the licence fee element of the purchase price. Birks justified this result on the ground that it involved the failure of a particular obligation within a valid contract.31 This apportionment of bases within a contract is certainly consistent with the notion that a partial failure of basis should be recognised.

D. Timing Birks’ analysis of absence of basis indicates that this should be determined either at the time of the initial receipt of the enrichment or subsequently, depending 27 28 29 30 31

See Section VI, below for discussion of this issue. Birks, above n 6, at 120. As Birks appears to acknowledge, ibid, at 132. Roxborough v Rothmans of Pall Mall Australia Ltd [2001] HCA 68, (2001) 208 CLR 516. Birks, above n 6, at 123.

482

Graham Virgo

on the type of absence of basis which is being considered. So, for example, if the enrichment is transferred in purported discharge of an obligation which does not exist, the absence of basis will exist from the outset. But the failure of basis may arise subsequently. For example, if money is paid in respect of a contract which the defendant subsequently breaches, the basis for the payment will have failed at the time of breach, even though the claimant has not yet discharged it.32 Further, if the enrichment is transferred in the belief that a basis for it will exist, but it is not forthcoming, the basis will fail. It is then more difficult to pinpoint precisely when the basis will have failed, although presumably this would ultimately be a question of fact.

E. Risk-taking For Birks risk-taking is an implicit part of the absence of basis question;33 for where there is an identifiable absence of basis it is still necessary to determine whether the claimant can be considered to have borne the risk of the basis being absent. So, where money has been paid in respect of a void contract, absence of basis is readily established, but the claimant will not be able to obtain restitution if he was aware of the possibility of the basis being absent but transferred the benefit nonetheless, because the claimant then bore the risk of invalidity.

III. Building the Pyramid34 Although Birks described the absence of basis principle as having a ‘surgical simplicity’35 it is neither as simple nor as radical as might first be thought. This is because he accepted that many of the traditional grounds of restitution remain relevant when identifying whether there is an absence of basis. Birks adopted a pyramidal structure to his analysis of the ‘unjust factor’ part of the unjust enrichment principle.36 At the base of the pyramid are the existing grounds of restitution which identify whether the claimant intended the defendant to be enriched,37 and other considerations which identify that the transfer of the enrichment was invalid, such as the invalidity of a contract. Establishing any of these grounds or considerations will determine whether there was an absence of basis for the transfer, which is at the second level of the pyramid, and, once this absence of basis has been identified, it can be concluded that the enrichment was unjust; which is

32 33 34 35 36 37

ibid, at 126. ibid, at 133. See further TA Baloch, ‘The Unjust Enrichment Pyramid’ (2007) Law Quarterly Review 636. Birks, above n 6, at 160. ibid, at 104. But not ‘policy-motivated’ grounds of restitution.

Demolishing the Pyramid—Unjust Enrichment

483

at the top of the pyramid.38 However, Birks’ shift of focus from positive grounds of restitution to absence of basis remains significant, since if the claimant can establish the absence of a basis for the transfer of a benefit to the defendant, the unjust element of unjust enrichment has been established and the burden will shift to the defendant to show that there is in fact a valid basis for the transfer. It follows from Birks’ new approach that much of the work which is normally done within the law of unjust enrichment would be done outside of it.39 Often there will need to be a prior determination of whether there is a legal basis for the transfer of the enrichment—within the law of contract or gift for example—and only once it has been decided that there is no such basis would the law of unjust enrichment be engaged. It follows that the law of unjust enrichment would become a much smaller subject, as reflected by the relatively slim volume of Birks’ treatment of unjust enrichment. This might be considered to be a good thing, but it simply means that the bulk of the law which is presently considered within the law of unjust enrichment would need to be considered in other disparate areas of the law.

IV. Demolishing the Pyramid In the vast majority of cases it will not matter whether the unjust enrichment claim is analysed as involving positive grounds of restitution or absence of basis, as reflected by the fact that Birks was able to reach the same result in many of the cases which relied on these specific grounds, by resorting to his absence of basis analysis. Indeed, in Kelly v Solari,40 one of the leading cases on restitution for mistake, Parke B adopted a dual analysis. He concluded that insurers who had mistakenly paid money under an insurance policy which had lapsed were able to obtain restitution both because the defendant was not entitled to receive the money and because the insurers’ mistaken belief as to the liability to pay meant that they did not intend the defendant to receive the money. Despite the consistency of result which would be achieved in most cases regardless of whether the positive grounds or absence of basis approach is adopted, Birks was adamant that a choice had to be made between them, both as a matter of principle and also because it is clear that sometimes different results will be achieved depending on which approach is adopted. It is therefore necessary to decide whether the absence of basis approach is preferable to the specific recognition

38

Birks, above n 6, at 104. RB Grantham, ‘Absence of Juristic Reason in the Supreme Court of Canada’ (2005) 13 Restitution Law Review 102, 105; see also A Burrows, ‘Absence of Basis: the New Birksian Scheme’ in Burrow and Rodgers, above n 15, at 36. 40 Kelly v Solari (1841) 9 M & W 45, 59 (Parke B). See Birks, above n 6, at 102. 39

484

Graham Virgo

of grounds of restitution. The writer’s preference is most definitely for the latter for a number of reasons.41

A. State of the Authorities Birks felt compelled to reconstruct his treatment of the grounds of restitution because he considered that recent decisions of the English courts meant that the absence of basis analysis was the only way of satisfactorily explaining the law. This was primarily because of the tranche of cases which he interpreted as recognising that money paid pursuant to void ‘interest rate swap’ transactions could be recovered simply by virtue of the fact that the transaction was void.42 But this approach has never been recognised by the courts, and despite Birks’ assertions to the contrary, the ‘swaps’ cases do not compel its recognition.43 Indeed, the orthodox approach requiring positive grounds of restitution to be established by the claimant has been specifically affirmed by the courts.44 Most importantly this was affirmed by the House of Lords in Deutsche Morgan Grenfell Group plc v IRC,45 where the identification of a mistake was vital to the unjust enrichment claim to enable the claimant to benefit from the extended limitation period under the Limitation Act 1980.

B. Different Results The biggest problem with the absence of basis theory is that it simplifies the analysis of the unjust factors in such a way that some of the subtleties and nuances of the existing law of unjust enrichment would be lost. For example, English law recognises a distinction between mistakes as to existing facts and mis-predictions as to what might happen in the future. Mis-predictions will not ground a restitutionary claim, because the claimant who has made a mis-prediction bears the risk as to future events occurring. This distinction is lost under the absence of basis approach, as shown by Birks’ conclusion that restitution should have been awarded

41 See also HW Tang, ‘Natural Obligations and Unjust Enrichment’ (2006) 6 Oxford University Commonwealth Law Journal 133; D Sheehan, ‘Unjust Factors or Restitution of Transfers Sine Causa’ (2008) Oxford University Comparative Law Forum 1. 42 See especially Guinness, Mahon & Co Ltd v Kensington and Chelsea Royal London Borough Council [1999] QB 215 (CA). 43 R Stevens, ‘The New Birksian Approach to Unjust Enrichment’ [2004] 12 Restitution Law Review 260, 271. 44 See Kleinwort Benson Ltd v Lincoln CC [1999] 2 AC 349 (HL) 408–9 (Lord Hope); R (Rowe) v Vale of White Horse DC [2003] EWHC 388 (Admin), [2003] 1 Lloyd’s Rep 418, 422; Cressman v Coys of Kensington (Sales) Ltd [2004] EWCA Civ 47, [2004] 1 WLR 2775; Uren v First National Home Finance Ltd [2005] EWHC 2529 (Ch) [17]; Sempra Metals Ltd v IRC [2007] UKHL 34, [2008] 1 AC 561. See also M Chen-Wishart, ‘In Defence of Unjust Factors’ in Johnston and Zimmermann, above n 8, at 159; and Krebs, above n 14. 45 Deutsche Morgan Grenfell Group plc v IRC [2006] UKHL 49, [2007] 1 AC 558.

Demolishing the Pyramid—Unjust Enrichment

485

both in Kerrison v Glyn, Mills, Currie and Co46 and Dextra Bank and Trust Co Ltd v Bank of Jamaica,47 even though in both cases the claimant had taken a risk as to the liability arising in the future.48 Further, there are circumstances where a basis, as defined by Birks, can be identified but restitution is still appropriate. This is illustrated by those cases involving legal compulsion,49 where the claimant has been legally compelled to discharge a debt owed by the defendant. In such cases the restitutionary claim against the defendant cannot be analysed in terms of absence of basis, because the claimant has discharged a real legal liability owed to the defendant’s creditor. Birks, however, treated this as a situation where the benefit to the defendant is a by-benefit, which he characterised as a non-participatory enrichment.50 But since the claimant voluntarily discharged the debt on behalf of the defendant it is especially artificial to describe it as ‘non-participatory’. A further difference between the ground of restitution and absence of basis analyses arises as regards the operation of the ground of total failure of consideration. Where a claimant transfers a benefit to the defendant pursuant to a contractual obligation and the contract is then discharged for breach, Birks concluded that restitution should be awarded because the basis for the transfer becomes absent, and he rejected any requirement that the claimant must not have received any benefit from the defendant, since this is dealt with under a defence that counter-restitution is not possible. But this is not consistent with the state of the authorities, especially the decision of the House of Lords in Stocznia Gdanska SA v Latvian Shipping Co,51 which clearly affirmed the total failure of consideration requirement. As long as that requirement exists, there is a significant obstacle to the recognition of the absence of basis principle. For where there is some consideration which has not failed there is surely some basis for the transfer. Birks had further difficulty in fitting the established principles of restitution following invalidity of a contract within the absence of basis principle. He stated that the contract-breaker cannot get restitution,52 but subsequently qualified this to show that the contract-breaker cannot obtain restitution automatically because of breach but rather only once the breach has been accepted by the other party as repudiating the contract.53 However, the other party can obtain restitution

46

Kerrison v Glyn, Mills, Currie & Co (1911) 81 LJKB 465. See text accompanying n 26 above. Dextra Bank & Trust Co Ltd v Bank of Jamaica [2002] 1 All ER (Comm) 193 (PC). See Birks, above n 6, at 145, although he did accept that the claim was correctly rejected because of the operation of the defence of change of position. 48 As to the nature of risk-taking, see text at Part VI and accompanying notes, below. Birks considered, ibid, that the claimants in Dextra Bank were not risk-takers because they had been deceived and were at cross-purposes with the defendants and for that reason there was no basis for the payment. 49 See text at section V.C. (ii) and accompanying notes, below. 50 Birks, above n 6, at 159. 51 Stocznia Gdanska SA v Latvian Shipping Co [1998] 1 WLR 574 (HL), a case which Birks does not even cite. 52 Birks, above n 6, at 126. 53 See Dies v British and International Mining and Finance Corp Ltd [1939] 1 KB 724 (KB). 47

486

Graham Virgo

immediately following breach because proof that the contract is terminable—and similarly if it is voidable—is sufficient to establish a failure of basis, without the need to show that it has been terminated or rescinded. But surely the contract remains effective until terminated or rescinded and constitutes a continuing basis for the defendant’s receipt of the enrichment. The emphasis on absence of basis does result in other dissonances with the common law by identifying cases where restitution should have been awarded even though it has been rejected on the facts. So, in CTN Cash and Carry Ltd v Gallaher Ltd54 the claimant was unable to obtain restitution from the defendant, which had mistakenly demanded payment and had threatened to withdraw credit facilities in the future if the money was not forthcoming. In such a case the money which the claimant had paid to the defendant was not due to it, but the Court of Appeal held that the claim for restitution failed because the requirements for economic duress could not be established, it being lawful for the defendant to withdraw credit facilities in respect of future dealings.55 However, the application of the absence of basis principle would mean that restitution should be awarded simply because there was no basis for the payment to be made. But this raises serious concerns about the extension of the law of unjust enrichment, which would undermine the security of transactions, especially in a commercial context.

