Risk and the Law

Risk and the Law

Natural and man-made risks have long been recognised as vital conditioning factors in the formation of social institutions and the conduct of social life. In this volume internationally recognised experts examine in detail the implications in practice of the modern concept of risk in particular legal fields. The chapters explore the ways in which the law in its many branches can accommodate, manage and reduce the extent of risk in the modern ‘Risk Society’, matters of pressing importance for the development of all branches of law in all jurisdictions. The fields of activity affected by the issues discussed include law, medicine, insurance, state security and public health. The collection also contributes to comparative legal studies in respect of risk and the law, presenting a perspective which has largely been neglected outside the works of general theory. Thus the topics considered range from the civil law of injuries in Germany, and the food law of the European Union, through sales of goods, including international sales, in English, German and French law, to the English law of torts. Risk and the Law, written by specialists who are authorities in their fields, will be of interest to academics and students who are interested in new developments and ideas regarding the relationship between risk, law and social change in many different fields. Gordon R. Woodman is Emeritus Professor of Comparative Law at the University of Birmingham, UK. Diethelm Klippel is Professor of Private Law and Legal History at the University of Bayreuth, Germany.

Risk and the Law

Edited by Gordon R. Woodman and Diethelm Klippel

First published 2009 by Routledge-Cavendish 2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN Simultaneously published in the USA and Canada by Routledge-Cavendish 270 Madison Ave, New York, NY10016 Routledge-Cavendish is an imprint of the Taylor & Francis Group, an informa business This edition published in the Taylor & Francis e-Library, 2008. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.” © 2009 selection and editorial matter, Gordon R. Woodman and Diethelm Klippel; individual chapters, the contributors All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data A catalog record for this title has been requested ISBN 0-203-89129-5 Master e-book ISBN

ISBN10: 0–415–47149–4 (hbk) ISBN10: 0–203–89129–5 (ebk) ISBN13: 978–0–415–47149–7 (hbk) ISBN13: 978–0–203–89129–2 (ebk)

Contents

The contributors Preface INTRODUCTION

1 Law and risk: an introduction

vii xi 1

3

TIM KAYE

PUBLIC LAW AND CRIMINAL LAW

2 Risk decisions in German constitutional and administrative law

21

23

PETER M. HUBER

3 The problem of de-individualisation in the risk society

36

OLIVER LEPSIUS

4 Risk decisions in cases of persisting scientific uncertainty: the precautionary principle in European food law

53

RUDOLF STREINZ

5 Risk and criminal law

75

GERHARD DANNECKER

PRIVATE LAW

6 The assumption of risk

83

85

ANSGAR OHLY

7 Risk and predictability in English common law

95

TIM KAYE

8 Transfer of property and risk of loss in French, English and German law ULRICH SPELLENBERG

114

vi

Contents

9 Transit risks in CIF contracts – meaning and categories

132

KOJI TAKAHASHI

10 Costs and risk: recent developments in the English law of costs

146

KEITH UFF

EMPLOYMENT AND SOCIAL SECURITY

159

11 The risk of sickness in German labour and social insurance law

161

DIETMAR BOERNER

12 ‘The butcher’s cart and the postman’s bicycle’: risk and employers’ liability

172

GEORGE APPLEBEY

13 The limits of individualisation in the risk society: social security in the customary laws of immigrant communities

184

GORDON R. WOODMAN

Index

201

The contributors

George Applebey is a Lecturer in Law at the University of Birmingham. He holds an LLB degree from the University of Glasgow, Scotland, and an MCL from Tulane University Law School, New Orleans. He is the author of a textbook on contract law and numerous writings on discrimination law, civil procedure (small claims) and health and safety at work. He served for six years as a member of the UK government’s Industrial Injuries Advisory Council. Dr jur. Dietmar Boerner has been a Researcher in Social Law and the Economics of Health Services, and a Privat-Dozent (Reader) at the Faculty of Law and Economic Sciences, University of Bayreuth. He is now a Judge of the Gemeinsamen Bundesausschuss, an institution which handles issues concerning health professionals, hospitals and similar institutions. Prof. Dr Gerhard Dannecker was Professor of Penal Law, Criminal Procedure and Information Law at the University of Bayreuth, until in 2007 he accepted a call to the University of Heidelberg to become Professor of Penal Law and Procedure with Special Reference to European and International Aspects. He has published extensively in his fields of specialisation. Prof. Dr Peter M. Huber was Professor of Public Law at the University of Bayreuth. In 2003 he became Professor of Public law and Legal Philosophy at the Faculty of Law at Ludwig Maximilians University, Munich. He has combined his academic posts with that of a Judge of the Thüringen Upper Administrative Court. Prof. Tim Kaye was a Senior Lecturer in Law at the University of Birmingham and is now a Professor of Law at Stetson University College of Law, Florida. He holds an LLB degree from the University of Sheffield and a PhD from the University of Warwick. He has published work in the fields of torts, jurisprudence and education law. Prof. Dr Oliver Lepsius has been Professor of Public Law and General and Comparative State Law at the University of Bayreuth since 2002. He was

viii The contributors formerly Professor of Public Law at the University of Heidelberg. He studied law at the Universities of Bonn and Munich and obtained an LLM from the University of Chicago. Among his recent publications are: Besitz und Sachherrschaft im öffentlichen Recht [Ownership of Land and Movables in Public Law], Tübingen: Mohr Siebeck (2002); Steuerungsdiskussion, Systemtheorie und Parlamentarismuskritik [Taxation Debates, Systems Theory and Parliamentary Criticism], Tübingen: Mohr Siebeck (1999); and Verwaltungsrecht unter dem Common Law. Amerikanische Entwicklungen bis zum New Deal [Common Law Constitutional Law. American Developments to the New Deal ], Tübingen: Mohr Siebeck (1997). Prof. Dr Ansgar Ohly studied law at the University of Bonn, obtained an LLM from the University of Cambridge and wrote his doctoral thesis as a scholar of the Max Planck Institute for Intellectual Property, Competition and Tax Law, Munich. He then became a Senior Researcher there, taught at the University of Karlsruhe, and in 2001 took up his present position as Professor of Private Law and Intellectual Property Law at the University of Bayreuth. His recent publications include: The Law of Comparative Advertising – Directive 97/55/EC in the United Kingdom and Germany, Oxford: Hart Publishing (2000, with M. Spence); ‘Volenti non fit iniuria’ – die Einwilligung im Privatrecht [Consent in Private Law], Tübingen: Mohr Siebeck (2002); and Privacy, Property and Personality – Civil Law Perspectives on Commercial Appropriation, Cambridge University Press (2005, with H. Beverly-Smith and A. Lucas-Schlötter). Prof. Dr Ulrich Spellenberg studied history at the University of Marburg and law at the Universities of Marburg and Göttingen, working then as a Researcher in the Universities of Bielefeld and Munich. In 1978 he became Professor of Civil Law at the University of Hanover, moving in 1979 to become Professor of Civil Law, International Private Law and Comparative Law at the University of Bayreuth, where he remained until his retirement in 2006. From 1979 he has also been Professeur invité at the University of Bordeaux I. He has published extensively on a wide range of subjects arising from his fields of interest. Prof. Dr Rudolf Streinz was Professor of Public, International and European Law at the University of Bayreuth until August 2003. Since then he has been Professor of Public and European Law at Ludwig Maximilians University, Munich. He has written extensively on food law in the European Union, including in 2006 ‘The Legal Situation for Genetically Engineered Food in Europe’ (with Jan Kalbheim), in Prof. Dr Knut J. Heller (ed.), Genetically Engineered Food (2nd edn), Weinheim: Wiley-VCH. Dr Koji Takahashi holds the degrees of LLB and LLM from Kyoto University and LLM and PhD from the London School of Economics. He was

The contributors

ix

a Research Fellow at the Institute of Maritime Law and the Law Faculty, University of Southampton, from 1999 to 2001, and a Lecturer in Law at the University of Birmingham from 2001 to 2003. He is now Professor of Law at Doshisha University Law School, Kyoto. He has published work in all his fields of interest, which include international trade law, private international law, international civil procedure and English law. He is the author of Claims for Contribution and Reimbursement in an International Context: Conflict of Laws Dimensions of Third Party Procedure, Oxford: Oxford University Press (2000). Keith Uff has been a Lecturer in Law at the University of Birmingham since 1972. He holds the degrees of BA, BCL and MA from the University of Oxford and is a Barrister-at-Law. He is Executive Secretary General of the International Association of Procedural Law. His recent publications include: ‘Common law arbitration: an overview’ (2004) Journal of International Dispute Resolution, 10–15; ‘Remedies of the defrauded principal after Attorney General for Hong Kong v. Reid’, in D. Feldman and F. Meisel (eds.), Corporate and Commercial Law: Modern Developments, Lloyds of London Press (1996); ‘Access to Justice: Lord Woolf’s Final Report, Procedure and Evidence’ (1997) 16, Civil Justice Quarterly, 17; and ‘Security for costs against European Community plaintiffs’ (1996) 15, Civil Justice Quarterly, 193. Prof. Gordon R. Woodman is Emeritus Professor of Comparative Law at the University of Birmingham. He holds the degrees of LLB, LLM and Ph.D from the University of Cambridge and Dr Jur. (h.c.) from the University of Bayreuth. He worked in law faculties in Ghana and Nigeria from 1961 until 1976, when he joined the University of Birmingham. His fields of interest are law in Africa, customary law, the diffusion of the common law, and the theory of legal pluralism. He is the editor of the Journal of Legal Pluralism. His more recent publications include: ‘Comments on the construction of political order: social contract theories and anthropological observation’, pp. 15–24 in Marie-Claire Foblets and Trutz von Trotha (eds.), Healing the Wounds: Essays on the Reconstruction of Societies after War, Oxford: Hart Publishing (2004); (co-edited with Ulrike Wanitzek and Harald Sippel) Local Land Law and Globalization: A comparative study of peri-urban areas in Benin, Ghana and Tanzania, Münster: LIT Verlag (2004); and ‘The Involvement of English Common Law with Other Laws’, pp. 477–500 in Christoph Eberhard and Geneviève Vernicos (eds.), La quête anthropologique du droit: Autour de la démarche d’Étienne Le Roy, Paris: Éditions Karthala (2006).

Preface

This volume is one of the fruits of a collaboration during the past 15 years between the Faculty of Law and Economics of the University of Bayreuth, Germany, and the Birmingham Law School at the University of Birmingham, United Kingdom. A link between the two institutions was formally established in 1994 with a programme for exchange visits by students, and was extended in 2001 to regularise and increase the numbers of exchange visits by members of the academic staff. We were encouraged by the success of this academic collaboration to organise a conference of members of the two institutions. This took place in March 2003 at Schloss Thurnau near Bayreuth. This volume consists of papers given in early draft form at that meeting and subsequently developed. The topic ‘Risk and the Law’ seemed apt to fulfil our objectives. These were to hold a discussion related to a current concern of legal development, but also to range over a wide variety of branches of law in both of the jurisdictions. The participants were already aware of the large volume of literature on the modern ‘risk society’, and most of us had already considered how our research in our own fields might be related to this. But we were aware also that many, if not all, branches of the laws of Germany and of England, of the European Union and of the international commercial community could be considered from this perspective. We believed that it would be useful to all of us to look outside our specialisms to learn how other branches of the law were dealing with the same categories of issues. The possibility of learning about corresponding developments in another national jurisdiction was a considerable bonus in this exploration. We did not aim to conduct highly detailed or specific comparative studies, but we hoped that, just as we might benefit from hearing about comparable issues arising in other branches of our national laws, we would also learn from hearing about these issues arising in other legal systems. We thus heard papers examining legal policies towards risk on topics ranging from consent in the civil law of injuries (in Germany) and food law (of the European Union) through sales of goods, including international sales (in English, German and French law), to labour law (both German and British), to mention only a selection. At the end of two days of intensive

xii

Preface

discussions on prepared papers we believed that our objectives had been attained to a high degree. It is hoped that readers of this volume will agree. We are grateful to all the contributors for preparing their papers for publication, and to Dr (now Professor) Tim Kaye for writing an introductory essay. Our gratitude is also expressed to the University of Bayreuth, which met all the expenditure required for us to meet at Schloss Thurnau. Our hope is that this book will prove as interesting and stimulating to readers as the exchange of ideas has been to the contributors. Diethelm Klippel University of Bayreuth Gordon R. Woodman University of Birmingham

Introduction

1

Law and risk: an introduction Tim Kaye1

From status to contract It was an Englishman writing in the nineteenth century, Sir Henry Maine, who famously observed that the transition from feudalism to industrial capitalism marked ‘a movement from Status to Contract’.2 Inherent in this claim was the notion of mobility, in two senses. First, improving communications meant that it was becoming easier to travel and thus to trade, so that wealth could be accumulated faster than ever before and without the need for bloodshed. Second, since these opportunities were becoming available to ever greater proportions of the population, people could no longer be expected necessarily to remain within the social class into which they had been born. Social status was thus no longer the fixed bedrock of society: generalised commodity exchange, facilitated by the law of contract, enabled everyone to climb the social ladder: Not many of us are so unobservant as not to perceive that in innumerable cases where old law fixed a man’s social position irreversibly at his birth, modern law allows him to create it for himself . . .3 But such opportunities, perhaps celebrated most openly across the Atlantic as the ‘American dream’, inevitably bring with them ever more numerous sources of potential harm. Most such hazards are not, of course, matters of life and death in the way that the perennial feudal issues of war, disease and famine so clearly were. But, as Gordon Woodman points out in his contribution, perceptions change. As life expectancy increases, new hazards arise. Many illnesses – like cancer or heart disease, for example – are not particularly dangerous to someone who is very likely to die before he or she reaches the age of 30 in any event. Yet they are significant hazards today. Moreover, these ‘new’ hazards are considerably more numerous and often difficult to envisage in advance. ‘In advanced modernity the social production of wealth is systematically accompanied by the social production of risks.’4 To put this another way, the movement from status to contract implies at the same time a movement from certainty to risk.

4

Risk and the Law

Given the centrality of risk-taking within industrial capitalism, a fact which has caused many non-lawyers to talk of the ‘risk society’,5 it is perhaps rather odd that, at least until recently, the role and concept of risk have received relatively little attention from lawyers. It seems instead that risk has been considered more suitable as an object of study for sociologists and economists. Perhaps this lack of systematic legal analysis can be at least partially explained by the fact that there is no one area of law by which it can be said to be more or less wholly encompassed. Unlike contract, which – irrespective of the particular national or supra-national jurisdiction – is self-evidently a legal construct with its own rules and principles, there is no clearly demarcated legal subject which can be said to deal solely and exclusively with risk. Legal studies obsessed with doctrinal analysis have therefore largely overlooked the role and significance of risk. This would not have impressed Maine. As Raymond Cocks has recently emphasised: He believed that in seeking to understand law the best results could be achieved by making constant references to non-legal topics. Ultimately, law had to be accounted for and criticised in non-legal terms. After all, a man who wrote about progress had at some stage to write about legal change and was then confronted by the fact that law did not create itself and was not changed by itself.6

Paucity of analysis of risk Moreover, where risk has been mentioned as a factor underlying the law, the analysis attempted has often been lazy and superficial. The word ‘risk’ is sometimes employed in English law, for example, as explaining in a throwaway line the basis for the doctrine of vicarious liability, according to which an employer is legally liable for the acts and omissions of his employees acting in the course of their employment. A typical example can be found in one of the leading student textbooks on torts: A more plausible explanation of strict liability is that it operates as a loss distribution mechanism. Accidental damage arising from the materialisation of a risk inherent in a particular activity is paid for by the person or enterprise carrying on the activity. That person is in the best position to spread the loss via insurance and higher prices for the products that the activity creates, and so the true social cost of these products is borne by the consumers in small amounts. Vicarious liability is a good example of this process.7 But not only does the author never use the word risk again, he fails to explain why – at least until the decision of the House of Lords in Lister v Hesley Hall Ltd8 – there was such confusion in the law as to when the concept of vicarious liability could be successfully invoked by a claimant eager to receive full

Law and risk: an introduction

5

compensation for harm or loss sustained as a result of someone else’s wrongdoing. An effective loss distribution mechanism would require that employers should not be able frequently to escape liability, as they could before Lister. It would also require that the law be clear as to the identity of the relevant defendant, for otherwise unnecessary costs are added into the equation. Yet, before Lister, it was extraordinarily difficult to predict the outcome of a case on vicarious liability. Interestingly, the judges in Lister itself hardly invoked the notion of risk to justify their decision to simplify the law, but instead based their views on the idea that the employer had somehow taken on a responsibility towards the victim (a line of argument more commonly associated with primary, rather than vicarious, liability). Moreover, as Cane has pointed out: just saying that the law is concerned with allocating risks does not answer the question of to whom particular risks ought to be allocated or why certain risks are allocated in a particular way. This question alerts us to the fact that although the language of risk-allocation is often associated with strict liability in tort law, the difference between faultbased and strict liability is not that the latter allocates risks of injury but the former does not, but that they allocate risks of harm in different ways.9 Accordingly, five chapters in this collection deal head-on with precisely the question of who does and/or should bear the risk in various situations. Dietmar Boerner explores the issue of who in Germany shoulders the risk of an employee being rendered unfit to work through illness or injury, whilst Ulrich Spellenberg and Koji Takahashi examine the position in various commercial transactions. Of course, whilst consent has always been one of the main threads running through the doctrine of the law of contract in both Germany and the UK, its role in the law of tort has always been considerably more problematic. Yet, as Ansgar Ohly argues, the way in which the law deals with apparent consent to the running of a risk remains a fundamental issue even outside contract law. This has major implications as to who bears the risk of being liable for the costs of litigation, an issue discussed by Keith Uff.

‘Fault-based’ and ‘risk-based’ liability A thoughtful analysis of the interaction between risk and law has been provided by the aptly named German, Erwin Deutsch. Echoing Maine’s analysis, Deutsch’s argument is that German law – at least, German civil law – has moved (and continues to move) from fault-based liability towards riskbased liability.10 In case this terminology should cause some confusion, it should be noted that Deutsch’s use of the term ‘fault’ is rather different from the way in which that term is commonly understood by English lawyers. To

6

Risk and the Law

him ‘fault-based liability’ is liability which is imposed when a legal person suffers injury as a result of another person’s inherently dangerous behaviour. In other words, it represents a legal view that risky activity is itself unwelcome (unless, of course, luck prevails and it causes no harm), so that when someone undertakes such behaviour they are automatically at fault. To many an English lawyer, this looks, ironically, like ‘no fault’ or strict liability. Yet to Deutsch, the ‘fault’ resides in the fact that the defendant chose to behave in a manner which put others at risk. A good example of this in English law is the rule in Rylands v Fletcher, where the construction of a reservoir and the storage of water within it was considered to be such a novel and dangerous activity – a ‘non-natural use of the land’ – that the occupier was held liable for any damage caused by the escape of water, even where the escape was not itself due to the conduct of the occupier.11 Liability which Deutsch would characterise as ‘risk-based’, meanwhile, is not founded on the view that risky activity is automatically reprehensible. According to this approach risk is simply a fact of life in a risk society, whether the activity involves the carrying out of well-established, routine tasks or something entirely new. The fact that harm has been done to a third party is not sufficient in itself to warrant the imposition of liability. When liability is risk-based, what matters is the degree or level of risk taken. Only if harm has resulted because an unacceptable level of risk has been taken will there be any imposition of legal liability. Translating this approach once again into English law helps to explain the modern interpretation of the rule in Rylands v Fletcher adopted by the House of Lords in Cambridge Water,12 where it was held that ‘the requirement of non-natural use [of land] embodies an evaluation of risk . . . where, inter alia, the gravity of the harm threatened is weighed against the utility of the defendant’s conduct’.13 This evolution of the rule in Rylands v Fletcher suggests that Deutsch’s analysis may be as applicable to English civil law as it is to German civil law. It also perhaps implies that it may represent a wider truth common to all legal systems operating within industrial capitalism, and that the movement from the fault-based to the risk-based model in civil law may have consequences in the fields of criminal and public law as well. Certainly the analyses of Peter Huber and Gerhard Dannecker suggest that it does, although this then opens up a new question as to which risks are best dealt with by each branch of the law. By way of contrast, the role of customary law is considered by Gordon Woodman. Deutsch does not claim that fault-based liability has been completely usurped by risk-based liability. He contends instead that the best approach is ‘to present risk-based liability as a complement to fault-based liability’.14 Indeed, just as status may still sometimes have more significance than any contractual agreement,15 there remain instances where fault-based liability is predominant, such as where the magnitude of the potential harm is very high (so that it is still felt that the activity is inherently dangerous)16 or the cost of prevention of harm is very low.17 One characteristic of duties based on

Law and risk: an introduction

7

Deutsch’s conception of fault – i.e. strict liability to English lawyers – is that they become highly specific, both in English and in German law.18 In this way they are identified as clear exceptions to the general norm of risk-based liability. Moreover, fault-based norms come to be identified by a symbolically important change in the language used to describe them. Thus in English, whilst risk-based liability implies issues of ‘law’, fault-based (strict liability) norms are downgraded to forms of mere ‘regulation’. Even criminal offences of this nature are rarely referred to as such but are instead normally described as ‘regulatory offences’, suggesting something more in the nature of a regular business expense than a ‘true’ crime. Whilst, as Dannecker points out, German criminal law – in theory, at least – avoids imposing criminal penalties where no fault can be proven, it is still the case that German civil law uses different terminology to reflect the fact that fault-based obligations are based on highly specific duties of care (Schutzpflichtungen) rather than on general norms (Verpflichtungen) which require respect for constitutionally enshrined, legally protected interests (Rechtsgüter).19

Liability and the ‘harm principle’ ‘In short,’ Deutsch argues, ‘one can say that fault-based liability prohibits, while risk liability permits’, so that ‘proper care [has been] redefined from risk-avoidance to risk-management’.20 This sounds remarkably like the guidance on risk assessment offered by the Health and Safety Executive (HSE) in the UK – and discussed at greater length by George Applebey – which, although asking whether it may be possible to eradicate a risk altogether, recognises that a more realistic approach is usually to focus on management of the risk. It advises: Evaluate the risks and decide whether existing precautions are adequate or more should be done . . . Even after all precautions have been taken, some risk usually remains. . . . ask yourself . . . how can I control the risks so that harm is unlikely?21 Nevertheless, Deutsch may be overstating the case when he argues that faultbased liability prohibits, while risk liability permits. Fault-based liability does not actually prohibit certain forms of conduct. What it does is automatically make the actor legally liable as soon as someone else suffers harm or loss as a result of that conduct. No defences can be offered to absolve the actor from legal responsibility. It would be more accurate, perhaps, to associate this type of liability with John Stuart Mill’s ‘harm principle’, according to which individuals are free to conduct themselves as they choose so long as they do no harm to others.22 If the conduct in question causes no harm, then there is no difficulty. But once harm is sustained, liability follows without question. Mill did not accept a defence of reasonable care any more than could be countenanced by the fault-based liability model. (The fault lies

8

Risk and the Law

in engaging in the activity in the first place, since either it was inherently dangerous or, if it was not, then it must have been undertaken very badly for otherwise no harm would have been caused.) The risk-based liability model, meanwhile, treads a very different path. It accepts that the victims of the conduct of others will have no legal basis for any complaint wherever it can be shown that the perpetrator of the harm took a justified risk (e.g. by taking reasonable care). This philosophy accepts that ‘accidents will happen’ and that the mere fact that someone has suffered an injury does not mean that someone else should pay. In other words, if there is deemed to be some social utility in the risk being undertaken, then this may at times override the interests of certain individuals who suffer as a result. At the same time, it should be recognised that an important limiting factor for Mill was that only physical harm justified the imposition of some form of sanction on the individual who had caused it. Nowadays it is generally recognised that other forms of harm – such as psychological injury and purely economic loss – may also merit the imposition of liability in certain circumstances, whilst the category of physical harm itself has also been extended well beyond the bounds of what Mill would have recognised. It is now the case, for example, that English law considers a wrongful denial of education to be a form of actionable personal injury.23 So whilst the move towards what Deutsch calls risk liability implies that the types of activity for which liability will automatically be imposed by the law may have diminished in number, the forms of harm for which liability can be imposed have concurrently – and significantly – increased. It might also be added that the number of people capable of being injured in a manner recognised by law through any one act has increased correspondingly. However, this does not mean that the movement from ‘fault-based’ to ‘risk-based’ liability involves a movement from the individualism of Mill to some form of collectivism. On the contrary, both systems of liability are essentially individualistic. Both depend upon analysis of the circumstances of the individuals involved to determine the legal position of the parties. But they look at different aspects of the relationship between actor and victim and, in so doing, cause the focus to shift from victim to perpetrator. Whereas fault-based liability is determined by whether or not the potential victim has suffered any harm, it is the conduct of the defendant which is the determining factor for the purposes of a system which operates risk-based liability. So long as this conduct is deemed acceptable, then there can be no liability. A recent English case which illustrates this point in striking fashion is Woodbridge School v Chittock.24 In 1996 Simon Chittock, then aged 17, suffered a serious injury in a skiing accident in an Austrian ski resort. He had been on a week’s school trip organised primarily for younger pupils, but which he and two other older boys had been allowed to join after an agreement between the school and the boys’ parents. This permitted the boys to ski – unsupervised but on designated runs – provided that they reported to a

Law and risk: an introduction

9

teacher at regular intervals. The evidence in the case was that the boys had skied competently all week. However, they had deliberately skied off-piste on one occasion, about which they were severely reprimanded by the teacher in charge. He reminded them that skiing off-piste was expressly not permitted in the school’s agreement with their parents and was also a risk not covered by the school’s insurance. The boys promised that it would not happen again. However, while skiing on a designated run the following day, Simon fell and fractured his spine, causing him to be permanently paralysed from the waist down. At the original trial the judge had found the teacher’s disciplinary response to Simon’s unauthorised skiing off-piste to have been negligent. He held that, after that incident, either the boys’ ski pass should have been confiscated, or else they should have been made to ski under supervision. This was reversed on appeal, however, where the Court of Appeal emphasised that the correct standard against which the teacher’s conduct should have been measured was that of a reasonable teacher in similar circumstances. This did not impose, in Auld LJ’s words, a duty to ensure a pupil’s ‘safety against injury from skiing mishaps’. It involved simply ‘a duty to take such steps as in all the circumstances were reasonable to see that [Simon] skied safely and otherwise behaved in a responsible manner’.25 Of course, reasonable teachers might each have responded differently to Simon’s off-piste skiing. As Auld LJ explained: The proper question, therefore, was not whether a more severe response than a reprimand would have been appropriate, but whether a reprimand was, in the circumstances and having regard to the School’s duty of care, within the reasonable range of options for [the teacher] to adopt. Thus, in order to succeed in establishing a breach of that duty, Simon had to establish that [the teacher’s] decision was one that no reasonable teacher in [the teacher’s] position could have reached.26 The Court of Appeal found unanimously that the teacher’s decision simply to reprimand the boys for skiing off-piste was indeed within the range of reasonable responses to the boys’ behaviour which a teacher might have made. Accordingly, the court held that the teacher involved had not been negligent and so it reversed the trial judge’s decision. The risk-based liability model is thus content to allow individuals to engage in certain forms of conduct despite the fact that they create a risk of harm to other people, even when such harm – including, as the case of Simon Chittock demonstrates, really serious harm – does in fact arise. What determines liability according to this model is the appropriateness or otherwise of the conduct of the defendant. This establishes the acceptable level of risk and inevitably implies that the law is prepared to tolerate a certain potential for harm to others. But having denied that Deutsch’s model of risk-based liability suggests a move to a more collectivist society, it needs equally to be emphasised that it

10 Risk and the Law also has little in common with Posner’s economic analysis of law at the other end of the political spectrum.27 As a German, it is perhaps not surprising that Deutsch does not accept Posner’s contention that case law is an inherently more appropriate means of resolving issues of liability than is legislation. On the contrary, he expresses no particular preference for one over the other but sees them both as legitimate sources of legal obligations. But, perhaps more importantly, he sees no reason to assert that the evaluation of conduct which causes harm to a third party must necessarily be driven by principles of economic efficiency. Clearly some socially acceptable standard will be required to inform the evaluation, because otherwise it will be impossible to know whether an individual’s risk management has been good enough to avoid being held legally culpable. But efficiency need not play a significant role at all. Other criteria, determined by non-economic factors such as morality and religion, or by environmental or other economic measures like ‘polluter pays’ or an evaluation of who can best control the risk, may be much more important instead.28 The basis on which the law may decide that harm or loss should be borne by a victim or wholly or partially re-distributed to others does not have to be so-called Pareto efficiency,29 even if it is clearly important that the outcome should not unduly stifle risk-taking and economic initiative.

What precisely is ‘risk’? Risk is therefore a rather slippery concept, because it tends to mean both different things to different people and different things in different contexts. As Rudolf Streinz demonstrates in the context of food law, some care needs to be taken to ensure that important issues of principle are not obscured simply by different usages of the term ‘risk’, or by its being confused with other, related terms such as harm, danger and hazard. Indeed, it is essential that the law is sufficiently sophisticated as to be able to disentangle these elements, for otherwise it is unlikely to be able to cope consistently and efficiently with the demands which a risk society places upon it. To attempt to avoid such linguistic confusion is not just a matter of semantics: there is in fact a significant, qualitative difference between a risk on the one hand and harm, danger and hazard on the other. Ideally, in modern society just as much as in days of yore, we would much rather avoid harm. But our attitude to risk is another matter altogether: it is the taking of risk that makes our world go round. As Beck points out, ‘risks are no longer the dark side of opportunities, they are also market opportunities’.30 Successful risk-takers are lauded as entrepreneurs and rewarded with profits, whilst unsuccessful risk-takers are encouraged to try again by laws which enable former bankrupts effectively to start with a clean slate. Moreover, many sports and leisure activities are considered attractive precisely (or at least partly) because of the inherent risk involved. It is certainly difficult to explain the appeal of an activity such as ‘free diving’ in any other way. There can be no risk if there is

Law and risk: an introduction

11

no hazard or danger to be overcome. But if – in principle at least – risk is good and harm is bad, it is vital that the law can distinguish effectively between the various concepts involved. Distinguishing risk from other concepts may also help us to focus on precisely what risk is. Niklas Luhmann argues that the concept of risk refers to the possibility of injury attributable to one’s own decision, whereas a hazard describes the possibility of harm caused by an external source. Thus ‘[t]he smoker may risk cancer, but for others it is a hazard. The car driver who takes a chance when overtaking behaves in just the same way . . .’31 For Luhmann, therefore, a risk is thus the product of a conscious decision, whereas a hazard or danger is something inflicted on someone without their prior agreement. Yet, in reality, it is hard to draw such distinctions. Most smokers do not smoke in pursuance of any conscious decision, but because they are addicted to nicotine, whilst those who choose to live with them have made a voluntary decision of their own. Driving a car too demands a conscious decision: does this not implicitly indicate a willingness to risk the possibility of poor driving by others? A preferable use of terminology, it is submitted, is to be found in the usage of the UK Health and Safety Executive. This defines hazard, for example, as ‘anything that can cause harm (e.g. chemicals, electricity, working from ladders, etc)’. Whilst the HSE does not define ‘danger’, it seems safe to treat it as a synonym for hazard, especially since both words tend to be rendered as ‘Gefahr’ in German. Hazards and dangers may either be inherent in the act concerned, or they may be merely incidental or entirely extraneous factors which come into play only if the act is carried out in an inappropriate manner. Risk, meanwhile, ‘is the chance, high or low, that somebody will be harmed by the hazard’.32 When harm results, we may refer to the incident as an ‘accident’, even if the harm done was intentional or reckless.33 A fuller analysis of the differences and interplay between these different concepts can be found in the contribution by George Applebey. Sometimes it will be the potential for harm to the actor himself which causes the act to be considered as risky, whereas on other occasions it is the possibility of harm to others which makes us want to use that term. Skiing, for example, is thought to be a risky activity primarily because it may lead to the death or injury of the skier herself, whereas surgery is considered risky – even when performed with great skill by the surgeon – because of the potential for harm to the patient. Driving a train, meanwhile, is considered risky not because of the activity in question – statistics consistently show that it is the safest form of motorised travel known to humanity – but because, on the relatively rare occasions when things do go wrong, large numbers of victims may be involved. Finally, whilst the risk-liability model is concerned with the possibility of harm, it does not always require that anyone need actually sustain harm before liability can be imposed. In some jurisdictions the mere exposure of someone to an unacceptably high risk of harm is itself sufficient for the imposition of liability. The criminal offence of ‘reckless endangerment’

12

Risk and the Law

recognised in many US jurisdictions is a good example of how the law may respond to unacceptable conduct which has not necessarily led to actual harm. In Washington, for instance: A person is guilty of reckless endangerment when he or she recklessly engages in conduct . . . that creates a substantial risk of death or serious physical injury to another person.34 It is not necessary that death or serious physical injury be actually inflicted in order for reckless endangerment to be proved to the satisfaction of a court. Gerhard Dannecker argues that the mere ‘endangerment of legally protected interests can be punishable’ under German law too. So the essence of risk is probability, chance or likelihood of harm, rather than the actual occurrence of harm. If the risk materialises it then translates a hazard or danger into harm (however narrowly or broadly that category may be drawn). This means that risk cannot be considered, in and of itself, to constitute harm.35 It is, rather, a measure of prediction of whether harm will actually result. However, this notion of risk is made more complex by the fact that, in many instances, it describes a cocktail of different degrees of probability. Indeed, although Perry defines it as the probability of harm occurring multiplied by the magnitude of the harm that might eventually occur,36 even this formulation involves an ambiguity because the concept of ‘magnitude’ of harm may refer either to the degree of harm suffered by one individual or to the number of victims affected. Yet risk is not predicated on fault: an activity can be risky without any blameworthy conduct by any individual. In combination, all these factors mean, for example, that (because of the establishment of a number of fail-safe systems) there may be a small chance of nuclear radiation being allowed to leak from a nuclear power plant. So the risk of legally inappropriate conduct is low. Even if it materialises, there may still be only a small chance that it will cause harm (perhaps because the location of the power plant is very remote). But if any harm does arise (perhaps because of strong winds carrying the radiation a long way), then its effects are likely to be extremely serious. Forms of conduct outside the field of nuclear energy production will of course yield different combinations of possibilities. Thus an activity may be said to be risky in one or more of four ways: a b c d

because of the likelihood that it could endanger someone if done badly; because of the likelihood that it could endanger someone whether done badly or not; because any harm suffered is likely to be very serious; and/or because any harm suffered is likely to be sustained by a large number of people.

Probability – and thus risk – is usually described in the terminology of

Law and risk: an introduction

13

either fractions or percentages. The theoretical parameters are thus zero or 0 per cent at the lower end – meaning that there is absolutely no risk of harm being sustained at all – and 1 or 100 per cent at the top end, in which case harm is a stone-cold certainty. But these parameters are entirely atypical precisely because the outcome in either case is a known fact. True risk implies uncertainty, and the chance of real-life risks actually materialising must therefore be found somewhere between these points: the higher the fraction or percentage, the greater the likelihood that a risk will crystallise into a real accident. Once it has been established how great the risk is, it is then possible to decide whether steps need to be taken to manage it more effectively. As was explained above, the move from fault-based to risk-based liability means that: Whereas previously proper care consisted in behaving so as to avoid danger, the concept of care has evolved through the twentieth century, so that now proper care consists in dealing appropriately with danger . . .37 As Beck puts it, the paradigm of the risk society is based on solving the questions of how can risk be ‘minimized, dramatized or channelled’.38

The perceptions and ‘reality’ of risk There is, of course, no rule which dictates how great the probability must be in order for the activity in question to be deemed sufficiently risky to warrant special measures to combat or manage the risk. Nor is there a level of risk which is universally considered too high to be worth running under any circumstances. The same problem underlies both these issues. The point of view from which judgements about risk must be taken is very much open to debate. It seems obvious that, in a risk society, those who take risks should have an accurate perception of the risks they are running, whether they be employers, employees or simply people engaged in fairly mundane everyday activities like driving a car. But deeper reflection shows that our notion of what is obvious may actually be highly problematic. In particular, it is extremely unclear what an ‘accurate perception of risk’ really means. When ‘accurate perception’ is urged upon us in any other context, it is taken to mean that individual beliefs should correspond closely with what actually happens. Thus public concern over crime rates in the UK, for example, is often said to be unwarranted because most measures of crime statistics suggest that crime is actually falling. If our perceptions became more closely aligned to the objective truth, then we would all be a little less anxious. Those stressing the importance of an ‘accurate perception of risk’ often take a similar line. Thus it has been suggested, for example, that: Research investigating workers’ risk perceptions has revealed that subjective risk estimates differ from objective risk data in a number of

14 Risk and the Law systematic ways. It is important that organisations are aware of biases in employees’ subjective assessment of risk, so they can develop mechanisms to reduce the impact of these biases.39 But this argument makes a fundamental mistake. ‘Objective risk data’ do not measure risks; they measure accidents. As was explained above, there is a fundamental distinction between a risk and an accident. An accident is an objective fact: it occurs when some form of risky conduct leads to harm. On the scale of probability, it counts as 1, or as 100 per cent: it is now a certainty because, as hindsight tells us, it did actually happen. The number of similar accidents can be added up for a given period of (say) a year and this total can be compared with previous or subsequent years to show whether the rate of accidents is increasing or declining. But this tells us only part of what we need to know in order to be able to quantify the risk associated with that same activity. Whether harm is sustained as the result of an activity is only one factor in the equation. Others include the level of endangerment where no harm was actually sustained, the number of people endangered or harmed in any one incident, and the frequency with which the activity is performed inappropriately. For these reasons, many organisations now log not just actual accidents but also so-called ‘near misses’. It has been objected that: studies suggest that the characteristics or features of a hazard, such as the amount that is known about the hazard, controllability of consequences and the impact on future generations, influence their perceptions. These hazard characteristics can be summarised under two headings, ‘dread’ and ‘unknown’. Nuclear reactor accidents are an example of a hazard that scores highly on the dread factor. Hazards that score highly on the unknown factor include electrical fields and DNA technologies. If a hazard scores highly on the dread or unknown factors, then people are more likely to misjudge the risk from this hazard.40 But there is actually no inherent misjudgement in such perceptions. Both ‘dread’ and ‘unknown’ factors imply that, if harm eventuates, the possibility of widespread and serious harm is significant. This is exactly the sort of thing that should be included within an evaluation of risk. Consequently, there appears to be no reason why human evaluations of risk, which try to take all these factors into account, should correspond closely with the single measurement of accident statistics. As Beck says, with more than a hint of irony: the seemingly harmless distinction between risks and the perception of risks gains importance . . . The monopoly on rationality enjoyed by scientific hazard definition stands and falls with this distinction. For it puts forward the possibility of objectively and obligatorily determining

Law and risk: an introduction

15

hazards in a specialized fashion and through expert authority. Science ‘determines risks’ and the population ‘perceives risks’. Deviations from this pattern indicate the extent of ‘irrationality’ and ‘hostility to technology’.41 Those who argue that our perceptions should follow more closely the ‘objective risk data’ also overlook another significant factor. In a society where risk liability predominates, we all learn to manage risks. This means not that we shy away from danger, but that we take it on: we try to deal with it in a manner we consider appropriate. As a result, irrespective of the other risk factors which they overlook, accident statistics need to be interpreted with a degree of sophistication. For it may be that they reveal more about our ability to manage the risks involved than about the level of risk, hazard or danger inherent in a given activity. It is therefore crucial to be sceptical of a simplistic interpretation of the statistics which demonstrate, for example, a long period of relatively accident-free operation of the railways in the UK until the late 1980s. Perhaps, by then, travel by rail had indeed become significantly less risky than earlier in the century. But it seems most likely that the diminishing number of accidents was less a consequence of reduced inherent risk and much more a product of an increased competence in managing those risks. The personnel just became more familiar with – and more expert in – handling a system of mass rail transit. Thus the period between the rail disasters at Clapham in 1988 and Potter’s Bar in 2002, during which a number of other serious accidents occurred, should not necessarily be seen as a period in which inherent risks proliferated – it is, after all, unlikely that they were significantly more numerous than before – but as a period in which the ability of relevant personnel to manage the risks sharply declined. The reasons for this decline are controversial – ‘risk habituation’ or complacency, loss of experienced staff, lack of finance – but irrelevant for present purposes. What is relevant is that they expose a major weakness with accident statistics. They are historical and so can tell us about the past rather than about the future. Yet a risk refers to some future possibility. As every competent investment adviser knows, extrapolating from the past may inform our view of likely future occurrences, but that is useful only if trends remain constant. Statistics can tell us very little about our current or future ability to manage risks, even where the level of hazard inherent in a particular activity apparently remains fairly static and relatively easy to identify. The trap of failing to acknowledge that the level of skill and success with which risks are managed should itself be considered a factor in risk assessment is all too easy to fall into if ‘objective risk data’ are treated as actual measures of risk.

Subjective and objective risk Ansgar Ohly describes what he calls ‘the long and winding road’ of German tort law as it has struggled to decide between a subjective and an objective

16

Risk and the Law

approach to the issue of risk-taking. This raises the perennial question of whether there can be such a thing as an ‘objective’ risk. It is submitted that the answer is both yes and no. Without an objective conception of risk it would be impossible for a court to adjudicate on whether an actor had taken appropriate steps to manage that risk. If we could not agree that there is a risk implicit in the use of an anaesthetic, for example, it would be impossible to evaluate the conduct of a particular surgeon or anaesthetist in an operating theatre. This may be the prime use of accident statistics within the context of risk evaluation. If it can be shown that accidents with anaesthetics have happened in the past, then it is clear that there must be some sort of risk against which some precautions should be taken by the professionals involved. But the role of objective risk is limited: as we have seen, such data cannot identify the level of the risk in question. Two things can, however, generally be established objectively. The first, as the example of the use of anaesthetics in surgery demonstrates, is whether or not some risk exists. The level or degree of risk may be controversial, but whether a risk does or does not exist is usually fairly clear (though not always, as explained further below). Second, it is often possible to make objective judgements as to whether the level of risk has risen or fallen. Again, this is not always true. But there is clearly general agreement that, for example, surgery is much more efficacious and less likely to lead to complications if practised with sterile equipment than with implements that have simply been washed in soap and water. Disagreements no doubt abound both as to what the precise level of risk in either case might be and whether it would be worth running the greater risk in extremis – such as in a war zone or remote environment where there is no facility for equipment to be sterilised properly – when balanced against the risk of further harm (or even death) were surgery to be delayed or avoided altogether. But such disagreements can in no way invalidate the proposition that it is possible to take an objective view on whether surgery is riskier with or without sterile implements. Surgeons in a British or German hospital who failed to use sterile equipment would clearly be guilty of serious professional malpractice and would be treated by the law accordingly, irrespective either of the precise level of added risk or of whether any harm actually resulted. Qualitative assessments about risk can thus be made objectively. Indeed, if objective evaluation were entirely impossible, it is arguable that the law would have very little role to play in regulating human risk-taking. The point is that, while even objective assessments of risks are always bound to be uncertain to some degree – and are frequently tied up with issues of professional judgement – they do provide a vital element of predictability. As Tim Kaye argues, uncertainty should not be confused with unpredictability. A risk is uncertain because it cannot be determined whether or not harm will actually result. But prediction does not deal in certainties: it deals in probability, which is precisely the currency of risk. A risk is thus objectively or qualitatively foreseeable even though it is impossible to know

Law and risk: an introduction

17

whether harm will in fact occur. For this reason, Kaye argues that any system of civil law within industrialised capitalism must reflect qualitatively objective assessments of risk and he provides a case study of the English law of negligence in an attempt to explain and justify this view. Quantitative judgements about risk are, however, a different matter altogether. The idea that it might be possible objectively to establish the chance of the occurrence of a specific risk is simply a contradiction in terms. Risk is, by definition, uncertain: the only certainties are where we know definitely that an accident either will or will not happen, in which case any talk of risk is meaningless. Since there is no means of calculating a risk in any quantitatively objective sense, it follows that all such evaluations of the level of risk are essentially subjective.42 In other words, it is for each actor to work out for himself what degree of risk he is likely to be taking on. Inevitably, we all make our calculations in a different way, just as we are prepared to tolerate different levels of risk. But perception is what counts here: risk assessments are made by subjective analysis of the information available.

The problem of lack of information This does mean, by definition, that any attempt at risk analysis – whether subjective or objective – meets a fundamental problem if there is a serious lack of reliable information about a particular issue. For there are some circumstances where it may not even be clear whether a risk really does exist. The abortive search for weapons of mass destruction in Iraq and the failure to detect any conspiracy to fly aircraft into the World Trade Center in New York on September 11 2001 are but two examples which may suggest that such situations are even increasing in number. As Peter Huber explains, German law explicitly recognises the possibility that a significant lack of information – for example, about alleged terrorist activity – may mean that it is simply not possible to know whether there is any danger or hazard at all. Without a hazard, there can be no risk. German law refers to such a situation as a ‘suspected danger’ (Gefahrenverdacht). One way that the law can deal with this lack of information is by adopting the ‘precautionary principle’ – explained by Rudolf Streinz – so as to ensure the protection of public health and safety so far as is practicable if what is often termed the ‘worst case scenario’ actually eventuates. The alternative to adopting the precautionary principle is to fall back on the standard approach of identifying an acceptable level of risk. The problems inherent in adopting this approach when there is a paucity of information about the potential hazard have been highlighted in typically colourful language by Ulrich Beck: A central term for ‘I don’t know either’ is ‘acceptable risk’. Let us spell out this term. In connection with risk distribution, acceptable levels for ‘permissible’ traces of pollutants and toxins in air, water and food have

18

Risk and the Law a meaning similar to that of the principle of efficiency for the distribution of wealth: they permit the emission of toxins and legitimate it to just that limited degree. Whoever limits pollution has also concurred in it. Whatever is still possible is, by social definition, ‘harmless’ – no matter how harmful it might be. Acceptable levels may indeed prevent the very worst from happening, but they are at the same time ‘blank checks’ to poison nature and mankind a bit. How big this ‘bit’ can be is what is at stake here. The question of whether plants, animals and people can withstand a large or a small bit of toxin, and how large a bit, and what ‘withstand’ means in this context – such are the delightful horror questions from the toxin and antitoxin factories of advanced civilization which are at stake in the determination of acceptable levels.43

Yet the precautionary approach too has its dangers, especially – according to Rudolf Streinz – when it is applied to issues outside the area of environmental law. Oliver Lepsius highlights precisely these dangers. He demonstrates how acting in a precautionary manner can lead to the law’s overcoming the problem of lack of information simply by asserting that a risk does exist. In so doing, it eliminates any possibility of subjectivism, leading to a different sort of problem, which Lepsius calls ‘de-individualisation’. This problem is compounded by the fact that, as Beck has pointed out: Precisely in dealing with risks, a variety of new social differentiations and conflicts emerge. . . . As the risk society develops, so does the antagonism between those afflicted by risks and those who profit from them. The social and economic importance of knowledge grows similarly . . . Thus new antagonisms open up between those who produce risk definitions and those who consume them. These tensions between business and the elimination of risks, and between the consumption and production of risk definitions, range across all areas of social action. Here lie the essential sources of the definitional struggles over the scale, degree and urgency of risks.44 (Emphasis in original) If nothing else, this collection of essays shows that every attempt to assess, manage or tackle a risk runs all sorts of risks of its own. But continuing to omit a discussion of risk from legal discourse is surely fraught with even greater dangers.

Notes 1 I am grateful to Gordon Woodman for his helpful and insightful comments on an earlier draft. 2 H.S. Maine, Ancient Law, London: John Murray, 1912, p. 182. 3 Ibid., p. 331. 4 U. Beck, Risk Society: Towards a New Modernity, trans. M. Ritter, London: Sage, 1992, p. 19.

Law and risk: an introduction

19

5 See especially Beck, n. 4 above; B. Adam et al. (eds.), The Risk Society and Beyond: Critical Issues for Social Theory, London: Sage, 2000. 6 R.C.J. Cocks, Sir Henry Maine: A Study in Victorian Jurisprudence, Cambridge: Cambridge University Press, 2004, p. 2. 7 M. Jones, Textbook on Torts, 8th edn, Oxford: Oxford University Press, 2002, p. 391. 8 [2001] 2 AC 215. 9 P. Cane, Atiyah’s Accidents, Compensation and the Law, 6th edn., London: LexisNexis Butterworths Tolley, 1999, p. 357. 10 E. Deutsch, ‘Das neue System der Gefährdungshaftungen: Gefährdungshaftung, erweiterer Gefährdungshaftung und Kausal-Vermutungshaftung’, Neue Juristische Wochenschrift, 1992, p. 73, trans. P. Larouche, reproduced in W. Van Gerven et al., Cases, Materials and Text on National, Supranational and International Tort Law, Oxford: Hart Publishing, 2000, p. 543. 11 Rylands v Fletcher (1868) LR 3 HL 330. 12 Cambridge Water Co Ltd v Eastern Counties Leather plc [1994] 2 AC 264. 13 S. Deakin et al., Markesinis & Deakin’s Tort Law, 5th edn, Oxford: Oxford University Press, 2003, p. 538. 14 Deutsch, n. 10 above, p. 543. 15 Maine asserted (n. 2 above, p. 331) that ‘indeed several of the few exceptions which remain to this rule [i.e. the movement from Status to Contract] are constantly denounced with passionate indignation’. 16 For example, a radiation leak from a nuclear power station, as in the Ionising Radiations Regulations 1999. 17 For example, the provision of basic safety equipment to employees, as in the Provision and Use of Work Equipment Regulations 1998. 18 See Deutsch, n. 10 above, p. 543 and, in English law, the highly detailed regulations issued under the Health and Safety at Work Act 1974. 19 See Bundesgesetzbuch § 823. 20 Deutsch, n. 10 above, p. 543. 21 Health and Safety Executive, Five Steps to Risk Assessment, Sudbury: HSE Books, 2002, p. 5. 22 See J.S. Mill, Utilitarianism, On Liberty and Considerations on Representative Government, London: J.M. Dent & Sons, 1972, p. 73. 23 Adams v Bracknell Forest Borough Council [2004] UKHL 29 (HL). 24 [2002] EWCA Civ 915. 25 Ibid., para 18(iii). 26 Ibid., para 21. 27 R. Posner, Economic Analysis of Law, 2nd edn, Boston: Little, Brown & Co., 1977. 28 Deutsch, n. 10 above, p. 545. 29 See V. Pareto, Manual of Political Economy, New York: Augustus M. Kelley, 1971. 30 Beck, n. 4 above, p. 46. 31 See N. Luhmann, ‘Verständigung über Risiken und Gefahren’, Die Politische Meinung, 1991, Vol. 4, 81, quoted in U. Beck, World Risk Society, Cambridge: Polity Press, 1999, p. 84. 32 Health and Safety Executive, Five Steps to Risk Assessment, Sudbury: HSE Books, 2002, p. 3. 33 For a similarly broad definition of the word ‘accident’, see Cane, n. 9 above, p. 3. 34 Revised Code of Washington, Title 9A, 36.050(1). 35 See S.R. Perry, ‘Risk, harm and responsibility’, in D.G. Owen, Philosophical Foundations of Tort Law, Oxford: Clarendon Press, 1997.

20 36 37 38 39 40 41 42 43 44

Risk and the Law Ibid., p. 322. Deutsch, n. 10 above, p. 543. Beck, n. 4 above, p. 19. M. Fleming and R. Lardner, ‘When is a risk not a risk?’, The Chemical Engineer, 8 July 1999, p. 1. Ibid.; see also The Royal Society, Risk, Analysis, Perception, Management, London: The Royal Society, 1992. Beck, n. 4 above, p. 57. For some reason Perry prefers the ugly word ‘epistemic’. See Perry, n. 35 above. Beck, n. 4 above, p. 64. Beck, n. 4 above, p. 57.

Public law and criminal law

2

Risk decisions in German constitutional and administrative law Peter M. Huber

INTRODUCTION To speak about risk decisions in German constitutional and administrative law means to look at these decisions from the perspective of public law, that is, of the legal order dealing with the relationship between the individual and the state. This limits the topic to the decisions, or lack of decisions, of the state and its authorities. In order to understand how public law has to deal with risks, it will be necessary first to clarify what is a ‘risk’ in the legal sense of the word, and second to place the notion of risk into the context of the constitution. Third, this chapter will look at the requirements which the legislator has to meet in making risk decisions, and to what extent the legislator is subject to judicial control. Having considered these, I shall conclude by showing how the legislator delegates risk decisions to the public administration, and what are the characteristics of risk decisions when they are taken by the administration.

THE DEFINITION OF ‘RISK’ To understand the dogmatic environment of the legal term ‘risk’, it is necessary to understand that it is closely linked to the legal term ‘danger’, which can be considered as a type of Archimedean point of the system of German administrative law.

The term ‘danger’ and the liberal concept of statehood After the turmoil of the French Revolution, the Napoleonic wars and the reestablishment of the ancien régime following the Vienna Congress in 1815, and under the influence of the liberal and democratic movement, the liberal concept of the state emerged. There had been signs of a trend towards a stronger social emphasis on social needs in political theory and legislation at that time. The French Constitution of 1793 contained a ‘right

24 Risk and the Law to work’ and the famous Allgemeines Landrecht für die Preussischen Staaten of 1 June 17941 contained many provisions which resembled modern labour laws, or social security directives of the European Union. However, this trend did not prevail. The liberal concept of the state saw it, not without reason, as the counterpart to the subject/citizen, whose life, liberty and property it was liable to threaten rather than protect. The state, only at this point in the history of continental Europe coming to be conceived of as a body corporate itself, separated from the person of the reigning monarch, was to be prevented from interfering with the rights, privileges and interests of its subjects/citizens except when expressly authorised by law (in the so-called legal reserve, Vorbehalt des Gesetzes). The only ground on which an interference was considered legitimate was the protection of people from danger and damage to life, liberty and property, either from outside or from inside society. This nightwatchman state (Nachtwächterstaat) was therefore allowed to have a military force because it had to fight danger from abroad. It was allowed a police force to fight dangers at home from criminals, fire, pollution from industrial plants and suchlike. Beside this, however, there was almost no legitimate function for the state. The most famous ruling reflecting this liberal concept of statehood, which still today can be considered the birth of modern public law in Germany, is the so-called Kreuzberg Judgment of the Prussian Supreme Administrative Court (Preussische Oberverwaltungsgericht) of 14 June 1882: In Berlin, in the district of Kreuzberg, there was – as there still is – a monument commemorating the victory over Napoleon in the wars of liberation. The police authority of Berlin had enacted an ordinance in which it was decreed that no building should be built so high that it blocked the view of this monument, because it was desired to maintain and foster patriotic feelings among the citizens. A land owner who wished to build a house of four storeys was therefore refused permission by the police authority. He sued the authority successfully. The Preussische Oberverwaltungsgericht ruled that the authority was not entitled under the general police provision (clause of para. 10, II, 17 ALR2) to stop land owners from building whatever they wanted. It held that the maintenance of an unspoilt view of a monument had nothing to do with the legal authority to counteract dangers to the public. i) Danger (Gefahr) Since that time German administrative law has developed the legal concept of danger (Gefahr) to a certain degree of perfection. It is used first and foremost in the area of police and security law, but not only there. Still today in many areas of administrative law the assumption that a ‘danger’ exists is decisive in providing legal power to state authorities. If a danger does not

Risk decisions in German constitutional and administrative law

25

exist, these authorities are in these areas not entitled to intervene by way of administrative acts. What, then, is a ‘danger’ in this respect? ‘Danger’ is considered to be a situation which, if not prevented from taking its course, can be expected with a sufficient degree of probability to lead to the violation of public security and order (öffentliche Sicherheit und Ordnung), i.e. damage to public or private goods. (‘Lage, in der bei ungehindertem Ablauf des Geschehens ein Zustand oder ein Verhalten mit hinreichender Wahrscheinlichkeit zu einem Schaden für das bzw. die Schutzgüter der öffentlichen Sicherheit (oder Ordnung) führen würde.’) The term ‘danger’ thus refers only to goods already in existence, so that police laws must not be used to achieve political goals such as the enhancement of social welfare or culture. In order to decide whether it is faced with a ‘danger’ justifying its action, an authority has to determine the probability of damage. This means that the authority has to make a forecast or prognosis. Seen from the point of view of the authority, the occurrence of damage must be likely. The test is not whether it is certain to occur or is imminent, but whether, according to the experience of a neutral observer, it is likely.3 In reaching a decision, it is necessary to observe a relationship between the value of the goods at stake and the degree of probability: the higher the value of the goods, the less may the degree of probability be, and conversely, the lower the ranking of the goods threatened, the higher the degree of probability must be. For this purpose the good of life ranks higher than liberty of movement, human dignity higher than freedom of assembly.4 ii) Nuisance (Belästigung) The term ‘danger’ as used in law has to be distinguished from the term ‘nuisance’, which we find in a number of statutes dealing with public security, but also in some concerned with environmental issues. The law traditionally speaks of ‘nuisance’ when referring to impacts which do not violate the rights and privileges of individuals but cause some sort of detriment. But unfortunately there are also references in legal contexts to ‘nuisances’ which amount to dangers. So much for the contribution of formal prescriptions to the certainty of the law. Two examples may be given. 1

2

§ 1 StVO (Federal Traffic Ordinance) requires that road users act in such a way that others are not annoyed to an unnecessary extent. However, if the traffic is affected by such a nuisance, this has to be considered a danger, with the result that police authorities are empowered to take administrative measures. § 24 read in conjunction with § 22 and § 3 I BImSchG (Federal Code on Air Pollution) states that authorities are entitled to issue administrative acts in order to avoid ‘substantial disadvantages’ (erheblicher Nachteile) or ‘substantial nuisances’ (erheblicher Belästigungen) emanating from

26

Risk and the Law such sources as church bells, fire brigade sirens and machines for the manufacture of concrete. The BImSchG thus classifies together dangers, substantial disadvantages and substantial nuisances.

