French Civil Liability in Comparative Perspective 9781509927272, 9781509927302, 9781509927296

Table of contents :
Preface
Contents
List of Contributors
List of Abbreviations
1. Introduction
I. The Existing Pattern of French Civil Liability Law
II. The Reform of French Civil Law
III. The Main Features of the Projet de Réforme
PART I: ‘CIVIL LIABILITY’, CONTRACTUAL AND EXTRA-CONTRACTUAL
2. A Common Framework for Civil Liability?
I. Duality, Unity and Compromise in the Framing of French Civil Liability
II. The Distinction between Contract and Tort in English Law
III. Conclusion
3. The Relationship between Contractual and Extra-Contractual Liability as between Parties to a Contract
I. The Duality of the Two Categories of Civil Liability: A Recurring Controversy
II. The Primacy of Contractual Rules
4. Liability of Contracting Parties Towards Third Parties
I. The Conditions of Admissibility of an Action by Third Parties
II. The Substantive Requirements for an Action by a Third Party
PART II: ‘FAULT’
5. The Definition of Civil Fault
I. Does the Text of the Draft Provision Already Say Too Much?
II. Should the Legislative Definition have Said More?
6. Crime, Breach of Legislative Duties and Fault
I. Introduction
II. Interfaces Between Tort and Crime
III. The Work that Legislative Duty and Fault Do Within the Legal System
IV. The Unity of Fault
V. Integrating Duties Across Tort and Crime
VI. How Unity and Disunity Develop
VII. Distinctive Characteristics of French Tort Law and What we Learn from them
PART III: LIABILITY WITHOUT FAULT
7. The Role of Liability without Fault
I. Mechanisation and Liability without Fault: The Deconstruction of a Myth
II. The Degree of Integration of Liabilities without Fault Within the System of Liability
8. Fait d’autrui in Comparative Perspective
I. Introduction
II. Setting the Scene
III. The French Regime
IV. The English Regime
V. The German Regime
VI. The Insurance Environment
PART IV: ‘HARM’
9. Loss and its Compensation in the Proposed New French Regime of Extra-contractual Liability
I. The Distinction between Dommage and Préjudice
II. The Necessary Characteristics of a Recoverable Préjudice
III. Préjudices Patrimoniaux and Préjudices Extrapatrimoniaux
IV. The Principle of Réparation Intégrale and the Calculation of Damages
V. Conclusion
10. The Concepts of ‘Harm’ in the French and Italian Laws of Civil Liability
I. The French Distinction between Dommage and Préjudice
II. The Italian Distinction between Danno-evento and Danno-conseguenza
III. The Italian Doctrines of Danno Biologico and of Danno Esistenziale
IV. The Concept of 'Harm' in the French and Italian Laws of Civil Liability
V. The Doubtful Compensation of 'Pure Economic Losses' and of 'Pure Non-economic Losses'
VI. The Compensation of 'Harms' in the Absence of Consequential 'Loss'
VII. Conclusion
11. Nuisance and Coming to the Nuisance: The Porous Boundary between Torts and Servitudes in England and France
I. The Tort of Nuisance and Troubles de Voisinage: Similarities and Differences
II. Coming to the Nuisance: The Significanceof the Defendant’s Prior Use of Land
III. Coming to the Nuisance as a Servitude
IV. Conclusions
PART V: CAUSATION
12. Liability for Alternative Causation and for the Loss of a Chance
I. Introduction: Liability for Uncertain, (Merely) Partial Causation under French, Austrian and German Law
II. Concurrence between Two Events which would Trigger Liability
III. Comparative Remarks on Article 1240 of the Projet de réforme
IV. Concurrence between an Event which would Trigger Liability and an Event which would not
V. Comparative Remarks on Article 1238 of the Projet de réforme
VI. Conclusion - An Interaction of Common Factors?
13. ‘Solidary’ Liability and the Channelling of Liability
I. From Obligatio in Solidum to Joint and Several Liability
II. A Common Trend in Comparative Civil Law
III. Conditions and Effects of the Proposed Solidarity Rule
IV. The Questionable Distinction between the Harm Caused by an Undetermined Member of a Group, and the Joint and Several Liability Case
V. The Drawbacks of a General Solidarity Rule in the Law of Torts
PART VI: DEFENCES
14. Defences to Tortious and Contractual Liability in French Law
I. General Issues Surrounding Defences
II. Specific Defences
III. Conclusion
15. Contracts Concerning Civil Liability
I. Introduction
II. The Lawfulness of Contracts Concerning Liability
III. The Rules on Contracts Modifying Civil Liability
PART VII: LIABILITY BEYOND DAMAGES
16. Comparative Reflections on Punishment in Tort Law
I. Introduction
II. Punitive Damages in the United States: Instrumentalism and Interpersonal Justice as Conflicting Approaches to Punishment and Tort Law
III. Exemplary Damages in English Tort Law … Or when Instrumentalism Clashes with Interpersonal Justice
IV. Punishment in French Tort Law: Is Instrumentalism the Whole Story?
V. Conclusions
17. Unjustified Enrichment and Civil Liability
I. The Main Characteristics of Unjustified Enrichment in French Law
II. The Distinction between Civil Liability and Unjustified Enrichment: Is Unjustified Enrichment an Independent Category?
III. The Boundaries between Civil Liability and Unjustified Enrichment: Which Should be the Basis of Recovery of Unlawful Benefits?
IV. Conclusion
18. Injunctions Requiring the Cessation of Unlawful Action
I. Introduction
II. Background
III. Injunctive Relief against Troubles Illicites in the Projet de Réforme
IV. Injunctive Relief in the English Law of Torts
V. Comparing English Law and Article 1266 of the Projet de Réforme
VI. Conclusion
PART VIII: BROAD THEMES
19. The Projet de Réforme du Code Civil Belge and the Reform of the French Civil Code: A Comparison of Selected Topics
I. Introduction
II. The Structure of the Avant-projet de Réforme
III. The Relationship between Contractual and Extra-contractual Liability
IV. Extra-contractual Fault
V. Harm
VI. Reparation of Harm
20. The Reform of Delict in the Civil Code and Liability in Administrative Law
I. Foundations
II. Fault Liability
III. Liability for Others
IV. No-Fault Liability
V. Conclusion
21. The Importance of Terminology in the Law of Civil Liability
I. The Importance of Terminology in the Law
II. The Importance of Terminology in the Law of Civil Liability
III. The Importance of Terminology in the French Law of Civil Liability
IV. Conclusion
22. Principles of Liability or a Law of Torts?
I. A Taste for General Rules
II. Diversity Beneath the Generality
III. The Increased Importance of the 'Special Regimes' of Liability and Compensation
IV. Concluding Observations
Appendix
Index

Citation preview

FRENCH CIVIL LIABILITY IN COMPARATIVE PERSPECTIVE The French law of torts or of extra-contractual liability is widely seen as ­exceptional. For long it was based on a mere five articles of the Civil Code of 1804, but on this foundation the courts and legal scholars have constructed liabilities for fault and strict liability of an extraordinary breadth and significance. While the rest of the general law of obligations (including contract) in the Civil Code was reformed in 2016 by executive ordonnance, this area was left aside, being the subject in 2017 of a proposal by the French Government for the legislative reform of the law of civil liability, a new legislative category to include both contractual and extra-contractual liability. This work considers important aspects of this developing area of French law in a series of essays by French lawyers and comparative lawyers working in French law and other civil law systems. In doing so, it provides insight into the doctrinal thinking and judgments of French lawyers as well as the possible directions in which this area of the law may be developed in the future. Volume 28: Studies of the Oxford Institute of European and Comparative Law

Studies of the Oxford Institute of European and Comparative Law Editor Professor Birke Häcker Board of Advisory Editors Professor Mark Freedland, FBA Professor Stephen Weatherill Professor Stefan Enchelmaier Recent titles in this Series Volume 19: The Unitary EU Patent System Edited by Justine Pila and Christopher Wadlow Volume 20: The EU Charter of Fundamental Rights as a Binding Instrument: Five Years Old and Growing Edited by Sybe de Vries, Ulf Bernitz and Stephen Weatherill Volume 21: The Images of the Consumer in EU Law: Legislation, Free Movement and Competition Law Edited by Dorota Leczykiewicz and Stephen Weatherill Volume 22: Passing Wealth on Death: Will-Substitutes in Comparative Perspective Edited by Alexandra Braun and Anne Röthel Volume 23: General Principles of Law: European and Comparative Perspectives Edited by Stefan Vogenauer and Stephen Weatherill Volume 24: The Future of Contract Law in Latin America: The Principles of Latin American Contract Law Edited by Rodrigo Momberg and Stefan Vogenauer Volume 25: The Code Napoléon Rewritten: French Contract Law after the 2016 Reforms Edited by John Cartwright and Simon Whittaker Volume 26: Discretion in EU Public Procurement Law Edited by Sanja Bogojevic, Xavier Groussot and Jörgen Hettne Volume 27: New Economic Constitutionalism in Europe George Gerapetritis

French Civil Liability in Comparative Perspective Edited by

Jean-Sébastien Borghetti and Simon Whittaker

HART PUBLISHING Bloomsbury Publishing Plc Kemp House, Chawley Park, Cumnor Hill, Oxford, OX2 9PH, UK HART PUBLISHING, the Hart/Stag logo, BLOOMSBURY and the Diana logo are trademarks of Bloomsbury Publishing Plc First published in Great Britain 2019 Copyright © The editors and contributors severally 2019 The editors and contributors have asserted their right under the Copyright, Designs and Patents Act 1988 to be identified as Authors of this work. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior permission in writing from the publishers. While every care has been taken to ensure the accuracy of this work, no responsibility for loss or damage occasioned to any person acting or refraining from action as a result of any statement in it can be accepted by the authors, editors or publishers. All UK Government legislation and other public sector information used in the work is Crown Copyright ©. All House of Lords and House of Commons information used in the work is Parliamentary Copyright ©. This information is reused under the terms of the Open Government Licence v3.0 (http://www.nationalarchives.gov.uk/doc/ open-government-licence/version/3) except where otherwise stated. All Eur-lex material used in the work is © European Union, http://eur-lex.europa.eu/, 1998–2019. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication data Names: Borghetti, Jean-Sébastien, editor.  |  Whittaker, Simon, editor. Title: French civil liability in comparative perspective / [edited by] Jean-Sébastien Borghetti, Simon Whittaker. Description: Chicago : Hart Publishing, an imprint of Bloomsbury Publishing, 2019.  |  Series: Studies of the Oxford Institute of European and Comparative Law; volume 28  |  Includes bibliographical references and index. Identifiers: LCCN 2019034066 (print)  |  LCCN 2019034067 (ebook)  |  ISBN 9781509927272 (hardback)  |  ISBN 9781509927289 (Epub) Subjects: LCSH: Liability (Law)—France.  |  Torts—France.  |  France. Code civil. Classification: LCC KJV2009 .F74 2019 (print)  |  LCC KJV2009 (ebook)  |  DDC 346.4402/2—dc23 LC record available at https://lccn.loc.gov/2019034066 LC ebook record available at https://lccn.loc.gov/2019034067 ISBN: HB: 978-1-50992-727-2 ePDF: 978-1-50992-729-6 ePub: 978-1-50992-728-9 Typeset by Compuscript Ltd, Shannon To find out more about our authors and books visit www.hartpublishing.co.uk. Here you will find extracts, author information, details of forthcoming events and the option to sign up for our newsletters.

PREFACE In 2016 the provisions of the French Civil Code dealing with the law of contract, the general regime of obligations, and proof of obligations, were reformed by ordonnance (a special form of governmental legislation). This reform of 2016 was the subject of an earlier work in this series, edited by John Cartwright and Simon Whittaker, The Code Napoléon Rewritten, French Contract Law after the 2016 Reforms (Bloomsbury, 2017). While the scope of this reform was very broad, the Ordonnance of 2016 left aside any substantive change to the law of extra-contractual liability (sometimes still called delictual liability or, in common law terms, the law of torts) for later attention. In March 2017, the French Ministry of Justice published a revised draft parliamentary bill, the Projet de réforme de la responsabilité civile, whose aim would be to reform the law governing ‘civil liability’, a new legislative category to include contractual as well as extra-contractual liability. Instead of the mere five articles of the original Civil Code of 1804, which have been supplemented mainly by the cumbersome implementation of the European­ Product Liability Directive, the proposed new law consists of some 83 articles which would be squeezed into the Code. In our view, this legislative reform proposal provided an excellent occasion to reconsider the French law of ‘civil liability’, even if the proposal never in fact becomes law. We therefore invited colleagues working on French civil liability in the UK, in France and in other continental laws to examine particular topics in this area from the perspective of comparative law and in the light of the Ministry’s proposals. The present volume in the series of the Studies of the Oxford Institute of European and Comparative Law is the product of the colloquium held at St John’s College, Oxford, in September 2018, at which we discussed just over 20 draft papers within this area. After an introductory chapter which provides a little more background to the French law of civil liability and to the Ministry’s proposed reform, this work therefore contains a series of chapters by legal scholars on many of the key topics arising from the proposed reform of French law and some going beyond its scope, such as on the relationship between private law liability and liability in the administration. To these discussions, we have added the French text of the Projet de réforme together with an English translation which we had earlier been commissioned to produce by the French Ministry of Justice. All the chapters were either written in English or have been translated into English for the purpose of this publication. The colloquium in September 2018 was organised by the Institute of ­European and Comparative Law and supported financially both by the Institute and by

vi  Preface St John’s College, Oxford. We are very grateful for this financial support. We are also most grateful to Jenny Dix, Administrator at the Institute of European and Comparative Law for her invaluable assistance throughout the preparation for, and running of, the colloquium, and to the President and Fellows of St John’s College for their hospitality. Our particular thanks go to the participants at the colloquium, both those who presented papers (and have reworked them for publication in this volume) and others who attended and made very significant contributions to the discussions. We would like to thank John Cartwright and Ciara Kennefick for helping us to translate some of the papers which were originally written in French. And we would especially like to thank Marco Cappelletti for his invaluable work on the technical editing of the papers without which this book would not have come to fruition. At a time when formal and legal connections between the UK and the legal systems of continental Europe are in a state of flux, we were pleased to find so many points of interest in common and issues to discuss. Comparative law remains, in our view, fundamentally a means to promote the understanding of the ‘other’ law and the ways in which other lawyers see their law and its purposes, but we find that the process in turn shines light on one’s own law, raising new questions and suggesting new approaches. We hope that, in its particular way, this volume contributes to this much larger venture. Jean-Sébastien Borghetti Simon Whittaker Paris and Oxford May 2019

CONTENTS Preface����������������������������������������������������������������������������������������������������������������������������v List of Contributors����������������������������������������������������������������������������������������������������� xi List of Abbreviations������������������������������������������������������������������������������������������������� xiii 1. Introduction�����������������������������������������������������������������������������������������������������������1 Jean-Sébastien Borghetti and Simon Whittaker PART I ‘CIVIL LIABILITY’, CONTRACTUAL AND EXTRA-CONTRACTUAL 2. A Common Framework for Civil Liability?��������������������������������������������������������15 Simon Whittaker 3. The Relationship between Contractual and Extra-Contractual Liability as between Parties to a Contract����������������������������������������������������������37 Yves-Marie Laithier 4. Liability of Contracting Parties Towards Third Parties��������������������������������������55 Philippe Stoffel-Munck PART II ‘FAULT’ 5. The Definition of Civil Fault��������������������������������������������������������������������������������79 Marie Dugué 6. Crime, Breach of Legislative Duties and Fault���������������������������������������������������99 Matthew Dyson PART III LIABILITY WITHOUT FAULT 7. The Role of Liability without Fault��������������������������������������������������������������������123 Jonas Knetsch 8. Fait d’autrui in Comparative Perspective����������������������������������������������������������143 Birke Häcker

viii  Contents PART IV ‘HARM’ 9. Loss and its Compensation in the Proposed New French Regime of Extra-contractual Liability����������������������������������������������������������������������������181 Dorota Leczykiewicz 10. The Concepts of ‘Harm’ in the French and Italian Laws of Civil Liability��������������������������������������������������������������������������������������������������205 Pietro Sirena 11. Nuisance and Coming to the Nuisance: The Porous Boundary between Torts and Servitudes in England and France�������������������������������������223 Ciara Kennefick PART V CAUSATION 12. Liability for Alternative Causation and for the Loss of a Chance���������������������������������������������������������������������������������������������������������247 Nuno Manuel Pinto Oliveira 13. ‘Solidary’ Liability and the Channelling of Liability�����������������������������������������263 Carlos Gómez Ligüerre PART VI DEFENCES 14. Defences to Tortious and Contractual Liability in French Law�����������������������285 Sandy Steel 15. Contracts Concerning Civil Liability�����������������������������������������������������������������309 Zoé Jacquemin PART VII LIABILITY BEYOND DAMAGES 16. Comparative Reflections on Punishment in Tort Law�������������������������������������329 Marco Cappelletti 17. Unjustified Enrichment and Civil Liability�������������������������������������������������������353 Mélodie Combot 18. Injunctions Requiring the Cessation of Unlawful Action���������������������������������377 Paula Giliker

Contents  ix PART VIII BROAD THEMES 19. The Projet de Réforme du Code Civil Belge and the Reform of the French Civil Code: A Comparison of Selected Topics�����������������������������397 Bernard Dubuisson 20. The Reform of Delict in the Civil Code and Liability in Administrative Law���������������������������������������������������������������������������������������425 John Bell 21. The Importance of Terminology in the Law of Civil Liability��������������������������445 Olivier Deshayes 22. Principles of Liability or a Law of Torts?�����������������������������������������������������������455 Jean-Sébastien Borghetti and Simon Whittaker Appendix��������������������������������������������������������������������������������������������������������������������477 Index��������������������������������������������������������������������������������������������������������������������������503

x

LIST OF CONTRIBUTORS John Bell QC (Hon), FBA is Professor of Law at the University of Cambridge and Fellow of Pembroke College, Cambridge. Jean-Sébastien Borghetti is professeur de droit at the université Panthéon-Assas (Paris II). Marco Cappelletti is a research student at the University of Oxford and Junior Research Fellow elect at St John’s College, Oxford. Mélodie Combot is a research student at the université Panthéon-Assas (Paris II). Olivier Deshayes is professeur de droit at the université Paris Nanterre. Bernard Dubuisson is professeur de droit at the université catholique de Louvain. Marie Dugué is professeur de droit at the université de Tours. Matthew Dyson is Associate Professor of Law at the University of Oxford and Fellow of Corpus Christi College, Oxford. Paula Giliker is Professor of Comparative Law at the University of Bristol. Birke Häcker is Linklaters Professor of Comparative Law at the University of Oxford and Fellow of Brasenose College, Oxford. Zoé Jacquemin is Maître de conférences at the université de Paris 13. Ciara Kennefick is Associate Professor of Law at the University of Oxford and Student (Fellow) of Christ Church, Oxford. Jonas Knetsch is professeur de droit at the université Jean Monet Saint-Etienne. Yves-Marie Laithier is professeur de droit at l’École de droit de la Sorbonne. Dorota Leckzykiewicz is Associate Professor of Law at the University of Oxford and Fellow of St Peter’s College, Oxford. Carlos Gomez Ligüerre is Professor titular at the Universitat Pompeu Fabra, Barcelona. Nuno Manuel Pinto Oliveira is Professor Catedrático at the Universidade do Minho. Pietro Sirena is Dean and Professor at Bocconi University, Milan.

xii  List of Contributors Sandy Steel is Associate Professor of Law at the University of Oxford and Fellow of Wadham College, Oxford. Philippe Stoffel-Munck is professeur de droit at l’École de droit de la Sorbonne. Simon Whittaker is Professor of European Comparative Law at the University of Oxford and Fellow of St John’s College, Oxford.

LIST OF ABBREVIATIONS A&E

Adolphus and Ellis’s Reports, Queen’s Bench (ER vols 110–113)

AC

Law Reports, Appeal Cases (Third Series) (1891–)

AcP

Archiv für die civilistische Praxis

ACQP

Principles of the Existing EC Contract Law (Acquis Principles) (2007–2009)

actu

actualités

ADC

Anuario de Derecho Civil

ALR

Australian Law Reports

AJDA

Actualité Juridique: Droit administratif

AJDI

Actualité Juridique: Droit immobilier

ALJR

Australian Law Journal Reports

App Cas

Law Reports, Appeal Cases (Second Series) (1875–1890)

AEPL (Gandolfi) Code

See Special Abbrevations below

APDC

Asociación de Profesores de Derecho Civil

Arch phil dr

Archives de philosophie du droit

Arr cass

Arresten van het Hof van cassatie

Art/art

Article

Ass plén

Assemblée pléniere de la Cour de cassation

B & Ald

Barnewall and Alderson’s Reports, King’s Bench (ER vol 106)

B&C

Barnewall and Cresswell’s Reports, King’s Bench (ER vols 107–109)

B&S

Best and Smith’s Reports, Queen’s Bench (ER vols 121–122)

BGB

Bürgerliches Gesetzbuch (Germany, 1900)

BGH

Bundesgerichtshof

BGHZ

Entscheidungen des Bundesgerichtshofs in Zivilsachen

xiv  List of Abbreviations Bing

Bingham’s Reports, Common Pleas (ER vols 130–131)

BJS

Bulletin Joly Sociétés

BOPC

Butlletí Oficial del Parlament de Catalunya

Bull ass

Bulletin des assurances [Belgium]

Bull civ

Bulletin des arrêts de la Cour de cassation, Chambres civiles

BVerfG

Bundesverfassungsgerichts

BVerfGE

Entscheidungen des Bundesverfassungsgerichts

BW

Burgelijk wetboek (Netherlands, 1992)

C assur

Code des assurances

C propr int

Code de la propriété intellectuelle

CA

Cour d’appel; Court of Appeal

Cass.

Cour de cassation de Belgique

Cass sez I, un

Corte di cassazione, Sezione I civile, Sezioni unite

Cass civ

Chambre civile de la Cour de cassation

Cass civ (1), (2) and (3)

Première, deuxième and troisième chambre civile de la Cour de cassation

Cass com

Chambre commerciale de la Cour de cassation

Cass mixte

Chambre mixte de la Cour de cassation

Cass soc

Chambre sociale de la Cour de cassation

CB

Common Bench Reports, Common Pleas (ER vols 135–139)

CBNS

Common Bench Reports, New Series, Common Pleas (ER vols 140–144)

Cc

Code civil [français]

Cc belge

Code civil belge [Belgium]

CC

Conseil Constitutionnel

Ccom

Code de commerce

Cconsom

Code de la consommation

CCQ

Code civil of Quebec

CE

Conseil d’État

CE Ass

Conseil d’État, assemblée du contentieux

List of Abbreviations  xv CESL

See Special Abbreviations below

Cf/cf

Compare

CFR

Common Frame of Reference

CGI

Code général des impôts

Ch

Law Reports, Chancery Division (1890–)

Ch/ch

Chapter

ChD

Chancery Division of the High Court of Justice;

chron

chroniques

Ch Réun

Chambres réunies de la Cour de Cassation

CISG

UN Convention on the International Sale of Goods (1980) (Vienna Sales Convention)

CJEU

Court of Justice of the European Union

CLJ

Cambridge Law Journal

CLP

Current Legal Problems

Cmnd

command (identifier for published parliamentary papers)

Cod civ

Codice civile [Italy]

COM

European Commission Documents

Com Cas

Commercial Cases

comm

commentaire

Comm com électr

Communication Commerce électronique

concl

conclusions

Contr impr

Contratto e impresa

Contr impr Eur

Contratto e impressa Europa

Contrats Concur Consom Contrats-Concurrence-Consommation CP

Code Pénal

CPC

Code de procédure civil

CPCE

Code des procédures civiles d’exécution

CPD

Common Pleas Division of the High Court of Justice; Law Reports, Chancery Division (1875–1890) Law Reports, Common Pleas Division (1875–1880)

CPP

Code de procédure pénale

xvi  List of Abbreviations D

Recueil Dalloz

D aff

Recueil Dalloz, Cahier droit des affaires

DCFR

See Special Abbreviations below

DDHC

Déclaration des droits de l’homme et du citoyen

Defr

Répertoire du notariat Defrénois

DH

Dalloz, Recueil hebdomadaire de jurisprudence (1924–1940)

Digest

Digest of Justinian

Dir

Directive

doctr

doctrine

DOGC

Diari Oficial de la Generalitat de Catalunya

DP

Dalloz, Recueil périodique et critique de jurisprudence, de legislation et de doctrine (1825–1940)

Dr & patrim

Droit & patrimoine

Dr circ

Droit de la circulation

E&B

Ellis & Blackburn’s Reports, Queen’s Bench (ER vols 118–120)

East

East’s Reports, King’s Bench (ER vols 102–104)

EBLR

European Business Law Review

EC

European Community

ed(s)

editor(s)

edn

edition

EEC

European Economic Community

Eg/eg

for example

EGLR

Estates Gazette Law Reports

ELJ

European Law Journal

ER

English Reports

ERCL

European Review of Contract Law

ERPL

European Review of Private Law

esp

especially

et al

and others, et alii

EU

European Union

Euro Dir Priv

Europa e diritto privato

EWCA Civ

Decision of the Court of Appeal (Civil Division)

List of Abbreviations  xvii EWHC

Decision of the High Court

F Supp

Federal Supplement (United States)

FamRZ

Zeitschrift für das gesamte Familienrecht

Fasc

fascicule

Foro It

Il Foro Italiano

Gaz Pal

Gazette du Palais

HL

House of Lords

Inst

Institutiones (Institutes of Justinian)

ICJ

International Court of Justice

ICLQ

International and Comparative Law Quarterly

J

Mr Justice

JT

Journal des Tribunaux

JCA

Juris-Classeur Administratif

JCP

Juris-classeur périodique, La semaine juridique

JCP E

Juris-classeur périodique, La semaine juridique, entreprises et affaires

JCP G

Juris-classeur périodique, La semaine juridique, édition générale

JCP N

Juris-classeur périodique, La semaine juridique, édition notariale

JETL

Journal of European Tort Law

JherJB

Jherings Jahrbücher für die Dogmatik des bürgerlichen Rechts

JL & Com

Journal of Law and Commerce

JLMB

Revue de jurisprudence de Liège, Mons et Bruxelles

JO Ass nat

Journal officiel, Assemblée Nationale

JO Sén

Journal officiel, Sénat

JO S(CR)

Journal official de la République Française, Sénat, Compte Rendu Intégral

JSC

Justice of the Supreme Court

jur

jurisprudence

JZ

Juristenzeitung

KB

Law Reports, King’s Bench Division (1901–1952)

Law Com

Law Commission

Leb

Recueil Lebon (properly, Recueil des décisions du Conseil d’Etat, statuant au contentieux)

xviii  List of Abbreviations LEDC

L’essentiel Droit des contrats

LGDJ

Librairie générale de droit et jurisprudence

LJ

Lord Justice of Appeal

Lloyd’s Rep

Lloyd’s List Law Reports

LPA

Les petites affiches

LQR

Law Quarterly Review

LR Ch App

Law Reports, Chancery Appeals (1865–1875)

LR Ex

Law Reports, Exchequer Cases (1865–1875)

LR HL

Law Reports, English & Irish Appeals (1866–1875)

LR QB

Law Reports, Queen’s Bench (First Series) (1865–1875)

LT

Law Times Reports

Ltd

Limited

Mon B

Moniteur belge

Moore PC

Moore, Privy Council Cases (Privy Council) (ER vols 12–15)

N, n

Note

NJW

Neue Juristische Wochenschrift

No(s), no(s)

Number(s)

Nuova giur civ comm

Nuova giurisprudenza civile commentata

obs

observation

OJ

Official Journal of the European Communities/ Union

Pan

Panorama

para(s)

paragraph(s)

Pas

Pasicrisie

PECL

See Special Abbreviations below

PETL

See Special Abbreviations below

PICC

UNIDROIT Principles of International Commercial Contracts (2010 edn); see also Special Abbreviations below

pr

principium

PSC

President of the Supreme Court

Pt

Part

PUAM

Presses Universitaires d’Aix-Marseille

List of Abbreviations  xix PUF

Presses Universitaires de France

QB

Court of Queen’s Bench Law Reports, Queen’s Bench Division (1891–1901, 1952–)

QBD

Queen’s Bench Division of the High Court of Justice; Law Reports, Queen’s Bench Division (1875–1890)

R

Rapport annuel de la Cour de cassation

RabelsZ

Rabels Zeitschrift für ausländisches und internationales Privatrecht

rapp

rapport

RCA

Responsabilité civile et assurances

RCJB

Revue critique de jurisprudence belge

RCLJ

Revue critique de législation et jurisprudence

RDA

Revue de droit d’Assas

RDC

Revue des contrats

RDI

Revue de droit immobilier

RDIDC

Revue de droit international and droit comparé

RDP

Revue du droit public et de la science politique en France et à l’étranger

reg

regulation

Req

Chambre des requêtes de la Cour de cassation

Rev arb

Revue de l’arbitrage

Rev prat soc

Revue pratique des sociétés civiles et commerciales

Rev soc

Revue des sociétés

RFDA

Revure Française de Droit Administratif

RGAR

Revue générale des assurances et des responsabilités

RGDC

Revue générale de droit civil belge

RIDA

Revue internationale des droits de l’antiquité

RIDC

Revue internationale de droit comparé

Riv Dir Civ

Rivista di diritto civile

Riv Dir Comm

Rivista del diritto commerciale e del diritto generale delle obbligazioni

Riv Trim Dir & Proc Civ

Rivista trimestrale di diritto e procedura civil

RJ

Repertorio de jurisprudencia

xx  List of Abbreviations RJ com

Revue de jurisprudence commerciale

RJQ

Recueil de jurisprudence du Québec

RLDC

Revue Lamy droit civil

RRJ RTD civ

Revue de la recherche juridique, Droit prospectif Revue trimestrielle de droit civil

RTD com

Revue trimestrielle de droit commercial et de droit économique

RTD eur

Revue trimestrielle de droit européen

RW

Rechtskundig Weekblad

S

Recueil Sirey

SC

Session Cases (Scotland)

Sched

Schedule

SCR

Supreme Court Reports (Canada)

somm

sommaires

s(s)

section(s)

T com

Tribunal de commerce

Taunt

Taunton’s Reports, Common Pleas (ER vols 127–129)

TC

Tribunal des conflits

TGI

Tribunal de Grande Instance

TLR

Times Law Reports

TR

Term Reports, King’s Bench (ER vols 99–101)

Tul L Rev

Tulane Law Review

TvR

Tijdschrift voor Rechtsgeschiedenis

UKHL

Decision of the House of Lords

UKPC

Decision of the Privy Council

UKSC

Decision of the UK Supreme Court

USA

United States of America

Ves Jun

Vesey Junior’s Reports, Chancery (ER vols 30–34)

vol(s)

volume(s)

WLR

Weekly Law Reports

ZEuP

Zeitschrift für Europäisches Privatrecht

List of Abbreviations  xxi

Special Abbreviations Avant-projet Catala

P Catala (ed), Avant-projet de réforme du droit des obligations et de la prescription: Rapport remis au garde des Sceaux (Paris, Documentation française, 2006); translation by J Cartwright and S Whittaker in J Cartwright, S Vogenauer and S Whittaker (eds), Reforming the French Law of Obligations (Oxford, Hart Publishing, 2009).