C. Analytical Complexity The biggest problem with Birks’ pyramidal analysis of the unjust factor is that it suggests something elegant when in fact it is unwieldy.56 The analogy with a pyramid is misplaced. The better analogy is with an iceberg, where nine-tenths of the object is hidden below the surface. This is because focusing on absence of basis alone cannot be sufficient to justify restitution. Where there is a potential basis for a transfer the grounds of restitution still need to be examined to determine whether the basis was valid. Even as regards the swaps cases, Birks acknowledged57 that it would be necessary to consider whether the claimant was mistaken when paying the money to the local authorities, since this would indicate that the claimant had not willingly taken the risk that the money was not due. But this emphasis on risk-taking, which appears to incorporate aspects of the law concerning the established grounds of restitution, undermines the elegant pyramidal structure of Birks’ approach to absence of basis. For, as regards his analysis of the swaps cases, he asserted that the unjust enrichment claim could be grounded simply on the invalidity of the underlying transaction. There is the absence of basis at the top of the pyramid, but we then need to return to the base of the pyramid to determine risk-taking with express reference to mistake. Now it may be easy to show in such 54 55 56 57

CTN Cash and Carry Ltd v Gallaher Ltd [1994] 4 All ER 714 (CA). See Baloch, above n 34. CTN Cash and Carry Ltd v Gallaher Ltd [1994] 4 All ER 714 (CA) 717–18 (Steyn LJ). See also Burrows, above n 39, at 45. Birks, above n 6, at 132.

Demolishing the Pyramid—Unjust Enrichment

487

a case that the claimant was not a risk-taker, but it is still a factor which needs to be taken into account. The role of risk within the law of unjust enrichment is significant and difficult, but the absence of basis approach does not make it any easier. For Birks, risk-taking is an implicit part of the absence of basis question,58 but it can be easily missed. Risk-taking needs to be brought out explicitly as a bar in its own right to unjust enrichment claims.59 The analytical complexity is reinforced by Birks’ assertion that a key example of absence of basis is where an express trust fails. In such circumstances a resulting trust will arise60 for the benefit of the settlor. Now, it is my view that the resulting trust cannot be usefully analysed with reference to the unjust enrichment principle, which I consider only to have personal and not proprietary consequences.61 That argument is, however, more difficult to defend if absence of basis is formally recognised. For if a settlor has constituted a trust and that trust fails, it is clear that there is no longer any explanatory basis in equity62 for the trustee’s retention of the property, which is consequently held on resulting trust for the settlor by virtue of the unjust enrichment principle. But if proprietary consequences are recognised within unjust enrichment here, logically they could be recognised in all cases where a basis has failed, which would result in excessive proprietary restitution.

D. Too Much Restitution The generally restrictive approach of the common law to unjust enrichment claims means that the defendant’s receipt of the benefit is secure save in the exceptional cases where the claimant establishes that the defendant’s enrichment is unjust. This principle of security of receipt is important to the law of restitution, especially in the commercial field where parties generally need to be certain that benefits have been effectively transferred and will not be upset too readily. A move to an absence of basis approach could destabilise commercial transactions.63 Birks recognised that this approach might result in too much restitution and, for that reason, he considered that identifying an absence of basis only means that a prima facie case for restitution has been established.64 The emphasis then shifts to the defences. But his analysis of the defences was much wider than the orthodox exposition and much of his analysis was far too vague. For example, he identified a vague principle of stultification where awarding a restitutionary remedy would

58

ibid, at 133. As discussed in text at Part VI and accompanying notes. Birks, above n 6, at 146. 61 See G Virgo, Principles of the Law of Restitution, 2nd edn (Oxford, Oxford University Press, 2006) 10. 62 There will be no failure of basis at law. 63 See also the analysis of CTN Cash and Carry Ltd v Gallaher Ltd [1994] 4 All ER 714 (CA). See text at Part IV.B and accompanying notes. 64 Birks, above n 6, at 263. 59 60

488

Graham Virgo

‘contradict a considered position elsewhere’,65 such as awarding a restitutionary remedy where the parties had made an illegal contract. But the vagueness of ‘stultification’ undermines the elegance of Birks’ analysis.

E. Uncertainty as to Basis A theory where the award of a restitutionary remedy depends on there being no basis for a transfer can only work effectively if the range of bases can be defined clearly and comprehensively.66 For, if it is not clear what a basis is or how it is defined, it will be impossible to determine whether there is any basis for the transfer, and it is an essential feature of Birks’ thesis that there is no basis at all for the defendant’s retention of the enrichment. This lack of clarity about the range and definition of particular bases is exemplified by the difficulty in defining gift—a basis of particular significance to the Birksian scheme. For example, must a gift involve the transfer of property or can it include the provision of services? What is the mental state of the donor to establish a gift? Further evidence of uncertainty about the range of bases derives from the Canadian experience, where there is a ‘no juristic basis’ test and where there has been a recent tendency to incorporate factors such as public policy or legitimate expectations of the parties to determine whether there was a basis for the receipt of an enrichment.67 This is surely far too uncertain to provide a workable test of basis. Sheehan has rightly described the Canadian approach as being ‘on the verge of chaos’ and the list of juristic reasons as being ‘somewhat eclectic and haphazard’.68 Consequently, Birks’ absence of basis theory should be rejected, since it is not supported by authority, principle or policy. But it certainly does not follow that the notion of absence of basis is alien to the law of unjust enrichment. In fact it forms a very important part of it and has long done so. The preferable view is that, once one of the recognised grounds of restitution has been identified, it is then necessary to determine whether there is any specific policy, rather than a defence, which should operate to bar the unjust enrichment claim. Two bars are particularly significant, namely that there was a basis for the defendant’s retention of the enrichment and, where no basis has been identified, that the claimant can be considered to have borne the risk of there being no basis.

V. Presence of Basis The true significance of basis is that, where there is a valid basis for the receipt of the enrichment by the defendant, this should bar the claim to restitution. This was specifically recognised by Lord Hope in Kleinwort Benson Ltd v Lincoln City 65

ibid, at 240. Sheehan, above n 41, at 2. Garland v Consumers’ Gas Distributors Inc [2004] 1 SCR 629; Pacific National Investments v Victoria [2004] SCC 75. 68 Sheehan, above n 41, at 10. 66 67

Demolishing the Pyramid—Unjust Enrichment

489

Council,69 who said that ‘the payee cannot be said to have been unjustly enriched if he was entitled to receive the sum paid to him’. Where there is a basis for a transfer, it is not for the law of restitution to undermine the transaction. It remains important then to determine what basis means. Whereas Birks focused on an essentially subjective notion of basis, the better view is that an objective interpretation should be adopted. The bar to restitution should apply where the receipt of the enrichment by the defendant can be considered to be legally effective in some way, in the sense that the transfer to the defendant has a legal consequence, as will occur for example where the transfer has discharged an existing obligation or can be considered to create an obligation other than the obligation to make restitution. The presence of basis principle has many advantages over the absence of basis principle. These include that it is consistent with the authorities and also that the legitimacy of the principle does not depend on whether it is possible to draw up a complete list of possible bases for receipt of an enrichment, as is required for the absence of basis principle. For the presence of basis principle, the restitutionary claim will only fail if a valid basis can be identified; it is not necessary to consider all possible bases to reach that decision. In practice what this means is that, once the elements of the unjust enrichment claim have been established by the claimant, restitution will follow save where a basis for the defendant’s retention of the enrichment can be identified by the defendant. This objective notion of basis can be established in a variety of ways.

A. Contract Undoubtedly the most significant basis for the transfer of an enrichment arises from contract. But the identification of a contractual basis can sometimes be fraught with difficulty.

(i) Continuing Contractual Regime Where a benefit is transferred to the defendant pursuant to a contractual obligation the continued existence of the contract will operate as a bar to the restitutionary claim.70 A contract will therefore constitute a basis for the payment, save where the contract can be shown to be void, has been rescinded or has been discharged for breach.71 This was recognised by Millett LJ in Portman Building Society v Hamlyn Taylor Neck:72 The continuing validity of the transaction under which the money was paid to the firm is, in my judgment, fatal to the society’s claim. The obligation to make restitution 69

Kleinwort Benson Ltd v Lincoln City Council [1999] 2 AC 349 (HL) 408. Guinness plc v Saunders [1990] 2 AC 663 (HL) 697–8 (Lord Goff); Taylor v Motability Finance Ltd [2004] EWHC 2619 (Com Ct) [23]; Mowlem plc v Stena Line Ports Ltd [2004] EWHC 2206 (QB); S & W Process Engineering Ltd v Cauldron Foods Ltd [2005] EWHC 153 (QB). 71 Barclays Bank Ltd v WJ Simms & Cooke [1980] QB 677 (Com Ct) 695. 72 Portman Building Society v Hamlyn Taylor Neck [1998] 4 All ER 202 (CA) 208. See also Britain v Rossiter (1879) LR 11 QBD 123 (CA) 127; Gilbert Partners v Knight [1968] 2 All ER 248 (CA) 250 (Harman LJ); Pan Ocean Shipping Ltd v Creditcorp Ltd [1994] 1 WLR 161 (HL) 164 (Lord Goff). 70

490

Graham Virgo

must flow from the ineffectiveness of the transaction under which the money was paid and not from a mistake or misrepresentation which induced it. It is fundamental that, where money is paid under a legally effective transaction, neither misrepresentation nor mistake vitiates consent or gives rise by itself to an obligation to make restitution.

The principle that a restitutionary remedy is not available if the contractual regime subsists is especially well illustrated by Taylor v Motability Finance Ltd,73 where the defendant employer terminated an employment contract with the claimant. The claimant argued that this constituted a repudiatory breach so that a restitutionary remedy could be awarded outside of the contract for the work he had done, grounded on total failure of consideration. This argument was rejected because the claimant had fully performed the contract by providing services to the defendant and so, although the primary obligation to perform had been revoked, the contractual regime subsisted in terms of the secondary obligation to pay damages for breach.74 This can be treated as consistent with the presence of basis principle in that the right to payment was considered to have already accrued under the contract. Where, however, a claimant has paid money to the defendant and received nothing in return, Taylor v Motability Finance suggests that, once the contract has been discharged, the contractual regime no longer applies and a restitutionary claim for the money will lie.75 Although this difference of approach depending on whether the enrichment involves money or services is artificial and difficult to defend, if that is put to one side the approach adopted in Taylor v Motability Finance is at least consistent with the presence of basis principle; if the contractual regime subsists then restitution should not be available. Sometimes, however, it has been recognised that a claim for restitution should lie despite the continued existence of the contractual regime. If this is correct, it would seem to undermine the presence of basis principle. This problem is especially well illustrated by the decision of the High Court of Australia in Roxborough v Rothmans of Pall Mall Australia Ltd,76 where the claimants sought restitution of a licence fee which had been paid to the defendant and which was then to be transferred to the state tobacco authority. There was no basis for the licence fee because it was demanded by reference to an invalid statute. The restitutionary claim succeeded on the ground of failure of consideration, even though the licence fee had been paid pursuant to a contract of sale which continued to operate. This was recognised by Kirby J, who dissented partly because the contract had not been terminated. Most of the majority failed to consider the requirement that the contract must be terminated before restitution could be awarded. Callinan J did consider the point and suggested that restitution would be available as long as performance of the contract was not possible and, if formal termination of the contract were 73

Taylor v Motability Finance Ltd [2004] EWHC 2619 (Com Ct). See Johnson v Agnew [1980] AC 367 (HL); Photo Products v Securicor Transport [1980] AC 827 (HL). 75 Taylor v Motability Finance Ltd [2004] EWHC 2619 (Com Ct) [25]. 76 Roxborough v Rothmans of Pall Mall Australia Ltd [2001] HCA 68, (2001) 208 CLR 516. See Cunnington, ‘Failure of Basis’ [2004] Lloyd’s Maritime and Commercial Law Quarterly 234. 74

Demolishing the Pyramid—Unjust Enrichment

491

required, the bringing of proceedings for restitution would be sufficient. But this would undermine the presence of basis principle, for if there is payment pursuant to a valid contract this should be sufficient to defeat the restitutionary claim. Beatson has argued that there may be exceptional circumstances where restitutionary relief can be awarded even though the contract has not been terminated, but only where this would not subvert the contractual allocation of risk.77 Even if this is correct, it could not justify the success of the restitutionary claim in Roxborough because there was no evidence that the risk of the licence fee being declared invalid had been placed on the defendant rather than the claimants. Consequently, Roxborough was a case where the award of restitution cannot be justified because there was a continued basis for the payment to be made. Indeed, Beatson’s acceptance that restitution can lie despite the continued validity of the contract is not consistent with the presence of basis principle and so should be rejected.