Suspected danger (Gefahrenverdacht) If authorities are unable to make a prognosis because they lack the necessary information, but if there are certain indications that a disturbance of public security and order may be caused, the case is classified as a ‘suspected danger’ (Gefahrenverdacht). The difference from a case of ‘danger’ is that where there is a ‘suspected danger’ the authorities must secure further information before they can properly decide that there is a ‘danger’. Legal doctrine is not clear about how ‘suspected dangers’ should be dealt with. Some say that any such case has to be handled like any case where there is no danger and therefore no acts of administration interfering with the rights and privileges of individuals are authorised.5 Jurisprudence and the majority of legal scholars have taken a different view however. They consider a suspected danger to be a danger in the legal sense6 because it is no different from a danger except that it is based on a lower degree of probability. A second reason for taking the preconditions of ‘danger’ as present in these cases is that otherwise authorities would have to wait idly by until some damage occurred and this would not be a very reasonable application of the law.7 In order not to destroy the boundaries set by police law on administrative acts, the courts and those others who propound this opinion try to limit the legal scope of administrative acts based only on a suspected danger. Such acts may only be aimed at clarifying the situation, finding proofs for a thorough prognosis, but must not be aimed at removing the possible danger itself. In other words, they are limited to the initiation of a danger investigation (the so-called Gefahrerforschungseingriff ). Thus, for example, in the case BGHZ 117, 303/306 ff. the authorities suspected that a herd of calves might have been fed with hormones. As it was not clear whether this suspicion was well founded, it was disputed whether the calves could be slaughtered. Although the BGH (Federal Civil Court) did not use the term suspected danger,8 it seems to have approached the issue in the way recommended. It held that the slaughter of the calves was lawful because there was no other way to clarify the situation.

Risk The most severe problem arises in those situations where there is either insufficient probability of damage or insufficient scientific insight and experience to decide whether there is a suspected danger and the necessary information

Risk decisions in German constitutional and administrative law

27

cannot be acquired in a predictable or reasonable time. These problems arise if authorities have to decide whether particular drugs, industrial colours, food preservatives, exposure to a certain level of radiation, etc. are or are not harmful to health. State decisions in such cases entail risks and therefore are called risk decisions. These examples show that risk decisions are only loosely linked to the precautionary principle and are not identical to laws or administrative acts which seek to diminish the probability that damage may occur to certain goods such as life, health and property. Although it is true that, within the precautionary principle, at the end of the scale we arrive at the so-called ‘inherent residual risk’ (Restrisiko), there is an important difference between risk decisions and administrative acts imposing this inherent residual risk. It is that Federal Statute imposes the latter as a necessary burden stemming, for example, from the use of nuclear energy. Nevertheless, there are links between these two institutions of administrative law, as the following questions may show: Do the legislator or public authorities have to counteract risks? Under what circumstances are they entitled to impose such risks on people? To what extent are they subject to judicial review in those areas?

THE CONSTITUTIONAL FRAMEWORK OF RISK DECISIONS

Political theory After the experience of religious and civil wars in seventeenth-century Europe Jean Bodin9 and Thomas Hobbes10 laid the grounds for modern statehood, the latter formulating the basic paradigm, which still today contains its atavistic and most essential justification: Homo homini lupus. Although John Locke was able later to mitigate Hobbes’ harsh alternative between anarchy and tyranny, it has remained the first and foremost function of the state to protect its subjects/citizens from all kinds of dangers and risks.11 This function influenced the Bills of Rights in New England12 and the US constitution of 1787.13 It has also shaped political thinking in Europe. The provision of safety legitimised the state’s monopoly of force (staatliches Gewaltmonopol14) and laid the grounds for the exercise of liberty in an open society. On the basis of this assertion, the Federal Constitutional Court, in its famous Brokdorf decision, the leading case concerning the right of assembly (Art 8 Basic Law), ruled for example: A legal order which, after the suppression of the medieval right of selfhelp, reserved to the state a monopoly of the exercise of force, not least in the interest of weaker minorities, must strictly insist upon prevention [of the use of force against people or goods]. That is a precondition for

28 Risk and the Law the guarantee of freedom of assembly as a means of active participation in the political process, and, as the experience of street battles during the Weimar Republic shows, is also on that account not to be renounced in a free democracy, because the averting of violence provides a justification for measures which limit freedom.15 The ECJ too held in the case Commission v France, concerning the frequent blockades of French farmers, that the Single Market for goods, today guaranteed by Art 28 EC, was – and still is – threatened by the lack of determination of the French government to ensure the free traffic of agricultural goods.16 Today in Germany there are several hundred laws – laws of the Federation as well as state laws – aimed at protection from risks, and there have even been amendments to the Basic Law to enable the state to act further to provide safety, both for the people and for itself.17 Thus the state’s function of protecting citizens from risks has been unaltered since the dawn of modern statehood. This function has been concretised and developed according to the needs of post-industrial society, dependent on an advanced technical infrastructure, and it has gained in importance.18 Although this function of the state is the ideological and constitutional basis for the justification of the state’s existence, and though its reality seems self-evident and plain, it is necessary to stress that in the constitutional architecture of the Basic Law it ranks in second place after the principle of freedom. Risk decisions such as those concerning the licensing of drugs, genetic engineering or nuclear plants have to be justified because they put the regulation of risk under state control at the expense of the freedom of entrepreneurs, scholars and others.

The prevention of risk and the rule of law It is hardly surprising that after the experience of the omnipresent and authoritarian Nazi state the adoption of the Basic Law was accompanied by a paradigmatic switch in the German legal system. Although the original idea of stating in Art 1 of the Basic Law that ‘the state exists for man’s sake’ was dropped, since 1949 the state has no longer been considered the Archimedean point of the legal system. Rather this is the individual. Legal thinking gradually stopped approaching constitutional issues from the perspective of the state, the state system and its functions and began to look at them from the point of view of civil rights and liberties and the necessity of an authorisation for every interference by state authorities. This switch of paradigm has been so fundamental that even after 50 years of constitutional practice not all its consequences are really clear or generally acknowledged. Article 1 para 1 Basic Law, proclaiming the dignity of man to be inviolate, and Art 1 para 3 Basic Law binding the executive, the judiciary and – for the first time in German history – the legislature to respect the fundamental

Risk decisions in German constitutional and administrative law

29

rights laid down in Arts 1–19 Basic Law, were the technical instruments for this change of perception. Consequently, every act of state has become relevant to the issue of fundamental rights, and almost every law affecting the rights and duties of individuals is considered an incursion (Eingriff) into the fundamental rights of the citizen. As such it must be justified. Justification however is possible if the legislator or administration can identify dangers, threats or risks to public or individual goods which the law aims to remove. If the legislator fails to do this, the law in question has to be declared null and void by the Federal Constitutional Court. Looking at the state along these lines of liberal tradition means considering it primarily as a threat to individual rights and interests and entails a tendency to keep the state out of society as far as possible. It (theoretically) points to a reduction of the state’s role within society and to a strict requirement of legitimation for any of its actions – protection from risks being one legitimating reason among others. The most famous case in this respect is the so-called Apotheken-Urteil of 11 June 1958,19 which every German law student has to learn in his or her first semester. In this opinion the Federal Constitutional Court nullified a Bavarian law which had sought to impose a restriction on the number of pharmacies in order to preserve the economic profitability of those already established. Because the effects of that law on the freedom to practise a profession were quite heavy, and there was no present and imminent danger to public health or any other discernible public good if the number of pharmacies were to increase significantly, the law was considered not only an interference with the right, but also a violation of Art 12 para 1 Basic Law. This decision not only gave birth to the constitutional principle of proportionality (Grundsatz der Verhältnismäßigkeit)20 but also established that on the one hand the state’s function to protect its citizens from danger remained unchallenged yet on the other it was subject to a differentiated system of legal restrictions. The scope of state action thus depends on the degree of probability of danger, the nature and extent of the goods threatened and the types of liberty at stake.

The discovery of Schutzpflichten (1975–1990) This changed when during the 1970s people became more aware of the dangers coming from within society – environmental pollution, the dangers of nuclear energy, terrorism and other threats. All of a sudden it became clear that life, liberty and property were not only and not even primarily threatened by the leviathan (the state) but also and principally by the activities of fellow citizens or foreigners. The stimulus in legal principle was the abortion issue. In its first opinion concerning this topic, the Federal Constitutional Court held that the state was obliged to use the criminal law to prevent pregnant women, fathers,

30 Risk and the Law families and doctors from engaging in abortion.21 It later reduced this requirement and held that Art 2 para 2 and Art 1 para 1 GG required only that abortion should not be routinely treated as lawful and financed by public health insurance.22 Nevertheless this ruling remains of lasting importance because it stated that the fundamental rights contained in the constitution, the rights to life (Art 2 para 2 GG) and to human dignity (Art 1 para 1 GG), compelled the state to shelter and foster those who were too weak to secure by their own efforts the effectiveness of their fundamental rights. The duty of protection (Schutzpflicht) was developed to its fullest extent. Legal doctrine extended it also from dangers to risks, holding that the state was not only obliged to counteract dangers but also to exclude any reasonable sort of risk. This may be illustrated by nuclear energy law (AtG). As danger, in the tradition of the ancient Prussian ALR, is defined as a situation in which there is a high probability of imminent harm to life, liberty or property or to the institutions of the state, nobody would have been willing to accept a situation in which state surveillance of nuclear plants would come into effect only at that level, that is, immediately before an accident occurs. Instead, the legislator and the courts developed the ‘precautionary principle’, which compelled the state to reduce the risks to such a low degree of probability that according to all experience it seems practically impossible that a nuclear plant will cause any harm (the so-called ‘inherent residual risk’, Restrisiko).23 This duty of protection, initially formulated in favour of the unborn child, was later applied also in favour of hostages taken by terrorists,24 neighbours of nuclear plants, airports and genetic engineering factories, as well as consumers. The Schutzpflicht doctrine was extended to almost every branch of modern life, so that we find it today in the civil obligations between landlord and tenant, in labour law, consumer protection law and many other areas. Through this development the concept lost its close and direct connection with the state function of counteracting dangers. It came to be seen as the constitutional basis for social legislation and environmental protection, which today itself is regarded as a new state function.25 After reunification, when Parliament discussed whether the goal of environmental protection should be explicitly laid down in the constitution, many lawyers argued that, given the Schutzpflichten for life, liberty and property, this was unnecessary. However, the figure of ‘danger’ has kept its significance. It is especially important in respect of administrative discretion, since if there is a danger and not only a risk, the discretion of administrative bodies, whether they act or not, may be reduced to zero. Even more important, ‘danger’ is relevant to ‘standing’. Laws which protect the individual from ‘dangers’ contain a corresponding ‘subjective right’. This means under § 42 para 2 Administrative Courts Procedure Act (VwGO) that the endangered individual is given standing to apply for a legal remedy if the administration violates the law. Notwithstanding that, according to the prevailing theory, laws that only put the ‘precautionary principle’ into concrete form, that is, which seek only to

Risk decisions in German constitutional and administrative law

31

minimise ‘risks’, protect only the general interest and therefore do not enable the individual to go to court.

The change of instruments (1980–1995) In the 1980s the state to a certain extent began to change the instruments with which it sought to fulfil its tasks and functions. Government began to prefer warnings, recommendations and information to the classical instruments of acts of administration. These informal instruments reach more people, are rather flexible and less subject to judicial control than traditional ordinances or administrative acts. They may on one hand be considered as a new vehicle for the state’s function of counteracting danger. But on the other hand, they shed a special light on the importance attached to this function within the political and legal system. For it is obvious that the state has to resort to these doubtfully effective means only because it lacks the practical and legal possibility of achieving the same results by using the classical instruments such as laws, ordinances and administrative acts. Legal theory is still struggling with the problem of how to insert these instruments into the constitutional order.26 The administrative courts, especially the Federal Court of Administration, have held that the application of these instruments constitutes an interference with the subjective rights and legally protected interests of individuals. Therefore it must be based on an authorisation provided directly in the constitution or at least in a parliamentary statute (‘legal reserve’).27 The Federal Constitutional Court in two opinions of 26 June 200228 stated however that the fundamental rights laid down in the constitution do not give protection against detriment caused by such informal acts of the ‘preceptoral state’.29

RISK DECISIONS IN THE LAW-MAKING PROCESS – THE PROGNOSIS PROBLEM The principles just examined require a forecast to be made as to whether any given concrete situation will turn into a disturbance of public security and order. State authorities have to judge whether there is a sufficient degree of probability, or at least a significant risk of such a result, on the basis of the scientific knowledge available and general experience. This has far-reaching consequences for judicial control, as any prognosis can be wrong. As nobody knows the future, the legislator too has to act on the basis of forecasts which later may turn out to have been wrong. Therefore, the standard for judicial control cannot be whether a forecast was right or wrong, but only whether it was made with sufficient diligence. If it was, it is lawful and based on a valid prognosis (gültige Prognose). An error concerning the facts of the situation, however the state of the art or general experience, results in an unlawful, that is, invalid, prognosis, which affects the law in question.

32 Risk and the Law In areas where there is not enough scientific knowledge available, or there is a lack of experience – for example in respect of the amounts and effects of radiation from mobile phones, telecom plants or pharmaceutical products – the legislator is entitled to engage in a sort of experimental legislation. As long as he acts on the basis of a valid prognosis, having collected all the available information before taking the decision, he may take a risk decision and allow risks, even if they later turn out to be concrete dangers, without violating the constitution and the fundamental rights at stake. If his knowledge improves later and a risk turns out to be a danger, or a threat under the precautionary principle significantly bigger than before, he has a duty to amend the legislation (Nachbesserungspflicht).

RISK DECISIONS IN ADMINISTRATIVE LAW In most cases, however, Parliament does not take the risk decision itself but delegates it to the public administration. As there are so many and such heterogeneous risks, from nuclear energy to thousands of chemical substances, drugs and food, and as knowledge and experience concerning those risks constantly change, the legislature would be simply overcharged if it had to take all risk decisions itself. Consequently, it has to refrain from settling the details and be satisfied with formulating the principles and the major guidelines. Against this background we find such guideline legislation in the major part of environmental law, such as in the Federal law dealing with chemical substances (ChemikalienG), in the Federal law protecting plants (PflSchG), in the Federal law concerning food (LFBG 2005), in the Federal law on genetic engineering (GenTG) or in the Federal Code on Drugs (AMG). During the last 10–15 years a new type of administrative decision has thus been created. This is profoundly analysed by Udo Di Fabio, Professor of Public Law in Bonn and Judge of the Federal Constitutional Court, in his habilitation paper Risikoentscheidungen im Rechtsstaat, 1994. He uses the example of drug concessions, which today are issued by the Federal Institute for Drugs (Bundesinstitut für Arzneimittel), §77 Abs.1 AMG, when the case does not fall within the competence of the Commission or the Council of the EU according to Art 3 EU regulation No. 2309/93. He has shown that in these cases prognoses as to a ‘danger’ are frequently impossible because the administration and the scientific community have no experience of the new substance of which the drug is composed. In a detailed analysis he has defined four characteristics of a risk decision in administrative law: 1

The laws empower the public administration to interfere with individual rights and liberties under certain conditions that do not fulfil the requirements of ‘danger’, and allow them to weigh the advantages against the risks of the measures which the licence is dealing with. One

Risk decisions in German constitutional and administrative law

2

3

4

33

consequence is that entrepreneurial activities are limited to a greater extent than under the old police laws and therefore require authorisation by a parliamentary statute, falling into the category of the legal reserve. The interference with individual rights and liberties is further intensified by the fact that the authority granted to the administration also entails scope for discretion, administrative prognosis, etc., which is not entirely subject to judicial control by the administrative courts. To understand this legal argument, one has to know that, according to the prevailing doctrine in Germany, judicial control is a substantial part of every individual right and liberty.30 Thus any restriction on the intensity of judicial control is also considered a restriction on the fundamental right or liberty itself. Licences, concessions and similar administrative acts which can be considered to be risk decisions do not have the same administrative standing (Bestandskraft) as regular administrative acts, which means that they cannot be relied on to the same extent. The reason is that the public administration, in a manner similar to that of the legislature, might have to amend its decisions when it turns out that the risk permitted was too great or was even a danger. The consequence of these characteristics is that the licences and concessions in question change their appearance and start to resemble prohibitions, from which exemptions are possible (Verbot mit Erlaubnisvorbehalt).

CONCLUSIONS Risk decisions pose numerous problems in German constitutional and administrative law. Their mere existence proves that peaceful coexistence under the conditions of an industrialised, technologically advanced society can no longer be guaranteed merely by the enactment of a law in terms of the nineteenth-century police law. Although this branch of public law has not lost its importance, especially after the events of September 11, the state today has to deal with uncertainty in constantly broadening fields of modern life. Risk decisions are one instrument to deal with this problem. They have only just been invented. And there is still a lot of legal work to do.

Notes 1 Allgemeines Landrecht für die Preussischen Staaten, II, 1 § 185 et seq., concerning maintenance between husband and wife. 2 Entscheidungen des Preussischen Oberverwaltungsgerichts (Decisions of the Prussian Supreme Administrative Court), 9, 353/355 et seq. 3 Entscheidungen des Bundesverwaltungsgerichts (Decisions of the Federal Administrative Court), 28, 310 et seq.

34

Risk and the Law

4 Entscheidungen des Bundesverwaltungsgerichts, 47, 31 et seq. 5 V. Götz, Allgemeines Polizei- und Ordnungsrecht, Göttingen: Vandenhoeck & Ruprecht, 2001, p. 128. 6 Entscheidungen des Bundesverwaltungsgerichts, 39, 190/193 et seq. 7 Entscheidungen des Bundesverwaltungsgerichts, 39, 190 et seq. 8 Entscheidungen des Bundesgerichtshofes in Zivilsachen (Decisions of the Federal Civil Court), 117, 303/306 et seq., ‘Verdacht oder Anschein’. 9 J. Bodin, Les six livres de la République, 1577, cited in A. Bergsträsser and D. Oberndörfer (eds.), Klassiker der Staatsphilosophie, Stuttgart: K.F. Koehler Verlag, 1962, p. 150; K. Doering, Allgemeine Staatslehre, Eine systematische Darstellung, Heidelberg: C.F. Müller Verlag, 2000, § 8, p. 190. 10 T. Hobbes, Leviathan (first published 1651), edited with an Introduction by Michael Oakshott, Oxford: Basil Blackwell, Chap 17, p. 109: And Covenants, without the sword, are but words, and of no strength to secure a man at all. . . . [I]f there be no power erected, or not great enough for our security; every man will, and may lawfully rely on his own strength and art, for caution against all other men. 11 Chr. Calliess, ‘Sicherheit im freiheitlichen Rechtsstaat’, Zeitschrift für Rechtspolitik, 2002, 1/4; Chr. Linck, ‘Staatszwecke im Verfassungsstaat – nach 40 Jahren Grundgesetz’, Veröffentlichungen der Vereinigung der Deutschen Staatsrechtslehrer 47, 1990, 7/27. 12 E.g. Virginia Declaration of Rights (1776), Section One: ‘all men . . . have certain inherent rights . . ., namely, the enjoyment of life and liberty . . . and pursuing and obtaining happiness and safety.’ 13 Preamble of the US Constitution: ‘We the People . . . establish Justice, insure domestic Tranquillity, provide for common defence, promote the general Welfare . . .’ 14 M. Weber, Wirtschaft und Gesellschaft, 5th edn, Tübingen: J.C.B. Mohr, 1972, pp. 821 et seq. 15 Entscheidungen des Bundesverfassungsgerichts (Decisions of the Federal Constitutional Court), 69, pp. 315–360 – Brokdorf: Auf deren Vermeidung muß eine Rechtsordnung, die nach Überwindung des mittelalterlichen Faustrechts die Ausübung von Gewalt nicht zuletzt im Interesse schwächerer Minderheiten beim Staat monopolisiert hat, strikt bestehen. Das ist Vorbedingung für die Gewährleistung der Versammlungsfreiheit als Mittel zur aktiven Teilnahme am politischen Prozeß und – wie die Erfahrungen mit den Straßenkämpfen während der Weimarer Republik gezeigt haben – für eine freiheitliche Demokratie auch deshalb unverzichtbar, weil die Abwehr von Gewalttätigkeiten freiheitsbegrenzende Maßnahmen auslöst. 16 Gerichtshof der Europäischen Gemeinschaft fur Kohle und Stahl (Court of the European Coal and Steel Community), 1997, I, p. 6959 – Commission v France. 17 Gesetz zur Änderung des Grundgesetzes vom 26.3.1998, Bundesgesetzblatt (Federal Law Gazette), I, 1998, 610, Art 13, paras 3–6, Basic Law; Gesetz zur Änderung des Grundgesetzes vom 28.6.1993, Bundesgesetzblatt, I, 1993, 1002, Arts 16a, 18, s 1 Basic Law. 18 Chr. Calliess, n. 11 above. 19 Entscheidungen des Bundesverfassungsgerichts, 7, p. 377 – Apotheken-Urteil. 20 P. Lerche, Übermaß und Verfassungsrecht, Köln: Heymann, 1961. 21 Entscheidungen des Bundesverfassungsgerichts, 39, p. 1.

Risk decisions in German constitutional and administrative law 22 23 24 25 26

27 28 29 30

35

Entscheidungen des Bundesverfassungsgerichts, 88, p. 203. Entscheidungen des Bundesverfassungsgerichts 49, pp. 89–137, 141 – Kalkar. Entscheidungen des Bundesverfassungsgerichts, 46, p. 160 – Schleyer. R. Schmidt, ‘Der Staat der Umweltvorsorge’, in P.M. Huber (ed.), Das ökologische Produkt, 1995, pp. 97–98; R. Steinberg, Der ökologische Verfassungsstaat, Frankfurt am Main: Suhrkamp, 1998. See Entscheidungen des Bundesverfassungsgerichts, 105, p. 252 – Glykol; ibid., 105, p. 279 – Osho; H. Bethge, ‘Zur verfassungsrechtlichen Legitimation informalen Staatshandelns der Bundesregierung’, Juristische Ausbildung, 2003, 327; D. Murswiek, ‘Das Bundesverfassungsgericht und die Dogmatik mittelbarer Grundrechtseingriffe’, Neue Zeitschrift für Verwaltungsrecht, 2003, 1–7; P.M. Huber, ‘Die Informationstätigkeit der öffentlichen Hand – ein grundrechtliches Sonderregime aus Karlsruhe?’, Juristenzeitung, 58, 2003, 290. Entscheidungen des Bundesverwaltungsgerichts, 71, p. 183; 75, p. 109; 82, p. 76; 87, p. 37; 90, p. 112; P.M. Huber, Allgemeines Verwaltungsrecht, 2nd edn, 1997, Heidelberg: C.F. Müller, pp. 240 et seq., with further references. Entscheidungen des Bundesverfassungsgerichts, 105, p. 252 – Glykol; 105, p. 279 – Sekten. U. Di Fabio, ‘Grundrechte im präzeptoralen Staat am Beispiel hoheitlicher Informationstätigkeit’, Juristenzeitung, 1993, 689. Entscheidungen des Bundesverfassungsgerichts, 24, pp. 367–407; 49, pp. 252–257; 83, pp. 130–148; 105, 109 et seq.; P.M. Huber, in H. Mangoldt, F. Klein and C. Starck (eds.), Kommentar zum Grundgesetz, 5th edn, München: Vahlen, 2005, vol. I, Art 19, para 4, 362.

3

The problem of de-individualisation in the risk society Oliver Lepsius

INTRODUCTION In the area of public law, one uses the term ‘risk’ mainly in a scientific sense, with reference to natural circumstances, for instance when referring to the potential and unforeseeable dangers of technical devices like nuclear energy, environmental protection or novel food.1 Uncertainty and incomplete information challenge the capacity of the law to ensure people’s safety and protect them from potential dangers.2 Hence risk creates legal problems with regard to the predictability of potential dangers. These dangers derive mainly from technical issues. In public law the term risk is rarely used in referring to human behaviour, although there may be nothing riskier than the human being: human behaviour is generally unpredictable and hence risky.3 But, of course, the term risk in this sense is just complementary to freedom; human behaviour is unpredictable because the human is a free individual and his or her freedom is legally protected by civil liberties and human rights. It is, therefore, strictly speaking incorrect to call human behaviour risky in the legal sense. It is rather the use of constitutionally protected individual freedom.4 The interesting and, with respect to individual liberties, alarming fact is that the term risk is, in German law, currently applied to human behaviour. In the field of police and security law we witness a development towards a perspective according to which the individual is conceived as a risk. The last decade has seen a shifting perception in what constitutes a risk. It is not only the unforeseeable effects of modern technologies that lead to a new perception of risks, it is also a new need to provide society with security against diffuse criminal organisations and networks which avail themselves of the globalised world’s benefits. This new development was triggered by two developments: a general tendency to increase the level of security in the aftermath of the fall of the iron curtain and the subsequent abolition of border controls between certain European countries (the ‘Schengen countries’); and the specific security requirements after the terrorist attacks of September 11.5 New security laws are made in reaction to new security needs in a globalised world in which border controls no longer furnish

The problem of de-individualisation in the risk society

37

protection against cross-border crime and terrorist acts are no longer the reaction to local conflicts but are connected to worldwide organisations or networks. The state feels no longer able to control certain risks: criminals easily move from country to country nowadays. Measures aimed at the prevention of terrorist attacks can no longer be limited to national authorities or be made effective by the apparatus of a single state. We also witness an important shift in the perception of risks. It is no longer only the physical (natural or man-made) world which is perceived to create risks. Human deeds tend to be treated more and more as risks. Crossborder crime and terrorism furnish the most important recent examples. In German law there has emerged a new tendency to assess human behaviour using the same notions of unpredictability and potential danger as are used in environmental law to assess natural events. On the one hand this is understandable, since human behaviour always was and, it is to be hoped, ever will be unpredictable. Free individuals do not follow predictable patterns. On the other hand it is not understandable because, since there has always been a problem of unpredictable individual behaviour, there are no essentially new challenges to security and police laws. Both terrorism and cross-border crime are merely new emanations of the well-known general problem of how to regulate the endless forms of human behaviour. Is the term risk, then, appropriate at all when the law has to deal with human behaviour? The use of the term risk and the whole idea of risk regulation by law creates a major danger to a free society if it results in the transfer of patterns of risk regulation from natural, technical problems to human behaviour. Both are unpredictable, but they should not be treated alike. German law reacts to these developments with a new notion which I want to call the de-individualisation of legal obligations.6 Recent German statutes against terrorism and crossborder crime furnish prominent examples which this chapter seeks to analyse and criticise.

NEW FORMS OF THE REGULATION OF SOCIAL RISKS IN GERMAN ADMINISTRATIVE LAW The de-individualisation of legal obligations represents the new tendency of German law to seek to regulate unpredictable human behaviour which is conceived as creating a risk to society. It is not a specific individual, a concrete person, which is chosen by the law as the subject of legal obligations. Rather the subject of legal obligations is the individual seen as an exchangeable part of the society as a whole. This development displays a shift in the view of who is a risk in the sense of being potentially dangerous. It is not the individual person any more who poses a risk but the individual as part of society in general. Therefore the presumption of risk in police and security laws becomes less dependent on individual behaviour and concentrates more on very general assessments of standardised situations which are

38 Risk and the Law considered dangerous. Two recent examples of statutory provisions in German police law are worth mentioning. Both use this new understanding of a de-individualised personality.

Police control competencies without suspicion (gefahrverdachtlose Polizeikontrollen, ‘Schleierfahndung’) In the 1990s a new security debate arose over the gradual dismantling of border controls under the Schengen agreement of the EU. This discussion was acknowledged by the enactment of new statutes. The traditional border controls between the states were abandoned and now persons can freely move between the different Member States in the so-called Schengen countries. However, border controls were not abolished completely; in reality they were functionally moved into the interior of the country. According to the new provisions, the amended German Federal Border Guard Act (Bundesgrenzschutzgesetz) and similar police statutes of the Länder, the police may now stop any person who is travelling on the autobahn, frequents an airport or a railway station or, more generally, finds himself within 30km of the federal border and check his or her ID.7 In any of these locations the police may check any person without any further suspicion. Before this statutory reaction to the Schengen agreement went into effect the police could check someone only when there was a reasonable suspicion that they constituted a danger. The person had to behave in a potentially dangerous manner or show some other sign of being a potential threat in order to be legally liable to police power. Now the situation is different. The individual’s duty to show identification and the power of the police to check a person at airports, railway stations, on the autobahn or within 30km of the border is not triggered by any personal conduct. The police may check a person regardless of what he or she has been doing or how he or she is behaving. Within these defined areas the presumption of risk has been legally reframed: it is not individual behaviour which makes one the subject of a legal duty. The subjection of a person to police powers is not the outcome of that individual’s behaviour. One’s conduct in this situation is irrelevant. It is rather the mere presence in certain locations that triggers the obligation. The only thing the person can do to avoid such police control is to avoid staying in these areas. This may be a possible course of conduct with regard to airports or railway stations but it is hardly an option with regard to presence in the 30km area. Take the case of an inhabitant of this area: that person has no choice whether to be present or not, he simply has to be there because he is living close to the border or his workplace is within the border area. In this case the legal liability has nothing to do with individual behaviour any more – and this is a new situation, since before the Schengen amendments to the German police laws it was impossible for the police to undertake personal checks without having grounds for suspicion.

The problem of de-individualisation in the risk society

39

What, then, is the rationale of these new laws? They are the reaction of the lawmakers on the federal and Länder level to the opening of the borders, the aim being to compensate for the loss of border controls with enlarged competencies for control within the borders. The police power is directed towards individuals and it can be satisfied by the individual through the presentation of valid identification papers. With these new provisions the lawmakers reacted against a new risk: the possibility of an increase in cross-border crime as a result of the abolition of border controls. Under the previous law, it was either the border that constituted a general risk, no matter how individuals behaved (the border itself was a dangerous place, so to speak), or it was a certain person who behaved suspiciously and was consequently liable, no matter where he or she was located. Under the new law the two are confused: the general apprehension of danger at borders is transferred to the interior and it is extended to persons instead of locations. Anybody who stays at certain places becomes the subject of an obligation regardless of their behaviour. These new police competencies constitute a problem for individual freedom.8 The stopping and checking close to the border of persons who have given no cause for suspicion addresses individuals whose conduct has not given any reason for the measures. The individual has neither caused a danger nor given a cause for suspicion. Furthermore, the citizen cannot contribute to the elimination of danger by staying away,9 since presence at, for example, a train station located within 30km of the border is as such without danger. Whether an individual is present or not has little bearing on measures taken to try to eliminate the threat. The purpose of the new police powers exercisable without grounds for suspicion is to be found in the fight against cross-border crimes, which is a rather general phenomenon without precise individual responsibility. The fight against diffuse criminal surroundings, however, is legally directed towards the individual who himself cannot contribute to that purpose. The individual is not responsible as a single person but rather is now held responsible as a member of a society, namely, the society which exists in general at certain places or close to the federal borders. The person in question is not addressed as an individual human being to whom individual responsibility can be assigned. Rather the person is perceived as an exchangeable member of a society whose presence at certain locations is deemed potentially dangerous. Yet the diffusion of cross-border crime in police terms is directed towards society at large, not the individual, while liability continues to be assigned to the individual. This liability is de-individualised because it affects all people at all train stations, airports and highways and as such can no longer be met by specific individual contributions to the removal of that threat. The individual is neither actor nor disturber of the status quo; his responsibility derives from being part of the general public. Hitherto police law had not included such powers of intrusion.10 The problem here is this: the law seeks to impute general phenomena to

40

Risk and the Law

individual responsibilities. With regard to such a general risk as cross-border crime, however, this is hardly possible. What does this entail with regard to the regulation of risks? First, there is a new perception of what is considered to be a risk. It is no longer a specific location or a specific person. Instead, abstract situations are deemed risky: a risk is perceived in a social condition in general, namely the general level of cross-border crime. Border areas have become generally dangerous and constitute risks. The old law regulated that risk rather simply: anyone who crossed the border was the subject of police controls. With the abolition of border controls, it became difficult to maintain the same level of precautions against cross-border crime. The potentially dangerous situation, and hence the risk, is now extended from single persons at the border to the society in the border area in general. Second, the individual is still the object of control, but no longer because of the individual’s intention to cross the border. The control competency now results from the fact that the individual is a part of the society. The interesting development we encounter here is a new imputation of risk: the individual has become liable for a social risk which has nothing to do with individual conduct.11 These new police powers have been scrutinised three times by German state constitutional courts, but not yet by the Federal Constitutional Court (Bundesverfassungsgericht). The Constitutional Court of MecklenburgVorpommern declared parts of the new statute in the state of MecklenburgVorpommern to be unconstitutional.12 In Saxony13 and in Bavaria, however, similar statutes survived judicial review. In a recent decision the Bavarian Constitutional Court declared the relevant statute in the Bavarian police code to be in accordance with Bavarian constitutional law.14 The question, whether such provisions in the police laws are constitutional or not, was thus decided differently in different German Länder. The Bundesverfassungsgericht has not yet decided the issue, although it would be helpful if it were to do so, given the different legal situations now in various parts of Germany.15 The constitutional court of Mecklenburg-Vorpommern declared police control competencies in the absence of suspicious behaviour to lack the constitutionally required nexus of individual behaviour and a legally protected good. According to the court this entailed the unconstitutionality of the statute. The individual in such circumstances has not sufficiently contributed to a risk, has not increased a danger and, hence, cannot in general be legally liable for the situation of risk. The court demanded from the lawmakers a more precise definition of the legal duty. Otherwise the statute would constitute an encroachment on civil liberties. The Bavarian court, however, did not recognise such a constitutional requirement. Basically the court was not troubled by provisions which just rearranged traditional legal duties in a new manner. The confusion of individual obligations with social risks, mentioned above, did not disturb the Bavarian court. In contrast to the court in Mecklenburg-Vorpommern, the

The problem of de-individualisation in the risk society

41

Bavarian judges limited their review to the mere question of the proportionality between the legal liability, which was deemed minor, and the potential threat, which was considered large. Neither the court in Bavaria nor that in Mecklenburg-Vorpommern, however, addressed the question whether the new pattern of de-individualised legal obligations was constitutional in general. This question remains open and requires further scrutiny.

Anti-terrorist laws (with specific reference to grid searches, ‘Rasterfahndung’) In the first example, cross-border crime as a general social phenomenon constituted the risk against which legal protection was needed. Let us now focus on a second example which manifests the same development: new statutes enacted following the terrorist attacks of September 11 2001.16 The German Bundestag with the consent of the Bundesrat enacted a bunch of new measures called ‘security packages’ (Sicherheitspakete).17 The first of these security laws concerned penal law and the law of private associations and is without further interest in our context here. The second of these ‘security packages’, however, contained a variety of new provisions in administrative law. Nearly 100 regulations in 17 different statutes were affected by the changes. The term ‘package’ is thus fitting. This second ‘security package’ or ‘anti-terror package’ reflects the conclusions drawn from a perceived global threat of Islamic terrorism. Its primary goal is the early detection by the security authorities of such threatening activities as a result of a stress on preventive protection.18 The new regulations enlarged the functions and powers of the security authorities, namely, the federal agency for the protection of the constitution (Bundesamt für Verfassungsschutz, BfV), the military counterespionage service (Militärischer Abschirmdienst, MAD), the federal intelligence service (Bundesnachrichtendienst, BND) and the federal criminal police office (Bundeskriminalamt, BKA). In addition they facilitated the exchange of data by various authorities. Further important new rules concern the laws on aliens and asylum. Entry into Germany by perpetrators of terrorist activities is to be prevented, measures to check incoming persons’ identities through the visa issue procedure and border controls are to be improved, and the use of armed Air Marshals from the Federal Border Guard on German flights is authorised. In addition, the law introduces precautionary measures which allow security checks on personnel in defence or other necessary installations, which allow for the inclusion of biometric data in personal identification documents and for the new form of grid searches by providing for the inclusion of specified social information. In a grid search (Rasterfahndung) an unlimited group of people is electronically scrutinised by computerised databank searches according to a grid of patterns which the police deem appropriate, such as gender, age, denomination, local origin, student status, receipt of social benefits, criminal

42

Risk and the Law

record, etc. The purpose of the search is to single out persons from a group of potentially dangerous people. These persons are then checked more intensely – although they may never have behaved suspiciously as individuals. The first question, again, refers to the perceived risk. At what kind of risk is this new terrorist legislation aimed? This is a reaction by the lawmakers triggered by the events of September 11, although not necessarily motivated by them. The attacks were seen not as the deeds of individual terrorists but as a de-individualised phenomenon, a new form of terrorist threat. This threat seemed to arise not from individuals but from a general development in a globalised world in which individuals are merely exchangeable tools in the hands of powers working in the background and assured of protection by certain states. The legislative initiatives in Germany were not meant to address the specific deeds of Al Qaeda. Rather they are aimed at what is perceived as an ever-present threat by Islamic terror organisations. To explain the legislative reaction to terrorism, it is important to understand the perception of the events of September 11 2001 and how it entailed the redefinition of a specific danger into an uncertain risk. The source of danger is no longer the individual perpetrator but rather impersonal networks and organisations lurking in the diffusion of Islamic fundamentalism. The new threat was grounded, like that in my previous example, in two developments: the elimination of a local context and the diffusion of an individual context of terrorist actions. Terrorism by networks seemed depersonalised and de-regionalised. The threat became global and could no longer be limited to a few perpetrators. The level of threat, therefore, was not attributed to individual terrorists but was seen as the consequence of a qualitatively new danger, emanating not from the individual attackers themselves but rather from the network of terror in the background. Terrorism was not perceived as the summation of individual deeds but rather as the result of collective, evil structures. Only by this change in perception could terrorism acquire a qualitatively new dimension, since the rights associated with democracy and freedom had been targets prior to the September attacks. The largely new character of these attacks, which also determined the legislative balancing of security and freedom, could be found only by not recognising the individual as responsible for the crimes committed. A qualitatively new threat seemed to demand qualitatively new powers. The new powers mentioned above reflect a legislative reaction to these newly perceived risks. They are directed not at individuals perceived as presenting risks but at the social setting as such that is deemed dangerous. How does the law impute this risk to the individual? Let us concentrate on just one example of the ‘security package’, the use of social data in grid searches (Rasterfahndung).19 The use of individual social data is intended to facilitate the prevention of the damage where the probability of its occurrence is highly doubtful. Can one generally assume that male persons from Arab countries under the age of 30 constitute a risk? The German authorities

The problem of de-individualisation in the risk society

43

thought so. There was an ordinance providing for the screening at German universities of all male students under the age of 30 who originated from specific countries (mainly in the Middle East) and who were of Islamic denomination. All these students were checked because of the conjecture that there might exist more hidden groups of terrorists like the group in Hamburg. A grid search covers every person with certain standardised characteristics which are thought to have a closer or less close relation to an assumed risk. The problem here is that Islamic students as such do not constitute a risk. Every individual, Arab or not, may possibly turn out to be a dangerous person under specific circumstances. Of course, theoretically, anybody can become a criminal. May the law treat anybody as a potential criminal just because they belong to a certain group of people which is suddenly supposed to constitute a risk? Whether one belongs to this group and hence is considered to be a risk does not, again, depend upon individual behaviour. Grid searches treat individuals like technical problems. The use of general social data in grid searches disconnects individual behaviour and danger. We witness a development similar to that of police controls without suspicion; it is not the specific individual who is seen to be dangerous. If it were otherwise, one could easily take action within the regular public or criminal law, traditionally directed toward unlawful human behaviour. Under the new pattern of security and perception of social risk, legal action is no longer a question of an individualisable danger but of a perceived general risk which cannot be imputed to concrete persons. Consequently, it is the collective and not the individual which is under surveillance now, even though the legal obligation traditionally falls upon the individual. So the collective group of all-male Arab students at German universities under the age of 30 is held liable for the purpose of the prevention of a risk, here terrorism. Risk has become a social or at least a collective phenomenon and, therefore, the powers have to address not the individual but the collective. However, the individual has no option as to whether he belongs to such a collective. His inclusion in the suspect group is completely independent of his personal conduct. He is not legally liable because of an individual contribution in increasing or decreasing a specific danger. Merely because he is loosely connected to a social group he is considered to be potentially dangerous. Parts of the powers to grid search have already been examined by various courts.20 Most concluded that searches grounded on a general suspicion based upon membership of specific religious communities or nationalities were unconstitutional. The district court Wiesbaden and the district court Berlin both declared the grid search for Islamic students at German universities to be unconstitutional.21 The court of appeals in Frankfurt22 and Düsseldorf23 also declared parts of the grid search unconstitutional. In North Rhine Westphalia all residents’ registration offices, universities, technical schools and the central registration office for foreigners were instructed to pass on data on all male individuals born between the years of 1960

44

Risk and the Law

and 1983 to police headquarters in Düsseldorf. The court of appeals in Düsseldorf held that, in order to aid the search for ‘sleepers’ of Islamic terror organisations, only personal data of citizens of suspected countries or of Muslim persons were allowed to be passed on, not data of German citizens. The current level of threat from German citizens is seen as too small to allow for the comprehensive grid search. The court in Frankfurt went even further, while other courts issued less restrictive judgments and did not doubt the constitutionality of the statutes and the concrete measures.24 Recent case law is far from coherent as far as the outcome is concerned. Yet in sum, the courts have tried to transform the imputation of risk to a social group into a rather conventional, individualisable suspicion of danger: a grid search with specific search terms is intended to individualise the threat. The criteria for the grid search have to apply only to a limited number of persons. The general tendency of the court decisions can be summed up as holding that the smaller a group is or the more precise the social data are, the less problematic is a grid search. The easier it is to specify the patterns for a grid search, the less problematic that search becomes. If the search, for instance, can be limited because it follows precise criteria to a group of, say, 100 rather than 10,000, the individualisation of the risk becomes once more feasible. Should one, however, still speak of a risk in a case where one is able to reduce the screening to such a limited number of persons as a mere 100? In this case it seems more plausible to speak of a potential danger instead of a risk. Put differently, when there is a mere risk, there is a problem that the measures may be unconstitutional. When there is a danger, the measures are permissible.

Assessment The concept of risk is a very dangerous concept with regard to personal behaviour because it treats individuals like technical devices according to non-influenceable patterns and cuts off any personal contribution. With regard to human behaviour the law should rather limit itself to the notion of danger. To establish danger it is necessary for there to be a traceable relationship between individual conduct and legal liability. The concept of risk in contrast denies this connection. It is the notion of risk itself which excludes such a requirement because otherwise the situation would not create a risk. The necessary unpredictability in risk requires that the possible effects are unforeseeable and hence it cannot provide for an individual contribution either to increase or to reduce the predictability of the risk. The idea of risk regulation, therefore, should not be extended to any form of human behaviour. As I said at the beginning, there is nothing riskier than the human being. Human behaviour is unpredictable because the human is a free individual. To treat individual behaviour according to the patterns of unpredictable technical or environmental risks would deny the human his foremost capacities: freedom and individuality.

The problem of de-individualisation in the risk society

45

CONSTITUTIONAL PROTECTION AGAINST DE-INDIVIDUALISED PATTERNS OF RISK REGULATION

Encroachment on civil rights The reasoning of the decisions concerned with grid searches displays a further interesting legal problem. Because generally the system of judicial review in Germany is individually based,25 the courts cannot address the issue of de-individualisation directly but have to reframe it as a problem of an intrusion on civil rights. Whether there is indeed an infringement of civil rights in these cases is not easy to determine. If a person is liable not because of his or her individual behaviour but because he or she is part of a social group or of society as a whole, one could question whether there is any infringement of rights at all. The problem here is that constitutional law may lose its criteria for controlling such legislative measures. Judicial review remains possible if there exists an individualisable context. Judicial review fails, however, if the person, although personally liable, remains an exchangeable part of the general society. Therefore, on the one hand the exercise of police control competencies in the absence of grounds for suspicion have now been held to a large extent constitutionally permissible, because the group from which the risk emanates is simply too large for these to be evidence that the control constitutes an infringement of the civil rights of any particular individual member. On the other hand, judicial review is more critical with regard to grid searches because it is not the whole society which is liable but a smaller group which can be identified by standardised patterns. One could argue that this solution is completely right from the standpoint of a concern with encroachment on civil rights. If there is no individual encroachment when a person is liable as part of the society and not because of their personal conduct, why should this be seen as a threat to civil rights? Cases, then, would come out differently depending upon the concrete level of personal encroachment. However, this ignores the critical point of de-individualisation. The law treats persons regardless of their personal behaviour as dangerous, as a general risk, and there are no constitutional safeguards because judicial review will not find individual infringements of civil liberties. This is the legal corollary of introducing the notion of risk into the realm of police and security law. Constitutionally the most suspect development is the new form of de-individualisation of liability, according to which the individual person is no longer treated as an individual but rather as an exchangeable element in an essentially dangerous environment. The safety valves of enforcement processes for basic rights cannot offset this encroachment and the individual is no longer perceived as essentially a law-abiding citizen but as a potential threat. Behind this new understanding lies a change in the perception of human nature. This change cannot be addressed constitutionally with

46 Risk and the Law concrete measures, although it is the decisive factor in the reorganisation and justification of the new security powers. New forms of risk regulation, therefore, give rise to a need for new forms of constitutional protection. The danger no longer emanates from individual culprits. Rather there is a diffuse level of threat that has to be preventively managed. This de-individualisation of danger has negative consequences for individual freedom. It is no longer individuals’ civil liberties that have to be balanced against encroachment on basic rights in individual cases. Now the focus is on collective security interests, which are weighed against collective rights.26 The weighing process has shifted away from the weighing of individual, subjective legal positions to a weighing of objective legal perspectives.27 Individual freedom in this constellation is no longer individually protected; its protection now merely rests on the reflex of a liberal society.28 The individual is part of society and shares its liberal status. One could argue that, since the freedom of society is threatened, the individual has to accept the possibility of limitations on his or her individual liberties, if they serve the purpose of securing societal freedom. The protection of freedom then ceases to be essentially the protection of the individual but becomes the protection of society, in which the individual can partake. His or her specific individual interests, which might not be generally recognised in society, are no longer sufficiently protected. Individual freedom becomes a freedom subject to the preservation of society. This collectivisation threatens especially the rights of minorities. The single person is no longer seen as an autonomous individual who engages in selfdetermined and responsible actions which are legally protected. Rather the individual is seen as an anonymous part of society and as such he or she reflexively partakes in collective security as the basis for general freedom. The de-individualisation of danger leads to a de-individualisation of responsibilities and duties, but also of rights, which are transferred as to their protection and defence function from the individual to the collective. Dangers can no longer be individualised; they arise from transnational organisations and networks. As a result it becomes nearly impossible to define legally the prerequisites and goals of counterterrorist measures, albeit such a definition would be legally useful. The definition of these dangers is important and the substitution of diffuse terms such as ‘threat’ or ‘risk’ should be avoided.

Affirmative protection duties (Schutzpflichten) This de-individualisation or collectivisation of rights and the accompanying loss of protected subjective-legal rights can also be seen in the constitutional justification for the statutory limitation of basic rights. In German constitutional law basic rights are trumped by legislation reservation clauses (grundrechtliche Gesetzesvorbehalte) that justify encroachments upon these rights. This justification is based on a good that enjoys higher legal protection. The encroachment must not be disproportionate. Infringements of civil

The problem of de-individualisation in the risk society

47

rights must be in accordance with the legislation reservation clause and have to be proportional. The justification for the encroachment on basic rights in the context of security concerns since the 1970s has relied primarily on a particular constitutional theoretical construction, the so-called protection duties (Schutzpflichten). In 1975 the German Federal Constitutional Court in its first abortion decision combined Art 2 with Art 1, section 1, sentence 2 of the Basic Law to declare a comprehensive duty of the state to protect each human life, especially from illegal encroachment by others.29 On the basis of this duty to protect human life, a basic right to protection from the state equal to the right of freedom was created by judge-made law. This legal duty could be used in the judicial balancing process to justify security-based encroachments on individual liberty. Statutory limitations of basic rights could now be justified not only by reference to conflicting interests in freedom but also by a constitutional duty of protection, which was distinct from individual civil liberties and weighed in on the side of the state.30 The state was given a duty, based on basic rights, to protect the legal commodity of ‘life’. This transformed the basic rights from individual rights into collective duties and the basic rights were modified by so-called objective-legal functions.31 Reliance on the affirmative duty of protection is needed only in cases where no conflicting individual interests exist. Given the comprehensive protection of individual liberties by the constitution, such a case is hardly imaginable, since a legislative measure always affects some individual’s freedom in some form, even if only the general freedom to act as expressed in Art 2, section 1 of the Basic Law. As such, the development of the affirmative duties was not really necessary. Only in one particular case could one not rely, in a constitutional theoretical sense, on the claims of conflicting individual liberties. In the abortion debate the foetus needs constitutional protection. A general freedom to act cannot be assigned to it and therefore it is limited to the protection granted in Art 2, section 2 of the Basic Law. This protection, which originally was individually derived, has to be guaranteed by the state. The German federal constitutional court referred to a commitment by the state to protect the developing life.32 But the debate about the duty of protection soon moved away from this individualised anchoring of the duty, which in the exceptional case of abortion had been used to maintain the idea of conflicting rights to freedom. Only a few years later the same duties of protection were used explicitly to anchor collective goods. With the help of the duties of protection it was now possible to place non-individualised legally protected goods in the balancing scales with individual civil liberties. The German Federal Constitutional Court used this justification nearly immediately to uphold an early anti-terrorist law in 1978 which limited the contact of a lawyer with his (terrorist) client.33 The general abstract protection against terrorist attacks could be used as justifications for individual, concrete encroachments upon basic rights, in this case rights of defendants

48

Risk and the Law

and their legal counsel in the criminal process. The affirmative duties have thus been diverted from their subjective-legal origin to an objective-legal principle.34 The development of the duties of protection is of central importance for the enlarged justification of encroachments upon basic rights. In a legal balancing with collective duties related to the protection of life, individual liberties seem to succumb. Affirmative duties lead to a fundamental unevenness in balancing processes.35 The collective protected interest is not weighed against a similarly collective protection against encroachments on basic rights. Since each individual liberty presupposes the functioning of the legal system, it cannot be thought of in a purely negative sense, and since it also necessarily presupposes that the laws have the effect of organising and supporting freedom, affirmative duties generally threaten to reduce civil liberties. One side of the weighing process now contains positive and negative components of freedom in general rather than subjective rights. Duties of protection shift the weighing from positions of individual basic rights to systematic constitutional decisions. The weighing no longer takes place between individually ascribable rights but rather between public interests. The duties of protection thus not only foster a certain de-individualisation but also support a certain de-legalisation of the doctrines of basic rights.

SUMMARY It is clear that the process of de-individualisation is twofold. It has taken place in the area of legal obligations and in the area of subjective rights. It has become part of administrative law as well as of constitutional law. And it is triggered by the idea that risk regulation is a function of the legal system, although it has turned out that neither legal duties nor constitutional rights can be adjusted to the concept of risk without losing their respective characters. The risk of terrorism and the risk of cross-border crime are not legally transferable to individual persons. The imputation of such risks is hardly possible within the doctrines of administrative law. With regard to individual behaviour one should not use the concept of risk. The outcome is a de-individualisation of legal obligations with major consequences for the effectiveness of administrative law and for the safeguards of constitutional law. The process of de-individualisation thwarts the constitutional safeguards for the protection of civil liberties. If police and security laws found legal obligations on a conception of the individual without any regard to individual behaviour or character but on a de-individualised concept of the individual as an exchangeable part of society, then it is quite natural that constitutional provisions protecting individual liberties will fail to perform their function of protecting the individual against acts done in furtherance of these obligations. The constitutional provisions concerning civil rights

The problem of de-individualisation in the risk society

49

protect only the individual use of specific freedoms but not the general freedom of society as a whole. They protect subjective rights, not social conditions. Therefore there arises a general threat to constitutional law. Individual freedom tends to be underprotected against new statutory provisions which base legal obligations upon a de-individualised concept of the individual. On the level of constitutional law, the individual is protected in his or her personal behaviour, while on the level of regular statutory law he or she is responsible as part of society as a whole without regard to any individual conduct. Because of all these problems I would argue, therefore, that the category of ‘risk’ should not be used with regard to human behaviour or social conditions which directly affect individual rights. Uncertainty and unpredictability as central concepts of the law in the risk society may not be used with reference to human behaviour. The concept of risk should be limited to purely technical issues.