Avant-projet Terré

F Terré (ed), Pour une réforme du droit des contrats (Paris, Dalloz, 2009); Pour une réforme du droit de la responsabilité (Paris, Dalloz 2011); Pour une réforme du régime général des obligations (Paris, Dalloz, 2013)

Chancellerie

the French Ministry of Justice: see www. justice.gouv.fr/histoire-et-patrimoine-10050/ chancellerie-garde-des-sceaux-aux-origines-desmots-23874.html

The Code Napoléon Rewritten

J Cartwright and S Whittaker (eds), The Code Napoléon Rewritten – French Contract Law after the 2016 Reforms (Oxford, Hart Publishing, 2017)

DCFR

C von Bar and E Clive (eds), Principles, Definitions and Model Rules of European Private Law: Draft Common Frame of Reference (DCFR) (Munich, Sellier, 2009)

Loi de ratification of 2018

Loi no 2018–287 of 20 April 2018 ratifiant l’ordonnance no 2016–131 du 10 février 2016 portant réforme du droit des contrats, du régime général et de la preuve des obligations.

Ordonnance of 2016 [or Ordonnance]

Ordonnance no 2016–131 du 10 février 2016 portant réforme du droit des contrats, du regime general et de la preuve des obligations as amended by loi no 2018–287 of 20 April 2018 ratifiant l’ordonnance no 2016–131 du 10 février 2016 portant réforme du droit des contrats, du régime général et de la preuve des obligations. Unless otherwise stated, the English translation used is that by John Cartwright, Bénédicte Fauvarque-Cosson and Simon Whittaker (2016 and 2018).

xxii  List of Abbreviations PECL

O Lando and H Beale, Principles of European Contract Law, Pts I and II (2000), Pt III (2003) (The Hague, Kluwer Law International)

PETL

European Group on Tort Law, Principles of European Tort Law. Text and Commentary (Wien/New York, Springer, 2005)

Projet de réforme [de la responsabilité civile]

Ministère de la Justice, Projet de réforme de la responsabilité civile March 2017, available www. justice.gouv.fr/publication/ Projet_de_reforme_ de_la_responsabilite_civile_13032017.pdf Unless otherwise stated, the English translation used is that by Simon Whittaker in collaboration with Jean-Sébastien Borghetti (and see Appendix to the present work). On the convention of citation of the Projet de réforme, see below, p 8.

Reforming the French Law of Obligations

J Cartwright, S Vogenauer, and S Whittaker (eds), Reforming the French Law of Obligations: Comparative Reflections on the Avant-projet de réforme du droit des obligations et de la prescription (‘the Avant-projet Catala’) (Oxford, Hart Publishing, 2009)

Report to the President [of the Republic] on Ordonnance of 2016

Rapport au Président de la République relatif à l’ordonnance n° 2016–131 du 10 février 2016 portant réforme du droit des contrats, du régime général et de la preuve des obligations, available at www.legifrance.gouv.fr/eli/rapport/2016/2/11/ JUSC1522466P/jo/texte

1 Introduction JEAN-SÉBASTIEN BORGHETTI AND SIMON WHITTAKER

This book considers important aspects of the French law of civil liability in the light of the recent Projet de réforme de la responsabilité civile, or Reform Bill on Civil Liability. It does so in a series of essays by French lawyers and comparative lawyers working in French law and in other civil law systems. Writing in English about a legal system which is itself expressed in a language other than English raises terminological issues, and we should address these at the outset. In this book, ‘civil liability’ translates the French responsabilité civile, which describes both contractual and extra-contractual liability. In the French context, responsabilité civile is distinguished both from responsabilité pénale, ie criminal liability, and responsabilité administrative, ie the liability of the State and public bodies (which French law calls ‘public persons’) for the harm that they may cause. The latter distinction means that civil liability, in the French sense of the term, only covers the liability of private persons.1 The reason why this is so is that civil liability is based on provisions of the Code civil, and that the Conseil d’État, which had (and still has) jurisdiction over claims for liability in the administration, deemed the law contained in these provisions inappropriate to govern the liability of public persons, given in particular that they act in the public interest.2 As is well known, the Code civil was adopted in 1804 under the authority of Napoleon as part of a broader codification movement. To this day, the Code’s provisions dealing with civil liability have remained almost unaltered. In this respect, ‘civil liability’ as a category is not as such identified in the Code. Instead, as promulgated, the Code civil dealt with damages arising out of non-performance of an obligation (‘des dommages et intérêts résultant de l’inexécution de l’obligation’, Section 4 of Chapter III of Title III of Book 3 as originally enacted) in a context which is wholly contractual, on the one hand, and with delicts and quasi-delicts (‘des délits et quasi-délits’, Chapter II of Title IV of Book 3 as originally enacted), on the other. The first category of damages progressively came to be known as 1 See further below, ch 20. 2 The key decision is seen as the arrêt Blanco, TC 8 February 1873, DP 1873, 3, 17 which denied the application of liability for fault under arts 1382 and 1383 Cc for public persons.

2  Jean-Sébastien Borghetti and Simon Whittaker responsabilité contractuelle, ie contractual liability; the second category was soon called responsabilité délictuelle (delictual liability) and has also been known for a few decades as responsabilité extracontractuelle (extra-contractual liability).3 In common law terms, this could be called tortious liability or liability in tort. The Projet de réforme, which is the main focus of this book, follows the modern terminology, referring to responsabilité civile in a general way to describe both responsabilité contractuelle and responsabilité extracontractuelle. In the various chapters in this book, some contributors use ‘tort’ or ‘tortious’ to refer to French extra-contractual liability; others stay more closely to the French terminology, preferring ‘contractual liability’, ‘extra-contractual liability’, and ‘civil liability’ to describe both. Later chapters will consider particular substantive provisions of the Projet de réforme, but this instrument can be understood only against the background of the existing pattern of French civil liability law and in the wider context of civil law reform in France. Having considered these in turn, we will outline its principal features.

I.  The Existing Pattern of French Civil Liability Law From a comparative perspective, the French law of extra-contractual liability is widely seen as exceptional and it deserves to be considered first. Of the original 2,281 articles of the Code civil, only five, articles 1382 to 1386, were devoted to this branch of the law. These articles were drafted in very broad terms, according to the typical, perhaps one could even say, the best style of the Code civil. They read as follows: Article 1382: Any human action whatsoever which causes harm to another creates an obligation in the person by whose fault it occurred to make reparation for it. Article 1383: Everyone is liable for harm which he has caused not only by his action, but also by his failure to act or his lack of care. Article 1384:[4] One is liable not only for the harm which one causes by one’s own action, but also for that which is caused by the action of persons for whom one is responsible, or of things which one has in one’s keeping. 3 To make matters more complicated, the expression responsabilité civile was first used by jurists instead of responsabilité délictuelle on the basis that it is less redolent of fault and therefore more appropriate for the range of liabilities imposed without fault: eg, B Starck, H Roland and L Boyer, Obligations, 1.Responsabilité délictuelle, 5th edn (Paris, Litec, 1996) 5 explaining that, properly, responsabilité délictuelle refers to liability for an intentional fault, whereas responsabilite quasi-délictuelle refers to liability for lack of care or negligence, but that usually responsabilité délictuelle refers to all ‘civil liabilities’. Cf J Carbonnier, Droit civil, Les biens, Les obligations (Paris, PUF, 2004) vol II, 2251ff, who contrasts responsabilité délictuelle with responsabilité contractuelle, but then sets out the law governing the former in terms of responsabilité civile. Nowadays, however, responsabilité civile is normally not used only to designate responsabilité délictuelle, and instead clearly refers to both extra-contractual and contractual liability. 4 The second and third paras of art 1242 were added in 1937, and the text of what is now para 4 was slightly altered in 1971 and then again in 2002 to reflect changes in family law.

Introduction  3 However, a person who holds on whatever legal basis all or part of immovable or movable property in which a fire has arisen shall not be liable as regards third parties for harm caused by that fire unless it is established that it must be attributed to his fault or the fault of persons for whom he is responsible. This provision does not apply to the relations between owners and tenants, which remain governed by articles 1733 and 1734 of the Civil Code. To the extent to which they exercise parental authority, a father and mother are jointly and severally liable for harm caused by their minor children who live with them. Masters and employers, for harm caused by their servants and employees within the functions for which they employed them. Teachers and artisans, for harm caused by their pupils and apprentices during the time which they are under their supervision. The above liability arises unless the father and mother and the artisans cannot prove that they could not have prevented the action which gave rise to this liability. As regards teachers, fault, lack of care or a failure to act invoked against those as having caused the harmful action, must be proved by the claimant at first instance following the general rule. Article 1385: The owner of an animal, or a person who uses an animal while he uses it, is liable for the harm which the animal has caused, whether the animal was in his keeping or whether it had gone astray or escaped. Article 1386: The owner of a building is liable for the harm caused by its ruin, where the latter occurred as a result of a lack of maintenance or defect in construction.

The small number of provisions on extra-contractual liability reflects the relative lack of importance of the subject for the drafters of the Code civil, and the decades immediately following the adoption of the Code confirmed the limited relevance of extra-contractual liability, with comparatively few published cases applying these provisions. However, as France changed over the course of the nineteenth century from a rural and agricultural society to a more urban and industrialised one, the number of tort claims brought before the courts started to increase. For this purpose, the open-ended character of articles 1382 to 1386 allowed the courts to extend considerably the range of extra-contractual liability. In this, articles 1382 and 1384 played a very prominent role. While article 1382 was initially intended to cover liability for intentional faults, and article 1383 liability for negligence, the former was progressively regarded as formulating a general rule of liability for fault, encompassing both liability for negligence and liability for intentional faults. Accordingly, article 1383 slowly faded away and was hardly ever cited in discussions or relied on by litigants. The general rule to be drawn from article 1382 can be restated as follows: a person who commits a fault must make reparation for the harm caused by that fault. This rule is as broad as it looks. Not only does it cover any type of fault, whether intentional or non-intentional, it also encompasses any type of harm, be it bodily injury, damage to physical property, the violation of a right or even pure emotional harm or pure economic loss. While in many legal systems judges and lawyers have often been concerned with keeping the ‘floodgates of liability’ closed, since the

4  Jean-Sébastien Borghetti and Simon Whittaker end of the nineteenth century French legal thinking on liability has been dominated by the idea that the main purpose of the law of extra-contractual liability is to grant compensation to the victims of harm. Political arguments regarding the limits which liability law imposes on the freedom of citizens, as well as economic arguments pointing to the cost of liability for society, have never received a warm reception in France, and have even tended to be frowned upon as showing disregard for the plight of those who have suffered harm. This concern for the compensation of ‘victims’, as claimants are significantly called by French lawyers5 (and, indeed, by the Projet de réforme itself), is perhaps best illustrated by the way in which courts have created new instances of liability of their own initiative – which is itself remarkable in a system where normally the law is created by the legislator, and not by judges. Of these instances, the most significant is strict liability for the ‘action of things’. The late nineteenth century in France witnessed many workplace accidents that could not be attributed to anyone’s fault, nor was there at the time a legislative scheme of workers’ compensation. Against this background, the Cour de cassation, France’s highest court in civil matters, relied on the first paragraph of article 1384 of the Code civil to create a new strict liability regime. This provision had not been intended to have any normative value, and instead served only as a transition between the two provisions on liability for fault (articles 1382 and 1383) and the provisions on liability for others (paragraphs 2 to 4 of article 1384 as enacted), and liability for harm caused by animals (article 1385) or ruinous buildings (article 1386). However, the courts relied on the very broad statement that ‘one is liable … for the harm … which is caused by the action … of things which one has in one’s keeping’ as allowing them to hold the ‘keepers’ (gardiens) of things strictly liable for any harm caused by things they have under their custody. While legislation on workers’ compensation soon deprived this new liability of its utility in the case of workplace accidents, article 1384(1) flourished in the context of road accidents, where in effect it made drivers strictly liable for harm caused by their vehicle. Even though in 1985 special legislation was passed on the compensation of victims of traffic accidents which has deprived the general liability for the action of things of most its practical significance, this ground of liability remains an ­important symbol both of the concern of French liability law for victims and of the major role of the courts in developing it. This is not the only example of a judge-made ‘tort’, as French courts have created a strict liability for an abnormal nuisance between neighbours, as well as two instances of liability for the actions of other people which are not mentioned in the Code but were found shelter under the welcoming legislative umbrella of article 1384(1)’s broad formulation. ­Moreover, far from opposing this move towards more liability, the legislator has created several strict liability regimes covering specific risks. The most important

5 See S Whittaker, La responsabilité pour fait personnel dans l’Avant-projet de réforme du droit de la responsabilité: donner voix aux silences du Code civil (RDC, 2007) 89, 99.

Introduction  5 in ­practice is the regime applicable to traffic accidents, which is to be found in special legislation (the ‘loi Badinter’ of 1985) outside the Code civil and which imposes a liability even stricter than the liability formerly imposed in this context under article 1384(1) Cc. France also implemented (albeit after a considerable delay) the 1985 European Directive on Product Liability into the Code civil (­articles 1386-1 to 1386-18).6 By comparison to these developments, the law governing French contractual liability is not as surprising to a common lawyer. While in French law, liability is usually strict, there can be situations where the debtor is liable only where the contractual non-performance was caused by his fault in the sense of a lack of care. In principle, only losses that were reasonably foreseeable at the time of the conclusion of the contract can be compensated, though French courts set no limit on the nature of these losses and therefore see no objection to compensating ‘moral’ or emotional harm arising from contractual non-performance. The most distinctive trait of French contractual liability is probably the so-called ‘non-cumul’ rule, which is in fact a no-option rule, by which a party to a lawfully concluded contract can claim against his contractual partner only on the basis of the rules of contract liability and not on the basis of the rules of extra-contractual liability, whenever harm arises in connection with the performance of that contract. This judgemade rule, which was developed principally from a concern that extra-contractual liability could upset the allocation of risks agreed between contracting parties, has resulted in contractual liability often being used to compensate losses arising out of personal injuries or damage to physical property. The courts ‘find’ a contractual obligation as to the safety of the other party or his property (une obligation de sécurité), determine whether it imposes strict liability (based on an obligation de résultat) or only negligence-based liability (based on an obligation de moyens) and then rule out recourse to any ground of extra-contractual liability. As a result, in common with extra-contractual liability, contractual liability has come to be seen first and foremost as a means of compensating harm, rather than as a way of vindicating the creditor’s expectation interest in the performance of the contract.7 The French law of civil liability, contractual and extra-contractual, therefore appears to be very broad. Not only does it allow recovery of almost all types of harm, but it also recognises many instances of strict liability. This situation is undoubtedly the result of a ‘pro-victim bias’ in legal actors (scholars, judges and the legislator itself), but it has also been made possible by the existence of a generous social security system and of a number of compensation funds which provide for the needs of victims of certain specific risks or products, such as asbestos, and by the fact that many defendants are covered by insurance.8 This means that the

6 Council Directive 1985/374/EEC concerning liability for defective products first implemented in French law by loi no 98-389 of 19 May 1998 relative à la responsabilité du fait des produits défectueux. 7 But see the discussions below by Whittaker (ch 2) and Laithier (ch 3). 8 J-S Borghetti, ‘The Culture of Tort Law in France’ (2012) 3 JETL 158.

6  Jean-Sébastien Borghetti and Simon Whittaker actual cost of compensating many losses is not borne by those who are in theory liable for them. It should also be stressed that French liability rules, whether they are of legislative or judicial origin, are formulated in very general terms. This gives considerable room in practice for judicial discretion and leads to the paradoxical result that French civil liability is in fact highly casuistic. Given that the formulation of the rules makes it difficult to anticipate how they will apply in any given case, one has to turn to earlier decisions to try and predict the result. In doing so, however, one may sometimes be rather disappointed as the judgments of French courts are extremely terse and the reasoning often distinctly opaque. In particular, the courts do not normally explain how their decision on the facts in front of them relates to decisions in earlier cases, whether they follow or whether they distinguish them, and this means that lawyers often have to guess the decisive elements justifying the different solutions adopted in very similar sets of facts. In other words, many of the concepts and distinctions that actually structure French liability law remain implicit, unstated in either the Code or the courts’ decisions. The question whether this should be changed has been a bone of contention among the various actors interested in the reform of civil law.

II.  The Reform of French Civil Law There have been discussions about reforming the Code civil’s provisions on contract and tort law since the beginning of the twentieth century. However, it was only on the celebration of the bicentenary of the Code in 2004 that the pressure for reform became decisive. A first academic reform project was drafted under the auspices of late Professor Pierre Catala. Known as the Avant-projet Catala, on its publication in 2005 it attracted considerable attention, both among lawyers in France and abroad, and on the part of the French Ministry of Justice, otherwise known as the Chancellerie.9 The draft instrument covered the general rules on contracts, civil liability, the general law of obligations (ie rules applicable to obligations in general, regardless of their source), and the law of prescription. A competing academic project known as the Avant-projet Terré, after the name its initiator, Professor François Terré, was drafted in the years that followed, covering the same subjects, with the exception of prescription law which was reformed by legislation in 2008.10 9 P Catala (ed), Avant-projet de réforme du droit des obligations et de la prescription: Rapport remis au garde des Sceaux (Paris, Documentation française, 2006); translation by J Cartwright and S Whittaker in J Cartwright, S Vogenauer and S Whittaker (eds), Reforming the French Law of Obligations (Oxford, Hart Publishing, 2009). 10 F Terré (ed), Pour une réforme du droit des contrats (Paris, Dalloz, 2009); Pour une réforme du droit de la responsabilité (Paris, Dalloz 2011); Pour une réforme du régime général des obligations (Paris, Dalloz, 2013). The law of prescription was reformed by loi no 2008-561 of 17 June 2008 portant réforme de la prescription en matière civile.

Introduction  7 The Chancellerie started to work on a reform project of its own, based on these two academic proposals. Nonetheless, the process of reform made little progress until 2014, when the government decided to split its intended reform into two stages. The government would first reform the general rules on contract and the general law of obligations by way of an executive ordonnance;11 the reform of the law of civil liability would take place later and would go through Parliament, as is normally the case for legislation of such importance. As regards the first stage, the government believed that it could proceed faster than Parliament, and it also wanted to avoid members of Parliament making unconsidered changes to provisions of the legislative bill drafted by the Chancellerie, especially in the case of highly technical matters concerning the transfer and assignment of claims. ­Moreover, it has traditionally been thought in France that the law of civil liability has a stronger policy component than the law of contract and the general law of obligations, so that it seemed appropriate that Parliament should concentrate on the former. However debatable these justifications may be, Parliament eventually consented to this approach and the reform of the general law of obligations (including contract) was carried out by the Ordonnance of 2016.12 The Ordonnance changed the structure of the Code civil and led to a renumbering of the provisions devoted both to contractual liability and to extra-contractual liability. Articles 1146 to 1153 on contractual liability therefore became articles 1231 to 1231-6, while the famous articles 1382 to 1386 on extra-contractual liability became articles 1240 to 1244.13 However, the substance of these provisions remained unaltered14 and is likely to remain so until the intended wider reform of civil liability law takes place. Two months after the reform of the general law of obligations, in April 2016 the government published a proposal for the legislative reform of the law of civil liability. It organised a consultation on the proposal, inviting all those interested to put forward their views and comments on it. Given the number of comments received, in March 2017 the Chancellerie published a revised proposal, the Projet de réforme de la responsabilité civile, which has been the starting point for this book.

III.  The Main Features of the Projet de Réforme The Projet de réforme consists of a proposal to reform the Code civil by replacing its existing articles 1231 to 1252. The form of the Projet itself consists of a single title 11 An ordonnance is a legislative instrument adopted by the government on a matter on which Parliament normally has jurisdiction, and upon prior authorisation of Parliament: see Constitution of the Fifth French Republic of 4 October 1958, art 38 (as amended). 12 Ordonnance no 2016-131 of 10 February 2016 portant réforme du droit des contrats, du régime général et de la preuve des obligations. 13 Arts 1386-1 to 1386-18 on product liability became arts 1245 to 1245-17. 14 While no changes whatsoever were made to the text of former arts 1382 to 1386, that of arts 1146 to 1153 was slightly modernised, but on the clear assumption that this should not affect their substance.

8  Jean-Sébastien Borghetti and Simon Whittaker containing a single article, paragraph 3 of which then sets out the proposed new Sub-Title of the Code civil on Civil Liability, which would become part of Book III of the Code, devoted to the sources of obligations. This means that the single article of the Projet de réforme contains 83 articles numbered and structured in the way that they would be on amendment of the Code civil. In the following discussions in the chapters of this book it is clearly important to distinguish between provisions of the existing Code civil and provisions of the Code civil as it would be if the Projet de réforme were enacted. At times this is done by referring to draft provisions of the Code civil as proposed by the Projet de réforme, but it is often more convenient to refer simply to the proposed provisions as though they are articles of the Projet de réforme itself. Hence, the Projet’s proposed new article 1233 of the Code civil is referred to simply as ‘article 1233 of the Projet de réforme’ or ‘article 1233 of the Projet’. Where provisions are cited as ‘article 1233 of the Code civil’ or ‘article 1233 Cc’, this refers to a provision in the existing Code civil. In common with the Chancellerie’s earlier proposal in 2016, the Projet de réforme draws heavily on the Avant-projet Catala and, to a lesser extent, on the Avant-projet Terré. Like these two academic projects, the Projet de réforme is mostly intended as a restatement of existing law, and it therefore purports to introduce into the Code civil the major developments that have affected civil liability over the last two centuries, whether as result of the work of courts or the legislator.15 Unfortunately, and unlike the recent Belgian extra-contractual liability reform project,16 the French Projet has been published without any accompanying travaux préparatoires explaining its proposals. This creates some uncertainly as to the exact scope or purpose of some of its provisions. While the Projet de réforme does not go very far in the way of setting out or elucidating the underlying distinctions of the French law of liability,17 it nevertheless seeks to clarify some conceptual issues. For example, it recognises the distinction between dommage (harm) and préjudice (loss), which can already be found in some existing legislative provisions,18 but which is absent from the historic articles of the Code civil.19 The Projet de réforme also follows the ­Avant-projet Terré in recognising that the specificity of liability for the actions of others does not lie in the action which gives rise to this liability (its fait générateur), which has been the way in which it is understood and presented in most textbooks to date, but rather in the attribution of this action to someone other than its author.20 The Projet de réforme is also innovative in the way in which it purports to develop what French lawyers often call the ‘non-compensatory’ functions of civil

15 See,

eg, its treatment of causation, below ch 12. below, ch 19. 17 On its modest explanation of the key notion of la faute see below, ch 5. 18 See eg art 2226 Cc on prescription. 19 On this terminology see below ch 9, ch 10 and ch 21. 20 See below, ch 8. 16 See

Introduction  9 liability, ie the prevention or deterrence of harm and the punishment of those committing acts attracting civil liability who appear as particularly b ­ lameworthy. Article 1266 of the Projet allows courts to order the cessation of an unlawful action, independently of any reparation of loss which may have been suffered.21 Moreover, article 1266-1, perhaps the most discussed provision in the Projet, provides that those who have caused harm by deliberately committing a fault with the view to making a gain or to saving money, can be ordered to pay a civil penalty.22 However, the most remarkable aspect of the Projet de réforme concerns the way in which it handles the relationship between contractual and extra-contractual liability. The Projet distinguishes clearly between the two types of liability and yet at the same time purports to unify the rules applicable to them as much as­ possible.23 This is made apparent in the title of the Projet itself and also in its structure. The title of the Projet de réforme constitutes the first formal recognition of civil liability as a category to include both contractual and extra-contractual liability as it refers simply to la responsabilité civile (civil liability). The sub-title which it creates in the Code civil is divided into six chapters: Introductory Provisions, Conditions of Liability, Grounds of Exoneration or of Exclusion of Liability, the Effects of Liability, Contract Terms Concerning Liability, and the Principal Special Regimes of Liability. While broadly mirroring the typical layout of French textbooks on extra-contractual liability, this structure revolves to a large extent around the relationship between contractual and extra-contractual liability, and seeks to underline the essential unity between them. Chapter 1 organises the relationship between extra-contractual liability and contractual liability. It confirms the well-established rule of non-cumul earlier noted,24 but then in effect reverses it in cases of personal injury where, in general, only extra-contractual liability can apply regardless of whether there exists a contract between the parties.25 This alteration of the non-cumul rule, which has long been advocated by an influential part of legal scholarship, is intended to avoid victims of personal injuries being treated differently depending on their contractual or non-contractual status. Chapter 1 also addresses the issue of liability of a contracting party to third parties in cases of breach of contract, which has been a contentious issue in French law over recent decades.26 Chapter 2, which is devoted to the conditions of liability, contains provisions common to contractual and extra-contractual liability (Section 1), provisions special to extra-contractual liability (Section 2) and provisions special to ­contractual



21 See

below, ch 18. below, ch 16. 23 See below, ch 2, ch 3 and ch 4. 24 Above, p 5. 25 See below, ch 3, esp pp 47–53. 26 See below, ch 4. 22 See

10  Jean-Sébastien Borghetti and Simon Whittaker liability (Section 3). The existence of provisions common to both types of liability is one of the reasons for creating the overarching concept of civil liability. However, the first section contains only six articles, dealing with reparable loss and causal relationship. The provisions special to extra-contractual liability cover the historic regimes set out by the Code civil as enacted or as developed by the courts: liability for fault,27 liability for the action of things,28 liability for abnormal nuisance between neighbours,29 and the various instances of liability for the actions of others, including those created by the Cour de cassation on the basis of article 1384(1) of the Code civil.30 On the other hand, the Projet de réforme does not follow the suggestion of both the Avant-projet Catala31 and the Avant-projet Terré32 of creating a new strict liability regime applicable generally to harm caused by dangerous activities. Lastly, Section 3 of Ch 2 contains three provisions special to contractual liability. Chapter 3 deals with four grounds of exoneration or of exclusion of liability, which are implicitly presented as common to both types of liability, contractual and extra-contractual.33 For reasons that are not entirely clear, contract terms excluding or limiting liability are addressed separately in Ch 5. The basic rule is that contract terms of this kind are valid regardless of the type of liability, even though there are more qualifications on this rule where extra-contractual liability is concerned.34 Chapter 4 is the longest (31 articles) and addresses the effects of liability, ie the way in which the reparation of harm and the compensation of losses should be carried out. The existing Code civil is almost totally silent on this major aspect of the law of civil liability, and the courts have therefore had to develop their own rules, which are to a large extent confirmed by the Projet de réforme.35 The provisions in this chapter are also presented as being common to both types of liability, even if almost half of them are in fact either explicitly36 or implicitly37 restricted to extra-contractual liability. Finally, Chapter 6 deals with the ‘principal special regimes of liability’, ie liability for motor-vehicle accidents and for defective products. The latter regime basically

27 See below, ch 5. 28 See below, ch 7. 29 See below, ch 11. 30 See below, ch 8. 31 Art 1362. 32 Art 23. 33 See below, ch 14. 34 See below, ch 15. 35 Some articles in ch 4 also reproduce existing legislative provisions on the right of recourse of those who have compensated victims for their losses and on the reparation of losses resulting from environmental harm. 36 Arts 1266 and 1266-1. 37 Arts 1267 to 1277, laying out special rules governing the reparation of losses resulting from personal injury.

Introduction  11 implements the European Directive on Product Liability,38 while the provisions on liability for motor vehicle accidents are mostly a restatement of the loi Badinter, combined with a slight broadening of its scope of application and a further tightening of the defences available to defendants.39 The presentation of these two regimes in a special chapter is probably intended to bear the idea that the liability they create transcends the division between contractual and extra-contractual liability. However, this is not the case and, despite their being applicable between parties to a contract, both these liabilities are extra-contractual in character, and so should have been dealt with in Chapter 2, along with the other instances of extracontractual liability. In the end, it is doubtful whether the Projet de réforme would unify contractual and extra-contractual liability as much as its presentation and structure suggest. What is certain, though, is that the division and relationship between these two types of liability will remain as a central feature of French law.

38 The Projet de réforme actually abandons some minor remaining differences between the Directive and its implementation in French law. 39 Arts 1285–1286 of the Projet de réforme.