(ii) Settlement and Compromise78 Where the claimant settles or compromises the defendant’s claim for payment the claimant will be barred from obtaining restitution from the defendant, since there will then be a basis for the payment. A compromise involves some degree of concession on each side, including forbearance to sue,79 and is binding regardless of the validity of the claim.

(iii) Voidable and Terminable Contracts It is a crucial aspect of Birks’ absence of basis thesis that, where money has been transferred pursuant to a contract which is voidable or terminable for breach, there will be no basis for the payment even though the contract has not yet been rescinded or terminated.80 This has been criticised as highly artificial.81 However, the artificiality of Birks’ analysis can be avoided if the presence of basis principle is applied instead. For, according to that principle, until the contract itself is avoided or terminated there continues to be a valid legal basis for the transfer and so restitution will not be available.

(iv) Unenforceable Contracts Where money has been paid in respect of an unenforceable contract it should not be recoverable because the contract provides a continuing basis for the defendant’s retention of the money. As Brett LJ recognised in Britton v Rossiter,82 ‘the contract exists, but no one is liable on it’. The significance of this is exemplified by Morgan 77

J Beatson, ‘Restitution and Contract: Non-Cumul?’ (2000) 1 Theoretical Inquiries in Law 83. NH Andrews, ‘Mistaken Settlements of Disputable Claims’ [1989] Lloyd’s Maritime and Comparative Law Quarterly 431. 79 Callisher v Bischoffsheim (1870) LR 5 QB 449 (QB); Cook v Wright (1861) 1 BS 559; Atlee v Backhouse (1838) 3 MW 633. 80 Birks, above n 6, at 125–6. 81 S Meier, ‘No Basis: A Comparative View’ in Burrows and Rodgers, above n 15, at 347. 82 Britton v Rossiter (1879) 11 QBD 123 (CA) 127. 78

492

Graham Virgo

v Ashcroft,83 where the claimant’s mistaken payment to the defendant was held to be irrecoverable because it was made pursuant to an unenforceable gambling transaction.84 It follows that restitution will only be available if the unenforceable contract has been set aside for some other reason. There are, however, cases where restitution has been awarded even though the contract was unenforceable, most notably the decision of the High Court of Australia in Pavey and Matthews Pty Ltd v Paul.85 In that case a builder had made a contract with Mrs Paul to do some building work for her. That contract was unenforceable under the Builders Licensing Act 1971. Despite that, the builder was able to recover the value of the work which had been done. This does not appear to be consistent with the presence of basis principle. However, Birks’ analysis of the case, which enabled him to explain it as being consistent with the absence of basis principle, can assist here too. He said that the refusal of Mrs Paul to pay for the work was a repudiatory breach: ‘It was because of the breach of that good but unenforceable contract that the basis of her enrichment failed’.86 Mere terminability cannot be sufficient to establish there was no basis,87 but assuming that the builder had accepted the breach as repudiating the contract, there was no longer any valid basis for the retention of the enrichment and the builder should be able to establish the unjust enrichment claim in the normal way. The significance of the presence of basis principle is especially important where money is paid pursuant to a contract which is illegal. Such contracts are unenforceable88 and so a basis exists for the defendant’s retention of the benefit. This can consequently be used to explain why money paid in respect of an illegal contract cannot be recovered on the ground of total failure of consideration, because the contract bars the claim. So, for example, in Parkinson v College of Ambulance89 money which had been paid in return for the procurement of a knighthood could not be recovered. Similarly, in Taylor v Chester90 the plaintiff had deposited a £50 bank note with the defendant as security for the price of refreshments supplied by the defendant in her brothel. This was a contract which was contrary to public policy. It was held that the plaintiff could not recover the note because he could not make out the case for restitution ‘otherwise than through the medium and by the aid of the illegal transaction’ to which he was a party. It appears that the basis provided by the illegal contract is of a peculiar kind because of the distinct policy issues involving illegality. The illegal contract itself cannot be relied on to establish a claim for restitution, but the contract will not operate as a basis for the transfer

83

Morgan v Ashcroft [1938] 1 KB 49 (CA). Gambling agreements are now enforceable following the Gambling Act 2005 s 335, so if the contract is otherwise valid there will be a legal basis for payments under it. 85 Pavey & Matthews Pty Ltd v Paul (1987) 162 CLR 221 (High Court of Australia). 86 Birks, above n 6, at 127. 87 See text at section V.A.(iii) and accompanying notes 80 and 81, above. 88 See HG Beale (ed), Chitty on Contracts, 30th edn (London, Sweet & Maxwell, 2004) 1090. 89 Parkinson v College of Ambulance [1925] 2 KB 1 (KB). 90 Taylor v Chester (1896) LR 4 QB 309 (QB). 84

Demolishing the Pyramid—Unjust Enrichment

493

if, for example, the claimant can establish the claim without relying on it, such as where the claimant was mistaken in making the transfer,91 or the claimant was forced to enter into the illegal contract,92 or entered into the contract as the result of misrepresentation,93 or where the claimant had withdrawn from the contract before any part of the illegal purpose had been fulfilled.94

B. Res Judicata When money has been paid to the defendant under a court judgment it cannot be recovered95 unless the judgment is set aside, for the judgment constitutes a basis for the payment.96 However, once the judgment has been set aside, the money paid in respect of it can be recovered on the ground of mistake of law. The judgment operates as a valid basis even if it has been obtained by fraud.97 It is only where such a judgment has been set aside that restitution will be awarded.98 Similarly, where a judgment has been issued for too much money, the claimant is unable to recover the excess money paid until the judgment has been rectified.99

C. Discharge of a Debt There are numerous situations where a claimant has paid money which operates to discharge a debt.100 The discharge of the debt constitutes the provision of a legally-effective basis and this will operate to bar the restitutionary claim, at least to the extent of the value of the debt. The application of this principle is particularly well illustrated by cases involving mistaken and compelled payments.

(i) Mistake In Barclays Bank Ltd v WJ Simms Son & Cooke (Southern) Ltd101 Robert Goff J recognised that the claimant would not be able to recover money on the ground of mistake if the defendant had provided good consideration for the payment, such as where the payment has the effect of discharging a debt which was owed 91

Oom v Bruce (1810) 12 East 225. Atkinson v Denby (1862) 7 H & N 934. 93 Hughes v Liverpool Victoria Legal Friendly Society [1916] 2 KB 482 (CA). 94 Tribe v Tribe [1996] Ch 107 (CA). 95 Moses v Macferlan (1760) 2 Burr 1005, 1009 (Lord Mansfield); De Medina v Grove (1846) 10 QB 152; and Huffer v Allen (1867) LR 2 Ex 15 (Court of Exchequer). 96 See Birks, above n 6, at 140. Birks also considered that a judgment could operate as a defence to a restitutionary claim: at 233. 97 De Medina v Grove (1846) 10 QB 152; Huffer v Allen (1867) LR 2 Ex 15 (Court of Exchequer). 98 De Medina v Grove (1846) 10 QB 152. 99 Huffer v Allen (1867) LR 2 Ex 15 (Court of Exchequer). 100 A debt will be discharged typically where the creditor accepts payment: see D Friedmann, ‘Payment of Another’s Debt’ (1983) 99 Law Quarterly Review 534, 537; P Birks and J Beatson, ‘Unrequested Payment of Another’s Debt’ (1976) 92 Law Quarterly Review 188, 201. 101 Barclays Bank Ltd v WJ Simms Son & Cooke (Southern) Ltd [1980] QB 677 (Com Ct) 695. See also Bank of New South Wales v Murphett [1983] VR 489 (Supreme Court of Victoria). 92

494

Graham Virgo

to the defendant. It is the discharge of the debt which constitutes the provision of good consideration by the defendant102 and which means that the defendant is no longer enriched at the expense of the claimant to the value of the discharged debt.103 This good consideration defence could be treated as a form of change of position, but that defence is not available where money paid by mistake is used to discharge a debt,104 simply because the effect is that one debtor is substituted for another. Rather, the good consideration defence is preferably analysed as being dependent on a legally-effective basis existing for the payment. The operation of the good consideration defence is illustrated by Steam Saw Mills Co Ltd v Baring Bros & Co Ltd,105 where the claimant paid money due under a contract to agents of a foreign government in ignorance of the fact that a revolution had broken out which subsequently led to the downfall of the government. The money was not recoverable even though the plaintiff would not have made the payment had it known of the revolution, because ‘the money was paid, not under a mistake of fact as to the existence of an obligation; it was paid in pursuance of an obligation which in fact existed’106 and was effective to discharge that obligation. Where, however, the claimant paid money under a contract to the defendant in ignorance of the fact that the contract had already been repudiated by the other party, it was held that the payment was recoverable because it did not discharge any legal obligation.107 Whether the claimant’s payment will be effective to discharge the debt will depend on whether it was an authorised payment. So in Barclays Bank v WJ Simms Son & Cooke (Southern) Ltd108 the claimant bank had mistakenly paid money to the defendant in respect of a cheque drawn on a customer’s account. If, which did not occur on the facts, the claimant’s mistake had simply been that the customer had sufficient funds in the account it would follow that, since the customer had authorised the payment, the debt which the customer owed to the defendant would have been discharged and the claimant would be forced to recover the money from the customer.109 But, since the customer had countermanded the payment, it ceased to be authorised and so did not discharge the debt.110 Consequently, the claimant could recover the money from the defendant by reason of the mistake. Where, however, the debt is discharged because it was

102 David Securities Pty Ltd v Commonwealth Bank of Australia (1992) 175 CLR 353 (High Court of Australia) 406 (Dawson J). 103 Lloyds Bank plc v Independent Insurance Co Ltd [2000] QB 110 (CA) 132 (Peter Gibson LJ). 104 Scottish Equitable plc v Derby [2001] 3 All ER 818 (CA). 105 Steam Saw Mills Co Ltd v Baring Bros & Co Ltd [1922] 1 Ch 244 (CA). 106 ibid, at 254. 107 British American Continental Bank v British Bank for Foreign Trade [1926] 1 KB 328 (CA). 108 Barclays Bank v WJ Simms Son & Cooke (Southern) Ltd [1980] QB 677 (Com Ct). 109 See, eg, Lloyds Bank plc v Independent Insurance Co Ltd [2000] QB 110 (CA). 110 See also Customs and Excise Commissioners v National Westminster Bank plc [2002] EWHC 2204 (Ch), [2003] 1 All ER (Comm) 327: unsolicited payment to a creditor’s bank account would not discharge a debt unless it was accepted as doing such.

Demolishing the Pyramid—Unjust Enrichment

495

an authorised—albeit mistaken—payment, the defendant is entitled to receive the payment and so will not be required to make restitution.111 That the good consideration defence involves the application of the presence of basis principle is particularly well illustrated by Aiken v Short,112 where the plaintiff bankers paid money to the defendant to discharge a debt which was owed to the defendant by a third party, Carter. The bank mistakenly thought that it was discharging a mortgage on property which it had purchased from Carter, when in fact Carter did not own the property. It was held that the claimant was unable to recover the money from the defendant since Carter had authorised the payment and so the debt was discharged.113 The discharge of the debt was held to constitute the provision of consideration by the defendant.114 It did not matter that the discharge of the debt benefited a third party and not the claimant, for it is enough that the defendant provided some consideration for the payment from the claimant, regardless of who benefited from it.115 This is certainly consistent with the principle that where there is a legally-effective basis for the payment the claim for restitution will be defeated to that extent. If the defence of good consideration succeeds in such a case, the claimant will be forced to bring a restitutionary claim against the third party debtor who was the actual beneficiary of the debt being discharged, the ground of restitution being the claimant’s mistake.