Notes 1 See C.R. Sunstein, Risk and Reason. Safety, Law and the Environment, Cambridge: Cambridge University Press, 2002; H. Hill, Risiko-Management in der englischen Verwaltung, Speyer: Deutsche Hochschule für Verwaltungswissenschaften, 2003. Major monographs on the subject in German are: U. Di Fabio, Risikoentscheidungen im Rechtsstaat, Tübingen: Mohr, 1994; M. Möstl, Die staatliche Garantie für die öffentliche Sicherheit und Ordnung, Tübingen: Mohr, 2002; P.-T. Stoll, Sicherheit als Aufgabe von Staat und Gesellschaft, Tübingen: Mohr, 2003; in comparative European perspective: E. Riedel (ed.), Risikomanagement im öffentlichen Recht, Baden-Baden: Nomos, 1997. 2 On the notion of uncertainty cf. in interdisciplinary perspective: C. Engel, J. Halfmann and M. Schulte (eds.), Wissen – Nichtwissen – Unsicheres Wissen, Baden-Baden: Nomos, 2002. 3 For a critical treatment of related questions from a social perspective see M.M. Wambach (ed.), Der Mensch als Risiko. Zur Logik von Prävention und Früherkennung, Frankfurt am Main: Suhrkamp, 1983. 4 On the problematic relationship between risk and freedom see G. Vobruba, ‘Prävention durch Selbstkontrolle’, in Wambach, ibid., pp. 29, 33, 40; D. Grimm, ‘Verfassungsrechtliche Anmerkungen zum Thema Prävention’, in D. Grimm (ed.), Die Zukunft der Verfassung, Frankfurt: Suhrkamp, 1991, pp. 197, 198, 200, 218; R. Münch, Risikopolitik, Frankfurt: Suhrkamp, 1996, pp. 34, 238; P.-A. Albrecht, Die vergessene Freiheit. Strafrechtsprinzipien in der europäischen Sicherheitsdebatte, Berlin: BWV, 2003; W. Brugger and C. Gusy, ‘Gewährleistung von Freiheit und Sicherheit im Lichte unterschiedlicher Staatsund Verfassungsverständnisse’, Veröffentlichungen der Vereinigung der Deutschen Staatsrechtslehrer 63, 2004, 101, 121, 129; 151, 155, 162; O. Lepsius, ‘Risikosteuerung durch Verwaltungsrecht’, Veröffentlichungen der Vereinigung der Deutschen Staatsrechtslehrer 63, 2004, 264, 283–290, 308. 5 On the legal effects of and reactions to September 11, see Kay Nehm, ‘Ein Jahr danach. Gedanken zum 11. September 2001’, Neue Juristische Wochenschrift, 2002, 2665; E. von Bubnoff, ‘Terrorismusbekämpfung – eine weltweite Herausforderung’, Neue Juristische Wochenschrift, 2002, 2672; C. Tomuschat,

50

6 7

8

9

10

11 12 13 14

15

16

Risk and the Law ‘Der 11. September 2001 und seine rechtlichen Konsequenzen’, DeutschAmerikanische Juristenvereinigung, Newsletter, 2002, 1. With regard to developments in international law see M. Kotzur, ‘ “Krieg gegen den Terrorismus” – politische Rhetorik oder neue Konturen des “Kriegsbegriffs” im Völkerrecht?’, Archiv des Völkerrechts, 40, 2002, 454–479; H. Schulze-Fielitz, ‘Nach dem 11. September: An den Leistungsgrenzen eines verfassungsstaatlichen Polizeirechts?’ in H.-D. Horn (ed.), Recht im Pluralismus. Festschrift für Walter Schmitt Glaeser, Berlin: Duncker & Humblot, 2003, 407; Adolf-Arndt-Kreis (ed.), Sicherheit durch Recht in Zeiten der Globalisierung, Berlin: BWV, 2003; M. Kötter, ‘Subjektive Sicherheit, Autonomie und Kontrolle. Eine Analyse der jüngeren Diskurse des Sicherheitsrechts’, Der Staat, 43, 2004, 371. See also O. Lepsius, Besitz und Sachherrschaft im öffentlichen Recht, Tübingen: Mohr, 2002, pp. 449–468, 482–486. See e.g. Bundesgrenzschutzgesetz (Federal Border Guard Act), § 22 Ia. The police laws of the Länder were amended with similar provisions. The provision in the Bavarian police law, Polizeiaufgabengesetz, is Art 13 I, Nr. 5; in North Rhine-Westphalia Polizeigesetz NW, § 31; in Hesse, Hessisches Sicherheits- und Ordnungsgesetz, § 26 I, IV. In a broader context see W. Hassemer, ‘Polizei im Rechtsstaat’, in G. Bemmann and D. Spinellis (eds.), Strafrecht – Freiheit – Rechtsstaat, Festschrift für G.-A. Mangakis, Athens: Sakkoulas, 1999, 621, 628 et seq., 640 et seq.; J. Limbach, Ist die kollektive Sicherheit der Feind der individuellen Freiheit?, Cologne: Heymann, 2002; Friedrich-Ebert-Stiftung (ed.), Sicherheit vor Freiheit? Terrorismusbekämpfung und die Sorge um den freiheitlichen Rechtsstaat, Berlin: Friedrich-Ebert-Stiftung, 2003. This constitutes the difference from so-called ‘dangerous places’, such as areas heavily frequented by drug users and dealers, at which even prior to the introduction of the new regulations police inspections were possible without individually founded suspicions. Mere presence at such locations constituted suspicious conduct. See the critical analysis of H. Lisken, ‘Verdachts- und ereignisunabhängige Personenkontrollen zur Bekämpfung der grenzüberschreitenden Kriminalität?’, Neue Zeitschrift für Verwaltungsrecht, 1998, 22; C. Möllers, ‘Polizeikontrollen ohne Gefahrenverdacht’, Neue Zeitschrift für Verwaltungsrecht, 2000, 382; E. Denninger, ‘Schleierfahndung im Rechtsstaat?’, Festschrift für Ekkehart Stein, Tübingen: Mohr 2002, p. 15. See Lorenz Schulz, Normiertes Mißtrauen. Der Verdacht im Strafverfahren, Frankfurt am Main: Klostermann, 2002, pp. 212–217. See judgment of Landesverfassungsgericht Mecklenburg-Vorpommern, ‘Urteil von 21.10.1999 – LVerfG 2/98’, Deutsches Verwaltungsblatt, 2000, 262– 269. Sächsischer Verfassungsgerichtshof, Entscheidung v. 10.7.2003, Vf. 43-II-00, www.justiz.sachsen.de/gerichte/homepages/verfg/docks/43-II-00A.rlf. See judgment of Bayerischer Verfassungsgerichtshof, Entscheidung v. 28.3.2003 – Vf. 7-VII-00, 8-VIII-00, www.bayern.verfassungsgerichtshof.de/7-VII-00. htm.; Bayerische Verwaltungsblätter 2003, 560, with comments by H.-D. Horn, Bayerische Verwaltungsblätter 2003, 545 et seq. The German Constitution provides for such circumstances the remedy of Divergenzvorlage, Art 100, III Grundgesetz. The Bavarian Constitutional Court did not have recourse to a Divergenzvorlage to the Bundesverfassungsgericht because it claimed to decide the matter purely with regard to Bavarian constitutional law. A comparative analysis of statutory reactions to 11 September 2001 and new anti-terrorist laws is provided in C. Walter et al. (eds.), Terrorism as a Challenge

The problem of de-individualisation in the risk society

17

18

19

20

21 22 23 24 25

26 27

51

for National and International Law: Security versus Liberty?, Berlin: Springer, 2004. Gesetzentwurf der Fraktionen SPD und BÜNDNIS 90/DIE GRÜNEN – Entwurf eines Gesetzes zur Bekämpfung des internationalen Terrorismus (Terrorismusbekämpfungsgesetz), Drucksachen des Deutschen Bundestages, 14/ 7386 (neu) v. 8.11.2001; Bundesgesetzblatt (Federal Law Gazette), I, 2002, 361; Bundesgesetzblatt, I, 2002, 3390. See also the justification for the ‘security package’ given by the federal government, Drucksachen des Deutschen Bundestages, 14, 7025, 7026 v. 4.10.2001. Overviews by M. Nolte, ‘Die Anti-Terror-Pakete im Lichte des Verfassungsrechts’, Deutsches Verwaltungsblatt, 2002, 573; S. Rublack, ‘Terrorismusbekämpfungsgesetz: Neue Befugnisse für die Sicherheitsbehörden’, Datenschutz und Datensicherheit, 2002, 202; E. Denninger, ‘Freiheit durch Sicherheit? Anmerkungen zum Terrorismusbekämpfungsgesetz’, Strafverteidiger 2002, 96; U. Kersten, ‘Internationaler Terrorismus’, Kriminalistik, 2002, 7; J. Saurer, ‘Die Ausweitung sicherheitsrechtlicher Regelungsansprüche im Kontext der Terrorismusbekämpfung’, Neue Zeitschrift für Verwaltungsrecht, 2005, 275–282. On the introduction of grid searches in the new statutes see H. Lisken, ‘Zur polizeilichen Rasterfahndung’, Neue Zeitschrift für Verwaltungsrecht, 2002, 490; W. Bausback, ‘Rasterfahndung als Mittel der vorbeugenden Verbrechensbekämpfung’, Bayerische Verwaltungsblätter, 2002, 713. Recent decisions on the permissibility of grid searches: ‘OVG Bremen, Beschluß v. 8.7.2002 – 1 B 155/02’, Neue Zeitschrift für Verwaltungsrecht 2002, 1530; ‘OVG Koblenz, Beschluß v. 22.3.2002 – 12 B 10331/02’, Neue Zeitschrift für Verwaltungsrecht 2002, 1528; VG Hamburg, ‘Beschluß v. 27.2.2002, 14 VG 446/02’, Datenschutz und Datensicherheit 2002, 370, VG Mainz, ‘Beschluß v. 19.2.2002, 1 L 1106101.MZ’, Datenschutz und Datensicherheit 2002, 303. ‘LG Wiesbaden, Beschluß v. 6.2.2002, 4 T 707/01’, Datenschutz und Datensicherheit 2002, 240; ‘LG Berlin, Beschluß v. 15.1.2002, 84 / 278 u.a.’, Datenschutz und Datensicherheit, 2002, 175. ‘OLG Frankfurt, Beschluß v. 8.1.2002, 20 W 479/01’, Neue Zeitschrift für Verwaltungsrecht, 2002, 627; ‘OLG Frankfurt, Beschluß v. 21.2.2002 – 20 W 55/02’, Neue Zeitschrift für Verwaltungsrecht, 2002, 626. ‘OLG Düsseldorf, Beschluß v. 8.2.2002 – 3 Wx 357/01’, Neue Zeitschrift für Verwaltungsrecht, 2002, 629. OVG Koblenz, n. 20 above; OVG Bremen, n. 20 above. See H. Bauer, Geschichtliche Grundlagen der Lehre vom subjektiven Recht, Berlin: Duncker & Humblot, 1986; D. Lorenz, Der Rechtsschutz des Bürgers und die Rechtsweggarantie, München: C.H. Beck, 1973; E. SchmidtAßmann, ‘Art. 19 Abs. 4 GG als Teil des Rechtsstaatsprinzips’, Neue Zeitschrift für Verwaltungsrecht, 1983, 1; R. Wahl, ‘Die doppelte Abhängigkeit des subjektiven öffentlichen Rechts’, Deutsches Verwaltungsblatt, 1996, 641; R. Wahl, ‘Vorbemerkung § 42 Abs. 2, Nr. 44–60’ in F. Schoch, E. SchmidtAßmann and R. Pietzner (eds.), Verwaltungsgerichtsordnung. Kommentar, Munich: Beck, 2004. The Bavarian Constitutional Court uses this style of reasoning in order to uphold the constitutionality of the Bavarian police law, see decision, n. 14 above, at V. A. 2. See T. Groß, ‘Terrorbekämpfung und Grundrechte. Zur Operationalisierung des Verhältnismäßigkeitsgrundsatzes’, Kritische Justiz, 2002, 1; E. Denninger, ‘Freiheit durch Sicherheit?’, Strafverteidiger, 2002, 96; W. Hoffmann-Riem,

52

Risk and the Law

‘Freiheit und Sicherheit im Angesicht terroristischer Anschläge’, Zeitschrift für Rechtspolitik, 2002, 497. 28 For this development in a broader context see O. Lepsius, Besitz und Sachherrschaft im öffentlichen Recht, Tübingen: Mohr, 2002, pp. 384–387, 396, 404–405. 29 Entscheidungen des Bundesverfassungsgerichts (Decisions of the German Federal Constitutional Court) 39, 1 at 42. 30 There is a general tendency in German constitutional law to regard legislative measures not only as a potential threat to individual liberty but also as helpful instruments in establishing liberty. See W. Cole Durham, ‘General Assessment of the Basic Law – An American View’, in P. Kirchhof and D. Kommers (eds.) Germany and Its Basic Law, Baden-Baden: Nomos 1993, pp. 37, 45: The German approach is fundamentally more sympathetic to a conception in which the state plays a role in facilitating the actualisation of freedom. Rather than being the key power that needs to be constrained if liberty is to be preserved, the state is seen as the vehicle for achieving freedom. 31 See H. Dreier, Dimensionen der Grundrechte. Von der Wertordnungsjudikatur zu den objektiv-rechtlichen Grundrechtsgehalten, Hannover: Hennies & Zinkeisen, 1993; E.-W. Böckenförde, ‘Grundrechte als Grundsatznormen’ in Staat, Verfassung, Demokratie, Frankfurt am Main: Suhrkamp, 1991, p. 159; D.P. Currie, ‘Positive and Negative Constitutional Rights’, University of Chicago Law Review, 53, 1986, 864, 880–882. 32 Entscheidungen des Bundesverfassungsgerichts, 39, 1. 33 Entscheidungen des Bundesverfassungsgerichts 49, 24 at 53. 34 The development of the affirmative duties has led to a lasting scholarly debate. The duties are accounted for partly by subjective legal reasoning and partly by objective legal reasoning. See J. Isensee, ‘Das Grundrecht als Abwehrrecht und als staatliche Schutzpflicht’, in J. Isensee and P. Kirchhof (eds.), Handbuch des Staatsrechts, vol. V, Heidelberg: C.F. Müller, 1992, § 111; G. Hermes, Das Grundrecht auf Schutz von Leben und Gesundheit: Schutzpflicht und Schutzanspruch aus Artikel 2 Absatz 2 GG, Heidelberg: C.F. Müller, 1987; C. Enders, ‘Die Privatierung des Öffentlichen durch die grundrechtlichen Schutzpflichten und seine Rekonstruktion aus der Lehre von den Staatszwecken’, Der Staat, 35, 1996, 351; R. Wahl and I. Appel, ‘Prävention und Vorsorge. Von der Staatsaufgabe zur rechtlichen Ausgestaltung’, in R. Wahl (ed.), Prävention und Vorsorge, Bonn: Economica, 1995, p. 1; P. Unruh, Zur Dogmatik der grundrechtlichen Schutzpflichten, Berlin: Duncker & Humblot, 1996; R. Wahl, ‘Die objektivrechtliche Dimension der Grundrechte im internationalen Vergleich’, in D. Merten and H.-J. Papier (eds.), Handbuch der Grundrechte, vol. I, Heidelberg: C.F. Müller, 2004, § 19, Nr. 52–55. 35 See P. Preu, ‘Freiheitsgefährdung durch die Lehre von den grundrechtlichen Schutzpflichten’, Juristenzeitung, 1991, 265.

4

Risk decisions in cases of persisting scientific uncertainty: the precautionary principle in European food law Rudolf Streinz

INTRODUCTION The ‘precautionary principle’ is quoted (but not defined) in the English version of the EC Treaty. According to Art 174 para 2 sentence 2, which has been introduced by the Single European Act of 1986,1 the Community policy on the environment shall be based on the precautionary principle and on the principle that preventive action should be taken.2 In general, the precautionary principle was initially developed in environmental law, on the national, especially German, basis as well as in European Community law and in international law.3 Since the presentation of the Green Paper on ‘The General Principles of Food Law in the European Union’4 in 1997 and especially since the Bovine Spongiforme Encephalopathy (BSE) crisis, the ‘precautionary principle’ has often been cited in EC food law affairs.5 In its judgments of 5 May 1998 on the validity of the Commission’s decision banning the exportation of beef from the United Kingdom to reduce the risk of BSE transmission,6 the European Court of Justice referred to Art 174 EC Treaty and in substance to the precautionary principle.7 In four cases the European Court of Justice and the European Court of First Instance explicitly quoted and dealt with the ‘precautionary principle’.8 In its judgment of 5 April 2001 on the Norwegian legislation which prohibited the import and marketing of ‘fortified corn flakes’ in Norway, the EFTA Court dealt in depth with the application of the precautionary principle emphasising the principle of proportionality which limits national measures having restrictive effects on the Common Market of goods.9 The European Commission presented general considerations on the precautionary principle in its ‘Green Paper on the General Principles of Food Law in the European Union’10 and in its Communication on Consumer Health and Food Safety of 1997.11 In its White Paper on Food Safety of 12 January 200012 the Commission said that it would bring forward a paper on the Precautionary Principle. This paper, the ‘Communication from the Commission on the precautionary principle’,13 was presented on 2 February 2000. It is of outstanding quality.

54

Risk and the Law

It contains a detailed explanation of the principle in detail and presents some practical guidelines for applying the precautionary principle. Finally, the precautionary principle has been part of existing EC food law since the coming into force of Art 7 of the Regulation (EC) No 178/2002 of the European Parliament and of the Council of 28 January 2002 laying down the general principles and requirements of food law, establishing the European Food Safety Authority and laying down procedures in matters of food safety,14 the (not officially) so-called ‘Basic Regulation’. First I will discuss some general problems of ‘precaution’. Then I will give a short overview of the precautionary principle in German food law. Subsequently, I will present the precautionary principle in EC food law according to Art 7 Basic Regulation and the Communication of the Commission. Finally I will give an outlook on present and foreseeable problems, especially in the field of world trade law.

GENERAL PROBLEMS OF THE CONCEPT OF ‘PRECAUTIONS’

The notion of the ‘precautionary principle’ Until some years ago, to find a substantial definition of the notion of the ‘precautionary principle’ or ‘precaution’ was not easy. Consulting general reference books proved not very successful. The famous Encyclopedia Britannica in 1977 mentioned under ‘Precaution’ only a novel by James Fenimore Cooper of the year 1820.15 Special reference books on law, too, give little information, if any. An attempt at a definition was made by the Codex Committee on General Principles of the Codex Alimentarius Commission. This Committee stated that the precautionary principle had been used more commonly in the context of international discussions on matters related to the environment, but might be extended to food safety aspects. Although there was no generally recognised definition, the precautionary principle was supposed to apply in cases where the scientific evidence was not conclusive enough to determine a level of protection, but where it seemed necessary to apply measures for the purposes of protecting public health. In the context of food safety, this would apply if risk assessment had not been completed, if specific difficulties appeared to qualify the risk, or if there might be a doubt as to the risk management measures to be taken. This would apply for instance to measures intended to control microbiological hazards where it was difficult to carry out a risk assessment due to uncertainty factors. In view of the serious hazards involved, strict control measures might be taken, although the scientific justification was not sufficient, as governments might need to apply emergency measures, for instance in the event of an outbreak of food-borne disease involving emerging pathogens. The Committee was careful to consider its own definition a mere

Risk decisions in cases of persisting scientific uncertainty

55

illustration of the precautionary principle and emphasised that this principle was generally linked to uncertainty in a risk assessment or specific difficulties in the establishment of risk management decisions. The precautionary principle was in any case related to health risks and was intended to address uncertainty or incomplete scientific evidence. If a scientific basis clearly existed – as it did in the view of the Committee in the case of BST (the hormone case) – there was no room for the precautionary principle. The view of the Codex Committee coincides quite clearly with the essence of the precautionary principle as it has been developed within the law of public security and in environmental law and finally has been extended to national and European food law, now being laid down in Art 7 of the Basic Regulation. It also gives an insight into some special problems of the precautionary principle.

The problem of international terminology and understanding The precautionary principle is ‘in principle’ well recognised in environmental law, not only in Germany but also cum grano salis in other Member States, in European law and in international law. It is, however, questionable, even if the terms could be regarded as synonyms, whether the understanding of this principle is the same in different legal orders based on different traditions. Similar problems arise regarding the terms ‘hazard’, ‘risk’ and ‘risk assessment’ as legal terms, unlike the same terms in a scientific sense. For example, a paper on the self-control of enterprises and Hazard Analysis and Critical Control Point (HACCP)16 demonstrated misleading differences in the English and the German version of the HACCP system of the Codex Alimentarius,17 the Council Directive 93/43/EEC on the hygiene of foodstuffs18 and the German Lebensmittelhygieneverordnung.19 Those terms, however, are central to and constitutive for the system of administrative law, especially the law of public security. In German law, the meaning of the term Gefahr – in English danger – has been developed in a long and difficult process of legislative activity, jurisprudence and academic discussion. This term has a twofold function: both to limit the powers of the state vis-à-vis its citizens and to create the powers necessary to assure public security and order. It is decisive for the German type of administrative law based on the rule of law. Probably the synonymous terms in other legal systems governed by the rule of law have a similar function, or are at least of similar importance – but within the specific traditions of each legal system. To develop, however, a common basis of understanding it is necessary to turn to comparative law.20

The development of the precautionary principle within the law of public security The development of modern administrative law concerning public security and order runs – especially in the field of environmental law but also, for

56 Risk and the Law example, in the field of pharmaceutical law – from the ‘classic’ warding off of danger (Gefahrenabwehr) to precaution (Vorsorge), from the term ‘danger’21 (Gefahr) to the term ‘reason for concern’ (Besorgnisanlaß) and to the term ‘risk’ (Risiko) as the general term. The tendency is to broaden both the reasons for and the restraints on public intervention.22 Nevertheless the general problem remains: even under the terms of the precautionary principle it is necessary to determine when reasons of precaution justify intervention. This is a decision about the limits of freedom. Whereas generally only an impending, specific, i.e. an evident, danger justifies an intervention, in special cases ruled by the precautionary approach an intervention is possible at an earlier stage. In these cases, however, the reference to precaution must be justified and the prerequisites of its application must be exactly determined. In environmental law a differentiation has meanwhile been established between ‘danger’ (Gefahr), which is to be warded off strictly, ‘reason for concern’ (in German terminology Besorgnisanlaß), which in the framework of precaution gives rise to a ‘duty to minimise the risks’ (in German terminology Minimierungsgebot) and the ‘residual risk’ (in German terminology Restrisiko), which is held to be an unavoidable feature of real life (sozialverträglich) and which therefore must be accepted.23 In this traditional differentiation the term ‘risk’ appears twice, firstly in the category of ‘residual risk’ and then as ‘risk’ in the framework of ‘precaution’. These ‘risks’ have to be distinguished from the ‘danger’ which has to be strictly warded off. Recent developments in German jurisprudence show other differentiations which use ‘risk’ as a general term referring, as in the sciences, to the whole scale of different results of the combination of the probability that damage will occur and the extent of that damage.24 Nevertheless, beyond all differences in terminology, in fact the differentiation mentioned above between ‘residual risk’, ‘simple risk’ as the field of precaution and ‘increased risk’, i.e. ‘danger’, remains the same. The main point is that in every case a discretionary decision weighing up the competing interests needs to be taken. The first decision must determine when the field of the residual risk which must be tolerated by everyone is left behind and the field of precaution is reached where the actors are obliged to minimise potential negative effects. The second decision must determine when the field of precaution is left behind and the field of the ‘increased risk’, i.e. in a legal sense ‘danger’, is reached where actions crossing this border line are not allowed and have to be banned by society. This differentiation is possible, however, only where maximum levels can be fixed by legislation, as has been done in EC food law for instance for pesticide residues.25 It must be emphasised that even such a fixing of maximum levels is a political decision, even though it is based on scientific research and advice. This is even more the case where such maximum levels cannot be fixed at all because serious scientific research and advice are unable to tell beyond which limit the risk begins. In these fields the differentiation is to be made between ‘residual risk’ and the ‘warding off of risks’, the criterion thereof being reasonableness (whether to

Risk decisions in cases of persisting scientific uncertainty

57

allow the residual risk or not). This term clearly demonstrates that this is a political decision, even if it be science based.

Conclusions: some particular problems of the precautionary principle Some particular problems of the precautionary principle are evident. While this principle is recognised at least ‘in principle’ in some areas of the law, especially in environmental law, it is difficult to define its consequences for individual cases, i.e. the application of the principle in practice. The reason for this is that the precautionary principle applies to situations of recognised uncertainty, where a risk assessment has been made but where, due to the limits of scientific recognition, this has not led to a clearly science-based decision, whereas serious risks, the existence of which cannot be excluded, require preventive, i.e. ‘cautious’, action. The determination of the situations which justify the application of the precautionary principle, and the appropriate extent of ‘caution’, are political decisions, even if they be partly based on scientific evidence. In this context it must be emphasised that the reference to ‘science’ is not necessarily a reference to objective data and presumptions. In the field of the protection of the environment as well as in the field of health protection, and therefore in relation to an important aspect of food law, science is often unable to say whether a line of causality exists or not. It is just for these cases of non liquet on both sides, as neither the risk nor the exclusion of any risk can be proven because absence of evidence is not evidence of absence, that the precautionary principle has been developed and is necessary. Finally, in some fields it is questionable which decision conforms to the requirements of the precautionary principle. In that case it can be shaped to support any cause when the protagonists are arguing about the future, since this future does not exist except in their imagination.26

THE PRECAUTIONARY PRINCIPLE IN GERMAN FOOD LAW The ‘classic’ aims of food law are to protect against health damage, to protect against deception and fraud, to ensure fair trading practices and to guarantee proper product information. Having in mind the principle of proportionality, the wide-ranging restrictions the precautionary principle imposes on the producer can be justified, if at all, only by the first aim, protection against health damage. Nevertheless, even in this field, national food law – and not only German food law – is governed by the principle that the person placing a product on the market is responsible for ensuring that the product is safe, which means that foodstuffs likely to damage human health or which are otherwise unfit for consumption may not be marketed (the so-called ‘principle to forbid the abuse’, in German

58

Risk and the Law

terminology Mißbrauchsprinzip).27 There are, however, special areas where food law and legislation can no longer confine itself as it once did to warding off immediate dangers in the meaning mentioned above under the German term (klassische) Gefahrenabwehr. Instead, it must be geared towards preventing potential health risks. Therefore German food law pursued – and in implementing European food law still pursues – in the field of additives the so-called ‘prohibition principle’ (in German terminology Verbotsprinzip). This means that the use of additives is allowed only by special permission. Having in mind the potential risks of additives, the principle of preventive health protection justifies or even requires such restrictions, given that it is a task of the state to protect its citizens against dangers to life and health. Although German food law does not expressly differentiate between the terms ‘danger’ and ‘risk’,28 the principle of preventive health protection as realised in the German rules on additives can in substance be regarded as guided by the precautionary principle.29

THE PRECAUTIONARY PRINCIPLE IN EC FOOD LAW

General approach Although the provisions on the free movement of goods of the EC Treaty, well known from the famous Cassis de Dijon jurisprudence of the European Court of Justice,30 and more than 40 years of legislative activity of the Community have replaced or determined national food law almost completely, until 21 February 2002, the date of entry into force of the Basic Regulation,31 there was no codification setting out the fundamental principles of food law within the Community and its Member States. The increasing calls for such a legislative framework led to the Commission’s ‘Green Paper on the General Principles of Food Law’,32 the White Paper33 and finally the so-called Basic Regulation laying down, inter alia, the general principles and requirements of food law. This contains the definition of ‘food’ (Article 2), the main definitions of ‘risk’ and risk-related matters (Article 3, No. 9–14), the method of risk analysis (Article 6) and the precautionary principle (Article 7). In general, EC law leaves the placing of foodstuffs on the market to the producer who is responsible for ensuring that the foodstuffs are safe (cf. Arts 14, 17 Basic Regulation). There is no prior authorisation necessary according to Community law, neither by Community authorities nor by national ones. There are, however, some exceptions regarding the use of additives and novel foods, where it is necessary to require prior authorisation in order to pursue the preventive protection of human health. This can be seen as implementing the precautionary approach. The precautionary approach can also be seen in the decision of the European Court of Justice concerning emergency measures against BSE,

Risk decisions in cases of persisting scientific uncertainty

59

the so-called mad cow disease. The ECJ found that the principle of proportionality had not been violated by the Commission Decision34 which led to a ban on exports from the United Kingdom. The ECJ said: At the time when the contested decision was adopted, there was great uncertainty as to the risks posed by live animals, bovine meat and derived products. Where there is uncertainty as to the existence or extent of risks to human health, the institutions may take protective measures without having to wait until the reality and seriousness of those risks become fully apparent.35 It is interesting that the Court held this approach to be backed, inter alia, by Art 174 (formerly Art 130r) paras 1 and 2 EC Treaty which contain the precautionary principle, as shown above.

Additives and novel foods According to Council Directive 89/107/EEC of 21 December 1988 on the approximation of the laws of the Member States concerning food additives authorised for use in foodstuffs intended for human consumption,36 additives need an authorisation. The substantive general criteria for the use of food additives are laid down in Annex II. Food additives can be approved only provided that, inter alia, there can be demonstrated a reasonable technological need and that they present no hazard to the health of the consumer at the level of use proposed, as far as can be judged on the basis of the scientific evidence available. To assess the possible harmful effects of a food additive or derivates thereof, it must be subjected to appropriate toxicological testing and evaluation. The evaluation should also take into account, for example, any cumulative, synergistic or potentiating, i.e. multiplying, effect of its use and the phenomenon of human intolerance to substances foreign to the body. All food additives must be kept under continuous observation and must be re-evaluated whenever necessary in the light of changing conditions of use and new scientific information. Approval for food additives must be limited to the lowest level of use necessary to achieve the desired effect and take into account any acceptable daily intake, or equivalent assessment, established for the food additive and the its probable daily intake from all sources. This concept is governed by the precautionary principle given the special uncertainties and the lack of knowledge regarding the effects of additives on human health.37 The precautionary approach is also followed in EC Regulation 258/97 of the European Parliament and the Council on novel foods and novel food ingredients of 27 January 1997.38 In order to protect public health it is, according to the terms of that regulation, necessary to ensure that novel foods and novel food ingredients are subject to a single safety evaluation through a Community procedure before they are placed on the market

60 Risk and the Law within the Community. Whereas Art 3 of this Regulation states only the selfevident provision that novel foods must not constitute a danger for the consumer, the precautionary approach can be seen in some parts of the procedure, which is differentiated according to different levels of risk. This is not the mere procedure of notification (Art 5) but the procedure of authorization (Art 6) in the case of genetically modified organisms (GMOs) combined with the procedure according to Council Directive 90/220/EEC (Art 9).39 The Directives on additives and GMOs, as well as in part the Novel Food Regulation, are based on the precautionary principle. It is another question whether their implementation by the Commission was led by this principle. This has been contested regarding the authorization of additives40 as well as regarding the authorization of GMOs.41 Whereas formerly only some Member States followed a very restrictive (‘cautious’) approach to GMOs,42 this trend has now been adopted by the European Commission and is the current policy of the Union. In October 1998 a de facto European moratorium on the approval of releases under Directive 90/220/EEC was introduced pending the drafting of new controls.43 In the meantime, Directive 2001/18/EC of the European Parliament and the Council of 12 March 2001 on the deliberate release into the environment of genetically modified organisms and repealing Council Directive 90/220/EEC,44 updating and strengthening the latter, was adopted. Nevertheless there is a controversy between the Member States whether there is enough ‘precaution’ or not.45 There are also controversies between the United States and the European Union as to whether the GMO Regulations of the EU and new proposals which go even further are compatible with World Trade Organization rules, and there are even such controversies between the Member States of the European Union.46

The precautionary principle according to the Commission Green Paper ‘The General Principles of Food Law in the European Union’, the Communication of the European Commission ‘Consumer Health and Food Safety’, the ‘Communication on the Precautionary Principle’ and the Basic Regulation With the ‘Green Paper on the General Principles of Food Law in the European Union’47 and its Communication on Consumer Health and Food Safety48 the EC Commission has laid the basis for a long-awaited discussion of the status and the development of Community food law. The approach was quite selfcritical; existing provisions were re-examined, undesirable developments admitted, obstacles revealed. In addition to general considerations, some specific proposals were put up for debate. But the importance of the general considerations should not be underestimated. It is not only the slogan ‘from stable to table’ (‘from farm to fork’)

Risk decisions in cases of persisting scientific uncertainty

61

that bears witness to a shift of emphasis to a holistic view of food law but also the newly developed approach to risk assessment with its stress on the principle of precaution. The precautionary principle was specified in the Communication of the Commission on this principle, the fourfold aim of which is to outline the Commission’s approach to using the precautionary principle, to establish Commission guidelines for applying it, to build a common understanding of how to assess, appraise, manage and communicate risks that science is not yet able to evaluate fully, and to avoid unwarranted recourse to the precautionary principle as a disguised form of protectionism. The Commission emphasises that the precautionary principle should be considered within a structured approach to the analysis of risk which comprises three elements: risk assessment, risk management and risk communication. It is particularly relevant to the management of risk and should not be confused with the element of caution that scientists apply in their assessment of scientific data. Recourse to the precautionary principle presupposes that potentially dangerous effects deriving from a phenomenon, product or process have been identified and that scientific evaluation does not allow the risk to be determined with sufficient certainty. The implementation of an approach based on the precautionary principle should start with a scientific evaluation, as complete as possible, and where possible identifying at each stage the degree of scientific uncertainty. Decision-makers need to be aware of the degree of uncertainty attached to the results of the evaluation of the scientific information available. Judging what is an ‘acceptable’ level of risk for society is an eminently political responsibility. Decision-makers faced with an unacceptable risk, scientific uncertainty and public concerns have a duty to find answers. Therefore, all these factors have to be taken into consideration. In some cases, the right answer may be not to act at all or at least not to introduce a binding legal measure. A wide range of initiatives is available if action should be taken, ranging from a legally binding measure to a research project or a recommendation. The decision-making procedure should be transparent and should involve as early as possible and to the extent reasonably possible all interested parties. Where action is deemed necessary, measures based on the precautionary principle should be, inter alia, proportional to the chosen level of protection, non-discriminatory in their application, consistent with similar measures already taken, based on an examination of the potential benefits and costs of action or lack of action (including, where appropriate and feasible, an economic cost/benefit analysis), subject to review in the light of new scientific data, and capable of assigning responsibility for producing the scientific evidence necessary for a more comprehensive risk assessment.49 This is the general approach formulated in general terms. A concrete conclusion is drawn in respect of emergency measures. Based on a decision of the

62 Risk and the Law European Court of Justice in the case of emergency measures against BSE,50 the Commission argues that paramount importance must be accorded to the protection of public health, even at the expense of serious, and possibly irreparable, damage to commercial and social interests. In general, the view of the Commission is influenced by the experiences it had during the BSE crisis which had led to harsh public criticism of the EC Commission and by the European Parliament which decided to set up a temporary Committee of inquiry to investigate alleged contraventions or maladministration in the implementation of Community law which were ascertained in that particular case.51 BSE, which for a long time had been viewed as a special British problem – erroneously in the not truly ‘BSE-free’ Germany and Bavaria – led to the inclusion of new and promising aspects in the Green Paper and the following special papers and finally the Basic Regulation. It is important that this approach will be not only a temporary action to please the public and the European Parliament but a durable one. The new approach regarding the role of scientific advice and political decision-making in risk management is remarkable in the context of the precautionary principle. The approach was specified in the Communication on Consumer Health and Food Safety and, as shown above, the Communication on the Precautionary Principle itself.52 The Commission accepts that scientific advice is of primary but not of exclusive importance. Community legislation had on a number of occasions recognised that other factors, in particular consumer needs and concerns, must also be taken into consideration during the decision-making process.53 This leads to the problem of combining the ‘classic’ goals of food law with ‘new concerns’ of consumers such as aspects of ecology, animal protection or employment. These are not immediate aspects of food law and must, while borne in mind, be separated from it.54 Furthermore, the Commission maintains that it is important to note the limits of the role of scientific committees within the scope of questions of food safety considered as questions of ‘real’ food law.55 A clear distinction has to be drawn between the concepts of risk assessment and risk management. Risk assessment is a scientifically based process consisting of the identification and characterisation of hazards, the assessment of exposure and the characterisation of the risk. Risk management, meanwhile, is the process of weighing policy alternatives in the light of the results of risk assessment and, if required, selecting and implementing appropriate control options, including regulatory measures. While the task of risk assessment might be delegated to scientific advisory bodies, the task of risk management remains the responsibility of the regulatory authorities. That means that at Community level it is the responsibility of the Council, the Commission and the European Parliament. This is a clear differentiation between the scientific basis of a political decision and the political decision itself which is mostly – but not exclusively – important where legislation for the uncertain future is required. This is, as demonstrated above, the field of application of the precautionary principle.

Risk decisions in cases of persisting scientific uncertainty

63

The Basic Regulation of 2002 lays down the principles of risk analysis and the precautionary principle. The Regulation states, inter alia, that where food law is aimed at the reduction, elimination or avoidance of a risk to health, the three interconnected components of risk analysis – risk assessment, risk management and risk communication – provide a systematic methodology for the determination of effective, proportionate and targeted measures or other actions to protect health. To inspire confidence in the scientific basis for food law, risk assessments should be undertaken in an independent, objective and transparent manner, on the basis of the scientific information and data available. It is recognised that scientific risk assessment alone cannot in every case provide all the information on which a risk management decision should be based. Other factors relevant to the matter under consideration can legitimately be taken into account, including societal, economic, traditional, ethical and environmental factors and the feasibility of controls. The methods of risk analysis are laid down in Art 6 of the Basic Regulation. The precautionary principle has been invoked to ensure health protection in the Community, giving rise to barriers to the free movement of food or feed. Therefore it is necessary to adopt a uniform basis throughout the Community for the use of this principle. In those specific circumstances where a risk to life or health exists but scientific uncertainty persists, the precautionary principle provides a mechanism for determining risk management measures or other actions in order to ensure the high level of health protection chosen in the Community.56 Therefore Art 7 para 1 of the Basic Regulation deals with the specific circumstances where, following an assessment of available information, the possibility of harmful effects on health is identified but scientific uncertainty persists. It provides that in this case provisional risk management measures necessary to ensure the high level of health protection chosen in the Community may be adopted, pending further scientific information for a more comprehensive risk assessment. Measures adopted on the basis of para 1 shall be proportionate and no more restrictive of trade than is required to achieve the high level of health protection chosen in the Community, regard being had to technical and economic feasibility and other factors regarded as legitimate in the matter under consideration. The measures shall be reviewed within a reasonable period of time, depending on the nature of the risk to life or health identified and the type of scientific information needed to clarify the scientific uncertainty and to conduct a more comprehensive risk assessment.

64

Risk and the Law

THE PRECAUTIONARY PRINCIPLE IN WORLD TRADE LAW

The view of the Appellate Body in the Hormone Case The background to the Hormone Case: the EC ban on the use of hormones in meat production The precautionary principle was the decisive argument of the EC Commission in defending the EC rules banning the use of artificial and natural hormones in meat production. The EC rules were motivated by concerns for consumer protection and food safety. However, the underlying problem of so-called beef mountains in the EC which threatened to be increased by the use of BST which promoted the growth of cattle was not far from the regulator’s mind in 1981 and 1988, when these natural hormones were banned.57 Therefore the major beef exporting countries, and in particular the USA and Canada, held the 1988 ban to be a disguised barrier to trade specifically designed to restrict access to the EC market.58 The legal background to the differences in opinion between the USA and Canada and the EC The USA and Canada challenged, within the WTO dispute-settlement system, the EC ban on the basis of the Agreement on Sanitary and Phytosanitary Measures (SPS Agreement).59 This Agreement is a part of the Agreement Establishing the World Trade Organization of 15 April 1994.60 It interprets the ordre public clause of Art XX lit. b of the General Agreement on Tariffs and Trade (GATT),61 which was included in GATT 1994, using a more objective approach, and employing relatively detailed standards of evaluation.62 The SPS Agreement differentiates, generally speaking, as follows. If a national measure – and ‘national’ measures in this context are also EC provisions such as the directives banning hormones in meat production or the Novel Food Regulation – conforms to international standards, guidelines and recommendations, e.g. of the Codex Alimentarius Commission, it is justified without more (cf. Art 3 para 1 SPS Agreement). It is, however, problematic if a state or the EC wants to deviate from them, as the EC did in the case of the growth-promoting hormones. The SPS Agreement expressly permits such action, but imposes the burden of justification on the state concerned or on the EC, which is the competent actor in this field (cf. Art 3 para 3 SPS Agreement). This justification must be based on scientific reasons which have to include the international standard. Of course, this international standard is not absolute. The SPS Agreement shows this by the mere fact that it expressly allows for deviations. The matter in question is the quality and intensity of the ‘scientific’ justification required for such deviations. Is this only an intensified onus of explanation,

Risk decisions in cases of persisting scientific uncertainty

65

or is it an onus of proof to the effect that the unsuitability of the international rule has to be proven? The answer to this question is decisive for the degree of restriction of political autonomy of the Member States of the WTO. Of particular importance in this context are the statements concerning the assessment of risks under Art 5 of the SPS Agreement, especially the interpretation of Art 5 para. 1 and Art 5 para 2, according to which WTO members: ensure that their sanitary or phytosanitary measures are based on an assessment, as appropriate to the circumstances, of the risks to human, animal or plant life or health, taking into account risk assessment techniques developed by the relevant international organizations. The view of the Appellate Body According to the Appellate Body, in their assessment of risk Member States are not restricted to those hazards to human health which can be proven by scientific means. In assessing risks Members are not restricted to the factors listed in Art 5 para 2 SPS Agreement, which is not to be interpreted as a closed list. They can also take into consideration elements of ‘risk management’, such as the potential abuse of control. Scientific inquiry and evaluation of risk are indeed only the starting point of any risk assessment under Art 5 SPS Agreement. The Appellate Body held that the Panel had been in error when it excluded from the scope of risk assessment in the sense of Art 5 para 1 SPS Agreement all matters not susceptible to quantitative analysis by the empirical or experimental laboratory methods commonly associated with the physical sciences.63 The approach of adopting a minimum magnitude of risk, i.e. a de minimis clause, was rejected.64 A measure is based on a risk assessment if it is reasonably supported by the results of scientific studies which means that there has to be a rational relationship between the measure and the risk assessment.65 In this respect it is not necessary that the scientific study which is put forward represents the ‘mainstream’ of opinion in the relevant scientific community. The divergent expertise of some qualified and respected scientists who have investigated the particular issue at hand carefully is sufficient.66 Whether or not the requirements of Art 5 para 1 SPS Agreement are fulfilled can be decided only on a case-by-case basis.67 This certainly is not a licence for a risk assessment based on superficial, weakly substantiated studies. The Appellate Body emphasised that the precautionary principle does not override the regulations of Art 5 para 1 and 2 SPS Agreement.68 But the precautionary principle is incorporated in Art 5 para 7 SPS Agreement, which, however, does not exhaust the relevance of this principle which is also reflected in the sixth paragraph of the preamble and in Art 3 para 3. These provisions explicitly recognise the right of Member States to establish their own appropriate level of

66 Risk and the Law sanitary protection. This level might be higher (i.e. more cautious) than the level implied in existing international standards, guidelines and recommendations.69 Conclusions concerning the precautionary principle In response to the EC’s claim based on the precautionary principle, Canada had claimed that the EC had not explained what was meant by this principle being ‘built into’ other provisions of the SPS Agreement. Although an exact definition of the very essence of the precautionary principle was not given, the Appellate Body has clearly stated two points: on the one hand this principle could not strictly override concrete provisions laid down in the SPS Agreement, but on the other hand, the precautionary principle determined the meaning of the provisions of the agreement. It is remarkable that the Appellate Body recognised the need for taking sanitary measures before a scientific proof for a human health hazard could be presented.70 The risk to be evaluated is not only risk ascertainable in a science laboratory but also ‘risk in human societies as they actually exist in the real world where people live and work and die’.71 This certainly does not replace a scientific risk assessment by a ‘consumer’s view’, for that would open the door for protectionism. The view of the Appellate Body may be interpreted to mean that Art 5 para 1 SPS Agreement leaves a far-reaching margin of appreciation to Member States as well as to the EC for precautionary measures as long as these are based on scientific findings. In respect of the decisive questions – what kind of scientific justification and what degree of scientific certainty the SPS Agreement requires – the Appellate Body held that respectable scientific minority views might constitute a sufficient justification. That holding may be regarded as a rule of doubt based on the maxim ‘safety first’. The problem is that the more advanced a technology is, the easier it will be for the defendant party to find experts of recognised standing who point out its dangers. So it may be comparatively easy to erect trade barriers and comparatively difficult to introduce new technologies into foreign markets.72 Therefore the Appellate Body decision can be only the beginning of the path to legal certainty – in so far as such certainty can be reached at all – concerning evidentiary requirements where WTO rules rely on scientific proof.

The controversies over GMOs: the dispute between the United States and the EC As a restriction on international trade the Novel Food Regulation of 199773 and the Regulations on labelling74 as well as the Regulations on GMOs75 must be in accordance with WTO Rules,76 like any regulation that restricts the liberty of economic activities. First of all, the United States objected to Commission Regulation (EC) No. 1813/97 of 19 September 1997 concerning the compulsory indication on the labelling of certain foodstuffs produced

Risk decisions in cases of persisting scientific uncertainty

67

from genetically modified organisms of particulars of other than those provided for in Directive 79/112/EC,77 which had been notified by the EC according to the Agreement on Technical Barriers to Trade (TBT Agreement). The Regulation required the labelling of genetically modified maize and soybeans. To impose an obligation to label simply because of the evidence of modified DNA or the insertion of a new protein would not, it was argued, be ‘scientifically’ justified and was therefore contrary to the principle of proportionality.78 Whereas the WTO Rules in the interpretation given by the Appellate Body in the Hormone Case leave a wide discretion to the EC authorities, there are problems, especially in the lack of precision of the central provisions, e.g. concerning the scope of application and the requirements for labelling. According to the principle of proportionality, safety controls carried out by third states have to be recognised as equivalent to those of the home state. The controversies went on in the field of marketing GMO products in the European Community and worldwide and may increase, bearing in mind the importance of GMOs in US export trade.79

The Cartagena Protocol on Biosafety The Cartagena Protocol on Biosafety to the United Nations Convention on Biological Diversity was finally adopted in January 2000 following several years of negotiations.80 This agreement is generally designed to address the uncertainty and incomplete nature of scientific knowledge of how Living Modified Organisms (LMOs)81 interact with biodiversity and the surrounding environment82 and incorporates the precautionary principle into its framework. It is specifically concerned with risk and trade aspects of LMOs and represents one of the first legally binding international agreements governing the transboundary transfer of genetically modified organisms. The text on precautionary decision-making in the Protocol represents an amalgamation of existing formulations in previous global agreements.83 The ‘precautionary’ approach established by the Protocol is broadly consistent with Art 5 para 7 of the SPS Agreement, the starting point for precautionary measures being a (scientific) risk assessment rather than the much feared non-scientific criteria for decision-making.84 Overall, however, the incorporation of the precautionary principle into the Protocol does little to dispel the uncertainties surrounding the specific content and scope of the precautionary principle, leaving the concept open to the criticism that it is both ‘bogus’ and ‘unscientific’.85

CONCLUSIONS The precautionary principle has been developed and is today relatively well established in the field of environmental law, not only in Germany but – perhaps in a different way, but at least substantially – in other countries and

68

Risk and the Law

especially in public international law. In this field the term ‘precautionary’ can be found in early English literature. In recent years this term has been used in food law also, especially on the international or supranational level like the Codex Alimentarius, the Council of Europe and the European communities. As the definition given in the framework of the Codex Alimentarius and Art 7 of the Basic Regulation of 2002 and the outline given by the EC documents, the Green Paper, the Communication on Consumer Health and Safety and the Communication on the Precautionary Principle demonstrate, it is in essence the same principle as the one developed in environmental law. It is supposed to apply in cases where scientific evidence is not conclusive enough to determine a level of protection, but where there is a need to apply measures for the purposes of protecting public health. This situation, however, must be recognised as such on a scientific basis. There must be uncertainty and there must be at least a potential risk to an outstanding good like human health which, if realised, would cause enormous damage. This leads to the principle of proportionality. The precautionary principle can be and has already been applied in the fields of additives and novel foods for instance. Furthermore, it may be applicable to products which come near to being pharmaceuticals, the so-called ‘nutriceuticals’, i.e. food supplements (in German the so-called Nahrungsergänzungsmittel) and to some extent to the so-called ‘designer food’.86 It is furthermore applicable in emergency cases like the BSE crisis. In some cases where the uncertainty itself is debated controversially – as is the case in the hormone dispute and regarding GMOs – the labelling approach may be a solution. This proposal has been developed for the field of divergent risk assessment concerning novel food87 and has also been brought into the debate on the hormone ban.88

Notes 1 Official Journal of the European Communities (OJ), 1987 L, 187/1. 2 In the German version the corresponding notions are Grundsätze der Vorsorge und Vorbeugung, in French les principes de précaution et d’action préventive, in Spanish los principios de cautela y de acción preventiva, in Italian principi della precauzione e dell’azione preventiva, in Dutch het voorzorgsbeginsel en het beginsel van preventief handelen. 3 Cf. R. Streinz, ‘The Precautionary Principle in Food Law’, European Food Law Review (EFLR), 1998, 8, 413, 418 et seq.; J. Gelbert, Die Risikobewältigung im Lebensmittelrecht, Bayreuth: PCO Verlag, 2001, pp. 135 et seq. with further references; Alberto Alemanno, ‘Le principe de précaution en droit communautaire’, Revue du Droit de l’Union Européenne (RDUE), 2001, 917, 921 et seq. 4 COM(97) 176 final, pp. viii, 37. 5 Cf. Streinz, n. 3 above, with further references. 6 Commission Decision 96/239/EC of 27 March 1996, OJ 1996 L, 78/47. 7 Case C-157/96 – National Farmers’ Union [1998] ECJ I-2211, grounds 63 et seq; Case C-180/96 – United Kingdom v Commission [1998] ECJ I-2265, grounds 98 et seq.

Risk decisions in cases of persisting scientific uncertainty

69

8 ECJ, Judgment of 8 January 2002, Case C-248/99 P – France v Monsanto Company and Commission [2002] ECJ I-1, ground 66; Judgment of 22 October 2002, Case 241/01 – National Farmers’ Union, grounds 1, 22 et seq., 40 (nyr); EC, Judgments of 11 September 2002, Case T-13/99, Pfizer Animal Health v Council, grounds 118, 121 et seq, 149, 156, 159; Case T-70/99 – Alpharma v Council, grounds 139, 142 et seq., 162, 169, 172, 359. 9 Case E-3/00 – EFTA Surveillance Authority v Norway, Zeitschrift für Lebensmittelrecht (ZLR), 29, 2001, 697. 10 COM(97) 176 final. 11 COM(97) 183 final. 12 COM(1999) 719 final. 13 COM(2000) 1. 14 OJ 2002, L 31/1. This article came into force on 21 February 2002 according to Art 65 Basic Regulation. 15 Encyclopedia Britannica, Micropedia, vol. VIII, 1977, p. 183. 16 M. Hartig and F. Untermann, ‘Betriebliche Eigenkontrolle und HACCP: Entwicklung der Rechtssetzung in der EU’, ZLR, 25, 1997, 602, 602 et seq. 17 Alinorm 97/14A. 18 OJ 1993, L, 175/1. 19 Bundesgesetzblatt (German Federal Law Gazette), 1997, I, 2008. 20 Cf. also Elizabeth Fisher, ‘Precaution, Precaution Everywhere: Developing a “Common Understanding” of the Precautionary Principle in the European Community’, Maastricht Journal (MJ), 9, 2002, 7, 8 et seq. 21 In English in everyday usage ‘danger’, ‘hazard’ and ‘risk’ are often regarded as synonyms – as to a certain extent also in everyday German are Gefahr and Risiko. For the purpose of a differentiated legal terminology ‘risk’ can be regarded as a ‘possible danger’, whereas ‘hazard’ seems to be understood as nearer to actual ‘danger’, but sometimes also to ‘risk’. Cf. e.g. The New Shorter Oxford English Dictionary, Oxford: Oxford University Press, 1993, and D. van Beseler and B. Jacobs-Wüstefeld, Law Dictionary English-German, 4th edn, Berlin, New York: de Gruyter, 1986. Here the term ‘danger’ is used for the German Gefahr, the term ‘risk’ for the German Risiko. In 1997 the Codex Alimentarius Commission adopted a number of definitions of risk analysis terms related to food safety. In this sense ‘risk’ is a function of the probability of an adverse health effect and the severity of that effect, consequential to a hazard in food, whereas ‘hazard’ is a biological, chemical or physical agent in, or condition of, food with the potential to cause an adverse health effect. 22 Cf. the detailed study of U. Di Fabio, Risikoentscheidung im Rechtsstaat, Tübingen: Mohr, 1994, pp. 65 et seq. and, regarding pharmaceutical law, pp. 166 et seq. 23 Cf. e.g. M. Kloepfer, Umweltrecht, 2nd edn, München: C.H. Beck, 1998, pp. 167 et seq., Marginal Number (MN), p. 171, MN, p. 16. 24 Cf. R. Wahl, ‘Risikobewertung und Risikobewältigung im Lebensmittelrecht’, ZLR 26, 1998, 275, 282 with further references. 25 Cf. e.g. Council Directive 96/33/EC amending the Annexes to Directives 86/362/EEC and 86/363/EEC on maximum levels for pesticide residues and foodstuffs of animal origin, OJ, 1996, L 144/35. 26 Cf. J. Adams, ‘The Precautionary Principle’, Economic Affairs, 1995, Vol. 16, 6. Cf. also Gelbert, n. 3 above, pp. 37 et seq. 27 Cf. The German Statute on Foodstuffs and Requisites (Lebensmittel- und Bedarfsgegenständegesetz – LMBG), § 38; Wahl, n. 24 above, 291 et seq. 28 Cf. Wahl, n. 24 above, 292. 29 Cf. the motives of the legislation, cited in W. Zipfel and K.-D. Rathke, Lebensmittelrecht, Loseblatt-Kommentar (Loose-leaf Commentary), München: C.H.

70

30 31 32

33 34 35 36 37 38 39

40 41

42

43 44 45 46

47

Risk and the Law Beck, 2002, vol. II, C 100, § 9 LMBG, MN 1, § 11 LMBG, MN 1. Cf. also D. Eckert, ‘Draft Framework Directive on Community Food Law’, in EUI Working Paper Law No. 94/4, 161, p. 164. Judgment of 20 February 1979, Case 120/78 – Rewe v Bundesmonopolverwaltung für Branntwein [1979] ECJ 649. An outline of some of the most important cases is presented in EFLR, 1993, 4, 183 et seq. See n. 14 above. See n. 4 above. On the process of evaluation of this Green Paper see ibid., p. 1, and F. Snyder, A Regulatory Framework for Foodstuffs in the Internal Market, EU Working Paper LAW No. 94/4; R. Streinz, ‘Trends und Perspektiven des europäischen Lebensmittelrechts im Zeichen des Grünbuchs und der Mitteilung der EG-Kommission “Gesundheit des Verbrauchers und Lebensmittelsicherheit” ’, ZLR 26, 1998, 145, 161 et seq. See n. 12 above. Commission Decision 96/239/EC of 27 March 1996, OJ, 1996, L, 78/47. ECJ, Case C-157/96 (n. 7 above), [1998] ECJ I-2211, ground 62 et seq. OJ 1989, L, 40/27. Cf. S. Schlacke, Risikoentscheidungen im europäischen Lebensmittelrecht, Baden-Baden: Nomos, 1998, 154 et seq.; Gelbert, n. 3 above, pp. 201 et seq. OJ, 1997, L, 43/1. Directive 90/220/EEC of the Council of 23 April 1990 on the deliberate release of genetically modified organisms into the environment, OJ, 1990, L 117/115. Cf. to this Directive in this context also O. McIntyre and T. Mosedale, ‘The Precautionary Principle as a Norm of Customary International Law’, Journal of Environmental Law, 1997, 221, 231. Cf. Schlacke, n. 37 above, 155. P. Sands, ‘In the Matter of Directive 90/220/EEC and in the Matter of the Authorisation by the European Commission of the Placing on the Market in the Community of Novartis Bt. Maize, Preliminary Opinion provided for the European Unit of Greenpeace International’, 1998, pp. 10 et seq. Austria has banned the GMO maize according to Art 16 of the Council Directive 90/220/EEC (n. 39 above). Cf. EU Food Law Monthly, No. 66/1997, 6 et seq, 10; No. 74/1998, 11 et seq. Luxemburg did the same. Other Member States, e.g. the United Kingdom, have recently followed a ‘precautionary approach’ too. It must be emphasised, however, that simply taking the ‘precautionary principle’ as a basis cannot replace a decision on the applications according to the rules of Council Directive 90/220/EEC (n. 39 above), having in mind, of course, the precautionary principle in so far as it influences and determines the provisions of this Directive. Cf. N. Salmon, ‘A European perspective on the precautionary principle, food safety and the free trade imperative of the WTO’, European Law Review, 27, 2002, 138, 142 et seq. OJ, 2001, L, 106/1. Directive 90/220/EEC was repealed on 17 October 2002. Cf. Salmon, n. 43 above, 142 et seq. Cf. e.g. on the position of France, EU Food Law Monthly, No. 122, 2002, 18; No. 123, 2002, 16. Cf. the position of UK and Germany, which have questioned whether the Commission proposals on traceability and GM labelling are compatible with WTO rules: EU Food Law Monthly, No. 121, 2002, 5 et seq. See also below, ‘The precautionary principle in world trade law’. The Council has reached (against the votes of the UK and the Netherlands, and Luxemburg which wants even stricter rules) a Common Position on regulating the traceability and labelling of GMOs and the traceability of food and feed products from GMOs: cf. World Food Law Monthly, 57, 2003, 6. COM(97) 176 final.

Risk decisions in cases of persisting scientific uncertainty

71

48 COM(97) 183 final. 49 COM(2000) 1, Summary, No. 2, 4, 5, 6. 50 ECJ Case C-180/96 R – United Kingdom/Commission, [1996] ECJ I-3903, para 90 et seq. See also the cases cited in n. 7 above. 51 See the Decision of the European Parliament of 17 July 1996, OJ, 1996, C, 239/1, the Report A 4–0020/97, and the Decision of 19 February 1997, OJ, 1997, C, 85/61. 52 Cf. also COM(2000) 1, Annex III. 53 COM(97) 176 final, p. viii. 54 Cf. ibid., p. 48. According to the Draft Framework Directive on Food Law, presented by A. Cleary, ‘EUI Working Paper Law No. 94/4’, pp. 138 et seq., these aspects should be included (ibid., p. 149, Art 2). For a holistic view on the basis of ‘real’ food law cf. R. Streinz, n. 32 above, 161 et seq. 55 COM(97) 176 final, pp. 35 et seq. 56 Basic Regulation (note 14), grounds 17–21. 57 Cf. the Council Directives quoted in G. Schohe, ‘The WTO Appellate Decision on Hormone-Treated Meat: Seriously Bridling EC Protectionism?’, EFRL 8, 1998, 253, 254, note 5. Cf. also Gelbert, n. 3 above, pp. 97 et seq. The ban has been upheld by Directive 96/22/EC of the Council of 29 April 1996, OJ, 1996, L, 125/3. On recent BST-problems cf. Gelbert, n. 3 above, pp. 110 et seq. 58 Cf. B. O’Connor, ‘EC measures concerning meat and meat products (hormones)’, EFLR 8, 1998, 143, 143. For the problem of scientific justification of the ban cf. also C. Jackson, ‘The Role of the European Parliament in the Control of Foodstuffs Legislation’, EFLR 1, 1990, 53, 61. Cf. also Gelbert, n. 3 above, pp. 93 et seq. 59 EC OJ, 1994, L, 336/40. 60 EC OJ, 1994, L, 336/1. 61 Text in L.B. Sohn (ed.), International Organisation and Integration, Dordrecht: Martinus Nijhoff, 1986, pp. 650 et seq. This article has been imitated by Art 30 [formerly Art 36] EC Treaty. Therefore the problems arising from both provisions are similar. 62 Cf. e.g. B. O’Connor and A. Bélen Castillas, ‘The Impact of the WTO on Trade in Food and Agricultural Products’, EFLR 7, 1997, 119. 63 Report of the Appellate Body, ‘EC Measures Concerning Meat and Meat Products (Hormones)’, WT/DS 26/AB/R and WT/DS 48/AB/R, No. 16. The report of the Appellate Body is partly published in Europäische Zeitschrift für Wirtschaftsrecht (EuZW), 1998, 157 (Nos. 181, 187, 206). 64 Ibid., No. 184, 186. 65 Ibid., No. 193. 66 Ibid., No. 194. 67 Ibid. 68 Ibid., No. 121, 125. 69 Ibid., No. 124, 253 lit. c. 70 Cf. ibid., No. 194 and 245. 71 Ibid., No. 187. 72 Schohe, n. 57 above, 261. 73 See n. 38 above. 74 Directive 90/218/EEC of the Council of 23 April 1990, which applies to the confined use of genetically modified micro-organisms. Regulation 49/2000/EC of the Commission of 10 January 2000 amending Council Regulation 1139/98 concerning the compulsory indication on the labelling of certain foodstuffs produced from genetically modified organisms of particular types other than those provided for in Directive 79/112, OJ 2000 L, 6/13; Regulation 50/2000/EC of the Commission of 10 January 2000, the labelling of foodstuffs and food

72

75 76 77 78 79

80 81 82 83 84 85 86

87 88

Risk and the Law ingredients containing additives and flavourings that have been genetically modified or have been produced from genetically modified organisms, OJ, 2000, L, 6/15. For EC legislation on Novel Foods see Confederation of the Food and Drink Industries of the EU (CIAA), Status Report on Food Legislation in the European Union, December 2002, pp. 39 et seq. Cf. nn. 39, 41. Cf. R. Streinz, ‘The Novel Foods Regulation – A Barrier to Trade in the International Movement of Goods?’, EFLR 8, 1998, 265. OJ, 1997, L, 257/7, repealed by Art. 3 Regulation 1139/98 (n. 72 above) and replaced by this Regulation. Cf. R. Streinz, n. 76 above, p. 286. Cf. e.g. World Food Law Monthly, No. 51/2002, 1 (‘Tough European stance on GM labelling increases the divide’) and No. 54/2002, 5 (‘Trade war between the US and EU’). Cf. also No. 55/2002 (‘Controversy for GM wheat in Asia’) and No. 58/2003, 20 (‘India rejects American GM-food consignment’). ‘Cartagena Protocol on Biosafety to the United Nations Convention on Biological Diversity’, UNEP/CBD/Ex COP/1/L.5, 28 January 2000. ‘Living modified organism’ means any living organism that possesses a novel combination of genetic material obtained through the use of modern biotechnology: Art 3(g) of the Protocol. Earth Negotiations Bulletin, Vol. 9, No. 137, 31 January 2000. A. Gupta, ‘Governing Trade in Genetically Modified Organisms: The Cartagena Protocol on Biosafety’, Environment, 42/4, 2000, 24. Ibid., 30; Salmon, n. 43 above, 152. H.I. Miller and G. Conko, ‘The Protocol’s Illusionary Principle’, Nature Biotechnology, 18, 2000, 360. Cf. Art 5, Directive 2002/46/EC of the European Parliament and the Council of 10 June 2002 on the approximation of the laws of the Member States relating to food supplements, OJ, 2002, L, 183/51. The precautionary principle is applicable as a general principle according to Art 7, Art 1 para 3, Art 4, para 2, Basic Regulation. Cf. R. Streinz, ‘Diverging risk-assessment and labelling’, EFLR 5, 1994, 155. Schohe, n. 57 above, 262.