12

part i ‘Civil Liability’, Contractual and Extra-Contractual

14

2 A Common Framework for Civil Liability? SIMON WHITTAKER

Following the French civil law tradition, the Code civil as enacted in 1804 divided its treatment of obligations between ‘contracts or obligations based on agreement in general’ (Book 3, Title III), and ‘engagements which are formed without agreement’ (Book 3, Title IV), the latter governing ‘quasi-contracts’ (Chapter I) and ‘delicts and quasi-delicts’ (Chapter II).1 Given this tradition, the distinction between contractual and delictual liability was long seen by French lawyers as one of ‘the elementary concepts which are like the ABC of the law’.2 By contrast, the Projet de réforme de la responsabilité civile treats contractual and extra-contractual liability within a single framework of ‘civil liability’ (la responsabilité civile),3 though

1 This division appears to be drawn directly from J Domat, Les loix civiles dans leur ordre naturel (Paris, Coignard, 1689) Pt 1 Book 1 (Des engagemens volontaires & mutuels par les conventions) and Book II (Des engagemens qui se forment sans convention), though the contents of these categories in Domat differ considerably. For the Roman texts, see Gaius’s Institutes 3.88; Digest 44.7.1 pr (attributed to Gaius’s Res Cottidianae); Justinian’s Institutes III.13.2. 2 J-É Labbé, S 1886, 4, 25 (note to Cour de cassation belgique, 8 January 1886). 3 The development of the French terminology here is interesting. In the first half of the ­twentieth century, French writers generally referred to la responsabilité contractuelle (meaning liability in damages for imputable non-performance of a contractual obligation) and to la responsabilité délictuelle (meaning liability arising from the various grounds set out by former arts 1382 to 1386 Cc following the name of Ch II of Title IV of Book 3 as originally enacted, ‘Des délits et des quasi-délits’). The expression la responsabilité civile was first used by jurists instead of la responsabilité délictuelle on the basis that it is less redolent of fault and therefore more appropriate for the range of liabilities imposed without fault: eg, B Starck, H Roland and L Boyer, Obligations, 1. Responsabilité délictuelle, 5th edn (Paris, Litec, 1996) 5 explaining that, properly, la responsabilité délictuelle refers to liability for an intentional fault, whereas la responsabilite quasi-délictuelle refers to liability for lack of care or negligence, but that usually la responsabilité délictuelle refers to all civil liabilities. J Carbonnier, Droit civil, Les biens, Les obligations (Paris, PUF, 2004) vol II, 2251ff (who contrasts la responsabilité délictuelle with la responsabilité contractuelle, but then sets out the law governing the former in terms of la responsabilité civile). It should also be noted that la responsabilité civile designates liability which is civil (ie belonging to droit civil) as opposed to criminal (la responsabilité pénale) or administrative (la responsabilité administrative or de la puissance publique). I will translate la responsabilité as ‘liability’ even though the French word (and its relationship to répondre—‘to answer for’) has a moral overtone absent from the English ‘liability’, as Carbonnier notes: ‘L’homme doit répondre’, citing Genesis 3 v 9 and 4 v 9.

16  Simon Whittaker it still reserves certain rules for one or the other type of liability.4 This paralleling of ‘contractual liability’ and ‘extra-contractual liability’ rests on an understanding that contractual non-performance can be seen as an (admittedly special) example of a fact or event which ‘generates’ liability (un fait générateur de responsabilité) in much the same way as the classic extra-contractual liabilities for personal fault, for the ‘actions of things’ or of certain persons within one’s control, or, more recently, for harm caused by defective products.5 It therefore also assumes that recovery of damages for contractual non-performance can properly be termed ‘contractual liability’ (la responsabilité contractuelle). In this chapter I will explain how the Projet de réforme reflects a compromise between the views of those scholars who argue that civil liability is fundamentally divided between contract and non-contract (the ‘duality thesis’) and those who instead argue that liability is essentially unitary (the ‘unity thesis’). I will then explain how the distinction between contract and tort arose in English law despite its general lack of Roman antecedents and note the significance of this distinction in modern law.

I.  Duality, Unity and Compromise in the Framing of French Civil Liability The roots of the debate as to the relationship between contract and delict in French law can be traced to the last quarter of the nineteenth century, which witnessed a remarkable liveliness in juristic debates as to the nature and purpose of these traditional categories. These debates had conceptual and practical normative aspects. Normatively, many jurists were concerned to establish strict liability as a basis for recovery in respect of accidents causing personal injuries or death in the machine age. This famously led to the ‘discovery’ of liability (without fault) for the ‘actions of things’6 but it also led to the suggestion that liability of an employer for accidents at work could be based on the failure in a contractual obligation on the employer to see that the employee is safe.7 The resulting expansion of liability (contractual and delictual) led to a much greater potential overlap and, ironically, to the recognition of the rule of non-cumul denying the intrusion of delict into the relations between contracting parties.8 In the course of these debates, the question arose as to the proper foundations of civil liability. On one side, there were jurists such as Labbé. In the 1880s he argued for the use of contract as a means of imposing strict liability on employers and also for the 4 Below, pp 21–22. 5 As will be explained, this understanding of damages for contractual non-performance remains controversial: see below, pp 19–20. 6 Former art 1384(1) Cc; Cass civ 16 June 1896, S 1897, 1, 17 note Esmein. 7 See below. 8 This rule was first recognised by the courts in Cass civ 11 January 1922, S 1924, 1, 105 note ­Demogue, but it had been advocated by Labbé in the 1880s: eg note S 1886, 4, 97.

A Common Framework for Civil Liability?  17 denial of any option in claimants to rely on delict against their contracting p ­ artners.9 As we have seen, for him the dichotomy between contractual and ­delictual liability was both traditional and fundamental,10 and this ‘duality’ in liability was reflected in the nature of the two ‘faults’ on which liability rested: contractual fault (la faute contractuelle, that is, imputable non-performance) is ‘special’ and has to be judged according to the expressed ‘wills’ of the parties giving rise to the contract, whereas delictual fault is general. Moreover, Labbé viewed contracts as entailing special and positive duties rather than the general duties of respect required by delict: In law we owe to other people in general respect, but not loyalty which is a matter for morality. By contrast, we can owe those with whom we have contracted our active service, our zeal, and our care undertaken for the furtherance of their interests.11

He noted that the positive law contained different rules governing contractual and delictual liability, notably as to the extent of liability and prescription.12 But this duality thesis did not go unchallenged. As early as 1886, Lefebvre derided la responsabilité contractuelle as ‘a defective expression, an erroneous form of words’13 and argued that all liability is based on delictual fault, in the case of contract this arising from breach of the loi created by the contracting parties.14 This position was then famously adopted by Grandmoulin, who argued that ‘there is not duality but a unity of liability … [and] liability is always delictual’.15 He supported his thesis with a number of arguments, but perhaps the most interesting is that, on non-performance, the original obligations (les obligations primitives) created by a contract are extinguished and instead the law imposes an obligation to pay damages by virtue of article 1382 Cc:16 ‘the infringement of 9 J-É Labbé in a series of notes in Sirey: see esp note to Luxembourg 27 November 1884, Bruxelles 24 April 1885, Bruxelles 28 April 1885, S 1885, 4, 25 and see also S 1885, 2, 201, S 1886, 2, 97, S 1886, 4, 25, S 1886, 1, 5 and RCLJ 1886, 433. This position was also taken by M Sauzet, ‘De la responsabilité des patrons vis-à-vis ses ouvriers dans les accidents industriels’ RCLJ 1883, 596; C Sainctelette, De la responsabilité et de la garantie (Bruxelles/Paris, Bruylant-Christophe, 1884) 115–19 (referring to contractual liability as la garantie). 10 Note to Cass civ Bruxelles 8 January 1886, S 1886, 4, 25 and see above, p 15. The translations in this chapter are the author’s own except in the case of the provisions of the Code civil reformed by the 2016 Ordonnance (which are by B Fauvarque-Cosson, J Cartwright and S Whittaker) and the Projet de réforme (which are by S Whittaker and J-S Borghetti). 11 ibid. 12 This was seen in former art 1150 Cc (limits on contractual liability). Although in the nineteenth century there was a general prescription period of 30 years for all actions (art 2262 Cc as enacted), there were a number of special periods applicable to particular contracts. 13 A-F Lefebvre, ‘De la responsabilité délictuelle, contractuelle – responsabilité du patron envers l’ouvrier’ RCLJ 1886, 485. 14 ibid 486–87, thereby invoking former art 1134(1) Cc. 15 J Grandmoulin, ‘De l’unité de responsabilité’ (Thesis, Rennes, 1882) 2–3. The ‘unity thesis’ is often associated with Planiol, but while he brought it to a wider readership, his own position was more moderate: see esp M Planiol ‘Classification des sources des obligations’ RCLJ 1904, 224, esp 225, 231 and 233 where he argued that obligations arose from two sources: la loi and le contrat. However, on non-performance of the original contractual obligation, a new obligation to pay damages arose which was identical to that which arises on breach of la loi. 16 Grandmoulin, ibid 7.

18  Simon Whittaker a right under an obligation [la créance] is a civil delict (1382) in the same way as the infringement of a property right’.17 The ‘unity thesis’ therefore denied the existence of ‘contractual liability’ as a distinct category and led to the suggestion that, in its turn, delictual liability was based on the non-performance of what Grandmoulin called a ‘general and negative obligation not to harm others’.18 In comparing the two theses, we can see that Labbé did not distinguish between two kinds of obligation, one original and one compensatory, but instead relied on the source of the obligation (two distinct faults) to determine both its classification and its effects, whereas Grandmoulin was concerned with the unity of liability. However, the unity thesis was not generally accepted either by legal scholarship or the courts and by the late 1920s had been abandoned,19 but the debate as to the proper identity and role of contractual liability revived towards the end of the twentieth century, and in two different ways. First, there was an increasing sense of the similarity in function of ‘civil ­liability’, inspired by the considerable expansions of liability, contractual (notably by obligations de sécurité, obligations d’information and actions directes) and noncontractual (notably, by liability for the ‘actions of things’), and by the recognition of new ‘mixed’ categories of liability (such as professional liability and product liability) which cut across this boundary. In the first edition of her influential work on the conditions of liability published in 1982, Viney therefore argued that it is ‘useless and even harmful’ to seek to force all the examples of liability into the two categories of ‘contractual’ and ‘delictual’.20 In order to avoid an ‘excessive fragmentation’ of the law into different regimes, it is important: to maintain as broad and unitary a regime of le droit commun [the law applicable in principle] as possible by reducing the differences of principle existing between contractual and delictual liabilities to the minimum required by respect for the contract and to reject derogations from this general set of rules except where they are genuinely demanded by the special character of the situation or the activity which gives rise to liability … To put this another way, it seems to me that, in future, the distinction between contractual and delictual liabilities is to lose its importance in favour of another distinction which now tends to become increasingly prominent between the ‘general law’ or ‘law applicable in principle’ and special regimes of civil liability.21

This vision shares with Grandmoulin’s thesis the idea that liability resulting from contractual non-performance and from an extra-contractual wrong is fundamentally the same phenomenon, and argues for the harmonising of their incidental features where ‘respect for the contract’ (essentially, I think, what the parties have agreed together with any incidents required by the nature of contract



17 ibid

3. 1. 19 L Josserand, note to Req 14 Dec 1926, DP 1927, 1, 105. 20 G Viney, Les obligations, La responsabilité: conditions, 1st edn (Paris, LGDJ, 1982). 21 ibid 299–300. 18 ibid

A Common Framework for Civil Liability?  19 in general or the particular contract) requires otherwise. On the other hand, it also ­acknowledges that the ‘general law of liability’ has become less important as the special regimes of liability (for road accidents, defective products, medical accidents, environmental damage etc) have grown in number and significance: the traditional distinction between contract and non-contract is therefore increasingly eclipsed by a distinction between the general law and the special law. Of course, the distinction between ‘general’ and ‘special’ laws is long and well-established within the French law of contract as special rules govern particular categories of contracts (both in terms of their subject-matter and of the status of their contracting parties) qualifying the position governing contracts generally.22 However, this movement towards the assimilation of contractual and extra-contractual liability into a wider category of ‘civil liability’ has not gone unchallenged. For towards the end of the last century a vocal group of scholars argued that ‘contractual liability’ is a false concept, thereby echoing Grandmoulin’s unity thesis according to which liability for contractual non-performance was delictual rather than ‘contractual’23 but drawing very different conclusions.24 For example, Le Tourneau and Cadiet argue that the source of the recovery of damages in the contractual context is the contract itself, rather than the non-performance of the obligation, and that what is usually called réparation is in reality: a mode of performance of the contract, different no doubt from the one which was foreseen given that it is by equivalent and often late, but a mode of performance just the same, or a mode of satisfaction [paiement] since satisfaction … is the performance of an obligation whatever the latter’s subject-matter.25

The (original) contractual obligation therefore is reflected in this substitute for performance and this explains the rule that the party in default should be liable only for foreseeable losses: ‘the equivalent should not go further than the promise’.26 This also means that damages awarded for contractual nonperformance must not extend beyond providing a monetary substitute for the

22 On the relationship of the French general and special laws of contract see S Whittaker, ‘Contracts, Contract Law and Contractual Principle’ in J Cartwright and S Whittaker (eds), The Code Napoléon Rewritten, 29, 34–38. It has been argued that the 2016 reforms sought to re-establish the significance of general contract law which had lost ground to special regimes: S Chassagnard-Pinet, ‘La réforme du droit commun des contrats – le vocabulaire’ Lextenso étudiant (2016) [2]. 23 Above, pp 17–18. 24 See esp P Rémy, ‘La “responsabilité contractuelle”: histoire d’un faux concept’ RTD civ 1997, 323; D Tallon, ‘L’inexécution du contrat: pour une autre presentation’ RTD civ 1994, 223; and recently, P le Tourneau and M Poumarède, ‘Pour en finir avec la responsabilité contractuelle …’ in Études en l’honneur du professeur Jérôme Huet (Paris, LGDJ, 2017) 269; O Deshayes, ‘La responsabilité contractuelle’ in P Jourdain and M Bacache (eds), Comparaison de la réforme du droit français de la responsabilité civile avec le nouveau Code civil roumain (Paris, IRJS, 2018) vol 1, 279. 25 P le Tourneau and L Cadiet, Droit de la responsabilité et des contrats (Paris, Dalloz, 2000/2001) 197 (original emphasis). On the last sentence, cf art 1342 Cc which provides that ‘le paiement est l’exécution volontaire de la prestation due’ (emphasis added). 26 ibid 255.

20  Simon Whittaker ‘benefit ­promised [by the debtor] and not received’:27 other losses caused by contractual non-performance may attract reparation only on the basis of extra-contractual liability.28 This approach was certainly arguable, given that the provisions of the Code civil governing awards of damages for contractual non-performance were expressed in language which did not include reference to the party in default being ‘­responsible’ for the losses thereby caused.29 Moreover, it has the merit of emphasising the special function of damages for contractual non-performance, that is, that they can provide a substitute for performance itself, but as a theory it has not gained general support among French scholars, not least because it does not fit with the assumptions of most jurists, nor with the practice of the courts for well over a century, both of which happily use the expression ‘contractual liability’ and accept that such a liability should in principle compensate all losses incurred by the creditor as a result of non-performance except those which were not foreseeable. Moreover, for some scholars, putting aside the case where the contractual obligation is itself to pay a sum of money, the idea that an award of money (damages) can be said to be ‘equivalent’ to performance undermines the important moral stance taken by French law which sees enforced performance in kind (l’exécution forcée en nature) as the primary means of giving effect to the binding force of contract, and even opens the way to the idea of ‘efficient breach’!30 From this perspective, the significance of la responsabilité contractuelle is that the contractual debtor must answer for (répondre) his given word and must bear the harmful consequences of his failure to respect his promise.31 In this way, again, the overall purpose of damages for contractual non-performance and damages arising from a ground of extra-contractual liability are assimilated, even while recognising that the source of the liability and some of its incidental rules may properly differ. To what extent are these various views reflected in the treatment of civil liability after the reform of the Code civil in 2016 and in the proposed Projet de réforme, both in terms of structure and in terms of the rules which it provides? As to the structure, the new Title III of Book III of the Code civil is ‘On the Sources of Obligations’ and its introductory provisions (articles 1100–1100-2) announce that ‘[o]bligations arise from juridical acts, juridically significant facts or from the sole authority of legislation’. It makes clear that juridical acts may be based on agreement (contract) or unilateral,32 while ‘[j]uridically significant facts [les faits juridiques] consist of behaviour or events to which legislation attaches legal consequences’33 and the obligations which arise from them are governed 27 ibid. 28 ibid 198. 29 Arts 1142, 1146–48 Cc. 30 M Fabre-Magnan, Droit des obligations, 1 – Contrat et engagement unilateral, 4th edn (Paris, PUF, 2016) 757–58. 31 ibid 758. 32 Arts 1100 and 1100-1 Cc. 33 Art 1100-2 Cc.

A Common Framework for Civil Liability?  21 ‘according to the circumstances, by the sub-title relating to extra-contractual liability or the sub-title relating to other sources of obligations’.34 The reformed law therefore states clearly that the sources of obligations do differ as between contract (the main example of juridical act) and extra-contractual liability (the first example of an obligation arising from a fait juridique, for example, the defendant’s personal fault or the action of a thing in his or her keeping).35 On the other hand, while the Ordonnance of 2016 changed the law governing the four other ‘sanctions’ for non-performance (the defence of non-performance, enforced performance in kind, price reduction, and termination),36 its changes to the provisions governing damages contained in a subsection ‘Reparation of loss resulting from non-performance of the contract’ were only superficial and, in particular, they continued to express the law without referring to any ‘responsibility’ in the party in default.37 In its Report to the President of the Republic on the Ordonnance of 2016, the Ministry of Justice explained that: Contractual liability [la responsabilité contractuelle] cannot be reformed in isolation from extra-contractual liability: it is generally accepted that, fundamentally, these two forms of liability are mechanisms of the same nature, resting on the existence of an action giving rise to liability [un fait générateur], harm, and a causal relationship between the two. Only differences of regime distinguish between them, based essentially on the special character of the action giving rise to liability in the contractual context …38

On these grounds, the Ministry left reform of contractual liability to a later wider reform of civil liability, contractual and extra-contractual.39 The Ministry’s proposals in the Projet de réforme reflect this intention. We notice first the change in heading of Sub-title II, from la responsabilité extracontractuelle (in the present Code civil) to la responsabilité civile and the fact that the latter can include both contractual and extra-contractual liability is immediately made clear by its first chapter, which explains the relationship between these two categories of liability, for example, denying in principle an option to a contracting party to rely on extra-contractual liability against the other party.40 Chapter II then sets out the ‘conditions of liability’, two of these being common to contract and noncontract: ‘reparable loss’41 (articles 1235–38) and the causal relationship between the defendant’s act and the harm42 suffered by the claimant (articles  1239–40).

34 Art 1100-2(2) Cc. 35 A fait juridique may also be found in the ‘purely voluntary actions’ which give rise to quasicontracts: see arts 1300–1303-4 Cc. 36 Arts 1217–1230 Cc. 37 Arts 1231–1231-7 Cc. 38 Report to the President on Ordonnance of 2016, 21. 39 ibid. 40 Art 1233 of the Projet de réforme. See also arts 1233-1 and 1234 of the Projet de réforme. 41 Le préjudice reparable. 42 On ‘harm’ and its relation to ‘loss’ see below, ch 9.

22  Simon Whittaker However, the Projet then distinguishes between extra-contractual liability (articles 1241–49, listing the actions which give rise to extra-contractual liability) and contractual liability according to which ‘[e]very non-performance of a contract which has caused harm to the creditor gives rise to an obligation in the debtor to be liable for it’, providing the traditional restriction of recovery to foreseeable harms and setting the requirement of notice to perform for damages for delay in performance.43 The following chapter (III) sets out the grounds of exoneration or exclusion of liability and these in principle govern both contractual and extra-contractual liability, though its common rule governing the effect of force majeure notes that the definition of the latter differs as between contractual and extra-contractual matters.44 In this way, the Projet de réforme provides a common framework for the two categories of liability, but distinguishes between them where it considers this normatively appropriate. For present purposes, though, most significant are the two preliminary provisions in Section I’s treatment of the ‘principles’ (explicitly so-identified)45 of Chapter IV on the effects of liability. Article 1258: The aim of reparation is to replace the victim as much as is possible in the situation in which he would have been if the harmful action had not taken place. It must cause him neither a loss nor an advantage. Article 1259: Reparation may take the form of reparation in kind or damages, these two types of measures being able to be combined so as to ensure full reparation of the loss.

Unlike some other provisions in Chapter IV (which themselves provide that they apply only to ‘extra-contractual matters’),46 articles 1258 and 1259 apply both to contractual and extra-contractual liability and they make clear that these liabilities share an aim in the ‘reparation’ of loss,47 whether this takes effect in kind (en nature) or by compensation (by way of damages). Having said that, the terminology used by the Projet de réforme is interesting and, to an extent, changes subtly. In setting out the conditions of liability, it distinguishes between extra-contractual liability where the heading of the relevant section uses the classic expression ‘an action giving rise to extra-contractual liability’ (le fait générateur de responsabilité extracontractuelle),48 whereas in the case of contractual liability, non-performance is not termed a fait générateur,49 article 1250 of the Projet de réforme stating simply 43 Arts 1250–1252 of the Projet de réforme. 44 Art 1253 of the Projet de réforme referring to the rule for contractual liability in art 1218 Cc. 45 Arts 1258–1266-1 of the Projet de réforme. 46 Art 1266 (cessation of unlawful action) and art 1266-1 (civil penalty) of the Projet de réforme. 47 cf art 1235 of the Projet de réforme. 48 Sous-Titre II, Ch II, Section 2, Sous-section 1. 49 It is, however, possible properly to refer to contractual non-performance as a fait générateur of liability, see, eg P Stoffel-Munck, ‘La singularité de la responsabilité contractuelle – Articles 1250 à 1252; 1253 alinéa 3 et 1263’ JCP G Suppl to nos 30–35 (25 July 2016) 30, [6] in relation to art 1250 of the Avant-projet de réforme de la responsabilité civile; preliminary note by G Viney to art 1363 of the Avant-Projet Catala (‘Le fait générateur de la responsabilité contractuelle est l’inexécution qui s’apprécie en fonction de la portée de l’engagement’).

A Common Framework for Civil Liability?  23 instead that ‘every non-performance of a contract which has caused harm to the creditor gives rise to an obligation in the debtor to be liable for it [à en répondre]’.50 This can be compared with article 1258, which sets out the purpose of reparation for both types of liability by referring to the hypothetical situation as if the ‘­harmful action’ [le fait dommageable] had not taken place, that is an event or action (including contractual non-performance) which has caused harm. Here, then, the assumption is surely that the ‘harmful action’ is also an ‘action which gives rise to liability’ with the difference that it has also satisfied the other two conditions of liability, ie reparable loss and a causal relationship between the ‘action attributed to the defendant’ (again, an expression which is neutral as to contract and non-contract).51 Moreover, article 1250 of the Projet de réforme uses the language of obligation at a secondary level, that is, at the level of liability: any non-performance of a contract which has caused harm to the creditor obliges the debtor to answer for it: this recognises an ‘obligation to answer for [obligation à répondre]’ non-performance of a contractual obligation. By contrast, the provisions of the Projet de réforme which set out the ‘actions giving rise [fait générateurs]’ to extra-contractual liability (articles 1241–49), avoid using the explicit terminology of obligation, despite the famous precedent provided by the Code civil as enacted. So, for example, while the Code civil originally provided that ‘any human action whatsoever which causes harm to another creates an obligation in the person by whose fault it occurred to make reparation for it’ (‘oblige celui par la faute duquel il est arrivé, à le réparer’),52 article 1241 of the Projet de réforme states that ‘a person is liable for the harm caused by his fault’ (‘[o]n est responsible du dommage causé par sa faute’). This new wording suggests that there is responsabilité without using the language of obligation. Moreover, there is, of course, no primary level of obligation in the French sense in the schema of French extra-contractual liability. For while extracontractual fault is explained by the Projet de réforme in terms of ‘a violation of a legislative requirement or a failure in the general duty of care or diligence’ (‘devoir général de prudence ou de diligence’),53 the reference to a ‘duty’ is significant as it does not bring with it the bilateral relationship of an ‘obligation’, with its combination of rights (droit de créance) and debts (la dette). The substantive provisions on the effects of liability in Chapter IV apply in p ­ rinciple to both types of liability and give expression to a defendant’s ­responsabilité primarily by an order for reparation in kind (articles 1260–61) or for damages (articles 1262–64).54 We can say, therefore, that the effects of

50 This is the first provision under Section 3 (‘Dispositions propres à la responsabilité contractuelle’). See also the new art 1231 inserted by art 1(2) of the Projet de réforme. 51 Arts 1235 and 1239 of the Projet de réforme. 52 Former art 1382 Cc and see identically new art 1240 Cc. 53 See ch 5 below. 54 On the former see ch 18 below, p 383. Provision is also made for the imposition of an amende civile, on which see ch 16 below.

24  Simon Whittaker the ­obligation à ­répondre imposed by law (article 1250) on non-performance of a contractual obligation seeks to ‘repair’ the consequences of this nonperformance, a position reminiscent of the views of Grandmoulin.55 On the other hand, the Projet de réforme recognises that there should be some differences in the rules governing contractual and extra-contractual liability. It therefore adopts the classic rule that a party to a contract is liable only for the consequences of contractual non-performance which were reasonably foreseeable at the time of the formation of the contract,56 whereas no such restriction applies to a person’s extra-contractual liability; the definition (and, indeed, role) of force majeure differs as between contractual and extra-contractual liability;57 and, conversely, in ‘extra-contractual matters’, a court ‘may prescribe reasonable measures appropriate to prevent harm or to see that an unlawful nuisance to which a claimant is exposed is stopped’ and also may in certain circumstances order the payment of a civil penalty, whereas these possibilities do not exist in relation to contractual liability.58 As will be seen, for an English lawyer, perhaps the most important distinction between liability for breach of contract and in tort concerns the measures of damages.59 The measure in contract is (typically and specially) to put the injured party in the same position as if the contract had been performed, reflecting the injured party’s ‘performance interest’, whereas the measure in tort is to put the injured party in the same position as if the tort had not been committed, reflecting the injured party’s status quo ante interest. In turning to French law, for long this distinction, or ones like it, were not discussed in French treatises on the law of obligations which, moreover, included relatively little discussion of the nature, measure or purposes of awards of damages.60 This can be explained by the relatively few provisions in the Code civil of 1804 governing damages for contractual non-performance and by the relative roles of the Cour de cassation and the lower courts, the juges du fond. For the Code civil did not concern itself with the identification of the ways by which damages for contractual non-performance should be measured. Instead article 1149 Cc (now article 1231-2) simply adopted the traditional distinction between ‘the loss which [the creditor] has incurred’ (or damnum emergens) and ‘the gain of which [the creditor] has been deprived’ (or lucrum cessans), which is not the same as the common law distinction between the reliance and performance measures, as while lucrum cessans does necessarily reflect a performance measure, damnum emergens may reflect either a performance or a reliance measure.61 Article 1149 Cc was interpreted as supporting a general 55 Above, pp 18–19. 56 Art 1251 of the Projet de réforme. 57 Art 1253 of the Projet de réforme, referring (as a matter of contract law) to art 1218 Cc. 58 Arts 1266 and 1266-1 of the Projet de réforme. 59 Below, p 32. 60 B Nicholas, The French Law of Contract, 2nd edn (Oxford, Clarendon Press, 1992) 225. 61 Nicholas (n 60) 226. Also G Viney, P Jourdain and S Carval, Traité de droit civil, Les effets de la responsabilité, 4th edn (Paris, LGDJ, 2017) 438.

A Common Framework for Civil Liability?  25 principle of ‘full reparation’ of all losses suffered by the creditor (le principe de la réparation intégrale)62 which, aside from the rule as to the foreseeability of the harm for contract,63 sets the single legislative rule governing recovery of contractual damages.64 Moreover, both jurists and courts have accepted that this principle applies as much to extra-contractual as to contractual liability,65 while the relative lack of legislative provision governing damages has encouraged the Cour de cassation to hold that the juges du fond have a very wide ‘sovereign power of assessment’ of the claimant’s loss and, therefore, of the damages recoverable.66 However, for well over a decade, French scholars have identified and emphasised the significance of the distinction between performance and reliance interest damages, typically borrowing the German terminological equivalent of ‘positive’ and ‘negative’ interests.67 In particular, Laithier advocated the use of this distinction, explaining that the ‘positive interest’ can be defined as the interest of the creditor in the proper performance of the contractual obligation by the debtor, to place the former in the same ‘patrimonial’ situation in which he would have been if the obligation had been completely and punctually performed.68 By contrast, the ‘negative interest’ is the one which the creditor would have had if he had not engaged in the contractual process and includes expenses inherent in the conclusion of the contract.69 Even so, it remains true that the distinction between positive and negative interest has no basis in the Code civil. Would this position change if the Projet de réforme were enacted? Under the Projet de réforme, the provisions at present in the last section of the sub-title on contractual non-performance and all the provisions on extra-contractual liability are abrogated, and Sub-Title II would be renamed La  responsabilité civile rather than La responsabilité extra-contractuelle treating contractual and extra-contractual liability in a common framework. In this respect, as earlier noted, article 1258 of the Projet de réforme states that: The aim of reparation is to replace the victim as much as is possible in the situation in which he would have been if the harmful action had not taken place. It must cause him neither a loss nor an advantage.

62 Eg C Larroumet and S Bros, Les obligations, Le contrat, 7th edn (Paris, Economica, 2014) 813. 63 Former art 1150 Cc; new art 1231-3 Cc. 64 Viney, Jourdain and Carval (n 61) 153ff. P Malaurie, L Aynès and P Stoffel-Munck, Droit des obligations, 8th edn (Paris, LGDJ, 2016) 565–66 state that this ‘principle dominates reparation in money for harm resulting from non-performance of a contract’. 65 Viney, Jourdain and Carval (n 61) 154–55. 66 ibid 176ff. 67 See, in particular, Y-M Laithier, Étude comparative des sanctions de l’inexécution du contrat (Paris, LGDJ, 2004); T Genicon, La résolution du contrat pour inexécution (Paris, LGDJ, 2007); Z Jacquemin, ‘Payer, réparer punir, Étude des fonctions de la responabilité contractuelle en droit francais, allemande et anglais’ (Thesis, University of Paris II, 2015) and J-S Borghetti, ‘Faut-il distinguer les dommages et intérêts compensatoires et les dommages et intérêts en lieu et place de la prestation?’ RDC 2016, 787. The debate is helpfully considered by Viney, Jourdain and Carval (n 61) 433ff. 68 Laithier (n 67) [110]. 69 ibid [125]–[128].