(ii) Compulsion Where the claimant has been compelled by the defendant to pay money, the value of that payment can be recovered on the ground of duress, whether duress of the person or of goods or economic duress. These grounds of restitution require the defendant to have made an unlawful threat. However, it will not be available where the defendant demanded a sum which was lawfully due to him, for then there is a basis for the payment. Restitution will only lie where that which the defendant seeks from the claimant is something to which the defendant is not entitled. Where some money is lawfully due to the defendant and he receives more because of duress, the claimant will only be able to recover the excess amount for which no basis exists.116

111 See Pollard v Bank of England (1871) LR 6 QB 623 (QB); Kleinwort Benson Ltd v Lincoln CC [1999] 2 AC 349 (HL) 408 (Lord Hope); Lloyds Bank plc v Independent Insurance Co Ltd [2000] QB 110 (CA) 132 (Peter Gibson LJ). 112 Aiken v Short (1856) 1 H & N 210. See also Porter v Latec Finance (Qld) Pty Ltd (1964) 111 CLR 177 (High Court of Australia). 113 Aiken v Short (1856) 1 H and N 210, 214 and 215. This was said to be a ‘crucial’ fact in the case: Barclays Bank Ltd v WJ Simms Son & Cooke (Southern) Ltd [1980] QB 677 (Com Ct) 687 (Robert Goff J). 114 See especially Pollock CB in Aiken v Short (1856) 1 H & N 210, 214. 115 This point was ignored by the High Court of Australia in David Securities Pty Ltd v Commonwealth Bank of Australia (1992) 175 CLR 353 where the majority of the Court consistently referred to the defence as involving the receipt of consideration: see esp at 383. 116 See Astley v Reynolds (1731) 2 Strange 915.

496

Graham Virgo

Another ground of restitution involving compulsion, called undue pressure,117 arises where the defendant’s demand is lawfully made, such as where the defendant threatens to publish information.118 But this ground will also be unavailable where the defendant has demanded a sum of money which is lawfully due to him.119 A third ground of restitution is founded on the principle of compulsion, but this raises some more difficult issues as regards its consistency with the presence of basis principle. This is the ground known as legal compulsion, which applies where a benefit is demanded by somebody to whom it is lawfully due. In fact, claims involving legal compulsion are consistent with the presence of basis principle as it was interpreted in the context of the good consideration defence,120 since the claim involves the legitimate discharge of a debt or liability. The typical scenario where legal compulsion arises is where the claimant is compelled to transfer a benefit to someone in circumstances where, if the claimant had not transferred the benefit, it would have been taken from him by recourse to legal process. So, for example, the claimant may be threatened with litigation by a third party if the claimant does not pay the money demanded by the third party. If the claimant pays this money and it is effective to discharge the liability, the claimant will be unable to sue the third party, because the discharge of the liability means that there is a basis for the payment. Further, the claimant will be unable to complain that the third party exercised unlawful pressure by threatening to sue him, because the threat of litigation is ‘a normal and necessary incident of social life’.121 However, if the claimant can establish that the liability to the third party was ultimately borne by the defendant, a claim for restitution against the defendant will not be barred by the provision of basis, since, as between the claimant and the defendant, no legal basis exists for the payment and so the defendant will be liable to reimburse the claimant.122 This is illustrated by Exall v Partridge,123 where the plaintiff had left his carriage on land which was leased by the defendant for the defendant to repair. The defendant had not paid his rent to his landlord, who entered the land and lawfully seized the plaintiff ’s carriage in distress of rent. The plaintiff was compelled to pay the rent to the landlord in order to recover his carriage. The plaintiff recovered the money he had paid to the landlord from the defendant on the ground that he had not paid the money voluntarily but had been

117 Lloyd’s Bank v Bundy [1975] QB 326 (CA) 338 (Lord Denning MR). See also Williams v Bayley (1866) LR 1 HL 200 (HL) 214 (Lord Chelmsford), and 216 (Lord Westbury). 118 Norreys v Zeffert [1939] 2 All ER 187. 119 Flower v Sadler (1882) 10 QBD 572 (CA); Mutual Finance Ltd v John Wetton & Sons Ltd [1937] 2 KB 389 (KB). 120 See text at section V.C.(i) and accompanying notes, above. 121 P Birks, An Introduction to the Law of Restitution (Oxford, Oxford University Press, 1985) 185. 122 Re Downer Enterprises Ltd [1974] 1 WLR 1460 (Ch) 1468. 123 Exall v Partridge (1799) 8 Term Rep 308. See also Johnson v Royal Mail Steam Packet Co [1867] LR 3 CP 38; Edmunds v Wallingford (1885) 14 QBD 811 (CA) and Whitham v Bullock [1939] 2 KB 81 (CA).

Demolishing the Pyramid—Unjust Enrichment

497

compelled to do so by the landlord seizing his property.124 As between plaintiff and landlord there was a legally-effective basis for the payment. As between the plaintiff and the defendant there was no basis and, since the defendant had been unjustly enriched at the plaintiff ’s expense, it was appropriate to award restitution. The same principle applies where the claimant’s liability is ancillary to that of the defendant, as where the claimant is acting as surety for the defendant’s debts and the creditor threatens to sue the claimant unless the debt is discharged. The claimant will be liable to the third party to discharge the defendant’s debt by virtue of the guarantee but, because the defendant bears the ultimate liability in respect of the debt, the defendant will be required to reimburse the claimant.125 Similarly, if the claimant and the defendant share a common liability, but the claimant has paid what he owed and some, or all, of what the defendant owed, the claimant will seek a contribution from the defendant. As with the reimbursement cases, contribution can be analysed as a situation where the claim is not barred by the presence of basis. Although the claimant in a contribution claim has discharged a legal liability, it follows only that the claimant cannot seek restitution from the person to whom that liability was owed. However, since the claimant has also discharged the other party’s liability, it is appropriate that that party makes a contribution, and as between the claimant and this party there is no basis for the payment. So, for example, where the claimant and defendant are co-sureties126 under a common liability to the same creditor in respect of the same debt,127 and the claimant pays a sum of money to the creditor in respect of that debt, the defendant is required to make contribution in respect of his own share of the liability.

D. Statutory Authority Where the claimant has transferred a benefit to the defendant pursuant to a valid statutory obligation, restitution will not lie because there is a legal basis for the defendant’s receipt.128 So, for example, payment of tax pursuant to a valid tax demand cannot be recovered.129 The identification of a valid statutory basis for tax payments proved to be a matter of particular controversy in Deutsche Morgan

124 See in particular Grose J: (1799) 8 Term Rep 308, 311. Le Blanc J described this as ‘compulsion of law’: ibid. 125 Anson v Anson [1953] 1 QB 636 (QB). See also Re a Debtor [1937] Ch 156 (CA) 163 (Greene LJ); Moule v Garrett (1872) LR 7 Ex 101 (Exchequer Chamber). 126 Deering v The Earl of Winchelsea (1787) 2 Bos & Pul 27. 127 Cooper v Twynam (1823) Turn & R 424. 128 Deutsche Morgan Grenfell v IRC [2006] UKHL 49, [2007] 1 AC 558 [89] (Lord Scott). See M Chowdry and C Mitchell, ‘Tax Legislation as a Justifying Factor’ [2005] Restitution Law Review 1. 129 In Kingstreet Investments Ltd v New Brunswick [2007] SCR 3 the Supreme Court of Canada recognised that it was open to the State of New Brunswick to suspend the declaration of invalidity and also to enact valid tax legislation and apply them retrospectively. In either case the tax demand would become authorised.

498

Graham Virgo

Grenfell Group plc v IRC,130 where the majority allowed a claim for restitution of tax paid prematurely. The claimant had not been able to claim group income relief which would have delayed its liability to pay the Advance Corporation Tax which was due to the Revenue. Lord Scott dissented on the ground that, since the relief had not been claimed, the claimant was actually liable to pay tax to the Revenue so that the payment was lawfully due.131 In determining whether Lord Scott was correct it is vital to remember the essential feature of the restitutionary claim. This was not a claim for recovery of the amount of tax which had been paid because, as was subsequently recognised in Sempra Metals Ltd v IRC,132 the tax which had been paid was eventually set off against mainstream corporation tax which was lawfully due. The taxpayer’s claim related simply to the early payment of tax in circumstances where the claimant would have wished to delay payment by means of the group income election.133 If we focus on the entitlement of the Revenue to receive the money when it did, rather than on the language of whether the tax was actually due to the Revenue, it is much easier to understand why the Revenue was required to make restitution: because the Revenue obtained an unlawful timing advantage through premature payment of the tax. For the period when the Revenue had the unlawful use of this money there was no valid legal basis for its receipt. However, once the tax was set off against the taxpayer’s mainstream corporation tax liability, the Revenue’s receipt became lawful and so, at that point, there was a basis for this payment and so the money paid could not be recovered.

E. Natural Obligations In Moses v Macferlan134 Lord Mansfield recognised that, where a claimant discharged a valid obligation in circumstances where he was not aware that the defendant could not have enforced it against him, the claimant could not obtain restitution simply because the money was technically not due, because he was morally or naturally bound to have paid the money to the defendant. This socalled ‘natural obligation’ would exist where money is claimed from the claimant as payable in point of honour and honesty although it could not have been recovered from him by any course of law; as in payment of a debt barred by the Statute of Limitations, or contracted during his infancy, or to the extent of principal and legal interest upon an usurious contract, or for money fairly lost at play: because in all these cases, the defendant may retain it with a safe conscience, though by positive law he was barred from recovering.135

130 131 132 133 134 135

Deutsche Morgan Grenfell Group plc v IRC [2006] UKHL 49, [2007] 1 AC 558. See R Stevens, ‘Justified Enrichment’ (2005) 5 Oxford University Commonwealth Law Journal 141. Sempra Metals Ltd v IRC [2007] UKHL 34, [2008] 1 AC 561. Deutsche Morgan Grenfell v IRC [2006] UKHL 49, [2007] 1 AC 558 [5] (Lord Hoffmann). Moses v Macferlan (1760) 2 Burr 1005. Moses v Macferlan (1760) 2 Burr 1005, 1012.

Demolishing the Pyramid—Unjust Enrichment

499

If money has been paid in such circumstances it would appear that there is a basis for its receipt and so restitution will not be available. The relevance of natural obligations today can be assessed with reference to each of Lord Mansfield’s examples. First, if money is paid to discharge a debt where the claim is time-barred, the claimant might assert that this was a mistaken payment, in the sense that, but for the mistake the money would not have been paid. But restitution should be denied, because the payment is effective to discharge the debt.136 In other words, there is a legally-effective basis for the payment, and the fact that the claim was time-barred is irrelevant. The analysis would only be different if the effect of the relevant limitation statute was to discharge the debt, for then the payment would not be legally effective since the payment would not have discharged the debt. Usually, however, limitation periods ‘bar only the remedy and not the right’137 and so there is no need to rely on a basis other than debt to bar the restitutionary claim. Lord Mansfield’s second category is where money is paid pursuant to a contract which is made during infancy. Again, restitution will be barred, but not because of some vague concept of ‘natural obligation’, but because the contract operates as a valid legal basis for the transfer. Contracts with minors for necessaries are valid. Contracts where the minor acquires an interest of a permanent or continuous nature, such as for the acquisition of shares or land, are voidable,138 and so remain a basis for the transfer until the contract is avoided. All other contracts with minors are unenforceable, and so constitute a basis until the contract has been set aside for some reason.139 So, again, the language of natural obligation is not of use here. The third category concerning usurious contracts is no longer relevant. The fourth category concerning gambling contracts is now covered by the Gambling Act 2005, which renders such contracts enforceable.140 Consequently, restitution of payments in respect of such contracts will be barred in the normal way by the existence of a valid contractual basis for payment.141 So if the specific heads identified by Lord Mansfield can now be explained using normal principles of presence of basis, is there any need to identify a ‘natural obligations’ head of basis? This will depend on how ‘natural obligations’ should be defined. Sheehan142 has described such obligations as arising where a claimant fulfils a duty which he had voluntarily agreed to undertake, even though the contract under which the claimant undertook the duty is void. But this does 136 See text at section V.C. and accompanying notes, above. See also HW Tang, ‘Natural Obligations and the Common Law of Unjust Enrichment’ (2006) 6 Oxford University Commonwealth Law Journal 133, 140. 137 ibid, at 141. 138 Steinberg v Scala (Leeds) Ltd [1923] 2 Ch 452 (CA). 139 See text at section V.A.(iv) and accompanying notes, above. 140 Gambling Act 2005 s 335. 141 See text at section V.A(i). and accompanying notes, above. Or it could be treated as a ‘voluntary gift from the loser to the winner’: Lipkin Gorman v Karpnale Ltd [1991] 2 AC 548 (HL) 562 (Lord Templeman), although this may put the natural interpretation of ‘gift’ under some pressure. 142 D Sheehan, ‘Natural Obligations in English Law’ [2004] Lloyd’s Maritime and Commercial Law Quarterly 172. Birks, above n 6, at 257–8, wrote about a ‘moral obligation to pay’.