ANNEX

Official Journal L 31/1 of 1.2.2001 REGULATION (EC) No. 178/2002 OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL of 28 January 2002 laying down the general principles and requirements of food law, establishing the European Food Safety Authority and laying down procedures in matters of food safety Article 2 Definition of ‘food’ For the purposes of this Regulation, ‘food’ (or ‘foodstuff’) means any substance or product, whether processed, partially processed or unprocessed, intended to be, or reasonably expected to be ingested by humans. ‘Food’ includes drink, chewing gum and any substance, including water, intentionally incorporated into the food during its manufacture, preparation or treatment. It includes water after the point of compliance as defined in Article 6 of Directive 98/83/EC and without prejudice to the requirements of Directives 80/778/EEC and 98/83/EC. ... Article 3 Other definitions For the purposes of this Regulation: ... 9. ‘risk’ means a function of the probability of an adverse health effect and the severity of that effect, consequential to a hazard; 10. ‘risk analysis’ means a process

consisting of three interconnected components: risk assessment, risk management and risk communication; 11. ‘risk assessment’ means a scientifically based process consisting of four steps: hazard identification, hazard characterisation, exposure assessment and risk characterisation; 12. ‘risk management’ means the process, distinct from risk assessment, of weighing policy alternatives in consultation with interested parties, considering risk assessment and other legitimate factors, and, if need be, selecting appropriate prevention and control options; 13. ‘risk communication’ means the interactive exchange of information and opinions throughout the risk analysis process as regards hazards and risks, riskrelated factors and risk perceptions, among risk assessors, risk managers, consumers, feed and food businesses, the academic community and other interested parties, including the explanation of risk assessment findings and the basis of risk management decisions;

74

Risk and the Law

14. ‘hazard’ means a biological, chemical or physical agent in, or condition of, food or feed with the potential to cause an adverse health effect; ... Article 6 Risk analysis 1. In order to achieve the general objective of a high level of protection of human health and life, food law shall be based on risk analysis except where this is not appropriate to the circumstances or the nature of the measure. 2. Risk assessment shall be based on the available scientific evidence and undertaken in an independent, objective and transparent manner. 3. Risk management shall take into account the results of risk assessment, and in particular the opinions of the Authority referred to in Article 22, other factors legitimate to the matter under consideration and the precautionary principle where the conditions laid down in Article 7(1) are relevant, in order to achieve the general objectives of food law established in Article 5.

Article 7 Precautionary principle 1. In specific circumstances where, following an assessment of available information, the possibility of harmful effects on health is identified but scientific uncertainty persists, provisional risk management measures necessary to ensure the high level of health protection chosen in the Community may be adopted, pending further scientific information for a more comprehensive risk assessment. 2. Measures adopted on the basis of paragraph 1 shall be proportionate and no more restrictive of trade than is required to achieve the high level of health protection chosen in the Community, regard being had to technical and economic feasibility and other factors regarded as legitimate in the matter under consideration. The measures shall be reviewed within a reasonable period of time, depending on the nature of the risk to life or health identified and the type of scientific information needed to clarify the scientific uncertainty and to conduct a more comprehensive risk assessment.

5

Risk and criminal law Gerhard Dannecker

THE DEVELOPMENT OF NEW RISKS Seriously antisocial behaviour is one of the chief concerns of criminal law. This raises the question whether the problems of the risk society1 can be solved by ‘risk criminal law’ (Risikostrafrecht) and whether or not a new definition of the functions of criminal provisions is necessary. This is an important subject matter of criminal political discourse in Germany. New dangers, both qualitative and quantitative, for example of atomic, chemical, ecological or genetic kinds, are typically characteristic of the risk society. The law, including criminal law, has to react to those dangers. In a proposal for a regulation concerning food law the European Commission states: In many cases, only criminal penalties will provide a strong dissuasive effect. The provision for such sanctions demonstrates social disapproval of a qualitatively different nature compared to administrative enforcement measures.2 In those circumstances only criminal law can guarantee strong preventive effects. Another characteristic of the risk society is that no single member of the society can any longer appreciate all existing risks all at once. Therefore it raises the question as to whether we need special regulations and principles such as strict liability offences, criminal liability of enterprises, etc.

THE RESPONSE OF THE LEGISLATOR TO THE RISK SOCIETY IN CRIMINAL MATTERS In Germany, criminal law is linked to the protection of legal interests. There are violations of the person, including the offences of causing death and wounding, certain public order offences, offences relating to safety standards at work and serious road traffic offences. Second, there are violations of general public interests. A third major sphere of liability comprises violations of

76 Risk and the Law the environment and the conditions of life, including various pollution offences, offences connected with health and purity standards and minor offences of public order. This shows that not only the infringement but also the endangering of legally protected interests can be punishable. The conditions to be fulfilled before an individual is convicted of an offence vary from one crime to another. Typically, only violations of interests are punishable. If criminal law protects social interests, the legislator normally decides that the endangering of a legally protected interest is a precondition for punishability (konkrete Gefährdungsdelikte). But alternatively a form of conduct by itself may be criminalised because of the consequences that may follow from it (abstrakte Gefährdungsdelikte). In the latter class of cases the interest which the law seeks to protect is not necessarily referred to in the definition of the criminal act; its protection can be just the motive of the legislator for the creation of the criminal provision. Thus, even if the accused can prove that he has not created any real danger, he will be convicted because the creation of a real danger is not necessary. During the twentieth century there was a growing class of crimes in which there was no necessity for harm. This was reflected in the broadening of the areas of offences covering environmental damage, white-collar crime and the law of genetic engineering. In the academic discussion in Germany there is a tendency to devote great attention to the rightness or wrongness of criminalising such forms of conduct. The boundaries of the criminal law are not treated as given. Justification must be provided in terms of sufficient reasons for invoking criminal sanctions against individual subjects. This raises the question of the circumstances in which a person, although he or she does not cause any harm, can nevertheless be held criminally liable. It is argued3 that problems of modern societies such as offences in the area of environmental damage, white-collar crime or criminality in matters of data processing, drugs, taxes and external trade cannot be eliminated by means of preventive criminal law. Behind this argument is the fear that the loss of substantial constitutional guarantees can result from the effective application of criminal law in these areas. Therefore, there are proposals4 to reduce criminal law in some core areas and there is a belief that the increasing problems of modern societies can be solved with the help of an intervention law that is located between criminal law and the law of administrative offences (Ordnungswidrigkeitenrecht). This new law of intervention would have guarantees and procedural regulations less distinctive than those of the criminal law. The sanctions should accordingly be less severe than regular criminal sanctions. But the limitation of criminal law to the core area of traditional offences stands for a return to a criminal law which favours the upper social class. A thief would be liable to criminal punishment whereas the well-off perpetrator of a white-collar crime or of an offence involving environmental

Risk and criminal law

77

damage would be confronted only with the lenient sanctions of the intervention law. In addition, another author, Schünemann,5 developed the following criticism. He says today is the era of plentiful trash, the production of which takes place through the constant, ruinous exploitation of natural resources. In these conditions the allocation of crimes against property to the core area of criminal law and simultaneous allocation of offences involving environmental damage to the field of administrative offences seems to be virtually conservative. Yet, criticism of the extension of modern criminal law is strengthened in so far as it is related to the tendency to apply criminal law in order to solve social problems even in cases in which it is impractical to use criminal law. In fact, the demand for the legislator to work out and present a programme of socio-political precautionary measures arises from the principle of subsidiarity. This, and not the creation of ineffective criminalisation, is the duty of the legislator. The great hazards in cases of inadequate dangerous acts must be tackled in the preliminary stage, before they cause any damage. Nowadays, the harmful chain of causation disappears in the anonymity of society. Under these circumstances, the criminal law must identify unacceptably dangerous acts in order to ensure the protection of legal interests. Therefore, an analysis of individual wrongful acts is necessary. If the prohibited behaviour is well defined, its relation to the legally protected interest is clear and the principle of ‘no guilt without fault’ is not infringed; it is legitimate to criminalise the preliminary stage of conduct. Even a crime like drink driving may not be excluded.

The exclusion of imputation in the case of a reasonable (permitted) risk Liability for negligence is created when there is said to be an unjustifiable risk of harm to others. In order to determine the degree of risk which gives rise to criminal liability for negligence one must take into account the fact that we all create some kind of risk to others in our everyday affairs without subjecting ourselves to liability. In fact, road traffic carries relevant risks to life, health and material goods, as accident statistics prove irrefutably. The legislator, however, permits road traffic because of the overriding public interest. The general permission to drive cars is based on a global balancing of interests. Driving is permitted even when it does not satisfy highly valued individual interests and even when it serves purposes which are disapproved of, e.g. the preparation of an offence. Thus, an act assuming a permitted risk does not constitute a criminal act. The infringement of a legally protected interest is not punishable if the rules of the road have been observed. This applies to offences committed intentionally as well as for those committed by negligence. In other words, although there is always some risk involved in

78

Risk and the Law

driving a car carefully, from the point of the law it is not unreasonable to take this small risk, bearing in mind the social utility of driving. In 1871, Ludwig von Bar described for the first time the cases of permitted risk.6 These cases have remained unaltered until the present time. He argued that there are business enterprises which are essential but surely dangerous and that it can statistically be assumed that, in all likelihood, over several years a number of persons will lose their lives as a result of these enterprises.7 Consequently the question arises, why those who run such business enterprises should be free of criminal responsibility despite their knowledge of the risks and if they have observed no more than ordinary care. Subsequently the restriction to businesses which are essential for the general public was abandoned and the range of cases of permitted risk, and the classes of interests, the satisfaction of which justified the taking of risks, were broadened. Besides essential interests, socially accepted and socially useful interests are now included. It is no longer necessary that interests of the general public are concerned since individual interests are also accepted nowadays. In addition, other cases are noted in the literature: risky rescue measures; risks in sports and games, where the players have agreed to the relevant acts and dangers; speculative deals by trustees which risk loss by an act not within the range of criminal breach of trust; and the exercise of rights in the furtherance of legitimate interests in the case of defamation. The legally defined concept of the permitted risk came into being in the era of industrialisation. At that time in history the human being was in a more secure and also a more insecure position than ever before. On the one hand, the expectation of life was growing because of technical and scientific progress. On the other hand, the risks, also resulting from technical progress, were bigger than ever because of an exponential increase in the number of hazardous activities and their potential for causing damage. These two conflicting aspects of technological progress may have been main reasons of the accompanying social and legal discussions concerning the costs and usefulness of technology. These facts may also have influenced the emergence of the legally defined concept of permitted risk. The field of permitted risk includes, inter alia, the whole area of public transport, the operations of industrial firms, especially the conduct of dangerous works, the exercise of sports fraught with risks, medical procedures, etc. The distinction between cases of permitted risks and cases in which no considerable risk is created can be quite difficult. But the uncertainties and the lack of clear distinction between these categories do not have practical consequences for the theory of imputation of criminal liability because the absence of the creation of a risk and the absence of causation of damage by the permitted risk exclude in similar ways the presence of the elements of a crime. In any case, a strong clue to identify instances where there is a permitted risk is the setting up of rules requiring the exercise of care. These are common in the fields of traffic, the running of technical installations, the

Risk and criminal law

79

exercise of dangerous sports, etc. The standardisation of protective measures proves the existence of a legally relevant risk.

Negligence In German criminal law, there is no criminal liability without fault. Even if the legislator sought to impose criminal liability for conduct unaccompanied by fault in order to help the prosecution to secure a conviction where intention, knowledge, recklessness or negligence is hard to prove, this would not be possible because of the principle nullum crimen sine culpa which is guaranteed by the German constitution (Grundgesetz). However, in Germany criminal liability generally does not require more fault than the degree of negligence which suffices for liability in tort. There does not exist a subjective test for criminal negligence, requiring that the person must know that his conduct is risky. However, he must, at least, be able to recognise that there is a risk. Moreover, the courts generally review the facts ex post facto in cases of negligent behaviour. Consequently, there is not a great difference in Germany between criminal liability for negligence and strict liability.

Liability of enterprises In Germany it is not accepted that a corporation may be held criminally liable for conduct performed by an agent of the corporation acting on its behalf within the scope of his employment, as is possible under English law. The traditional view in Europe was that a corporation could not be guilty of a crime: it had no mind and thus was incapable of the criminal intent then required for all crimes; it had no body and thus could not be imprisoned. This view has changed with the growth and the development of the corporate entity in the modern business world. In most European countries today it is conceded that a corporation may be criminally liable for the acts or omissions of its agents on its behalf. Although that view is taken in the Netherlands, France, Denmark and other countries, in Germany it is not yet accepted. Yet German law knows administrative sanctions in those circumstances. These are penal sanctions in a wide sense and are comparable to the sanctions of the EU in cartel law. Enterprises can be fined when the board of directors or a high-ranking manager acting on behalf of the enterprise participates in the prohibited conduct. The law was changed in 2001, so that it is now even sufficient if a person in middle management acted for the enterprise or failed to control the employees who acted for the enterprise. Fines can be imposed on the enterprise in addition to the managers’ punishment. In general, it can be stated that all statutes can be applied to corporations. The arguments for punishing enterprises are: Sometimes no one other than the corporation can be convicted because

80

Risk and the Law the division of responsibility within a corporation may be so great that it is difficult to fix it on an individual. Even when an individual agent could be convicted, it might be unjust to single out one person for substantial punishment when the offence resulted from habits common to the organisation as a whole. Fines should be paid by those who received the fruits of the illegal conduct in order to prevent unjust enrichment. Proceeding against the corporation serves to link the offence with the corporation in the public mind.

Often an officer of a corporation is not personally liable for the crimes of the corporation or of other corporate employees merely by virtue of the fact that he is an officer. The usual principles of accountability ordinarily apply, so that before he can be held liable it must be shown that the criminal acts were done under his direction or with his permission.

Potential fields of the risk dogmatic The scope of application of the risk dogmatic is not confined to the general part of criminal law. Thus, there are no appreciable counter-arguments to the opinion that the risk of subsequent illness and of foreign infection with an illness constitute forms of personal injury.8 Furthermore, the inclusion of the endangering of property in the category of property loss is almost generally accepted.9

CONCLUDING OBSERVATIONS Modern criminal law is characterised by the ‘bringing forward of criminal law in the sphere of endangerment’. The principle of criminal law that only legal interests are protected consequently is not observed any more. The problems of mechanisation can be solved only by measures of criminal law in cases in which the criminal law is an appropriate and indispensable instrument. When the criminal law is used, the basic principles of criminal law have to be observed. Therefore, it is not the legitimacy of modern criminal law in general which is questionable, but the legitimacy of concrete single provisions and specific dogmatic legal institutions. Nevertheless, we must raise the question to what extent criminal law has the capacity to counteract modern risks to life using traditional constitutional-liberal tools. Criminal law should not be regarded as a primary means of protecting individual and social interests. In terms of prevention of damage, more can probably be achieved through various techniques of situational crime prevention, social crime prevention and general social

Risk and criminal law

81

and educational policies. Nevertheless, no modern industrially developed country seems able to dispense with criminal law.

Notes 1 2 3 4 5 6

7 8 9

U. Beck, Risikogesellschaft. Auf dem Weg in eine andere Moderne, Frankfurt am Main: Suhrkamp, 1986. European Commission, COM(2003) 52 final, at para 55. F. Herzog, Gesellschaftliche Unsicherheit und strafrechtliche Daseinsvorsorge (Habilitation), Heidelberg, 1991. Zeitschrift fur Rechtspolitik, 1992, 210. Goltdammer’s Archiv für Strafrecht, 1995, pp. 210 et seq. C.L. von Bar, Die Lehre vom Kausalzusammenhang im Recht, besonders im Strafrecht, Leipzig, 1871; J. Rehberg, Zur Lehre vom ‘Erlaubten Risiko’ (Dissertation), Zürich, 1962, pp. 21 et seq.; F. Meckel, Die strafrechtliche Haftung für riskante Verhaltensweisen (Dissertation), Giessen, 1968, pp. 2 et seq.; H. Roeder, Die Einhaltung des sozialadäquaten Risikos, Berlin: Duncker & Humblot, 1969, p. 28; W. Preuss, Untersuchungen zum Erlaubten Risiko im Strafrecht, Berlin: Duncker & Humblot, 1974, pp. 15, 32 et seq. von Bar, n. 6 above, p.14. Entscheidungen des Bundesgerichtshofes in Strafsachen, 36, pp. 1 et seq., 6. Compare C. Prittwitz, ‘Das “Aids-Urteil” des Bundesgerichtshofs’, Strafverteidiger, 1989, pp. 123–128, 126. Compare A. Schönke et al. (eds.), Strafgesetzbuch Kommentar, Berlin and München: C.H. Beck, sect. 263, para 143 et seq.

Private law

6

The assumption of risk Ansgar Ohly

INTRODUCTION Better than any abstract definition, a case decided in New York in 1929 highlights the meaning of the term ‘assumption of risk’. The case is known as the ‘Flopper case’.1 A ‘Flopper’ is an attraction which could then and still can be found at fairgrounds. It consists of a belt moving upwards on which people step and either sit down or try to remain upright. Many do not manage to keep their balance and fall down, to the amusement of everyone standing by. The plaintiff, whom the judgment characterised as a ‘vigorous young man’, did just that: he stepped on the belt, fell down and suffered a fracture of his kneecap. Afterwards he sued the operators of the amusement park for damages. Cardozo CJ did not have much sympathy for the plaintiff: A fall was foreseen as one of the risks of the adventure. There would have been no point to the whole thing, no adventure about it, if the risk had not been there. The very name, above the gate, ‘the Flopper,’ was warning to the timid . . . One who takes part in such a sport accepts the dangers that inhere in it so far as they are obvious and necessary . . . The plaintiff was not seeking a retreat for meditation. Visitors were tumbling about the belt to the merriment of onlookers when he made his choice to join them. He took the chance of a like fate, with whatever damage to his body might ensue from such a fall. The timorous may stay at home.2 On the basis of this quite typical case, the following definition can be proposed: X can be said to ‘assume a risk’ when he takes a risk of which (a) he is aware or should be aware and which (b) has been caused or is controlled by Y. The crucial question is whether X can claim damages for personal injury3 or whether X’s ‘assumption of risk’ provides a defence for Y. Although ‘assumption of risk’ seems to be the most appropriate term to describe this situation, several synonyms exist. German lawyers usually refer to Handeln auf eigene Gefahr (acting at one’s own risk),4 while the maxim volenti non fit injuria has been used as a synonym for the voluntary assumption of risk in US5 and English law.6

86

Risk and the Law

Several typical factual situations discussed under this heading can be distinguished.7 First, the Flopper case is an instance of injuries occurring at dangerous places or due to dangerous events such as injuries suffered at fairgrounds,8 in swimming pools,9 on ski slopes,10 by spectators at sports events11 or by people attending rock concerts.12 Second, many cases concern passenger injuries resulting from dangerous driving in situations in which the passenger was aware of the dangers.13 A third category, to be distinguished from the first group of cases, concerns sports injuries suffered by the players.14 Fourth, some authors discuss occupier’s liability under this heading.15 However, this question will be excluded here as it is submitted that different principles apply to it. Finally, one case is rarely mentioned in this context: medical operations are not normally discussed as cases of ‘assumption of risk’,16 although the patient’s consent not only refers to an intentional interference (such as the opening of the abdominal wall) but also to the risk that negative consequences materialise from a treatment which is in accordance with the medical standard of care.17 As will have become evident, the legal issues concerning the assumption of risk are manifold. Two restrictions appear necessary. First, although in some of the categories mentioned above a contract will exist as between the parties, this chapter will analyse only tort liability. Second, while it is acknowledged that similar problems exist in English18 and US law,19 the following analysis is restricted to German law. Writing about German law in English in itself involves a risk. Many of the concepts used in the German Civil Code and in German legal doctrine are highly abstract and thus almost impossible to translate and hard to explain.20 At the risk of occasional imprecision, this chapter will omit some of the niceties of German legal doctrine and will instead focus on typical cases decided by German courts and present a model which may be partly transferable to other systems of tort law, thus trying to keep the risks of German conceptualism within reasonable bounds.

DEVELOPMENT OF GERMAN LEGAL DOCTRINE: THE LONG AND WINDING ROAD BETWEEN THE SUBJECTIVE AND OBJECTIVE APPROACHES Shortly after the Bürgerliches Gesetzbuch (the Civil Code of 1900 – BGB) had entered into force, the Reichsgericht (German Supreme Court before 1945) delivered its first judgments in cases concerning the assumption of risk. Although the line adopted in these early decisions is far from consistent, the court generally rejected the proposition that the claimant’s Handeln auf eigene Gefahr provided a full defence. Rather, the court subsumed these cases under the doctrine of contributory fault, which has been codified in § 254 BGB.21 However, in a judgment handed down in 1933,22 which concerned the case of a minor injured as a passenger in a motor cycle accident, the Reichsgericht

The assumption of risk

87

changed its approach. Impressed by an article by Flad,23 the court held that Handeln auf eigene Gefahr was a type of consent, which had to meet certain requirements in order to be valid. While the BGB remains silent on the elements of consent, the Reichsgericht held that consent was a Willenserklärung, i.e. a ‘statement of will/intention’. This is a typical example of an abstract concept used in the BGB. A ‘statement of will’ is a statement intended to bring about legal consequences. Offer and acceptance in contract law are typical ‘statements of will’, but unilateral declarations such as the unilateral termination of a contract are also within the scope of this term. The first book of the BGB, which contains general provisions valid for all branches of private law, contains validity requirements for ‘statements of will’ in §§ 104 et seq. BGB. These provisions concern inter alia statements made by minors (§§ 104 et seq. BGB), prescribed forms (§§ 125 et seq. BGB) or statements conflicting with statutory rules or public policy (§§ 134, 138 BGB). In its judgment of 1933, the Reichsgericht distinguished consent from the contractual exclusion of liability, which affects only the remedies rather than the lawfulness of the act itself, and held that consent was a unilateral declaration which resulted in a full justification of the permitted act. However, in order to be valid, consent had to fulfil the requirements imposed by §§ 104 et seq. BGB. § 111 BGB provides that statements made by minors which are not part of a contract (einseitige Rechtsgeschäfte) are valid only if the legal guardians have given their consent beforehand. Since evidently this had not happened in the case decided by the Reichsgericht, consent was invalid and the defendant was held liable. This approach can be characterised as subjective: the focus is not on the defendant’s objective duties of care but only on the subjective will expressed24 by the claimant. This approach was widely criticised by academic writers. Under the influence of this criticism the Bundesgerichtshof (Federal Supreme Court since 1945), which had initially followed the Reichsgericht’s approach, departed from this position in a judgment of 1961.25 Again, the case concerned a road accident. Four young men had travelled by car. Only the driver had a valid driver’s licence, whereas the three others were aged between 16 and 17 years and thus minors not yet allowed to drive cars.26 Nevertheless the driver allowed one of his passengers to take the wheel. Shortly afterwards, the new driver lost control of the car and caused an accident in which the third person was injured. According to the Reichsgericht’s approach, the injured person’s consent would have been invalid pursuant to § 111 BGB. The Bundesgerichtshof, however, rejected the consent approach with six arguments.27 First, the implication of a consent to the injury (as opposed to the act of driving the car) regularly amounted to a fiction in cases of this type. Second, §§ 134, 138 BGB and the criminal law provisions on manslaughter committed at the request of the victim (§ 228 Criminal Code – StGB), according to which no person can give valid consent to being killed, led to inappropriate results in cases concerning fatal accidents. Third, the requirement that statements of will must be communicated to the recipient excluded

88 Risk and the Law justification when the driver was too drunk to be able to receive the message. Fourth, in cases concerning minors the requirement of prior consent expressed by the legal guardians regularly led to the invalidity of consent since many risk decisions by minors were made spontaneously. Fifth, the doctrine of contributory fault provided a sliding scale for the attribution of liability which seemed preferable to the ‘all-or-nothing’ approach of the consent theory. Finally, the consent approach differed from the approach prevailing in criminal law which was undesirable given that the same act could trigger both civil and criminal sanctions. In the same year, Hans Stoll published his seminal thesis Das Handeln auf eigene Gefahr, in which he proposed an entirely objective approach. He shifted the focus from the victim’s consent to the duties of care of the person doing the injurious act: the claimant’s acting at his own risk started where the defendant’s duties of care ended. On this basis, Stoll distinguished between two situations. Whenever the defendant observed his duties of care he was not liable and thus the claimant acted at his own risk (echtes Handeln auf eigene Gefahr – genuine acting at one’s own risk). When, however, the defendant neglected his duties of care, he was liable in principle. In this situation, the idea of assumption of risk was reduced to the application of the doctrine of contributory fault (unechtes Handeln auf eigene Gefahr – false acting at one’s own risk). Stoll’s theory is an entirely objective one; the claimant’s ability to give valid consent and his intentions are immaterial. What matters are the defendant’s objective duties of care. Even in situations in which the claimant voluntarily takes a risk, the objective approach prevails: only volunteers ride the ‘Flopper’, a fact which is already taken into account when the operator’s duties of care are determined. For an account of the present legal situation, it is necessary to distinguish between the various groups of cases mentioned above. Judgments concerning the first two categories of ‘dangerous events’ and ‘passenger accidents’ generally follow the objective approach. A judgment delivered by the Bundesgerichtshof in a case in which a spectator was injured by a puck during an ice-hockey match28 is a typical example. The court does not inquire about the spectator’s intentions or his ability to consent. The reasoning concerns only the duties of care of an ice hockey club, particularly the necessary height of the transparent screen which protects the spectators in ice-hockey stadiums. When players are injured in games involving bodily contact (the third case), the courts draw a distinction. Players are not liable for injuries caused by acts which are in accordance with the rules of the game. Even liability for slight fouls of the type which regularly occur in the heat of the game is regularly excluded. Serious violations of the rules, however, result in full liability, which may be reduced under the doctrine of contributory fault. While this result seems to be generally accepted, the doctrinal basis remains unclear and disputed by academics. In a leading case concerning football injuries,29 the Bundesgerichtshof deliberately left this question open.

The assumption of risk

89

An area in which the subjective approach applies is the last case, that of medical law. An operation is lawful only with the patient’s prior informed consent. This consent has to satisfy certain validity requirements concerning issues such as whether the necessary information has been provided to the patient, or the issues concerning consent given by minors. It should be noted, though, that the patient’s consent does not replace the objective duties of care, but only modifies them. Whereas without the patient’s explicit consent treatment is allowed only in emergency situations, consent regularly permits treatment but only that which is in accordance with the medical standards of care. At this point it seems intriguing to ask whether a patient can also explicitly allow a treatment which is not in accordance with the prevailing standards of care. In a criminal case decided by the Bundesgerichtshof,30 a woman had asked her dentist to extract all of her teeth because she expected this operation to stop the permanent headache she was suffering from. The dentist informed her that her expectation had no medical foundation whatsoever. When she nevertheless insisted, he performed the operation. The Bundesgerichtshof convicted him for negligent battery, but most critics reject this decision. Indeed, this paternalist attitude is hard to sustain, given that the woman was given all the necessary information and given that teeth can be replaced. However, this issue leads away from the assumption of risk to the general doctrine of consent, particularly to the unresolved issue of how far consent can go.

ANALYSIS: THE SUBJECTIVE APPROACH (PARTLY) REHABILITATED

Starting point It has become evident that two approaches conflict. The maxim volenti non fit injuria represents the subjective approach: no harm is done to a person who permits the injury. The concept underlying this theory is personal autonomy. A person can surrender or risk his property as he sees fit, as long as the rights of third persons are not affected. To some extent the same is true for bodily integrity, although here consent is arguably subject to a reasonableness test:31 bodily integrity can be surrendered or risked when this appears reasonable given the result that is to be expected from the dangerous act.32 At least slight injuries can also be permitted for motives that may appear unreasonable to others, e.g. for sexual gratification. Both in German and in English law the limits of personal autonomy cause intricate problems.33 The objective approach does not look at the injured person’s voluntas but at objective rules regulating dangerous social contacts. Risks are controlled by objective rules which are imposed by legislation or case law. These rules are the result of a balancing exercise influenced by various parameters

90

Risk and the Law

including the protection of potential victims and the intensity of the risk on the one hand and the freedom of action and the costs of prevention on the other hand. The injured person’s intentions have only an indirect function; the duties of care may be less strict when the persons who are exposed to the danger can be expected to take their own precautions.34 It is submitted that both approaches should be combined. The cases under analysis here regularly concern the undesired outcome of a desirable social contact.35 This contact can be regulated by objective rules, but respect for personal autonomy requires that these rules may be modified by subjective agreement or unilateral permission.

Consent to risk Consent is an act of communication by which a person allows an interference with his rights which would otherwise have been unlawful. Put in simple words, consent is equivalent to the statement ‘You may’. Consent provides a full defence, the jurisprudential legitimation of which is strong because it is not derived from a utilitarian balancing exercise but from personal autonomy.36 In order to justify the far-reaching consequences of consent it must satisfy strict validity requirements. In many jurisdictions courts have been tempted to imply consent when this implication led to objectively desirable results.37 This temptation should be resisted, both in cases concerning intentional interferences and in cases concerning risks. In this respect the Bundesgerichtshof deserves approval for its judgment of 1961. Not every participation in a dangerous situation amounts to a consent. However, this does not mean that consent is necessarily restricted to cases of intentional interference.38 Whenever a person has had the chance to balance the desirability of the act and the potential damage and reaches the conclusion that the act should be permitted regardless of its potentially damaging consequences, this permission is also an expression of personal autonomy. This idea may be illustrated by an example taken from the case law of the Reichsgericht in criminal matters.39 Two persons intended to cross a river in spite of floods and stormy weather. The ferryman rejected their proposal and pointed out the dangers. However, they insisted and challenged him by doubting his courage. Eventually he gave in. Halfway across the river, the boat capsized and both passengers were killed. It is submitted that the passengers gave their consent.40 They certainly did not intend to commit suicide but they had had a chance to weigh the risks of the enterprise against its potential benefit and they communicated their decision to the ferryman. His objective duties of care prohibited him to go out under the dangerous weather conditions, but the passengers’ permission modified his duties. This example can be generalised: it is submitted that an explicit permission given in knowledge of the potential undesired consequence can be regarded as consent.41 Thus, in particular, a patient’s ‘informed consent’ is a consent in the proper sense of the word, also insofar as it relates to the

The assumption of risk

91

risk of negative consequences caused by an operation which is performed lege artis. It is true that the distinction between a ‘consent to risk’ and mere conscious participation in a dangerous situation may be hard to draw. However, the distinction must be made in all consent cases, also in those concerning intentional interferences. Consent must not be easily implied; mere inactivity never amounts to consent. Finally it should be pointed out that there is a difference between consent and the contractual exclusion of liability. As the Reichsgericht correctly pointed out,42 consent justifies the act in question whereas the exclusion of liability affects only the remedies. There are several cases concerning the implied exclusion of liability. Generally German courts have adopted a strict analysis. The exclusion of liability generally requires an explicit agreement, and an implicit exclusion can be assumed only under exceptional circumstances.43

Duties of care According to the theory outlined above, consent is restricted to situations in which the claimant (a) knows about the risk and (b) explicitly allows the defendant to perform the dangerous act. Consequently in most situations referred to in the introduction there is no consent because the parties do not discuss the dangers and because the defendant’s behaviour remains the same regardless of what the claimant explains. The operator of an icehockey stadium is not interested in which risks individual spectators are prepared to allow. In these situations, the objective approach as explained by Stoll prevails. Liability depends on whether the defendant observes his duties of care. Thus, in the ice-hockey case the Bundesgerichtshof was right to restrict its investigation to the objective duties of ice-hockey clubs. If the screen had been regarded as sufficiently high44 it would not have mattered if the injured spectator had been a minor or a drunk and thus a person unable to give valid consent. In this situation it would be misleading to refer to the maxim volenti non fit injuria; the claimant’s voluntas is immaterial. Finally, it should be pointed out that the objective approach needs to be supplemented by the doctrine of contributory fault. Even if the defendant has neglected his duties of care, the amount of damages can still be reduced if the claimant’s behaviour has contributed to the damage. However, mere participation in a dangerous activity regulated by objective rules will rarely amount to contributory fault; a spectator of an ice-hockey game can trust that the necessary security measures have been taken.

92

Risk and the Law

SUMMARY: THE BRAVE AND THE TIMOROUS The result is that the two approaches should be combined. Consent is not restricted to cases of intentional interference. What matters is not the nature of the interference but the nature of the permission. An explicit permission to perform a dangerous act given in knowledge of the potential undesirable consequences is a consent in the proper sense of the word. The maxim volenti non fit injuria should be restricted to this situation because only in this constellation are the claimant’s intentions material. If, however, the claimant has merely participated in a dangerous situation without consenting to its potential consequences, the implication of consent would amount to a fiction. Liability can thus be excluded only under the objective approach. Risks are regulated by objective duties of care and a defendant who observes these duties is not liable. According to this theory, the claimant in the ‘Flopper case’ did not consent to his injury. While he should have been aware of the danger, he did not permit the defendant to behave in a certain way. Had he given his permission to the operator’s representatives they would not have changed their behaviour. The ‘Flopper’ is a typical instance of a dangerous situation regulated by objective rules. Fairground attractions have to conform to certain safety standards. If they do, liability will be excluded. Thus, while the result of the Flopper case is to be approved, the subjective approach adopted in the reasoning is not. To sum up, the brave may be able to negotiate the terms on which a dangerous social contact takes place. Even the timorous are often protected by objective duties of care. But when they are not, they will have to stay at home.

Notes 1 2 3 4

Murphy v Steeplechase Amusement Co., 166 NE 173 (NY 1929). Ibid. at p. 174. Which is usually at issue here. See Entscheidungen des Bundesgerichtshofes in Zivilsachen, 34, 355 at 364; H. Stoll, Das Handeln auf eigene Gefahr, Tübingen: Mohr, 1961; H. Schlosser, ‘Handeln auf eigene Gefahr’, Jura, 1985, 554; E. Deutsch, Allgemeines Haftungsrecht, 2nd ed, Köln, Berlin, Bonn und München, 1996, § 20 II (S. 328 et seq.); H. Lange and G. Schiemann, Schadensersatz, 3rd edn, Tübingen: Mohr Siebeck (2003), § 10 XIV (pp. 641 et seq.). 5 See Murphy v Steeplechase Amusement Co. (see n. 1 above) at 174. Meanwhile the term ‘assumption of risk’ seems to have become commonly used: see § 496B, Restatement (Second) of Torts; K.W. Simons, ‘Reflections on Assumption of Risk’, (2002) 50 UCLA L Rev 481. 6 Cf. Freeman v Home Office (No. 2) [1984] QB 524, p. 557 per Sir John Donaldson, MR; W.V.H. Rogers, Winfield & Jolowicz on Tort, 15th edn., London: Sweet & Maxwell, 1998, pp. 844, 852; R.A. Jaffey, ‘Volenti non fit injuria’ [1985] CLJ 87. 7 See also the typology developed by Stoll, n. 4 above, pp. 14 et seq.

The assumption of risk

93

8 See the Flopper case, n. 1 above; and AG Olpe, Monatsschrift fur Deutsches Recht, 1959, 302 (injuries suffered while dancing at a local beerfest). 9 See e.g. Bundesgerichtshof (BGH) Neue Juristische Wochenschrift (NJW) 1982, 1144. 10 See e.g. BGH Neue Juristische Wochenschrift-Rechtsprechungs report: Zivilrecht (NJW-RR) 2002, 1056; BGH NJW 1985, 620. 11 See e.g. BGH NJW 1984, 801 (injuries from being hit by puck during ice hockey match). 12 See e.g. BGH NJW 2001, 2019 (hearing damage); Oberlandesgericht Hamm, Monatsschrift fur Deutsches Recht 2002, 518; and Landesgericht Hechingen, NJW-RR 2003, 84 (stage diving). 13 See e.g. Entscheidungen des Reichsgerichts in Zivilsachen, 141, 262; Entscheidungen des Bundesgerichtshofes in Zivilsachen (BGHZ) 34, 355. 14 See e.g. BGHZ 63, 140 (football); BGH NJW 1976, 2161 (basketball). 15 See Stoll, n. 4 above, pp. 81 et seq. 16 But see E. Deutsch, ‘Das Persönlichkeitsrecht des Patienten’, Archiv fur die civilistische Praxis 192, 1992, 161, 167 and in his Medizinrecht, 4th edn, Berlin and Heidelberg: Springer, 1999, para 104. 17 Ibid. 18 See n. 6 above. 19 See n. 5 above. 20 For an English perspective on the language of the German civil code see B.S. Markesinis, The German Law of Obligations, Vol. II, The Law of Torts: A Comparative Introduction, 3rd edn, Oxford: Clarendon, 1997, p. 24. 21 § 254 (1) provides: If any fault of the injured party has contributed to the occurrence of the damage, the duty to compensate and the extent of the compensation to be made depend upon the circumstances, especially upon how far the injury has been caused predominantly by the one or the other party. (Translation in Markesinis, n. 20 above, p. 18) 22 Entscheidungen des Reichsgerichts in Zivilsachen (RGZ) 141, 262. 23 F. Flad, ‘Handeln auf eigene Gefahr. Zugleich ein Beitrag zur Beurteilung der sogenannten Gefälligkeitsfahren’, Das Recht. Rundschau für den deutschen Juristenstand, 1919, Vol. 23, cols 13–20. 24 Only the claimant’s will as communicated to the defendant is relevant (§§ 133, 157 BGB). Merely internal reservations which are not expressed are immaterial (§ 116 BGB). 25 BGHZ 31, 355. 26 In Germany the minimum age for driving is 18. Thus it now corresponds to the age of majority, whereas at the time the case was decided the age of majority was 21. 27 See n. 25 above at pp. 360 et seq. 28 BGH NJW 1984, 801. 29 BGHZ 63, 140. 30 BGH NJW 1978, 1206. 31 See A. Ohly, ‘Volenti non fit iniuria’ – die Einwilligung im Privatrecht, Tübingen: Mohr Siebeck, 2002, pp. 410 et seq., 421 with further references. 32 Pursuant to § 228 of the German Criminal Code, consent to a battery will be regarded as invalid if the act violates bonos mores (die guten Sitten). On possible interpretations of this provision see H. Niedermair, Körperverletzung mit Einwilligung und die Guten Sitten, München: C.H. Beck, 1999, pp. 92 et seq.; W. Stree in A. Schönke and H. Schröder (eds.), Strafgesetzbuch (Kommentar),

94

33 34 35 36 37

38 39 40

41

42 43 44

Risk and the Law 26th edn, München: C.H. Beck, 2001, § 228 n. 7; on the applicability in private law see Ohly, ibid., pp. 419 et seq. For borderline cases in German law see the dentist’s case (n. 30 above) and Ohly, n. 31 above, pp. 415 et seq.; in English law the issue was raised by the House of Lords in the criminal case of R. v Brown [1994] 1 AC 212. See Stoll n. 4 above, p. 367. See G. Jakobs, Strafrecht, Allgemeiner Teil, 2nd edn, Berlin: de Gruyter, 1991, para 7/126. Ohly, n. 31 above, pp. 65 et seq. For US law see Florida Publishing Co. v Fletcher 340 So.2d 914 (1976) (journalists taking photographs in private home after owner’s daughter had been killed by a fire held to be justified by ‘implied consent’). For English law see: Zino Davidoff v A & G Imports [1999] RPC 631 (parallel importation into EU held to be justified by implied consent although trade mark owner had explicitly ruled out this possibility in contract of sale), and see the critical comments by W.R. Cornish, ‘Trade Marks: Portcullis for the EEA’ European Intellectual Property Review (EIPR), 1998, 172, 177, R. Swift, Case Comment EIPR, 376, 378; and the ruling of the ECJ [2002] 1 CMLR 1 (on the issue of consent see para 45). For German law see: BGHZ 115, 123 at 127 et seq.; 116, 268 at 274 et seq.; 148, 97 (on the question of a client’s or patient’s implied consent to the use of his personal data after sale of lawyer’s firm or doctor’s practice). As many academic authors assume. See for German law Lange and Schiemann, n. 4 above, p. 648; for English law, Rogers, n. 6 above, p. 845. Entscheidungen des Reichsgerichts in Strafsachen 57, 172 – Memel case. Another issue is whether this consent was valid, given that § 216 of the Criminal Code states that even manslaughter with the victim’s consent is a crime. The prevailing view in criminal law is that this provision only applies to intentional interferences. The distinction can be further clarified by reference to the classification of the types of wilful behaviour in German criminal and private law. There are three types of wilful behaviour (Vorsatz): (1) where an injury is both intended and desired (Absicht), (2) behaviour where an injury is intended without being desired (dolus directus 2. Grades) and (3) behaviour where an injury is accepted as a possible consequence of an act performed for another purpose (dolus eventualis). This classification can also be applied to the victim’s behaviour, see BGHZ 31, 355 at 360. Whenever the victim permits an act while accepting its possible negative consequences, his state of mind is equivalent to dolus eventualis, although this term directly only applies to the offender. RGZ 141, 262 at 264. See H. Heinrichs in O. Palandt (ed.), Bürgerliches Gesetzbuch, 62nd edn, München: C.H. Beck, 2003, § 254, paras 70 et seq.; Lange and Schiemann, n. 4 above, § 10 XVI 2 (pp. 652 et seq.). Which it was not: see BGH NJW 1984, 801 at 802.

7

Risk and predictability in English common law Tim Kaye

CERTAINTY AND JUSTICE It is often said that the biggest challenge facing private law is that posed by the continual need to choose between certainty and justice. Certainty is a worthy goal, it is said, because everyone knows exactly where they stand and judicial decision-making can proceed from clearly articulated statements of legal principle. It also ensures that like cases are treated alike. The pursuit of certainty, however, can lead to anomalous outcomes and may not allow for sufficient differentiation between different circumstances. If pursued blindly, certainty can cause the law to ossify, making it appear stuck in a time warp and bearing no apparent connection to modern-day reality. Thus justice must step in to enable the law to achieve a fairer, more reasonable result – an interesting paradox since justice is usually portrayed as being administered blindfold! But the problem with justice is that it introduces uncertainty into the law. Legal principles are modified or overturned; policy arguments of extra-legal origin are admitted. Even morality and politics may find an echo in the courtroom. New legal principles are developed, but they contain elements of uncertainty within them so that, for example, parties may be entitled to a remedy only if they have behaved ‘reasonably’, though no corresponding hard-and-fast definition of ‘reasonableness’ is provided. The wholesale pursuit of justice, it is said, takes us back to a realm of uncertainty and potentially arbitrary decision-making, which is precisely the mischief which the rule of law is designed to overcome. After all, is not certainty a form of justice in its own right? And so the certainty–justice dilemma confronts us once again, albeit that it may have been reconstituted on a different plane or in a new context. This account may be thought so familiar as hardly to be worth recounting. Indeed, it may be claimed that the certainty–justice dichotomy is an inherently insoluble dilemma, so that there is little point in discussing it in the abstract: the issue simply needs to be confronted in relation to a specific set of facts, when it is possible to argue that either certainty or justice is the more appropriate in that particular context. There is some force in this

96 Risk and the Law contention, but perhaps not as much as is generally assumed. In fact, the certainty–justice dichotomy presents the problems outlined above only if an issue is viewed from a peculiar combination of perspectives. Philosophically, the dichotomy is problematic only to legal positivism and allied schools of thought. But, even then, it presents difficulties only if viewed from a particular temporal perspective as well: that is, after the events in question have already taken place and have become known.

LEGAL POSITIVISM Legal positivism is the philosophy of law which takes the view that ‘law’ and ‘legal principle’ are distinct from such things as morality, policy, religion and politics. Its theory of law is essentially descriptive, in that it is concerned principally to describe what the law is. But it also has a prescriptive element, in that it holds that extra-legal factors should not be allowed to impinge on legal decision-making. It would, however, be unfair to characterise legal positivism as suggesting that law is (or should be) hermetically sealed from what goes on in the rest of the world. A better metaphor, perhaps, is that legal positivism sees law as creating its own walled garden. Special tools are required to work within it, but there are many interesting products: some have considerable utility, others are even beautiful – though there are, admittedly, a few weeds. But to enter this distinctive world, it is important to use the gate provided: namely legislation. Thus morality, policy, religion and politics can and do have a part to play in the law, but only if first transformed into an appropriate legal form so as to demonstrate the necessary provenance.1 Without legislative intervention they are inappropriate tools for a judge, and any judge who does seek to employ them rather than ‘proper’ legal rules and principles is liable to be criticised for abusing his or her position and deciding a case on the basis of his or her personal standpoint rather than according to law.2 It is therefore obvious that legal positivism strongly favours certainty over justice. But a theory of law entirely shorn of notions of justice is a wholly unattractive proposition and few legal positivists today are prepared to embrace the theory in so bold a form. Perhaps the most famous example of a legal positivist’s attempting to ameliorate this stark approach is Hart’s allowance that any true legal system contains a ‘minimum content of natural law’.3 In any event, once it is accepted that morality, policy, religion and politics can be introduced into the legal system via legislation in order to produce outcomes which are more just, then it is obviously pertinent to ask: ‘Why wait?’ If both the mischief and an effective remedy can already be easily identified, why should a court decline to do justice in the case before it simply on the formalistic basis that there is not yet any legislation to justify it? Otherwise the court is legitimating socially unacceptable conduct, not just by the defendant involved in the instant case but also by anyone else

Risk and predictability in English common law

97

who engages in similar conduct in the future. Moreover, it may be that the legislature will be prompted to intervene only after a court has drawn attention to a deficiency in the law by acting in a non-positivistic fashion. A good example in English law is the case of Khorasanjian v Bush,4 in which the Court of Appeal was prepared to grant an injunction to a young woman who was being pestered by repeated telephone calls from her exboyfriend on the grounds that this amounted to a form of private nuisance. Private nuisance is a tort which has been in existence for well over a century, but any claimant before Khorasanjian had always needed to demonstrate to a court that he or she had suffered some damage to a proprietary interest (or some loss of its enjoyment) in order to mount a successful claim. Yet the claimant in Khorasanjian had no proprietary interest; she was merely a licensee in her parents’ house. The rationale for the court’s decision was effectively that this was a remedy required by justice, for there was no other means available by which the court could inhibit or prohibit the continual telephone calls. A few years later the UK Parliament intervened by passing the Protection from Harassment Act 1997, which would have provided the claimant in Khorasanjian with a worthwhile remedy if it had been in force at the time. Being aware of this legislation, the House of Lords felt able to declare in a subsequent case, Hunter v Canary Wharf,5 that a claimant was indeed, after all, required to prove damage to – or loss of enjoyment of – a proprietary interest in order to succeed in a claim of private nuisance. ‘Spouses and children of tenants who complained of dust and poor television reception caused by the erection of a large building nearby’ were thus denied any compensation.6 Though it might be argued that dust and poor television reception were mischiefs needing correction in their own right, it is clear from a number of obiter remarks that their Lordships felt that private nuisance was essentially a ‘tort to land’ and that claims for personal injury and inconvenience should properly be brought under the heading of negligence7 and that, if they would not succeed in negligence, they should not succeed at all. Now that the Protection from Harassment Act 1997 had filled the gap in the law exposed by Khorasanjian, the common law had done its bit for justice and could now safely revert to a more widely understood, and thus more certain, course. It is arguable that, without Khorasanjian, there might never have been a Protection from Harassment Act 1997 or, at least, that the Act might not otherwise have been drafted in the way that it was. So most legal positivists are prepared to concede that, from time to time, extra-legal considerations must be taken into account in order to do justice, but that certainty needs to be restored to the legal system as soon as possible thereafter – as has clearly happened in this area through a combination of the Protection from Harassment Act 1997 and Hunter. Indeed, a subsequent ironic twist makes it possible that Hunter itself may now need to be revisited because of the enactment of the Human Rights Act 1998, which requires inter alia that

98 Risk and the Law public authorities (including the courts) show respect for private and family life under Art 8 of the European Convention on Human Rights and Fundamental Freedoms. Arguably loss of television reception, for example, interferes with private and family life; and the Convention does not limit its ambit to those with a proprietary interest in where they live. So it may be that the desire to return to the prior degree of certainty within the law of private nuisance has itself introduced a degree of uncertainty; Khorasanjian may have been right all along.8

Non-positivistic theories But the certainty–justice dilemma need never arise for other schools of thought.9 Natural lawyers and Marxists, for example, simply need to evaluate the outcome of each instance of litigation according to the ultimate principle or goal which they espouse. A particular judicial decision will be correct for natural lawyers if it corresponds with their conception of reason or of the natural law. For Marxists a decision is correct if it provides the best way of promoting the goal sought. Certainty is not really valued by Marxists at all; for natural lawyers it is useful for everyday, run-of-the-mill decisions and for assisting in the promotion of justice, but it cannot become more important than justice. Marxism and natural law theory are, of course, at the other end of the philosophical spectrum from legal positivism. But even Ronald Dworkin, whose theories have a good deal in common with positivism, never needs to confront the issue. By defining ‘principle’ more broadly than is traditional and suggesting that this term should be used to reflect not just explicitly espoused legal doctrines but also a view of law which can be deduced by searching for a general theory which ‘best fits’ a particular area of law, he is able to claim that the law automatically contains within it a requirement that everyone is deserving of ‘Equal Concern and Respect’.10 Thus the claimant in Khorasanjian – as well as potential claimants in analogous situations in future – could be said to be self-evidently entitled to a remedy, so that whether that remedy was provided under the rubric of the common law of private nuisance or through legislation was merely a subsidiary issue of form rather than a primary matter of substance.

Reasoning after the event But even for positivists, the certainty–justice dilemma can arise only once the events in question have actually taken place and are known. This is partly because, until the facts of a case are established, it is impossible to decide the outcome.11 For example, even if the claimant in Khorasanjian had had a proprietary interest in her parents’ home, it would have been impossible to decide whether her former boyfriend’s conduct amounted to a nuisance until it was established just how many telephone calls he had made over what

Risk and predictability in English common law

99

period of time, together with the gist of what he had said. Until this information was known, things would inevitably remain uncertain. Indeed, it would not even have been possible to advise the claimant definitively ex ante how many unwanted telephone calls would establish a nuisance, or over how long or short a period she must put up with them before she would be able to seek redress through the law. And indeed it might be difficult to work out a just solution until after the events have occurred. But the tradition of reasoning ex post facto also occurs because, until a judgment has been interpreted in subsequent cases, it may be very difficult to know precisely what it means. As Postema has said: The judge in a prior case does not unilaterally and finally fix the scope or meaning of a rule through his or her decision, regardless of how carefully crafted the language of the opinion is . . . Law, on this view, is a matter of convention.12 The fact that the certainty–justice dilemma presupposes that we view the matter with hindsight should really come as no surprise. It simply reflects the oft-quoted dictum of Oliver Wendell Holmes that ‘[t]he life of the law has not been logic; it has been but experience’.13 Obviously the notion of experience presupposes that something has already happened. Indeed, the essence of the distinction between the common law and civil law views of the legal enterprise is that, whereas the latter establishes a series of generalised principles in advance which is to be interpreted and made more concrete in individual situations as they arise, the common law simply reacts to events and establishes what the law is by incremental progression from case to case. Thus textbook writers in the common-law tradition take as their primary objective the accurate description of past cases. But just as it is permissible to cast doubt on whether the dominance of the positivist perspective on the common law is justified, so too is there validity in questioning the assumption that cases are best understood only once they have been decided. Indeed, since any textbook on the English law of torts currently opens with some lament about the complex uncertainties of the law, particularly in the law of negligence, and is accompanied by rhetorical questions as to how justice can really be done within such a legal minefield, it is clear that even those who adhere to the traditional approach currently find its limitations somewhat perplexing. It is therefore interesting to consider whether an alternative approach might prove more worthwhile.

RISK The notion of risk is predicated on viewing cases not after they have occurred but in advance. If we know already that something has – or has not – happened, then it ceases to be a risk. In both cases its status is certain,

100 Risk and the Law so there is nothing risky involved. But when considering a matter in advance, then our lack of omniscience means that we are, by definition, unable to be certain about whether or not it will occur. This is when we have to deal with risks. In mathematical terms, the certainty that something has happened equates to 1, whilst the certainty that nothing has happened equates to 0. The risk of something happening must therefore be somewhere between 0 and 1, and is usually rendered as a decimal fraction or as a percentage. Insurance companies, for example, can calculate the percentage chance that a driver of a particular type – typically defined according to age, sex and vehicle – will cause a road accident in which someone else is injured. Accurate calculations of this sort enable the insurance industry to determine the appropriate level of premiums for drivers of that type. Thus if there is a 20 per cent chance of drivers of that type being involved in a motor accident, they will – all other things being equal – be charged a premium which is double that charged to drivers of a type who are predicted to have an accident rate of only 10 per cent. It has become a commonplace observation that the degree of utility of the law of torts in England really depends on the ability of the insurance industry to compensate victims of misfeasance. Indeed, this observation is frequently cited as the major explanation of why, even though the law of torts is fault based, it cannot generally act either as a genuine deterrent from wrongdoing or as a retributive system imposing penalties on those who deserve them, since tortfeasors depend in practice on their insurers to foot the bill. The widespread availability of insurance – indeed, the fact that third-party insurance is compulsory in the very situations where torts are most likely to occur (namely at work and on the roads) – means, it is said, that compensation is the dominant objective of the tort law system in England and Wales. Strangely, the argument appears more or less to come to a halt at this point, observing on the way that compensation can often be achieved more efficiently and at considerably less cost through a state-backed, no-fault insurance scheme such as that in existence in New Zealand.14 But while the common law in England and Wales continues to depend so heavily on the availability at reasonable cost of private third-party insurance policies in order to finance the payments demanded as a result of judicial decisionmaking,15 it would surely be odd if the law itself had not been shaped significantly by the requirements and practices of the insurance industry.

The ‘deep pockets theory’ Hitherto the only genuine attempt to put insurance centre-stage has been work prompted by Atiyah’s seminal Accidents, Compensation and the Law,16 which has led to the regular citation of the so-called ‘deep pockets theory’ as an explanation of certain judgments which are not perhaps easily explained by traditional positivistic reasoning from legal principle. But the

Risk and predictability in English common law

101

‘deep pockets theory’ is itself of questionable validity unless cases are viewed with the benefit of hindsight. It can explain why some cases are decided in favour of an uninsured claimant17 and why others are decided against an insured defendant when other potential defendants are arguably at least as much to blame (if not more)18 or when there appears to be no other means of providing a remedy.19 The theory can also explain the widening of the scope of vicarious liability in Lister v Hesley Hall.20 But it fails to explain why many insured defendants are held not to be liable; nor can it explain why many private claimants with no realistic prospect of being able to insure themselves nevertheless fail to obtain judgment in their favour.21 It also has nothing to say about how a case should be decided when both parties will have had a reasonable opportunity to insure themselves: sometimes the burden is put on the defendant, as in Nettleship v Weston 22 and Condon v Basi,23 and sometimes on the claimant, as in Caldwell v Maguire and Fitzgerald.24 Thus the deep pockets theory can be invoked only once it is known what a court has actually decided. If the court’s decision imposes liability on the party better able to bear it – whether because of the availability of insurance or otherwise – then the deep pockets theory can be cited as the justification (or, at least, explanation) for this result. But if the more impecunious party is left having to bear the burden of whatever loss or damage has accrued, then it can safely be said that the deep pockets theory did not determine the outcome of the case. The best that the deep pockets theory can really offer is that it may explain why a broader range of possible decisions is available to a court than traditional emphasis on legal principle might otherwise suggest. In fact, the deep pockets theory suffers from two fundamental flaws. The first is philosophical. If the common law is really concerned so much with having a given loss borne by the party that can best afford it, it makes no sense for either the law of tort – or significant areas of the law of contract – to insist on the principle that liability must be predicated on fault: strict liability coupled with a system of immunities for certain classes of defendant would achieve this objective much more easily. Indeed, the fact that the deep pockets theory has utility only ex post facto demonstrates that it is an inadequate response to the fundamental role which insurance plays in giving practical meaning and effect to the judgments of the courts in contract and tort. As explained above, because insurers need to be able to predict risk, they have to consider alternative possibilities before they happen rather than afterwards, when it is already too late. The fact that the deep pockets theory has little predictive power but is regularly invoked after cases have actually been decided simply betrays the fact that it originates from positivistic lawyers who are familiar with this perspective. But there is a second, empirical flaw in the deep pockets theory which is just as fundamental. The fact is that it has almost nothing useful to say about the impact of insurance on the common law because it actually implies that insurance is self-defeating. If courts really were motivated in more than the

102

Risk and the Law

occasional case to come to certain decisions on the grounds that this places the burden of the loss on the party which can best afford it, then there would be no point in any private company actually offering insurance for anything other than act of God. For, by the very act of providing insurance, it would make the insured client more likely to be found liable for any loss with which it is connected during the life of the policy, so long as any other party involved did not have at least an equal opportunity to take out insurance. But refusing to provide cover would make potential clients much less likely to be found liable. If this were true, there would be a very small insurance market in the UK and those companies which did offer insurance for man-made risks would be struggling to make a profit. In fact, the converse is the case: new insurance products are constantly being developed, while the insurers which have suffered significant losses in recent years are those which have had substantial exposure to natural events such as storms and floods on the one hand and life assurance on the other. Companies insuring man-made risks have proved far more profitable and have seen their financial stability threatened only by the diminution in value of their portfolios of stocks and shares, a fate shared by all shareholders no matter what the status of the company in which they have a stake.