26  Simon Whittaker We can see that this formal expression of the principle of full reparation70 ­abandons reference to ‘losses incurred’ and ‘gains of which one has been deprived’ and even uses the notions of la perte (loss) and le profit (advantage or gain) to make expressly the point that reparation must be full but it should not put the injured party in a better position.71 However, what is particularly interesting is that when applied to contractual and extra-contractual liability, the formula in article 1258 of the Projet generates different measures of damages. In the case of extra-contractual liability, the purpose of reparation is to put the injured party as far as possible in the situation he would have been if the fait dommageable, ie the fait générateur which has caused loss, had not taken place. This leads to what English law would call the status quo ante measure.72 In the case of contractual liability, though, this same test leads to a measure of damages to put the injured party (the creditor) in the position as though the contractual non-performance which has caused harm had not taken place, ie as though the contract had been performed.73 Article 1258 could therefore be seen as a legislative recognition of the performance measure in contract damages, though of course it could be more explicit and refer to the proper performance of the contract.74 While to an English lawyer this is entirely welcome, to a French lawyer it can appear inaccurate as it does not reflect the situation where damages are combined with termination for contractual nonperformance (la résolution), where the logic of termination with restitution and counter-restitution argues against recovery of the ‘lost bargain’ and instead permits only damages to put the injured party in the position as though the contract had not been made.75 Overall, therefore, the framework for civil liability proposed by the Projet de réforme represents a compromise between those who affirm the fundamental unity 70 Eg F Leduc, ‘Les règles générales régissant la réparation du dommage’ JCP 2016 Suppl to nos 30–35 (25 July 2016) 36, 37 (in relation to the Avant-projet de loi). 71 This formula has been used by the Cour de cassation, eg Cass civ (2) 28 May 2009, Bull civ II no 131 glossing over former art 1147 Cc by reference to ‘le principe de la réparation intégrale du préjudice sans perte ni profit’. 72 This is indeed, also, the expression used by Leduc (n 70). 73 Deshayes (n 24) 284 observes that the two measures for contract and delict share the purpose of putting the creditor in the position in which he ought to have been and concludes that they share an identical function. 74 Leduc (n 70) 41 suggests as an alternative formulation that ‘[e]n matière contractuelle, la réparation doit tender à placer le créancier dans la situation qui eût été la sienne si le contract avait été dûment exécuté’. Similarly, Borghetti (n 67) [3]: ‘[c]hercher à englober les deux hypothèses [of the measures for contract and non-contract] dans un concept unique de réparation, comme le fait l’article 1258 [of the avant-projet de réforme], n’aide malheureusement pas à bien saisir la function et l’enjeu des règles proposées’. 75 Leduc (n 70) and see also M Fabre-Magnan, ‘Termination of Contract: A Missed Opportunity for Reform’ in J Cartwright, S Vogenauer and S Whittaker (eds), Reforming the French Law of Obligations, ch 8, 175–76 in relation to the Avant-projet Catala and the criticisms of this approach by S Whittaker, ‘“Termination” for Contractual Non-performance and its Consequences: French Law Reviewed in the Light of the Avant-projet de réforme’ ibid, ch 9, 200–03. See further T Genicon, La résolution du contrat pour inexécution (Paris, LGDJ, 2007) 721ff.

A Common Framework for Civil Liability?  27 of contractual and extra-contractual liability and those who consider that there may well be good reasons to distinguish between them for particular purposes. This reflects the position of Viney – which is not surprising, given her prominent role in the Avant-projet Catala, whose general structure of ‘civil liability’ the Projet de réforme adopts.76 In doing so, the first provision on the principles governing the effects of liability (contractual and extra-contractual) appear at first sight to give no more than a legislative expression to the established principle of la réparation intégrale, but they may well have a particular importance in terms of the measure of recovery for contractual non-performance.

II.  The Distinction between Contract and Tort in English Law While legislation has affected the content of the rules governing certain aspects of contract law and certain types of contract and has amended or replaced certain torts, the English law governing contract and tort does not have any general legislative basis, codified or otherwise. This area of English law remains typically common law in the sense of the law developed by the courts, but it is also common law in the sense of owing relatively little to the Roman civil law tradition. There was certainly no general reception into English law of the Roman law of obligations, its rules developing instead incrementally by way of particular actions recognised by the courts in relation to contract and torts.77 Moreover, the absence of an overall legislative framework means that modern English law does not possess the kind of formal structure in, or relationship between, the categories of liability visible in the Code civil as promulgated, after its reform in 2016 or after the further reforms proposed by the Projet de réforme. More generally, in looking at the law as exposed in English treatises on contract and tort (which is typically found in separate works), while the law of contract has been subjected to a degree of classical ­reconstruction,78 the law of torts still represents English law at its most gothic. Given all this, we face something of a puzzle: why do English lawyers distinguish between ‘contract’ and ‘tort’ (or perhaps contract law and the law of torts) as overarching legal categories in a recognisably similar way to the civil law pattern seen in the Roman institutional works and reflected in classical French law, even

76 Viney chaired the working sub-group on la responsabilité civile in the Avant-projet Catala and wrote the exposé des motifs to the relevant section of the avant-projet itself, ie arts 1340–86. For her introduction to the provisions in the Avant-projet Catala see J Cartwright, S Vogenauer and S W ­ hittaker (eds), Reforming the French Law of Obligations (Oxford, Hart Publishing, 2009), Appendix (with an English translation by J Cartwright and S Whittaker) 808ff. 77 D Ibbetson, A Historical Introduction to the Law of Obligations (Oxford, Oxford University Press, 1999) Pt 3 and esp 153–54. 78 See AWB Simpson, ‘Innovation in Nineteenth-century Contract Law’ (1975) 91 LQR 247.

28  Simon Whittaker though the contents of those legal categories (especially as regards torts) are very different? It is particularly puzzling given that what became the main action sanctioning breach of contract (the action of assumpsit) emerged from the ‘action for trespass on the case’, that is, tort.79 The answer is complex and historical, and involves a shift in understanding of the distinction from form to substance combined with a growing theoretical sense of the voluntary character of contracts. By the eighteenth century, English courts had recognised a series of personal actions which a plaintiff could bring in the context of civil wrongs as very broadly understood. For example, if A had promised B something and had failed to keep the promise, B could (depending on the circumstances) bring an action of debt (for a sum certain), an action in covenant (if the promise was made under seal) or an action for damages on the promise (known as assumpsit) if the promise was supported by ‘consideration’.80 Moreover, if A had harmed B in some way quite apart from any contract (for example, by reason of a trespass to his person, land or goods, by defrauding him, by defaming him, or (in certain circumstances) by negligently causing him injury or damaging his property), then the law would grant him an action for damages. In this respect, the distinction between contract and tort became significant in two ways. First, in 1609 in Pinchon’s Case the Court of King’s Bench decided that the action of assumpsit (which later became the main contractual action) could lie against an executor, avoiding the rule against the transmissibility of actions on death which governed both trespass and the action for debt.81 This rule which distinguished in practice between contract and tort remained law until 1934.82 Secondly, by the middle of the eighteenth century, courts and legal writers began to distinguish between personal actions ex contractu or upon contract and those ex delicto or upon a tort.83 To begin with, the significance of this distinction was fairly marginal, even for a legal author concerned with classification such as B ­ lackstone (who mentioned the distinction, referring to the Institutes of ­Justinian),84 but by the end of the century, there were a number of procedural rules (such as joinder of the parties to litigation and joinder of ‘counts’, ie the formal bases of a claim) which turned on the distinction between actions ‘in form ex contractu’ and ‘in form ex delicto’.85 This principally procedural character of the distinction supported the view that a plaintiff could in principle choose which action to bring, 79 ibid ch 7. 80 Ibbetson (n 77) Pt 3, 153ff. 81 Pinchon’s Case (1609) 9 Co Rep 86b. 82 Law Reform (Miscellaneous Provisions) Act 1934 s 1. 83 An early example may be found in Denison v Ralphson (1682) 1 Vent 365 in the context of joinder of ‘causes upon contract’ and ‘upon a tort’. 84 Sir William Blackstone, Commentaries on the Laws of England, 1st edn (Oxford, Clarendon Press, 1765–69), Book III, ch 8, 117 quoting J Inst 4.6.1 and not referring to the procedural significance of the distinction. 85 J Chitty (the elder), A Practical Treatise on Pleading and on the Parties to Actions and the Forms of Actions, 1st edn (London, W Clarke & Sons and J Butterworth, 1809) vol I, ch 2, 87, which sets out the forms of actions ex contractu and ex delicto.

A Common Framework for Civil Liability?  29 whether in contract or in tort, subject only to the necessary conditions for the relevant action being present.86 For example, a passenger injured while travelling by stagecoach could choose to sue the carrier in tort (on ‘the common custom of the realm’ or in ‘case upon negligence’) or in contract (in assumpsit for breach of a contractual duty) on the basis that the tort was a wrong independent of contract.87 In Brown v Boorman in the 1840s, the House of Lords even appeared to hold that a plaintiff could always sue in tort in the alternative to breach of contract.88 There, the plaintiff sued his commodity broker in case upon negligence (tort) on the ground that he had sold his goods and delivered them without payment to the wrong (and later bankrupt) person, but the broker countered that the claim should have been brought in contract (assumpsit). In the Exchequer Chamber, Tindal CJ held that the action brought in tort could go ahead, observing that: the action of case upon tort very frequently occurs where there is a simple nonperformance of the contract … The principle in all these cases would seem to be that the contract creates a duty and the neglect to perform, or the nonfeasance, is a ground of action in tort.89

In the House of Lords, Lord Campbell expressed similar views.90 To a modern eye, these dicta appear to allow a claim in tort whenever there is a breach of contractual duty, a position redolent of the arguments of French writers such as ­Grandmoulin,91 but one should not forget the procedural context both of the distinction between contract and tort and also of the issue in Brown v Boorman, and indeed both judges clearly saw the broker’s defence as a highly technical plea. Certainly, the decision was seen by later writers or courts merely as authority for recognition of an option between breach of contract and an independent tort established on the facts.92 However, on the fundamental reform of civil procedure in the middle of the nineteenth century the distinction between contract and tort changed its ­significance from form to substance. The new law created ‘a simple and uniform process for commencing actions’ and the old procedural rules attaching to the distinction between actions in a form tort and in a form contract were swept away so as to avoid ‘the numberless unnecessary formalities, which pervaded every suit’.93 The distinction between actions ‘founded on contract’ and those

86 Brown v Dixon (1786) 1 TR 274; Ansell v Waterhouse (1817) 6 M & S 385. This position was contested: Orton v Butler (1822) 5 B & Ald 652 (rejecting joinder of counts in contract and tort); Powell v Layton (1806) 2 Pos & Pul New Rep 365 (no option to sue in case instead of a claim for nondelivery of goods under a contract). 87 Ansell v Waterhouse (n 86) 390. 88 (1842) 3 QB 511; (1844) XI Clark & Finnelly 1. 89 (1842) 3 QB 511, 526. 90 (1844) XI Clark and Finnelly 1, 44. 91 Above, pp 17–18. 92 eg Hyman v Nye & Sons (1882) 6 QBD 685, 689; Baylis v Lintott (1873) LR 8 CP 345 and, more recently, Midland Bank Ltd v Hett, Stubbs & Kemp (a firm) [1979] Ch 384, 410, 420, 431–32. 93 JCFS Day, The Common Law Procedure Acts, 3rd edn (London, H Sweet, 1868) ii.

30  Simon Whittaker ‘founded on tort’ was used to determine new, if relatively minor, procedural questions, such as those governing costs and in applying this distinction the courts considered the substance of the plaintiff ’s claim, distinguishing between a breach of contract and a ‘wrongful act’ arising without such a breach.94 Moreover, wider changes were taking place.95 Some English lawyers such as Sir Frederick Pollock developed a taste for classification and were concerned with what they perceived as the substance of legal concepts.96 With the decline of trial by jury in civil matters traceable to the reforms of the mid-century,97 more issues were decided by a judge alone and this led to issues being seen as legal rather than factual. Finally, English lawyers increasingly saw contracts as based on the intentions (or ‘wills’) of the contracting parties, and this underlined the difference between contractual obligations – which were voluntary – and tort duties, which were not, a development inspired to a considerable extent by influences from the civil law.98 Gradually the law itself also changed, with liability in tort for negligence in particular growing in ambit and in importance. As courts desired to expand liability beyond the boundaries of privity of contract, the drawing of a sharper line between contract (to which privity applies) and the law of torts (to which it does not) became more attractive. Indeed, in this light the freeing of liability in tort for negligence from the ‘privity fallacy’ by reference to the independence of tort from contract was the most important aspect of the famous decision in D ­ onoghue v Stevenson.99 Overall, the change from a procedural and formal to a substantive understanding of the distinction between contract and tort and the release of liability for negligence in tort from imprisonment within the citadel of privity of contract by appealing to the independence of tort from contract both enhanced the importance of the distinction between the two categories themselves.100 What, then, is the modern position as to the overall relationship between contract and tort in English law? First, the position expressed by Winfield in 1931 remains the orthodox ­starting-point.101 In his view, liability for breach of contract is distinguished from liability in tort in that: (i) the duties in tort are primarily fixed by the law while in contract they are fixed by the parties themselves; and 94 Bryant v Herbert (1878) 3 CPD 189 & 398, 390, 392 seeing a claim for detinue (ie in substance founded on the claimant’s right of property) as ‘founded on tort’ as it was not ‘founded on contract’. Bramwell LJ (at 392) referred to the old law governing the classification of personal actions according to their form as ‘useless, and worse than useless, learning’. 95 See further S Whittaker, ‘The Relationship between Contract and Tort: a Comparative Study of French and English Law’ (Thesis, University of Oxford, 1987) ch 10. 96 Sir F Pollock, The Law of Torts, 1st edn (London, Stevens and Sons, 1887) 1 and see also ch XIII and Appendix A, Historical Note on the Classification of the Forms of Personal Actions by FW Maitland. 97 Day (n 93) xii–xiii. 98 PS Atiyah, The Rise and Fall of Freedom of Contract (Oxford, Oxford University Press, 1979) 408. 99 Donoghue v Stevenson [1932] AC 562, 568–69, 594, 596–97, 610, thereby disapproving the earlier restrictive approach in Winterbottom v Wright (1842) 10 M & W 109. 100 cf WL Prosser, ‘The Fall of the Citadel’ (1966) 50 Minn L Rev 791, 817. 101 PH Winfield, Province of the Law of Tort (Cambridge, Cambridge University Press, 1931) 380.

A Common Framework for Civil Liability?  31 (ii) in tort the duty is towards persons generally while in contract it is towards a specific person or persons.102 This approach still holds true as a general statement, but it is usually qualified in a number of ways.103 In particular, as Winfield allows, contractual obligations are not all negotiated or even agreed as they may be imposed by the law directly or via ‘implied terms’ whereas, conversely, some tortious duties (notably, those under the ‘broad principle of assumption of responsibility’ in Hedley-Byrne) contain a voluntary element. How, though, do these general statements relate to the substance of the modern law? First, while there is indeed a general law of contract applicable to all types of contract and guided by the overarching principles of freedom of contract and the binding force of contract, the English law of torts lacks the kind of broad and general bases of liability which are found in French law and instead consists of a series of particular, ‘nominate’ torts, each with their own structure, rules and focus.104 This remains true even after the considerable expansion of liability in the tort of negligence over the last half-century or more, as the duty of care concept allows the courts to define the situations in which people can or cannot be liable for harm caused by their negligence. In this respect, there are two key features, for liability for (pure) omissions and liability for pure economic loss remain ­exceptional in the tort of negligence, outside a handful of recognised situations (such as the liability of professionals such as lawyers, accountants, surveyors etc). This very considerably reduces the potential area for overlap of claims in tort and for breach of contract, where claims for pure economic loss are universally recognised and remain typical. Unlike French law, where extra-contractual liability for fault is unrestricted as to type of harm or context, and therefore potentially all-engulfing of contract, in English law, liability in tort for negligence is restricted as to type of harm and context. Secondly, there are a number of differences in substantive rule depending on whether liability is based on a tort and on breach of contract.105 These include the rule that a contractual right, for example, to a certain sum due under a contract, can generally be assigned, but a right of action in tort generally cannot;106 rules of the conflict of laws governing both jurisdiction107 and applicable law;108 and the 102 See similarly Robinson v P E Jones (Contractors) Ltd [2011] EWCA Civ 9, [2011] BLR 206 [79] (Jackson LJ with whom Stanley Burnton and Maurice Kay LJJ agreed). 103 See H Beale (gen ed) Chitty on Contracts, 33rd edn (London, Sweet and Maxwell, 2018) vol 1, [1-154]–[1-156] (S Whittaker). 104 See also ch 22 below, pp 456–457. 105 Chitty on Contracts (n 103) [1-157]–[1-160] for a summary. 106 ibid [1-160] and [19-050]. 107 Regulation (EU) 1215/2012 on jurisdiction and the recognition and enforcement of judgments in civil and commercial matters (recast) (‘the Brussels I bis Regulation’) [2012] OJ L351/1 arts 7(1) and 7(2). 108 See in particular Regulation (EC) 593/2008 on the law applicable to contractual obligations (‘Rome  I’) [2008] OJ L177/6; Regulation (EC) 864/2007 applicable to non-contractual obligations (‘Rome II Regulation’) [2007] OJ L199/40.

32  Simon Whittaker law governing the capacity of minors who are in principle liable for their torts, but only to a limited extent on their contracts.109 Moreover, although actions for breach of contract and in tort must both in general be brought within a limitation period of six years, the period begins to run ‘from the date on which the cause of action accrued’ and this may vary as between the two.110 However, the most important differences between liability in tort and for breach of contract concern the rules governing damages.111 These include differences in the types of losses which may be compensated in damages and of remoteness of damage, but the fundamental difference lies in the measures of damages as these reflect the different nature of the interests protected by the two categories of liability. In tort the measure of damages is both retrospective and negative. According to Lord Blackburn in Livingstone v The Rawyards Coal Co (which concerned a claim for damages for the tort of trespass to land), there is a general rule that, where any injury is to be compensated by damages, in settling the sum of money to be given for reparation of damages you should as nearly as possible get at that sum of money which will put the party who has been injured … in the same position as he would have been in if he had not sustained the wrong for which he is now getting his compensation or reparation.112

This is negative in the sense that the damages are to make up for the worsening of the claimant’s position by the defendant’s civil wrong; it is retrospective in that its focus is the position as it would have been if the wrong had not taken place, sometimes called the status quo ante. In the case of contract, the typical measure protects the ‘performance’ or ‘expectation’ interest of the injured party: its primary function is to put him as far as possible in the position in which he would have been had the contract been performed; whereas the function of damages in tort is to put the injured party in the position in which he would have been if the tort had not been ­committed.113 The performance measure in contract reflects a general understanding that contracts are intended by the parties to change and (hopefully, at least at the time of the conclusion of the contract) to improve the positions of the c­ ontracting 109 Chitty on Contracts (n 103) [9-053]–[9-054]. 110 Limitation Act 1980 ss 2 & 5. There is a 12-year period for claims brought on a ‘specialty’ (ie promises contained in a deed, s 8) and a three-year special period for actions in respect of personal injuries which applies explicitly to actions in contract as well as in tort: s 11. On the date of accrual see Battley v Faulkner (1820) 3 B & Ald 288; Pirelli General Cable Works Ltd v Oscar Faber & Partners [1983] 2 AC 1 and Chitty on Contracts (n 103) [28-031]–[28-032]. 111 McGregor on Damages, 20th edn ((London, Sweet & Maxwell, 2017) [22-001]ff; A Burrows, ­Remedies for Torts and Breach of Contract, 3rd edn (Oxford, Oxford University Press, 2004) ch 2. 112 Livingstone v The Rawyards Coal Co (1880) 5 App Cas 25, 39. 113 Robinson v Harman (1848) 1 Exch 850, 855 (contract); Livingstone v Raywards Coal Co (1880) 5 App Cas 25, 39 (tort); Morris-Garner v One Step (Support) Ltd [2018] UKSC 20, [2018] 2 WLR 1353 at [31] (Lord Reed contrasting the two measures). See generally McGregor on Damages (n 111) [22-001]ff; Burrows (n 111) ch 2; and (on the contract measure) D Friedmann ‘The Performance Interest in Contract Damages’ (1995) 111 LQR 628.

A Common Framework for Civil Liability?  33 parties; and damages for breach of contract (being the primary remedy of breach of contractual obligations other than to pay money)114 are seen as a way of allowing the injured party to obtain a substitute for performance. This is not to say that the injured parties cannot recover out-of-pocket expenses (often called ‘­reliance damages’ based on a ‘reliance interest’),115 that is, to put them in the position as though the contract had not been made, but they cannot avoid a bad bargain by claiming reliance damages, as the party in breach may defeat such a claim by establishing that the losses so claimed would have been incurred even if the contract had been performed.116 In this sense, the performance interest is primary: an award of damages should not put the injured party in a better position than if the contract had been performed.117 Either way, it will be seen that the performance interest is both positive and prospective, looking to the position as though the primary obligations of the party in breach had been performed. Very recently, in Morris-Garner v One Step (Support) Ltd Lord Reed explained the performance measure in contract damages by reference to the well-known distinction between primary and secondary obligations put forward by Lord Diplock in Photo Production Ltd v Securicor Transport Ltd,118 to the effect that: a contract is the source of primary legal obligations upon each party to it to procure that whatever he has promised will be done is done. Leaving aside the comparatively rare cases in which the court is able to enforce a primary obligation by decreeing specific performance of it, breaches of primary obligations give rise to ‘substituted or secondary obligations’ on the part of the party in default. Those secondary obligations of the contract breaker arise by implication of law: ‘The contract, however, is just as much the source of secondary obligations as it is of primary obligations.’119

In Lord Reed’s view, ‘[d]amages for breach of contract are in that sense a substitute for performance. That is why they are generally regarded as an adequate remedy’.120 This raises several points of comparison with the French approach. Most obviously, Lord Diplock and Lord Reed adopt a similar two-fold analysis of contractual non-performance and its effects as we have seen in the Projet de réforme,121 114 These are remedied by the action for the agreed contract sum available at common law and as of right. 115 This terminology became prominent following the article by L Fuller and WR Perdue, ‘The ­Reliance Interest in Contract Damages’ (1936) 46 Yale LJ 52 and 373. 116 C & P Haulage v Middleton [1983] 1 WLR 1461; CCC Films (London) Ltd v Impact Quadrant Ltd [1985] QB 16; Omak Maritime Ltd v Mamola Challenger Shipping Co [2010] EWHC 2026 (Comm), [2011] 2 All ER (Comm) 155 esp at [65]. 117 Flame SA v Glory Wealth [2013] EWHC 3153 (Comm). 118 Photo Production Ltd v Securicor Transport Ltd [2018] UKSC 20 at [34] referring to [1980] AC 827, 848–49. Lord Diplock had earlier used the distinction to explain the proper approach to the construction of suretyship contracts after termination by the creditor of the principal contract: Moschi v Lep Air Services Ltd [1973] AC 331, 350. 119 [2018] UKSC 20 at [34] quoting Lord Diplock in [1980] AC 827, 849. Lady Hale PSC, Lords Wilson and Carnwath JJSC agreed with Lord Reed. 120 [2018] UKSC 20 at [35]. 121 Above, pp 22–23.

34  Simon Whittaker but they reject one of its possible consequences in that they consider that the ­secondary obligations arising on breach of contract find their source in the contract ‘just as much’ as do the primary obligations, even though the former are imposed ‘by implication of law’. In their view, this is tied to the fundamental function of contract damages, namely to act as a substitute for performance itself. This is reminiscent of those French jurists who deny the idea of ‘contractual liability’ on the basis that contractual damages are merely a sort of satisfaction or ‘discharge by performance’ (paiement),122 but there are two differences. The first is that the English analysis does not lead either to a rejection of the terminology of ‘liability’ as regards contract damages nor to the idea that damages for breach of contract are restricted to recovery of a substitute for the object of the contracting party’s main obligation (in French terms, the prestation): the ‘performance interest’ measure allows, in principle, recovery of damages to put the injured party in the position as though the contract had been performed, so that, for example, damage caused to existing interests of the injured party (personal injury or damage to property) are equally covered. The second difference is with the relationship to specific enforcement or ‘enforced performance in kind’ of the contract’s primary obligations. In English law, specific enforcement of contractual obligations (other than to pay money),123 remains exceptional and discretionary and this is explained by Lord Reed in terms of the ‘adequacy’ of damages allowing recovery of a substitute for performance. By contrast, under the reformed Code civil, a party faced with contractual non-performance (the creditor) may claim ‘forced performance in kind’ (l’exécution forcée en nature) from the party in default unless this is impossible or there is a manifest disproportion between its cost to the debtor in good faith and its interest for the creditor.124 As we have seen, for some French jurists, this primacy of performance by the debtor reflects the moral basis of contractual obligation and argues against the idea that damages for contractual non-performance are themselves a substitute for performance itself.125 Returning, though, to English law, the contrast in the formulae for the recovery of the performance interest typical of contract and the status quo ante interest in tort may be misleading in that they do not compare like with like. For if one applies the tort measure to the ‘wrong’ on which a claim in contract is based (the breach of contract), this leads to the performance measure of contract damages: it is to put the injured party in the position as though the breach of contract had not taken place, that is, as though the contract had been performed. This suggests that the difference in the nature of awards of damages for breach of contract and in tort does not lie in the formulae used to identify them (their formal measures),

122 Above, p 19. 123 Monetary obligations are specifically enforced by the action for the agreed sum and are available at common law and as of right. This remedy (the modern successor to a claim in debt) is seen as distinct from a claim for damages. See, notably, White and Carter (Councils) Ltd v McGregor [1962] AC 413. 124 Art 1221 Cc (as amended in 2018). 125 Fabre-Magnan, above, p 20.

A Common Framework for Civil Liability?  35 but rather in the nature of the interest ‘wronged’. In a claim in tort, in principle the claimant seeks redress for harm caused to an interest (be it, for example, in personal integrity, reputation, property) which predates and exists independently of any action by the defendant or dealing between the parties. In a claim for breach of contract, the claimant is seeking a substitute for what the defendant agreed (or was otherwise obliged) under the contract to do for the claimant: in the special case, the claimant’s interest prejudiced by breach was itself created by the contract.126 Having said this, one should not be misled by broad judicial dicta such as Lord Blackburn’s quoted above127 into thinking that English law recognises a general principle of ‘full reparation’ in the sense in which it is recognised by French law and explicitly stated in the Projet de réforme.128 Indeed, Lord Blackburn himself acknowledged that his statement of principle ‘must be qualified by a great many things which may arise’.129 This is as true of contract as of tort, though the qualifications are not identical. So, unlike French law, there are rules excluding certain types of harm from recovery depending on the nature of the claim (notably, for mental distress or lost reputation in claims for breach of contract); different rules of remoteness of damage applicable both to contract and tort and differing ­treatments of the relevance of contributory negligence in the claimant. However, the English law governing damages for breach of contract has developed in ways which have nuanced, but also expanded the practical calculation of the performance interest. So, the classic rules excluding damages for distress or loss of reputation have been qualified where the nature of the contract or of the particular contractual obligation have so required;130 the time at which the injured party’s loss is to be quantified has been held to be generally (but not always) the date of breach;131 where the party in breach has performed defective services or property, the injured party may, in principle, recover the difference in value or, if there is none, the cost of bringing the result of performance up to standard, but not where this would be unreasonable given the relative cost to the party in breach of doing so and the interest of the injured party in conforming performance;132 and, most

126 See to similar effect Morris-Garner (n 113) [31] (Lord Reed). It should be noted that in some cases at the borderline between contract and tort the mere classification of liability as contractual may not affect the measure of damages: see, eg, in the context of liability for pre-contractual statements (on the basis of breach of warranty or in tort) Esso Petroleum Ltd v Mardon [1976] QB 801 as explained by Chitty on Contracts (n 103) [1-168]. 127 Above, p 32. 128 Art 1258 of the Projet de réforme. 129 (1880) 5 App Cas 25, 39, giving the example (in the context of trespass) of a distinction between the innocent and the malicious trespasser. 130 Jarvis v Swann Tours Ltd [1973] QB 233 (damages for disappointment in relation to a ­holiday); Ruxley Electronics and Construction Ltd v Forsyth [1996] AC 344; Farley v Skinner (No 2) sub nom ­Skinner v Farley [2001] UKHL 49, [2002] AC 732 (loss of amenity); Malik v Bank of Credit and Commerce International SA [1998] AC 20. 131 Chitty on Contracts (n 103) [26-088]. 132 Ruxley (n 130).

36  Simon Whittaker recently, the Supreme Court has clarified the proper approach to the recovery of ‘negotiating damages’, that is, the amount which a claimant could reasonably have demanded as a quid pro quo for the relaxation of the obligation in question.133 The measures of damages in English law, whether for tort or contract, are seen as a starting-point for the assessment of the claimant’s award; there is a lot more law to be considered before the award can be determined.