500

Graham Virgo

involve a much wider notion of ‘basis’ than those previously encountered since it is not necessarily a legally-effective basis for the payment. Tang has adopted a narrower interpretation of the natural obligations principle which has regard to ‘whether the policies behind the doctrine or law which avoided the obligation will be furthered by allowing or denying restitution’.143 But the crucial issue is whether transfer of a benefit pursuant to a void obligation should ever be barred. Application of the presence of basis principle suggests that it should not be. For if there is no legally-effective basis for the transfer because the contract is void, there is no reason why the restitutionary claim should be defeated.144 Vague references to underlying policies cannot assist. Once the unjust enrichment claim has been established, and mistake of law will usually be readily identifiable where there is a void obligation, restitution should follow unless there is some other basis which can be identified for the transfer. Consequently, there is no need to treat ‘natural obligations’ as providing a basis distinct from any of the other bases.

F. Gift Where the receipt of a benefit can be characterised as a gift, this should constitute a legally-effective basis for the transfer and restitution should be denied.145 It is consequently important to be able to identify when a transfer can be considered to be a gift. But this is not straightforward because the key test turns on the nature of the claimant’s mental state at the time of the transfer. Clearly if the claimant intended the transfer to be gratuitous then this will be a gift. Birks considered that the notion of oblique intent will suffice, so that if the claimant foresaw that the defendant might receive a benefit as a virtually certain consequence without the claimant receiving anything in return, this could be considered to be gratuitous. Birks even considered that the claimant’s knowingly taking a risk that he might not be paid for the provision of a benefit could mean that the claimant had made a gift to the defendant.146 So, for example, the busker who wants to be paid for his music was considered by Birks to have made a gift of his services to those who chose not to pay. But surely it is artificial to describe the busker as a donor of his services. The real problem with the busker is that he is a risk-taker and the notion of risk has become unacceptably wrapped up with Birks’ absence of basis analysis. Risk-taking should not be treated as an implicit part of the identification of basis question. Consequently, the concept of gift should be interpreted restrictively to encompass only a purpose to benefit gratuitously. Where the benefit has been

143

Tang, above n 136, at 133. Save if the claimant bore the risk of invalidity: see text at Part VI and accompanying notes, below. 145 Birks considered that a transfer which was intended by the claimant to be a gift could not be recovered: Birks, above n 6, at 104. 146 ibid, at 130–31. 144

Demolishing the Pyramid—Unjust Enrichment

501

provided recklessly then whether restitution should be awarded should depend on the application of the different principle of risk-taking.147 Where the claimant has purported to make a gift to the defendant there will not be a valid basis for the transfer if there are particular circumstances which undermine the validity of the gift, such as where the claimant was mistaken or had been compelled to make the gift to the defendant. However, there is some evidence, in the context of mistake at least, that it is not enough to show that but for the mistake the gift would not have been made; some additional reason must be established, such as misrepresentation. This was especially emphasised by Lord Scott, obiter, in Deutsche Morgan Grenfell Group plc v IRC148 where he recognised that [m]y present opinion is that unless there were some other reason, such as a misrepresentation by B, to enable the gift to be set aside, the mistake made by A would not suffice, notwithstanding that the payment had not been made pursuant to any legal obligation and that but for the mistake it would not have been made. But [the resolution to this question] can be left for another day.

The presence of basis principle can only operate effectively if it is clear when there will be a basis for a transfer. It follows that it is a matter of real significance to determine when a gift will be valid and a lot more work needs to be done to determine when the basis of a gift can be identified and why.149

G. Voluntary Transfers150 Where the claimant has acted voluntarily in transferring a benefit to the defendant which was not due then, even though it appears that there was no legallyeffective basis for the transfer, the claimant will not be able to obtain restitution. This principle was recognised by Lord Reading in Maskell v Horner:151 If a person with knowledge of the facts pays money, which he is not in law bound to pay, and in circumstances implying that he is paying it voluntarily to close the transaction, he cannot recover it. Such a payment is in law like a gift, and the transaction cannot thereafter be re-opened.

Lord Reading’s analogy with a gift is consistent with the idea that a voluntary transfer can in fact be considered to be a legally-effective basis in its own right, which should bar the restitutionary claim. The difficult issue, of course,

147

See text at section VI and accompanying notes. Deutsche Morgan Grenfell Group plc v IRC [2006] UKHL 49 [207] 1 AC 558 [87]. See also Wilson v Thornbury (1875) LR 10 Ch App 239 (CA); Ogilvie v Littleboy (1897) 13 TLR 399 (CA). 149 Some of this work has been done by HW Tang, ‘Restitution for Mistaken Gifts’ (2004) 20 Journal of Contract Law 1. 150 David Securities Pty Ltd v Commonwealth Bank of Australia (1992) 175 CLR 353 (High Court of Australia) 374. 151 Maskell v Horner [1915] 3 KB 106 (CA) 118. 148

502

Graham Virgo

concerns the identification of when the claimant can be considered to have acted voluntarily. Within the traditional analysis of the grounds of restitution the question of whether the claimant had acted voluntarily is generally indistinguishable from the question of whether the ground of restitution vitiated the claimant’s intention to transfer the benefit. For example, if the claimant was compelled to pay the defendant, the claimant will only be considered to have acted voluntarily if the compulsion had not actually caused the claimant to make the payment. But if the compulsion did not cause the claimant to make the payment it cannot constitute a ground of restitution anyway. The burden is consequently on the claimant to prove that he acted involuntarily, in other words that the compulsion caused the payment to be made.152 Similarly, if the claimant paid the defendant as a result of a mistake, the claimant will only be able to obtain restitution if the mistake was a sufficient cause of the payment and, if it was not, the claimant can be considered to have acted voluntarily. This was recognised by Robert Goff J in Barclays Bank Ltd v WJ Simms Son & Cooke (Southern) Ltd,153 when he said that restitution for mistake would be denied where ‘the payor 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’. However, even if it appears that the claimant had acted involuntarily due to mistake or compulsion, it may still be possible to characterise the transfer as voluntary for other reasons. For example, where the defendant has made an unfounded claim accompanied by a threat to sue, if the claimant mistakenly satisfies the claim he will not be able to obtain restitution because the payment will be deemed to be voluntary.154 This has sometimes been called ‘submission to an honest claim’, but it does not depend on the defendant having made an explicit claim; the claim may be implied. Payment made in response to such compulsion is deemed as a matter of law to have been voluntary simply because the claimant was given the opportunity to contest his liability in proceedings but gave way and paid the defendant.155 It makes no difference whether the claimant thought that the money was not lawfully due to the defendant or mistakenly believed that it was due. The voluntariness of the claimant’s actions has also proved to be significant where the claimant voluntarily became the surety for the defendant’s liabilities owed to a third party without being requested to do so by the defendant. If the third party compels the claimant to discharge the defendant’s liability it has been held that the claimant is not able to obtain restitution from the defendant, even though the defendant has benefited from the discharge of liability and bore the 152

Mason v New South Wales (1959) 102 CLR 108 (High Court of Australia) 144 (Windeyer J). Barclays Bank Ltd v WJ Simms Son & Cooke (Southern) Ltd [1980] QB 677 (Com Ct) 695. See also Parke B in Kelly v Solari (1841) 9 M & W 54, 59 and Lord Abinger CB (at 58). 154 Moore v Vestry of Fulham [1895] 1 QB 399 (CA); Kleinwort Benson Ltd v Lincoln CC [1999] 2 AC 349 (HL) 382 (Lord Goff). 155 Moore v Vestry of Fulham [1895] 1 QB 399 (CA) 402 (Lord Halsbury). 153

Demolishing the Pyramid—Unjust Enrichment

503

ultimate liability, because the claimant voluntarily assumed the suretyship.156 Barring restitution in such circumstances has been criticised,157 but it can be regarded as consistent with the presence of basis principle. In particular, it is appropriate to treat the discharge of liability in these circumstances as being akin to a gift, so that a basis for the payment can be identified. This analysis is not obviously consistent, however, with those cases which have recognised that voluntary actions bar restitution. So, for example, in Owen v Tate158 the plaintiff, without consulting the defendant, deposited a sum of money with the defendant’s creditor as security for a loan. The reason for this was that the defendant’s previous surety wished to recover the title deeds of her property which had been surrendered to the creditor and, as a favour to her, the plaintiff agreed to become surety. When the defendant discovered that the plaintiff had become the surety it protested, but when pressed for payment by the creditor it asked the creditor to have recourse to the money which the plaintiff had deposited. The plaintiff sought restitution from the defendant but the claim was unsuccessful because it was held that the plaintiff had undertaken the risk of being sued by voluntarily becoming a surety. But it is a significant feature of Owen v Tate that the plaintiff became a surety to benefit a third party and not the defendant; if there was any intent to make a gift it was to the third party and not the defendant. So can a basis for the payment be identified even though it relates to a third party and not to the defendant? It was suggested earlier that this would not be sufficient to bar the claim as against the defendant who benefited from the discharge of liability.159 This consequently suggests that the Owen v Tate principle is not consistent with the presence of basis principle, because as between the claimant and the defendant there was no basis for the payment to the creditor, and indicates that Owen v Tate is incorrect. Similarly, in Macclesfield Corp v Great Central Railway160 the defendant was under a statutory duty to repair a bridge over its canal, but refused to repair it when the claimant highway authority called upon it to do so. The claimant then did the work itself but was unable to obtain restitution of the costs incurred since it was under no legal liability to repair the bridge, and accordingly had acted as a volunteer. Again, it is difficult to characterise the claimant as having intended to make a gift to the defendant. Voluntariness does not defeat restitutionary claims in other areas where the ground of restitution is legal compulsion. So, for example, in Edmunds v Wallingford161 the fact that the claimant’s goods which were seized had been left voluntarily on the defendant’s premises was held to be irrelevant. It is also inconsistent with the law on contribution whereby a surety is entitled to contribution 156 Owen v Tate [1976] QB 402 (CA). See also Re National Motor Mail Coach Co Ltd [1908] 2 Ch 515 (CA) 520; Esso Petroleum Ltd v Hall Russell & Co Ltd [1989] AC 643 (HL). 157 See Birks, above n 121, at 311. 158 Owen v Tate [1976] QB 402 (CA). 159 See text at section V.C. and accompanying notes. 160 Macclesfield Corp v Great Central Railway [1911] 2 KB 528 (CA). 161 Edmunds v Wallingford (1885) 14 QBD 811 (CA).

504

Graham Virgo

from his co-surety regardless of whether the claimant was requested by the defendant to become a surety. Consequently, the better view in these three-party cases is that, where the claimant voluntarily placed himself in a position to be legally compelled to discharge a liability owed by the defendant to a third party, then, as between the claimant and the defendant, there is no basis to bar the restitutionary claim.162

VI. Risk-Taking Where a claimant has transferred a benefit to the defendant in circumstances where there is a ground of restitution but no legal basis for the defendant’s receipt, clearly restitution will not be defeated by the presence of basis bar. However, restitution may still be barred by a distinct principle, where the claimant can be considered to have taken the risk that there might not be a valid basis for the defendant’s receipt of the enrichment. Although the language of risk-taking can be identified in a variety of cases and commentaries, there has been little progress in developing any coherent jurisprudence as to the identification and allocation of risk. In fact, risk-taking may arise in two different ways, depending on the claimant’s mental state at the time of the transfer and whether the risk has been allocated by the parties.