Risk and predictability If the view that the common law essentially depends on the efficacy of insurance is genuinely to be taken seriously, it is important that a form of analysis considerably more sophisticated than that offered by the deep pockets theory be undertaken to determine what impact insurance has had on the development of the common law. In particular, it is vital that this new analysis be undertaken from the same perspective as that required within the insurance industry, namely before the event. In other words, what is being sought is a theory with significant powers of prediction, so that we can identify and quantify risks with a reasonable degree of confidence. The distinction being drawn here between theories which are employed ex post and those for use ex ante is not the same distinction as that made by Jody Kraus, despite the fact that Kraus uses precisely these labels.25 Both schools of thought identified by Kraus purport to identify the method by which a judge should adjudicate on the case before him or her. He uses the term ex post to describe the effect of a court’s decisions retrospectively on pre-existing rights, whilst ex ante decision-making is concerned with how a judgment might prospectively affect future behaviour. But in the context of the argument here, both these schools of thought employ ex post reasoning since both require that the facts have already been determined and are known to the judge. The prediction of risk involves ex ante reasoning which must take place before the individual concerned has actually done anything. This is certainly well before any litigation has even been contemplated, let

Risk and predictability in English common law

103

alone brought all the way to a court. Kraus’s terminology serves only to highlight the limited horizons of traditional positivist thinking. It should also be emphasised that to focus on the prediction of risk by no means implies a reversion to a positivistic emphasis on certainty. A simple example from the world of insurance should suffice to make this clear. If it is known that an event will occur, causing significant loss to a particular individual, then that person would be well advised to attempt to pass on the loss by means of subrogation. A typical means of doing so would be through insurance. But no insurance company would be prepared to offer cover in such circumstances; if it knows that it will be liable, it knows too that the policy will not be profitable. Moreover, even if the insurer is actually unaware that the event in question will occur (from its point of view transforming the event from a certainty into a risk), the terms on which it will provide the cover will nevertheless exclude causes known to the person insured. Thus insurers do not offer travel insurance, for example, in respect of losses which may accrue as a result of an illness or disability from which the insured person is known to be suffering at the time that the policy is taken out. Yet if it is known with certainty that a particular individual will not be required to pay for certain losses, then although an insurer would no doubt be keen to offer cover in these circumstances – effectively getting money for old rope – there would be absolutely no sense in the potential client seeking insurance for that eventuality (unless compulsory insurance is, for some bizarre reason, required by law).

Objective and subjective risk It is sometimes argued that risks can be classified as either objective or subjective.26 Objective risk is concerned with predicting future events on the basis of records of what has happened in the past. Thus insurance companies keep records which inform them about the frequency of car accidents. Subjective risks, meanwhile, are risks which are perceived by individuals. These may or may not correlate strongly with objective risk assessment. The relationship between individuals’ fear of crime and actual crime rates, for example, is highly problematic, no matter how the latter are measured. It might be thought that insurance is typically concerned with objective risks. After all, insurance companies collect all sorts of data. But this data is not genuinely objective; it simply reflects the history of those people whom that company has insured in the past. The data cannot predict whom the company will insure in the future, even though it might be used as a reason for seeking to promote the company to certain types of individuals and to refuse to offer cover to others. Moreover, the data may well be objectively misleading. For example, for any given year there will be a certain number of road accidents caused by drivers under the age of 21, which will mean in turn that a certain percentage of drivers in that category were responsible for road accidents. But it will be a matter of chance as to whether an individual

104

Risk and the Law

insurer happened to find that the same percentage of drivers under 21 whom it had insured actually made a claim in that year. More likely, some insurers will find that they received fewer claims than the objective data would suggest, whilst other insurers received more claims. Whilst objective data, if it is available, may inform an insurer’s future decisions as to what policies to write and how much to charge, much more significant is their perception of all the relevant sets of data, together with their view of the current state of the law, their own financial position and their likely exposure to claims of entirely different kinds which may need to be cross-subsidised. On the whole, therefore, insurance is not really concerned with objective risk. What matters is the perception of risk by both the potential client and potential insurer as they each try to work out whether cover is required or offered at all and, if so, how extensive it should be and what price they should pay or charge. ‘A hazard has no meaning except in human terms.’27 Insurance is therefore concerned with so-called subjective risk. Thus it is possible to insure against latent damage precisely because its existence is not known at the time when the policy is taken out. It cannot therefore be classed as a certainty but as a risk to the parties involved, even though objectively the damage has already occurred. Indeed, many so-called objective risks are so broadly defined as not really to be risks at all but empirical certainties. There is sufficient data, for instance, to enable us to be certain that someone in the UK will die every day. Similarly, it is known that there will be a road accident, that some aspect of medical treatment will fail, and that one or more trains will arrive late or not at all. Re-casting these scenarios may convert them into genuine objective risks, but this conversion serves only to demonstrate that the very concept of objective risk has little utility. Even if I know that there is a risk that I may be killed or injured in a road accident or as a result of medical treatment, or that I may be late for a law lecture, I can make a meaningful decision as to whether – and how – to adjust my conduct to take account of these risks only once I myself have considered them. If I perceive them as remote, I may decide to ignore them even if, on the basis of evidence unknown to me, that would not appear to be a wise thing to do. For all these reasons a theory of the common law based on the predictability of risk must therefore adopt a subjective notion of risk.

Micro and macro An important effect of shifting the focus of analysis away from the certainty–justice dichotomy and towards the prediction of risk is that the level of analysis is forced to change as well. Concentration on certainty and justice implies an obsession with the minutiae of each individual case, something for which common lawyers are well known (though whether that reputation constitutes fame or notoriety depends on one’s point of view). A focus on prediction of risk, meanwhile, inevitably involves a shift of

Risk and predictability in English common law

105

emphasis towards the bigger picture. Instead of taking the traditional common law view that each case is unique because no set of facts can ever be replicated,28 it demands that each case be seen simply as an instance of a category of risks. Analysis thus shifts from a micro to a macro level. The macro slant means that it is not so important to predict the outcome of each individual case as to foresee how particular areas of law are likely to develop. Implicit in this approach is an acceptance that some predictions about individual cases will inevitably turn out to be wrong. In other words, uncertainty is not the problem – to be avoided wherever possible – which it is for traditional, ‘after the fact’, positivistic analysis; it is naturally endemic to the whole approach and simply reflects the fact that no one is omniscient. Prediction is not knowledge. However, that is not to say that accuracy of prediction of individual cases is unimportant. A theory which enables the outcomes of a larger number of cases to be accurately predicted is obviously normally to be preferred over a competing theory with less predictive power. But considerable care may be needed; as the school of American Realism pointed out many years ago, some cases may simply defy any attempt to predict them in advance. Others may well not have stood if they had been appealed; like everyone else, judges can and do make mistakes. And, of course, the full facts of a case may not be apparent in advance. A predictive theory therefore concentrates on trends and tendencies rather than claiming to provide a fool-proof way of determining the outcome of every single case.

Making new distinctions The emphasis of the prediction of risk approach also means that certain distinctions between different areas of the common law which are currently thought to be fundamental suddenly cease to be meaningful. It is often said, for example, that the tort of negligence is conduct based, whereas the tort of nuisance is fairness based.29 The reasoning behind this drawing of the boundaries between the two torts is that, in negligence, conduct is deemed unreasonable (and thus unlawful) if it fails to meet an objective standard (so that it will be unlawful whatever the context), whereas conduct can be deemed a nuisance only after it has been weighed in the balance against the damage done to the claimant’s rights and found wanting. Thus the identity of the claimant in negligence can have no impact on the standards required of the defendant, since it is the defendant’s own conduct per se which is material. In nuisance, meanwhile, it is impossible to determine the lawfulness or otherwise of the defendant’s conduct until something is known of the nature of the claimant’s own rights and behaviour. It can thus be seen that nuisance is a tort which par excellence traditionally requires positivistic ex post facto explanation. But when viewed ex ante this apparently obvious distinction between the two torts appears a little contrived. Instead, it becomes clear that each branch of the common law is simply seeking to determine the risk of a

106

Risk and the Law

person’s being injured through the inappropriate behaviour of a potential defendant. As Ripstein has put it: Where a risk is inappropriate, the person who imposes it does so at his own risk. That is, should the risk ripen into an injury, the result is the injurer’s to deal with. If a risk is not inappropriate, however, its costs simply lie where they fall; it is one of the risks of ordinary life . . .30 It is important to glean as much information as possible about all the parties involved in order to facilitate a more accurate calculation of the risks involved. Clearly both the conduct of the potential defendant and the status of any potential claimant are important factors to be borne in mind when assessing a risk. From this perspective the emphasis on an assumed distinction between conduct and fairness seems rather misplaced. Indeed, both conduct and fairness are relevant not just to torts but also to the law of contract for, as Zipursky has said: The entitlement [to a remedy] is inferred from what the defendant has and has not done to the plaintiff, and how the plaintiff has been affected, and how the plaintiff has interacted with the defendant.31 When viewing these different common law actions ex ante, the differences between them can be seen to lie not in whether conduct or fairness alone (or a combination of both conduct and fairness) is relevant but in what sort of risks are being considered. This does not mean that contractual risks are of a type agreed upon by the parties whereas risks in tort are restricted to those imposed by the law irrespective of individual agreement. Even legal positivists are prepared to accept that many obligations falling within the ambit of the law of contract are imposed upon the parties involved, whether by statute (as in the Supply of Goods and Services Act 1982) or by the common law itself, as exemplified by the mutual obligations of trust and confidence expected of employer and employee.32 Similarly, it is well known that many instances of liability in the law of torts arise out of agreements.33 Some cases may thus be approached in both contract and tort. Indeed, this is virtually inevitable wherever a contract requires that one or more parties act with reasonable care and skill.34 When viewing a situation ex ante, it is immaterial whether a risk arises as a result of agreement between the parties or by authority of the general law. What matters is simply that the nature and degree of the risk involved can be identified. Ironically, this does not mean that it is irrelevant whether a risk is conceptualised in contract or tort, nor that the divisions between torts are unimportant. The fact that emissions of pollution, for example, are more likely to be covered by the law of nuisance than by other branches of the common law, whereas failures to deliver on time are a quintessentially contractual risk, is not without significance. Indeed, the magnitude of a

Risk and predictability in English common law

107

risk – sometimes its very existence – may be wholly or partially determined by how it is perceived or classified by the law. A risk classified as contractual which does not involve an obligation to act with reasonable care and skill will be of a significantly different order from a risk classified as tortious: the former implies the prospect of strict liability whereas the latter will involve some conception of fault before liability can be imposed. Similarly, whether a defendant’s potential obligations are conceived in negligence or private nuisance will determine who is likely to qualify as a potential claimant and will therefore be important in quantifying a potential defendant’s exposure to risk.35

Three issues in identifying risk In order to be able to predict a risk of legal liability with any degree of reliability, it is actually important to tackle three issues. These are: 1 2 3

The risk of doing (or omitting to do) something which the law deems inappropriate. The risk that such acts or omissions will cause damage or loss. The risk that those suffering such loss will be legally entitled to bring a claim.

It is probably true to say that (2) is an issue of fact which is primarily determined by natural, environmental or ‘organic’ factors36 – such as a potential claimant’s susceptibility to harm or the current state of knowledge within an industry or profession – over which the law has relatively little control. Nevertheless, even this issue of risk is at least partially influenced by the various doctrines of causation and remoteness of damage which operate in contract and tort. Risks (1) and (3) are, in any event, clearly determined by the law. If a matter is conceived as potentially lying within the province of contract law, then (1) is dealt with by the express and implied terms of the contract involved, whereas (3) concerns the question of privity, as modified by the Contracts (Rights of Third Parties) Act 1999. In negligence, (1) is dealt with by the principle of standard of care and (3) by the doctrine of duty. In nuisance (1) brings in the concept of ‘unreasonableness’; (3) leads to an examination of whether or not the claimant has suffered either damage to, or loss of enjoyment of, a proprietary interest (in private nuisance) or ‘special damage’ (for public nuisance). It is therefore clear that each of the three types of risk identified above cannot be accurately predicted without a clear understanding of the law. Moreover, since what is important is that risks should be capable of being predicted with as high a level of accuracy as is humanly possible, it is obvious that trends of judicial decision-making need to be clearly understood by those whom the law may affect (or, at least, their legal advisers) so as to allow these people to make meaningful calculations as to exposure to

108

Risk and the Law

potential risk and thence decide whether, and in what manner, to adjust their own conduct.

Consistency of outcome It may now be objected that the argument has gone full circle and that, having castigated legal positivism for its emphasis on certainty, the idea of focusing on the prediction of risk actually turns out to demand the self-same thing. But that criticism misses the point entirely. Certainty requires that we know what the law actually says. Prediction requires only that we can identify the likely effects of the law. That is why the previous paragraph talked of transparent ‘judicial decision-making’. It does not really matter to the predictive approach what reason is given for a particular judicial decision, still less whether that reason is internally coherent or meshes neatly with other legal principles. What matters is simply the outcome: who won and what remedy did they get? The phraseology used by the judges may be a useful pointer in explaining the decision, but equally it may amount to a form of verbal camouflage, setting up a smokescreen to hide what is really going on (and whether this camouflage is deliberate or – as seems more likely – subconscious, is not really material either). Legal positivists would say that this is precisely their point whenever they comment that a judge has arrived at a particular decision for reasons not of legal principle but of policy (though principle is what he or she may be claiming to espouse in his or her judgment). But a different point is being made here. The predictive approach is not concerned with whether a decision is made for reasons of policy or principle – and is not even necessarily interested in which is permissible or more appropriate. What matters is that outcomes of – and not reasoning employed in – cases be consistent. Thus the positivistic emphasis on principle holds no particular attraction to the predictive approach. So long as outcomes can be predicted with a high degree of accuracy, it is frankly irrelevant whether the explicit or implicit reasons for a particular decision are principle or policy based. Indeed, it should be noted that the concepts of fairness and justice which cause so much of a problem for legal positivism because they open the way to policy-based reasoning are actually implicit to the predictive approach, though they operate at a different level from the actual process of prediction. It is clearly impossible for the law to take a view on what constitutes appropriate or inappropriate conduct, as required for the prediction of risk (1), unless it makes a judgement as to what is fair or just in particular circumstances. But whether the concept of fairness or justice employed is based on legal principle or on policy is unimportant so long as, in the vast majority of cases, it produces an outcome which was predictable ex ante. Moreover, it is irrelevant whether the source of a particular view of what is just is statute or the common law itself, so long as it is clear what the effects (if any) of this view may be.

Risk and predictability in English common law

109

The ‘floodgates’ argument The law of negligence is particularly fertile ground for showing how the predictive approach differs in practice from the ex post facto positivistic emphasis on certainty. As mentioned at the beginning of this chapter, virtually all textbooks on the law of tort currently lament the uncertain state of the law of negligence and particularly the policy-based reasoning which is claimed to underlie so many of the decisions on the existence of a duty of care. This doctrine of duty, as explained above, deals with the third type of risk to be predicted, namely the risk that whoever suffers loss or damage as a result of an inappropriate act or omission will be legally entitled to mount a claim. The leading case in this area is without doubt Caparo Industries v Dickman,37 in which the House of Lords upheld the earlier Privy Council case of Yuen Kun-Yeu v Attorney-General of Hong Kong 38 to the effect that the test of whether a duty exists or not actually involves three stages, namely reasonable foreseeability, proximity and ‘fairness, justice and reasonableness’. Traditional common-law analysis has picked over the various judgments and found them wanting: the concept of proximity is said to be too slippery as to be meaningful and so acts as an implicit invitation for the use of policy arguments, whilst that of ‘fairness, justice and reasonableness’ actually welcomes policy with open arms. Hence, it is said forlornly, the law is chaotically uncertain. But an approach based on predictability of risk takes a very different view of Caparo. It also explains the true status of the notorious ‘floodgates’ argument. The most succinct exposition of the floodgates argument was given by Cardozo CJ, who said that it involved a determination to avoid ‘liability in an indeterminate amount for an indeterminate time to an indeterminate class’.39 What is particularly interesting about this argument in the context of the present chapter is that it is actually a clear example of an ex ante approach to law based on predictability of risk: the problem is always conceived in terms of whether the floodgates will or would open if a particular claim were to be allowed. Indeterminate liability of the sort described by Cardozo CJ is anathema to the prescriptive approach because it makes it impossible to calculate risks and hence to make decisions as to whether certain courses of conduct are worth pursuing. To take insurance companies as an example, it makes their business impossible to conduct because the indeterminate nature of a risk implies that the premium to be charged is equally incapable of being determined, whilst an infinite risk implies an infinite premium.

A brief example of the predictive approach The predictive approach therefore assumes that leading cases must establish some basis on which meaningful predictions of risk can be made. It is not enough that both Yuen Kun-Yeu and Caparo can be seen as allowing the

110

Risk and the Law

defendants to escape liability even if their actions may have contributed to the claimant’s losses. Whilst such a view prevents the risk being infinite, it is still not determinate. Nor would the predictive approach be prepared to adopt the view of some commentators who wrote at the time that these cases suggested that the carrying out of certain public functions (such as the licensing of banks, as in Yuen Kun-Yeu) effectively gave the body responsible an immunity from the law of tort. Immunity suggests zero risk, which the predictive approach treats as inherently implausible; for the insurance industry, zero risk implies zero premium and no profit. Thus Caparo must be seen as establishing liability in certain predictable situations, namely where the risk is clearly limited. By focusing on the outcome of Caparo rather than on the reasoning, it is clear that the auditors are held not to be liable because they do not have a direct relationship with the shareholder company which is seeking to recover compensation for the loss it suffered through relying on the accuracy of the auditors’ report. Whether this means there is no ‘assumption of responsibility’ or whether there is a lack of proximity of relationship – the two alternative formulations given in the judgments – is not really material. Indeed, just because these are the phrases used by the judges does not mean that the true decision could really be subsumed under one of the other two limbs of the duty of care doctrine. After all, it could be argued that a lack of direct relationship implies a lack of reasonable foreseeability or a situation where it would not be ‘fair, just and reasonable’ to hold the auditors liable. It must at once be conceded that, once it is felt that the future implications of a decision can be accurately predicted, then it will inevitably become necessary to explain that view to others. Labels too will then become inevitable. But whether the labels have to stem from the phraseology of the courtroom is a moot point. Textbook writers Cheshire and Fifoot, for example, were not only responsible for the ironic phrase ‘innominate term’ to describe a contractual term that is neither a condition nor a warranty; they also identified a whole concept which explained the outcomes of cases far better than that previously tried by the judiciary.40 The best way of understanding the law on the duty of care for negligently inflicted economic loss, it is argued, is to start from the basis that economic loss has always been the stock-in-trade of contractual claims. The advantage of contract over tort from the perspective of predictability is that there is often a written document in existence, spelling out the relevant terms. But if this model works so well in contract, is it not likely that judges will subconsciously resort to its methodology when dealing with economic loss in negligence? On this view a duty of care will be found to exist where factors are present which equate to the doctrines of agreement, consideration and intention to create legal relations which are found in contract. ‘Intention’ in contract is well known to be a misnomer, since such intention is actually deemed by law to exist wherever the parties are acting in a commercial context unless they specifically indicate to the contrary. This is

Risk and predictability in English common law

111

thus an aspect of contract law which can be transposed into tort without undue difficulty. So far as agreement is concerned, it is impossible actually to make an agreement with someone unless each party can identify the other(s) – or, at least, a class of persons from which the other(s) will be drawn. Thus a duty could not exist in Yuen Kun-Yeu because there was no means by which the licensing authorities could have identified the claimant. Third, consideration means that each party provides something of economic value. It cannot be that this has to be directly replicated in tort because otherwise there would always have to be an actual contract in existence before there could be a potential case of negligence. Instead, what must occur is that either each party provides something of value, but via a third party, as in Smith v Bush,41 or just one party provides the economic value. Finally, the basis of the relationship must be reasonably clear, just like the terms of a contract. Thus the shareholders in Caparo lost their claim because the basis of the relationship was more limited than that on which they have been able to claim a remedy, whilst the basis of the relationship in Law Society v KPMG Peat Marwick 42 was precisely such as to found liability of the defendant towards the claimant. Interestingly, the parallel between contract and tort seems to have occurred to the House of Lords in Smith v Bush, for their Lordships actually went out of their way to make explicit the three contractlike conditions which would have to apply before a buyer could rely on the valuation of a property prepared at the behest of the mortgagee. It is submitted that this approach of recasting cases of negligently inflicted economic loss as though they were cases of contract is capable of explaining all the well-known appellate cases in this area except one. Utilisation of this methodology thus makes the outcomes of future cases highly predictable and easy to factor in to calculations as to whether a risk is worth running or not. In particular, it makes insurance viable. On a technical note, it also suggests that the contentious case of Junior Books v Veitchi 43 was correctly decided. Indeed, it could be argued that Junior Books is essentially the equivalent in negligence of the contract case of Shanklin Pier v Detel Products.44 (Indeed, it may be that founding liability in tort in such circumstances involves less of a fiction than the device of collateral contract employed in Shanklin Pier.) The one case which the predictive approach to risk cannot explain is White v Jones.45 But if the move from an ex post facto positivistic perspective – focusing on judicial formulations of reasoning – towards an ex ante perspective of risk assessment – based on the actual outcomes of decided cases – means that the law of negligence currently appears not to be in chaos but sports just one anomalous case, surely it has something going for it and deserves to be explored further.

Notes 1 Ronald Dworkin prefers to see this as a test of ‘pedigree’ rather than provenance: see Taking Rights Seriously, London: Duckworth, 1977.

112

Risk and the Law

2 See J. Bentham, Of Laws in General, H.L.A. Hart (ed.), London: Athlone Press, 1970, pp. 184–195. 3 See H.L.A. Hart, The Concept of Law, 2nd edn, Oxford: Clarendon Press, 1994. 4 [1993] 3 WLR 476. 5 [1997] 2 All ER 426. 6 See M. Jones, Textbook on Torts, 8th edn, Oxford: Oxford University Press, 2002, p. 334. 7 See also F.H. Newark, ‘The boundaries of nuisance’, Law Quarterly Review, 65, 1949, 480. 8 See McKenna v British Aluminium Ltd, 25 April 2002, The Times (Queen’s Bench Division). 9 Note the use of the word ‘need’. By incorporating some elements of positivism, writers who claim to be in other camps sometimes still end up putting themselves on the horns of the same dilemma. See e.g. J. Finnis, Natural Law and Natural Rights, Oxford: Clarendon Press, 1980; D. Beyleveld and R. Brownsword, Law as a Moral Judgement, Reprint, Sheffield: Sheffield Academic Press, 1994. 10 Dworkin, n. 1 above; R. Dworkin, Law’s Empire, London: Fontana, 1986. 11 ‘Fact scepticism’ is, of course, a well-known school of thought within legal realism. 12 G.J. Postema, ‘Philosophy of the Common Law’, in J. Coleman and S. Shapiro (eds.), The Oxford Handbook of Jurisprudence and Philosophy of Law, Oxford: Oxford University Press, 2002, p. 601. 13 O.W. Holmes, The Common Law, Boston: Little, Brown & Co., 1881, p. 1. 14 I. Campbell, Compensation for Personal Injury in New Zealand: Its Rise and Fall, Auckland: Auckland University Press, 1996. 15 By which I mean paying not just for the damages actually awarded after litigation but also for compensation agreed by the parties involved (including the insurance companies themselves) as a result of incidents which never lead to litigation, or where cases are settled out of court, because of the parties’ understanding of the current legal position. 16 P. Cane, Atiyah’s Accidents, Compensation and the Law, 6th edn, London: Butterworths, 1999. 17 See e.g. McLoughlin v O’Brian [1982] 2 All ER 298 (HL); Smith v Eric S. Bush; Harris v Wyre Forest District Council [1989] 2 WLR 790 (HL). 18 See e.g. Smolden v Whitworth unreported, 17th December 1996 (CA). 19 See e.g. Alcock v Chief Constable of South Yorkshire [1991] 4 All ER 907 (HL); White v Jones [1993] 3 All ER 481 (HL). 20 [2002] 1 AC 215; [2001] 2 All ER 769; [2001] ELR 422 (HL). 21 See e.g. McFarlane v EE Caledonia Ltd [1994] 2 All ER 1 (CA). 22 [1971] 2 QB 691 (CA). 23 [1985] 1 WLR 866 (CA). 24 [2001] EWCA Civ 1054 (CA). 25 See J. Kraus, ‘Philosophy of Contract Law’, in Coleman and Shapiro (eds.), n. 12 above, p. 701. 26 See e.g. F.E. Warner and D. Slater (eds.), The Assessment and Perception of Risk, London: Royal Society, 1981. For reasons best known to himself, Perry prefers to divide risk into ‘objective and epistemic conceptions’: see S.R. Perry, ‘Risk, Harm and Responsibility’, in D.G. Owen (ed.), Philosophical Foundations of Tort Law, Oxford: Clarendon Press, 1997, p. 322. 27 T.R. Lee, ‘The public’s perception of risk and the question of irrationality’, in Warner, n. 26 above, pp. 6–7. 28 It may be argued that the impossibility of replicating facts was taken to the point

Risk and predictability in English common law

29 30 31 32 33 34

35 36 37 38 39 40 41 42 43 44 45

113

of absurdity in The Wagon Mound (No 2) [1967] 1 AC 617, where the Privy Council reached a different conclusion from that which it had reached in The Wagon Mound (No 1) [1961] AC 388, despite the fact that both cases arose out of the same incident. Whether the explanation of Lord Reid, namely that the evidence in the second case was substantially different from that led in the first case, does indeed justify the Privy Council’s reaching different conclusions is perhaps a moot point. See e.g. P. Cane, ‘Justice and justifications for tort liability’, Oxford Journal of Legal Studies, 2, 1982, 30–62; M.A. Jones, n. 6 above, p. 334. A. Ripstein, ‘Philosophy of Tort Law’, in Coleman and Shapiro (eds.), n. 12 above, pp. 662–663. B. Zipursky, ‘Philosophy of Private Law’ in Coleman and Shapiro (eds.), n. 12 above, p. 647. See e.g. Bliss v South East Thames Regional Health Authority [1985] IRLR 308 (CA); Malik v Bank of Credit and Commerce International (in liquidation) [1997] IRLR 462 (HL). See e.g. Smith v Bush; Harris v Wyre Forest District Council [1989] 2 WLR 790 (HL); White v Jones [1993] 3 All ER 481 (HL). Some situations may, of course, also give rise to the prospect of a remedy in both private and public law, but that is an issue beyond the scope of this chapter. Examples include Phelps v Hillingdon [2000] 4 All ER 504 (HL); Tate & Lyle Industries v GLC [1983] 1 All ER 1159 (HL); Gillingham BC v Medway (Chatham) Dock Co Ltd [1992] 3 All ER 923. See e.g. Lippiatt v South Gloucestershire CC [1999] 4 All ER 148 (CA); Holbeck Hall Hotel v Scarborough BC [2000] 2 All ER 705 (CA); Bybrook Barn Centre Ltd v Kent CC [2001] BLR 55 (CA). ‘Organic’ is used here in the sense that Durkheim famously employed as part of his concept of ‘organic solidarity’: E. Durkheim, trans. G. Simpson, The Division of Labour in Society (1893) Gencee, IL: Free Press, 1947. [1990] 1 All ER 568 (HL). [1987] 2 All ER 705 (PC). Ultramares Corporation v Touche (1931) 174 NE 441, 444. See Hong Kong Fir Shipping Co Ltd v Kawasaki Kishen Kaisha [1962] 2 QB 26 (CA). [1989] 2 WLR 790 (HL). [2000] 4 All ER 540 (CA). [1982] 3 WLR 477 (HL). [1951] 2 KB 854. [1993] 3 All ER 481 (HL).

8

Transfer of property and risk of loss in French, English and German law Ulrich Spellenberg

I will not try to explain precisely what should be called generally ‘a risk’. For the following remarks it may suffice to ask who is in the end economically poorer if goods which were the object of a contract of sale or the like perish by accident or force majeure. And I will not ask why the Codes here speak of ‘risk’.

FRENCH LAW

Res perit domino In speaking about loss of goods without anybody’s fault, that is by accident or force majeure, a good starting point seems to be the old Roman saying res perit domino: the loss of goods is for their owner. And indeed if my car or my house perishes by fire or through some other accident, nobody is obliged to indemnify me, except of course an insurer. That is what insurance is for. And, at least in French law, it makes no difference if the owner has voluntarily lodged his goods with a depositee. Art 1929 code civile (c.civ.) states that the depositee is liable only if he is at fault, unless he is late in returning the goods. Art 1302 c.civ. confirms that those who owe specific goods are liable only if at fault. Art 1137 c.civ. says in the same sense that an obligation to give (obligation de donner) includes the obligation to guard the goods, but the debtor has only to observe the diligence of a good dominus or housefather.1 One may conclude that here the creditor carries the risk not to get his due because of the accidental destruction of the object. But title to (or ownership of) goods may change hands and it seems good logic that the loss generally falls on the current owner. Thus, the outcome depends on whether the destruction happens before or after the title to the goods is transferred. We may then say that the risk is linked to the transfer of property and does not, for example, depend on possession. And that is what Art 1138 § 2 c.civ. says: Il [le contrat de vente] rend l’acheteur propriétaire et met la chose à ses

Property transfer and loss in French, English and German Law

115

risques dès l’instant où elle a dû être livrée, encore que la tradition n’en ait point été fait, à moins que le débiteur ne soit en demeure de la livrer; auquel cas la chose reste au risque de ce dernier. (The contract of sale makes the buyer the owner and puts the goods at his risk from the moment when they should have been delivered even if delivery is not made and the price not paid, except where the debtor fails to deliver on time; in which case the risk rests with him.) Article 1138 § 2 applies in all contracts to transfer title to goods, including sale, exchange, donation coupled with an obligation and, it is believed, simple promise of a gift. It is often said that Art 1138 § 2 is an exception to the rule res perit debitori (the loss of a thing is for the debtor)2 which applies in all other contracts where services, works or other performances are promised for consideration. The rule res perit debitori is not stated in the civil code in general terms but is said to be a generalisation of some rules for special contracts. The main instance of this rule is Art 1722 which states that the landlord of a house is discharged of his obligation under the lease if the house is destroyed by accident but also that the lessee no longer has to pay the rent. A parallel rule is Art 1790 relating to contracts for work. The contractor who has undertaken to execute work on materials or goods belonging to the customer is discharged from his obligation if the materials or goods disappear without his fault, but also cannot ask for his remuneration for his work. But contrary to Art 1138 c.civ. all these contracts do not transfer title to goods. Most authors argue that the rule that the debtor who loses the object of the contract is discharged from his obligation3 but can no longer claim the promised consideration is the consequence of the reciprocity in these contracts.4 In contracts for consideration the reason (la cause) of each promise is seen in the promise of the other party. And if one of these promises is extinguished by the other, the consideration loses its cause and is therefore void, too,5 and if the consideration was already paid it is to be restored.6 But since the seller in a contract of sale has by the fact of the formation of this contract already and automatically executed his obligation to transfer ownership, he is no longer a debtor and can, in consequence, no longer bear a risk as a debtor, so there is no longer a reciprocity of obligations. But the debt of the buyer continues.7 This seems to be close to the view of the authors of the c.civ., who most probably simply applied the res perit domino principle instead of res perit creditori or emptori (the seller) because in their eyes the buyer was no longer a creditor after the transfer of the property as far as the ownership is concerned. Of course, the seller may still have possession of the object and he is responsible if it gets lost by his fault. Article 1624 with Art 1302 c.civ. state that his obligation to deliver will expire only if the object is destroyed without his fault and he has the burden of proof.8 Article 1138 § 2 c.civ. has two parts. First, it states that the contract of sale

116

Risk and the Law

itself and alone transfers the title to the goods which are the object of the contract, and second, that from this moment onwards the buyer bears the risk of a loss by accident. The latter seems to be the logical consequence of the former. Article 1138 § 2, stating that the title passes solo consensu by the contract alone, is but an example of the general principle in Art 711 which says that the transfer of property is par l’effet des obligations, by the effect of any obligation to give and deliver. Thus, risk is seen as a consequence of ownership and not of possession of the goods in question. This is confirmed by Arts 1929 and 1137 c.civ. already cited. All this seems good logic if one follows the rule res perit domino, and indeed the courts and in principle also the literature draw these consequences.

Consequences in valid contracts Loss before delivery Delivery in Arts 1604 and 1606 c.civ. is understood as delivery of the goods into the hands of the carrier, not their arrival at the place of the buyer.9 And it seems that there is not even any discussion of the possibility that a loss of the goods by accident even before their delivery to a carrier could not be at the risk of the buyer on the ground that he is already the owner. To decide otherwise would not be consistent with the abovementioned Arts 1136 and 1137 nor with Art 1929. It is argued that the seller is still in possession of the goods only because the buyer did not require them to be delivered or collected them himself.10 Thus, the seller could be regarded as depositee, who, as we have seen, is not responsible for loss by accident. The risk of loss will be differently placed only if delivery has been required by the buyer and the seller has failed to comply (Art 1138 § 2 in fine). It is, of course, necessary that there are goods the ownership of which can be acquired and transferred. Thus, in the case of the sale of unascertained goods, ownership can be transferred only simultaneously with their appropriation to the contract. Article 1585 c.civ. says, if the goods are sold by weight, number or measure, the sale is not effective in the sense that the risk of the goods sold lies on the buyer until they are weighed, counted or measured, and, as it is necessary to add, marked in some way as allocated to the buyer. Until then, the seller bears the risk and has to replace any lost goods. For the sale of real estate, there is a special rule in Arts 1601–3: the buyer acquires the property only on completion of a construction and thus the seller bears the risk that the construction disappears before its completion. This is as generalised and applied to the sale of goods which are to be produced in the future, too.11 Thus, in summary, the buyer bears the risk of a loss of the goods sold from the moment of the formation of the contract even if the goods are still in the hands of the seller, provided that the goods are appropriated to the contract and in existence.

Property transfer and loss in French, English and German Law

117

Carriage risks More important in practice, of course, are the risks of carriage. There is no doubt that they are borne by the buyer as owner (as Art 100 code de commerce expressly states).12 Of course the parties can contract otherwise, and in commerce, especially international commerce, that is very often done. But the general rule applies if there is no stipulation, as may often be the case outside commercial contracts. Additional justification is often seen in the fact that, unless there is an agreement to the contrary, the place of delivery is that of the seller, and therefore it is the duty of the buyer to take the goods away. Of course, the seller is liable for any default on his part in delivery to the carrier. Sale under reservation of title If the risks of accidental loss are linked strictly to ownership, the buyer will not bear them before he becomes the owner of the goods. Normally that is the case at the moment of the formation of the contract of sale, but the rule that the property passes solo consensu is not obligatory. And a stipulation that the title will be acquired by the buyer only when the price is completely paid is of course very common. The clause came into usage early in the last century with the sale of automobiles. But in 1934 the Cour de Cassation gave two leading decisions to the effect that this reservation of title would not be effective against other creditors in a bankruptcy of the buyer. Through this development clauses of retention of title lost the major part of their practical usefulness. This changed in 1980 and 1985 as a result of legislation making these clauses effective in bankruptcy (‘opposable’) and since then they are to be found in most contracts. These clauses can be interpreted either as conditions precedent or as conditions subsequent. I think they are mostly meant by the parties to be suspensive and the consequence is that without the payment of the price the seller is still the owner and consequently bears the risk of the loss of the goods.13 Thus, in the case of the sale of a nickel-plating machine with reservation of title, the machine was delivered to the buyer, installed and started to work properly and then it was destroyed in an accidental fire. The seller claimed the price of around FF1 million but his action was dismissed because he still was the owner, on the basis of res perit domino, there being no special stipulation in the contract as to the bearing of risks.14 At least today commercial contracts often include such a stipulation.15 Most commonly the buyer assumes the risk and takes out insurance, but that had not been done in that case. The reason may have been that the seller was a German enterprise and in German law a special clause would not have been necessary because the buyer bears the risk in any event, as will be seen later.

118

Risk and the Law

Some authors think that this effect of Art 1138 in the case of sale with reservation of title is not just and equitable.16 One author proposes17 to hold that a buyer bears the risk because the seller has performed all his duty and is no longer a debtor in the sense of the general rule res perit debitori. As far as I see, this proposition has not met with approval. It is quite probable that the authors of the code civile meant the vendor to bear the risk as proprietor.18 And even if the views and the will of the authors of the code are not binding, Art 1138 § 2 clearly links the risk and the ownership. But, most important, the rule res perit creditori is to be found in the code only in articles applying to unilateral obligations, e.g. Art 1929, on deposition. It is argued in the present class of cases that there is still reciprocity of obligation, since the property has not yet passed to the buyer.

Void contracts Res perit domino If the buyer assumes the risk because he becomes the owner of the goods by virtue of the formation of a contract of sale, he consequently should not bear the risk if this contract is void or avoided retroactively. The sale being ineffective from the beginning the buyer never was the owner, and if the goods have disappeared without his fault, even while he was in possession, it is the risk of the vendor who has to make restitution of the price received without getting back his goods. Article 1376 c.civ. obliges those who receive what is not their due to make restitution, but by Art 1379, if they received in good faith they are responsible only for loss caused by negligence. Their obligation to restore what they got is therefore largely similar to the obligation of depositees and other debtors of specific obligations who also are liable only if they have been at fault, even if fault may be defined in somewhat different ways in the different instances. At least as in the other cases of specific obligation the burden of proof is on the debtor. Yet if the buyer receives in bad faith he is by Art 1379 liable even in case of force majeure. The chamber of commerce of the Cour de Cassation decided, in a case where a sale of a flock of sheep was annulled because the sheep were probably infected with a certain illness and died after delivery and payment of the price, that the seller was to pay back the full price received without recovering the sheep.19 The decision was approved in the literature, but only with the qualification that it was restricted to the situation in which the cause of the nullity of the contract and of the disappearance of the sheep was identical.20 One may perhaps say that in this case the cause of the disappearance of the object of the contract existed before the formation of the contract and therefore was within the risk carried by the seller who delivered defective goods. In other situations, therefore, the decision could and should be different.21

Property transfer and loss in French, English and German Law

119

But if Art 1138 § 2 c.civ. and res perit domino are seen as an exception to the more general rule of res perit debitori, it seems good logic to apply this special rule of risk also for void or retroactively avoided contracts. But there is astonishingly little case law on the question.22 The literature mostly criticises the mechanical application of the rule res perit domino on void contracts. The arguments are quite different but mostly they propose to allocate the risk of accidental loss of the object of restitution on the buyer. Some older authors would not allow the buyer to demand restitution of the price paid if he could not restore the goods received.23 There are some judgments of the Cour de Cassation in this sense,24 but apparently the goods did not get lost by accident but by voluntary acts of the buyer. Anyway, if the buyer is refused the right to demand restitution he bears the loss or the risk of the thing even if in law he has never – retroactively – been its owner. But today if need be, he will have to restore the value,25 which arrives at the same result. It is also said that one should take into consideration that, in spite of its nullity, the contract was nevertheless executed on both sides, obliging both parties to restore what they got, and that this situation is a kind of inverse reciprocal contract.26 The goods and the price having been exchanged, they should be re-exchanged in a similar way. The purpose of restitution for nullity of a contract is said to be to put both the parties back into the situation existing before the contract, that is restitutio in integrum. Some authors therefore speak of restitution équilibrée (balanced restitution). By arguing in this way, the writers seek to rejoin the other general rule that the debtor bears the risk, i.e. that if the object of the contract perishes without his fault, his obligation will be terminated, but so also will his rights against the other party. It is generally argued that this is a consequence of the ‘notion of cause’, according to which each obligation in a contract for consideration is the reason for the counter-obligation.27 A debtor whose obligation is discharged loses his goods and does not get the price or some other consideration. In a similar way here the buyer cannot claim the price without restoring the goods or their value.28 One may argue that la Cause continues in a reversed way.29 Agostini and Diener argue that one should not distribute the risk of the goods but the risk of the contract,30 thus allowing one to take into consideration the causes of the voidness of the contract,31 the destruction of the thing. Thus, Poisson-Drocourt32 will not allow restitution if the contract is void for immorality and allows minors to demand restitution even if they cannot restore the goods they bought.

Conclusions One may say that all the conclusions drawn by the courts are good logic on the basis of Art 1138 § 2: title passes solo consensu and res perit domino.

120 Risk and the Law But nevertheless there is much to criticise in cases of valid contracts just as in cases where parties are required to restore what they have received in execution of a void reciprocal contract. Indeed, this rather mechanical application of res perit domino is not convincing. If one wants other results one has to part with res perit domino as a general rule. Valid contracts In the case of a valid contract of sale there are no grounds for criticism, it would seem, of the rule that the buyer bears the risks during carriage and the risk of loss if the goods disappear while still in the hands of the seller. This imposition of risk on the buyer may be based not only on his ownership but also on the argument that he had a duty to take the goods away.33 But it is said to be inequitable that the seller bears the risk when the goods are sold with reservation of title. The argument is often put that the seller, even though he is still owner, is no more in possession and control of the goods. The Cour de Cassation also does not take sufficiently into consideration that it was the common intention that the goods should definitely stay with the buyer. The application of res perit domino in this case against the seller is too mechanical an approach. The imposition of risk on the buyer by requiring him to pay the price would be a better solution. But it is difficult to see how this can be done under the code civil, unless by an argument used in German law to the effect that the buyer had already received the essential part of his due when he received the possession and decides alone whether the condition for his title will be accomplished. Void contracts In the case of a void contract the attribution of risk to the seller, as being still the owner of the goods, is criticised from several points of view. One objection could be that the seller not being in possession could not watch over the goods. But to shift the risk to the buyer on the ground that he is in possession would be contrary to the rule that depositees and similar possessors are liable only if at fault. Above all, there is a specific rule in Art 1379 c.civ. that the buyer in good faith if obliged to restore is liable only for his faults, while the buyer in bad faith is liable even in case of force majeure. The problem is a lack of differentiation mainly due to the often regretted lack of a general theory of restitution in French law. The second argument is that the courts should consider that the restitution is a kind of reverse reciprocity and that therefore the buyer should not be entitled to recover his price if he cannot restore the goods he received without restoring their value, e.g. the car he bought and lost in an accident which was not his fault.34 Thus one could apply the rules of mutuality or, let us say, quasi-mutuality to the effect that the buyer bears the risk of loss even without being at fault. One may argue further that Art 1379 c.civ.

Property transfer and loss in French, English and German Law

121

discharging the debtor in case of accidental loss applies only to restitution in unilateral relationships, which could well be what the fathers of the code civil had in mind. But even accepting this starting point, one is tempted to ask whether to make the buyer bear the risk is just, if the nullity of the contract arose from the fault of the seller, or if the reason for the nullity stems from his sphere. An example would be a case where the nullity of the contract arose from a misrepresentation of the seller as to the age of the car, for example. In an interesting way Agostini combines the aspect of mutuality with the general rule of torts.35 The purpose of restitution is, he says, to put the parties in the situation they were in before the contract. Article 1382 c.civ. states that whoever by his fault causes damage to another has a duty to compensate. He puts forward the argument that the subject matter of the contract and restitution only could perish while being with the buyer because of the void contract. In that way he will hold the buyer responsible even for loss of the goods by accident if, but only if, it was his fault that the contract was void. If it was the fault of the seller, then he will get his money back even if he is not able to restore the car. But the writers provide no clear results. A general theory of restitution would be welcome in France to fill gaps. Because of the inadequacies of the doctrine of res perit domino, they nevertheless agree in principle that the risk of accidental loss of the object of a sale and eventually its restitution should not be linked to ownership but should depend on various elements, especially on the responsibility for the nullity of the contract.

SHORT REMARKS ON ENGLISH LAW We may conclude that French law links the risk of loss to actual ownership but is not happy in a significant number of cases where the title has passed as a consequence of contracts transferring property, and especially but not only in situations where a right of restitution arises from void contracts. Indeed, it is quite difficult to integrate risk seen as a proprietary notion into a system of reciprocal contracts because the c.civ. rather clearly did not consider this.

Valid contracts The English Sale of Goods Act s 20 takes the same starting point, perhaps under the influence of French law. In Comptoir D’Achat et de Vente du Boerenbond Belge v Luis de Ridder Limitada (The Julia) 36 Normand, LJ, referring to an older judgment declared the ‘rule res perit domino is generally an unbending rule of English law, arising from the very nature of property’ but ‘it has been found necessary to provide for the passing of the risk to the buyer before the property passes to him if the parties so agree’. And that may be above all ‘if the buyer rather than the seller has an immediate

122

Risk and the Law

interest’ in the subject matter of the contract. In the absence of a special agreement, risk is imposed as in French law, linked to property and therefore on the buyer, so that he bears the risk from the time of the formation of the contract. That concerns, above all, the risk of carriage37 but also the case where the sold goods perish while still in the hands of the vendor.38 Where the goods are sold with reservation of title and the goods perish before payment of the price, the seller still was their owner and therefore should bear the risk provided that there is not the usual agreement to the contrary. But as far as concerns the allocation of risk in the execution of valid sales in English law, it is said that there is much recourse to reliance on implied terms to avoid applying the rules of the Act if they would produce undesired consequences,39 especially to shift the risk to the buyer, for example, from the moment of delivery. I think this is a device rather typical of English law. It would not be forbidden to take this course in French law, too, but I have the impression that judges do not take as much freedom to construe agreements in ways which probably were not really meant by the parties. Thus for example in the case of the machine sold with reservation of title, the Cour de Cassation discussed whether the reservation clause was intended to operate as a resolutive condition which would have allowed the attribution of risk to the buyer, but found that the parties in fact intended it to be a suspensive condition so that the seller still being the owner as such bore the risk.40

Void contracts Implied terms are less helpful with void or rescinded contracts. We arrive there at the law of restitution which seems to me even more differentiated than German law. For reasons of space I will not endeavour to explain in detail. I can make only short indicative remarks with the warning that they are in many respects incomplete. One important point is the rule that restitution has to be possible and therefore, on principle, a buyer may not rescind a contract when restitution in integrum is impossible.41 This rule applies to rescission for mistakes but today is less strict. It suffices that the substantial identity of the subject matter is restored.42 If it has deteriorated, the court may order compensation.43 Thus, if the buyer has totally lost or sold the goods received he cannot any longer rescind the contract, not even for fraud.44 He should be able to make substantial restitution in specie of the property which he has received. Technically there is not attribution of the risk of the object of restitution perishing but refusal of rescission of the contract, but the final effect is that the buyer bears the risk of the accidental loss because he cannot claim his money back because he lost the goods he received. But even if the buyer can get back only the money paid he bears the risk of the accidental loss of the goods he received. If rescission is not available or not claimed, the defrauded party could claim damages in an action for deceit. In this case the problem of risk in our

Property transfer and loss in French, English and German Law

123

sense does not present itself and the question is pushed to the domain of the action for damages. But whether or not the buyer may claim restitution of the money paid depends inter alia on the grounds of the invalidity of the contract. On principle it is excluded for money paid on an illegal or immoral contract.45

GERMAN LAW German law could not fall into the trap, if trap it is, of res perit domino because unlike French or English law the title to the goods does not pass automatically with the formation of the contract of sale. The sale only obliges the seller to transfer the title in order to accomplish his obligation. To do this he has to deliver the goods and to agree with the buyer that he shall acquire the title (§ 929 Bürgerliches Gesetzbuch, Civil Code). Thus title and possession are, in principle, both with the seller or with the buyer. One effect of such a system is that after the formation of the contract of sale there still exists a reciprocity of two obligations, namely of the seller to pass the title and the buyer to pay the price. Since both obligations persist there can be in German law, unlike French law, a supervening event rendering impossible the passing of the property after the conclusion of the contract.46 We may say that res perit domino before the formation of a contract of sale and after the transfer of the title but if the object of the contract gets lost by accident between the formation of the contract and its fulfilment this risk is attributed on the basis of responsibility for non-performance and reciprocity (Synallagma) of the obligation. By paras 275 and 280 BGB the seller is discharged if the transfer of title promised in the sale is impossible and neither owes compensation, and by § 326 BGB the buyer who gets nothing is discharged, too.

Valid contracts Res perit debitori The basic rule is that a debtor is discharged by supervening impossibility of performance if it was not his fault and if performance was not overdue (paras 275, 280, 287 BGB). Because of the mutuality of obligations the creditor is also released and thus the risk is carried by the debtor, i.e. the seller, who loses his goods without his fault and does not get the price because he can no longer perform his obligation. Transport risks The rule being that the title passes only with delivery into the hands of the buyer, it should follow that the seller bears the risk not only if the goods

124 Risk and the Law disappear when they are still with him but also if this happens when they are in transit. Normally the seller would be discharged by the loss if it is not his fault but would also lose his price since he cannot any more pass the title to fulfil the contract. But the BGB intervenes, § 447 shifting the risk to the buyer. The authors of the BGB believed this to be more equitable because transport increases the risk of loss and takes place at the request of the buyer. § 269 BGB provides that, if there is no other stipulation, the place of delivery is to be that of the seller where therefore the buyer has to fetch the goods. Under res perit domino the risk would lie with the seller who in German law stays the owner until the buyer gets the goods. Reservation of title In the case where the buyer is already in possession of the goods before the property passes to him, § 446 BGB states that he bears the risk, that is, that he has to pay the price even if it is no longer possible for him to obtain the title because the thing has perished. The most common and important situation is that of sale with reservation of title. The reason for the rule is rather controversial. To argue that the buyer being in possession could better control and guard the goods is not totally convincing because, by definition, he, too, cannot prevent the loss by accident or force majeure. Therefore some authors explain that the buyer who has possession and the right of usage already holds the essence of what is due to him and therefore should bear the risk even if the rest, to wit the title, is still outstanding.47 Whatever the justification, the law is clear. Even if the seller is still the owner, he does not bear the risk. The inverse situation, that the buyer already has the title and nevertheless leaves the goods with the seller, is possible but of no practical importance.

Void contracts If the risks are distributed as just described, mainly on the basis of a possibility to perform and of reciprocity in contracts for consideration, there are greater problems in the case of a void or rescinded contract when the question arises as to whether what both parties received in performance of this contract is to be restored. It is assumed that they were both not aware of the invalidity. Unlike French law the BGB regulates restitution in detail and there is a detailed, perhaps too diversified, theory of restitution. Unjust enrichment The basic rule is that whoever acquires property which was not due to him has to retransfer it. As stated already, since the contract of sale does not pass the title, normally the issue whether it is void does not affect the transfer of the property by the separate transaction in performance of the contract

Property transfer and loss in French, English and German Law

125

believed valid. Thus the parties have to restore not only possession but also title. If the goods originally received are no longer in the hands of the buyer, he has to restore their value (§ 818 II BGB). Exception for loss of value In the case of a sale, in principle, both have to restore what they got: the buyer has to restore the goods and the title and the seller to return the price received. But paras 818 III, 819 BGB state that the debtor who received in good faith has to restore only as far as he is still enriched irrespective of the reason why it is no longer of benefit. In other words, even if he deliberately destroyed the goods he does not have to restore their value. The justification is that he in good faith believed himself to be their owner and as such he could have done whatever he liked with them. Only if he knew that he had received them without justification (sine causa) will he be responsible and even then only if the goods have perished by his own fault (§ 989 BGB). Thus, even if the debtor is in bad faith the creditor bears the risk of an accidental loss of the goods. Only when the debtor is late with the restitution does he bear the risk of an accidental loss (§ 819 I with paras 990 II, 287 BGB). This regulation is rather favourable to debtors of restitution. It resembles Art 1379 c.civ. in France. Finally it is the creditor who bears the risk of the loss of the things he should get back. Probably the authors of the BGB had in mind only unilateral relations and did not consider the situation when a reciprocal contract performed by both parties was void or retroactively avoided. And indeed one can read paras 812 and ss BGB in the way that both rights to restitution are regulated independently. ‘Theory of balance’ But the Reichsgericht held inequitable the result of an independent application of this rule to reciprocal contracts enabling the buyer to claim the price paid without restoring the value of the goods received.48 The problem is obviously created by an application of the rules of § 818 III BGB concerning loss of value independently on the two obligations. This neglects the fact that the situation came into existence through the performance of a reciprocal, even if void, contract. As a way out the Reichsgericht chose to define enrichment in a new way. It held that only the balance between the price and the value of the destroyed goods should be the unjust enrichment for which an order for restitution might be made. If, for instance, the car that he bought by a void contract for
10,000 perishes in an accident, the buyer may claim only
10,000 minus its real value, let’s say
9,000, i.e.,
1,000. Consequently the buyer bears the risk and the circumstances of the car’s destruction make no difference. It may as well be by the fault of the buyer.49 The aim was to avoid or at least to

126

Risk and the Law

limit § 818 III BGB, so that the buyer, even if he had acted in good faith in the matter affecting the validity of the sale, might not claim his money without restoring the goods or their value, thus bearing the risk of their loss. The effect resembles the English rule on restitutio in integrum. The difference is that the buyer always gets back the profit which the vendor made. This so-called Saldotheorie (theory of balance) is not only difficult to find in the text of the BGB but is also no longer favoured by doctrine. Even the BGH admits that there have to be exceptions, e.g. if the buyer is a minor, the contract was avoided for the fraudulent misrepresentation of the seller, or, as some argue, the delivered goods were already defective.50 The BGH argues now that this rule is a matter of equity and must be applied in a flexible way.51 A prominent author bluntly declares the Saldo-Theory dead and a non-theory because the theory does not allow one to explain its proper exceptions.52 The object of reducing the effect of § 818 III BGB is generally accepted and indeed it seems that the fathers of the BGB at the time of drafting it took into account only unilateral relations. What is criticised is the method of balancing which makes the buyer bear the risk even in situations where this is unjustifiable. Merely defining enrichment does not allow the necessary differentiations according to the reasons for the voidness of the contract or the destruction of the goods. A common case, as several times decided by the BGH, is the sale of secondhand cars which are then destroyed while with the buyer who then claims back the price. Inverse mutuality The BGH has often argued that the restitution has to be mutual because of the reciprocity of the void contracts for consideration, because even if those contracts prove to be void the parties had performed their ‘duties’ reciprocally.53 This recalls the French theory of inverse reciprocity. The effect of the BGH’s theory of balance is that the buyer finally bears the loss of the goods which have perished accidentally54 because their value is deducted from the price he paid. The dominant opinion today in Germany no longer follows the BGH and its theory.55 It is believed to be too simple and mechanical always to set off the value of the goods received and now destroyed. While in some situations it is unfair to the vendor if this is not done and the buyer gets his money back without restoring their value, in other situations it is unfair to the buyer if he has to. Also the BGB allows exceptions to this approach. We cannot define enrichment for all purposes simply as the balance between the price and the value of the object of the – void – contract if we have to accept different results.56 Many authors refer to the similar case of the right of withdrawal from the contract (§ 346 III Nr. 3 BGB). The buyer exercising this right has to restore

Property transfer and loss in French, English and German Law

127

the value of the object received if it was destroyed by his negligence and not if it disappeared by accident. Reasoning from here one does, nevertheless, not apply the rule of § 326 res perit debitori but refers to the special rule of § 346 III Nr. 3 BGB, regulating restitution after rescission or withdrawal of a contract in the case of inter alia delivery of defective goods. This rule is that the buyer may rescind without restoring the value if the goods have perished without his fault. If it was by his fault, he has to restore the value. ‘Fault’ is here understood in a larger and somewhat special sense. The basis of this § 346 BGB is indeed the assumption of a kind of reciprocal relations of restitution. The result of this theory is therefore that the seller bears the risk of accidental loss and not the buyer as by the BGH’s theory of balance. It is argued that the rules for unjust enrichment in the case of void contracts for consideration must not contradict the rules of restitution of reciprocal contracts and thus the causes of the destruction of the subject matter become important. Consideration of reasons for nullity But this rule alone, too, has not proven totally satisfactory. Therefore the possibly dominant doctrine starts again with the – older – notion of the two rights to restitution after a void contract (Zwei-Kondiktionen-Lehre) and tries to handle the problem right where it lies, that is § 818 III BGB which allows a debtor in restitution to plead that he is no longer enriched. § 818 III BGB refuses this privilege only if the party, i.e. the buyer, knows that what she received was not her due, and the discussion is how to confine the theory of balance or whether there are other situations when the buyer should not be allowed to plead the loss of his enrichment that is the loss of the goods received.57 Canaris’ theory takes into consideration not only the causes of the disappearance of the goods but also the reasons for the nullity of the contract and the purpose of the invadidating rules. If, for example, the contract was rescinded for seller fraud, he has to bear the risk; if it was a buyer default, he has to restore the value of the goods; if the reasons are neutral, such as lack of form or consensus or perhaps mutual mistake, it is proposed that the buyer bears the risk.58 Mostly accepted by the BGH, too, is that the buyer need not restore the value of the things received if he was a minor or defrauded by the seller, always provided of course that the buyer did not know of the nullity or avoidability of the contract. But there is much controversy here and it is therefore difficult to determine the current state of the law precisely. Nevertheless one may safely state that the risk of accidental loss of the subject matter is neither in valid nor in invalid contracts attributed on the basis of ownership.