III. Conclusion Both the structure and contents of the Projet de réforme reflect a compromise between the two contrasting visions of the formal relationship between contractual and non-contract associated with the dualist and unity theses. It accepts the idea of ‘contractual liability’ and this allows it to set out a number of issues common to liability arising on contractual non-performance and on the extracontractual grounds of liability. However, the Projet recognises that some rules should govern only extra-contractual liability or, as the case may be, contractual liability. As noted at the beginning, there is not the same need for an English lawyer to set out the overall relationship between contractual and tortious liability as there is with the French codified system of civil law, but I think that the compromise position reached by the Projet would accord reasonably well with the working understandings of liability in contract and tort used by most English lawyers. However, there is one point on which the Projet is, from the English perspective, too laconic. Although article 1258 of the Projet de réforme can be interpreted as allowing French law to distinguish between the interests protected by awards of damages in tort and in contract, it remains true that, despite the arguments of French scholars, the new law would not make clear the special and ‘positive’ ­character of the creditor’s performance interest in contract damages. If this difference in the interests protected of claimants in contract and in tort is indeed central, then it should surely be clearly expressed in the reformed French Code civil.



133 Morris-Garner

(n 113) esp [95].

3 The Relationship between Contractual and Extra-Contractual Liability as between Parties to a Contract YVES-MARIE LAITHIER

The relationship between contractual and extra-contractual liability is ‘among one of the most complex and finely-balanced questions of the law of obligations’.1 As is often the case, this complexity results to a large extent from history, in this case in the way in which the law of civil liability and the law of contracts have been elaborated over the centuries. In this respect, the history of the broadly equivalent English law is particularly interesting to French jurists since the action of assumpsit – which allowed an injured party to obtain compensation for losses caused by breach of a promise – originated in the law of torts, even though modern English law is presented (especially in the writings of continental scholars) as distinguishing carefully between the consequences of the commission of a tort and of breach of contract.2 Such an apparent paradox incites particular interest. Having said this, in French law the ups and downs of history do not explain everything. The acute difficulty in determining the relationship between contractual and extra-contractual liability in a clear and accurate way can be explained by a wide range of factors including the refinement of modern conceptual analyses, the complexity of the factual situations in which this relationship arises (for example, where carrying out the same commercial operation depends on the performance of a series of distinct contracts), the growth of special legislative schemes of liability, and various and often contradictory policies. In this paper, I will consider my topic around three key notions: the ‘contracting parties’; ‘­contractual liability’ (la responsabilité contractuelle); and the ‘relationship’ which the latter bears with extra-contractual liability. Let me explain each in turn. 1 P Wery, Droit des obligations, 2nd edn (Bruxelles, Larcier, 2011) [619]. In the following, ‘­extra-contractual liability’ and ‘delictual liability’ will be used synonymously. 2 See ch 2 above, pp 27–28. See also AWB Simpson, A History of the Common Law of Contract. The Rise of the Action of Assumpsit (Oxford, Clarendon Press, 1975) 273ff.

38  Yves-Marie Laithier First, my topic’s reference to the ‘contracting parties’ logically excludes c­ onsideration of the relationship of contracting parties with third parties, whether in terms of the liability which such a third party may bear if he makes himself an accessory to breach of a contract or in terms, conversely, of the liability which a party owing a duty under a contract (a contractual ‘debtor’) may bear towards a third party. In practice, drawing the boundary between what is or is not a matter for contract often gives rise to considerable hesitation,3 whether one is concerned with the existence of a contract (notably, in cases of the wrongful breaking-off of precontractual negotiations),4 with the application of the concepts of a ‘party’ or a ‘third party’ (given that in French law a ‘party’ is not restricted to persons who concluded the contract as, for example, in the case of universal successors of title, such as an heir, or assignees),5 or with the content of the contract. Secondly, the topic rests on a distinction between ‘contractual liability’ and ‘extra-contractual liability’. This distinction is usual, but a small proportion of French legal scholars would see it as a biassed presentation of French law in that it accepts as correct a distinction which, for them, is conceptually mistaken. For these scholars, the proper dividing line is between the law of civil liability (which is necessarily extra-contractual) and the law of contract. Thirdly, I cannot explore here all the points in common, nor all the differences between contractual and extra-contractual liability,6 and will instead focus on an examination of their relationship. In theory, this relationship could be fixed in a number of different ways, with the rules of one or other system of liability being excluded, combined together or made the subject of a choice between them. As is well-known, the principle accepted by the Cour de cassation is that, as regards the relationship between contracting parties,7 the creditor does not have the possibility of relying against the debtor on the rules of delictual liability:8 any option is therefore firmly excluded. In this way, as between the parties, this rule termed non-cumul (no accumulation) determines the relationship between contractual and extra-contractual liability. Although its working in practice in the hands of the courts always gives rise to a number of issues,9 the principle of non-cumul provides a direction for them – this is, indeed, one of its merits. The principal justification of non-cumul is the desire not to undermine the agreement of the contracting parties and in particular the extent of the risks 3 For a profound account see G Viney, Traité de droit civil, Introduction à la responsabilité, 3rd edn (Paris, LGDJ, 2008) [181]ff. 4 On the relevance of contractual liability in this situation see J Ghestin, G Loiseau and Y-M Serinet, Traité de droit civil, La formation du contrat, 4th edn (Paris, LGDJ, 2013) vol 1, [739]. 5 See eg J Flour, J-L Aubert and E Savaux, Les obligations. 1, L’acte juridique, 16th edn (Paris, Sirey, 2014) [431]ff. 6 On the special character of contractual liability in the Avant-projet de réforme de la responsabilité civile (2016), see P Stoffel-Munck, ‘La singularité de la responsabilité contractuelle – Articles 1250 à 1252; 1253 alinéa 3 et 1263’ JCP G Suppl to nos 30–35 (25 July 2016) 30. 7 Cass com 9 July 2002 no 99-19156. 8 See eg Cass civ 11 January 1922, in F Terré, Y Lequette and F Chénedé, Grands arrêts de la jurisprudence civile, 13th edn (Paris, Dalloz, 2015) [182]. 9 Viney (n 3) [226]ff.

The Relationship between Contractual and Extra-Contractual Liability  39 which the debtor has agreed to bear in undertaking his obligations:10 in this way the refusal of an option to the creditor protects the function and effectiveness of contracts as un acte de prévision, that is to say, a juridical act by which the parties can plan and make provision for their future relationship.11 Moreover, this protection is particularly necessary in French law, given the breadth of the rules of extra-contractual liability.12 Unlike other laws, the French law of extra-contractual liability is not built on a system of special nominate torts, and the rules on which it is founded are extraordinarily general, whether they govern its conditions or its effects. This creates a far from theoretical risk that the contractual regime may be circumvented to the advantage of the extra-contractual regime. As Ancel rightly observes: it is difficult to imagine how one could do without [the principle of non-cumul] as this would allow the law of delictual liability to open the way to a right to reparation of harm in a majority of cases of contractual non-performance.13

Without denying this justification for non-cumul (which in my view is convincing), could it be that prohibiting an option is related to the nature of contractual liability in other ways? In other words, has the debate about the duality of the two orders of liability had any impact on the primacy of the contractual regime ensured by the rule of non-cumul? Or, instead, does this primacy in fact depend on other factors? These two questions will be addressed in the following sections in turn.

I.  The Duality of the Two Categories of Civil Liability: A Recurring Controversy Apart from some brief periods of calm, for more than a century the analysis of the consequences of non-performance of a contractual obligation has been a subject of controversy within French legal scholarship. The terms of the debate have changed, but the distinctiveness of the rules governing contracts has always stood at the centre of the argument. For a long time the controversy turned on the question of whether the regime governing contractual liability was distinct from

10 See eg P Brun, Responsabilité civile extracontractuelle, 3rd edn (Paris, LexisNexis, 2014) [106]; P Malaurie, L Aynès, P Stoffel-Munck, Les obligations, 10th edn (Paris, Defrénois, 2018) [1006]; Viney (n 3) [216]. See also J-S Borghetti, ‘La responsabilité du fait des choses, un régime qui a fait son temps’ RTD civ 2010, 1, [58]ff where the author explains the interactions between the rule of non-cumul, the generalisation of liability for the actions for things and the appearance of obligations de sécurité de moyens (contractual obligations to take care of safety of the other contracting party). 11 See F Terré, P Simler, Y Lequette and F Chénedé, Les obligations, 12th edn (Paris, Dalloz, 2018) [897], esp at 973. 12 See also J-S Borghetti, ‘L’articulation des responsabilités contractuelle et extracontractuelle – ­Article 1233 alinéa 1er et 1234’ JCP G Suppl to nos 30–35 (25 July 2016) 15, [7]. 13 P Ancel, ‘Le concours de la responsabilité délictuelle et de la responsabilité contractuelle’ RCA 2012, 2, dossier no 8, [11].

40  Yves-Marie Laithier the one governing delictual liability and whether, as a result, it was independent from the latter.14 The emphasis which was placed on the source of an obligation (on the will of the parties in the one, on legislation in the other) today gives these initial discussions a rather dated air. However, the analytical frameworks which were used (whether resting on a single obligation or instead on an original obligation followed by a secondary obligation) remain of interest. More recently, from the 1980s and 1990s, it has been the concept of ‘contractual liability’ (la  responsabilité contractuelle) which has been denounced. This criticism is presented as more radical, but its purpose is exactly the same, that is, to preserve the special character of contract by liberating it from ways of thinking borrowed from extra-contractual liability. I shall first provide an overview of the debate so as to allow us to consider its possible effects on the position of parties to a contract.

A.  An Overview of the Debate Although the supporters and opponents of the notion of ‘contractual liability’ ­disagree profoundly, in fact neither has adopted an extreme position. On the one hand, no one doubts the existence of extra-contractual liability nor in particular the right of a party to a contract to sue the other contracting party on this basis if the conditions of contractual liability are not satisfied, for example, because the harm which was caused by that other party does not result from a failure to perform his contractual obligations15 or because the contract has been annulled as a result of dol (fraud or other dishonesty). On the other hand, and conversely, no one maintains that ‘contract is dead’ and has been entirely absorbed by delictual liability.16 The special character of the regime governing contractual non-performance has certainly been played down, perhaps at times rather too much, but never to the point of denying its existence or wiping away the ancient distinction between contract and delict to the advantage of the latter. There are, however, a number of approaches between these two pole positions, reflecting the more measured views within French legal scholarship. Briefly, the following are the three main ones. A first approach understands civil liaibility as a notion which is split into two  branches: contractual liability on one hand and everything else on the 14 On the theses of C Sainctelette and J Grandmoulin and their respective influence see Viney (n 3) [162]ff and also ch 2 above, pp 16–18. 15 See eg Cass com 24 October 2018 no 17-25672 (where it was decided in relation to the principle of non-cumul that ‘this principle only forbids the creditor of a contractual obligation from relying on the rules of delictual liability against the debtor of this obligation, and does not forbid the bringing of a distinct claim based on article L 442-6.I.5° of the Code de commerce, which is aimed at the reparation of a loss resulting not from a contractual failure but from the brutal breaking-off of an established commercial relationship’ (emphasis added)). In French law, the ‘brutal breaking-off ’ of an established commercial relationship (now art L 442-1.II of the Code de commerce) is seen as a fault distinct from non-performance of a contract. 16 cf G Gilmore, The Death of Contract (Columbus, Ohio State University Press, 1974) 95ff.

The Relationship between Contractual and Extra-Contractual Liability  41 other, that is to say, extra-contractual liability. Non-performance of a (primary) ­contractual obligation is analysed as a legally significant action or event (un fait juridique) which creates a (secondary) obligation to make reparation for the harm which flows from it. Moreover, such an obligation to make reparation for harm satisfies the definition of civil liability at least as regards its essential function. This analysis does not require the contractual regime to be completely alligned with the extra-contractual regime, though the former borrows the latter’s characteristic features. Whether as regards delict or contract, the purpose of reparation is to place the claimant in the position: in which he ought to have been … This identity of function allows … contractual and extra-contractual liability to be viewed as two branches of the same tree of civil liability.17

This conclusion is exactly the same as that reached by the second analysis, set out first by le Tourneau18 and later by Rémy in a well-known article.19 These authors denounce the intellectual and juridical encroachment of civil liability on voluntary engagements and argue that the tendency to assimilate the regime governing contractual non-performance with the regime governing delictual liability should be eliminated. In their view, ‘contractual liability’ does not exist: it is a ‘false concept’. If a contractual debtor fails in his obligation and the creditor therefore claims damages, the latter represent the value which should have been (but was not) provided to the creditor. Damages are a monetary substitute for the benefit to which the creditor was entitled. Since their role is not to make reparation for harm, they can be called ‘substitutionary damages’ or, following the most usual French expression (which has the same meaning), ‘performance by equivalent’. For those authors who adopt the third analysis, the controversy which has just been explained is based on a false dichotomy. They maintain that in the contractual context, damages are not exclusively either a form of reparation or instead performance by equivalent, but they may be one or the other, depending on the particular contractual obligation which has been broken.20

17 O Deshayes, ‘La responsabilité contractuelle’ in P Jourdain and M Bacache (eds), Comparaison de la réforme du droit français de la responsabilité civile avec le nouveau Code civil roumain (Paris, IRJS, 2018) vol 1, 279, esp 284–85, who adds that, for this reason, art 1258 of the Projet de réforme is well-drafted. For an opposing view which regrets the lack of clarity in this provision and its failure to distinguish between the meaning of ‘full reparation’ in the contractual and in the extra-contractual contexts, see F Leduc, ‘Les règles générales régissant la réparation du dommage – Articles 1258 à 1264’ JCP G Suppl to nos 30–35 (25 July 2016) 36, [7]. See also ch 2 above, pp 25–27. 18 See the successive editions of his work, Droit de la responsabilité civile, since 1982 (Paris, Dalloz). 19 P Remy, ‘La “responsabilité contractuelle”: histoire d’un faux concept’ RTD civ 1997, 323. 20 See J Huet, ‘Responsabilité contractuelle et responsabilité délictuelle. Essai de délimitation entre les deux ordres de responsabilité’ (Thesis, University of Paris II, 1978). See also Z Jacquemin, ‘Payer, réparer, punir, Etude des fonctions de la responsabilité contractuelle en droits français, allemand et anglais’ (Thesis, University of Paris II, 2015) [144]ff, who distinguishes between damages in lieu of performance (whose function is satisfaction (paiement)) and compensatory damages (whose function is reparation).

42  Yves-Marie Laithier In my view, it is helpful to add that, while this debate is expressed in general or even abstract terms in the French legal literature, in fact a general policy broadly in favour of compensation for personal injuries always lies in the background and guides the views put forward, whether overtly or not. Moreover, it may well be that this discussion has lost or is about to lose its raison d’être, and this for at least two reasons. The first is that, even though the Cour de cassation cannot be unaware of the controversy, it continues to subject awards of damages in the contractual context to proof of harm caused by a fault. Thus, to take just a single example, the Cour de cassation has recently criticised a court of appeal which, having found that a contract for plant-hire had been broken off in an unfair and brutal manner, had ordered the defending party to pay a sum of €7,000, even though it had found that the claimant had not established any economic loss. Its reasons for doing so are telling. The Cour de cassation first declared that ‘contractual fault does not necessarily in itself imply the existence of a harm caused by this fault’ and it added that ‘damages cannot be granted except where in giving judgment the court finds that a loss resulted from contractual fault’.21 In other words, even if it is shown that a debtor has failed in his contractual obligations, the creditor cannot obtain damages unless he establishes harm at the date of the court’s giving judgment. Harm is not essentially identical with nonperformance: the two elements are distinct. The regime governing contractual damages is therefore based on reasoning drawn from civil liability. The second reason is that the Projet de réforme formally presented by the Minister of Justice in March 2017 adopts a dualist structure.22 In the Projet, civil liability appears as a genus of which contractual and extra-contractual liability are two species and are subject to common rules (giving some support to the unitary approach to civil liability) as well as to their own particular rules. In this way, the reform proposes that the Code civil should include the notion of contractual liability, and therefore reflects a dualist approach even though this is not itself explicitly stated. Moreover, in reading the Projet, it becomes apparent just how much extracontractual liability remains the point of reference both at an intellectual and normative level.23 Nevertheless, the critics of the concept of contractual liability have not given up the fight. Certainly, they concede, ‘contractual liability’ appears by that name in the Projet de réforme; and it is also true that article 1250 of the Projet, which provides that ‘[e]very non-performance of a contract which has caused harm to the creditor gives rise to an obligation in the debtor to be liable for it’, is evocative both in its form and the terms which it uses of what is now article 1240 Cc, according to which ‘[a]ny human action whatsoever which causes harm to another creates an 21 Cass civ (1) 22 November 2017 no 16-27551. 22 See P Brun, ‘La distinction des deux ordres de responsabilités dans le projet de réforme de la responsabilité civile’ in Etudes en la mémoire de Philippe Neau-Leduc (Paris, LGDJ, 2018) 125, esp 126; Terré, Simler, Lequette and Chénedé (n 11) [897]. See also ch 2 above, pp 21ff. 23 See eg art 1233-1(1) of the Projet de réforme.

The Relationship between Contractual and Extra-Contractual Liability  43 obligation in the person by whose fault it occurred to make reparation for it’.24 But, they add, this formal recognition of contractual liability is more apparent than real: the terminology is there, but not the substance. In reality, they observe, the Projet suppresses the very elements on which contractual liability is founded. In support of this assertion, they cite in particular the fact that non-performance of a contractual obligation is not depicted as a fault, the disappearance of the distinction between obligations de moyens (obligations to take care) and obligations de résultat (obligations of accomplishment), the absence of any reference to liabilities for the actions of things or of other people, and the fact that obligations de ­sécurité (safety obligations) are ‘decontractualised’. Overall, therefore, in their view the Projet would allow a return to: a kind of original purity. A contractual liability for the action of non-performance of a contractual obligation which will no longer be a special regime of extra-contractual liability since their domains can no longer overlap, but an entirely separate regime whose conditions and domains are distinct from extra-contractual liability.25

This last observation is crucial. In law, conceptual rigour is not an end in itself: the life of the law relies on practical know-how, rather than abstract science. Those who oppose the concept of ‘contractual liability’ therefore take pains to underline the practical significance attached to the position which they defend.

B.  The Wider Repercussions of this Debate In the case of the relationship between parties to a contract, the debate about the nature of contractual liability possesses two wider repercussions. The first concerns its domain. The question can be formulated in the following way: in the case of non-performance of a contractual obligation, does the way in which contractual liability is conceptualised modify its relationship with extra-contractual liability? In the present state of the law, this leads one to ask if the principle of non-cumul should be questioned. I will return to this issue later.26 The second repercussion concerns the regime of contractual liability and, in particular, its effects. This is, indeed, the main point on which the critics of contractual liability insist: in the case of non-performance of a contractual obligation, damages are not a form of reparation but are instead a mode of performance of the contract. But this proposition has a practical significance only if the principles for the assessment of damages depend on the nature of contractual liability – which is debatable. This therefore raises a further series of questions.

24 See P le Tourneau and M Poumarède, ‘Pour en finir avec la responsabilité contractuelle …’ in Etudes en l’honneur du Professeur Jérôme Huet (Paris, LGDJ, 2017) 269, esp 270–71. 25 P le Tourneau (ed), Droit de la responsabilité et des contrats, Régimes d’indemnisation, 11th edn (Paris, Dalloz, 2017) [3213.33]. 26 See section II.B below, p 51ff.

44  Yves-Marie Laithier Should we treat contractual damages as a matter of performance or of reparation for non-performance? For the opponents of the concept of contractual liability, contractual damages do not make reparation for a harm but instead constitute: a mode of performance of the contract, imperfect given that it is no doubt different from what was foreseen in that it is by equivalent and often late, but a mode of performance just the same, or a mode of satisfaction [paiement] since satisfaction … is the performance of an obligation whatever the latter’s subject-matter.27

The purpose of damages is therefore to liquidate the contractual obligation which has been broken. Since in its ‘original form’ a contract is a ‘promise of a designated benefit’,28 damages represent the equivalent in money of the benefit that the complete and timely performance of the obligation which has been broken would have procured for the creditor. This equivalent is not limited to the value of the subject-matter of the broken obligation (its prestation); it also includes consequential damages as long as they are direct and foreseeable. Putting this another way: [p]erformance by equivalent cannot be arbitrarily limited to the equivalent of the prestation which was promised but not performed; it ought to extend to the equivalent of the creditor’s forseeable interest in performance of the contract; this is the measure of enforced satisfaction by equivalent [le paiement forcé par équivalent], the only one that can fulfill the obligation, or, if one prefers, give effect to the creditor’s right.29

Understood in this way, the notion of performance by equivalent does not lead to any different results from reparation. It is true that the technical path differs as reparation rests on the identification of an obligation to compensate arising from the legally significant fact (le fait juridique) of imputable non-performance of the contract by the debtor, but the result is appreciably the same. In terms of principle, there is no fundamental difference between the aim of ‘performance by equivalent’, which is to restore the creditor by means of money to the position which should have resulted from the punctilious performance of the contractual obligation, and the aim of civil liability, which is to replace the contractual creditor in the position in which he should have been if the harmful act had not taken place. Moreover, and most significantly, as soon as the foreseeable consequences of non-performance of the prestation are taken to form part of the notion of performance by equivalent, the distinction between performance and reparation becomes particularly tenuous. Without doubt this degree of overlap may be explained by reference to the fact that the benefit promised to the creditor is not limited to the supply of the prestation but includes all the advantages which he expected to gain from it. It remains the case, though, that this broad conception of the notion of performance by



27 le

Tourneau (n 25) [3213.42] (original emphasis). (n 19) 354. 29 le Tourneau and Poumarède (n 24) 285–86. 28 Remy

The Relationship between Contractual and Extra-Contractual Liability  45 equivalent blurs its line with the notion of reparation which equally covers losses consequential on non-performance or defective performance of the prestation.30 The significance of this debate would be more easily seen if the distinction drawn by French legal scholars between performance by equivalent and reparation had been properly integrated with a further distinction prevalent in the analysis of damages in common law systems, that is, the distinction between the positive interest (called the expectation interest or, more accurately, the performance interest in the English context) and the negative interest (or reliance interest).31 It could be thought that, on the one hand, the notion of ‘performance by equivalent’ refers to the interest in performance of the obligation, and so corresponds to the amount of money which would put the creditor in the financial position in which he would have been if performance had been perfectly achieved; whereas, on the other hand, ‘reparation’ refers to the interest of a creditor injured by non-performance of a contractual obligation in being put in the financial position in which he would have been if the contract had not been made, that is, to erase the harm resulting from the reliance placed by the creditor on a debtor unworthy of it. However, this distinction according to the types of protected interests is not integrated into the positions adopted either by the supporters or the opponents of contractual liability. In practice, everyone considers that in the contractual context the creditor has a right to damages corresponding to the positive interest. This is completely clear in the writings of the opponents of contractual liability, since it is exactly the way in which they define performance by equivalent, but it is just as true of those who support the notion of contractual liability and therefore the notion of reparation. For the latter, to maintain that contractual damages must ‘repair’ the loss means that: they must be calculated in a way so as to compensate losses caused by the failure to perform (damages) and to satisfy the interest which the creditor had in performance (the positive interest): damages must place the creditor in the position in which he would have been if the contract had been properly performed.32

And if the distinction between positive interest and negative interest is to have a real place in French positive law (which at present officially it does not), its putting into effect would have to respect other criteria, notably, whether or not the contract is ineffective (for example, on a ground of invalidity) or whether or not it has been ‘destroyed’ as a consequence of terminated for non-performance.33 In this way, one reaches something of an impasse. The principal support relied on by the opponents of the concept of contractual liability is the special 30 See eg J Flour, J-L Aubert and E Savaux, Droit civil, Les obligations 3. Le rapport d’obligation, 9th edn (Paris, Sirey, 2015) [178]. 31 See ch 2 above, pp 32–33. 32 Malaurie, Aynès and Stoffel-Munck (n 10) [977]. 33 See G Viney, P Jourdain and S Carval, Traité de droit civil, Les effets de la responsabilité, 4th edn (Paris, LGDJ, 2017) [304]ff.

46  Yves-Marie Laithier and ­justifiable nature of the rule which limits damages for non-performance of a contractual obligation to foreseeable harm. In doing so, they break with the position adopted both by legal scholarship and the courts dating back to the last century which saw the requirement of foreseeble damage as an anomaly, an unwelcome breach of the sacrosanct principle of full reparation. The renewed acceptance by the courts of foreseeability of harm is very much to the advantage of those authors who seek to demonstrate the unbridgeable gulf between the commission of a delict and non-performance of a contract. Putting this particular aspect aside, the debate has not led to any improvement in the rules used to assess damages. Practical difficulties of assessment do not find their solution in choosing one or other conception. To illustrate this, let us take a simple example drawn from a New Zealand case.34 A contract of sale of bottles of wine was concluded between a producer and a wine merchant. As a result of a mistake made by the producer, the wine in the bottles was of a quality lower than indicated on their labels. Unaware of this, the wine merchant resold all the bottles to his customers without a discount. Later, the wine merchant discovered that the goods were not in conformity with the contract and claimed damages from the producer, even though none of his own customers had made a claim against him. Could he recover damages and, if so, what should be their proper measure? The answers are by no means clear but – and this is the important point here – the debate as to the nature of contractual liability does not help us to decide.35 As regards performance by equivalent, as has been seen, this notion can be understood more or less extensively. If performance by equivalent refers only to the promised prestation which has not been correctly performed, damages are due according to the difference in value of the goods as agreed and the goods of lower quality actually supplied. But if performance by equivalent refers instead to ‘the equivalent of the foreseeable interest which the creditor had in performance of the contract’,36 then there is no room for any award of damages, at least where it was clear in the above example that the goods were intended for resale: the wine-merchant obtained the profit which he expected from performance of the prestation. As he had not agreed either to a reduction or reimbursement of the prices paid by his customers, the buyer was in exactly the same financial position in which he would have been if the goods had been in conformity with the contract. As regards reparation, damages must be assessed in a way so as to restore the creditor to the position in which he should have been if the harmful act had not been committed. There remains the question of determining which position should act as the point of reference. If the ‘position’ refers only to the relationship between the producer and its own buyer, the latter has suffered a loss. Here, the 34 See McSherry v Coopers Creek Vineyard Ltd (2005) 8 NZBLC 101. 35 ibid [20], the court decided that the buyer had suffered no harm and as a result his claim for damages was rejected. 36 le Tourneau and Poumarède (n 24) 286.

The Relationship between Contractual and Extra-Contractual Liability  47 harm corresponds to the difference in value between the goods as agreed and the lower-quality goods which were delivered. But if the ‘position’ refers to the one resulting from the resale for which the bottles of wine were intended, then no damages should be awarded, as the goods had been sold at the price which the wine merchant would have made if the goods had been in conformity with the contract; he therefore obtained the profit which he expected from performance of the contract. Here, there was no harm, as it had been avoided. In situations of this kind, it is not the choice between the existence or nonexistence of contractual liability which determines the assessment of damages. The true difficulty in this case is knowing whether in defining the creditor’s harm (dommage), one should take account of the possibility that the wine merchant/ buyer has passed on to third parties (his customers) the consequences of his own seller’s non-performance, given that overall it was these sub-purchasers who ultimately paid for goods that were not of the agreed quality. And the solution to this problem depends on other considerations, for example, whether it is legitimate in calculating damages to take into account the price of resale where this was fixed by the initial contract. Moreover, in French law, a case of this kind raises the question of the proper relationship between two distinct but compatible sanctions of nonperformance: damages and price reduction,37 the latter of which has long been recognised by the law of sale where the property is defective and which was generalised by the new article 1223 Cc. Thus, in our earlier example, the wine merchant/ buyer could obtain a price reduction even if he had resold the bottles of wine at a reduced price. This result can be justified as reflecting the strict commutative logic which lies at the heart of the mechanism of price reduction.38 As will be seen, the debate as to the legitimacy of ‘contractual liability’ has had greater influence on the question whether the extra-contractual regime as well as the contractual regime should apply as between contracting parties.

II.  The Primacy of Contractual Rules One of the most famous rules of the law of obligations, and one often presented as particularly characteristic of French law, is the rule known as non-cumul (or, in full, non-cumul des responsabilités contractuelle et délictuelle), according to which a creditor victim of the non-performance of a contractual obligation is f­ orbidden

37 These two sanctions may be awarded concurrently, but any reduction in price affects the assessment of damages since, in determining the loss for which reparation is to be made, account must be taken of the fact that, ex hypothesi, the price has been reduced and that, if already paid, part of it has been returned. See, eg, as regards non-performance of a contract of sale: O Deshayes, obs on Cass civ (3) 14 December 2017, RDC 2018, 29, esp 30–31. See also A Bénabent, Droit des contrats spéciaux civils et commerciaux, 12th edn (Paris, LGDJ, 2017) [244]. 38 See O Deshayes, T Genicon and Y-M Laithier, Réforme du droit des contrats, du régime général et de la preuve des obligations. Commentaire article par article, 2nd edn (Paris, LexisNexis, 2018) 558ff.

48  Yves-Marie Laithier from bringing a claim in damages on the basis of extra-contractual liability. This principle is the technical means by which the primacy of the contractual order is maintained in relations between contracting parties. Here, I will first outline its content in the positive law and then consider its future.