A. Claimant’s Mental State If the claimant knew that there was no basis for the transfer of an enrichment to the defendant then it cannot be recovered,163 and similarly where the claimant suspected that there was no basis for the transfer.164 Various reasons have been suggested as to why restitution should not be available where the claimant is suspicious about the legitimacy of the demand. These were considered by the House of Lords in Deutsche Morgan Grenfell Group plc v IRC.165 Lord Brown,166 in the context of a claim for mistaken payment, said that a claimant who was suspicious about the validity of the payment would not be mistaken. So, for example, a claimant who paid tax suspecting that it was not due because, for example, it was being challenged in the courts, might not be able to rely on mistake to recover the tax, because suspicion negated the mistake. However, other judges in Deutsche 162

Although restitution might still be denied by virtue of the distinct bar of risk-taking. Brisbane v Dacres (1813) 5 Taunt 143, 159–60 (Gibbs J); Brennan v Bolt Burdon [2004] EWCA Civ 1017, [2005] QB 303. 164 Kleinwort Benson Ltd v Lincoln City Council [1999] 2 AC 349 (HL) 410 (Lord Hope); R (Rowe) v Vale of White Horse District Council [2003] EWHC 388 (Admin); Brennan v Bolt Burdon [2004] EWCA Civ 1017, [2005] QB 303; Cobbold v Bakewell Management Ltd [2003] EWHC 2289 (Ch). 165 Deutsche Morgan Grenfell Group plc v IRC [2006] UKHL 49, [2007] 1 AC 558. 166 ibid, at [162]. 163

Demolishing the Pyramid—Unjust Enrichment

505

Morgan Grenfell considered that a suspicious claimant could be considered to be mistaken but restitution might still not be available, either because the effect of the claimant’s suspicions is that the mistake may not have been the cause of the transfer,167 since the claimant may have taken a calculated risk about the existence of a liability to pay money,168 or simply because somebody who has taken a risk as to the existence of a basis for the payment should not be able to get restitution. The latter approach was advocated by Lord Hoffmann,169 although he did not consider that doubts as to liability would necessarily mean that the claimant had taken the risk that money might not be due. Rather, he considered that whether the claimant was a risk-taker would depend on ‘the objective circumstances surrounding the payment as they could reasonably have been known to both parties, including … the extent to which the law was known to be in doubt’.170 This will, however, depend on the degree of doubt; while some doubt will not prevent recovery,171 the greater the doubt the less likely it is that recovery will be ordered. The key significance of the difference between the judges in Deutsche Morgan Grenfell is that the first approach, which suggests that suspicion negates mistake, is not of general application throughout the law of unjust enrichment where other grounds of restitution apply. The alternative approach, focusing on risk-taking, could apply generally throughout the law of unjust enrichment regardless of the nature of the ground of restitution being relied on. However, the application of the risk-taking principle may depend on the nature of the ground of restitution involved. For example, if a claimant has paid tax which was not due to the Revenue and seeks to recover the overpaid tax using the Woolwich172 principle, should this claim be barred if the claimant was aware of the possibility that the tax might not be due? Surely in this context the fact that the claimant is a risk-taker should not bar the claim for restitution, since the constitutional rationale which underpins the Woolwich principle should support restitution being granted regardless of the claimant’s suspicions. Similarly, could the risk-taking principle defeat a claim founded on duress, or will the fact that the defendant has exerted or threatened to exert unlawful pressure be sufficient to negate the claimant’s exposure to the risk of compulsion? It is clear that risk-taking by the claimant may bar a claim to restitution. This will undoubtedly be affected by the claimant’s mental state, but Lord Hoffmann’s recognition of the significance of ‘objective circumstances’ potentially expands the boundaries as to when a restitutionary claim might be defeated because the

167

ibid, at [65] (Lord Hope). ibid, at [175] (Lord Brown). ibid, at [27]. 170 ibid, at [28]. 171 Charfield v Paxton (1799) 2 East 471. See also Kleinwort Benson Ltd v Lincoln City Council [1999] 2 AC 349 (HL) 401 (Lord Hoffmann), 412 (Lord Hope); Westdeutsche Landesbank Girozentrale v Islington LBC [1994] 4 All ER 890 (CA) 934 (Hobhouse LJ) (need for ‘an actual conscious appreciation’ for recovery to be barred). 172 Woolwich Equitable Building Society v IRC [1993] AC 70 (HL). 168 169

506

Graham Virgo

claimant can be deemed to have taken the risk that there was no basis for the defendant’s receipt of the enrichment. It will be important to identify which considerations will be relevant to the deeming of risk-taking, but the nature of the ground of restitution will no doubt be significant.

B. Allocation of Risk Where the risk that there is no legally-effective basis for a transfer has been allocated by the parties it is not for the law of restitution to reallocate it.173 Whether the risk has been allocated will turn on all the evidence, including the nature of any bargaining between the parties, but such a conclusion should only be drawn where the risk has clearly been allocated to the claimant. Whether the risk has been allocated is especially well illustrated by the cases which consider whether a payment from the claimant can be characterised as a non-refundable deposit or simply an advance payment. If it is the former it follows that the payer takes the risk of his default in the performance of the contract but does not take the risk of the defendant’s default.174 If it is an advance payment the claimant does not bear the risk of any default.175 Similarly, where the claimant transfers a benefit to the defendant pursuant to an agreement which is subject to contract, this operates to exclude the right to restitution as a general rule, because the inclusion of the phrase ‘subject to contract’ allocates the risk of a contract not being made to the person who transfers a benefit before a contract is made.176

VII. Application of the Principles The application of the presence of basis and risk-taking principles is especially well illustrated by the decision of the High Court of Australia in Lumbers v W Cook Builders Pty Ltd (in liq).177 Had these principles been applied explicitly the same result would have been achieved, but in a way which would have been much easier to understand. The case involved an oral contract made between Lumbers and Sons to build a house. Without Lumbers’ knowledge, Sons entered into a sub-contract with Builders, an unlicensed builder, to do most of the building work. Builders then sued Lumbers for a quantum meruit. The Supreme Court of South Australia had held that Lumbers was liable to make restitution because it had been incontrovertibly benefited, it would be unconscionable for it to keep the benefit without paying for it and the contract between it and Sons was irrelevant to its liability to Builders. All 173 See Pan Ocean Shipping Co Ltd v Creditcorp Ltd [1994] 1 WLR 161 (HL) 166 (Lord Goff of Chieveley). 174 Howe v Smith (1884) 27 Ch D 89 (CA); Monnickendam v Leanse (1923) 39 TLR 445. 175 Palmer v Temple (1839) 9 Ad & E 508. 176 See, eg, Regalian Properties plc v London Docklands Development Corp [1995] 1 WLR 212 (Ch). 177 Lumbers v W Cook Builders Pty Ltd (in liq) [2008] HCA 27.

Demolishing the Pyramid—Unjust Enrichment

507

the judges in the High Court overruled this decision. Four judges178 delivered a joint judgment. This was wide-ranging, with much of the analysis focusing on the identification of an enrichment, but essentially the decision focused on the fact that the contract between Lumbers and Son barred any restitutionary liability to Builders.179 But it was not made clear why this contract would be relevant in this way. In fact, there was a much easier and convincing route to reach the same result, a route which was essentially adopted by Gleeson CJ in his judgment. The best way of analysing this route is by returning to first principles. Builders was seeking to establish an unjust enrichment claim against Lumbers. It could show that Lumbers had been enriched at its expense and, presumably, the relevant ground of restitution would be total failure of consideration,180 in that it had provided a benefit without receiving anything in return. The question then was whether this claim was barred. First, was there a legally-effective basis for the receipt of the enrichment by Lumbers? There was a contract between Lumbers and Sons, which had not been terminated, but the provision of the benefit by Builders could not be attributed to this contract. Further, although there was a contract between Sons and Builders, the provision of the benefit could not be regarded as a legallyeffective basis as between Lumbers and Builders because there was no contractual relationship between them. This fact pattern is very similar to that which was considered previously in respect of a claimant discharging the debt owed by the defendant to a third party.181 In that three-party situation, the discharge of the debt was legally effective as between claimant and third party but not as between claimant and defendant. So, in Lumbers v Cook, the provision of the benefit by Builders discharged a contractual liability as between Sons and Builders,182 but this had no effect on the non-contractual relationship between Lumbers and Builders. However, it does not follow that the contractual relationship between Sons and Builders was irrelevant to the restitutionary claim. Although there was no legallyeffective basis for Lumbers’ receipt of the enrichment, the fact that Builders provided the benefit as a result of the contract with Sons meant that Builders should be regarded as having borne the risk that Sons would not pay for the benefit provided. This was specifically recognised by Gleeson CJ who said:183 The contractual arrangements that were made effected a certain allocation of risk; and there is no occasion to disturb or interfere with that allocation. On the contrary, there is every reason to respect it. There was no mistake or misunderstanding on the part of Builders. It was accepted on both sides in argument that in the ordinary case a building

178

Gummow, Hayne, Crennan and Kiefel JJ. Lumbers v W Cook Builders Pty Ltd (in liq) [2008] HCA 27 [77]. Or possibly even mistake. 181 See text at section V.C. and accompanying notes, above. 182 Gleeson CJ did say that ‘[i]f any party has been enriched at the expense of Builders, it is Sons’: [2008] HCA 27 [27]. But it does not follow from this that Builders would have a restitutionary claim against Sons, because Sons’ contractual liability to pay for that enrichment remained valid. In other words there was a legal basis for the enrichment. 183 ibid, at [46]. 179 180

508

Graham Virgo

subcontractor does not have a restitutionary claim against a property owner, but must look for payment to the head contractor.184

The other judges made matters far too complicated by considering that the contract between Lumbers and Sons (ie between defendant and third party) was relevant to the claim for restitution brought by Builders against Lumbers.185 This was the wrong contract. The risk was allocated by the contract between Builders and Sons (claimant and third party) and, as Gleeson CJ said,186 ‘the scope of restitutionary claims [should not be extended] beyond the bounds set by legal principle, especially where to do so would be to cut across or disturb contractual relationships and established allocation of risk’. Although the other judges were right to say that the identification of rights and obligations ‘requires close attention to what contractual or other obligations each owes to the other’,187 it is vital to be clear as to why a particular contractual relationship is considered to be relevant. This is much easier to understand if the principles of basis and risk-taking are considered carefully.

VIII. Conclusion To return to where we started, the law of unjust enrichment is properly analysed as a linear formula. But to the elements of enrichment at the claimant’s expense with a recognised ground of restitution and no available defence should be added that there was no legally-effective basis for the transfer and that the claimant had not taken the risk that there might be no basis. So, if the claimant has paid money to the defendant as a result of a mistake, restitution will be denied if there was a basis for the payment, such as discharge of a debt, or the claimant took the risk that the money had been paid by mistake. Focusing on the presence of basis and developing a much clearer jurisprudence of risk-taking is consistent with the cases, and with unjust enrichment theory and policy. Many commentators have considered Birks’ Unjust Enrichment to be a radical re-working of the law of unjust enrichment through its focus on absence of basis. But it was not. Many have criticised Birks’ approach as not being grounded in the cases. But much of it was. The most serious criticism is that Birks sought to create a new pyramidal structure when none was needed. There is plenty of evidence that the courts have for many years had regard to whether there was a basis for a transfer, but this is an adjunct to and not a replacement of the traditional unjust factors. Despite this a number of commentators and a few judges have supported the pyramid. If it exists, it should be demolished; if it does not yet exist, planning permission should never be given for its construction. 184 See also Pan Ocean Shipping Co Ltd v Creditcorp Ltd [1994] 1 WLR 161 (HL) 166 (Lord Goff of Chieveley). 185 Lumbers v W Cook Builders Pty Ltd (in liq) [2008] HCA 27 [80], [124]–[127]. 186 ibid, at [48]. 187 ibid, at [127].