128

Risk and the Law

CONCLUDING REMARKS The old Roman rule of res perit domino is appropriate – and obvious – when there are no obligations to transfer or to deliver. If the obligation is unilateral, German and French law agree that the debtor is discharged by accidental loss if the object perishes by accident59 and one may say that the creditor bears the risk. It does not make a difference whether the creditor is, as usually in French law, the owner60 or whether he has, as in some situations in German law, only the right to get the property. It seems just and fair to treat in the same way those who lose what they have and who lose what is their due. Of course the creditor has the right to damages or compensation if the thing perishes by fault of the debtor or if he was late in delivery. But in French and German law one speaks of risk mainly with regard to contracts to transfer property for consideration. And the situation is different. It is not so much the question that the buyer does not get the promised goods because they do no longer exist but whether he nevertheless has to pay for them. In France he has to pay if the contract was for specific or for ascertained goods. The argument is that the civil code by Art 1138 § 2 c.civ. exceptionally links the risk to the transfer of title by the formation of the sale. In German law, when the goods arrive with the buyer he either acquires the title and now, of course, has to pay even if the goods perish, or, without becoming their owner, he bears the risk by § 446 BGB.61 And what concerns the dangers of carriage, the risk of transport is attributed in both laws to the buyer either in France as being already the owner of the goods or in Germany by the special rule of § 447 BGB. Differences of practical importance are on the one side when the goods sold are still in the hands of the vendor and on the other side when they are with the buyer but sold under reservation of title. In the first situation in French law the buyer nevertheless has to pay and in the second he is discharged because he does not get property which is his due. In German law, on the contrary, in the first situation the buyer is discharged, while he has to pay in the second by virtue of § 446 BGB. It is not easy to decide what is more just and fair. As to reservation of title, however, it is interesting to note that the French doctrine is not happy with the result of res perit domino and strongly recommends contractual stipulations to shift the risk to the buyer. But in French law – and similarily in English Law – the risk is basically linked to property by statute. As long as they are not reformed one has to recommend contractual agreements. More difficult is the situation when a contract was performed by both parties and then proves void or is retroactively rescinded. The – scarce – cases of the Cour de Cassation apply the rule res perit domino and attribute the risk to the seller because he is still the owner. That is not at all evident because Art 1138 § 2 c.civ., which serves as the basis, deals with valid contracts of sale and the dominant opinion in doctrine is against and refers to the fact that the contract nevertheless was performed as a reciprocal

Property transfer and loss in French, English and German Law

129

contract. Since the c.civ. here is less definite one may propose to decide the issue without regard to who has the property. But as the discussion in Germany concerning the BGH’s theory of balance shows, it is also not always just and equitable to oblige the buyer, who, without fault, lost the goods he had received, to restore their value. One has to decide by considering the reasons for the nullity of the contract and also the causes of the loss of the goods. In any case, whether the buyer claiming restitution of the price paid was the owner of the goods is not decisive.

Notes 1 For the meaning and even the existence of this kind of obligation see M. Fabre-Magnan, ‘Le mythe de l’obligation de donner’, Rev. trim. de droit civil (RTDciv), 1996, 85; contra G. Pignarre, ‘A la redécouverte de l’obligation de praestare’, RTDciv, 2001, 41. The problem is that by Art 711 and 1138 s 2 any obligation to transfer a title by itself transfers the ownership and therefore after formation of the contract, it is argued, there is no such obligation left. 2 E.g. F. Terré, P. Simler and Y. Lequette, Droit civil, Les obligations, 8th edn, Paris: Dalloz, 2002, No. 643. 3 As is generally said in Art 1302 c.civ. 4 Terré, Simler and Lequette, n. 2 above, No. 641; J. Carbonnier, Droit civil, vol. 4 Les obligations, 22nd edn, Paris: Presses Universitaires de France, 2000, No. 191. 5 Cabonnier, ibid., No. 59; but there is quite some discussion about this theory without practical consequences. 6 E.g. P. Malaurie and L. Aynès, Droit civil, vol. 6 Les obligations, 9th edn, Paris: Editions Cujas, 1999, No. 761. 7 Terré, Simler and Lequette, n. 2 above, ibid.; contra H. Mazeaud et al., Leçons de droit civil, vol. III, Obligations, 10th edn, Paris: Montchrestien, 2000, No. 1118. 8 For the meaning of fault here see M-L. Morançais-Demeestre, ‘La responsabilité des personnes obligées à restitution’, RTDciv. 1993, 773. 9 Cass. com. 8.10.1998, Bulletin des Arrets de la Chambre Civile de la Cour de Cassation IV, No. 229. 10 Carbonnier, n. 4 above, No. 193; Mazeaud et al., n. 7 above. 11 H. Mazeaud et al., Leçons de droit civil, vol 3/2 Principaux contrats, 6th edn, Paris: Montchrestien, 1984, No. 909. 12 Cass. civ. 19.11.1991 La Semaine Juridique: Juris Classeur Periodique (J.C.P.) 1992 IV 274; Cass. Civ. 10.10.1995 D. 1995 IR 246; Cass. com. 17.5.1983 Bull. IV No. 146; Cass. com. 7.12.1991 Bull. IV No. 466. 13 Under a condition subsequent the buyer will bear the risk as owner if the thing perishes before the condition takes place: B. Starck, H. Roland and L. Boyer, Droit civil, Les obligations 2. Contrat, 6th edn, Paris: Litec, 1998, No. 2020. 14 Cass. Com. 19.10.1982 D. 1983. IR 482, note B. Audit; Cass. civ. 4.7.1995 Gaz. Pal. 1996 pass 88. 15 Terré, Simler and Lequette, n. 2 above, No. 646. 16 Audit n. 14 above; P. Simler, J. Cl. Dr. civil Art 1136–1145, No. 60; D. von Breitenstein, Revue Trimestrielle de Droit Civil 42 1980, No. 5. 17 Von Breitenstein, ibid. 18 Terré, Simler and Lequette, n. 2 above, No. 643. 19 J. Chevallier, Cass. civ. 6.12.1967, RTDciv., 1968, 708 n.

130

Risk and the Law

20 Chevallier, n. 19 above, p. 709; E. Poisson-Drocourt, D. 1983. 86. 21 Expressis verbis Poisson-Drocourt, ibid. 22 Cass. civ. 21.7.1975 D. 1976. 582 note Agostini et Diener is sometimes cited as putting the loss on the buyer in spite of a retroactive nullification by the state. But the decision is very unclear. 23 E.g. M. Planiol and G. Ripert, Traité pratique de droit civil francais. Obligations, 2nd edn, Paris: L.D.G.J., 1954, No. 241. 24 Cass. civ. 7.12. 1928 D.H. 1929. 52; 23.2.1970. 1970. 604. 25 Poisson-Drocourt, n. 20 above, 86 no. 8; Paris 31.10.1973 D. 1974. 583 note P. Malaurie. 26 Carbonnier, n. 4 above, No. 107; Agostini and Diener D. 1976.583; PoissonDrocourt, n. 20 above, 86; Malaurie and Aynès, n. 6 above, No. 586, 762. 27 See n. 4 above. 28 It seems now accepted that the buyer may demand restitution at least by offering restitution in value: M. Malaurie, Les restitutions en droit civil, Paris: Cujas, 1991, p. 139. 29 For a similar notion in German law see below on the ‘Theory of balance’. 30 Agostini and Diener, n. 26 above, ibid. 31 Brun Gaz. Pal. 1974 1. 286, distinguishes between cases where the contract is void for reasons of public (ordre public) and private interest. 32 Poisson-Drocourt, n. 20 above, 86. 33 E.g. Mazeaud et al., n. 7 above, 1118. 34 But he may offer the value of the car: Poisson-Drocourt, n. 20 above, 86; contra M. Sayag, J.C.P.1972. I.2451. 35 Agostini and Diener, n. 26 above, ibid. 36 [1949] AC 293, 319. 37 A.G. Guest (ed.), Benjamin’s Sale of Goods, 6th edn, London: Sweet & Maxwell, 2002, No. 6–002; Pignataro v Gilroy [1919] 1 KB 459; Healey v Howard [1917] 1KB 337; Wardar’s (Import & Export) Ltd. v W. Norwood & Sons Ltd. [1968] 2 QB 663. 38 SGA, s 20: but when the property in them is transferred to the buyer, the goods are at the buyer’s risk whether delivery has been made or not. 39 M.G. Bridge, Sale of Goods, Oxford: Oxford University Press, 1998, p. 114; Benjamin’s Sale of Goods, n. 37 above, No. 6–003. for contractual agreements as to the time for passing of property and risk, see Underwood v Burgh Castle Brick and Cement Syndicate [1922] 1 KB 343; Carlos Federspiel & Co. v Charles Twigg & Co. [1957] 1 Lloyd’s Rep. 240. 40 Audit, n. 14 above; for construction of the contract in this sense von Breitenstein, n. 16 above. 41 Erlanger v New Sombrero Phosphate Co. [1878] 3 App Cas 1218, 1278 et seq; Redgrave v Hurrel (1881) 20 Ch D 1 (all concerning sale of land); A. Burrows, The Law of Restitution, London: Butterworths, 1993, pp. 32 et seq. 42 Spence v Crawford [1939] All ER 271, 288 et seq; Atlati v Kruger (1955) 94 Ch 218, 223. 43 G. Jones, Goff and Jones: the Law of Restitution, 6th edn, London: Sweet & Maxwell, 2002, No. 9–026. 44 Smith New Court Securities Ltd. v Scrimgeour Vickes [1994] 1 WLR 1271, 1282 (Nourse LJ). 45 Shaw v Shaw [1965] 1 WLR 537; J.K. Grodecki, ‘In Pari Delicto Potior est Conditio Defendentis’, Law Quarterly Review 71, 1955, 254. 46 To some extent this resembles a contract to sell in the SGA and its frustration. 47 See J. von Staudinger et al., BGB, Berlin: De Gruyter, 2004, para 446 No. 7, combining both arguments. 48 Entscheidungen des Reichsgerichts in Zivilsachen 54, 137, 142; Entscheidungen

Property transfer and loss in French, English and German Law

49

50 51 52 53 54 55 56 57 58 59 60 61

131

des Bundesgerichtshofes in Zivilsachen (BGHZ) 53, 144; BGHZ 57, 137; BGHZ 72, 252; BGHZ 78, 216. Also the question whether a misrepresentation on their defects contributed to the destruction of the things (mostly cars bought second hand) is perhaps not taken into consideration: W. Lorenz, in J. von Staudinger, Kommentar zum Bürgerlichen Gesetzbuch, Buch 2, Recht der Schuldverhältnisse, Berlin: de Gruyter, 1994, para 818 No. 43. Perhaps only if the defects caused the destruction. BGHZ 53, 144, 147 et seq. C.-W. Canaris, ‘Die Gegenleistungskondiktion’, in B. Pfister and M.R. Will (eds.), Festschrift für W. Lorenz zum 70. Geburtstag, Tübingen: Mohr, 1991, pp. 19–63, 58. E.g. BGHZ 147, 152, 158; BGHZ 78, 246, 229. If he sold them he has of course the price. See e.g. Lorenz, n. 49 above, para 818 No. 41; Canaris, n. 52 above, p. 19 et seq.; contra Martinek, in: J. von Staudinger, BGB: Eckpfeiler des Zivilrechts, München: Sellier/de Gruyter, 2005, p. 781. W. Flume, ‘Die Entreicherungsgefahr und die Gefahrtragung bei Rücktritt und Wandlung’ Neue Juristische Wochenschrift, 1970, 1161. B. Dauner-Lieb et al., Münchener Kommentar zum BGB, 4th edn, München: C.H. Beck, 2004 para 818 No. 101, 123 et seq; especially Canaris, n. 52 above. Lorenz, n. 49 above, para 818 No. 41 p. 310, differentiating Canaris, n. 52 above, p. 37 et seq. Paras 275 I, 280 I BGB. Comp. Art 7111 c.civ: ‘the property is acquired by . . . the effect of obligations’. Case of Sale with retention of title.

9

Transit risks in CIF contracts – meaning and categories Koji Takahashi

THE RISKS INVOLVED IN INTERNATIONAL SALES International sales involve various risks, such as the foreign exchange risk, the risk of failing to obtain export or import licences, the legal risks involved in bringing and enforcing claims abroad and the risk of the contracting partner’s default (the credit risk). This chapter will focus on the transit risks, i.e. the risks arising in the course of the carriage of goods, and will examine them in the context of CIF (cost, insurance and freight) contracts.1 The CIF and C&F (cost and freight, also referred to as CFR) contracts, along with FOB (free on board) contracts, are the most frequently used form of contract for international sale of goods. They are shipment contracts in the sense that the transit risks pass at the port of shipment. They can be contrasted with the arrival contracts, such as DES (delivered ex ship) and DEQ (delivered ex quay), under which the transit risks pass at the port of discharge. The transit risks contemplated are the risk of loss of goods and the risk of damage to goods. What traders and lawyers are less aware of are that there are other categories of transit risks and the usefulness of distinguishing them. The purpose of this chapter is to give a systematic treatment to different categories of transit risks, articulate their meaning and illustrate the usefulness of distinguishing them. Although the transit risks usually contemplated are the risk of loss of goods and the risk of damage to goods, there are in fact other transit risks. This chapter will illustrate the usefulness of distinguishing different categories of transit risks such as the risk of increased cost of carriage, the risk of deterioration, the risk of damage, the risk of loss inflicted by external causes, the risk of loss caused by the inherent nature of goods and the risk of delay. Too often the words ‘loss’, ‘damage’ and ‘deterioration’ are used without defining them. In a majority of cases, no problem arises, but in some cases it is useful to define and distinguish them. This volume, on the theme of ‘risk and the law’, provides a fitting opportunity to give a systematic treatment to the different categories of transit risks and to articulate their meaning.

Transit risks in CIF contracts – meaning and categories

133

Risk of increase in the cost of carriage The transit risks pass as from shipment also in FOB contracts. CIF contracts are distinguished from FOB contracts by the fact that the freight for the main sea carriage is paid by the seller, whereas in FOB contracts it is paid by the buyer. In other words, in CIF the cost of carriage from the seller’s premises to the buyer’s premises is divided at the port of discharge. CIF contracts therefore contain two critical points: the port of shipment where the transit risks pass and the port of discharge where the cost of carriage is divided. Which matters concern the risk and which matters concern the cost of carriage are usually obvious. Thus where an additional freight is charged under the contract of carriage due to a market increase during transit, it is the seller rather than the buyer who is to bear the additional freight since it is a matter concerning the cost of carriage. But what if a casualty affecting the ship – such as strandings, collisions, strikes, government directions – should take place and has caused additional cost to be charged under the contract of carriage? Who – as between the seller and the buyer – is to bear the additional cost? The question can be translated into the language of risk and be phrased: where does the risk of unexpected increase in the cost of carriage pass from the seller to the buyer? Does it pass as from shipment, together with the risk of loss or damage? Or is the increased cost, being part of the cost of carriage, to be divided at the place of discharge? A detailed definition of the CIF contract is to be found in the Incoterms (International Commercial Terms, revised from time to time by the International Chamber of Commerce). The Incoterms are applicable where they are incorporated into the contract. The Incoterms 2000 provides in the pertinent part: B6 Division of costs The buyer must . . . pay . . . – all costs and charges relating to the goods whilst in transit until their arrival at the port of destination, unless such costs and charges were for the seller’s account under the contract of carriage; . . . The ICC Guide to Incoterms 2000 elaborates the point:2 While the seller has to pay all costs required to bring the goods to the port of shipment and to deliver the goods onboard the vessel (as well as unloading charges at the port of discharge, provided they have been included in the freight), the buyer has to pay any further costs which may arise after the seller has delivered the goods onboard the vessel. In this sense, the transfer of the risk also determines the division of costs. If something occurs as a result of contingencies after shipment – such as strandings, collisions, strikes, government directions, hindrances

134

Risk and the Law because of ice or other weather conditions – any additional costs charged by the carrier as a result of these contingencies, or otherwise occurring, will be for the account of the buyer. [Emphasis added]

It follows that where the Incoterms 2000 version of CIF is applicable, the risk of increase in the cost of carriage passes as from shipment. That risk should be recognised as a category of transit risk since it is, by definition, not covered by the risk of loss of or damage to goods and also because it would underscore the point that an unexpectedly increased cost of carriage is to be divided in CIF at the port of shipment, in contrast to the cost of carriage included in the freight, which is to be divided at the port of discharge.

Risk of deterioration Deterioration is distinguishable from damage. Though they both affect the quality of goods, damage is inflicted by external causes such as contamination, whereas deterioration is caused by the inherent nature of the goods, such as discoloration of bananas.3 The risk of deterioration so understood remains with the seller until the port of discharge in English law. The leading authority is Mash & Murrell v Joseph I. Emanuel.4 In this case a quantity of Cyprus potatoes was sold on C&F terms. When they arrived at Liverpool, they were found to be rotten. The buyer claimed that the seller was in breach of implied warranty. Diplock J allowed the buyer’s claim, holding that when goods were sold under a C&F, CIF or FOB contract, the goods must be shipped in such a state that they could endure the normal journey and be in a merchantable condition on arrival. The emphasis is on the condition at the time of shipment, in compliance with the nature of CIF as a shipment contract. This requirement now finds an express statutory basis. Section 14(2) of the Sale of Goods Act 1979 requires the goods sold in the course of a business to be of satisfactory quality and subsection (2B), which was inserted after the Mash & Murrell case, though not because of it, specifically mentions ‘durability’ as one of the aspects of quality to which regard is to be had. The requirement that the goods must be durable at the time of shipment is consistent with the nature of the CIF contract as a shipment contract, since regard is had to the condition of the goods at the time of shipment rather than at the time of discharge. Translated into the language of risk, this requirement means that the risk of deterioration happening during a normal course of voyage remains with the seller until discharge. This is in contrast to the risk of damage, which passes as from shipment. The risk of deterioration should, therefore, be recognised as an anomalous type of transit risk.5 Then, what if the voyage is not normal? In Mash & Murrell, an appeal was made to the Court of Appeal.6 The court did not overrule Diplock J’s ratio but ascertained the facts differently and reached a different conclusion. The court found that the voyage to Liverpool was not a normal voyage since

Transit risks in CIF contracts – meaning and categories

135

the potatoes were not properly ventilated during the voyage and accordingly held that there was no sufficient ground for inferring that potatoes were not fit to travel at the time of shipment. Translated into the language of risk, this decision simply confirms the rule that in CIF contracts the risk of damage inflicted by external causes passes as from shipment, because the incidents such as the lack of ventilation which make the voyage unusual should be regarded as external causes which inflict damage to goods.

OUT-TURN CLAUSE We have now examined the usefulness of recognising the risk of increase in the cost of carriage and the risk of deterioration as categories of transit risks, and of distinguishing them from the risk of loss or damage. We will now consider the usefulness of distinguishing the risk of loss from damage and of splitting the risk of loss by examining the interpretation of out-turn clauses in CIF contracts. In the typical CIF contract the price is calculated on the basis of the quantity agreed in the contract. The seller must ship the goods in the contractual quantity, which is supposed to be recorded in the bill of lading. But the seller does not promise that the goods will arrive at the port of discharge since the CIF contract is a shipment contract. Some CIF contracts, however, contain a clause, sometimes called an ‘out-turn’ clause or a ‘net landed weight’ clause, which calls for the price to be calculated on the basis of the quantity actually discharged. There are several possible interpretations where an out-turn clause is contained in a CIF contract, which will be examined in turn.

Denying effect to the out-turn clause First, an out-turn clause in a CIF contract may not be given effect on the ground that it is repugnant to the nature of the CIF contract as a shipment contract. Thus in Law & Bonar Ltd v British American Tobacco Company Ltd,7 a CIF contract contained a printed clause stating that the goods were at the seller’s risk until discharge. That clause was held inapplicable in the particular transaction entered into in that case on the ground that it was repugnant to the nature of the CIF contract. It is submitted that great caution should be exercised in adopting this interpretation since the CIF buyers may have legitimate interests in inserting an out-turn clause. If the seller ships goods in less than the contractually agreed quantity, bills of lading are supposed to provide the buyer with evidence of the quantity actually shipped. But in practice, printed forms of bills of lading commonly contain the so-called ‘weight and quantity unknown’ clause which destroys their evidential value.8 The CIF buyers therefore often have good reason to wish to pay for only so much of the goods as have actually arrived.

136

Risk and the Law

The decision in Law & Bonar was in fact narrowly interpreted in later cases.9 Similarly, an out-turn clause should be denied effect only where it is part of a printed form added to other negotiated terms and where no alternative interpretations, examined below, can reasonably be adopted.

Denying effect to the CIF/C&F label Where a contract expressed to be a CIF or C&F contract contains an outturn clause and other strong elements proper to the arrival contracts, the CIF/C&F label may be ignored. Then, the contract may be interpreted as being in fact an arrival contract and no difficulty in giving effect to the outturn clause will arise. As the traders are not always legally well informed, they may not choose the most appropriate trade term. Thus in the House of Lords case, The Julia,10 a quantity of rye was sold on CIF terms but contained an out-turn clause, which stipulated as follows: The grain to be weighed . . . Seller and buyer to have the right of supervision both as to weighing and delivery. Any deficiency on bill of lading weight to be paid for by seller and any excess over bill of lading weight to be paid for by buyer at contract price. The ship never reached the port of discharge. The court allowed the buyers to recover the price they had paid, holding that the contract was in fact not a CIF contract but a contract to deliver at the port of discharge. The contract exhibited a strong character of an arrival contract.11 In particular, there was a clause indicating that the risk of damage remained with the seller until the port of discharge, which read as follows: Condition guaranteed on arrival . . . Samples to be taken and sealed at port of discharge jointly by the agents of the shippers and of the holders of the bill of lading or shippers’ delivery order.

Giving effect to both the out-turn clause and the CIF/C&F label We have now seen cases in which effect is denied either to the out-turn clause or the CIF/C&F label. But there is a third way. The contract may be regarded as possessing a mixed character and effect may be given to both the out-turn clause and the CIF/C&F label. The trade terms such as CIF, C&F, FOB, DES and DEQ are merely convenient shorthand in order to save the trouble of negotiation on detailed terms. In the absence of contrary agreement, the meaning of those trade terms is supplied by the governing national law or, where applicable, Incoterms. But their meaning may be modified by special terms agreed to in order to cater for particular parties’ needs. An attempt should therefore be made to give effect to both the special terms and

Transit risks in CIF contracts – meaning and categories

137

the usual meaning of the parties’ chosen trade term. To this end, we will examine below a few ways of giving effect to both an out-turn clause and the CIF/C&F label. Distinguishing the risk of loss from the risk of damage The contract may contain a clause which expressly treats the risk of damage differently from the risk of loss, although it may not necessarily say that it passes as from shipment. So in Produce Brokers New Company (1924) Ltd v Wray, Sanderson & Co Ltd,12 the contract was for sale of oil at a price CIF Hull and contained the following out-turn clause: DELIVERY:

. . . Buyers to ascertain the weight within seven days after discharge under sellers’ superintendence. PAYMENT: Buyers to pay 98 per cent of provisional invoice amount in exchange for shipping documents . . . Balance to be paid or refunded . . . after delivered weight is ascertained. Part of the oil was spilled during discharge at Hull. It was held that the contract was of a mixed character and that the oil remained at the sellers’ risk. The seller was therefore held to have no right to receive or retain payment in respect of the spilled oil. As regards the risk of damage, there was a clause which purported to allocate it by agreement or arbitration. It read as follows: Should the quality and/or condition of the oil on arrival not prove equal to the above guarantees, or should the oil contain seawater or other admixture, this contract is not to be void, but the oil is to be taken with an allowance to be agreed upon or fixed by arbitration, provided always that the oil shipped shall be of the description contracted for. The contract may be silent on the risk of damage. If a CIF contract is found to be of mixed character due to an out-turn clause, the risk of damage will presumably pass as from shipment while the risk of loss will remain with the seller in accordance with the out-turn clause. Thus in The Gabbiano,13 the contract was expressed to be a CIF contract but contained an out-turn clause which provided: If, after loading, any steamer stemmed under this contract is lost, or is, for any reason unable to deliver the cargo or any part thereof, the quantity of ore so undelivered by such steamer shall be written off the contract quantity . . . The court upheld the validity of this clause while recognizing it as being inappropriate to a CIF contract proper. But it held that the contract remained,

138

Risk and the Law

as expressed to be, a CIF contract with variations. The report mentions no provision on the risk of damage but the court held that if the circumstances concerning the risk of loss never arose, the contract would be performed according to its tenor as an ordinary CIF contract. This would mean that if the contract contained no clause on the risk of damage, that risk passed as from shipment. Splitting the risk of loss Where a CIF or C&F contract contains an out-turn clause but the rest of the clauses exhibit a strong character proper to the CIF and C&F contracts, the clause may be given a restrictive interpretation by limiting its effect to only a certain type of risk of loss. That would enable a certain risk of loss to remain with the seller until discharge in accordance with the out-turn clause while the rest of the risk of loss to pass as from shipment in accordance with the strong CIF and C&F character. Limiting the effect of the out-turn clause to partial loss In Soon Hua Seng Co Ltd v Glencore Grain Ltd,14 the entire cargo of rice sold was lost when the ship ran aground. The contract was headed ‘C&F full outturn weight’ and contained a clause saying ‘Price: cost and freight linerterms Rotterdam in bulk . . . full outturn weight at port of destination’. But the court found that the contract had the hallmarks of a true C&F contract in contrast to the contract in Produce Brokers, examined above. The court observed that the contract in the latter case was full of reference to delivery at the port of discharge, containing obligations regarding discharging, pumping, strikes and quality and condition. By contrast, the contract in the instant case provided inter alia: 4. The Rice to be at Buyers’ risk from warehouse to warehouse . . . 5. Sellers to deliver the Rice overside and Buyers to take the Rice. . . . 6. . . . Remarks: Quality, condition fumigation final at time of shipment as per certificate independent surveyors . . . In view of the strong elements proper to the CIF contracts, the court, excluding the operation of the out-turn clause in the case of the total loss, held that if the goods were lost and did not arrive at all, the risk of loss was on the buyers and no question of any adjustment to the payment could arise. An exclusion of an out-turn clause in the case of total loss may be expressly provided in CIF and C&F contracts. Thus clause 117 of the CIFFO (CIF Free Out) contract of the Sugar Association of London reads: In case of total loss of a consignment where the contract provides for settlement on landed weight . . ., the Buyer shall be invoiced at the

Transit risks in CIF contracts – meaning and categories

139

contract price for the shipped weight of that consignment plus five per cent plus the premium for bags when applicable. Giving effect to an out-turn clause in the case of the total loss would result in the total release of the buyer from his obligation to pay. This result might be thought too blatantly contradictory to the nature of the CIF and C&F contracts, especially where the contract contains strong elements proper to the shipment contracts. In such cases, the exclusion of the operation of an out-turn clause in the case of the total loss may be defensible. Then, partial loss cases are left to be covered by the out-turn clause. But should all of them be covered? If the application of an out-turn clause were to depend upon the simple distinction between the total and partial loss, wholly arbitrary results would be produced. Thus if 99 per cent of the goods were lost, the buyer would have to pay only for 1 per cent in accordance with the outturn clause, whereas if 100 per cent of the goods were lost, the buyer would have to pay for 100 per cent. It is submitted that a better distinction on which to rest the application of an out-turn clause is the one between the loss inflicted by external causes and the loss caused by the inherent nature of goods. Limiting the effect of the out-turn clause to the loss caused by the inherent nature of the goods The loss of goods is sometimes inflicted by external causes such as leakage and spillage. In other cases, loss is caused by the inherent nature of the goods such as evaporation and shrinkage.15 If the effect of an out-turn clause is limited to the loss caused by the inherent nature of the goods, the total loss is almost automatically excluded since the inherent nature of goods is unlikely to result in the total loss. A closer reading of the decision of Seng v Glencore Grain suggests that the court’s intention was in fact to limit the out-turn clause to the loss caused by the inherent nature of the goods. The court held that the contract distinguished between the risk of accidents, to be covered by insurance, and out-turn weight differences arising in other ordinary circumstances. The court also agreed with the view expressed in Schmitthoff’s Export Trade,16 which in turn refers to section 2–321 of the United States Uniform Commercial Code. That section reads: Under a contract containing a term C.I.F. or C. & F. (1) Where the price is based on or is to be adjusted according to ‘net landed weights’, ‘delivered weights’, ‘out turn’ quantity or quality or the like, unless otherwise agreed the seller must reasonably estimate the price. The payment due on tender of the documents called for by the contract is the amount so estimated, but after final

140

Risk and the Law adjustment of the price a settlement must be made with commercial promptness. (2) An agreement described in subsection (1) or any warranty of quality or condition of the goods on arrival places upon the seller the risk of ordinary deterioration, shrinkage and the like in transportation but has no effect on the place or time of identification to the contract for sale or delivery or on the passing of the risk of loss. (Emphasis added)

The official comment for this section says that those provisions provide for a shift to the seller of the risk of quality and weight deterioration during shipment without changing the legal consequences of the CIF or C&F term as to the passing of marine risks to the buyer at the point of shipment. Although the terminology used is different from the one adopted in this chapter, those provisions seem to limit the effect of out-turn clauses to deterioration and the loss caused by the inherent nature of the goods. In English law, deterioration does not need to be covered by the out-turn clause since the risk of deterioration remains with the seller at any event in accordance with Mash & Murrell, examined above. The buyers in Seng Co. v Glencore Grain pointed out the difficulty of proving the cause of loss. But the court did not see this as a problem, holding as follows: simple loss of weight appears unlikely to result from an accident at sea affecting a rice cargo, unless one postulates examples of obvious accidents affecting the integrity of the vessel or requiring the jettisoning of cargo. Increase in weight, due to additional moisture, is a more likely result of an accident; but in that event the extra moisture would probably be readily discernible. The proper allocation of the burden of proof seems to be for the buyer to prove the quantity actually discharged and then for the seller to prove that any loss resulted from external causes.

ARRIVAL-TIME CLAUSE In the typical CIF and C&F contracts, the passage of risk of loss at the port of shipment means that the sellers do not promise that the goods will arrive at the port of discharge, let alone by a certain date. However, CIF and C&F contracts sometimes contain a clause referring to the arrival time.

Transit risks in CIF contracts – meaning and categories

141

Interpreting the clause as relating to shipment One way to make sense of such a clause is to interpret it as requiring the seller to make or procure such shipment as would normally arrive at the port of discharge by the specified time. If the clause is interpreted as relating to shipment, it complies with the nature of the shipment contract. Thus in The Wise,17 a C&F sale of motor spirit provided inter alia: 9. Delivery: Vessel TBN [i.e. to be nominated]. Arrival March 15–30 1986. 12. Title and risk: Passes at vessel’s manifold flange at loadport. 14. Vessel: vessel nominated to be subject to buyers’ acceptance. The sellers nominated the Wise and the buyers accepted it. But the vessel was then hit by a missile and the cargo never arrived. The court held that the contract did not wear the air of a contract designed to procure the guarantee of delivery within the period stipulated. On that basis, it was held that the sellers’ obligation was to nominate such vessel as would in the ordinary course of events arrive at the port of discharge within the stipulated period. The Wise was found to fulfil that requirement because, but for the missile attack, she would have arrived at the port of discharge by 30 March 1986.

Distinguishing the risk of delay from the risk of loss or damage A clause referring to the arrival time cannot be interpreted as relating to shipment if, unlike the contract in The Wise, the contract itself relates to the goods already on board a particular vessel which has already sailed. Thus in The Jambur,18 a contract for the sale of gas oil contained a clause which read: DELIVERY:

Latest by 30th April 1990, CIF basis one safe berth/port Kaohsiung, Taiwan, as full cargo per mt ‘JAMBUR’, which sailed from Constanza 10:00 am 28th March 1990.

A collision and arrest made it impossible for the Jambur to reach Kaohsiung by 30 April. The buyers pointed out that the delivery clause in the present case could not, unlike the clause in The Wise, sensibly relate to the choice of the vessel since the contract itself had chosen the Jambur, which had loaded the cargo and had sailed. It was common ground that, apart from the delivery clause, the contract had all the features of a classic CIF contract. It provided inter alia: PRICE: US Dollars 184.50 per metric ton CIF basis . . . on B/L weights. INSPECTION: If possible, mutually agreeable independent inspectors to

appointed by seller at loadport for quantity/quality ascertainment.

be

142

Risk and the Law

Notwithstanding the strong CIF character, the court did not accept the sellers’ submission that the delivery clause should be disregarded. By interpreting the delivery clause as expressing the seller’s promise to deliver at the port of discharge by 30 April, the court held that the buyers were entitled to treat the contract as repudiated on 10 April for the sellers’ anticipatory breach. It has been pointed out by some commentators19 that in the sale of oil, the time of the arrival can be crucial to the buyer, who has to make advance arrangements for the provision of terminal facilities needed for the reception of oil. It was also held in The Orient Prince20 that the dates of delivery of oil products in Rotterdam were matters of great importance, because oil traders base their profit on the margin between buying and selling prices, and Rotterdam had historically been the basis for reckoning prices since large quantities of oil were delivered there for consumption or storage. In that case a CIF sale of naphtha contained a clause saying ‘Delivery Feb 15/March 15 Basis Rotterdam’. The court interpreted this as meaning that the seller must deliver the cargo in Rotterdam by midnight on 15 March if Rotterdam was nominated as the discharging port, or alternatively within a reasonable time after 15 March if another port was nominated. In the sale of other goods, too, the arrival time may be important. Thus in Cargill International SA v Bangladesh Sugar & Food Industries,21 a C&F sale of sugar contained a clause unequivocally promising arrival by a specified date. It read: Special clause (i) The arrival period/time is the essence of the contract. Therefore the seller shall strictly adhere to the arrival period/time stipulated in this contract. If the seller fails to do so, the buyer shall be entitled . . . to terminate the contract and . . . forfeit the performance bond. As the vessel arrived late, the buyers rejected the shipment and made a call on the bond. It was held that the buyers were entitled to make the call for the sellers’ breach of the arrival-time clause, though, on a point irrelevant to the theme of the present chapter, they were held entitled to retain only an amount equal to the loss actually suffered. These cases show a great tendency of the courts to give effect to arrivaltime clauses in CIF and C&F contracts by interpreting them as meaning that the risk of delay remains with the seller until the port of discharge. In the archetypal CIF and C&F contracts, the sellers do not promise that the goods will arrive at the port of discharge by a certain date and, therefore, the risk of delay passes as from shipment together with the risk of loss or damage. But in the cases where a CIF or C&F contract contains an arrival-time clause, it may be useful to distinguish the risk of delay from the risk of loss or damage as this enables the risk of delay to remain with the seller until discharge in accordance with the arrival-time clause and the risk of loss or damage to pass as from shipment in accordance with the CIF/C&F label.

Transit risks in CIF contracts – meaning and categories

143

Thus the contract in The Jambur, seen above, indicated that the risk of loss and damage passed as from shipment by the clauses which provided that the payment of the price was to be on the bill of lading weights and that an inspection of the goods was to be carried out at the port of shipment for quantity and quality ascertainment. The court seems to have endorsed (although not unequivocally) the effect of those clauses by holding that ‘[t]he Buyers carried the risk of contamination of the cargo, the risk of leakage between the point of discharge and the shoretank and, possibly, the risk of a degree of loss in transit’. If the risk of delay remains with the seller until discharge and the risk of loss passes as from shipment, the consequences are that, if the goods are lost in transit, the seller can still claim the price, but if the goods are only delayed the seller cannot claim the price. The seller may then be motivated, when the vessel is likely to arrive late, to cause the goods to be lost. But if he does so, he would be in breach of his obligation not to prevent the goods from arriving. Although the CIF and C&F sellers do not have an obligation to deliver the goods at the port of discharge, he is still under a negative obligation not to interfere with the carriage so as to prevent the buyer from receiving the goods.22

SUMMARY This chapter has examined different categories of transit risks and sought to illustrate the usefulness of recognising and distinguishing between them. The main points made were as follows: 1

2

3

The risk of increased cost of carriage caused by a casualty affecting the ship – such as strandings, collisions, strikes, government directions – is a distinct category of transit risk since it is by definition not covered by the risk of loss of or damage to goods and also because this would underscore the point that an unexpectedly increased cost of carriage is to be divided at the port of shipment. This is in contrast to the cost of carriage included in the freight, which is to be divided at the port of discharge. Deterioration is distinguishable from damage. Though they both affect the quality of goods, damage is inflicted by external causes such as contamination, whereas deterioration is caused by the inherent nature of the goods, such as discoloration of bananas. The risk of deterioration so understood remains with the seller until the port of discharge under CIF and C&F contracts. This is in contrast with the risk of damage, which passes as from shipment. The risk of deterioration should, therefore, be recognised as an anomalous type of transit risk. Loss is distinguishable from damage as it affects the quantity (weight or volume), as opposed to quality, of goods. In the typical CIF and C&F contracts there is little merit in distinguishing the risk of loss from the

144

4

5

Risk and the Law risk of damage since they both pass at the port of shipment. But where a contract expressed to be CIF or C&F contains an out-turn clause (a clause calling for the price to be calculated on the basis of the quantity actually discharged), the distinction may be useful as it enables the risk of loss to remain with the seller until discharge in accordance with the out-turn clause while the risk of damage passes as from shipment in accordance with the CIF/C&F label. Where a contract expressed to be CIF or C&F contains an out-turn clause but the rest of the clauses exhibit a strong CIF or C&F character, it may be useful to split the risk of loss. The out-turn clause may then be given a restrictive interpretation by limiting its effect to a certain type of risk of loss. That would enable the certain risk of loss to remain with the seller until discharge in accordance with the out-turn clause and the rest of the risk of loss to pass as from shipment in accordance with the strong CIF or C&F character. For this purpose the distinction between the total loss and partial loss is arbitrary. A better distinction is the one between the loss inflicted by external causes such as leakage and spillage and the loss caused by the inherent nature of goods such as evaporation and shrinkage. In the typical CIF and C&F contracts, the passage of risk of loss at the port of shipment means that the seller does not promise that the goods will arrive at the port of discharge, let alone by a certain date. The risk of delay, therefore, passes as from shipment together with the risk of loss or damage. But where a contract expressed to be CIF or C&F contains a clause calling for arrival of goods by a certain date (an arrivaltime clause), it may be useful to distinguish the risk of delay from the risk of loss or damage as this enables the risk of delay to remain with the seller until discharge in accordance with the arrival-time clause while the risk of loss or damage passes as from shipment in accordance with the CIF or C&F label.

Notes 1 C&F contracts will be treated as identical to CIF contracts for the purpose of this chapter since the only difference between them, i.e. the existence or absence of an obligation on the seller to conclude an insurance contract, is not material for the purpose of the discussion here. 2 The comment to B6 of CFR, to which reference is made from the comment for CIF. 3 Other commentators may prefer other expressions. For example, Guenther Treitel, in A.G. Guest (gen. ed.), Benjamin’s Sale of Goods, 6th edn, London: Sweet & Maxwell, 2002, para 18–222, uses the words ‘risk of necessary deterioration’ to refer to the deterioration which any goods of the contract description necessarily suffer in the course of contemplated transit and the words ‘risk of extraordinary deterioration’ to refer to deterioration due to some accident or casualty. 4 [1961] 1 Lloyd’s Rep 46.

Transit risks in CIF contracts – meaning and categories

145

5 Recognition is given to the risk of deterioration by the Sale of Goods Act 1979 in the case where, unlike in CIF contracts, the seller agrees to deliver specific goods at his own risk. In that situation, the buyer nevertheless takes ‘any risk of deterioration in the goods necessarily incident to the course of transit’ (sn 33) unless otherwise agreed. 6 [1961] 2 Lloyd’s Rep 326 (CA). 7 [1916] 2 KB 605. 8 See The Atlas [1996] 1 Lloyd’s Rep 642; The Mata K [1998] 2 Lloyd’s Rep 614. 9 The Julia [1949] AC 293; The Jambur, 14 November 1990, unreported (transcript available on Lexis). 10 [1949] AC 293. 11 It contained, for example, a clause allowing a delivery order to be substituted for a bill of lading and a certificate of insurance for a policy and a clause stating ‘[a]ll average to be for seller’s account’. 12 (1931) 39 Ll L Rep 257. 13 [1940] P 166. 14 [1996] 1 Lloyd’s Rep 398. 15 In J.J. Lightburn and G.M. Nienaber, ‘Out-turn clauses in c.i.f. contracts in the oil trade’ Lloyd’s Maritime and Commerce Law Quarterly, 1987, 177, the authors state that the distinction between the transportation loss (due to unavoidable causes) and the marine loss (which is fortuitous and extraordinary) is clearly recognised in the oil trade. It is interesting to note that they include spillage as an example of transportation loss, while acknowledging that there may be no spillage if perfectly new equipment is used. 16 C.L. Schmitthoff, Schmitthoff’s Export Trade: The Law and Practice of International Trade, 9th edn, London: Sweet & Maxwell, 1990, p. 54. 17 [1989] 1 Lloyd’s Rep 96 (reversed by the Court of Appeal on other grounds [1989] 2 Lloyd’s Rep 451). 18 14 November 1990, unreported (transcript available on Lexis). 19 E.g. G. Treitel, Chapter 19 in A.G. Guest (gen. ed.), Benjamin’s Sale of Goods, 6th edn, London: Sweet & Maxwell, 2002, para 19–071. 20 [1985] 1 Lloyd’s Rep 389. 21 [1998] 1 WLR 461. 22 See e.g. Peter Cremer v Brinkers Grondstoffen NV [1980] 2 Lloyd’s Rep 605; The Playa Larga [1983] 2 Lloyd’s Rep 171; The Rio Sun [1985] 1 Lloyd’s Rep 351.

10 Costs and risk: recent developments in the English law of costs Keith Uff

Many readers of this chapter will be aware that in April 1999 the Supreme Court1 and the county courts of England and Wales adopted new procedural rules as the result of recommendations made by Lord Woolf of Barnes in his report to the Lord Chancellor entitled ‘Access to Justice’.2 The new Civil Procedure Rules 1998 (CPR)3 replace the former Rules of the Supreme Court (RSC), which applied in the Court of Appeal and the High Court, and the County Court Rules (CCR), which applied in the county courts.4 Lord Woolf’s original intention was to replace all the former rules simultaneously, but this proved to be impractical and some of the former rules were kept in force (and in some cases remain in force) for the time being by Schedule 1 to the new rules. Such of the old rules as remain in force are being progressively replaced by new rules as quickly as the Rule Committee can draft them and the courts’ bureaucratic machinery can be adapted to accommodate them.5 The rules relating to costs were in the second batch of new rules to be adopted and are now fully in force in all the courts. In this chapter I shall be concentrating on the provisions in the CPR, but it will be necessary occasionally to refer to the old rules. In such cases I shall refer only to the RSC, unless there was some significant difference between the RSC and the CCR. Lord Woolf’s reforms also effect some important, and often rather confusing, changes in the technical vocabulary of English procedural law (claimant for plaintiff; disclosure for discovery, etc.). Where it becomes necessary to draw attention to such cases I shall do so thus: claimant [plaintiff]. One remarkable feature (remarkable to English lawyers, that is) of Lord Woolf’s reforms is the way in which the CPR set out an explicit policy objective – the ‘Overriding Objective’. CPR Part 1 states that the Rules are a new procedural code with the overriding objective of enabling the court to deal with cases justly,6 a concept which explicitly includes saving expense and dealing with the case in a way which is ‘proportionate’. ‘Proportionality’ means that the case must be dealt with in ways which are proportionate to the amount of money involved, the importance of the case, the complexity of the issues and the financial position of the parties.7 In this chapter I shall begin by giving an account of the system for

Costs and risk: recent developments in the English law of costs

147

awarding the costs of litigation in outline and then explain some important recent changes, which do not form part of the Woolf reforms, in the funding of litigation. I shall then analyse in rather more detail the changes which Lord Woolf introduced into the costs system as part of his reforms and conclude by considering whether these changes are likely to prove effective in promoting the objectives which Lord Woolf set out as informing the reform process.8 The theme of risk, which is the subject matter of our colloquium, is never far away when the subject of costs is mentioned. The risk of having to pay costs, whether one’s own or a fortiori those of the other party, is a very significant factor both in making the decision to litigate and in the way in which the litigation itself is conducted. The question which is raised in this chapter is whether it is possible to promote desirable procedural objectives by a radical adjustment in the incidence of this risk.

THE ENGLISH COSTS SYSTEM IN OUTLINE

The power to award costs The following comments are confined to costs between parties to ordinary adversary litigation (known as party and party costs). ‘Costs’ for this purpose means all the costs of litigation including court fees, out-of-pocket payments by a party’s solicitor (disbursements) and fees payable to the lawyers (barristers and solicitors) engaged in the case. As CPR r 43.2 puts it: ‘Costs’ includes fees, charges, disbursements, expenses, remuneration, reimbursement allowed to a litigant under rule 48.6, any additional liability incurred under a funding arrangement and any fee or reward charged by a lay representative for acting on behalf of a party in proceedings allocated to the small claims track. It should be mentioned at this early stage that in English practice, unlike that in Germany, there is normally no fixed relationship between the costs recoverable from another party and the amount in issue in the litigation, although, as previously mentioned, some of the rules tend indirectly to promote the idea that the amount of costs recoverable should be ‘proportionate’ to the amount in issue. In cases allocated to the Small Claims Track (those in which the claim is for £5,000 or less), no costs are normally recoverable from the losing party, subject to some exceptions. The power to award costs is conferred by the Supreme Court Act 1981, s 51 (as amended), which provides as follows: (1) Subject to the provisions of this or any other enactment and to rules of court, the costs of and incidental to all proceedings in

148

Risk and the Law (a) the civil division of the Court of Appeal; (b) the High Court, and (c) any county court, shall be in the discretion of the court. (2) Without prejudice to any general power to make rules of court, such rules may make provision for regulating matters relating to the costs of those proceedings including, in particular, prescribing scales of costs to be paid to legal or other representatives. . . . (3) The court shall have full power to determine by whom and to what extent the costs are to be paid.

It will be noted that the court is given a very wide discretionary power in relation to the award of costs, but this power is made subject to such rules as the Rule Committee may make concerning the exercise of the discretion. In general, costs can be awarded only against someone who is a party to the litigation, but it was held by the House of Lords in Aiden Shipping Co. Ltd v Interbulk Ltd 9 that the power conferred by s 51 can, in an appropriate case, be exercised against a person who is not a party to the proceedings (in that case, a person for whose benefit the proceedings had been brought but who was not a party on the record).

Orders for costs In order to recover costs a party must (with very few exceptions) get an order from the court, hence the litigator’s maxim: ‘If you don’t ask, you don’t get.’ The court will usually be asked by the winning party to make an order at the conclusion of the trial of the action and the parties’ counsel will address the judge on the terms of the order which he is being invited to make. The judge is not obliged to make an order for costs and in some circumstances may decide to make no order. There is a special technical vocabulary used in relation to orders for costs. In a simple case where one party is the outright winner of the case at trial the court will make an order for claimant’s [plaintiff’s] costs or defendant’s costs as the case may be. In pre-trial hearings there is a wide range of orders which can be made, such as the following: Costs in cause – the costs of the hearing will be payable in accordance with the order for costs made at the trial (this is the order usually made in pre-trial hearings, unless there is some reason for making a different order). Claimant’s (or defendant’s) costs in cause – the claimant or defendant as the case may be will not have to pay the other party’s costs of the hearing and may get his costs if he is successful at the trial.

Costs and risk: recent developments in the English law of costs

149

Costs in any event – the party receiving the benefit of this order gets the costs of the hearing irrespective of the outcome at trial and can enforce the order immediately. Costs thrown away – costs wasted as a result of some procedural error committed by a party. That party will have to pay those costs to the other party irrespective of the final outcome of the case. Costs reserved – the judge conducting the pre-trial hearing leaves it up to the trial judge to decide which party should pay the costs of the pretrial hearing. By use of these orders the court can reward procedural virtue and penalise procedural misconduct. In extreme cases the court has the power to order a party’s lawyers to pay the costs personally.10 The order for costs contains two elements: (a) it states which party is to pay the costs, and (b) it states the basis of assessment [taxation]. The basis of assessment is the test which the official who assesses the costs has to apply to determine whether any particular item is allowable or not. There are two bases of assessment, the standard basis and the indemnity basis. In an assessment on the standard basis the court will: (a) only allow costs which are proportionate to the matters in issue; and (b) resolve any doubt which it may have as to whether costs were reasonably incurred or reasonable and proportionate in amount in favour of the paying party.11 In an assessment on the indemnity basis the court: will resolve any doubt which it may have as to whether costs were reasonably incurred or were reasonable in amount in favour of the receiving party. The indemnity basis is therefore more generous to the receiving party. In ordinary adversary litigation the court will normally order assessment on the standard basis, but may make an order for assessment on the indemnity basis if, for example, it considers that a party has misconducted himself in the conduct of the litigation. It may make such an order in respect of the costs of part of the litigation only,12 but will not do so merely because a party has been unsuccessful.13 ‘The decisions of the Court of Appeal in Raja v Rubin14 and Baron v Lovell15 show that the court had been concerned with some part of the paying party’s conduct of the litigation which merited the disapproval of the court. The usual order on the standard basis should be made unless there is some element which deserves some mark of disapproval. It is not just to penalise a party for running litigation which it has lost.’16 It is reasonably clear from the number of reported cases on this issue

150

Risk and the Law

since 1999 that parties are making more applications for indemnity costs under the new rules than was the case under the old ones, but there is no empirical way of demonstrating whether they are more likely to be successful. Orders for costs are enforced using the usual mechanisms for the enforcement of judgment.

Exercising the discretion to award costs The basic principle on which costs are awarded is that costs are in the discretion of the court, but normally ‘the costs follow the event’, i.e. the winner of the litigation is entitled to receive his costs from the loser. This is sometimes referred to as the indemnity rule of costs17 and is shared by many legal systems, the most significant exception being the United States where each party has normally to bear their own costs.18 The CPR state this principle as follows: If the court decides to make an order about costs – (a) the general rule is that the unsuccessful party will be ordered to pay the costs of the successful party; but (b) the court may make a different order.19 Under the old rules this principle was interpreted in a way which was very favourable to a party who had achieved any degree of success in the litigation, so that it was rare for a party to be deprived of his costs if he was the overall winner of the litigation, even if he had fought the case on several grounds and been successful only because of a favourable outcome in one of them. Although the new rule preserves the general rule that the unsuccessful party will be ordered to pay the costs of the successful party, Lord Woolf was anxious to move away from this position and he therefore envisaged far more partial (or ‘split’) orders for costs which more accurately reflect the level of success achieved by the receiving party.20 This has resulted in many more partial orders for costs being made in which the party who is successful overall is deprived of his costs on a particular issue where he has been unsuccessful on that issue, or even may be ordered to pay the costs of the unsuccessful party on that issue. It is not necessary to show, as it usually was under the old rules, that the successful party has acted unreasonably or improperly.21 It has been said22 that in determining how to apportion costs between the losing litigant and his partially successful opponent a number of factors have to be considered: (1) the successful party’s reasonableness in pursuing the issue on which he was defeated; (2) his conduct in relation to that issue and the litigation generally; (3) the extra costs incurred by running that issue, in terms of preparation and court time; and (4) the fairness of allowing the successful party some of his costs.

Costs and risk: recent developments in the English law of costs

151

This changed policy has the attraction of discouraging parties from advancing weak arguments in the hope that, if they are successful overall, they will not have to meet the cost of doing so, but it has the significant disadvantage that the winner of the litigation may be deprived of part, or even the whole, of the fruits of the litigation which are rightly his by virtue of costs which are irrecoverable or payable to the other side. The risk element in costs is increased, because a party cannot be sure that if he is successful overall he will be able to recover all the costs of his success from the other party. The court then has two dissatisfied customers instead of the usual one.

Assessment (taxation) of costs When the court has made an order for costs it is necessary, unless the parties can agree an appropriate figure or the case is one where there are fixed costs,23 for there to be an assessment (formerly known as a taxation24). If the matter is large or complicated the court will order a detailed assessment. This involves the winning party drawing up a bill of costs in a set form (there are very complicated rules about how this has to be done)25 and having it considered at a hearing by a specialised corps of judicial personnel who are referred to collectively as ‘costs officers’. These are costs judges [taxing masters] in the High Court in London, and District Judges in the county court and in the High Court outside London. Some of their powers may be delegated to court officials of non-judicial rank. At the assessment the costs officer will go through the bill, line by line if necessary, and decide whether particular items are allowable and if so for what amount, applying the basis of taxation (see above) ordered by the court. The assessment is done by reference to a complicated mixture of amounts fixed by the CPR in respect of some matters (a rate per item)26 and hourly rates of charge in respect of the work of the solicitors.27 A feature of Lord Woolf’s reforms is the encouragement of the use of summary assessment of costs, that is, an assessment carried out by the trial judge himself when he makes the order for costs at the end of the trial. In order to facilitate this process, the CPR require the parties to provide the court with estimates of their costs to date at intervals during the pre-trial procedure and for a party claiming costs to provide certain further information before the summary assessment takes place. Although summary assessment of costs by the trial judge was a long-standing feature of procedure in the county court (which traditionally dealt only with the smaller cases), Lord Woolf hoped to transpose it into the High Court and thereby save the costs of formal assessment. In fact, this does not so far seem to have been a successful aspect of his reforms. There has been resistance to summary assessment in all but very straightforward cases and High Court judges, who are generally unfamiliar with the minutiae of costs calculations, are reluctant to engage in the process.

152

Risk and the Law

RECENT CHANGES IN THE FUNDING OF CIVIL LITIGATION28 At about the same time as Lord Woolf was conducting his enquiry into civil justice which resulted in ‘Access to Justice’, important changes were occurring in the funding of civil litigation which have had a profound impact on the system as a whole, perhaps exceeding in their effects those of the Woolf reforms. Before the Courts and Legal Services Act 1990 the funding of civil litigation in England could be summarised as follows. Litigants could fund their cases in one of three ways. First, they could pay the costs out of their own resources. This was the only method available to corporate litigants who did not have liability insurance in respect of the claim in question. In practice, of course, insurance companies were the principal source of funding of costs on the defendant’s side in many cases, especially claims for damages for personal injury,29 and also funded the claimant’s side of the case where the insurer had indemnified his insured against a loss and was exercising his right of subrogation to sue in the name of the insured. Second, a litigant might be ‘assisted’ in meeting the costs of litigation. Trade unions, for example, often support their members in funding actions for personal injury arising out of industrial accidents. There has also been a considerable growth in recent years in private legal expenses insurance. This is often sold by insurance companies as an optional ‘add-on’ to other types of policy, such as motor or house contents policies. It has been held that an assisting party is entitled to the benefit of the costs order made in the case if the assisted party is successful, but is liable for the other party’s costs if it is unsuccessful.30 Third, an impecunious litigant might receive legal aid under the terms of the Legal Aid Act 1988. The civil legal aid system was exceptionally favourable to litigants. It was available in almost all types of civil litigation and was state funded, subject to a contribution from the legally aided party.31 This system, however, became too expensive for the state to fund and in the case of claims for money has now been almost entirely replaced by the Conditional Fee Agreement (CFA), or ‘no win, no fee’ agreement. Under this type of arrangement a claimant has to find a solicitor who is willing to act for him on the basis that if the case is lost the client will pay nothing, but if the claim is successful the lawyer gets the assessed costs in the normal way and a ‘success fee’ agreed between the lawyer and the client. The client is allowed to add the success fee to the costs which he can recover from the losing party. If the claim is unsuccessful the claimant has to pay the costs of the other party (because the indemnity principle has not been abrogated). It has therefore become usual for claimants who have entered into CFAs to take out After The Event (ATE) insurance to insure themselves against this risk. The success fee and the premium for the ATE are recoverable as part of the costs if the claim is successful.32 The use of CFAs, which may be regarded as a watered-down version of the American contingency fee system, has

Costs and risk: recent developments in the English law of costs

153

proved extremely controversial and given rise to many practical problems. Some people think that adopting the American system would be a better way forward.33

THE WOOLF REFORMS AND COSTS Lord Woolf’s reforms have made radical changes in the English civil justice system, the most obvious of which is the emphasis placed by the CPR upon case management – the idea that the court, not the parties, should control the pace at which the litigation is conducted and should take an active part in deciding the procedural steps to be taken in order to bring the case to trial. This idea (which might seem to lawyers educated in the civilian tradition to be too obvious to mention) was highly controversial when first proposed by Lord Woolf and there were many prophets of doom who predicted that case management – at least if operated by the English judiciary – would lead to disaster. Although it is too early to come to any firm conclusion, the evidence so far suggests that the prophets of doom were wrong and the casemanagement aspects of the Woolf reforms seem to have worked rather well. A less remarked aspect of the Woolf reforms was some important changes in the rules concerning costs. Lord Woolf, who was presented with evidence that in many cases the costs exceeded the amount recovered, was concerned to reduce the costs of litigation by encouraging earlier settlement (compromise) of cases. It had long been a feature of the English system that many cases did not settle until a very late stage in the pre-trial [interlocutory] process when most of the costs had already been incurred. Settlements literally at the door of the court were very common, particularly in claims for damages for personal injury. It followed that substantial savings in the costs of litigation could be achieved if the parties could be induced to reach a settlement at an earlier stage of the litigation. Lord Woolf was convinced that late settlement was the result of the parties having inadequate information about the case in its early stages and that therefore early settlement could be encouraged by requiring a much earlier exchange of information between the parties so that they are better informed about the strengths and weaknesses of their position. The CPR give effect to this policy, as explained below. Another technique which is adopted is adjusting some of the rules concerning costs so as to put the risk of paying costs on to a party who is failing to accept an offer to settle on terms which the court (retrospectively) considers to have been reasonable.