A.  The Content of the Principle Denying an Option In order to grasp the significance of the principle denying an option as a matter of substantive law, we must assume a situation in which the conditions for the application of both contractual and extra-contractual liability are met: a single action causing harm has occured which can justify the application of either of the two regimes. Let us take by way of example the case of a builder who breaks the legal rules governing disabled access and for this reason supplies a building which does not correspond to what was agreed. Where a term of the building contract reduces the prescription period of an action seeking to establish contractual liability, can the creditor either claim damages on the basis of extra-contractual liability or, instead, claim on the basis of contractual liability while relying, if need be, on the prescription period applicable in the extra-contractual context? The answer in both cases is no: the creditor can neither opt in favour of the extra-contractual regime as a whole39 nor can he combine the rules of one or other regime as they favour his position.40 Where the conditions for the application of contractual liability are fulfilled this rules out the application of extra-contractual liability in the relationship between the contracting parties. The benefit which a creditor would draw from having an option is not enough to grant it to him.41 As  the Cour de cassation has consistently held, ‘the creditor of a ­contractual

39 See, eg, the censure by the Cour de cassation of a court of appeal which had imposed liability on the basis of the former art 1384(1) Cc (liability without fault) where there was a contract for hospital services: liability in the body providing these services could not be sought other than on the basis of its failure in its contractual obligations of safety and supervision: Cass civ (2) 8 March 2018, JCP G 2018, 639 note L Perdrix, RTD civ 2018, 674 obs P Jourdain. 40 For an example from an abundant case law see Cass civ (3) 14 May 1997, Bull civ III no 105. 41 As Borghetti observes, the necessary application of contractual liability which is required by the rule of non-cumul is not insignificant to creditors who may find in contracts an effective means of excluding delictual liability in their relations with their contractual partners. It appears therefore that ‘the prohibition of limitation clauses in the delictual context is in part a trompe-l’oeil. Certainly, a contract which would have for its sole effect a limitation of delictual liability would in theory be a nullity as being contrary to public policy [ordre public], but the precise result of the presence of any other contract is the prohibition of delictual liability as a sort of side-effect. It is therefore false to state that delictual liability is a matter of public policy in French law and that this is the justification for the prohibition of limitation or exclusion clauses in this context. On the contrary, delictual liability is so little a matter of public policy that any contract whatsover absolutely excludes the possibility of reliance on delictual liability, even where this would have a vital significance for the individual in question. As with other laws, French law in fact accepts that contract can modify the application of delictual liability. It does so only in a way which is more subtle, by permitting the pure and simple paralysis of delictual rules where there is a contract, independently of the interests at stake’: J-S Borghetti (n 10) [59]–[60].

The Relationship between Contractual and Extra-Contractual Liability  49 obligation cannot rely on the rules of delictual liability against the debtor of this obligation even where he would have an interest in doing so’.42 There are a few exceptions to the principle of non-cumul, that is, where the creditor enjoys an option between the two regimes. If one leaves aside contracts for third parties implied for the benefit of those close to a direct victim of an accident (although this is a technique which is hardly used any more),43 an example can be found in the choice recognised in the owner of a building ordered to compensate a third party for harm caused by a construction defect, since in this situation the owner may sue the builder either on a contractual basis or on a delictual basis by way of recourse action as subrogated to the rights of the third-party victim.44 Another example may be found in the choice enjoyed by a creditor of a contractual obligation where his debtor has commited a criminal offence, for example, a breach of confidence, as the Criminal chamber of the Cour de cassation holds that a civil action (action civile)45 brought before a criminal court must obey the rules of extra-contractual liability, this having the effect that the choice between suing in the civil court or in the criminal court leads to a choice between contractual and extra-contractual liability.46 Once it was also thought that an option was possible in the case of dishonest or deliberate fault (faute dolosive), for example where a debtor lies in relation to the insurance which he ought to take out to cover a risk linked to performance of the contract: in its way of thinking, this was not very far from the American case law which accepted a tort of bad faith breach of contract.47 In the end, though, this approach was rejected in the French context48 and so contractual fault does not give rise to any option however serious it may be. Even so, in the case of intentional or gross fault (faute intentionnelle ou lourde), legislative or contractual limitations on the damages recoverable are ineffective so that the scope of reparable harm more or less follows the lead set by extracontractual liability. As will be seen, this considerably reduces the significance of the denial of an option.49 Turning to the operation of the principle of non-cumul in the civil process, in a number of situations, it may not be easy to decide whether the conditions of contractual liability are satisfied. Were the parties bound by a contract and, if so, did the action which caused the claimant’s harm constitute a failure to perform a contractual obligation? For example, did a defendant who used an industrial

42 See Cass com 26 September 2018 no 17-15306; Cass com 10 March 2015 no 13-10003; Cass com 14 February 2012 no 11-10346; Cass civ (1) 3 March 1993 no 90-19790; Cass civ (1) 4 November 1992, Bull civ I no 276; Cass civ (1) 11 January 1989, Bull civ I no 3. 43 See Viney (n 3) [188-2]. 44 ibid [225]. For a critique of these decisions, see S Carval, obs on Cass civ (3) 20 December 2006, RDC 2007, 754. 45 On the action civile, see ch 6 below, pp 110–113. 46 Viney (n 3) [223]. 47 EA Farnsworth, Contracts, 4th edn (New York, Aspen Publishers, 2004) [12.8]. 48 Viney (n 3) 222. 49 Malaurie, Aynès and Stoffel-Munck (n 10) [1007].

50  Yves-Marie Laithier tool, developed with the claimant, break his contractual obligation forbidding its exploitation (thereby attracting contractual liability), or did this instead constitute unfair competition (which would otherwise attract extra-contractual liability)? The boundary between the contractual and the extra-contractual spheres is often rather unclear.50 In order to reduce the uncertainty which this produces, a person who seeks the imposition of liability is allowed to put his claim principally on a contractual basis but alternatively (and subsidiarily) on the basis of the extracontractual rules.51 This does not weaken the principle denying an option since consideration of the claim based on delictual liability is subordinated to a prior decision that contractual liability does not apply. To this can be added the fact that in French law the rule known as ‘the authority of the matter adjudicated’ (l’autorité de la chose jugée) (whose English equivalent is known as res judicata) forbids the admissibility of a new claim based on the same facts and pursuing the same legal purpose even where it rests on a different legal basis. So, for example, a claim for compensation against a doctor based on contractual liability is inadmissible where an earlier final judgment rejected a claim to the same end based on delictual liability.52 Given the need to present all possible grounds of recovery in any initial claim (referred to as the need to ‘concentrate’ these grounds), a claimant who has difficulty in identifying the proper basis of his claim has therefore every reason to invoke both regimes of liability in order not to lose his right to compensation, the contractual regime principally but subsidiarily on the basis of the extra-contractual.53 The importance of making an alternative subsidiary claim on an extra-­ contractual basis becomes clear when one looks at the role of the court. A claimant who by mistake or ignorance fails to ask for the imposition of extra-contractual liability on a defendant in addition to contractual liability cannot rely on the court to correct this omission. In other words, a court which rejects a claim based on contractual liability on the ground that this does not apply is not bound to consider if the conditions of delictual liability are instead satisfied.54 Similarly, if a creditor 50 Eg Cass com 24 September 2003, Bull IV no 145, where the dock at a port operated by the consignee of goods was damaged during the manoeuvring of a ship at the dock in the course of performance of a contract of carriage between the carrier and the port operator. The court held that the claim for reparation for this damage was necessarily and exclusively governed by the rules of contractual liability. The consequence of this in the case itself was to render a choice of jurisdiction of the English courts inapplicable to the claim for reparation. The judgment pushes the boundary of contractual obligation a long way as the dock did not form part of the goods carried. Cf Cass civ (3) 16 March 2005, Bull civ III no 67, where the court’s approach was more restrictive in the context as regards extra-contractual liability for non-performance of a legislative obligation to restore and to cleanse polluted property (whose breach attracted criminal sanctions) ‘notwithstanding the existence of a private law relationship’ between the parties. 51 See Cass com 13 July 2010 no 09-14985. 52 See Cass civ (2) 25 October 2007, Bull civ II no 241, RDC 2008, 1143 obs O Deshayes, RTD civ 2008, 159 obs R Perrot. 53 See Ancel (n 13) [15]. 54 See eg Cass civ (3) 4 November 2009 no 08-13192, which held that ‘if art 12 of the Code de p­ rocédure civile, which is among the guiding principles of procedure, requires the court to give or to restore to the

The Relationship between Contractual and Extra-Contractual Liability  51 sues the other party to the contract by relying exclusively on delictual liability, his claim may be rejected without the court being bound ‘to check whether the action could have been successful on another legal basis’.55 All these rules encourage contractual creditors to be extremely careful! If denial of an option is accepted in principle by French legal scholars on the ground of the structure of the law and the very open nature of extra-contractual liability, several proposals have been put forward aimed at introducing a degree of flexibility in the rule or reducing its scope.

B.  The Future of the Principle Denying an Option The controversy about the concept of contractual liability has had only a minor significance on the principle of non-cumul, particularly given that one could expect a sharper divide between the two sets of protagonists. On the one side, there are the supporters of contractual liability who could be expected a priori to oppose an option for the creditor in order to preserve the distinctiveness of contract, whether in their discussions of the positive law or of the Projet de réforme. On the other side, there are the opponents of contractual liability who could be expected to be open to the relaxation of the rule of non-cumul on the basis that contractual damages and reparation for harm caused by a delict do not exist at the same level, for if the function of damages is only to provide an equivalent to performance, there is no inconsistency in allowing an extra-contractual action in cases where non-performance of the prestation properly so-called is concurrent with the commission of a delict. However, in reality the positions actually adopted by either group of protagonists are similar. The supporters of the notion of contractual liability agree with the rule of non-cumul as they are not in favour (or are no longer in favour) of the unification of civil liability. Indeed, as soon as the distinction between contractual and extra-contractual liability is seen as valid so as to preserve the distinctiveness of contract in relation to delict, the prohibition of an option appears, at least in principle, to be logical. As Ancel has observed, [i]f the law thinks it necessary to set specific rules for the reparation of contractual harm, it is logical that it would also want to avoid them being excluded by the application of the rules of delictual liability.56

facts or juridical acts which are relied on by the parties in support of their claims in proceedings their correct legal characterisation, in the absence of special rules, it does not require the court to change the characterisation nor the legal basis of their claims; that, in deciding that M. X … and SCI Gard X … had brought a claim seeking the imposition of contractual liability against M. Y … the court of appeal was not bound to consider if their action could succeed on the basis of delictual liability.’ 55 See Cass soc 29 March 2017 no 15-27938. 56 Ancel (n 13) [11].

52  Yves-Marie Laithier This resonates with the views of the critics of contractual liability who similarly consider that the principle forbidding an option is justified as it ‘prevents the law of delict from disturbing the contractual regime’.57 Moreover, they add that this principle is ‘sound’ as it deprives the parties who have freely chosen the contents of their contract, of the possibility of ‘avoiding the yoke of their own special law as soon as they find it too burdensome, even though it is a law to which they have voluntarily consented: volenti non fit injuria’.58 The clearest likely development in the principle of non-cumul may be found in the Chancellerie’s Projet de réforme. The change which is envisaged there reveals its drafters’ hesitation as between two models of civil liability. According to the first model, civil liability is above all based on breach of a duty which is imposed by the law and, accordingly, it is the defendant’s fault or unlawful act which occupies a central place. According to the second model, civil liability is above all based on the recognition of a certain number of rights or interests. These are not all of equal importance in the scale of social values and so the recognition of liability (in terms of its conditions or its effects, but also in terms of the rules governing proof, prescription, procedure etc) depends on the type of harm caused, that is to say, on the type of interest harmed (for example, an unlawful interference with the health of another person). The Projet de réforme proclaims its loyalty to the first model, but it is clearly aware of the second. Certainly, at first sight, the Projet does not explicitly categorise nor does it rank the interests protected by civil liability:59 and while the notion of an interest does appear as a general category in the Projet (in its definition of le préjudice (loss) in article 1235), there are no visible sub-divisions of this category. But in reality, the ranking of protected interests plays an unheralded role in the draft law by way of its favourable treatment of the victim of personal injury, this treatment including the rules governing the relationship between the two orders of liability. In the result, while the rule of non-cumul is formally approved (co-contractors are forbidden on non-­ performance from opting in favour of rules specific to extra-contractual liability (article 1233 of the Projet)), there is a new exception to this rule for the situation where the claim is made for personal injury. In this case, reparation is to be governed by extra-contractual liability even if the loss resulting from a personal injury is caused ‘in the course of performance of a contract’ (article 1233-1(1) of the Projet), although the victim may rely on express stipulations which are more favourable to him (article 1233-1(2) of the Projet). The assumption behind this is that there can be no recovery of corporeal losses by means of an obligation de

57 Remy (n 19) [47], 354. 58 le Tourneau (n 25) [3213.321]–[3213.322] thereby invoking the classic idea of a contract as a special law for the parties found in former art 1134 Cc and now art 1103 Cc. 59 In favour of the approach of the Projet is that it remains loyal to French tradition and does not give way to ‘the sirens of comparative law and the illusions of European harmonisation’; see M Bacache, ‘La  recodification des principes classiques – Articles 1235 à 1238; 1241 à 1249; 1253 à 1256’ JCP G Suppl to nos 30–35 (25 July 2016) 20, [12].

The Relationship between Contractual and Extra-Contractual Liability  53 sécurité apart from where this is contained in the contract itself, a position which is itself controversial.60 Be that as it may, concern for the victim of personal injuries led the drafters of the Projet de réforme to qualify the principle rejecting an option as it has been applied in the positive law. Even if, in practice, the chances of express contract terms actually applying are slim, as ‘contracts which explicitly provide for o ­ bligations for the protection of the physical integrity of the creditor imposed on the debtor are rare’,61 the express acceptance of an option in the Code civil is itself significant. It gives technical expression to a policy in favour of the protection of victims, the purpose here being to allow the creditor who has suffered personal injury to escape the uncertainty62 generated by the distinction between obligations de moyens and obligations de résultat,63 while leaving him with the ability to rely on contractual stipulations in circumstances where it is to his advantage to do so.64 This is proof that, even in France, pragmatism sometimes takes precedence over elegance.65

60 See eg Ancel (n 13) 7, who emphasises that ‘for a carrier to be under an obligation to ensure the safety of its passengers is not exactly the same as being bound to everyone in general to take care not to put their safety at risk’. Cf Brun (n 22) 133 and le Tourneau (n 25) [3213.31], who is entirely happy that personal injuries should be a matter for extra-contractual liability. 61 J-S Borghetti, ‘Un pas de plus vers la réforme de la responsabilité civile: présentation du projet de réforme rendu public le 13 mars 2017’ D 2017, 770, [6] who refers to the ‘hazard’ (l’aléa) faced by a claimant. See also Brun (n 22) 131; S Carval, ‘Le projet de réforme du droit de la responsabilité civile’ JCP G 2017, 401 (who gives the example of a contract for the supervision of children or elderly people which may include a obligation de sécurité de résultat, but who adds that apparently this is rarely the case in practice). 62 For recent illustrations, see J-S Borghetti, obs on Cass civ (1) 30 November 2016, RDC 2017, 235. 63 See D Mazeaud, ‘Réflexions sur le projet de réforme de la responsabilité civile du 13 mars 2017’ in Etudes en la mémoire de Philippe Neau-Leduc (Paris, LGDJ, 2018) 711, esp 719. 64 In favour of the option thereby given to a creditor despite its detrimental effect on the principle of non-cumul see J Knetsch, ‘Faut-il décontractualiser la réparation du dommage corporel?’ RDC 2016, 801, [37]; idem, ‘Le traitement préférentiel du dommage corporel – Articles 1233, alinéa 2; 1240; 1254; 1281, alinéa 2’ JCP G Suppl to nos 30–35 (25 July 2016) 9, [14.] 65 Cf Lord Goff in Henderson v Merrett Syndicates Ltd [1995] 2 AC 145, esp 186.

54

4 Liability of Contracting Parties Towards Third Parties PHILIPPE STOFFEL-MUNCK

The purpose of this chapter is to set out the circumstances in which – in French law and, to a certain extent, in English law – a party to a contract may be held liable to compensate harm caused to a third party by a contractual failing (default/ breach) attributable to the party. By ‘third party’ we simply mean someone who is not a party to the contract whose performance is in question. From the French perspective, this means someone who has not agreed to be subject to the contract, that is to say, bound by the legal norms it lays down or gives rise to, and in particular by the obligations that it creates. The English perspective, however, is somewhat different, because to be party to a contract it is not enough to consent to it: one must also provide ‘consideration’.1 With that proviso, at first glance our topic suggests that we should distinguish at least three situations. First, the claimant, although a third party, holds a right of action which is derived from the contract, and it is this that forms the basis of his claim. He derives his rights from one of the contracting parties – whether directly or indirectly, and whether or not by the party’s will. This situation can result from an assignment of rights2 or some other form of transmission, such as the transfer of accessory rights that French law recognises in the case of chains of contracts by which ownership is transferred. It may also result from a newly created right, as in the case of stipulations for third parties. It is easy to see that a third party of this kind has a contractual action. This will therefore not be the subject of our study. Secondly, the claimant participates in what is called in French law a ‘group of contracts’, that is to say a set of contracts contributing to the accomplishment of some definite overall economic operation, such as the construction of a ­building 1 J Cartwright, Contract Law. An Introduction to the English Law of Contract for the Civil Lawyer, 3rd edn (Oxford, Hart Publishing, 2016) 237. 2 The assignee of rights under an obligation does not become a party to the contract from which the rights derive, but he can bring a contractual claim against the contracting party who bears the burden of the obligation if the latter does not perform it.

56  Philippe Stoffel-Munck or the production of an entertainment. The situation to be considered here is that in which the claimant is party to one contract that contributes to the o ­ verall economic operation, which includes also the contract from which his claim arises. This is, for example, the situation of an actor employed to play a character in a play in relation to the contract between the organiser of the event and the theatre in which it takes place. There is a close proximity between the third party and the latter contract; it makes sense to consider that he has an action in case of a default that causes him harm. However, in technical terms this is not so ­obvious, because although he is a sort of ‘party’ to the overall operation, he is not party to the contract on which he bases his claim for a sanction for defective performance. Thirdly, the claimant is a ‘complete’ third party, which means – to put it simply – that he is in neither of the two previous situations. This is the situation that will be our main focus. In each of these three cases, English and French law have very different approaches to the possibility of the third party relying on a breach of contract. In the first of the above cases, French law considers that contractual actions – both the general action asserting liability for contractual non-performance and actions based on the legal guarantee in sale – are transmitted along the chain together with the ownership in the thing.3 By extension of the rule against the concurrence of liabilities (the so-called rule of non-cumul), the tort action is therefore not available to the sub-purchaser. By contrast, English law denies the sub-purchaser a contractual action,4 but allows a tort-based action against a person who is a higher link of the chain if, but only if, it is shown that he has committed a tort. This is not straightforward, as we shall see.5 Of course, as in French law, an action is available on the basis of liability for defective products deriving from the 1985 Directive.6 However, this is a specific subject, too specific to come within the scope of our discussion. In the second case, the French position has evolved and will no doubt evolve further. Initially, only a delictual action was available to the third party, which implied that the actions attributed to the defendant deserve to be considered as ‘faults considered in themselves and irrespective of any contractual point of view’.7 3 Ass plén 7 February 1986 no 83-14631 and no 84-15189; BAP no 2: ‘But whereas the employer under a building contract, just like a sub-purchaser of goods, enjoys all the right and actions attached to the thing which belonged to its author; he therefore has a direct contractual action against the manufacturer, based on the non-conformity of the thing that has been delivered.’ 4 O Deshayes, La transmission de plein droit des obligations à l’ayant-cause à titre particulier (Paris, LGDJ, 2004) [428]–[429]. 5 Below, p 66ff. 6 Council Directive 1985/374/EEC of 25 July 1985 on the approximation of laws, regulations and administrative provisions of the Member States concerning liability for defective products implemented in French law in arts 1248–1248-17 Cc and in English law by the Consumer Protection Act 1987 esp Pt 1. 7 eg Cass civ (1) 5 October 1983, Bull civ I no 219, RTDC 1984, 505 obs J Huet: the owner of a car entrusted it to a garage which repaired it badly; the vehicle was subsequently sold, and the defect in the repair caused harm to the purchaser. The latter had only had a delictual action against the garage.

Liability of Contracting Parties Towards Third Parties  57 Then, in 1988, the First Civil Chamber of the Cour de cassation held that only a contractual action could be brought within the double limit of the rights which the third party enjoyed under his contract, and of what the defendant could be held liable for under his own contract.8 This allowed the defendant to avail himself of limitations of ­liability applicable under the contract on which he was being sued. Its Third Chamber, however, retained the traditional position.9 From 1991, the Assemblée plénière of the Cour de cassation rejected the initiative of the First Chamber and held in the famous arrêt Besse that an employer’s action against its building subcontractor could not be contractual, given the absence of any contractual relationship between them.10 It therefore returned to delictual liability, with the claimant having the obligation to identify the delictual fault on the part of the defendant within the meaning of article 1382 Cc. However, the Assemblée plénière considerably relaxed this requirement 15 years later when, in the even more famous arrêt Boot Shop it held starkly that: a third party to a contract may invoke, on the basis of delictual liability, a contractual failing where the failing has caused him harm.11

By its generality, this ruling in the arrêt Boot Shop covered both the second and third of the situations referred to above, but the facts related more to the situation of an action brought within a group of contracts. In the case itself, the owners of a building had leased part of it to the company Myr’ho for the operation of its business; by a management agreement (location gérance), Myr’ho entrusted the Boot Shop company with the running of the business it had established in the leased premises. The landlords, who were unaware of this fact, had allowed the premises to deteriorate following the settlement of squatters in the building. This would have allowed the lessee to bring a contractual action against them. However, the

8 Cass civ (1) 8 March 1988 no 86-18182, Bull civ U no 66, JCP G 1988 II 21070 note P Jourdain, RTDC 1988, 531 obs P Rémy and 741 obs J Mestre: ‘Whereas, when the debtor of a contractual obligation has entrusted performance of that obligation to another person, the creditor has against that person only an action which is necessarily contractual in nature, which he may pursue directly but subject to the double limit of his own rights and of the extent of the undertaking of the person who has been substituted as debtor.’ Cass civ (1) 21 June 1988 no 85-12609, Bull civ I no 202, D 1989, 5  note C Larroumet, JCP G 1988 II 21125 note P Jourdain, RTDC 1989, 74 obs J Mestre and 107 obs P  Rémy: ‘Whereas, in a group of contracts, the contractual liability regime necessarily governs claims for compensation brought by all those who have suffered harm by reason only of their link with the initial contract; in this case, since the debtor should have foreseen the consequences of his default in line with the applicable contractual rules, the victim has against him only an action which is ­contractual in nature, even though there is no contract between them.’ 9 Cass civ (3) 13 December 1989, RTDC 1990, 287 obs P Jourdain. 10 Ass plén 12 July 1991 no 90-13602, Bull ass plén no 5, JCP G 1991 II 21743 note G Viney, D 1991, 549 note J Ghestin, RTDC 1991, 750 obs P Jourdain. 11 Ass plén 6 October 2006 no 05-13255, Bull ass plén no 9. See F Terré, Y Lequette and F Chénédé (eds) Les grands arrêts de la jurisprudence civile, 13th edn (Paris, Dalloz, 2015) [177], RDC 2007, 269 obs D Mazeaud (and numerous other contributions at pp 537 to 633), JCP G 2006 II 10181 avis A Gariazzo, note M Billiau, D 2006, 2825 note G Viney, RLDA December 2006, 70 note P Jacques, RLDC January 2007, 5 note P Brun, LPA 22 January 2007, 16 note C Lacroix, RTDC 2007, 115 obs J Mestre and B Fages, RTDC 2007, 123 obs P Jourdain, JCP G 2007 I 115 no 4 obs P Stoffel-Munck.

58  Philippe Stoffel-Munck person who suffered commercial loss as a result of this default was not so much the lessee as the operator of the business – that is, Boot Shop. The lessee and the shop manager (locataire-gérant) therefore together brought an action seeking, on the one hand, damages by way of reparation of the commercial loss and, on the other hand, measures in kind aimed at reinstating the premises. The lower courts allowed their claim in full. The landlords brought a complaint to the Cour de cassation, arguing in particular that the Cour d’appel should have identified how the failing alleged against them constituted ‘a delictual fault considered in itself and irrespective of any contractual point of view’. The Assemblée plénière limited its reply to this point alone. After stating the principle set out above, it held that: having decided for good reasons that the access ways to the demised property had not been maintained, that the entry gate had been condemned, that the service lift did not function, and that it was therefore impossible to make proper use of the demised premises, the Cour d’appel, which had thus identified the harm caused by the landlords’ failings to the shop manager of the business carried on in the demised premises, had legally justified its decision.

Thus, the third party was entitled to pursue a delictual claim against a defaulting debtor as soon as (‘where’) this default had caused him harm, that is to say without having to prove further (i) a contractual failing attributable to the defendant, (ii) harm and (iii) a causal relationship between them. The arrêt Boot Shop was given in the context of a group of contracts, since all those involved were party to contracts for the commercial exploitation of the premises; and because of this, many commentators did not criticise the particular solution in the case on the facts but only its underlying reasoning. However, the judgment made no mention of this context. Its aim was to settle a long-standing disagreement between the Commercial Chamber and the First Civil Chamber of the Cour de cassation about the relativity of contractual fault, to which the former continued to adhere, whilst the latter had abandoned it. By formally establishing the latter position, the statement of principle set out by the Assemblée plénière has the abstraction and absoluteness that are typical of the French normative grand style. For this reason, the principle that it stated could not be limited to actions within a group of contracts, where its legitimacy is more assured. As we shall see, it will indeed not be limited to that situation. There is a clear contrast with English law. At common law, it seems that a contracting party can be liable in tort to a third party only where the courts have recognised the existence of an independent duty of care towards the third party in the tort of negligence or some other independent basis in tort.12 ­ Admittedly, the existence of a factual context in which a French lawyer sees a ‘group of contracts’ may favour such a finding, even though the English cases do 12 H Beale (ed), Chitty on Contracts, 33rd edn (London, Sweet & Maxwell, 2018) [1-214]–[1-226] and [18-024]ff.

Liability of Contracting Parties Towards Third Parties  59 not recognise a general category of ‘group of contracts’. Thus, a duty of care may be found on the part of a sub-bailee in favour of the bailor,13 or of a building subcontractor for the physical harm (though not the pure economic loss) which he causes to the building owner in the performance of his task.14 However, in English law, there is no question of breach of contract systematically giving rise to a claim in the tort of negligence and, in particular, English courts have been very restrictive in finding duties of care in a contracting party to a third party who has suffered ‘pure’ economic loss, that is, economic loss not consequential on the claimant’s own personal injury or damage to property.15 It would therefore be all the more strange to assimilate breach of contract to a tort without regard to the nature of the harm caused, whereas French law holds a contractual debtor responsible for any kind of harm that his non-performance may cause to the third party, not only personal injury or damage to property. When the third party does not participate in the same ‘group of contracts’ with the parties, and has not acquired the benefit of an action arising from their relationship, it seems even more difficult to give him an action for compensation simply based on a failing that has occurred in this relationship. In principle, English law does not accept such a claim: the doctrine of privity of contract very clearly rejects any claim by a third party based on breach of contract itself by one of the contracting parties and this reasoning is sometimes seen as applying by analogy to claims in the tort of negligence for pure economic loss as the courts may accept that the imposition of a duty of care in tort beyond privity of contract is inconsistent with the contractual structure put in place by the parties.16 The main exceptions to this picture that we encounter relate to situations that are close to the idea of a ​​stipulation for the benefit of third parties, such as where a­ professional has a contractual obligation to accomplish a service which he knows will benefit a third party. An example is the lawyer who fails to fulfil the duties of care entrusted to him by his client in carrying out his last wishes, which causes harm to the intended beneficiary.17 In French law, by contrast, the principle laid down in the arrêt Boot Shop appears to allow a complete stranger to rely on contractual non-performance as being a delictual fault. A judgment of the First Civil Chamber of the Cour de cassation illustrates this clearly. In this case, a building abutting a cliff wall had been ravaged by fire, and the insurer of the owner of the premises had refused to

13 Moukataff v BOAC [1967] 1 Lloyd’s Rep 396. 14 British Telecommunications plc v James Thomson & Sons (Engineers) Ltd [1999] 1 WLR 9 (HL(Sc)); Murphy v Brentwood DC [1991] 1 AC 398 (HL). 15 Chitty on Contracts (n 12) [18-033]. 16 Simaan General Contracting Co v Pilkington Glass Ltd (No 2) [1988] QB 758 and see Chitty on Contracts (n 12) [1-218]. 17 Ross v Caunters [1980] Ch 297; White v Jones [1995] 2 AC 207 (HL) and see below, p 68. Another example may be found in the case of the liability of a valuer employed by a mortgage lender to value the property and whose report was passed on to the claimant who had bought the property: Smith v Eric S Bush, Harris v Wyre Forest DC [1990] 1 AC 831 (HL).

60  Philippe Stoffel-Munck advance the funds necessary for the most urgent reinforcement works because it suspected that the claim arose from a deliberate act. This delay allowed a retaining wall to collapse, which in turn caused the collapse of part of the cliff. However, a school had been built at the edge of the cliff, and the collapse carried away part of its playground. The lower courts, followed by the Cour de cassation, held the insurer liable to the owner of the school, on the ground that ‘by postponing in a purely dilatory manner payment of the indemnity to which its insured was entitled, the insurer had committed a fault which was the source of the loss suffered by [the third party]’.18 In this case, the third party victim had no direct or indirect link with the insurance contract. At first glance, therefore, it seems that French law accepts that any contractual failing can form the basis of liability to any third party, and for any harm. Conversely, English law excludes this solution in principle, on the basis that liability in tort is independent from breach of contract and that, at least as regards pure economic loss, general liability in the tort of negligence should not undermine the principle of privity of contract. In what follows, we aim to develop this first impression further, but also to qualify it, particularly since it seems that French law will have to move to a more moderate position. To do this, we shall first examine the procedural aspect of the question by examining the conditions required for this claim for damages to be heard: this involves identifying those who have the right to bring the action. We shall then compare the substantive conditions of each legal system for the success of such an action.