INDEX

A v Bottrill 276–7 Abatiello v Monsanto 395 academic moralism 66–7 account of profits 342–3, 373–80 contract/fiduciary overlap 378–9 as diversion of commercial opportunities remedy 374–5 insolvency implications 375–7 action rights corrective justice and 120–2 as external pluralism 118–19 Aiken v Short 495 Allied Maples Group Ltd v Simmons & Simmons 183–4 American Law Institute Restatement of the Law of Restitution 415–19 Restatement of the Laws of Tort, Second 169 Restatement of the Laws of Tort, Third: Liability for Physical Harm 84–5, 88–9, 108 Anns v London Borough of Merton 90–1, 289–90, 302 Anson, W 199 Aquinas, St Thomas 116 Aristotle 196, 212 Nicomachean Ethics 54, 57–8 Aston Cantlow v Wallbank 72–3 Atiyah, PS 195 Attorney-General for Hong Kong v Reid 372, 376–7, 379–80, 384 Attorney-General v Birmingham Borough Council 411–12 Attorney-General v Observer Ltd 277 Australia, restitution law 380–2, 398 autonomous legal person 459–66 Avon County Council v Howlett 470 Barclays Bank Ltd v WJ Simms Son & Cooke (Southern) Ltd 493–5, 502 Barker v Corus (UK) Ltd 167, 172–3, 177–8, 188, 189 Batchelor v Tasmania 295–6 Bazley v Curry 268, 274–5 Becker, GS 319 Becker, L 354, 356 Beever, A 115–16, 175, 185–6 benefits, self-regarding conferral of 439–43 bipolarity constraint 272–9

Birks, P 195–6, 202, 207, 397–9, 441 comparative law use 474–6 event-based classification 215–16, 222–30 as internalist 195–6 on obligations 445n restitution see under unjust enrichment as follows absence of basis, basis, narratives Blackstone, W, Commentaries 31, 34, 425, 428 BM v British Columbia 296–7 Boardman v Phipps 375 Bolton v Stone 99 Bonnington Castings Ltd v Wardlaw 168 boundary disputes 224 Branch v Mobile Oil 395 breach of contract 226 Brimelow v Casson 161–2 Britton v Rossiter 491 Brooks v Commissioner of Police for the Metropolis 287 Brownsword, R 3 Brudner, Alan 102–3 Burlington Northern & Santa Fe Ry Co v Grant 396 Burrows, A 349, 460, 463–4 Calabresi, G 83 Caltex Oil (Australia) Pty Ltd v The Dredge ‘Willemstad’ 267 Canada Charter of Rights and Freedoms 32, 294, 306 gains-based remedies 365–6, 383–8 law of restitution 380–1 Supreme Court 40–4 Canada Inc v Strother (3469420) 366, 373–4, 375, 377, 378–9, 381, 383–4 Canada Trust Co v Ontario Human Rights Commission 25 Canadian Aero Service v O’Malley 375 Canadian National Railway Co v Norsk Pacific Steamship Co 257 Cane, P 198, 217, 270 Canson Enterprises Ltd v Broughton & Co 378 Caparo Industries plc v Dickman 268 Carr-Saunders v Dick McNeil Associates Ltd 391, 398 Carter, JW 398 Cattenach v Melchior 264

510

Index

causation concepts 243 fiduciary allowances 345–8 causation indeterminacy background 167 case law 168–73, 180–3 compensation/deterrence 187–90 corrective justice explanations 174–9, 187 evidential damage 175–7 loss of chance 179–87 argument 180–3 evaluation of argument 183–7 risk imposition 177–9 tort goals and 165–7 Chambers, R 196 Chan v Zacharia 335 chance, loss of see under causation indeterminacy Chaplin v Hicks 183 Chen-Wishart, M 461, 463 Children’s Foundation case 43–4 Chin, NY 462–3 Chirnside v Fay 353–4, 360 Christian legal scholarship 198 classification 215–37 aims 216–17 animal taxonomy and 471–4 background 215–16 by causative events 222 construction issues 225–30 deficiencies 230–2 dispositive 222–5 conclusions 236–7 dispositive 219–21, 222–5 expository 217–19 likeness issues 221, 232–6 taxonomist approaches 196 coherence 141, 255–6, 271–2 Coleman, J 51–2, 83, 85, 92, 402–3 Collins, H 198 common law constraints 269–70, 272 commutative justice 312–13 comparative law use 474–6 compensation causation indeterminacy 187–90 as duty 136–7 personal injury 204–5 and risk 98–9 as tort goal 166 see also damages Compensation Act 2006 177 compulsion 495–7 consequentialist considerations 263–4 constructive trust, unjust enrichment and 366–72 contract as basis for restitution continuing regime 489–91 unenforceable contracts 491–3 voidable/terminable contracts 491

breach of 226 contents of 205–6 goals 3 married women 34–8 penalty default 273 rights 134, 153–6 Roman law 200, 202, 207 unenforceable 491–3 Cook v Lewis 169 Cooper v Hobart 91, 110–11, 302 corrective justice action/court-ordered/ordinary rights and 120–2 approaches 261, 275–7, 314 basis 51–2, 114–15 causation indeterminacy 174–9 deterrence and 315 distributive justice, separation 52–4 equality and 54–7 as normative foundation 50–1 pluralism and 114–17 separation thesis and 52–4, 57–8 theorist approaches 196–7 as tort law approach 81 tort law as 166 see also duty of care, rights-based correlative rights 86–7, 150–1 court-ordered rights availability rules 129–30 corrective justice and 120–2 as external pluralism 119–20, 122–9 Kantian explanations 122–9 new right creation rules 130–3 rubber-stamp view of 129 Cran v New South Wales 300 created rights 133–5 crime victims, negligence claims 285–98 Commonwealth position 285–92 ombudsman function 292–4 see also domestic violence victims, negligence claims CTN Cash and Carry Ltd v Gallaher Ltd 486 Dagan, H 198, 273–4, 461, 469, 472 damages, see also compensation Deglman v Guaranty Trust of Canada 380 deontological considerations 263–4 desert deterrence and 321–3 fiduciaries and see fiduciary allowances deterrence causation indeterminacy 187–90 class identity issues 316–17 corrective justice and 315 desert-based 321–3 distributive justice and 315–16, 323 fiduciaries see fiduciary duties notional harm 319–21 private law role 311–12

Index relevance 317–18 systemic deterrence 328–9 tax principle 323–5 as tort goal 166 Deutsche Morgan Grenfell Group plc v Inland Revenue Commissioners 478, 484, 497–8, 501, 504–5 Devenish Nutrition Ltd v SanofiAventis SA 400–1 Dextra Bank and Trust Co Ltd v Bank of Jamaica 485 difference principle 318 Direct Payment Rule of compensation 98–9 discharge of a debt 493–7 disgorgement damages 375 meaning 341 see also fiduciary allowances dishonesty, fiduciary allowances and 360–3 dispositive classification 219–21 distress damage feasant 156–7 distributive justice corrective justice, separation 52–4 deterrence and 315–16, 323 meaning 313–14 Dobson v Dobson 22–3 domestic violence victims, negligence claims 294–8 see also crime victims, negligence claims Donoghue v Stevenson 79, 89–90, 101, 104, 110–11 Drummond Wren, Re 23–4 duty of care, rights-based background 79–84 correlative rights 86–7, 150–1 economic efficiency-based goal 87–90 foreseeability 107–12 proximity issues 90–1 rights debate 84–5 risk and see risk Dworkin, R 66, 71–2, 197, 246, 403 economic efficiency goal 87–90, 113, 139–40, 151–2, 198–9 Edelman, J 333, 343, 346, 375 Edmunds v Wallingford 503–4 Edwin Hill v First National Bank 161 Eldridge v British Columbia 42 Elguzouli-Daf v Commissioner of Police for the Metropolis 299 enrichment acceptability 423, 443 benevolent 426–8 survey of 423–8 taxable 424–6 see also restitution; unjust enrichment Epstein, R 101, 402 estoppel claims 226–7 ethic minorities, use of negligence claims 307–9

511

European Convention for the Protection of Human Rights 146–9 European Union, harmonisation of law 207–8 Evans v City of Johnstown 396 Eve, Re 23 evidential damage 175–7 Exall v Partridge 496–7 expository classification 217–19 external pluralism see action rights; court-ordered rights; ordinary rights external/internal approaches see obligations, internal/external approaches Fairchild v Glenhaven Funeral Services Ltd 167, 169–73, 174–9, 182, 189, 266 Farah Constructions Pty Ltd v Say-Dee Pty Ltd 381–2 fiduciary allowances background 341–2 causation 345–8 desert deterrence and 350–1 as justification of allowances 355–60 meaning 352–5 dishonesty and 360–3 honest defendant 363–4 justification 344 limitation of 360–4 remoteness 348–50 see also account of profits fiduciary obligations 325–38 constructive trust and 366–72 deterrence 325–6, 329–32, 339 systemic 328–9 gains-based remedies 365–6, 383–8 no conflict duty 332–8 positive/performance distinction 326–8 see also account of profits; obligations Fleming, J 174 Fletcher, G 51–2, 83, 85 formalism, instrumentalism and 261–2, 272–4 Forsyth-Grant v Allen 393–4 Fourteenth Amendment 24–5 fraudulent misrepresentation 319–20 Frazer Edmiston Pty Ltd v AGT (Qld) Pty Ltd 358–9 free acceptance 465–6 free-riding 427–8, 439–40 Fried, C 195 Friedmann, D 401 Fuller, L 1 functionalism 4–5 gains-based remedies/relief 365–6, 383–8 see also under nuisance, gain-based relief Gambling Act 2005 499 Gardner, J 100 gift 500–1

512

Index

goals coherence 141 legal debate 1–2 levels of analysis 2 relevance 4–5 shaping process 2–4 Goff, R 441, 457 Goldberg, John 83–4, 108–9 Gordley, J 197 Gordon, WJ 427–8, 440 Grantham, RB 349, 474 Gray v Motor Accident Commission 276–7 Gregg v Scott 180–3, 186, 189 Guinness plc v Saunders 344, 360 Hand formula 87, 151–2 Hardy, Thomas, Mayor of Casterbridge 35 harm anti-harm wrongs 397–400 notional 319–21 relative frequency 97–107 Restatement of the Laws of Tort (American Law Institute), Third: Liability for Physical Harm 84–5, 88–9, 108 Harris, J 353, 355 Harris v Digital Pulse Pty Ltd 361–2, 378 Hart, HLA 193 Hedley, S 467, 473 Heward v Eli Lilly & Co 399–400 Hill v Hamilton-Wentworth Regional Police Services Board 284, 285–8, 289–92 passim, 295, 296, 299–300, 301–7, 310 Hohfeld, WN 142–3, 404 Holmes, OW 2–3, 84, 112 Honoré, T 404 Hotson v East Berkshire Area Health Authority 93–6, 97–8, 104, 175, 180–3, 185, 186–7 human rights as civil rights 146–9, 163 Human Rights Act 1998 20, 72–3 Hurd, HM 104–6 immanence of right 314 innate rights 133 insider/outsider distinctions 193–4 instrumentalism formalism and 261–2, 272–4 as tort law approach 81–2, 87–90 internal pluralism 133–8 internal/external approaches see obligations, internal/external approaches interpretive theorist approaches 141, 197, 246–8, 250–1 Irish Constitution 20 Jaggard v Sawyer 392, 405–6 Jane Doe v Metropolitan Toronto (Municipality) Commissioners of Police 22, 289–92, 295

JD v East Berkshire Community Health NHS Trust 162 Jones, G 441, 457 Kamloops (City) v Nielsen 289 Kant, I 55–7, 87, 117, 212 doctrine of right 122–9 Katz, A 273 Keech v Sandford 368, 369, 372, 374 Kelly v Solari 483 Keren-Paz, T 198 Kerrison v Glyn, Mill, Currie and Co 480, 485 Kienzle v Stringer 265 Kleinwort Benson Ltd v Lincoln City Council 488–9 Kraus, J 402–3 Kronman, AT 314 LAC Minerals Ltd v International Corona Resources Ltd 365, 366, 368–70 law as integrity 197 Law Reform (Married Women and Tortfeasors) Act 1935 37, 38 Lawrence, Stephen 387 legal classification see classification legal personhood background 27–8 definition 28n physical integrity protection 32–4 private law content 30–8 importance 38–9 property/contract 34–8 public law equality and 28–30 importance 39–44 reasonable person test 41–2 Lewin on Trusts 335 liberties, rights and 149–53 Linden, A 3, 4, 165–6, 283 Lion Laboratories Ltd v Evans 277–8 Lipkin Gorman v Karpnale 457 Lister v Stubbs 372, 376, 384–5 loss of chance see under, causation indeterminacy Lumbers v W Cook Builders Pty Ltd 506–8 McCarthy, D 98–100 Macclesfield Corp v Great Central Railway 503 MacCormick, N 5, 70–1, 271–2 McFarlane v Tayside Health Board 263–4, 271 McGhee v National Coal Board 167, 168–73, 178, 188–9 MacPherson v Buick Motor Co 79–80, 104, 108–10 Maitland, F 200 March v E & MH Stramare Pty Ltd 265 Marmo v Tyson 394–5