‘Front-end loading’ of costs The CPR contain a number of rules designed to encourage an early exchange of relevant information between the parties. In many types of litigation there is a Pre-Action Protocol which sets out a list of information which the

154

Risk and the Law

parties are expected to exchange, even before the action has been started. When the action has been started there are further provisions which encourage early disclosure, such as a rule permitting a party to annex witness statements to his statement of case [pleading]. Although these rules have the desirable objective of encouraging early settlement, they have the disadvantage of requiring the parties to incur many of the costs of the litigation at an earlier stage of the action than was formerly the case, so-called ‘front-end loading’ of costs. It is unclear at the moment whether this has had the effect of encouraging earlier settlement, as Lord Woolf hoped, although there is clear evidence that fewer actions are being started, which may be because more are being settled.

Part 36 payments and offers It has long been a feature of the English procedural system that a defendant in a claim for debt or damages can make a payment into court of a sum which he is prepared to pay to settle the action. The practice was formerly governed by RSC Ord. 22. If the plaintiff decided to accept the sum paid in (he had three weeks in which to decide whether to do so), he received that sum, plus the taxed costs up to the date of the payment in. If he decided not to accept the sum paid in and the action proceeded to trial, the effect of the payment on costs depended upon the amount which the plaintiff recovered. If he recovered more than the amount paid in, he got a judgment for the amount awarded by the court, plus his taxed costs in the usual way. But if he recovered no more than the amount paid in, he (the plaintiff) had to pay the costs (including the defendant’s costs) after the date of the payment in. This rule therefore had a very vicious sting in its tail – a plaintiff who failed to take out a payment in and recovered less at trial would often see the benefit of whatever sum the court had awarded immediately lost by having to pay costs to the defendant. This principle was a source of much anxiety to plaintiffs, particularly in personal injury cases. Defendants’ advisers devoted a great deal of time and effort to deciding whether to make a payment in and if so when and how much to pay. In this way the risk of having to pay the costs is used as a powerful weapon to encourage settlement. The CPR replace the old system of payment into court with ‘Part 36 payments’ and ‘Part 36 offers’. The significant difference between the old rules and the new ones is that CPR Part 36 can be used by claimants as well as defendants. The previous practice in relation to payments in by defendants is continued, but a claimant can now make an offer to the defendant indicating what sum he would be prepared to accept in order to settle the action. If the claimant recovers more than his Part 36 offer at trial he receives interest at a penal rate on the costs to which he is entitled under the normal rule. If a defendant makes a Part 36 payment, the costs consequences depend, as before, upon whether the claimant accepts it or not. If he accepts the offer the case is settled for that amount plus the claimant’s costs up to the

Costs and risk: recent developments in the English law of costs

155

date of the offer. If he rejects it and recovers less than the amount offered at trial he will normally be ordered to pay the defendant’s costs after the date of the offer.34 It is too soon to say whether this system has had the effect of encouraging defendants to make realistic offers in settlement at an earlier stage than before, although the number of reported cases about technical problems arising under the new rules suggests that claimants are making extensive use of the new powers. Where the claim is not for a money judgment but for some other kind of order, such as an injunction, there is a procedure by which a party can make an offer indicating the terms of an order of the court to which he would be willing to consent. This procedure was formerly known as a ‘Calderbank letter’,35 which was sent on the basis that it was ‘without prejudice save as to costs’, i.e. it could not be mentioned to the court until it was considering its order for costs. The new practice is contained in CPR r. 38, rr. 1, 2 and 19. If the claimant has obtained an order on terms less favourable than the terms offered in the letter, the expectation is that the court will award the defendant his costs after the date of the letter.

‘Proportionality’ It was noted above that one of the criteria set out in the CPR for achieving the Overriding Objective is that of ‘proportionality.’ This concept is intended to convey the idea that the parties should not be free to expend whatever they wish on the costs of the litigation irrespective of the amount in issue, but that the steps which they take in the litigation and their expenditure upon them should be in proportion to the size and importance of the case. Thus the court, exercising its case management powers, may forbid a party to take a step in the pre-trial procedure which he would otherwise be entitled to take, or may disallow the costs of a step taken by a party either on the grounds that the step itself, or the party’s expenditure on it, was not proportionate. This principle, while not producing a rigid connection between the amount in issue and the costs recoverable, does have the effect of discouraging parties from expending excessive amounts.36

CONCLUSION The risk of having to pay the costs of litigation is very significant in civil litigation in England. Lord Woolf’s reforms have made some attempt to bring the cost more into line with the size and significance of the case, although the early signs are that they have not been successful in this respect.37 The attempt to use the risk of paying costs as a weapon to encourage early settlement of the case is controversial and it is impossible to say whether it will prove to be a success in the long run. So far, the sharp fall in the number

156

Risk and the Law

of cases started is the only clear statistical evidence of the effects of the new rules. Whether this represents early settlement of cases, as Lord Woolf hopes, or a defeatist attitude on the part of potential litigants towards the risk of paying costs, it is impossible to say.

Notes 1 The Supreme Court is the collective name used in the legislation for the High Court and the Court of Appeal. It does not exist as a separate court. The House of Lords is the highest court in the English legal system and thus the nearest to the idea of a supreme court as understood in other common law jurisdictions. 2 London: HMSO, 1996. 3 The power to make the new rules is conferred by the Civil Procedure Act 1997, s 1, and the body which makes them, the Civil Procedure Rule Committee, is created by s 2. The former Supreme Court Rule Committee and the County Court Rule Committee are abolished. 4 The use of the lower-case ‘c’ in ‘county courts’ is not an error. This is the way it is printed in the County Courts Act 1984. 5 There have already been over 30 statutory instruments amending the original CPR. Some of these amend the original rules, which contained some drafting errors, and others have added new Parts, replacing old rules in the RSC and CCR. Lord Justice May et al. (eds.), Civil Procedure 2002, London: Sweet & Maxwell, 2002, contains all the rules and relevant statutes with massively detailed, semi-official commentary. It is usually referred to as ‘The White Book’ from the colour of its binding. 6 CPR r. 1.1(1). 7 See below, The Woolf reforms and costs. 8 The Lord Chancellor’s Department is conducting a study of the effects of the Woolf Reforms. The latest publication is Further Findings, A continuing evaluation of the Civil Justice Reforms, Lord Chancellor’s Department (August 2002). The Civil Justice Council and The Law Society have also commissioned some research in this area. See T. Goriely, R. Moorhead and P. Abrams, More Civil Justice? The Impact of the Woolf Reforms on Pre-action Behaviour, London: The Law Society, 2002. These studies tend to show that costs of civil litigation have, if anything, risen as a result of the Woolf Reforms. 9 [1986] AC 965, HL 10 This is known as a wasted costs order. The power to make such an order is derived from the Supreme Court Act 1981, s 51(6). This power has proved extremely troublesome to the courts. The practice is regulated by CPR r. 48.7. 11 CPR r. 44.4(2). 12 In Clark v Associated Newspapers Ltd. (The Times, 28 January 1998), where cross-examination of a claimant took the form of a totally uncalled-for personal attack, Lightman J made an order for indemnity costs in favour of the claimant for that portion of the trial. 13 Raja v Rubin (The Times, April 14, 1999), where the Court of Appeal held that a judge was wrong to award indemnity costs against a claimant who was held to have no sufficient interest to make an application to the court under the Insolvency Act 1986. 14 Ibid. 15 The Times, September 14, 1999. 16 Civil Procedure 2002 44.4.2.

Costs and risk: recent developments in the English law of costs

157

17 The use of the word ‘indemnity’ here should not be confused with the indemnity basis of taxation referred to above. 18 Formal accounts of civil procedure in the United States may give the impression that they too have the indemnity principle, but it should be understood that in the US ‘costs’ usually means only court fees and not lawyers’ fees or disbursements. 19 CPR r. 44. 3(2). 20 See, A.E.I. Rediffusion Music Ltd. v Phonographic Performance Ltd. [1999] 1 WLR 1507. Civil Procedure 2002, 44.3.1 discusses the relevant cases. 21 See Re Elgindata (No. 2) [1992] 1 WLR 1207. This pre-Woolf decision of the Court of Appeal marked a change in practice in this matter. 22 Antonelli v Allen, The Times, 8 December 2000, per Neuberger J. 23 Fixed costs are awarded in straightforward debt-collecting cases and are at a rate fixed by statute. 24 Lat. taxatio, a limit. 25 Many solicitors use a specialised professional group called Law Costs Draftsmen to draw their bills for them. Costs draftsmen usually charge as their fees a fixed percentage of the bill as assessed. 26 In some types of litigation, such as actions for damages for personal injuries, there are agreements between representatives of ‘major players’ (claimants’ solicitors and insurance companies in the case of personal injury litigation) as to how much is to be allowed for particular items. 27 Barristers’ fees are a disbursement because they are paid by the solicitor to counsel, not directly by the client. The solicitor should be reimbursed by his client. 28 M. Zander, ‘Where are we heading with the funding of civil litigation?’, Civil Justice Quarterly, 22, 2003, 23 is a critical survey of recent developments. 29 Such cases constitute a very large proportion of cases which go all the way to trial. 30 Bourne v Colodense [1985] ICR 291 (CA). It should be noted that an ‘assisting’ party is not a party on the record in the case and the court is formally unaware of its existence, but in practice judges are well aware of the likelihood that a party is being assisted. The assister’s right to take the benefit of any order for costs arises from an express or implied contractual right to be indemnified by the party represented. The liability to pay the costs of an unsuccessful assisted party would arise under the Aiden Shipping Co. Ltd. v Interbulk Ltd. [1986] AC 965 (HL) principle noted above, but in practice few problems arise in respect of this. 31 The size of the contribution (if any) varied with the legally aided party’s financial means. 32 Callery v Gray [2002] UKHL 28; [2002] 1 WLR 2000. Critical Note by M. Zander, ‘Where are we now on conditional fees? – or why this emperor is wearing few, if any, clothes’, Modern Law Review, 65, 2002, 919. The effect of the Callery v Gray decision has been confirmed by the Access to Justice Act 1999, s 29. 33 Zander, n. 28 above, pp. 35–40. 34 The rules require that the existence of a payment by the defendant be kept from the trial judge because he might otherwise be influenced in the amount of damages which he awards, see CPR r. 36.19. 35 From the case Calderbank v Calderbank [1976] Fam 93. 36 Rather oddly, proportionality is not mentioned explicitly in CPR Part 44, which contains the general rules about cost assessment, but Lord Woolf himself has said that ‘the considerations which should be taken into account when making an order for costs are redolent of proportionality’ (Lownds v The Home Office (unreported), 2002). EWCA Civ 365, (cited in Zander, n. 28 above, p. 32). 37 See the studies cited in n. 8 above.

Employment and social security

11 The risk of sickness in German labour and social insurance law Dietmar Boerner

INTRODUCTION In all the stages of an employer–employee relationship problems will emerge whenever an employee gets sick. This applies to the conclusion of a contract of employment, to its performance and to the termination thereof. However, my reasoning is intended to centre on the question of how the employee’s income will be affected by sickness. Who, in this respect, is bound to take the sickness risk? The employer, the social insurance fund or the employee himself? The actual importance of this question is revealed by taking a look at statistics. It is true that sickness figures in Germany have fallen slightly in recent years and at the time of writing have reached a satisfactory 4 per cent of employed persons.1 Nevertheless the costs of income maintenance are at a high level. Consequently, employers had to spend a total of almost
27 billion in 2001 for continued remuneration in case of sickness. In addition, the sickness benefit payment to be defrayed by the health insurance scheme amounted to almost another
8 billion.2 INCOME RISK OF EMPLOYED PERSONS To what extent does an employee’s sickness affect his income? Two steps must be taken to answer this question. First of all it is necessary to find out how the Civil Code distributes the risk of loss from sickness. Afterwards it must be ascertained to what extent the distribution of this risk under civil law is modified by labour and social security legislation.

Distribution of risk according to the Civil Code According to the systematic arrangement of the Civil Code two problem areas must be distinguished in the event of non-performance due to an employee’s sickness. In the first place it is necessary to clarify to what extent the sickness will affect the employer’s claim for performance. In the second place a problem is posed by the question as to whether an employed person

162

Risk and the Law

is entitled to claim the continued payment of wages in spite of work not rendered. Employer’s claim for performance The future development of the claim for performance to be raised by the employer has been laid down in the Civil Code, Art 275, subsection 1. This provides that such claim shall be excluded insofar as it is impossible for the obligor or for anyone else to render the work required. If an employee is unable to work due to sickness, performance will usually become impossible. This must be attributed to the fact that the performance is normally to be rendered at a certain period of time set by contractual arrangement or by the employer; consequently, it has the character of an absolute fixed-date obligation. The failure of an employee to render his work at the fixed time will not only entail a delay in performance but already result in impossibility of performance.3 The legal consequence ensuing from this impossibility has been laid down in the Civil Code, Art 275, subsection 1, to the effect that the employer loses his claim for performance; hence, the employee is discharged from his job liability. According to this provision the employer bears the performance risk. Employee’s claim for remuneration The consequences resulting from the impossibility of performing the work with respect to the employee’s claim for remuneration are implied, in principle, in the Civil Code, Art 326, subsection 1, sentence 1. It follows from this article that the claim to remuneration becomes void insofar as the employee is released from his obligation to perform due to impossibility. Consequently, the employee bears the remuneration risk. Thus the principle of ‘no work, no wages’ has taken shape in Art 326, subsection 1, sentence 1, Civil Code, as a result of the mutuality of the contract.4 Final result When considered in the light of the Civil Code the following distribution of risk is operative in the event of sickness of an employee. The employer has to shoulder the performance risk whereas the remuneration risk is borne by the employed person.

Distribution of risk according to labour and social insurance law A modification of the distribution of the sickness risk as provided for in the Civil Code has been brought about by labour and social insurance law in favour of the employee. It is true that none of these special provisions

The risk of sickness in German labour and social insurance law

163

changes the fact that the employer forfeits his claim for work performance, but – under certain conditions – the employee can now assert a claim for remuneration or for a compensation payment. So the employee is normally entitled to continued payment from the employer during the first six weeks of his incapacity for work. Subsequently, the employee has the right, for the most part, to draw sick pay from the statutory health insurance. If the incapacity for work was caused by an accident at work the sick pay is to be replaced by an injury benefit or a pension, to be drawn from the statutory accident insurance fund, and calculated by reference to the amount of the pension to which the injured person will be entitled.5 Insofar as the sickness results in a permanent impairment of earning capacity the employee may ultimately be entitled to a pension from the statutory pension scheme.6 The benefits to be drawn from the statutory accident insurance and those arising from the statutory pension scheme exhibit to a large extent distinct features. For this reason I shall not describe these in a more detailed way, since it would otherwise become necessary to go far beyond the scope of this chapter. Instead, I now look more closely at the prerequisites of a continued payment of the remuneration and the sick pay. I would like to begin with an exposition of the provisions for continued payment of remuneration. Continued payment of remuneration by the employer Pursuant to Art 3, subsection 1, sentence 1, of the Law on the Continued Payment of Remuneration, the employee has the right to a continued payment of remuneration by the employer if the employee is prevented from performing his work by incapacity due to sickness arising otherwise than through his own fault. The said claim exists from the time of the incapacity for work until the termination of a six-week period. As far as this provision applies the employee will not incur the risk of loss of remuneration. Consequently, the said regulation implies also a breach of the principle of ‘no work, no wages’.7 a

b

Pursuant to Art 3, subsection 1, sentence 1, of the Law on the Continued Payment of Remuneration, the first prerequisite of the claim for continued payment of remuneration is the employee’s sickness. That is supposed to mean any irregular physical or mental condition. In this respect the cause of sickness is to be left out of consideration. So the sickness may be traced back, in a particular case, to, for example, a work accident, a sports injury or a road accident.8 The sickness as such, however, does not give rise to a claim for continued payment of remuneration, since it is not just any form of sickness which ipso facto prevents the employee from performing his work. If an employee catches a mild cold, for example, he will undoubtedly not be prevented from pursuing his work. Another prerequisite for a continued payment of remuneration therefore is the employee’s incapacity for

164

c

Risk and the Law work. Such incapacity for work must be assumed if an employee is incapacitated to perform the work he is obliged to do under the employment contract or if he could perform such work only by running the risk of worsening his state of health in the foreseeable future.9 Hence, the yardstick for appraising the employee’s incapacity for work is related to the characteristics of his job requirements. So a postman’s foot injury may render him unfit for work whereas an office clerk will retain his ability to work if he incurs the same damage. Moreover, the employer will have to bear the employee’s sickness risk only if the incapacity for work is not due to the employee’s fault. Sickness caused by the employee’s own fault may suffice, since the employee is by no means under an obligation to the employer with respect to keeping in good health. However, the employee is not entitled to assert a claim for the continued payment of remuneration if he, by his conduct, grossly violates the standard which can be expected of a sensible person in pursuit of his own interest. Therefore, slight negligence will not suffice ipso facto; the nature of his conduct must involve a special hazardous thoughtlessness, wilful recklessness or actual malice.10

The question whether the employee’s conduct amounts to wilful recklessness can be answered by clarifying the circumstances of the case. Taken from the vast number of court rulings, which is too large to be measured, the following examples illustrate the problematic nature of sports accidents. There are three different ways of being responsible for sports injuries. First, the employee may infringe the standing rules of the particular sport by acting with gross negligence and carelessness. Second, a person who engages in a sport in a way exceeding his physical strength and abilities to a large degree acts culpably. Finally, the Federal Labour Court will also take active negligence as established whenever a so-called dangerous sport is practised. The last case is conceived as involving a sport in which risk cannot be avoided even with good training regulations and strict observance of all the rules. It must be admitted, though, that up to the present no sport has been classified as falling into this category of dangerous sports. The judges have not been able to recognise the practice by the employee even of risk-related sports like boxing, motorcycle racing, hang-gliding and parachuting as from the outset a culpable way of acting. In view of these liberal judicial decisions the criterion of the practice of a dangerous sport proves to be of little significance. Doubts about this criterion have also been expressed on the ground that the practice of a specific sport in itself scarcely allows the drawing of conclusions as to the employee’s culpable way of acting. On the contrary, all the circumstances under which the particular sporting accident happened must be considered relevant. In particular, the employee’s skill in the sport is included.11 If the prerequisites for a continued payment of remuneration are met, the employer shall, in principle, continue the payment of the employee’s normal

The risk of sickness in German labour and social insurance law

165

wages under Art 4, subsection 1, of the Law on the Continued Payment of Remuneration. Therefore, the employee is to be regarded as if he has continued to be gainfully employed.12 In summary it can be stated that the employee’s sickness risk during the first six weeks of incapacity for work is, in principle, transferred to the employer by virtue of the Law on the Continued Payment of Remuneration. Only if the employee is guilty of gross negligence will the shifting of the sickness risk not take effect. Sickness benefit to be drawn from the statutory health insurance In the event of a long-lasting period of incapacity for work the employee is normally entitled to draw sickness benefit from the statutory health insurance scheme after the expiration of the six-week period of continued payment of remuneration.13 Pursuant to the Code of Social Law, volume 5, Art 44, subsection 1, sentence 1, the employee may make a claim for sickness benefit where he is incapacitated for work due to sickness. The sickness benefit together with the continued payment of remuneration serves the purpose of providing economic security for the sick employee. With respect to the prerequisites and legal consequences, however, the two forms of income maintenance differ in significant respects. a

b

c

d

This applies only to the category of persons entitled to make a claim. Whereas the employer is in principle bound to continue to pay remuneration to all his employees, only persons covered by the statutory health insurance may claim sickness benefit. Consequently, employees exempt from compulsory insurance cannot assert a claim under the present head. On the one hand, this category comprises employees drawing a monthly income in excess of
3,900;14 on the other hand, also lowincome earners, that is, employees with wages not exceeding
400 per month, belong to the said category.15 Consistently with the definition laid down in the Law on the Continued Payment of Remuneration, the constituent fact of ‘sickness’ is described here as an irregular physical or mental condition; the cause of such sickness is again left out of consideration. In the opinion of the Federal Social Court an insured employee is incapacitated for work if he is no longer able to do his current job or to pursue a similar kind of gainful employment due to his state of health or if continued activity would result only in a worsening of this state.16 Pursuant to Art 47 of the Code of Social Law, volume 5, the sickness benefit amounts to only 70 per cent of the earnings the employee has normally been paid before he fell sick; this regulation is in contrast to that of the continued payment of remuneration. Under Art 48 of the Code of Social Law, volume 5, the sickness benefit will be paid for a 78-week period in case of incapacity for work due to one and the same sickness.

166 e

f

Risk and the Law As well as in the case of the continued payment of remuneration the amount of the sickness benefit may be reduced if the sickness turns out to be caused by the entitled person’s own fault. However, gross negligence on the part of the employee will in that case not be ipso facto a sufficient reason. Moreover, the reduction of the claim will not take effect automatically by operation of law. On the contrary, the prerequisite for the reduction of the benefit under Art 52 of the Code of Social Law, volume 5 is the fact that the insured employee has caused the sickness wilfully or in connection with the commission of a criminal offence. In that case payment of the sickness benefit in whole or in part is to be left to the discretion of the health insurance. In summary, it can be stated that the insured employee’s sickness risk, from the 7th to the 78th week, is to be borne by the statutory health insurance to the extent of 70 per cent and as to the remaining 30 per cent by the employee. If the sickness has been caused wilfully the employee’s share may be increased up to 100 per cent. Those employees who are not members of the statutory health insurance will always have to bear the entire risk.

Final result When briefly outlining the provisions laid down in the civil law as well as in the labour and social law with respect to the distribution of the sickness risk, the following picture can be formed. During the first six weeks of the incapacity for work the employer has in principle to shoulder the entire sickness risk. From the 7th to the 78th week the risk is attributed to the statutory health insurance to the amount of 70 per cent. The insured employees and their employers are the funding parties which provide the economic basis for meeting the sickness costs. Each of the funding parties is required to pay one-half of the health insurance contributions.17 However, the risk for each and every sickness caused through a considerable fault of his own must be shouldered by the employee himself. Moreover, between the 7th and the 78th weeks the insured employee incurs approximately 30 per cent of his loss of earnings; the employee without health insurance coverage will even suffer a loss of earnings to an amount of 100 per cent.

Reasons for the distribution of risk This differentiated regulation leads us on to the question: what are the reasons which have induced the legislator to formulate this distribution of risk? Risk bearing by the employee In this respect the reductions of benefit provided for by the law in the event of a considerable fault on the part of the employee do not cause any

The risk of sickness in German labour and social insurance law

167

difficulties. In cases like these it would be unfair for the employer or the mutual benefit body of the statutory health insurance to be required to shoulder the costs of the compensation of earnings.18 Distribution of risk between employer and the statutory health insurance In contrast, the reasons for the distribution of risk between the employer and the statutory health insurance are not easily comprehensible. The obligation of the employer to continue the payment of remuneration seems to be reasonable only in so far as the incapacity for work results from a work accident, that is, can be seen as the realisation of an operational hazard.19 In other cases the employer bears the risks of all contingencies in the employee’s life. It is therefore not surprising that different views are held as to the justification of the obligation to continue the payment of remuneration. These views have been underpinned partly by legal and partly by politicaleconomic considerations. (a) Possibilities of legal obligations beyond the Law on the Continued Payment of Remuneration (i) Continued payment of remuneration as consideration due from the employer The pertinent literature defines in part the continued payment of wages as a debt the employer is obliged to pay by the nature of contract and the circumstances. The employer utilises the employee’s working capacity for the purpose of reaching an economic goal of his own. It is for this reason that, by the very nature of the situation, he is bound to render a consideration in order to guarantee security for the employee’s personal subsistence. The continued payment of remuneration is not a denial of the idea of ‘no work, no wages’; on the contrary, this remuneration is considered to be a consideration for the work the employee has already performed and for the overall work to be performed by him in the future.20 This opinion is not convincing, at least from the legal point of view. It has already been seen that the Civil Code, Art 326, subsection 1, sentence 1 involves the principle of ‘no work, no wages’. The Law on the Continued Payment of Remuneration thwarts this principle by granting the employee a claim for remuneration even for the period of his incapacity for work. With respect to the period of sickness the continued payment of remuneration therefore implies ‘wages without work’.21 Nothing is altered by this characterisation of the continued remuneration even when a period exceeding the period of sickness, perhaps even exceeding the total period of the employment, is taken into account. It can be seen that the breach of the principle of equivalence of contractual obligations caused by the sickness is

168 Risk and the Law also reflected in the overall obligations of employee and employer respectively. The continued payment of remuneration cannot be defined realistically as the consideration for a performance already rendered nor for the employee’s total performance. Such an interpretation would not be in accordance with the principle of mutuality of the contractual arrangement. This view is consequently not suited to provide a justification in law of the obligation to continue the payment of remuneration, but, at the most, to provide a socio-political justification. The same applies to the derivation of the obligation to continue the payment of remuneration from the nature of the situation; this also cannot legally be justified. (ii) Continued payment of remuneration as an aspect of the employer’s duty to provide for the welfare of the employee According to a view formerly widely held but currently rarely maintained,22 the obligation to continue the payment of remuneration arises from the employer’s duty to provide for the welfare of the employee. However, the objection to this explanation is that this duty – inasmuch as this term is still in use at all – implies today only such obligation to provide protection and support as can be derived from Art 242 of the Civil Code. This provision implies only obligations accessory to a contract but not a liability to pay remuneration as the primary obligation of the employer.23 Consequently, the liability for the continued payment of remuneration cannot be based upon the employer’s obligation to provide for the welfare of the employee.24 (iii) Final result As an interim result it must therefore be stated that the employer’s obligation to continue the payment of remuneration does not arise from reasons embedded in the dogmatic nature of private law which might have priority over the Law on the Continued Payment of Remuneration. On the contrary, the obligation is fully constituted by the Law on the Continued Payment of Remuneration itself. This brings us to the question of the specific aims which the legislator had in mind when laying down this regulation. Stated objectives of the Law on the Continued Payment of Remuneration (i) Income maintenance of the employee Naturally, the Law on the Continued Payment of Remuneration aims primarily at the financial cushioning of the employee in the initial stage of his incapacity for work.25 In the absence of a claim to the continued payment of remuneration the sick employee would be paid sickness benefit only by the health insurance system. This benefit has always been lower

The risk of sickness in German labour and social insurance law

169

than the amount payable under the Law on the Continued Payment of Remuneration. (ii) Relief of the statutory health insurance fund In addition, the continued payment of remuneration aims at the financial relief of the statutory health insurance. This objective finds expression in two amendments of the law from 1930/31 which still govern the distribution of risk between the employer and the health insurance.26 On the one hand, avoidance of the legal provisions regarding the continued payment of remuneration was abolished, at least in favour of employed persons.27 The claim for continued remuneration could henceforth no longer be excluded or limited by the employment contract, as had previously often happened. On the other hand, it was determined that payments made by the employer were to be credited in full to the health insurance as the payment of benefits.28 By means of this regulation the legislator has given the continued payment of wages priority to the sickness benefit. By virtue of these innovations the legislator of the time effected a deliberate shift of obligations from the health insurance to the individual employer. (iii) Final result The above reasoning makes it clear that the present distribution of the sickness risk between the employer and the statutory health insurance does not originate from the dogmatic principles of private law or labour law. On the contrary, the employer’s obligation to continue the payment of remuneration is exclusively derived from the socio-political intention of the legislator, first to guarantee the employee a better social cushioning than that provided by the sickness benefit and second to give financial relief to the statutory health insurance.

Prospects for the future My previous statements concerned the distribution of risk under the current law. In the run-up to the 2004 health care reform a redistribution of the sickness risk has commenced. The aim of this is again to afford financial relief to the health insurance schemes. Unlike that provided under the regulation laid down 70 years ago, this relief is definitely not intended to be to the disadvantage of the employer. For example, nobody considers the possibility of extending the six-week period of the continued payment of remuneration so as to set a longer period of suspension of the claim for sickness benefit.29 Instead, Federal Chancellor Schröder announced in his government policy statement of 14 March 2003 a plan to abolish the sickness benefit in its entirety. The employee would be expected to secure insurance on a private basis against loss of earnings due to sickness.30 By this

170

Risk and the Law

measure the Federal Chancellor aimed to shift the sickness risk of an employee from the health insurance system to the person involved. It is uncertain, however, in which way the employer’s obligation to continue the payment of remuneration would be affected by the abolition of the sickness benefit. Would it not be reasonable also to repeal the Law on the Continued Payment of Remuneration as part of the process of the health reform? In that case the employee would naturally also have to shoulder the remuneration risk for the first six weeks of the sickness. There is one argument that seems to speak for such an innovation. This relies on the fact that the continued payment of remuneration would lose its function of providing relief to the health insurance system following the abolition of the sickness benefit. However, there are likewise arguments that speak against the repeal of the Law on the Continued Payment of Remuneration. These are derived from the fact that the social protective function of the continued payment of remuneration would become all the more evident after the abolition of wage compensation under social welfare law. Moreover, the repeal of the Law on the Continued Payment of Remuneration in practice would not in many cases discharge the employer from the obligation to continue the payment of remuneration since the obligation has been incorporated and defined to a large extent in the provisions of collective agreements.31 It appears therefore that the forthcoming restructuring of the welfare state will probably result in a redistribution of the employee’s sickness risk. Whereas up to the present it has been primarily the employer and the statutory health insurance system that were bound to bear the risk, the employer and the employee himself seem to have to shoulder this risk in the future. Thus, in the relationship between personal responsibility and solidarity, the personal responsibility might regain considerable importance.

Notes 1 H. Kohler, ‘Krankenstand – Ein beachtlicher Kostenfaktor mit fallender Tendenz’, IAB Werkstattbericht, Nr. 1/2002, 3, pp. 3 et seq. 2 Bundesvereinigung der Deutschen Arbeitgeberverbände, Soziale Selbstverwaltung 2002, 39, p. 39. 3 U. Preis et al. (eds.), Erfurter Kommentar zum Arbeitsrecht, 3rd edn., München: C.H. Beck, 2003, § 611 BGB sec. 837–839; M. Gotthardt, Arbeitsrecht nach der Schuldrechtsreform, München: C.H. Beck, 2002, sec. 75 with further references. 4 ‘Gegenäußerung der Bundesregierung auf die Stellungnahme des Bundesrates zum Entwurf eines Gesetzes zur Modernisierung des Schuldrechts’, Deutscher Bundestag Drucksache 14/6857, 42, pp. 47 et seq. 5 §§ 45–52 and §§ 56–62 Sozialgesetzbuch (SGB) VII. 6 § 43 SGB VI. 7 U. Schulte-Mimberg, in B. Schulin (ed.), Handbuch des Sozialversicherungsrechts, Band 1, Krankenversicherungsrecht (HS-KV), München: C.H. Beck, 1994, § 13 sec. 49; J. Schmitt, Entgeltfortzahlungsgesetz, 4th edn, 1999, ‘Einleitung’, sec. 1; W. Boecken, ‘Entgeltfortzahlung im Krankheitsfall’, in R. Richardi and O. Wlotzke (eds.), Münchner Handbuch zum Arbeitsrecht, 2nd edn, München: C.H. Beck, 2000, § 82 sec. 1, § 83 sec. 40.

The risk of sickness in German labour and social insurance law

171

8 Boecken, n. 7 above, § 83 sec. 25–29; Schmitt, n. 7 above, § 3 sec. 34–38. 9 Entscheidungen des Bundesarbeitsgerichts Arbeitsrechtliche Praxis (BAG AP), Nr. 42 (Blatt 879) and Nr. 52 (Blatt 542) § 616 BGB; BAG AP Nr. 62 § 1, LohnFG, sheet I 1; BAG AP Nr. 1 § 74 SGB V, sheet II 1. 10 H. Oetker, in J. von Staudinger (ed.), Kommentar zum Bürgerlichen Gesetzbuch, Berlin: de Gruyter, 2002, § 616 sec. 240–251; H. Kaiser, H. Dunkl, D. Hold and G. Kleinsorge, Entgeltfortzahlungsgesetz, 5th edn, Köln: Dr Otto Schmidt, 2000, § 3 sec. 92–117; Boecken, n. 7 above, § 83 sec. 92–122; Schmitt, n. 7 above, § 3 sec. 84–118; H-J. Dörner in Preis et al. (eds.), n. 3 above § 3 EFZG (Entgeltfortzahlungsgesetz) sec. 46–63. 11 Cf. Houben, SpuRt 2000, 185; Boecken, n. 7 above, § 83 sec. 119–121; Dörner, n. 10 above, § 3 EFZG sec. 52. 12 Kaiser, Dunkl, Hold and Kleinsorge, n. 10 above, § 4 sec. 7 et seq.; Schmitt, n. 7 above, § 4 sec. 19 et seq. 13 Cf. § 49 Abs. 1 No. 1 SGB V. 14 § 6 Abs. 1 No. 1 SGB V and Abs. 6 Satz 1 SGB V. 15 § 7 Abs. 1 Satz 1 SGB V i. V. m. §§ 8 and 8 a SGB IV. 16 Entscheidungen des Bundessozialgerichts 46, 190 (191); 47, 47 (51). 17 § 249 Abs. 1 SGB V. 18 Kaiser, Dunkl, Hold and Kleinsorge, n. 10 above, § 3 sec. 94; Schmitt, n. 7 above, § 3 sec. 85, 88. 19 Cf. § 52 No. 1 SGB VII. 20 H. Wiedemann, Das Arbeitsverhältnis als Austausch- und Gemeinschaftsverhältnis, Karlsruhe: C.F. Müller, 1966, pp. 15 et seq.; K. von Koppenfels, ‘Die Entgeltfortzahlung im Krankenheitsfall an der Schnittstelle von Arbeits- und Sozialrecht’, Neue Zeitschrift für Socialrecht, 2002, 241, 245–247; I. Koller, Die Risikozurechnung bei Vertragsstörungen in Austauschverträgen, München, C.H. Beck 1979, pp. 399–401; cf. E.A. Kramer, Arbeitsvertragsrechtliche Verbindlichkeiten neben Lohnzahlung und Dienstleistung, Wien: MANZ’sche, 1975, p. 52 et seq.; P. Schwerdtner, ‘Fürsorge- und Treuepflichten im Arbeitsverhältnis’, Zeitschrift fur Arbeitsrecht, 1979, 1, 19 et seq. 21 von Koppenfels, n. 20 above, 241, 245 et seq. 22 BAG AP Nr. 25, 34, 68 and 72, § 1 LohnFG; A. Hueck and H.C. Nipperdey, Lehrbuch des Arbeitsrechts, vol. 1, 7th edn, Berlin and Frankfurt am Main, Franz Vahlen 1963, p. 329; E. Kreßel, ‘Lastenverteilung im Arbeitsrecht – dargestellt am Beispiel der Entgeltfortzahlung’, in M. Heinze and J. Schmitt (eds.), Festschrift für Wolfgang Gitter zum 60. Geburtstag, Wiesbaden: Chmielorz, 1995, 491, pp. 500 et seq. 23 W. Blomeyer, in R. Richardi and O. Wlotzke (eds.), n. 7 above, § 94 sec. 12–16; Oetker, n. 10 above, § 616 sec. 174; Preis et al., n. 3 above, § 611 BGB sec. 760–762. 24 Koller, n. 20 above, pp. 398 et seq.; Oetker, n. 10 above, § 616 sec. 174; von Koppenfels, n. 20 above, 241 pp. 244 et seq. 25 Boecken, n. 7 above, § 82 sec. 30 with further references. 26 Entscheidungen des Bundesgerichtshofes in Zivilsachen 7, 30 (40, 47); Boecken, n. 7 above, § 82 sec. 6–8, 31–34; Schulte-Mimberg, n. 7 above, § 13 sec. 21–24; B. Schulin, in: M. Heinze and A. Söllner, Arbeitsrecht in Der Bewahrung: Festschrift Fur Otto Rudolf Kissel Zum 65. Geburtstag, München: C.H. Beck, 1994, 1055 (1062 et seq.); Oetker, n. 10 above, § 616 sec. 170. 27 Cf. § 12 EFZG. 28 Cf. § 49 Abs. 1 Nr. 1 SGB V. 29 Cf. § 49 Abs. 1 Nr. 1 SGB V. 30 Frankfurter Allgemeine Zeitung, 15.3.2003. 31 Boecken, n. 7 above, § 82 sec. 23–25.

12 ‘The butcher’s cart and the postman’s bicycle’: risk and employers’ liability George Applebey

INTRODUCTION

Hazard and risk 1 Risk, of course, is not the same as hazard.2 The British Management of Health and Safety at Work Regulations 1999 attempt to give a definition of the distinction: (a) a hazard is something with the potential to cause harm (this can include articles, substances, plant or machines, methods of work, the working environment and other aspects of work organisation); (b) a risk is the likelihood of potential harm from that hazard being realised. The extent of the risk will depend on: (i) the likelihood of that harm occurring; (ii) the potential severity of that harm . . .; and (iii) the population which might be affected by the hazard, i.e. the number of people who might be exposed.3 The notion of risk diverges when used in different areas of law which deal with accident and illnesses at work. Under the criminal regulatory system the primary objective is to prevent injury and to a lesser extent to deter wrongdoing. Prevention involves identifying risks in advance and seeing their potential for harm. The Management of Health and Safety at Work Regulations 1999, based upon the European Framework Directives of 1989, carry this into practice in Britain. Risk is also relevant in tort law, especially negligence, where foreseeability of risk is well established. In contract law the existence of risk is one of the motivating forces of making a contract. By allocating risks (if they are free to do so) the parties seek to organise and clarify their liabilities and potential risks in advance. Finally in social security law, there is a vital distinction between ‘occupational’ and ‘nonoccupational’ risks of accidents and of the causes of conditions and diseases. The latter is largely a matter for the science of epidemiology to determine and for medical experts to argue over questions of aetiology. Only injuries

Risk and employers’ liability

173

which are occupational may qualify. If the risk is common to all human kind in largely equal measure then it does not count. In each of these separate fields of law which deal with injuries at work in the United Kingdom, the notion of risk varies both in scope and in its purpose.

Injury, harm, guilt and liability If we were to describe related aspects of a paradigmatic civil action the requirements of liability would be hazard creating a risk, leading to injury, resulting in liability. This would be the normal pattern in a claim where injury was a necessary precondition or element. However, under the regulatory system harm to an individual is not a prerequisite of prosecution. This risk of injury is sufficient. Thus, in R v Board of Trustees of the Science Museum,4 the Kensington Museum was held to be guilty of an offence under the Health and Safety at Work Act 1974 s 2 by merely permitting the existence of legionella pneumophilia, the bacterium causing legionnaire’s disease, in the water system, creating a potential risk to the public though nobody became ill. Under the Health and Safety at Work Act there may even be enforcement in respect of potential hazards. In this case not only injury but even risk could be absent if a danger existed at an accident site. In Railtrack plc v Smallwood,5 a prohibition notice regarding a potential hazard on a railway was upheld even though the railway line in question to Paddington Station was closed at the time in the aftermath of a major rail disaster. The expressions ‘activities’ and ‘at work’ were given an extended and purposive interpretation even though in the context at the time neither appeared to be taking place.

Causation, law and social policy The relationship between hazard, risk, injury and guilt or liability therefore depends on several factors: (a) causation (i) in fact (ii) or in law, (b) legal rules and (c) public or legal policy. Risk becomes the key factor linking hazards and danger in the workplace to ultimate guilt or liability. As such it is not only vital for employer liability in its various legal ramifications, it also plays a significant role in economic and policy considerations as well as in legal choices. In the social security system the government has a clear (perhaps even an overriding) concern to keep the cost of the system low. The factors involving the calculation of risk for occupational injuries are many, but outstanding amongst them must be the following: (a) the existence of a legal duty of care; (b) the standard of care required, which might be strict, absolute, based on reasonable care or practicability, or limited to using ‘best endeavours’ etc.; (c) the seriousness or gravity of the risk; (d) the likelihood or probability of the risk occurring; (e) the effect of the risk materialising on individuals or on groups of the population; (f) the costs of dealing with the risk; (g) the foreseeability of risk; (h) the social and economic impact of risk

174

Risk and the Law

allocation; (i) contributory or other causative factors; (j) the social and economic impact of risk allocation; (k) the means, in terms of money, time and energy, of avoiding risks; (l) knowledge or awareness of risk both by persons concerned but also by the state of scientific knowledge at any time; (m) the scientific calculation of risk, which might be based on statistics, epidemiology or even ‘informed’ guesswork and common sense. The concept of risk is therefore a creature with many heads, capable of applying in different contexts in distinct areas of law. We turn to this in the following sections.

THE BRITISH REGULATORY SYSTEM

The Health and Safety at Work Act 1974 Laws governing the health of workers in Great Britain can be traced back to the beginning of the nineteenth century. By the middle of that century, Factories Acts had a secure (albeit controversial) place as regulators of industrial activity. By the early 1970s this law had grown and developed into so many different pieces of legislation that it had become a mess. The Health and Safety at Work Act 19746 established a new structure and laid down a new set of general duties. Sections 2–8 of the Act lay down wide obligations on a cross-section of persons, employers, contractors, those who occupy building for work purposes, those who conduct undertakings and also on employees themselves. Significantly the 1974 Act is designed to protect not only workers but also others who may be affected by the activities of persons at work. This includes neighbouring property owners and occupiers, and even the general public. The two key sections of the 1974 Act are s 2 and s 3. The former requires every employer to ensure the health and safety at work of his employees ‘so far as is reasonably practical’. Under s 3 those who conduct an undertaking are also under a duty towards those at risk of injury and must take measures again so far as is ‘reasonably practicable’. Failure to comply with the general duty sections can lead to fines of up to £20,000 in magistrates’ courts or unlimited fines in the Crown Court. The interface between ss 2 and 3 and who is under the main duties imposed by the Act, i.e. the employer or others, has been the subject of a good deal of case law.7 The Act requires that the prosecution establish a prima facie breach of the statute or of one or more of the regulations made thereunder. (It is very common for more than one breach of the regulations to be charged in proceedings arising out of one incident.) Thereafter the burden of proving that it was not reasonably practicable to comply with the Act or regulations shifts to the defendant. The defendant need only prove this on the lower civil standard of ‘balance of probabilities’ rather than the normal criminal standard of proof beyond a reasonable doubt.8 There is a growing case law on this

Risk and employers’ liability

175

reversal of the burden of proof in strict liability offences. The principle has been challenged as infringing Art 6 of the Human Rights Act giving a right to a fair trial including the presumption of innocence. The courts have distinguished a reverse ‘legal burden’ and a mere ‘evidential burden’.9

The phrase ‘reasonably practicable’ The meaning of the phrase ‘reasonably practicable’ is generally traced back to an early case interpreting the words as used in a pre-1974 piece of legislation. In Edwards v National Coal Board,10 Asquith LJ stated: ‘Reasonably practicable’ is a narrower term than ‘physically possible’, and implies that a computation must be made in which quantum of risk is placed in one scale and the sacrifice involved in the measures necessary for averting the risk (whether in money, time or trouble) is placed in the other, and that, if it be shown that there is a gross disproportion between them – the risk being insignificant in relation to the sacrifice – the defendants discharge the onus upon them. Moreover, this computation falls to be made by the owner at a point of time anterior to the accident. Reasonably practicable means that the risk is essentially based upon a negligence standard, i.e. balancing of the likelihood and extent of harm against the measures necessary to restrict or eliminate the risk entirely. The Act’s qualification of the central duties to a standard of reasonable practicability has led to a divergence of views on the extent of the duties imposed. Are they (1) in fact absolute and qualified by a defence of reasonably practicable or (2) in essence no more than duties to do what is reasonable to guard against risks? There has been a divergence of judicial views on this point.11

Relationship between British and European Union law The British concept of risks modified to a standard of reasonably practicable has led to difficulties and questions over implementing European directives. The majority of European directives in this area lay down what appear to be absolute objectives, albeit subject to defences, whereas the British approach has historically been detailed and prescriptive but qualified to a reasonableness standard. The directives have been made a part of British law as regulations under the 1974 Act. The questions of whether the directives have been implemented correctly have been controversial in several cases. The European Union has been highly influential in adding new law in the field of health and safety. Starting with the Framework directive in 1989 (Council Directive 89/391), there were added at the same time a further five key directives which were made part of British law in 1992. These were the Workplace Regulations, Use of Work Equipment, Display Screen Equipment

176 Risk and the Law Regulations, Manual Handling of Loads Regulations 1992. The most important, however, are the Management of Health and Safety Regulations 1992. These are the most widely used regulations in the UK and introduced the requirement of the ‘risk assessment’ (which we discuss shortly).

Improvement and prohibition notices This is an area where the concept of risk is clearly apparent. The 1974 Act introduced a new means of enforcing health and safety in the workplace. Inspectors may now issue notices even without injury to an individual or even in some cases without having to prove a breach of the law. There are two types of notices. (1) Improvement notices require a party to bring their workplace into line with the law within a stated time scale. An offence must be specified (usually a breach of regulations). It is an offence to ignore or fail to comply with a notice. If the recipient objects to the notice he may appeal to an employment tribunal and from there to the High Court. (2) The second category of notices are prohibition notices. The inspector need only form the opinion that there is a ‘risk’ of ‘imminent personal injury’. The notice may be ‘immediate’ or ‘deferred’. The process or indeed the entire workplace may be closed on this basis. A notice may even be issued where no work is taking place.12 Notices are used extensively in British health and safety law – thousands are issued annually. This is now the principal means for enforcing health and safety legislation.

Risk and sentencing Even after a person has been found guilty of a criminal offence, the risk involved may be relevant to the level of fine imposed. In R v Howe and Son (Engineers) Ltd [1999] 1RLR 434 a small precision engineering company in Bristol was found guilty of an offence under the Health and Safety at Work Act. An employee using a type of electric vacuum machine known as a ‘Freddy’ was electrocuted due to the machine becoming live as a result of the cable being in a dangerous state of disrepair. The Court of Appeal Criminal Division held that among the factors to be taken into account in determining the proper fine was ‘the level of risk involved in the task being performed at the time’.13 Risk can therefore be an aggravating or mitigating factor in setting the appropriate penalty.

Risk assessments: the Management of Health and Safety Regulations 1999 Regulation 3 requires every employer to make a ‘suitable and sufficient’ assessment of (a) the risks to the health and safety of his employees to which they are exposed whilst they are at work and (b) the risks to persons not in his employment arising (i) out of or (ii) in connection with the conduct of his

Risk and employers’ liability

177

undertaking. The purpose of risk assessment is to identify measures to comply with the requirements and prohibitions under the relevant statutory provisions. (It is typical of the confused use of the word ‘risk’ that risk assessment is largely about identifying and dealing with hazards at work.) These regulations are made under the 1974 Act. They also include approved Codes of Practice and Guidance Notes for their application. This type of ‘soft law’ is very common in UK health and safety rules. The practical test under the 1999 regulations is that under the risk assessment system, there should be a goal-setting regime for health and safety risks. Those on whom duties are imposed should undertake their working operations at levels of risk which are ‘as low as reasonably practicable’ (ALARP).

CIVIL LIABILITY

Tort action in negligence Employers’ liability claims follow the basic principles of negligence where risk is obviously the central tenet of liability. Duty of care, the standard of reasonableness and reasonable foreseeability all apply. The defence of assumption of risk (volenti non fit injuria) has not applied to employers’ liability actions for many years and was rendered ineffective in cases of personal injuries generally by ss 2 (1) and 2(3) of the Unfair Contract Terms Act 1977. The leading case establishing negligence in actions for employers’ liability is Wilsons and Clyde Coal Company v English.14 The duty is fourfold. The employers must provide (a) competent and properly trained staff, (b) adequate plant and equipment, (c) a safe place of work and (d) a safe system of work. In the case of faulty work equipment which causes injury the law was amended in 1969 by the Employers’ Liability (Defective Equipment) Act which renders the employer liable, even if the fault in question was that of the manufacturer of the tool or equipment. This has been applied to everything from a steel chisel to a supertanker.15 For decades actions in negligence have been commonplace for accidents at work, but have increasingly been used in recent years for widespread occupational health hazards, for instance, occupationally caused hearing loss (deafness), conditions such as repetitive strain injuries and now stress at work.16 A recent case has now set out clearly the parameters of liability for occupational stress claims.17 The risk in employers’ negligence cases is of course taken over by insurance in the UK. Under the Employers’ Liability (Compulsory Insurance) Act 1969 it is a criminal offence not to insure against claims by a workplace. Private insurance companies and their solicitors therefore still play a huge part in handling claims as well as defending and settling legal actions for personal injuries by employees. In the United Kingdom many trades unions provide free legal services for their members, including the not-inconsiderable

178 Risk and the Law costs of litigation. The trades unions have over the years been of enormous assistance to workers who otherwise could not have afforded the risk of embarking upon litigation in the civil courts. They deserve a great deal of credit for this benefit.

The tort of breach of statutory duty As a result of the failure to develop an adequate tort action for injuries at work in the nineteenth century (see next section) the courts in England developed a new tort which has become the main action for injuries at work. Starting with Grove v Lord Wimbourne 18 the courts adopted the view that, if it could be shown that Parliament intended that a criminal statute was also intended to give victims a civil action that could be applied in an action for injuries at work. This has become its principal importance in British law. The claimant must prove (a) that there was a breach of the statute in question, (b) that he or she suffered an injury and (c) that the injury was caused by a breach of statute. Breach of statutory duty is commonly pleaded with negligence. The principal difficulty is showing that a civil action (as well as criminal sanctions) was intended. Most legislation is silent on the matter. The Health and Safety at Work Act’s general duties do not give a civil action to those injured as a consequence of its breach but regulations made thereunder do, unless they state otherwise.19 An interesting recent case in this area is Stark v Post Office.20 The claimant was a postman in Hull who was propelled over the handlebars of his bicycle on which he was delivering letters. The reason for the accident was that the brake stirrup broke and lodged in the front wheel, causing it to lock. The bicycle was 14 years old but the employer had done its best to maintain it in good working order. The cause of the breakage was probably metal fatigue. The Court of Appeal held that the plaintiff should succeed in recovering damages from the defendants, the Post Office, not for negligence but for breach of the Provision and Use of Work Equipment Regulations 1992. It was held that these imposed an absolute obligation to maintain the bicycle in efficient working order. The reason for this was not only the preceding European Work Equipment Directive21 but also the history of British law in this area. Certain key statutory provisions had required strict liability and the directive was intended to maintain or improve standards, not to make them worse. Although not strongly argued in the case itself, under British law s 15 of the Health and Safety at Work Act allows only for regulations that maintain or improve health and safety requirements. However, in Horton v Taplin Contracts Ltd,22 the Court of Appeal held an employer not to be in breach of the Work Equipment Regulations or the Construction Regulations 1996 when a worker was pushed off scaffolding by a fellow worker even though the scaffolding did not have a stabiliser or a guard rail. This was because the injury was caused by an event which was not reasonably foreseeable.

Risk and employers’ liability

179

CONTRACT LAW AND OCCUPATIONAL RISKS Traditionally English contract law espoused the principle of freedom of contract. Nowhere was this more obvious than in relation to employment. The nineteenth century referred to this as the law of master and servant. The courts put few impediments in the way of the employer’s freedom. In 1836 an overloaded butcher’s cart filled with meat gave way in the street at Peterborough. An employee sitting on the cart was thrown to the ground and sustained a broken leg. He sued his employer for the fault of another employee in overloading the cart. The House of Lords held his action must fail. The plaintiff had assumed the risk of injuries at work. There was an implied term that the employee took the risk when he agreed to do the work. This was known as the doctrine of common employment.23 Today the position is the reverse. The risk is no longer with the employee but with the employer. Risk cannot be allocated to an employee. Under the Unfair Contract Terms Act s 2(3) awareness of or agreement to even an express term in a contract cannot be taken to assumption of risk. Furthermore, there is an implied term in every contract of employment that the employer will take reasonable care for his employee’s health and safety.24 In the last decade one of the controversial areas of health and safety has been the regulation of hours of work of employees across the European Union. The Working Time Directive was challenged by the UK government as not being a health and safety issue.25 The UK government was forced to back down on the issue and has implemented the Working Time Regulations 1998. The English courts have subsequently taken this a step further. In Barber v RJB Mining (UK) Ltd,26 the High Court held that the Working Time Regulations imposed a contractual obligation on an employer not to require an employee to work more than 48 hours during any reference period. (There are, of course, many derogations and exceptions to these regulations.) Junior hospital doctors claimed in Johnstone v Bloomsbury Area Health Authority 27 that a term requiring working hours and being ‘on call’ for over 100 hours a week infringed the implied term of health and safety in their contracts. It was therefore an exemption clause subject to the Unfair Contract Terms Act and rendered void. Lack of sleep could be a health issue for doctors and a risk to patients. The working hours of younger hospital staff still remain over the European limits however. Contract law is still a classic area of risk theory. Allocation of risk by express or implied term stands outside the realm of regulations and determines liability or otherwise. It is one of the primary objects of contracts. By contract employers frequently assign risks or agree indemnity and disclaimer clauses. Depending upon the context the risk may or may not be transferred.28

180

Risk and the Law

THE NOTION OF RISK IN UK SOCIAL SECURITY LAW In Britain we usually credit Germany and Chancellor Bismarck with introducing the first pieces of social insurance legislation in western Europe. In Britain the first example of this process was the Workmen’s Compensation Act 1897. This was a scheme of supposedly no-fault liability for accidents at work. It still exists in many English-speaking countries such as the United States and Australia. In Britain it became part of the welfare state benefits scheme in 1948. It now forms part of the social security system. Although the funding and payment of benefits are now based upon social security, the rules of the system have remained remarkably similar over the 100 years. There are two routes by which a claimant may obtain industrial disablement benefit. The first is ‘by accident arising out of and in the course of employment’. This requires an event (or a series of events) which can be proved to have a specific date or time attributed to it. Otherwise the claimant must belong to an occupation or work in a business or industry for which a disease or condition has been prescribed in advance as qualifying for benefit. He or she may not bring a claim based solely on their own illness or injury if it has not been prescribed. There are therefore two important aspects of this system for the purposes of a consideration of risk: (a) the meaning of ‘employment’ as opposed to a ‘non-employment’ risk and (b) the principles by which diseases are prescribed for the awarding of the state benefits. This lies in the hands of the Industrial Injuries Advisory Council, referred to again below. Persons injured at work are entitled to claim both damages in the courts and social security benefits. If their civil action succeeds, however, benefits for up to five years will be deducted from the damages by the Department of Social Security.

What is an ‘employment risk’? The phrases ‘in the course of’ and ‘arising out of’ the employment remain the statutory test for recovery by accident for industrial disablement benefit. These are regarded as being separate requirements. ‘In the course of’ is defined by the time, space and activity of work. ‘Arising out of’ demands a causal connection between the accident and the nature of the job or work being undertaken. For instance, a heart attack at work would be in the course of employment but would not arise out of the employment unless the claimant’s heart condition resulted from the work, which would be very unlikely and almost impossible to prove. Once it has been established that the accident would not have occurred ‘but for’ the employment, it will generally be held to arise ‘out of the employment’ ‘unless a claimant added or created a different risk and this different risk was the real cause of the accident’.29

Risk and employers’ liability

181

To assist claimants there is now a statutory prescription that ‘an accident arising in the course of an employed earner’s employment shall be taken in the absence of evidence to the contrary, also to have arisen out of that employment’.30 For the presumption to benefit the claimant he must not only prove that the accident occurred in the course of employment but also that there is no evidence to the contrary. In nearly all disputed cases there is ‘evidence to the contrary’ so the presumption does not materially improve the position of claimants. There are a number of problematic examples of ‘employment risk’ worth considering. Pre-existing conditions The work must contribute to a material degree to the risk. Once that has been established it does not matter that a pre-existing health risk made a person more vulnerable to injury.31 Common risks These are risks which are common to mankind but which the claimant would not have suffered had the employment not brought him to the particular spot (where, for example, he was stung by a wasp). The industrial injuries commissioners have developed the idea of ‘locality risks’ – that is that the employment took the claimant to the spot where the hazard was encountered. In the circumstances the claimant can recover benefit. Special risks Misconduct by others, behaviour of animals, and dealing with an emergency ‘in or about’ the place of employment are all treated as arising out of the employment. Contributory risks If a claimant ‘adds or creates a different risk’ then he or she cannot receive benefit if ‘this different risk was that cause of the accident’.32

Prescribed diseases If a claimant does not qualify by accident his or her only claim is through a schedule of prescribed diseases shown to be typical risks for certain specified categories of employment. In order to qualify the claimant has to show that (i) he or she suffers from a prescribed disease, (ii) the disease is prescribed in relation to his occupation and (iii) the disease developed in relation to that occupation. If (i) and (ii) are established then the claimant is given the benefit of a prescription that the disease is occupationally caused. This can

182

Risk and the Law

be rebutted by the adjudication officer proving on the balance of probabilities that the disease was not due to the nature of his occupation. A claimant cannot bring a claim on its own merits if a disease has not previously been prescribed. In 1981 the Industrial Injuries Advisory Council (IIAC) recommended a system of ‘individual proof’. This has not been enacted.

The Industrial Injuries Advisory Council This is the quango (quasi-autonomous non-government organisation) which advises the UK government on industrial injuries and among other things deals with the prescription of diseases. It was my privilege to serve on it as a member from 1987–1993. A disease may be prescribed for social security benefits in relation to any employed earners if: (a) the disease ought to be treated having regard to its causes and incidence and any other relevant considerations, as a risk of their occupations and not as a risk common to all persons; and (b) it is such that, in the absence of special circumstances, the attribution of particular cases to the nature of the employment can be established or presumed with reasonable certainty.33 The IIAC prescribed diseases with a clear occupational origin from an early date. Now the focus has shifted and includes many conditions caused by non-occupational factors. As a ‘rule of thumb’ the IIAC requires a minimum of a doubling of risk before a disease is prescribed. Since such a clear increased risk is difficult to prove with reasonable certainty the process of prescribing new diseases is slow, although often also controversial. If a condition develops incrementally over a period of time or as a result of wear and tear it is categorised as a ‘process’. This is neither an accident nor a prescribed disease and therefore fails to qualify for benefit.34

CONCLUSION Risk is a central element in health and safety law in the United Kingdom. It applies to risk assessments seeking to prevent injuries, in establishing liability to pay compensation, in the sentencing process for the proper level of punishment for criminal offences, and in the concept of what is ‘occupational’. The importance of risk is a central concern of health and safety at work lawyers.