I.  The Conditions of Admissibility of an Action by Third Parties French law and English law will be considered in turn.

A.  French Law In French law as it stands today, actions for damages are, in principle, ordinary actions, which means that for a claim to be admissible it is enough for a claimant to have an interest in bringing the action within the meaning of article 31 CPC. This provides: An action is available to all those who have a legitimate interest in the success or dismissal of a claim, without prejudice to those cases in which legislation confers the 18 Cass civ (2) 10 May 2007 no 06-13269, Bull civ II no 127 (‘qu’en différant de façon purement dilatoire le versement de l’indemnité pourtant acquise à son assurée, l’assureur avait commis une faute qui était à l’origine du préjudice subi par [le tiers]’); JCP G 2007 I 185 no 5, obs P Stoffel-Munck; RDC 2007, 1137 obs S Carval.

Liability of Contracting Parties Towards Third Parties  61 right of action solely upon persons whom it authorises to make or defend a claim, or to defend a particular interest.19

An ordinary action (action banale) is contrasted with so-called ‘authorised’ actions (actions attitrées), which are covered by the second part of article 31. An action is said to be ‘authorised’ when legislation gives a right of action to only a limited category of persons, defined in advance. For example, an action for divorce is an authorised action in the sense that only the spouses can bring it, even though others might have an interest that they think legitimate in the dissolution of the union. Some actions for civil liability are similarly ‘authorised’, such as the action for damages under article 266 Cc relating to the particularly serious consequences that a divorce can involve. So, too, the action for contractual liability, which is limited to a person who is party to the contract or who is creditor of the obligation whose non-performance grounds the action.20 The action claiming delictual liability based on article 1240 Cc (formerly 1382 Cc), on the other hand, is an ordinary action: it is enough to have a legitimate interest in bringing the action. This condition of admissibility is hardly distinguishable from an assessment of the existence, prima facie, of harm suffered by the claimant. This condition makes it possible to declare inadmissible an action that is non-personal,21 vindictive or absurd, brought when no actual or possible harm on the part of the claimant is even plausible. This does not often happen, and so in the ordinary course of events the requirement of admissibility does not play a significant role in restricting claims. As a result, it can be said that French case law, before and since the arrêt Boot Shop, allows any person to bring a delictual action against the debtor who is at fault as long as he can plausibly allege that this default has caused him harm. Opening up actions claiming delictual liability to anyone in this way makes it difficult to assess in advance the risks which attach to contractual non-performance. The debtor, in most cases, cannot identify precisely the circle of those who might be harmed as a result of his non-performance – and even less so if we adopt a broad theory of causation as to consequential harm, such as one which treats all causes equally, however remote they may be (the theory of the ‘equivalence of conditions’). Even if the circumstances allowed a contractual debtor to identify the first circle of persons with an interest in the accomplishment of his act of p ­ erformance,

19 ‘L’action est ouverte à tous ceux qui ont un intérêt légitime au succès ou au rejet d’une prétention, sous réserve des cas dans lesquels la loi attribue le droit d’agir aux seules personnes qu’elle qualifie pour élever ou combattre une prétention, ou pour défendre un intérêt déterminé.’ 20 eg Cass civ (1) 19 June 2007 no 05-21678, Bull civ I no 433, holding that only the assignee of the benefit of a right can bring an action for liability resulting from the non-performance of the assigned obligation, and the assignor no longer has standing to sue. 21 An example of this is the case law relating to actions in delict brought by partners for reparation of the loss in value of their shares by reason of harm suffered by the firm. It is well established that the action is not admissible unless the partner makes a plausible allegation that he has suffered loss separate from that of the firm, and this cannot be the loss consisting of depreciation of his shares by reason of the harm suffered by the firm.

62  Philippe Stoffel-Munck ­ pening up the potential action so broadly may lead to risks that are ­disproportionate o to what his contract is worth to him. Take, for example, a maintenance company that undertakes the repair of a lift in a 10-storey residential building. If it fails to intervene within the agreed time, or does not effectively remedy the defect, this would create difficulties for all the residents who live above the second or third floor. A number of forms of harm may result which, taken together, would not be commensurate with what the company receives by way of payment, and, ­moreover, it could not predict how significant they would be. Given that unforeseeable harms remain fully recoverable in a delictual action, these issues raise the question of the insurability of the contracting party’s liability. This subject does not yet seem to be a sensitive one, because actions by third parties do not seem to be significant in number. So it is necessary to compare the reality of what we see in actual litigation with what is theoretically possible given how the law stands. Between the two there seems to be a cultural phenomenon which causes French courts, and even practitioners, not to make systematic or full use of the unreasonable possibilities that the law offers them. Empirically, we know that the peculiar generosity of the rules of the French law of civil liability is tempered by the equally peculiar restrictive approach in practice to the assessment of a claimant’s harm. It could be that a similar phenomenon is at play here. Someday, practice may, however, take the theoretical position seriously, and our thinking about the future law should therefore be aimed at limiting the circle of those who can bring an action. In this regard, the Projet de réforme de la responsabilité civile devotes an article, which would become article 1234 of the Code civil, to the liability of a contracting party towards third parties. In its latest version this provision reads as follows: Where non-performance of a contract causes harm to a third party, the latter can claim reparation of its consequences from the debtor only on the basis of extra-contractual liability, and subject to that third party’s establishing one of the actions giving rise to liability targeted by Section II of Chapter II. Nevertheless, a third party who has a legitimate interest in the proper performance of a contract can equally invoke, on the basis of contractual liability, a contractual failing where the latter has caused him harm. The conditions and limitations on this liability which are applicable in the relations between the contracting parties may be set up against him. Any contract term which limits the contractual liability of a party to the contract in relation to a third party is deemed not written.

The first paragraph of this provision allows any third party to bring an action, but subordinates its success to an exacting substantive condition: proof of one of the actions giving rise to extra-contractual liability, for example a fault in the defendant. In other words, the mere finding of the debtor’s default is not sufficient to establish delictual fault. This is how the provision has been understood by legal scholars, and it is logical to understand it in this way when one compares it with the second paragraph which presents a ‘contractual failing’ (un manquement contractuel) as an alternative basis of liability with its own regime.

Liability of Contracting Parties Towards Third Parties  63 The ­requirement of a fault distinct from simply a contractual default is, as will be seen below, much more restrictive than the rule laid down in the arrêt Boot Shop. The second paragraph of the provision does not require proof of an action giving rise to extra-contractual liability since it allows the third party to bring a contractual action. However, only a third party ‘who has a legitimate interest in the proper performance of a contract’ can bring such an action. This is apparently a special condition of admissibility of the action, as its relationship to the terms of article 31 CPC is clear. However, if this condition were seen as merely repeating the condition in article 31, its inclusion in article 1234 would be pointless. Logically, it should be understood that the provision adds something to article 31 CPC. What can article 1234(2) proposed by the Projet de réforme be adding? The breadth of the concept of ‘legitimacy’ could cover a number of things. Bearing in mind the consensus that exists amongst writers that the scope of the arrêt Boot Shop should be reduced, and drawing some lessons from continental comparative law, we could make this condition play the same role as the ‘proximity’ condition (Vertragsnähe) established by German law for a third party to bring an action.22 This condition means that the third party ‘must be as exposed to the risk of non-performance as the creditor or must be entitled to benefit from the act of performance that is provided’.23 English law adopts a requirement of proximity for the success of an action in the tort of negligence (as part of the elements establishing a duty of care) rather than for its admissibility, and takes a restrictive approach where the harm for which the third party seeks reparation is neither personal injury nor damage to property.24 It would therefore be quite consistent to adopt a restrictive condition of this nature in French law. This would not limit the exposure of the lift engineer in the example given above, but it would at least exclude from recovery those who suffer only fortuitously from non-performance. The notion of ‘legitimacy’ could also cover the second condition set by German law, namely ‘the interest of the creditor’.25 From this perspective, the only third party who can bring an action is one whom the creditor of the unperformed obligation had an interest in benefiting from the act of performance. This covers the case of actions within a group of contracts, where each person involved in the common operation has an interest in the proper performance by others following his own intervention, as well as in the diligence of his other contracting parties who performed before him. For example, in the renovation of an office, the painter has an interest in the electrician or the plumber having done their job well

22 R Wintgen, Etude critique de la notion d’opposabilité: les effets du contrat à l’égard des tiers en droit français et allemand (Paris, LGDJ, 2004) [317]. 23 ibid. 24 Chitty on Contracts (n 12) [18-034]. 25 Wintgen (n 22) [318].

64  Philippe Stoffel-Munck because it would be necessary to repaint the walls if they had to be damaged; their ­non-performance would have repercussions for him. It is therefore in the painter’s interest that his colleague’s contractors have been diligent and punctual. The law today already takes into account this specifically contractual interaction between parties who co-operate, imposing on each an obligation to inform others that are involved.26 The addition of these two conditions would fit the contractual nature of the action that is available to the third party and make it more logical to allow him to bring an action without having to prove more than a dissatisfied creditor. It is certainly more justifiable to make this easy in the context of groups of contracts and in the case of contracts for third parties. For other third-party victims, the action would remain strictly delictual, with the requirements of proof that such an action involves.

B.  English Law Unlike article 31 of the French CPC, English law does not impose a special requirement of admissibility for bringing an action. If, therefore, a person claims to be the victim of a tort, he has standing to bring an action. The trial will decide whether his claim is good or bad, but it will be heard, as long as it is not an abuse of process or discloses ‘no reasonable grounds for bringing or defending a claim’.27 In contractual matters, English law provides, in principle, that only a contracting party has standing to claim performance of the contract or other remedies arising from its breach: this is the force of the principle of privity of contract. As for the action for reparation of harm caused by the breach, a third party will face no obstacle to have his case heard as such; it is rather at the stage of substantive requirements for the success of such an action that English law is restrictive, and much more so than French law.

II.  The Substantive Requirements for an Action by a Third Party At the most general level, English and French law agree on the essential points as regards an action for damages: the success of the action presupposes the demonstration of harm, an action giving rise to liability (in this case, the French will say a ‘fault’, and the English will require more specifically ‘negligence’)28 and a causal

26 eg Cass com 6 December 1983 no 82-13924, Bull civ IV no 339, RTD civ 1984, 523 obs P Rémy. 27 Civil Procedure Rules r 3.4(2). 28 The statement in the text reflects the typical case in which a third party seeks damages against a contracting party in respect of the behaviour which constitutes breach of contract. In principle, a

Liability of Contracting Parties Towards Third Parties  65 relationship between the two. As soon as we dig a little deeper into the detail of these requirements, however, this impression of convergence disappears – which also confirms that the Devil (ie the one who ‘divides’) is well settled there. It soon becomes apparent that English law is subject to considerably more qualifications and is much more overtly complex than its continental neighbour, at least in its current state. This is why it is better to approach each system separately, because each constitutes a whole.

A.  The Requirements for Success in an Action by a Third Party in English Law As has already been said, the basic principle is that English law does not give a contractual action to those not party to the contract. The effect of the doctrine of privity of contract is that the remedies available in case of breach cannot be used by the third party,29 except in special cases such as the stipulation for third parties that is now governed by the Contract (Rights of Third Parties) Act 1999.30 The third party must therefore bring an action in tort, which will allow him to obtain compensation for his harm. The various conditions for the existence of the different torts differ between them, but in practice the most important tort is the tort of negligence. As regards the latter, the claimant must establish that the defendant owed him a ‘duty of care’ whose breach (‘negligence’) caused his harm. Underneath the condition of the existence of a duty of care, English courts have controlled the scope of liability for negligence, including in cases where the defendant’s action which is complained of also constitutes a breach of contract. For these purposes, the courts draw a number of distinctions in a complex case law, but one of the key sets of distinctions relates to the nature of the harm suffered by a claimant. Generally, the courts have been more willing to recognise the existence of a duty of care in the tort of negligence where the victim of negligence has suffered death, personal injury or physical damage to property whereas they have been much more restrictive as regards psychiatric injury and pure economic loss and where the defendant has committed a mere omission rather than an action.31 In this respect, in the context of the liability of a contracting party to third parties, the courts have typically considered the question of the existence of a duty of care by asking whether the defendant can be said to have undertaken an assumption of responsibility towards the third party who brings the action. Where this is the case, third party can claim on the basis of any tort against a contracting party subject only to establishing the various specific conditions for the application of the tort in question. 29 The leading case is Beswick v Beswick [1968] AC 58 (HL). 30 Note that this statute imposes a special test of intention (s 1(1) and (2)) and requires that the third party be ‘expressly identified in the contract by name, as a member of a class or as answering a particular description’: s 1(3). 31 For a general account see WE Peel & J Goudkamp, Winfield & Jolowicz on Tort, 19th edn (London, Sweet & Maxwell, 2014) ch 5.

66  Philippe Stoffel-Munck the third party may recover pure economic loss, but not merely personal injury or damage to property.

(i)  The General Case and its Limitations Since the famous decision in Donoghue v Stevenson,32 it is well-established that the tort of negligence may be the basis of an action brought by a third party who suffers personal injury as a result of the negligence of a contracting party, there, a manufacturer. For this purpose, the concept of negligence itself is much narrower than a mere contractual breach that may result in one contracting party’s liability to the other.33 It consists of the violation of a duty of care that the law imposes for reasons of policy unrelated to any desire to promote the accomplishment of the contractual transaction as the parties could expect and consists of a failure to take reasonable care of the safety of the claimant in the circumstances. Thus, there is little risk of confusion between the standard by which the contractual responsibility of a contracting party is judged, and that used to assess whether there has been negligence. In particular, liability does not generally arise in the tort of negligence by reason of an omission, although this is counterintuitive to a French lawyer simply as a matter of language as négligence suggests a failure to act rather than the commission of an action. There are exceptions to this general rule, which even allow recovery of pure economic loss, as we shall see below; but the rule remains.34 Therefore, in general, it is incumbent on the third party to demonstrate what positive action it attributes to the defendant as characterising breach of the duty of care in relation to the person of another or his property. For example, it does not seem that a wrongful repudiation of a contract can constitute an action of a kind which would attract liability for these purposes.35 In French law, by contrast, a contracting party may incur delictual liability to third parties merely on this basis.36 The harm which attracts a duty of care is normally limited to personal injury or damage to the claimant’s property and does not appear to extend usually to nonphysical damage37 such as pure economic loss. Although the notion of personal injury seems broadly similar in the two systems, the concept of damage to property requires some clarification. This head of damage includes the value of a thing that has been destroyed, or the cost of its repair if that is possible. There is still a question about how to value this. However, it seems that the calculation of the

32 Donoghue v Stevenson [1932] AC 562 (HL(Sc)). 33 Chitty on Contracts (n 12) [18-033]. 34 ibid [18-027]. 35 ibid. 36 Cass com 21 October 2008 no 07-18487, JCP G 2009 I 123 no 6 obs P Stoffel-Munck, RTD civ 2009, 318 obs B Fages, RDC 2009, 508 obs J-S Borghetti. 37 We may wonder here whether English law recognises this category more clearly than French law, given that it is common in practice.

Liability of Contracting Parties Towards Third Parties  67 value can include the profit that the thing would have generated directly. The case which suggests this possibility, while distinguishing it from pure economic loss, is Muirhead v Industrial Tank Specialities Ltd.38 In this case, the owner of a fish farm raising lobsters wished to install pumps in order to increase his yield. He therefore contracted with a specialist for their installation. The pumps did not work due to a fault in their electric motors, which led to the death of some of the lobsters. The farmer brought an action against the supplier of the motors. He received compensation for his property damage (the destruction of part of the lobster stock) and the resulting financial consequences, ie the loss of profits expected from the sale of those lobsters. However, he did not obtain compensation for loss of the profits which he had expected to make from an increase of his yield. Such loss was designated as pure economic loss. It could have been compensated in a contractual claim because contractual liability allows compensation for the positive or performance interest of the contract and typically consists of pure economic loss. It differs to that extent from liability in the tort of negligence. When it can be inferred from the circumstances that the contracting party intended to protect the interest of the third party through his performance of the contract, compensation for pure economic loss may be a little more conceivable, though we are then a little less distant from the idea of s​​ anctioning a promise made to the claimant. But this situation remains exceptional.

(ii)  The Exceptional Case: An Assumption of Responsibility There is a series of situations in which English law considers that a contracting party has ‘assumed’ responsibility towards third parties, which gives rise to a duty of care towards them. In these cases, which are grouped under the phrase ‘assumption of responsibility’, the basis of the action of the third party against the contracting party is easier to establish. Indeed, in such a case an omission may be sufficient to give rise to liability. Such an assumption of responsibility is of particular importance as regards the recognition of duties of care in respect of pure economic loss. This assumption of responsibility is exceptional. There is a question whether it can be found in every kind of contract, depending on the circumstances and on an analysis of the defendant’s will in the particular case. But it is certainly admitted in various types of contracts by professionals, as long as the necessary circumstances come together, in particular that the harm was foreseeable by the defendant and the relevant third parties were identifiable.

38 Muirhead v Industrial Tank Specialities Ltd [1986] QB 507 (CA) noted S Whittaker (1986) 49 MLR 369, DW Oughton [1987] JBL 370.

68  Philippe Stoffel-Munck An example is provided by contracts for the provision of legal services by ­solicitors.39 The key case is White v Jones in 1995.40 In this case, after a family quarrel, a man had executed a will in March 1986 that disinherited his two daughters. After a reconciliation, he decided to draw up a new will bequeathing a certain sum to each. He instructed his law firm in writing on 17 July 1986. The lawyer assigned to this task did nothing until 16 August, and even then made little progress. The testator died on 14 September without having formally revoked his previous will, which therefore remained in force. His daughters claimed that the defaulting firm were liable for the sums that their deceased father had decided to give them and which they would have received if he had changed his will. However, the only fault that they could allege against the defendant solicitor was a negligent omission (the failure to take reasonable care to execute the change), and the two would-be legatees had lost nothing from their existing assets. Yet the House of Lords, following the Court of Appeal, held that their claim in the tort of negligence was well founded on the basis that, in these circumstances, the contract between the solicitor and the testator expressed (or should be deemed to reflect) an assumption of responsibility for their benefit.41 A similar conclusion has been reached with respect to other professions involved in the provision of professional services such as advice or auditing.42 An extension of this doctrine was made by the decision in Junior Books in 1982, in the context of building contracts.43 This case involved the building of a factory and the owner had appointed a specialist subcontractor to make and lay the floor of the production area. There was no direct contractual link between them. The floor proved defective, cracked and had to be replaced, which led to a halt in production and loss of profits. The subcontractor’s fault did not cause personal injury or damage to property previously owned by the owner. The loss of profit and the replacement cost were thus pure economic loss. The House of Lords accepted that the final head of damage was recoverable even though there was no contract between the parties because a combination of circumstances meant that the subcontractor knew how much the owner was relying on his special skill, how much depended on his performance, and the harm that would result from its failings. They were thus in such a relationship of ‘proximity’ that the result was a duty of care which could cover purely financial losses. This solution, which seems rather obvious to a French lawyer, gave rise to such a wave of criticism that, despite the rule of binding precedent, several later decisions seem to have undermined its authority.44 It has been portrayed as ‘discredited’,

39 A further important example may be found in Henderson v Merrett Syndicates Ltd [1995] 2 AC 145 (HL). 40 White v Jones (n 17). 41 See further Chitty on Contracts (n 12) [1-222]. 42 ibid [18-024]. 43 Junior Books [1983] 1 AC 520 (HL(Sc)). 44 See esp Murphy v Brentwood DC (n 14).

Liability of Contracting Parties Towards Third Parties  69 ‘virtually extinguished’ and ‘as aberrant, indeed, as heretical’.45 It does not bear thinking about how English commentators would receive the arrêt Boot Shop. The authority of the latter case is certainly criticised in France, but in a much less radical way.

B.  French Law The solution given in the arrêt Boot Shop is so criticised today that it seems clear that the law is likely to evolve significantly. After considering the criticism, we shall consider the direction that the development might take.

(i)  Contemporary Criticism of the arrêt Boot Shop The arrêt Boot Shop holds that ‘the third party to a contract may invoke, on the basis of delictual liability, a contractual failing where the failing has caused him harm’. The terms of the principle as thus stated tend to equate delictual fault and ‘contractual failing’. The formula removes any need to investigate whether the actions of the defendant which form the basis of the claim would have constituted fault outside a contractual context. In addition, by referring to the term ‘contractual failing’, the Cour de cassation has chosen the widest term in the vocabulary of contractual liability. It covers not only the lack of diligence, technical skill or loyalty in the performance of the contract but also any default; a simple non-performance or a simple delay is equally a ‘failing’. The limiting effect of imposing a requirement of an action giving rise to liability is thus reduced to a minimum. As for the requirement of harm, we know that the notion cannot be broader than it is in French law, which accepts, for example, reparation for the loss of even the tiniest chance, and places no limit on the concept of intangible loss. In the case of contract, it is only foreseeable loss that is reparable, but this limitation does not apply to delict, where the principle of full reparation prevails. We are therefore apparently faced with a paradox: a contracting party has a greater exposure to liability in favour of third parties than in favour of the person with whom he has entered into his contractual undertaking. The requirement of a causal link was not specified by the decision in the arrêt Boot Shop, either, so that, if we adopt the theory of the ‘equivalence of conditions’ (which treats all necessary causes equally), a contractual debtor under French law is responsible to everyone for every harm that would not have occurred without his contractual failure. Of course, although that is what the decision says, no doubt it is not what it means nor what litigants have understood to mean because, otherwise, litigation would have exploded over the last 12 years. In reality, it may be thought that most

45 See

the references in Chitty on Contracts (n 12) [18-029].

70  Philippe Stoffel-Munck of those who practise French tort law have worked out the distinction between the abstract formulation of the rule and what it allows in real terms. Through their experience and culture, they have a sense of the cases in which we can more or less make it effective, depending on the interests at stake, the social acceptability of the various available solutions, context, and of course the character of the judges. The lack of definition of the concepts, and in particular the flexibility of the requirement of a causal connection, certainly gives the court every latitude to reject a claim which seems unreasonable, without having to give any specific reason. Moreover, if the court accepts the claim, its sovereign power over the evaluation of the claimant’s harm will allow it to moderate this evaluation as it thinks fit. We can thus leave it to the wisdom of judges to avoid the arrêt Boot Shop becoming a means of blowing up the contract. Never mind if from time to time this wisdom is absent, as these cases remain marginal at the macroeconomic level. In principle, however, it is not satisfactory that the rule is anything other than what a simple reading of it would suggest. At first reading, the general principle formulated by the arrêt Boot Shop exposes debtors in such a demonstrably excessive manner that the vast majority of writers have been critical of it, seeking ways to give a better definition of the requirements for a contracting party’s liability towards third parties. Moreover, the courts themselves have sometimes chosen to depart from the terms of the decision. In 2017, two notable decisions, one by the Commercial Chamber and the other by the Third Civil Chamber, distanced themselves from the somewhat simplistic formulation given in the arrêt Boot Shop. The first decision was on 18 January 2017. It ruled on a misrepresentation made by the transferor in connection with the transfer of control of a company. The company had suffered harm, and it therefore brought an action for reparation against the transferor on the basis of article 1382 Cc by invoking the ‘failing’ which consisted in the inaccuracy of the facts whose accuracy the transferor had guaranteed in the instrument of transfer. Its case having been dismissed as a matter of substance (au fond), the company brought a complaint to the Cour de cassation on the basis of the arrêt Boot Shop case law. It was dismissed on the ground that: receiving from the company a claim made on the basis of the delictual liability of the transferors by reason of a failing in the undertakings they had given in favour of the transferees in the instrument of transfer, and the harm which resulted to the company in consequence, without it being shown or even alleged that this contractual failing constituted a quasi-delictual fault against it, the Cour d’appel rightly held that the company was not entitled to rely on a violation of the contractual stipulations in the instrument of transfer.46

46 Cass com 18 January 2017 no 14-16442 (‘saisie d’une demande de la société fondée sur la responsabilité délictuelle des cédants en raison d’un manquement aux engagements souscrits par eux envers les cessionnaires dans l’acte de cession et du dommage qui en était résulté pour elle, sans qu’il soit établi ni même allégué que ce manquement contractuel constituait une faute quasi-délictuelle à son égard, c’est à bon droit que la cour d’appel a retenu que la société n’était pas fondée à se prévaloir d’une violation des stipulations contractuelles de l’acte de cession’); D 2017, 1036 note D Mazeaud, AJ Contrats 2014,

Liability of Contracting Parties Towards Third Parties  71 This decision therefore asserts as a matter of law (‘rightly’) that the two faults (contractual and delictual) are not automatically identical: it must be established positively how the failing also constitutes a fault in the sense required for delictual liability. The second decision, given by the Third Civil Chamber, does not merely endorse the distinction between contractual and delictual issues: it clarifies the relationship between them by holding that ‘reasons … based solely on the failing in a contractual obligation of result (obligation contractuelle de résultat) to deliver work that conforms with the contract and is free of defects, are insufficient to establish a delictual fault’ on which third parties may rely.47 The First Civil Chamber, for its part, keeps to the line that it inspired. At most, in a decision in September 2017, it adopted a form of words which is a little less categorical than that of the Assemblée plénière, in which the court simply holds that ‘the relative effect of contracts does not prevent a third party from relying on a contractual failing committed by a party, as long as that failing is directly at the source of the loss he suffers’.48 In the same period, the Ministry of Justice published the Projet de réforme which also took the opposite approach to that of the arrêt Boot Shop by proposing a new article 1234 of the Code civil, the terms of which were set out above.49 In view of the convergence of these indications of development, apart from the case law of the First Civil Chamber of the Cour de cassation, it may be useful again to make a critical assessment of the issue. Comparative law makes it possible, in particular, to identify the key points on which we should strike a balance between the interests of third parties and those of contracting parties.

(ii)  The Direction of Development Article 1234 of the Projet de réforme outlines well enough the two ways in which a contracting party might be liable, although it fails to prioritise them. If we exclude the idea that he should on principle be immune in relation to third parties, the contracting party’s liability may be either extra-contractual or contractual. These are paragraphs 1 and 2 of the proposed provision. 191 note A Lecourt, JCP G 2017, 1174 [2] obs P Stoffel-Munck, RDC 2017, 425 obs J-S Borghetti, Gaz Pal 2017 no 292 note D Houtcieff, RTDC 2017, 651 obs H Barbier. 47 Cass civ (3) 18 May 2017 no 16-11203, Bull civ III no 64 (‘des motifs … tirés du seul manquement à une obligation contractuelle de résultat de livrer un ouvrage conforme et exempt de vices, sont impropres à caractériser une faute délictuelle’), RDC 2017, 425 obs J-S Borghetti, RCA 2017 comm 212 note L Bloch, RTDC 2017, 651 obs H Barbier, AJ Contrat 2017 no 8, 377 F Chénédé, RDI 2017 no 7, 349 P Malinvaud, Gaz Pal no 24 (27 June) 14 obs D Mazeaud, RGDA 2017, 364 note P Dessuet, JCP G 2017, 1174 [2] obs P Stoffel-Munck, Dr et patrim 2017 no 273, 55 obs A Hontebeyrie, D 2017, 1225 note D Houtcieff. 48 Cass civ (1) 20 September 2017 no 16-20456 (‘l’effet relatif des contrats n’interdit pas au tiers à une convention de se prévaloir du manquement contractuel commis par une partie, dès lors que ce manquement est directement à l’origine d’un préjudice subi par lui’), RCA 207 comm 302. 49 Above, p 62.

72  Philippe Stoffel-Munck (a)  A Delictual Action It will readily be admitted that a contractual failing may be coupled with a d ­ elictual fault. However, the social need to promote economic activity points towards not allowing his first minor failing in performance to expose the debtor to liability which might be disproportionate to his interest in the contract. Along the same lines, if we accept that entering into a contract is based on foresight, we betray its essence if we allow a third party to bring an action without any regard for what the debtor could have foreseen. So why should we allow the third party’s action to bypass an exclusion clause when there is no basis for criticising the debtor, no gross negligence, and certainly no fraud or other dishonesty (dol)? Finally, the principle of the relative effect of contracts, even at it its most basic level, suggests that we should not give third parties the same rights as the person to whom the contractual undertaking was given and who, in most cases, has given something in return. If we accept these arguments, it follows that the proclaimed equivalence between delictual fault and contractual failing is inappropriate and illogical. Is it enough then to say, as article 1234(1) does, that a delictual fault is sufficient to give rise to a liability that will ignore the terms of the contract and the expectations of the parties? This is a return to the solution adopted by some of the case law before the arrêt Boot Shop, and it would face the same complaint of vagueness, uncertainty and even arbitrariness. But can we think of a better solution? First, it seems possible to define fault in a little more detail. Under the current draft provision, a court may well take the view that, on the facts of the case, a contractual failing amounts in itself to a delictual fault. The court only has to say that this is the case. However, for the reasons explained above, it seems to us that this equivalence is not the case. In particular, Borghetti has shown convincingly how extreme it was to hold any breach of the principle of pacta sunt servanda as a delictual fault.50 And we have to keep reminding scholars who tend to take ethics for law that there is a serious – and even a tragic – truth behind Blaise Pascal’s aphorism that ‘the man who wants to act like an angel, acts like the Beast’. It would therefore certainly be helpful to specify that the fault capable of giving rise to the contracting party’s liability towards third parties cannot be based solely on the finding of a contractual failing. More is needed. It is not so much that the actions alleged against the debtor must be considered ‘irrespective of any contractual point of view’, as the case law before Boot Shop stated.51 The idea is that one must prove more than a simple non-performance or, to put it another way, more than a failing in an obligation of result – a position which was exactly what was held by the Third Civil Chamber in its decision of 18 May 2017.52 50 Observations on Cass com 21 October 2008, RDC 2009, 506, 512. This opinion is also held in England: see F Pollock, The Law of Torts, 13th edn (London, Stevens, 1929) 448ff; Chitty on Contracts (n 12) [1-158]. 51 Above, n 11. 52 Above, n 47.