Index

513

married women contract 34–8 legal personhood see legal personhood tort law and 32, 37 Married Women’s Property Act 1882 36–8 Marshall, Donald 308 Maskell v Horner 501–2 Mason, K 398 Mayor of Casterbridge (Hardy) 35 Menefee, SP 35 merger doctrine 32–3, 34–5, 37 Mill, JS 200 mistake as basis of restitution 493–5 payments 436–8 Mitchell, Charles 350 monist rights, explanations 113, 114 moral injury theory 276 morality academic moralism 66–7 correlative rights 86–7, 150–1 moral rights 144–5 non-instrumentalist 82–3 of risk 92–6 More, MS 104–6 Morgan v Ashcroft 491–2 Moses v Macferlan 498–9 Mount Isa Mines Ltd v Pusey 268–9 Muir v Alberta 42–3 Murad v Al-Saraj 326

precise norms derivation 406–7 premises 401–2 property as absolute 405–6 unitary conception 403–5 proportionality 414 vagueness of norms 410–12 OBG decision 252–3 obligations definition 445n fiduciary see fiduciary obligations imposition 274–5 internal/external approaches emergence 195–9 interaction 193–5, 213–14 polarisation issues 208–13 public/private blending 203–8 Victorian approach 199–202 rules of 449–51, 456 ombudsman function in tort law 283–4 for victims of crime 292–4 for wrongly accused 305–7 ordinary rights corrective justice and 120–2 as external pluralism 118 Osman v Ferguson 286–7 O’Sullivan v Management Agency and Music Ltd 357–8 Owen v Tate 503

Naffine, N 461 narratives see unjust enrichment, narratives National Lottery Rule of compensation 98–9 NC Corff Partnership Ltd v OXY USA Inc 395–6 necessity 157–8 negligence law negligent (police) investigation 284 see also duty of care, rights-based Neyers, J 252–3 Nicomachean Ethics (Aristotle) 54, 57–8 non-instrumentalist approaches 82 notional harm 319–21 Nozick, R 425 nuisance, gain-based relief 389–419 all benefits/wrongs availability 408–10 anti-enrichment/anti-harm wrongs, distinction 397–400 background 389–90 case law treatment 391–7 conclusion 419 internalising externalities 413–14 market price considerations 412–13 meaning 390–1 permanent deprivation of rights 414–15 practice proposals 415–19 property rights condition 400–7 exceptions 407 legal rules 402–3

Palsgraf v Long Island Railroad 79–80, 89–90, 92, 101, 103–4, 108–10 Parkinson v College of Ambulance 492 Pavey and Matthews Pty Ltd v Paul 492 penalty default 273 Perdue, W 1 Perre v Apand Pty Ltd 257 personal injury compensation 204–5 personhood see legal personhood Peter v Beblow 367 Pettkus v Becker 365–6, 366–8, 369, 377, 380–2 Phipps v Boardman 357 physical integrity, married women 32–4 pluralism 114, 140–1 corrective justice and 114–17 external see action rights; court-ordered rights; ordinary rights internal 133–8 police negligence liability 283–310 accountability promotion 309–10 crime victims see crime victims, negligence claims defensive policing issue 304 discretionary nature of investigations 302–4 domestic violence victims 294–8 litigation flood issue 304–5 negligent investigation 284 ombudsman function of tort law 283

514

Index

policy 261–80 background 261–2 bipolarity and 272–9 coherence and 271–2, 279 common law method 269–70, 272 consequentialist/deontological considerations 263–4 consistency and 271 definition 262–3 denial of rights 277–9 fairness/justice considerations 265–7 formal considerations 264–5 institutional constraints 268–9 remedies and 275–7 role 267–8 political culture 67–9, 74 Pollock, F 199 Portman Building Society v Hamlyn Taylor Neck 489–90 Posner, R 66–7 presumptive positivism 452–5 Pretto-Sakmann, A 196 priority disputes 158–9 private law, theories of 312–14 private-public law distinction argument for 48–9 background 47–8 conclusion 74–5 hard cases 70–2 legal doctrinal 49–52 philosophical-normative 65–9 political culture 67–9, 74 project pursuits 58–65 separation thesis 52–4, 57–8 use 72–4 see also public-private law relationship privileges, rights and 144, 156–7, 163 property law, married women 34–8 property rights 134, 156–63 see also under nuisance, gain-based relief Prosser, W 84 public-private law relationship 17–45 constitutional law issues 17–18 direct/indirect horizontal approaches 20–1 debate 17 hierarchy of values approach 19–25 interaction 26–8, 45 interactive approach 26–8 mutually-constitutive approach 18–19 personhood and see legal personhood see also private-public law distinction Quarles v Little Energy River Co R v Jackson 34 Rawls, J 318 Reading v Attorney-General 336 Reid v Ford Motor Co 398–9

395–6

remedies commercial opportunities 374–5 extent 275–7 gains-based 365–6, 383–8 see also under nuisance, gain-based relief rights 135–8 self-help 157 see also compensation; damages remoteness, fiduciary allowances 348–50 Restatement of the Law of Restitution (American Law Institute) 415–19 Restatement of the Laws of Tort (American Law Institute) Second 169 Third: Liability for Physical Harm 84–5, 88–9, 108 restitution damages 375 rules approaches to 449–51 guidelines 428–9, 451–6 mistaken payments 436–8 in practice 435–6 predictability 429–32 presumptive positivism 452–5 public values 432–5 self-regarding conferral of benefits 439–43 see also enrichments; unjust enrichment Rickett, CEF 349, 474 rights action see action rights approach 141–2, 261 conflict 143–4 contract 134, 153–6 correlative 86–7, 150–1 court-ordered see court-ordered rights created 133–5 definition 142–3 denial as policy 277–9 duty of care and see duty of care, rights-based explanations for categories 113–14 instrumentalist/non-instrumentalist 113, 114, 138 human rights 20, 72–3 as civil rights 146–9, 163 innate 133 liberties and 149–53 monist rights 113, 114 moral 144–5 necessity 157–8 ordinary see ordinary rights priority disputes 158–9 privileges and 144, 156–7, 163 property 134, 156–63 see also under nuisance in relation to law 139–42

Index remedial 135–8 sceptical explanations 113 third-party 159–63 tort and see under tort law rights theorist approaches 196 Ripstein, Arthur 100, 101 risk duty of care and 97–107 forseeability and 108 imposition 177–9 morality of 92–6 risk-taking see unjust enrichment, risk-taking Riss v New York 293–4 Robert Addie & Sons Ltd v Dumbreck 158–9 Roman law 200, 202, 207 comparative law use 474–6 Roxborough v Rothmans of Pall Mall Australia Ltd 380–1, 381–2, 481, 490–1 rule-deterrence 315, 321–2, 334 rules background 445 definition 446 departure from 448–9 determinate/indeterminate 446–7 see also restitution, rules Russell-Stanley Corp v Plant Industries Inc 396 Rylands v Fletcher 102 Said v Butt 160–1 Saiman, C 382 Samuel, G 473 sceptical rights explanations 113 Schauer, F 452–3 Scheppele, KL 458 Scotland Act 1998 73 Scottish Equitable plc v Derby 470 self-help remedies 157 self-regarding conferral of benefits 439–43 Sempra Metals Ltd v IRC 498 separation thesis 52–4, 57–8 Serhan Estate v Johnson & Johnson 399 Sheehan, D 499–500 Shelley v Kramer 24–5 Sher, G 356 Sherwin, EL 385–6 Skyring v Greenwood 469 slavery codes 30–1, 35, 39, 144–5 Smith v Chief Constable of Sussex Police 287–8, 298 Smith v Eric S Bush 257 Soulos v Korkontzilas 366, 368, 372, 377, 379, 384 South African Constitutional Court 21–2 Stapleton, J 184–5, 198, 258, 274 Steam Saw Mills Co Ltd v Baring Bros & Co Ltd 494

515

Steyn, J 2 Stocznia Gdanska SA v Latvian Shipping Co 485 Stoke-on-Trent Council v W & J Wass Ltd 392–3 storytelling see unjust enrichment, narratives Summers v Tice 169, 174 Surrey CC v Bredero Homes 392, 400 Sutcliffe, Peter 292 systemic deterrence 328–9 Tamanaha, B 5 Tame v New South Wales 299–300 Tang, HW 500 tax principle 323–5 taxable enrichments 424–6 taxonomist approaches 196 see also classification Taylor v Chester 492–3 Taylor v Motability Finance Ltd 490 third-party rights 159–63 Thompson, JJ 101 Thompson, Re 329 tort law concepts 248–59 background 240 choice factors 243–7 classical-form 239–40 coherence and 255–6 interpretive theorist approaches 246–8, 250–1 policy-based approaches 256–7 prototype-form 240–3 rights-based strictures 252–7 scepticism 257–8 summary 259 corrective justice as 166 goals 3–4, 81–2, 165–6, 248–50 instrumentalist or corrective justice approaches 80–3 limits on 149–53 married women and 32, 37 moral rights 145 ombudsman function 283–4 relational character 85–6 undue influence 462–3 United States Civil code 294 Constitution 20, 24 dominant approach to tort 79–80, 83–4 slavery codes 30–1 unjust enrichment absence of basis analysis 477–9, 508 common law and 484–6 complexity 486–7 pyramidal nature 482–4 risk-taking and 482

516 security of receipt 487–8 settlement/compromise 491 state of authorities 484 uncertainty 488 autonomous legal person 459–66 basis 477, 508 application 506–8 compulsion 495–7 contract see contract, as basis for restitution definition 479 discharge of a debt 493–7 explanatory 480 gift 500–1 mistake 493–5 natural obligations 498–500 objective 488–9 partial failure 481 res judicata 493 statutory authority 497–8 subjective assessment 480–1 timing 481–2 voluntary transfers 501–4 by subtraction 365–6, 379–80 by wrongs 365–6 causative classification 223–4, 236–7 constructive trust and 366–72 contract see contract, as basis for restitution free acceptance 465–6 narratives autonomous legal person 459–66 mapping metaphor 471–6 shopper scenario 467–71 use 457–9, 476 risk-taking 482 allocation of risk 506 application 506–8 mental state of claimant 504–6 security of receipt 487–8 settlement/compromise 491 themes 458–9 as top-down reasoning 380–2 unjust factors 459–61, 466, 474–5 pyramidal structure 482–3 see also enrichment; restitution utilitarianism 314

Index value pluralism 430–1 vicarious liability 319–20 victims see crime victims; domestic violence victims Vincent v Lake Erie Transportation Co 157–8 vindicatio 224 Virgo, G 346, 474 voluntary transfers 501–4 Waddams, S 314 Wagon Mound (No 2) 99 Warman International Ltd v Dwyer 345–6, 359, 360 Watson, A 475–6 Weinrib, EJ 79 on corrective justice 115, 117, 197, 253, 255, 275 on court orders 122–9 on damages 136 on deterrence 315 on goals 1–2 on moral rights 82–3, 87, 102–3 policy considerations 263, 273 on restitutionary liabilities 432–5 separation thesis 52–7, 58, 74–5 Weir, T 184 West, R 461–2, 466 White v Jones 257 wife sales 35 Wightman, J 198 Williams, G 4 Wilsher v Essex Health Authority 171, 175, 178, 181–2, 189 women see married women Woolwich Equitable Building Society v IRC 505 wrongly accused, negligence claims 299–309 Commonwealth position 299–305 ombudsman function 305–7 use by ethic minorities 307–9 Wrotham Park Estate v Parkside Homes 391–2, 400 Yorkshire Ripper 292 Yuen Kun Yeu 110 Zhu v Treasurer of New South Wales 161 Zipursky, B 83–4, 100, 101, 108–9