Notes 1 The conference paper at Schloss Thurnau of which this chapter is an altered version was translated into German by Professor Eberhard Eichenhofer of the

Risk and employers’ liability

2

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34

183

University of Jena and is published as ‘Ein Überblick über Risiken und unterschiedlichen Fragen nach Occupational Health and Safety Law for Injuries at Work in the United Kingdom’ in ZESAR (Zeitschrift für europäisches Sozial- und Arbeitsrecht), 2003, 312–317. For an intellectual account of ‘risk’ and the history of this idea in economics see P.L. Bernstein, Against the Gods: The Remarkable Story of Risk, New York, etc.: John Wiley & Sons, 1996. Unfortunately this book does not deal with the subject matter of this chapter, the risks of suffering personal injuries while at work. Management of Health and Safety at Work Regulations 1999, Approved Code of Practice, para 11. [1993] 1WLR 1171. [2001] 1CR 714. The Act is based upon the report of a committee set up to examine the law known as the Robens Report. ‘Safety and Health at Work’, London: HMSO, 1972, Cmnd 5034. R v Swan Hunter Shipbuilders and Telemeter Installations Ltd [1981] 1 RLR 403. Davies v Health and Safety Executive [2003] 1RLR 170. R v Lambert [2001] 3 WLR 206, DPP v Sheldrake [2004] UKHL 43. [1949] 1KB 704. See Steyn LJ in R v British Steel plc [1995] 1 WLR 1356. Railtrack plc v Smallwood [2001] EWHC Admin 78 [2001] 1 CR 714. See also R v Friskies Petcare (UK) Ltd [2000] 2 Cr App R 401, and R v Fresha Bakeries [2003] 1 Cr App R (s)44. [1938] AC 57. The Derbyshire (lost in the Pacific Ocean with all its crew): see Coltman v Bibby Tankers [1988] AC 276. Walker v Northumberland County Council [1995] ICR 702. Hatton v Sutherland [2002] 2 All ER 1. [1898] 2 QB 402 s 47 [2000] 1CR 1013 89/655/EEC The Times 25 November 2002 (1837) 3 M and W 1. See Lister v Romford Ice and Cold Storage Co [1957] AC 535. United Kingdom v EU Council [1996] All ER (EC) 877. [1991] IRLR 308. [1991] 2 All ER 293. See Thompson v T Lohan (Plant Hire) Ltd [1987] 1 WLR 649 and Phillips Products Ltd v Hyland [1987] 1 WLR 659. R(1) 2/63. See generally A.I. Ogus, E.M. Barendt and N. Wikeley, The Law of Social Security, 5th edn, London: Lexis/Nexis, 2002, pp. 306–328. Social Security Claims and Benefits Act 1992 s 94(3). R(1) 11/80. R(1) 2/63. Social Security Contributions and Benefits Act 1992 s 108(2). Faulds v Chief Adjudication Officer [2000] 1 WLR 1035 provides an example of the operation of this important lacuna in the industrial injuries scheme.

13 The limits of individualisation in the risk society: social security in the customary laws of immigrant communities Gordon R. Woodman THE NATURE OF RISK IN MODERN SOCIETY AND THE LIMITS OF THE COMMON CONCEPTION OF SOCIAL SECURITY This chapter suggests that it is useful to give the notion of ‘law’ in the title of this volume, Risk and the Law, a broader meaning than merely that of state law. It contends that we should take account of what some call non-state law, or alternatively informal law, unofficial law, folk law (Volksrecht), living law (lebendes Recht) or customary law, and the ways in which these other laws give protection against risk. More particularly it is suggested that there are fruitful possibilities for the investigation of social security provisions in the non-state laws of immigrant communities in the United Kingdom, especially in the recent past and the present. Any findings may have practical importance for the development of public and private state law regarding protection against risk. A preliminary comment may be needed on a common assumption concerning the extent of risk to the individual of suffering serious damage in modern society. The assumption is that this risk is now much reduced. It is of doubtful validity. There has been a reduction in certain risks of extreme damage such as death through starvation or common disease. In consequence the need for social institutions which in the past were designed to reduce these risks, or to reduce the damage they caused to others than the immediate victims, such as those orphaned at a young age, may have declined in the industrialised world. However, developments in modern societies have increased people’s potential well-being and their expectations regarding their well-being. The result has been perceptions of risks of forms of damage which were previously not seen as serious or as avoidable. Thus there is far greater concern today than in the past about the risks of poverty as a result of unemployment, or of the loss, through failure to obtain an educational qualification, of the prospect of lucrative employment, or even of a loss of the material wealth which follows from profitable investment through a decline in the value of stock market investments. People are less directly dependent on the natural world for their survival and well-being,

Social security in the customary laws of immigrant communities

185

but more dependent on their accommodation within existing social processes. Partly because of this some speak of modern western society as ‘the risk society’,1 notwithstanding that an observer from the poorer parts of the world might wonder whether this emphasis was justified. In this risk society, must members look for protection only to state law? It is another common assumption that the complex, sophisticated and nationwide law of the modern state is the appropriate institution to protect against the risks which loom largest in that society, often through the legal devices discussed in other chapters in this volume. Only the state could have developed the social security systems which were created in the nineteenth and twentieth centuries. Hence the common understanding of social security is of a form of state protection against loss. Thus one standard definition of social security is: The protection which society provides for its members, through a series of public measures, against the economic and social distress that otherwise would be caused by the stoppage or substantial reduction of earning resulting from sickness, maternity, employment injury, unemployment, invalidity, old age and death; the provision of medical care; and the provision of subsidies for families with children.2 This summarises the conclusions of years of debate. It includes the heads of social security benefit set out in the Social Security (Minimum Standards) Convention 1952 of the ILO (ILO 102). But it may present an unduly narrow notion of social security law in several respects. Here it will suffice to note that the reference to the provision of social security by ‘society . . . through a series of public measures’ appears to be a reference to the law and institutions of the state and the ‘members’ of a society to mean the citizens of a state. Two grounds may be suggested for asking whether this notion of legal measures to reduce risk is not too narrow. First, it implies that, before the law of the modern state has been developed for a population, those people can have no social security law, nor indeed law in any field. This is contrary to the usage of many legal and social historians who study and write about law before the rise of the state, including legal norms directed to reducing risk. It is also contrary to the findings of anthropologists, who suggest that societies which have not been brought under the rule of state law in many areas of life nevertheless frequently have provided protection against risk. This protection is commonly afforded to an individual by the social group or groups to which he or she belongs, according to a system of norms which merits the name of law. Second, the notion that legal measures are always state measures implies that, when today a state law does exist for a population, any other normative orders which may be observed among them can be clearly distinguished from the law of the state. Thus, while there may be various forms of protection

186

Risk and the Law

against risk, there is a dividing line between state and non-state protection and only state law provides legal protection. Much of the writing about legal pluralism suggests that this is mistaken, for reasons which cannot be examined in depth here.3 In this chapter one set of examples of non-state social security is considered. This consists of customary norms of immigrant communities in the UK. The term ‘customary law’, used to refer to norms which emerge without formal legislation from the practices of a community, seems appropriate, provided that ‘customary’ is not taken to imply that the norms are of ancient origin, as required by the doctrines of some state laws. It is hoped to establish the possibility that these customary norms constitute currently operative protection against risk for many persons who would otherwise be especially vulnerable and who are not effectively protected by state law. It is suggested that members of these communities are often in a special position regarding both the risks to which they are subject and the modes in which they may be protected from these risks. Consequently the provisions made by state law to protect persons from risk may be less effective for them.

IMMIGRANT COMMUNITY LAWS IN THE UNITED KINGDOM The topic of risk and customary law in western states has been the subject of almost no research or debate. In much of what follows in the remainder of this chapter there will be discussion together of those immigrant groups about which information is available in the literature. This runs the risk of implying that all immigrant groups responded to their situation in the UK in similar ways. It is certain that this is not correct, since immigrants come from very diverse societies of origin. However, in the present state of research, if one wishes to go beyond reporting and analysing the circumstances of one small group, it is possible only to indicate the trends that are visible in those societies which have, to a certain extent, been studied, and to point out any similarities or contrasts which appear. There are some apparent similarities which suggest some general hypotheses, but all that follows must be qualified as highly tentative. It might be argued that a study of immigrant groups is unlikely to indicate significant general patterns of law because they are, almost by definition, exceptional, minority groups. Immigrant societies, in part because they are frequently excluded from the majority society, may be likely to develop and insist upon their own, distinctive laws. However, a study of these could provide insights which would throw light on the effectiveness of state laws. It is likely that there are customary laws within majority groups, coexisting with the state laws. This type of study might enable us more easily to distinguish these customary laws of majority groups from the state laws. This is a more difficult distinction to draw because state laws are frequently

Social security in the customary laws of immigrant communities

187

designed to meet the interests and aspirations of the majority and so do not differ markedly from the customary laws of that class. Research does suggest that in many immigrant groups conditions exist which are likely to produce their own customary laws. They frequently remain cohesive within their alien environments, the forces from which, rather than dissolving them, operate to stimulate greater cohesion. They frequently encounter discrimination, especially in housing and employment, which leads them to look for assistance within their own communities to overcome these obstacles. Norms which the groups generate or adopt define and regulate their internal relations and are effectively enforced. But, while anthropologists and sociologists have conducted research into immigrant societies, their research reports show that they have given little attention to their laws. For example, some literature on the laws of Roma (gypsy, traveller) communities seems to establish a persuasive case for the view that some and probably all such communities have laws.4 But in the leading work on these communities in Britain a search in the subject index for Law reveals as adjacent headings: ‘Language, Romany’; ‘Law, British’ – but not ‘Law, Romany’.5 Many works on recent immigrants focus primarily on issues of discrimination, studying attitudes and behaviour in the host community rather than the structures and practices of immigrants’ communities. This perhaps reflects that fact that the primary value promoted by liberals in the majority community seems often to be tolerance, apparently with the eventual objective of assimilation, rather than sustained recognition of immigrants’ normative orders. Nevertheless, it is possible to discover from the literature the tendency of immigrants to form their own communities. Britain has for many centuries contained substantial, distinctive immigrant communities, within which social order has been maintained by their own laws in the form of customary norms. In the course of generations some of these have merged with the larger society. This has occurred through processes of assimilation in which the immigrant community has adopted the social character and the law of the society surrounding them, although there has sometimes been some adoption of social characteristics in the other direction. However, some communities have tenaciously retained or even developed their own group identity over long periods. It seems likely that this tendency has become more marked in the recent past. The UK may today contain more cultural diversity (sometimes called multiculturalism) than ever before. We may note the course of immigration over the centuries. There is today little trace of the distinctive characteristics of the laws of some of the earliest immigrant communities of the distant past, such as the Saxons (from 450 CE onwards), the Danes (principally in the ninth century6) and the Normans (in the eleventh century7). This applies also to the Huguenot immigrants of more modern times. Their story arose from the religious conflicts following the reformation and the persecution of protestants in France, although the term came to be used

188

Risk and the Law

of all French-speaking protestant immigrants, including those from the low countries. The first large influx of such refugees occurred in 1562–1563, but the process continued at variable rates for nearly two centuries. After a period of abatement there was a brief revival of immigration in the 1620s, then a marked intensification in the 1670s, with a peak reached in the 1680s as a result of the events leading to and following from the Revocation of the Edict of Nantes, the law which had guaranteed certain freedoms to protestants, by Louis XIV in 1685. It has been estimated that between the late 1670s and 1710, 40,000–50,000 Huguenots came into England. Thereafter the rate greatly diminished, although there was a further, relatively slight phase of immigration in the late 1740s and early 1750s.8 In the early stages of their immigration Huguenots formed their own communities in the south-east and south-west of England, most notably in Canterbury. In the first and second generations they generally married within their own community and created, as far as they were allowed, their own church and work environment.9 The Privy Council was designated to concern itself with the government of this community and it adopted a policy of governing them through their own members and church organisation in an early version of the colonial policy of indirect rule.10 The organisation of their churches was elaborate and effectively regulated members’ lives.11 It was reported that by 1718 the Huguenots in London had 35 churches and had developed virtually a new town in the Spitalfields quarter.12 There was some hostility towards them. Their community rule requiring them to marry in their own churches gave rise to some conflict with English churches in the seventeenth century.13 In 1710 there were strong, publicly expressed complaints about their continued use of the French language and generally about the internal cohesion of their communities.14 Eventually, and partly because of this hostility, they looked towards and achieved integration into the wider community. The numbers of Huguenot churches declined. The records suggest that, while second-generation settlers often contracted marriages within the community, the third generation often did not.15 They became assimilated and indistinguishable from the bulk of the population. The process of assimilation would seem to have been aided by the fact that, because of the circumstances in which they left their earlier homes, they almost never maintained contact with these places of origin. This may not be the case with immigrant communities today. There have been many other groups which have settled in particular localities. For example, the merchants of Hansa, economic migrants, are said to have lived ‘in austere seclusion’ in Steelyard, the Hanseatic London trading depot, and to have developed ‘what would later be called extraterritorial rights, managing their affairs through their own aldermen and guildhall’.16 There have been many occasions when large numbers of immigrants have been induced to enter England, often on the encouragement of entrepreneurs seeking workers for particular projects, such as phases in the draining of the

Social security in the customary laws of immigrant communities

189

fens in East Anglia and Lincolnshire.17 It would appear that for periods of time they formed their own closed communities in England, but that eventually they became absorbed into wider English societies. Yet some long-established communities have retained their own laws. Roma/gypsies/travellers seem first to have arrived in the United Kingdom in the early fifteenth century.18 They have had a presence in the country ever since, although their itinerant lifestyle has presumably meant that many individuals have moved between the UK and other countries. An essay of 1827 not only evidences the antagonism towards them on the part of the English bourgeoisie but also suggests that, at least in the common view, they observed a law of their own. They are referred to as an ‘idle, profligate people, who infest all the countries of Europe and live in the midst of governments in a kind of commonwealth by themselves’.19 They appear always to have had a precisely defined kinship system.20 By the latter half of the eighteenth century a substantial community of black immigrants, apparently almost all from the Caribbean, lived in London. They met together regularly for social occasions. A number attended the Court of King’s Bench on the day in May 1772 when Lord Mansfield gave his judgment in Somerset v Stewart,21 the case brought to secure the release of a slave who had been brought to England from the Caribbean and had subsequently been imprisoned by his owner on a ship in order that he might be returned to Jamaica. A large number met immediately afterwards to celebrate the decision to order the release of Somerset.22 Jewish communities existed in England before the advent of Roma, but they were expelled from England in 1290 and not re-admitted until the last years of the Commonwealth and the early years of the Restoration of 1660. Thereafter immigrants most often joined already established Jewish communities. It has been said that they had their own ‘miniature republics, with councils and law courts of their own’.23 There are a number of works on the history of Jews in England which examine their law and its application in England.24 There is also a significant volume of law reports and other legal literature expounding and discussing cases in the state courts which have involved the recognition of Jewish law. Parts of this community have remained unintegrated with the rest of English society in some important respects. But its law has been accorded recognition, especially in respect of the rules governing family relations and religious observance. There have been small numbers of studies of other immigrant groups, sometimes giving slight indications of their communal systems of mutual support in the face of uncertainty. There has been Italian immigration over many centuries, at certain periods following a chain migration pattern (referred to below). A community has been maintained.25 Irish immigration has been considerable since the period of the Irish famine of the 1840s and gave rise to Irish quarters in some cities.26 German immigrants were the largest group after the Irish during most of the nineteenth century. They also were largely economic migrants.27 The Second World War brought a

190

Risk and the Law

significant influx of Polish immigrants. Studies suggest that they maintained a degree of social cohesion in the first generation, although there appears to have been a good deal of assimilation.28 Immigration in the period since the end of the war is distinctive. In the late 1940s immigration of large numbers from the English-speaking Caribbean countries began, encouraged both by the hopes of the immigrants for employment and by prospective employers seeking labour in conditions where other sources of cheap labour had dried up. Immigration from the Indian sub-continent was more limited in the early years, but increased greatly in the later 1960s, especially in the last few months before the Commonwealth Immigrants Act 1962 introduced controls. After the Act immigration continued at a high level through the issue of immigration vouchers and the admission of dependants of those already in the country. In the late 1960s and early 1970s numbers of East African Asians exercised the right which they still held by virtue of having British nationality and moved to Britain because of deteriorating conditions in the countries of East Africa. The total numbers have been raised by smaller numbers of immigrants from Cyprus, Hong Kong and other territories, coming largely for economic reasons, and from Iran, Somalia and other countries coming as refugees. Today many members of these communities are second- or third-generation immigrants, born in Britain. One result of the new form of immigration has been the emergence of relatively large communities adhering to religions which have not historically been followed in Europe. The largest of these are the Muslim communities. Although some Muslims have lived in England since the late eighteenth century at least, and distinct Muslim communities existed in that century in Liverpool and London, the numbers increased sharply after 1950.29 There are today in London, south-east England and Birmingham large, varied, cosmopolitan Muslim communities; elsewhere in England ‘the Muslims are almost synonymous with communities of Indian subcontinental origin’.30 The Hindu and Sikh religions are the others most strongly represented as a result of recent immigration. There is currently debate as to whether Rastafarianism should be characterised as a religion, but it seems that it has some of the characteristics of a religion with a system of law.31 These instances of immigration have occurred in different circumstances and have taken different forms from the earlier instances. One difference is that all such movements until recent times involved immigrants who were mainly urban dwellers in their countries of origin. Recent immigrants from the Caribbean, India, Pakistan, Bangladesh and Hong Kong have been more often rural dwellers.32 A possibly more significant difference is that many immigrants have been able to maintain close relations with their communities of origin. They can continue to participate in discussions and decision-making in those communities by electronic communication and by making regular visits. In some immigrant communities these opportunities are used a great deal.33 They are thus often effective members of

Social security in the customary laws of immigrant communities

191

communities in two continents. Many of them do not wish to adopt ways of life very different from those of their places of origin and so in the foreseeable future will choose to continue to belong to minority communities in England and successfully resist becoming assimilated into the majority community.34 At the same time ease of communication within Britain enables cohesion to be maintained between different groups of immigrants from a community of origin who, for economic reasons, have settled in different locations in Britain. As a result of these developments it seems likely that ethnic minority communities today are more compact societies, with a greater sense of identity than in the past. The result has been summarised by a leading scholar as the emergence of ‘self-created worlds’, constituting a ‘home from home’, in which the maintenance of family ties in particular has resulted in the ‘generation of a sense of communal solidarity’.35 Nevertheless, for these recent immigrant groups also there is considerable variation in the degrees of cohesion within communities.36 We have some evidence of the forms of social organisation of groups within immigrant communities. It seems clear that immigrant communities are most frequently delimited and delimit themselves in terms of their places of origin. This provides members with a group identity which they can readily recognise and accept, giving a sense of stability in a situation which can be threatening and disorienting. It also usually provides members with a body of legal norms which they find familiar, although this law often needs change and adaptation to be practicable in an alien environment. Within immigrant communities the most important group appears usually to be the kin group, although membership of a significant group may also be determined, in conjunction with or independently of kin relations, by original residence in a common locality (the ‘village’ community), allegiance to a common leader, nationality, religious adherence, or some other characteristic. The kin group is normally a larger group than that which is thought of as a ‘family’ in the west today. It includes persons whose relationship to the other members is relatively remote and sometimes even fictitious. Further, in the case where there are few who are closely related, it is even further expanded by the inclusion of persons with yet more remote or more doubtful relationships. The composition of the kin or other group varies between different societies, but the membership of each observes a body of norms which enables them systematically to perceive who are rated as their co-members and then equally systematically impose duties of mutual assistance on them. These norms are observed, and effectively structure social life. It is difficult to find any characteristic distinction between these enforced and observed norms and the norms of state law. Even in respect of more recent immigrants there has been relatively little research into their law. Nevertheless, lawyers who reflect on the legal issues specific to these groups in British society have noted that the law of the British state falls to be applied within the context of a functioning customary law. Thus Pearl, writing from an extensive

192

Risk and the Law

experience, noted at early as 1986 that ‘for most South-Asian families living in the UK, the nuclear family . . . only exists in the context of a very complex network of obligations, rights and duties, extending outside to the wider and more important group. Often this group includes those still in the home country.’37

SOCIAL SECURITY IN THE CUSTOMARY LAWS OF IMMIGRANT COMMUNITIES There is a good deal of evidence that in many immigrant communities communal norms give protection against the otherwise considerable risks inherent in the initial stages of immigration. There is less evidence as to the continuation of communal social security after that stage, although it seems to continue strongly in some communities.

Initial immigration and settlement The initial process of immigration into and settlement in the new society is necessarily a risky process, given the considerable difference in cultures which often exists between the place of origin and the host community, and the lack of material resources of most immigrants on their arrival. Migrants in modern times have rarely come from the most destitute sections of their societies, but they still have needed help to make the journey possible and moderately safe. The process is significantly eased for the new immigrant by assistance from other members of the community of origin, including both those still in the place of origin and those who have already migrated to the UK. This is noticeable in the case of ‘chain migration’, whereby a small community in the place of origin collectively finances the immigration of one man to the UK and he, once settled and earning an income, finances one or more further immigrants from the village, who in turn assist more of their relatives or fellow villagers to come. It has been stated of Pakistani migrants that ‘few left [home] without the active support of their extended families’ and that the ‘vast majority [arrived] in cascading chains along increasingly well-worn paths of kinship and friendship’.38 Many other accounts have been given of chain migration.39 Whether or not immigration is enabled in this fashion, it is not uncommon in the early period of migration from a particular district for migrants to construct a scheme of reciprocal obligations among each other, effectively extending, with some adjustments, the law of their home community to the immigrant community.40 Arrangements made or provided at the moment of arrival merge into continuing relations of mutual dependency in some communities, reducing the risk of unemployment, or at least being perceived as reducing the risk.

Social security in the customary laws of immigrant communities

193

Thus Ballard and Ballard write about networks, relatively organised systems, for assisting new immigrants. They mention the case of a wholesaler who built up his business by supplying fellow immigrants as retailers to whom he extended credit and gave other assistance.41 It is common for members of a particular immigrant community to enter the employment of a member of the same community42 and for assistance to be given to newcomers in the intricacies of British bureaucracy, with assistance in finding jobs, completing official forms, finding accommodation and general welfare.43

Mutual support systems subsequent to settlement under immigrants’ customary laws It has been noted that research on immigrant communities gives little attention to their laws. Much of that which has been done has focused on that law which most commonly receives observance, namely that defining and regulating kinship and marriage.44 Even less has been devoted to investigating their laws providing protection to members against risk. How far does there appear to be observance of such norms? It appears that within such groups there has in the past often been, and is often today, a recognised, often expressly reiterated, duty of members to give assistance to the fellow member who is at risk of loss. This duty is effectively enforced, partly by principles of reciprocity, partly by the threat of sanctions, including in extreme cases exclusion from all the benefits of membership of those who disregard it. Literature concerning instances of migration shows that, for example, the Huguenot community operated a developed system of internal poor relief.45 Gypsies, perhaps because of their well-defined kinship system, seem long to have observed obligations of mutual assistance.46 Immigrants of CaribbeanAfrican descent gave assistance on occasion to members of their community in need.47 German immigrants of the nineteenth century maintained philanthropic organisations within their communities to provide assistance to their poorer members.48 However, in some communities there has been relatively little communal protection against risk. In the Italian immigrant community the mutual aid where given seems to have been often to secure employment for each other but not otherwise to assist those in difficulty, although there are records of communal assistance to some members who needed to be repatriated to Italy.49 The studies of Polish immigrants from the Second World War do not suggest that the communities performed any social security functions.50 The existence of the welfare state in the UK removes the need for a few measures which might otherwise be taken. Thus it has been suggested that, whereas Montserratian immigrants in New York and Boston, USA, operated a formally constituted mutual aid society which provided originally funeral costs and sickness and unemployment insurance, ‘[t]he fact that the British “welfare state” covers many of these functions undoubtedly impeded the development of voluntary associations’.

194

Risk and the Law

Ballard finds that in more recent instances ‘networks of reciprocity’ were set up, following from the initial assistance which enabled the immigration to take place. He finds that the groupings are based on ‘parochial’ allegiances, no more extensive than caste, sect and descent-group,51 and he has warned against assuming that ‘Asian’ immigrants constitute ‘a single homogeneous entity, even in the context of a single restricted locality.52 Ballard and Ballard record the practice in a Sikh community whereby established immigrants purchased housing, which they then rented out to newer arrivals, establishing patron–client relationships which had a long-term normative effect, based in part on strongly held, socially enforced notions of izzat (personal honour).53 Ballard found that in the same community an immigrant’s first employment or the initial capital to set up in business was commonly provided by more senior members of that immigrant community.54 In the context of a community governed by a substantial body of norms, this may be seen as giving rise to regular legal relationships of patronage, even if the initial assistance arises from generosity rather than obligation. Ballard recorded that there might be competition between patrons to gather the greater following of clients, which suggested a normatively structured social system creating power relations. Watson found that in a Chinese community there were less intimate or binding connections, although here also businesses had been built up through community connections, especially kin.55 In this case there was strong social bonding within the Chinese community settled in the UK and also their relatives in their place of origin, but this was based largely on kinship. Voluntary associations had failed to remain in existence unless their membership had more in common than their place of origin.56 Reciprocal gift-giving practices are sometimes taken over from the home community, and systems of patronage are created in the new environment, both of which provide some degree of social security to the vulnerable.57 Practices such as rotating credit associations (‘boxes’ in Montserrat) have been established.58 In the Montserrat case the obligations of members of a ‘box’ to keep up payments were enforced by the ‘implicit sanction . . . that the news of such a misdemeanour would be quickly communicated among Montserratians in London and also back to the home island. The offender would be given a bad name which could adversely affect his future relations with other islanders’.59 Where employment from another member of the immigrant community had been initially obtained, social opinion within that community then might tend to maintain the terms of the relationship and to prevent either party from having recourse to the employment law of the state.60 Werbner found an established practice among Pakistani immigrants living in Manchester of loans for business or emergencies. The groupings were of friends rather than of relatives. The normative basis for the system seems to have been clear: a request for a loan could not be refused, the terms of a loan when made were determined by regular rules, and there were no known

Social security in the customary laws of immigrant communities

195

cases of defaults in repayment. She also found a ‘customary system of exchange’ of gifts, and rotating credit societies. A strong case for the existence of a customary law seems to be made. However, the primary ground for all these transactions seems to have been the advancement of business interests rather than assistance in emergencies. This appears to be a customary law for mutual advancement first, and only secondarily a customary law for the reduction of risk.61 It might be supposed that common residence in a particular locality could serve to hold a community together. This may well be so, and it has been found that some immigrant groups do tend to settle initially in specific localities.62 However, subsequently, as members of the community prosper, they may move their places of residence, typically from inner-city districts to the more prosperous suburbs. This does not necessarily weaken in any degree the networks of hospitality and mutual assistance which have been constructed.63 Chinese immigrants, however, have generally not tended to settle in particular localities.64 Many voluntary associations have been formed by minority ethnic groups to provide for mutual assistance. These have in most cases had a formal status within state law, such as that of a registered charity.65 It seems unnecessary to consider these in the present context, although we may speculate that the customary laws of the groups concerned may influence the manner in which such an association functions. Many immigrants remit a high proportion of their incomes to their homes. Often, if they wish to return eventually to live there, they are obliged to do this by social opinion in their home communities.66 This tends to bind together members of immigrant communities and inhabitants of immigrants’ places of origin, both transnationally and within individual states, so enhancing the effectiveness of any customary norms of social security.

CONCLUSION: AN ALTERNATIVE, NON-INDIVIDUALISTIC LAW OF SOCIAL SECURITY Thus there are significant normative orders specific to immigrant communities in the UK, which give their members some degree of protection against various risks. These are largely risks which in nature or degree arise from the special circumstances of immigrant groups, such as the risks of homelessness or unemployment. Frequently these communities’ customary laws make good in some measure their exclusion from protections which are afforded by the state to members of the majority and which are unavailable to those who lack nationality, linguistic ability or simply familiarity with the ways of the country. While state social security is directed mainly at individuals’ needs, the customary laws of immigrant groups are communal, operating through networks of reciprocity in which all members are linked. There is a primacy of communitarian concepts of the individual as a member

196 Risk and the Law of a community of this type, not primarily as a subject or citizen of a state. Individualisation is not a feature of this customary protection against risk. The foregoing generalisations probably do not hold good for all categories of immigrant communities. More specific generalisation is subject to even stronger qualification. We shall not know much about immigrant communities until far more research has been done (and it may be remarked that anthropological methods are very likely to be the most productive of useful research). As far as one can discern from the literature currently available, a few more, tentative suggestions may be attempted. It would seem that the effective communities for the purposes of mitigating risk to immigrants are quite small. Their membership is not entirely static, for history shows that over time members tend to become assimilated into the majority society – although there are grounds for predicting that this is happening to a lesser extent with the immigrant groups from outside Europe of the past halfcentury. Even within these small groups the intensity of commitment varies from one individual or household to the next. The customary laws of these communities are linked to those of their communities of origin, but are by no means reproductions or transplants of them. Immigrants are of necessity highly adaptive. Moreover, they adapt their customary laws not only to the circumstances and ways of the society they meet on their arrival but also subsequently to later changes in their social environment in the UK (and quite possibly as a consequence of later changes in the customary practices of their places of origin, with which they remain constantly in contact). Knowledge of these issues is relevant to the formulation of state legal policy. It is necessary to examine relatively neglected issues as to the relationship between the state and the customary laws of minority groups. This includes the design of state laws to take account of these other customary laws, including the forms of ‘recognition’ which the state may accord to such laws. The law of social security is one of the fields in which the state needs to recognise the challenge of the cultural diversity and hence the legal pluralism now flourishing in the UK. The policies which state law is to adopt as a result remain open for research and debate. There is also a need to consider issues concerning the evaluation of customary laws against principles of general morality or human rights, but making the greatest possible effort to avoid ethnocentric bias and remembering that exactly the same needs to be done with respect to state law. It is not argued here that the state should necessarily leave communities with their own systems outside the scope of state protection, and certainly not that the state should attempt to suppress or supplant these customary systems. But what does seem clear is that there needs to be an awareness of the true social situation when the state system is planned. Immigrant communities seem clearly significant in this respect, but these are not the only minorities with their own customary laws. It is not possible here to investigate other groups which, unable to benefit from the norms and

Social security in the customary laws of immigrant communities

197

institutions of state law, have formed their own communities with their own, unofficial laws.67 It may merely be noted that this observation strengthens the primary proposition put here, that law is not merely the law of the state, and that there are other, customary laws which should always be studied in any investigation of a particular field of law and society. This chapter has attempted to give a brief indication of some ways in which studies of risk and the law might be so broadened.

Notes 1 E.g. U. Beck, Risk Society: Towards a New Modernity, London: Sage, 1992. 2 International Labour Organisation, Introduction to Social Security, Geneva: ILO, 1984, pp. 2–3. 3 See e.g. J. Griffiths, ‘What is Legal Pluralism?’, Journal of Legal Pluralism, 24, 1984, 1; J. Vanderlinden, ‘Return to legal pluralism: twenty years later’, Journal of Legal Pluralism, 28, 1989, 149; J. Vanderlinden, ‘Vers une nouvelle conception du pluralisme juridique’, Revue de la recherche juridique – droit prospectif, XVIII, 1993, 573; G.R. Woodman, ‘Ideological combat and social observation: recent debate about legal pluralism’, Journal of Legal Pluralism, 42, 1998, 21; K. von Benda-Beckmann, ‘Legal Pluralism’, in Tai Culture, VI, 2001, (O. Aaendchen (ed.), 18; J. Griffiths, ‘Legal Pluralism’, in N.J. Smelser and P.B. Baltes (eds.), International Encyclopedia of the Social and Behavioral Sciences, Oxford: Pergamon, 2001, p. 8650; A. Griffiths, ‘Legal Pluralism’, in R. Banakar and M. Travers (eds.), An Introduction to Law and Social Theory, Oregon: Hart Publishing, 2002, p. 289. 4 W.O. Weyrauch (ed.), Gypsy Law: Romani Legal Traditions and Culture, Berkeley: University of California Press, 2002. See also: T. Acton, S. Caffrey and G. Mundy, ‘Theorizing Gypsy Law’, American Journal of Comparative Law, 45, 1997, 237; T. Acton, ‘A three-cornered choice: structural consequences of value-priorities in Gypsy Law as a model for a more general understanding of variations in the administration of justice’, American Journal of Comparative Law, 51, 2003, 639. 5 J. Okely, The Traveller-Gypsies, Cambridge: Cambridge University Press, 1983. 6 F.M. Stenton, Anglo-Saxon England, 3rd edn, Oxford: Oxford University Press, 1971. 7 W.S. Holdsworth, A History of English Law, Vol. II, 3rd edn, London: Methuen & Co., 1923, p. 12. 8 R.D. Gwynn, Huguenot Heritage: The History and Contribution of the Huguenots in Britain, London: Routledge and Kegan Paul, 1985, pp. xi, 5, 26, 30. 9 Gwynn, ibid, p. 53. 10 Gwynn, ibid, p. 43. 11 Gwynn, ibid, p. 98. 12 V.G. Kiernan, ‘Britons Old and New’, in C. Holmes (ed.), Immigrants and Minorities in British Society, London: George Allen & Unwin, 1978, pp. 23, 38–39. 13 Gwynn, n. 8 above, pp. 163–164. 14 Gwynn, n. 8 above, pp. 167–168. 15 Gwynn, n. 8 above, p. 162. 16 Kiernan, n. 12 above, p. 27; see also H. Kellenbenz, ‘German immigrants in England’, in C. Holmes (ed.), n. 12 above, 63. 17 Kiernan, n. 12 above, pp. 33–34, passim.

198

Risk and the Law

18 Kiernan, n. 12 above, p. 31. 19 Sir Roger de Coverley, The Spectator, Vol. 2 (1827) p. 319, quoted in C. Holmes, ‘The German Gypsy question in Britain, 1904–1906’, in K. Lunn (ed.), Hosts, Immigrants and Minorities: Historical Responses to Newcomers in British Society 1870–1914, Folkestone: Wm Dawson & Sons, 1980, pp. 134, 136. 20 Okely n. 5 above, pp. 170–200. See also: Acton et al., n. 4 above; Acton, n. 4 above. Other writings which include some information on Roma in England are: J.P. Clebert, The Gypsies (trans. Charles Duff), London: Vista Books, 1963; D. Kenrick and S. Bakewell, On the Verge: the Gypsies of England, London: Runnymede Trust, 1990; A. Fraser, The Gypsies, Oxford: Blackwell, 1995; S. Poulter, Ethnicity, Law and Human Rights: The English Experience, Oxford: Clarendon Press, 1998, Chap. 5. These contain historical information, but almost no discussion of Roma laws. 21 (1772) 98 ER 499. 22 P. Fryer, Staying Power: The History of Black People in Britain, London: Pluto Press, 1984, Chap. 4. 23 Kiernan, n. 12 above, p. 26. 24 See especially: A. Hyamson, A History of the Jews in England, London, 1908; H.S.Q. Henriques, The Jews and the English Law, Oxford: Oxford University Press, 1908; C. Roth, A History of the Jews in England, Oxford: Clarendon, 1941; M. Freedman (ed.), A Minority in Britain: Social Studies of the AngloJewish Community, London: Vallentine, Mitchell & Co, 1955; L.P. Gartner, The Jewish Immigrant in England 1870–1914, 2nd edn, London: Simon Publications, 1976. 25 R. Palmer, ‘The Italians: Patterns of Migration to London’, in J.L. Watson (ed.), Between Two Cultures: Migrants and Minorities in Britain, Oxford: Basil Blackwell, 1977, p. 242. 26 E.g. F. Finnegan, Poverty and Prejudice: A Study of Irish Immigrants in York 1840–1875, Cork: Cork University Press, 1982. 27 P. Panayi, German Immigrants in Britain During the 19th Century, 1815–1914, Oxford: Berg Publishers, 1995, Chap. 4: ‘Ethnic organisations: religion, philanthropy, culture and politics’. 28 S. Patterson, ‘The Poles: an exile community in Britain’, in Watson (ed.), n. 25 above, p. 214. 29 J.S. Nielsen, Muslims in Western Europe, 2nd edn, Edinburgh: Edinburgh University Press, 1995, pp. 4–6. Chap. 4 in this work contains a detailed account of the immigration of Moslems into the UK. 30 Nielsen, ibid., p. 43. 31 T. Modood (ed.), Church, State and Religious Minorities, London: Policy Studies Institute, 1997; T. Modood et al., Ethnic Minorities in Britain: Diversity and Disadvantage, Fourth National Survey of Ethnic Minorities, London: Policy Studies Institute, 1997; Poulter, n. 20 above. 32 C. Holmes and K. Lunn, ‘Introduction’, in Lunn (ed.), n. 19 above, 2. On Hong Kong, see also J.L. Watson, ‘The Chinese: Hong Kong Villagers in the British Catering Trade’, in Watson (ed.), n. 25 above, 181. 33 Modood et al., n. 31 above, p. 313 stated that nearly half the members of all the ethnic minority groups studied had visited their family’s country of origin in the previous five years, many more than once. 34 R. Ballard, ‘Preface’, in R. Ballard (ed.), Desh Pardesh: the South Asian presence in Britain, London: Hurst, 1994, vii, p. ix. 35 Ballard, ibid., p. vii. The expression Desh Pardesh is explained to mean both ‘home from home’ and ‘at home abroad’ : R. Ballard, ‘Introduction: The Emergence of Desh Pardesh’, in Ballard (ed.), n. 34 above, 5.

Social security in the customary laws of immigrant communities

199

36 Cf. above cases and also: G. Dench, Maltese in London: A Case Study in the Erosion of Ethnic Consciousness, London and Boston: Routledge and Kegan Paul, 1975, pp. 34–46; F. Anthias, Ethnicity, Class, Gender and Migration: Greek Cypriots in Britain, Aldershot: Avebury, 1992, p. 113. 37 D. Pearl, Family Law and the Immigrant Communities, Guide and Practice Series, Bristol: Family Law, Jordan & Sons, 1986, p. 10. 38 Ballard, n. 34 above, pp. 11–13. 39 B. Dahya, ‘Pakistanis in Britain, transients or settlers’, Race, 14, 1973, 246; B. Dahya, ‘The Nature of Pakistani Ethnicity in Industrial Cities in Britain’, in A. Cohen, (ed.), Urban Ethnicity, London: Tavistock, 1974, pp. 77–118; S.B. Philpott, ‘The Montserratians: Migration Dependency and the Maintenance of Island ties in England’, in Watson (ed.), n. 25 above, pp. 90, 104–105; N. Foner, ‘The Jamaicans: Cultural and Social Change among Migrants in Britain’, in Watson (ed.), n. 25, pp. 120, 125–126; Watson, n. 32 above, pp. 189–191; R. Palmer, ‘The Italians: Patterns of Migration to London’, in Watson (ed.), n. 25 above, pp. 242, 251–255, stating that this practice was followed by Italians migrating to London in the mid-19th century; A. Shaw, A Pakistani Community in Britain, Oxford: Basil Blackwell, 1988, pp. 22–25; A. Shaw, ‘The Pakistani community in Oxford’, in Ballard (ed.), n. 34 above, pp. 35, 39–40; K. Gardner and A. Shukur, ‘ “I’m Bengali, I’m Asian, and I’m living here”. The Changing Identity of British Bengalis’, in Ballard (ed.), n. 34 above, pp. 146–148, 151. 40 R. Ballard and C. Ballard, ‘The Sikhs: The Development of South Asian Settlements in Britain’, in Watson (ed.), n. 25 above, pp. 21, 28–31. 41 Ballard and Ballard, n. 40 above, pp. 23–24 (noting similar processes in the 19th century), 30–31, 35–38; V.F. Kahn, ‘The Pakistanis: Mirpuri Villagers at Home and in Bradford’, in Watson (ed.), n. 25 above, pp. 57, 67, 77; Philpott, n. 39 above, pp. 108–109; Gardner and Shukur, n. 39 above, pp. 146–148, 151. 42 Anthias, n. 36 above, pp. 53–66, 103–105, on Greek Cypriots in Britain. 43 Philpott, n. 39 above, pp. 108–109; A. Shaw (1994), n. 39 above, pp. 40–41; MA. Pérouse de Montelos, ‘A Refugee Diaspora: When the Somalis Go West’, in K. Koser (ed.), New African Diasporas, London: Routledge, 2003, p. 37 (on ‘short-term needs’). 44 See e.g.: Pearl, n. 37 above; R. Burghard (ed.), Hinduism in Great Britain: The Perpetuation of Religion in an Alien Cultural Milieu, London: Tavistock, 1987; W. Menski, ‘Asians in Britain and the question of adaptation to a new legal order: Asian laws in Britain?’, in M. Israel and N.K. Wagle (eds.), Ethnicity, Identity, Migration: The South Asian Context, Toronto: Centre for South Asian Studies, University of Toronto, 1993, p. 238. 45 Gwynn, n. 8 above, pp. 107–108. 46 Okely, n. 5 above, pp. 170–200. 47 Fryer, n. 22 above, p. 70 notes newspaper reports of 1773 that, when two black men were imprisoned in London for begging, more than 300 other members of this community visited them and ‘contributed largely towards their support during their confinement’, and notes other evidence of members’ provision of mutual support. 48 P. Panayi, n. 27 above, Chap. 4: ‘Ethnic organisations: religion, philanthropy, culture and politics’. 49 Palmer, n. 25 above, p. 242. 50 Patterson, n. 28 above, p. 214. 51 Ballard, n. 35 above, p. 4. 52 R. Ballard, ‘Common law and uncommon sense: the assessment of “reasonable behaviour” in a plural society’, www.arts.manchester.ac.uk/casas/papers/ pdfpapers/commonsense.pdf, last accessed 1 November 2007. This publication contains, in a section entitled ‘Britain as a plural society’ a summary of Ballard’s

200

53 54 55 56 57 58 59 60 61 62 63 64 65

66 67

Risk and the Law findings as to the salient features of ‘the processes of ethnic colonisation in which South Asian settlers in Britain have been engaged’. Ballard and Ballard, n. 40 above, pp. 31–33. R. Ballard, ‘Differentiation and Disjunction among the Sikhs’, in Ballard (ed.), n. 34 above, pp. 88, 104–106. Watson, n. 32 above, pp. 192–193. Watson, n. 32 above, pp. 196–200. Shaw, n. 39 above (1988), pp. 111, 134–155; Watson, n. 32 above, pp. 203– 204. Philpott, n. 39 above, pp. 110–111. For a more general survey, see A. Herbert and E. Kempson, Credit Use and Ethnic Minorities, London: Policy Studies Institute, 1996, pp. 30–34, 57–61, 84–85. Philpott, n. 39 above, p. 111. Anthias, n. 36 above, pp. 53–66, 103–105, on Greek Cypriots in Britain. P. Werbner, ‘From Commodities to Gifts: Pakistani Migrant Workers in Manchester’, in A. Rogers and S. Vertovec (eds.), The Urban Context: Ethnicity, Social Networks and Situational Analysis, Oxford: Berg, 1995, p. 213. Philpott, n. 39 above, p. 108. Ballard and Ballard, n. 40 above, pp. 42–43. Watson, n. 32 above, pp. 181–182. M. McLeod, D. Owen and C. Kharmis, Black and Minority Ethnic Voluntary and Community Organisations: Their role and future development in England and Wales, London: Policy Studies Institute, 2001, especially Chap. 2: ‘Black and Minority Ethnic Organisations and Their Role in the Community’. However, this work does not consider organisations with a religious base. McLeod, Owen and Kharmis, ibid., pp. 112–116; Watson, n. 32 above, pp. 207–209. See e.g. the study of the customary laws of Brazilian favelas in B. de S. Santos, ‘The law of the oppressed: the construction and reproduction of legality in Pasargada’, Law & Society Review, 12, 1977, 5, contained in a greatly revised form in B. de S. Santos, Toward a New Legal Common Sense: Law, Globalization, and Emancipation, 2nd edn, London: Butterworths, 2002, Chap. 4.

Index

accident and risk, difference between 14 administrative law see German constitutional and administrative law affirmative protection duties 46–8 arrival-time clause in CIF contracts 140–3, 144 assumption of risk 85–94, 177, 179 biodiversity 67 black immigrants in United Kingdom 189, 190, 193 border controls 38–41 breach of statutory duty 178 BSE 53–4, 58–9, 62 C&F (cost and freight) 132, 134, 136, 138–44 carriage risks see CIF contracts, transit risks in Cartagena Protocol on Biosafety 67 causation, law and social policy 173–4 certainty 95–6, 98–100, 103–4, 108 CIF (cost, insurance, freight) contracts, transit risks in 132–45 arrival-time clause 140–3, 144 C&F (cost and freight) 132, 134, 136, 138–44 cost of carriage, risk of increase in 133–4, 143 delay from risk of loss or damage, distinguishing 141–3 denying effect to out-turn clause 135–6 deterioration, risk of 134–5, 143 giving effect to both out-turn clause and CIF/C&F label 136–40, 144 Incoterms 133–4

inherent nature of goods, limiting the effect of out-turn clause to 139–40 interpretation of clause as relating to shipment 141 loss and damage, risk of 132, 137–8, 141–4 out-turn clause 135–40, 144 partial loss, limiting the effect of out-turn clause to 138–9 splitting the risk of loss 138 civil rights, encroachment of risk on 28–9, 31–3, 45–9 collectivism 8–10, 46 conditional fee agreements 152–3 consent 4, 87–92 constitutional law see German constitutional and administrative law contract 4, 106–7, 110–11, 115–27, 179 see also CIF contracts, transit risks in contributory fault 88, 91 corporate criminal liability 79–80 cost of carriage, risk of increase in 133–4, 143 costs in English law 4, 146–57 Calderbank letters 155 Civil Procedure Rules 146–7, 153–5 discretion 150–1 ‘front-end loading’ of costs 153–4 funding of litigation, recent changes in 152–3 indemnity basis 149–50 orders for costs 148–50 Part 36 payments and offers 154–5 pre–action protocols 153–4 proportionality 146–7, 155 settlement, encouragement of 153–4 standard basis 149

202

Index

costs in English law – Contd. summary assessment 151 taxation of costs 149, 151 terminology 147–8 Woolf reforms 146–7, 151–5 crime, risk and 75–81 cross-border crime 36–41, 48 enterprises, liability of 79–80 environment 76–7 falls in crime rates 13 Germany 75–81 harm 76 imputation in case of reasonable risk, exclusion of 77–8 legislator, response of 75–7 negligence 77–9 white–collar crime 76–7 customary laws of immigrant communities, social security in 184–200 assimilation 190–1 chain migration 192 countries of origin, close links with 190–1, 195, 196 discrimination 187 gift-giving practices 194–5 initial immigration and settlement 192–3 kin groups 191–3 mutual support systems after settlement 192–5 United Kingdom 186–92 voluntary associations 193, 195 damages 122–3, 153 danger German constitutional and administrative law 24–9, 32–3 hazards, distinguished from 10–11, 55–6 nuisance 25–6 prediction 36 probability of damage 25 suspected danger 26–7 decision-making, harm resulting from one’s own 11 deep pockets theory 100–2 definition of risk 10–13, 23–7 de-individualisation in the risk society 18, 36–52 affirmative protection duties 46–8 assessment of human behaviour 37–40, 44

civil rights, encroachment of risk on 28–9, 31–3, 45–9 constitutional protection against de-individualisation of regulation 45–8 cross-border crime 36–41, 48 Germany 36–52 individual as source of risk 36–7, 47 police 37–43, 45, 48 predictability of danger 36 Schengen Agreement, border controls and 38–41 security laws 36–7, 48–9 social data in grid searches, use of 42–5 social risks, regulation of 37–44 terrorism 36–7, 41–4, 47–8 delay in carriage, risk of 141–3 deterioration, risk of 134–5, 143 duties of care 88, 91–2 EC law BSE 53–4, 58–9, 62 employer’s liability 175–6, 178, 179 food law 53–63, 75 free movement of goods 28 genetically modified organisms (GMOs), dispute with EC over 60, 66–7 precautionary principle 53–63 economic loss 109–11 employers’ liability in English law 7, 172–84 assumption of risk 177, 179 breach of statutory duty 178 burden of proof 174–5 causation, law and social policy 173–4 codes of practice and guidance notes 177 common risks 181 contract law 179 EC law 175–6, 178, 179 equipment, defective 177 hazards 11, 172–3 Health and Safety at Work Act 1974 174–5 improvement and prohibition notices 176 Industrial Injuries Advisory Council 182 injury, harm, guilt and liability 173 insurance 177–8 negligence 177–8

Index pre–existing conditions 181 prescribed diseases 181–2 prosecutions 173, 174 reasonably practicable, meaning of 174–5 regulation 174–7 risk assessments 176–7 sentencing 174, 176 social security law 172, 180–1 unfair contract terms 177, 179 working time 179 employment see employers’ liability in English law, risk and; income risk of employed persons in Germany English law see also costs in English law; employer’s’ liability in English law certainty and justice 95–6, 98–9, 103–5, 108 common law 95–113 consistency of outcome 108 contract law 106, 107, 121–3 customary laws of immigrant communities, social security in 184–200 damages 122–3 deep pockets theory 100–2 economic loss 109–11 floodgates argument 109 harassment 97–8 Human Rights Act 1998 97–8 identifying risk, issues in 107–8 insurance 100–4, 109 legal positivism 96–9, 106, 108 micro and macro 104–5 negligence 99, 105, 109–11 nuisance 97–9, 105–7 objective and subjective risk 103–4 proximity 109–10 restitution 122–3 risk and predictability in common law 95–113 television reception, loss of 97–8 transfer of property and risk in 121–3 enterprises, liability of 79–80 environment 10, 18, 25, 29–30, 32, 76–7 exclusion of liability 91 fault–based liability 5–8, 79, 88, 91, 114, 127, 164–7 floodgates argument 109 FOB (free on board) carriage 132, 133–4 food law 53–63, 75 freedom of assembly 27–8

203

French law, transfer of property and risk in 114–21, 128–9 carriage risks 117 loss before delivery 116 res perit domino 114–16, 118–19 reservation of title 117–18 valid contracts and void contract 116–21 genetically modified organisms (GMOs) 60, 66–7 German constitutional and administrative law abortion 29–30 change of instruments 31 constitutional framework of risk decisions 27–31 danger 24–9, 32–3 definition of risk 23–7 delegation of risk decisions 32–3 freedom of assembly 27–8 inherent residual risk 27 instruments, change of 31 law-making process, risk decisions in 31–2 nuisance 25–6 political theory 27–8 precautionary principle 27, 30–1 prevention of risk 28–9 probability of danger 29 prognosis problem 31–2 proportionality 29 protection, state’s duty of 29–30 recommendations, warnings and informations, change to 31 risk decisions 23–35 rule of law 28–9 Schutzpflichten, discovery of 29–31 social risks, regulation of 37–44 statehood, liberal concept of 23–6 German law see also German constitutional and administrative law; income risk of employed persons in Germany assumption of risk 85–94 civil rights, encroachment of risk on 28–9, 31–3, 45–9 contract 123–7 criminal law 75–81 de-individualisation 36–52 fault-based liability 5–6 food law 57–8 police 37–43, 45, 48 res perit debitori 123

204

Index

German law – Contd. reservation of title 124 restitution 125–7 security 41–4 sickness in labour and social insurance law, risk of 161–71 social risks in administrative law, regulation of 37–44 subjective and objective risk 15–16 terrorism 41–4, 47–8 tort law 15–16 transfer of property and risk in 123–7, 128–9 transport risks 123–4 unjust enrichment 124–7 valid and void contracts 123–5 harassment 97–8 harm 7–13, 14, 76 hazards dangers, distinguished from 10–11, 55–6 employer’s liability 172–3 Health and Safety at Risk Act 1974 174–5 health insurance 167–8, 169 hormone case 26, 55, 64–6 Huguenot communities in United Kingdom 187–8 Human Rights Act 1998 97–8 immigrant communities, social security and customary laws of 184–200 improvement and prohibition notices 176 income risk of employed persons in Germany 161–70 continued payment of remuneration 162–5, 167–9 distribution of risk 162–9 fault of employee 164–7 health insurance 163, 167–70 distribution of risk between employer and statutory insurance 167–8 insurance fund, relief of 169 long-term incapacity 165 maintenance of income 168–9 remuneration, claims for 162 risk bearing by employees 166–7 sickness benefit 165–6, 168–70 sports injuries 164 welfare of employee, employer’s duty to provide for 168

Incoterms 133–4 Indian sub-continent and East Africa, immigrants from 190, 192, 194–5 individuals de-individualisation in Germany 36–52 fault-based liability 8 harm principle 8 social risks in German administrative law, regulation of 37–44 industrial diseases 181–2 Industrial Injuries Advisory Council 182 information, lack of 17–18, 32, 36 insurance 100–4, 109, 152–3, 167–70, 177–8 see also CIF (cost, insurance, freight) contracts, transit risks in Islamic terrorism 42–4 Jewish communities in United Kingdom 189 justice and certainty 95–6, 98–9, 103–5, 108 justified risks 7–8, 17–18 kin groups 191–3 law-making process, risk decisions in 31–2 legal funding 152–3 legal positivism 96–9, 106, 108 liberal concept of statehood 23–6, 29 life, protection of 47–8 Living Modified Organisms (LMOs) 67 loss or damage, risk of 5, 109–10, 114–16, 132, 137–8, 141–3 market opportunities 10–11 Marxism 98 medical treatment informed consent in 86, 89, 90–1 objective risk 16 Muslims 42–4, 190 natural law theory 98 negligence 77–9, 88, 91–2, 99, 105, 109–11, 118, 177–8 no-fault liability 6, 100 nuisance 25–6, 97–9, 105–7 objective and subjective approach to risk 13–14, 15–17, 86–91, 103–4 out-turn clause in carriage contracts 135–40, 144

Index Part 36 payments and offers 154–5 perceptions and reality of risk 13–15 physical injury 8 police 24, 26, 37–43, 45, 48 precautionary principle 17–18 additives and novel foods 59–60 BSE 53–4, 58–9, 62 Cartagena Protocol on Biosafety 67 EC food law 53–63, 75 General Food Law Regulation 178/ 2002, text of 73–4 genetically modified organisms (GMOs) 60, 66–7 German constitutional and administrative law 27, 30–1 German food law 57–8 hormone case 55, 64–6 meaning 54–5 security law, within 55–7 World Trade Organization law 64–7 prediction 36, 95–113 prescribed diseases 181–2 probability 11–13, 14, 16–17, 25, 29–30 prognosis problem 31–2 proportionality 29, 59, 68, 146–7, 155 protection, state’s duty of 29–30 proximity 109–10 qualitative and quantitative risk assessment 16–17 rail disasters 15 reckless endangerment 11–12 regulation constitutional protection against deindividualisation 45–8 employer’s liability 174–7 social risks in German administrative law 37–44 res perit domino 114–17, 118–21, 123, 124, 128 reservation of title 117–18, 122, 124 residual risk 27, 56 restitution 115, 118–23, 125–7 risk assessment 16–17, 54–7, 65–7, 106, 176–7 risk management 7, 10, 15, 54–5, 65 Roma in United Kingdom 187, 189 rule of law, prevention of risk and 28–9 Rylands v Fletcher 6 Sanitary and Phytosanitary Measures (SPS) Agreement 64–6, 67

205

Schengen Agreement, border controls and 38–41 Schutzpflichten, discovery of 29–31 security de-individualisation 36–7, 48–9 Germany 41–4 precautionary principle 55–7 terrorism 41–4 sickness in German labour and social insurance law, risk of 161–71 social risks, regulation in German administrative law of 37–44 general assessments of standard situations 37–8 individual behaviour 37–44 Schengen Agreement, border controls and 38–41 terrorism laws 41–4 social security employer’s liability 172, 180–1 immigrant communities, customary law of 184–200 income risk of employed persons in Germany 161–71 sickness in German labour and social insurance law, risk of 161–71 sports injuries 8–9, 11, 86, 88, 91, 164 statehood, liberal concept of 23–6, 29 strict liability 6–7, 107 subjective and objective approach to risk 13–14, 15–17, 31, 86–91, 103–5 suspected danger 17, 26–7 Technical Barriers to Trade (TBT) Agreement 67 television reception, loss of 97–8 terrorism de-individualization 36–7, 41–4, 47–8 Germany 41–4, 47–8 Muslims 42–4 September 11, 2001 attacks on United States 17, 36–7, 41–2 social risks, regulation in German administrative law of 41–4 tort 4, 15–16, 99–100, 106–7, 121, 178 see also negligence transfer of property and risk 114–21, 122–7 transport risks 15, 78–9, 86–8, 90, 103–4, 117, 123–4 see also CIF contracts, transit risks in unfair contract terms 177, 179

206

Index

United States genetically modified organisms (GMOs), dispute with EC over 60, 66–7 September 11, 2001 attacks 17, 36–7, 41–2 unjust enrichment 124–7

vicarious liability 4–5, 101 volenti non fit injuria 85, 89, 91–2, 177 Woolf reforms 146–7, 151–5 working time 179 World Trade Organization (WTO) law 64–7