Liability of Contracting Parties Towards Third Parties  73 Delictual fault involves the characterisation of behaviour. It is therefore ­appropriate to ­characterise the defendant’s behaviour and to identify in what way it falls below what a party outside the contract can expect from an ordinary person acting in similar circumstances. This test presupposes that we can identify a reference standard applicable to the defendant: a certain degree of care, skill, probity and civility to which everyone must be held whether or not he is in a contractual relationship. This reference standard is distinct – if not in kind, at least in degree – from that required in the relationship between contracting parties, since a third party cannot be placed on an equal footing with the contracting party who is the creditor of the obligation. This is not only because the legitimate expectations of the contractual debtor must be respected, but also because the third party has not sacrificed anything to obtain the benefit of such a standard. This observation reflects in broad terms the English idea that a third party cannot enjoy the benefits of a contracting party because he has not given any consideration.53 The liability of a contracting party towards a third party also deserves to be particularised in terms of the causal relationship. The theory that all necessary causes should be treated equally, which is used routinely in relation to delictual liability, does not take into account what the person could foresee when he committed the actions that are complained of against him. However, there are good arguments for incorporating a requirement of foreseeability into the assessment of the contracting party’s liability towards third parties. These arguments are political and logical (see above). They are also fit with both English and German law. In both of these systems, liability to third parties presupposes that the contracting party could have foreseen the harm that might result from the alleged default.54 The theory of causation that looks to the appropriate cause(s) of the harm (causalité adequate) takes this forward-looking element into account. It would be right to make it the rule in this context. Finally, it would be helpful to establish the priority as between the delictual action provided by paragraph 1 and the contractual action referred to in paragraph 2 of article 1234 of the Projet de réforme. Under German law, the action of a third party against the contractual debtor is subsidiary and so it is not possible if the third party can bring an action for the reparation of his harm against one of his own debtors. Thus, it has been held that a subtenant who suffers disturbance in the peaceful enjoyment of the premises had no action against the landlord because the person who is normally answerable for such an event was his own contracting party, the tenant. This solution could be transposed into groups of contracts more broadly. In the context of the arrêt Boot Shop it would have led the shop manager to sue the party with whom he had contracted, even if the latter then may bring

53 Chitty on Contracts (n 12) [18-018]. 54 Wintgen (n 22) [319]. English law appears also to consider this issue in deciding whether or not a duty of care is owed to the claimant.

74  Philippe Stoffel-Munck a recourse action against the landlords so as to pass on the cost of the liability. There are two advantages of such a requirement for bringing an action. On the one hand, it allows everyone to receive compensation by reference to what his contract provides. This avoids giving third parties more rights than they had under the contracts by which they undertook the activity in the course of which the harm occurred. In this way, not only the defendants’ expectations, but also those of the claimant, would be respected. On the other hand, this subsidiarity would make it possible to block attempts to obtain double compensation. Indeed, in the absence of such subsidiarity, a victim who first sued his own contractual debtor could still sue a second time against another member of the group, either in the hope of obtaining a better assessment of the value of his harm (since the first decision would not have the authority of res judicata given the different parties that are involved) or even of taking the risk of keeping quiet about the existence of the first proceedings. Neither of these two situations is really desirable. If it were considered appropriate, this procedural rule could also be applied to the contractual action provided for by article 1234(2) of the Projet de réforme. (b)  A Contractual Action First, it should be recalled that article 1234(2) of the Projet de réforme provides that, as an alternative to delict (‘nevertheless’): a third party who has a legitimate interest in the proper performance of a contract can equally invoke, on the basis of contractual liability, a contractual failing where the latter has caused him harm. The conditions and limitations on this liability which are applicable in the relations between the contracting parties may be set up against him. Any contract term which limits the contractual liability of a party to the contract in relation to a third party is deemed not written.

This is an addition compared to the first version of the Projet de réforme, which only provided for the third party to have a delictual action.55 This was strongly criticised.56 In principle, this contractual action appears to allow the defendant to limit his exposure as far as his foresight allows, either by an exclusion clause or by reference to article 1231-3 Cc (former article 1150), according to which ‘[a] debtor is bound only to damages which were either foreseen or which could have been foreseen at the time of conclusion of the contract, except where non-performance was due to a gross or dishonest fault’. On the other hand, the debtor is exposed to the possibility of liability as regards any contractual failing. There is therefore a question of balance – and this is also clear when we examine the condition of 55 See article 1234 as initially drafted in the ‘Avant projet de loi de réforme de la responsabilité civile’ published on 29 April 2016, www.textes.justice.gouv.fr/art_pix/avpjl-responsabilite-civile.pdf. 56 See esp O Deshayes, ‘La nouvelle mouture de l’avant-projet de loi de réforme de la ­responsabilité civile: retour sur la responsabilité des parties à l’égard des tiers’ RDC 2017, 238; J-S Borghetti, ‘La responsabilité des contractants à l’égard des tiers dans le projet de réforme de la responsabilité civile’ D 2017, 1846.

Liability of Contracting Parties Towards Third Parties  75 admissibility of the action which limits it to ‘a third party who has a legitimate interest in the proper performance of a contract’. However, the three protections offered to the contractual debtor may seem unsatisfactory, and this would undermine somewhat the impression of balance. Let us consider how. First, is the requirement of admissibility of the action really that? It appears so, but we guess that there will be contrary views, and so if this is the intention of the legislator, it would do well to express it more explicitly. In addition, this requirement must have a specific content which contrasts with that of article 31 CPC. The wording of article 1234(2) even suggests this; indeed, it does not refer to an interest in bringing an action, but rather an interest ‘in the proper performance of a contract’. This may appear to be the same, because evidence of loss can retrospectively serve as an interest which one had in the proper performance. But it may also seem different. For example, it might be understood as requiring that the claimant expected to benefit positively from the performance of the contract. However, between the two, the balance logically leans in favour of the specificity of the requirement, because if article 1234(2) does not add anything to article 31 CPC, it might as well be removed, as earlier explained.57 This observation leads us to add that the requirement should be assessed as at the date of the contract: it is at that time that the third party must have had an interest in its proper performance, which presupposes that he was aware of it. The reason is that this requirement is intended (or so it seems) to strike a balance between the expectations of third parties and those of the parties. But these were fixed at the time of consent. If the expectations of third parties were to be taken into account irrespective of the date on which they developed, the parties’ expectations would be disregarded. Those who have speculated on the proper performance of the contract after the parties have committed themselves will have only a tort action, and that should suffice. Secondly, the limitation of damages to only those harms which were foreseeable at the time of the conclusion of the contract is a rule which does not provide a great counterbalance to the third party’s ability to rely on any contractual failing. Indeed, the case law relates this foreseeability to abstract heads of damage which the contractual debtor could imagine occurring in case of default on his part. But once we reach a certain level of abstraction, everything – or almost everything – becomes foreseeable. If the provision were to specify that the debtor is liable only for the harm that he could concretely anticipate occurring for the third party, things would be different. However, a form of words that reflects this idea is difficult to compose. Finally, giving the debtor the benefit of the terms of the contract raises serious issues of logic. The question is not so much about clauses that passively define the scope of the third party’s action for damages, either as regards time, through a reduced prescription period, or at the jurisdictional level, such as an arbitration

57 Above,

p 63.

76  Philippe Stoffel-Munck clause. In both of these cases, it may seem logical to continue to give the debtor the benefit of these arrangements because this is precisely one of the objectives of the scheme leading to the award of a contractual action to third parties. The most awkward example is drawn from clauses limiting the amount of compensation. If the debtor has capped his liability at 100, and a third party brings a claim for 100, will the debtor still have any further obligation when his own creditor or another third party brings an action against him? This question, and some others, may not be answered by a concise formula of the kind that French civil lawyers like to use. But however it is done, the question will have to be answered.

part ii ‘Fault’

78

5 The Definition of Civil Fault MARIE DUGUÉ

In 2016, a year notable for the reform of contract law, a change was also forced upon specialists in the law of civil liability. Article 1382 of the Code civil, that jewel of French law,1 which provided that ‘[a]ny human action whatsoever which causes harm to another creates an obligation in the person by whose fault it occurred to make reparation for it’, is no more! Even though its words remain, we had to mourn the loss of its iconic numbering, and become accustomed to referring to its ‘replacement’, article 1240 Cc, however much it may pain us to do so. But how long will that last? No one knows for certain, but everyone urges that it should probably not be long. The reform of civil liability, heralded for some years, seems imminent – and a reform of ‘fault’ is included. The latest draft published by the French Government, the Projet de réforme of 13 March 2017, refers extensively to the notion of fault, which is mentioned in no less than 16 of its articles (although in the contractual context, the term ‘non-performance’ is preferred).2 That, however, is not surprising: even when it is confined to the sphere of delictual liability, fault plays a number of different roles. It is the basis on which liability arises, but also defines its effects, for example in relation to a party’s possible exoneration from liability, or in defining the basis of claims for contribution between parties who are jointly liable. Fault is ‘ubiquitous’.3 The most remarkable provisions concerning fault, however, remain those contained in Chapter II of the Projet relating to the conditions of civil liability, and in particular in Sub-section 1 of Section 2 relating to actions giving rise to extracontractual liability. There we find § 1, laconically titled ‘Fault’, which devotes three articles to the question. Article 1241, which states that ‘[a] person is liable for the harm caused by his fault’, carries forward the spirit of former article 1382, while modernising its form. Article 1242-1 appears not yet to be settled, given the square brackets around it, but even this provision is instructive: by stating that 1 H Mazeaud, ‘L’ “absorption” des règles juridiques par le principe de responsabilité civile’ DH 1935 chron 5, emphasising that ‘the principle set out in article 1382 Cc is one of the great rules of justice – rules which can encapsulate the whole of the law on their own’. 2 The relationship between the notions of ‘fault’ and non-performance are still to be settled, however. 3 P Brun, Responsabilité civile extracontractuelle, 4th edn (Paris, LexisNexis, 2016) [293].

80  Marie Dugué ‘[f]ault in a legal person results from fault in one of its organs or a failure in its organisation or its functioning’, it specifies the conditions under which an action of an individual is to be imputed to a legal (non-physical) person.4 Finally, there is article 1242, which is intended to provide a definition of fault for the first time, in these terms: ‘A violation of a legislative requirement or a failure in the general duty of care or diligence constitutes a fault’. If we could study only one of these provisions, which should it be? We might jump upon article 1241, given that it restates the general principle of liability for fault which is a peculiarity of French law, especially when compared to its German or English counterparts.5 Yet, as such, the article does not say much. In particular, although it certainly presupposes that ‘any fault, whatever its gravity and whatever the source of the duty that has been violated, engages the liability of its author in the same way, and gives rise to an obligation in him to make reparation for the entirety of the harm caused to the victim’,6 it does not prevent the qualification of the scope of this principle at the stage at which fault is determined. If fault is itself conceived in a limited way, can we still speak of a general principle of liability according to which all harms are treated in the same way?7 If we do not know the boundaries of fault, how do we know, in the final reckoning, how much value to place on article 1241? The definition of fault is therefore of great interest and this interest can only be increased by its analysis through the prism of comparative law, and especially of English law. It might seem strange to say this, because it is clear that fault is not a very commonly used concept in English law.8 The explanation is probably to be found in the origins of the law of tort, which derives directly from the procedural system of writs. To bring an action, it used to be necessary to obtain a writ from the Lord Chancellor and the number of these writs was limited, with each having its own requirements of both form and substance. Over time, the writs were given names and thus different situations of liability, different ‘torts’, came into being.9 Today, there are dozens of them, each with its own regime and aimed at the protection of a particular interest.10 This explosion of civil liability by means of a multitude of 4 See C Corgas-Bernard, ‘Focus sur quelques non-dits du projet de réforme du droit de la responsabilité civile’, RLDC 2017, 152. 5 G Viney, ‘Pour ou contre un “principe général” de responsabilité pour faute? Une question posée à propos de l’harmonisation des droits civils européens’ in Le droit privé français à la fin du XXe siècle: Études offertes à Pierre Catala (Paris, Litec, 2001) 555. 6 M Bacache, ‘La recodification des principes classiques – Articles 1235 à 1238; 1241 à 1249; 1253 à 1256’ JCP G Suppl to nos 30–35 (25 July 2016) 22. 7 Some provisions, formulated in quite similar terms and sometimes even described as ‘general clauses’, would not be thought of in these terms by French lawyers. This would include, in particular, those which do not take into account the principle of full reparation of harm, which traditionally correlates with the general principle of liability for fault. See eg art 4:101 Principles of European Tort Law. 8 C von Bar, The Common European Law of Torts (Oxford, Clarendon Press, 2000) vol 2 [209]; WV  Horton Rogers, ‘Fault under English Law’ in P Widmer (ed), Unification of Tort Law: Fault (The Hague, Kluwer Law International, 2005) 69. 9 C van Dam, European Tort Law, 2nd edn (Oxford, Oxford University Press, 2013) [502-1]. 10 P Remy, ‘Réflexions préliminaires sur le chapitre Des délits’ in F Terré (ed), Pour une réforme du droit de la responsabilité civile (Paris, Dalloz, 2011) 15, 28. In the end, their only common feature was

The Definition of Civil Fault  81 individual wrongs makes it impossible to find a single concept of fault as a condition for recovery of compensation. Looked at more closely, it can be seen more as a justification for imposing liability.11 One qualification on this general picture must be made, however, owing to the tort of negligence, which experienced tremendous growth during the twentieth century, and where liability is based on ‘breach of a duty of care’. This is reminiscent of the fault with which French lawyers are familiar – and it makes comparison less difficult. This is good, because the comparison is quite fruitful. As a matter of intuition, we could expect that the French definition would seem rather restrained to the lawyer in a common law system, for whom the establishment of negligence is a subtle exercise.12 Before showing a ‘breach’, it is first necessary to establish the existence of a ‘duty of care’, which brings together a number of requirements. As a result, these two stages alone take up a striking number of pages in English textbooks, on a point on which French law has much less to say. Explanations can certainly be offered for this disparity. In particular, it should be recalled that English tort law originates in case law and this could explain why the tort of negligence has been subject to refinement. By contrast, the French law of civil liability was condemned to a brief exposition of the law because its source is legislative. Is not the function of legislation ‘to fix, from broad perspectives, the general maxims of law’,13 in order not to freeze it? Moreover, why would one have taken the trouble to specify the notion of fault more particularly, given that it took its language very largely from that which was current in 1804?14 This is not, however, a convincing explanation. Not only is legislation not static on the other side of the Channel,15 but the rise of the French law of civil liability is also not to belong to any other category, such as breach of contract: S Whittaker, Liability for Products: English Law, French Law and European Harmonisation (Oxford, Oxford University Press, 2005) 159. 11 See P Giliker, Tort, 6th edn (London, Sweet & Maxwell, 2017) [1-005]; see also C von Bar, The Common European Law of Torts (Oxford, Clarendon Press, 1998) vol 1, [263], [264], discussing the relationship between fault and the tort of private nuisance. 12 This is shown by the famous ‘Learned Hand test’, established in an American case, United States v Carroll Towing Co 159 F 2d 169 (2d Circuit 1947). According to this, a breach of duty may be established from the following equations: cost of measures necessary to avoid the harm (C) < probability that the harm will occur (P1) x loss suffered if the harm occurs (P2) = liability C > P1 x P2 = no liability. See Giliker (n 11) [5-015]. 13 JEM Portalis, Discours Préliminaire sur le Projet de Code Civil (1799), reproduced in JEM Portalis, Ecrits et Discours Juridiques et Politiques (Presses Universitaires d’Aix-Marseille, 1988) 21, 26. 14 See J-S Borghetti, ‘The definition of la faute in the Avant-projet de réforme’, in J Cartwright, S  Vogenauer and S Whittaker (eds), Reforming the French Law of Obligations (Oxford, Hart, 2009) 271, 272. 15 See eg Compensation Act 2006, s 1: ‘A court considering a claim in negligence or breach of statutory duty may, in determining whether the defendant should have taken particular steps to meet a standard of care (whether by taking precautions against a risk or otherwise), have regard to whether a requirement to take those steps might (a) prevent a desirable activity from being undertaken at all, to a particular extent or in a particular way, or (b) discourage persons from undertaking functions in connection with a desirable activity.’

82  Marie Dugué the work of the courts, which have not hesitated to deprive liability for one’s own personal actions of its original moral significance. However, it is clear that the courts have generally refused to define fault,16 or to set out clearly the criteria for its assessment, even though it constitutes a legal concept over which the Cour de cassation has long exercised control.17 Other justifications must therefore be sought. In doing this, it should be noted that liability for fault, and especially for ‘negligence’, occupies a considerably larger place in English law than in French law. This may be demonstrated by the fact that most of the leading cases which set the scope of the tort of negligence would not have been decided in France on the basis of the former article 1382 Cc.18 On the one hand, this is due to the reluctance shown by English law as regards strict liability, which was developed to a tremendous extent in French law from the end of the nineteenth century; on the other hand, it is also explained by the fact that English law brings within tort law which for us falls within the ambit of the public law of administrative liability.19 Nevertheless, it would be going too far to infer from this that liability for fault has become obsolete in French law, making it no longer of interest. The potential of fault for growth has been noted for quite some time,20 as has its propensity to conquer new areas of application.21 Because it has a subsidiary but universal role, liability for fault remains paramount: ‘where there is no other legal way forward, there it is, ready for use, with all its aspects well-tested’.22 Indeed, this impression has been reinforced by the recognition of its constitutional character, even though this is subject to various limitations.23 This is why one cannot help thinking that the definition of fault in fact provided by article 1242 of the Projet de réforme pursues a kind of avoidance strategy. It does not refuse to give a definition – how could it, given that the purpose of the reform is to promote ‘law that is understandable, transparent and provides

16 There are some exceptional, isolated decisions: see esp CA Paris 30 June 2006, RG no 04/06308, Bull Joly 1 September 2006, 594 note Schmidt: ‘Although it is mentioned, fault is not defined by ­articles 1382 and 1383, but it can apply in a case of omission as well as an act: the state of mind which it presupposes may take the form of an intention to cause harm, or it may be unintentional and constitute the contravention of a rule set by legislation or by custom, or even professional standards contained in codes of practice, and cannot be justified‚ such as by the proper exercise of a legal right; finally, any fault, however slight, gives rise to a right to full reparation of the harm caused – but only harm which is certain, existing and caused by the fault.’ 17 Cass civ 15 April 1873, S 1873, I, 174. 18 In particular, many would have been decided on the basis of liability for the action of things: see eg Bolton v Stone [1951] AC 850. 19 This point is made by Whittaker (n 10) 156. 20 P le Tourneau, ‘La verdeur de la faute dans la responsabilité civile (ou la relativité de son déclin)’ RTD civ 1988, 505. 21 F Arhab, ‘Les nouveaux territoires de la faute’, RCA 2003 chron 16. 22 P le Tourneau (ed), Droit de la responsabilité et des contrats: Régimes d’indemnisation, 11th edn (Paris, Dalloz, 2017) [011.16] (‘lorsqu’aucune autre voie juridique n’existe, elle est là, prête à l’emploi, aux tenants et aboutissants éprouvés’). 23 See M Bacache-Gibeili, Les obligations, La responsabilité civile extracontractuelle, 3rd edn (Paris, Economica, 2016) [129ff].

The Definition of Civil Fault  83 legal certainty’?24 But the Projet takes only a minimal step forward. Using general terms, its definition puts forward two kinds of fault – fault which is ‘a violation of a legislative requirement’; and fault which is ‘a failure in the general duty of care or diligence’ – and in this way promotes a distinction based on the nature of the duty violated. But a number of others could have been mentioned, for example, so as to distinguish according to the subject matter of the duty, according to the seriousness of the violation, between faults of omission and faults of commission, or between negligent and intentional fault. The question therefore inevitably arises as to whether the right choices were made. No doubt the English perspective on this issue will not be the same as the French. From the French point of view, the first thought will undoubtedly be whether the text of the Projet de réforme already says too much. Do we not miss the silence of the old law, which has provided the opportunity for the judicial creativity (and its reverse) that we have seen? From the English point of view, the question is rather whether it should have said more.25 In the following, we shall explore these two perspectives in turn.

I.  Does the Text of the Draft Provision Already Say Too Much? Is it going too far to define fault as a ‘violation of a legislative requirement or a failure in the general duty of care or diligence’? The first answer that comes to mind is no. At first sight, article 1242 of the Projet de réforme provides a definition of fault which is reassuring but, on reflection, may in fact be misleading.

A.  The Definition is Certainly Reassuring The definition provided by article 1242 of the Projet is reassuring because it is both traditional and flexible.

(i)  A Traditional Approach The notion of ‘personal action giving rise to liability’, which is adopted by the Projet de réforme, is doubly traditional, because it is in line with the main trends not only of the case law, but also of academic writing.

24 J-J Urvoas, ‘Présentation du projet de réforme de la responsabilité civile’ (speech to the Académie des Sciences morales et politiques, presenting the Projet de réforme, 13 March 2017). 25 cf S Whittaker, ‘La responsabilité pour fait personnel dans l’avant-projet de réforme du droit de la responsabilité: donner voix aux silences du Code civil’ RDC 2007, 92.

84  Marie Dugué As far as the case law is concerned, it must be remembered that one of its outstanding achievements has been to make fault an objective concept. Starting with a series of decisions on 9 May 1984,26 which took their lead from an earlier legislative development,27 the Cour de cassation now seems to treat fault as synonymous with mere unlawful conduct, without considering further whether it is imputable. To put this another way, fault may be committed by a person who is unable to appreciate the significance of his actions, such as an individual with a mental disorder or a very young child. In defining the concept of fault as it does, article 1242 certainly seems to confirm that it has been reduced to an objective element. There is no reference to any requirement of awareness. This interpretation is further confirmed by two subsequent provisions of the Projet. First, article 1242-1 provides – following the existing law – that fault may be imputed to a legal person where it results from a failure in its organisation or its functioning, without this being first shown to be the responsibility of a physical person acting as one of its organs, and this implies that there is no requirement of discernment on the part of the person who commits fault.28 Secondly, article 1255 of the Projet provides that ‘[u]nless it bears the characteristics of force majeure, fault in a victim who lacks discernment has no exonerating effect’, and this confirms that a person lacking discernment may commit fault. Simply put, any fault in such a person cannot undermine their right to reparation unless it satisfies the requirements of force majeure, which are quite stringent.29 This new provision is welcome: although now widely accepted, the idea that fault is objective still leaves a bitter taste. It was required to protect victims, but it worked against very small children who had been partly responsible for their own harm, as their behaviour could easily be established as objectively at fault. Article 1242 of the Projet is also entirely in line with academic writings. The definition that it uses has the advantage of being widely accepted by jurists. It seems to come straight from Planiol, who saw in fault ‘the contravention of a pre-existing obligation’.30 This approach to liability for one’s own action still remains the best-known, but it has certainly had its critics. In particular, its terminology has given rise to some reservations: what exactly is this ‘obligation’, given that 26 Ass plén 9 May 1984 (Lemaire and Derguini decisions), Bull AP no 2 and Bull crim no 164, D 1984, 525 concl Chabannes, note F Chabas, JCP G 1984 II 20255 note N Dejean de la Bâtie, JCP G 1984 II 20256 note P Jourdain, RTD civ 1984, 508 obs J Huet. These decisions show that a child may be liable without it having to be shown that he was capable of understanding the consequences of his actions. 27 The loi of 3 January 1968 had abandoned the principle that the insane (‘déments’) were not legally responsible, by adding into the Code civil a new art 489-2 (which is now art 414-3), providing that ‘One who has caused harm to another whilst under the influence of a mental disturbance nevertheless has the obligation to make reparation’. 28 See earlier, in relation to the Avant-projet Catala, J-S Borghetti (n 14) 276. 29 This development should certainly not be underestimated, given the extent to which the effect of the victim’s fault is transformed, even though the definition remains unchanged. 30 M Planiol, ‘Études sur la responsabilité civile’ RCLJ 1905, 283; idem, Traité élémentaire de droit civil, 5th edn (Paris, LGDJ, 1909) vol II, [863] (‘la contravention à une obligation préexistante’).

The Definition of Civil Fault  85 it has no identifiable debtor nor any real existence in advance of its breach?31 This problem could no doubt be avoided by using the term ‘duty’ instead of ‘obligation’, but the definition remains open to criticism as being abstract. Far from really defining fault, in the end we are left with the question of what these pre-existing duties are, breach of which would give rise to liability.32 An alternative proposal, seeking to define fault as misconduct that would not have been committed by a reasonable person in the same circumstances,33 says no more as it is still necessary to know what rules of behaviour have to be followed by this reference model in order to evaluate the misconduct. This explains why many writers today support the version proposed by the definition in the Projet de réforme,34 which had also been preferred by all the earlier reform proposals without any particular opposition.35 This is because, in addition to its traditional character, the definition has the advantage of being very flexible.

(ii)  A Flexible Approach Since civil liability is an area of the law that has been developed essentially by the courts, a number of writers have expressed the fear of writing it in the marble of legislation and thus preventing it from adapting to social changes.36 Would the definition in article 1242 of the Projet compel the courts to explain further the characteristics of fault, by indicating how a person’s conduct must contravene legislation or the general duty of care or diligence? Even if the provision were 31 See esp J Brethe de la Gressaye, ‘La théorie de la responsabilité en droit pénal et en droit civil’ Rev Gén Dr 1928, 193: ‘What kind of obligation is it that would have a specific individual as creditor, but the rest of humanity as debtor?’ 32 G Marty, ‘Illicéité et responsabilité’ in Études L Julliot de la Morandière (Paris, Dalloz, 1964) 339, 344; idem, ‘L’expérience française en matière de responsabilité civile et les enseignements du droit comparé’ in Mélanges offerts à Jacques Maury, vol 2, Droit comparé, théorie du droit et droit privé (Paris, Dalloz & Sirey, 1960) 173, 180; H and L Mazeaud and A Tunc, Traité théorique et pratique de la responsabilité civile délictuelle et contractuelle, 6th edn (Paris, Montchrestien, 1965) vol I, [392]. 33 Mazeaud and Tunc (n 32) [439]; G Marty and P Raynaud, Les obligations, 1 – Les sources, 2nd edn (Paris, Sirey, 1988) [457]. 34 G Viney, P Jourdain and S Carval, Les conditions de la responsabilité, 4th edn (Paris, LGDJ, 2013) [443]; Bacache-Gibeili (n 23) [143]; le Tourneau (n 22) [221-13]. 35 Art 1352(2) Avant-projet Catala: ‘Fault consists of breach of a rule of conduct imposed by legislation or regulation or failure to conform to a general duty of care and diligence’ (‘Constitue une faute la violation d’une règle de conduite imposée par une loi ou un règlement ou le manquement au devoir général de prudence ou de diligence’); art 5 Avant-projet Terré: ‘Fault consists in committing an unlawful action, voluntarily or by negligence. An action is unlawful when it contravenes a rule of conduct imposed by legislation or by the general duty of care and diligence’ (‘La faute consiste, volontairement ou par négligence, à commettre un fait illicite. Un fait est illicite quand il contrevient à une règle de conduite imposée par la loi ou par le devoir général de prudence et de diligence’); art 1242 of the Bill presented by the Chancellerie on 29 April 2016: ‘The violation of a rule of conduct imposed by legislation or a failure in the general duty of care or diligence constitutes fault’ (‘Constitue une faute la violation d’une règle de conduite imposée par la loi ou le manquement au devoir général de prudence ou de diligence’). 36 See eg Bacache (n 6) 21.

86  Marie Dugué to be interpreted in this way,37 it must be admitted that it would not complicate the work of the courts greatly and would still give them a great deal of freedom. Article 1242 is certainly very accommodating, whichever limb of the definition we consider. In the first place, fault may consist of ‘a violation of a legislative requirement’. It might be surprising, at first sight, that the Projet de réforme refers to legislation (la loi), given that fault is routinely inferred by the courts as the contravention of administrative regulations (les règlements).38 However, this omission is without doubt an indication that the term ‘legislative’ would have to be given its extended meaning; ‘legislation’ certainly does not refer only to rules voted by Parliament in the area reserved to it by article 34 of the Constitution (le domaine législatif in a strict sense as distinct from le domaine réglementaire), but can also be understood as referring to all the rules of written law originating from the State.39 Thus, fault would probably arise from the disregard of any provision (texte) that imposes a duty of conduct on individuals, directly or indirectly. The Projet provides only one exception, as article 1234 states that: [w]here non-performance of a contract causes harm to a third-party, the latter can claim reparation of its consequences from the debtor only on the basis of extra-contractual liability, and subject to that third-party’s establishing one of the actions giving rise to liability.40

This means that article 1103 (formerly 1134(3)) of the Code civil’s provision that ‘[c]ontracts which are lawfully formed have the force of legislation for those who have made them’ will no longer allow contractual non-performance to be regarded as a delictual fault: non-performance of a contractual provision is not in itself a delictual fault. Under article 1242 of the Projet, fault may also lie in ‘a failure in the general duty of care or diligence’. This is simply a reference to a formula that French civil lawyers know well and, although they sometimes use different words, the same idea underlies the expressions ‘general norm of civilised behaviour’,41 ‘fundamental norm of behaviour’,42 or general duty of care and/or diligence: that is, ‘beyond and outside existing legislation’, there is ‘an unknown legal territory, or a

37 This is certainly not settled, as emphasised by J-S Borghetti, ‘L’avant-projet de réforme de la responsabilité civile’ D 2016, 1442, discussing, however, the draft of 29 April 2016: ‘it is not clear that the Cour de cassation will be inclined to depart from its present brevity, and the definition contained in article 1242, even if it is one day given legislative force, may have only, in essence, doctrinal value’ (‘il n’est pas certain que la Cour de cassation soit disposée à abandonner son laconisme actuel et la définition de l’article 1242, à supposer qu’elle acquière un jour force de loi, pourrait ne conserver qu’une portée essentiellement doctrinale’). 38 See eg Cass civ (3) 29 November 1983, Bull civ III no 247. 39 G Cornu, Association Henri Capitant, Vocabulaire juridique, 12th edn (Paris, PUF, 2018), ‘loi’. 40 On this provision and its background see further Ch 4, above, esp at pp 62–64. 41 J Darbellay, Théorie générale de l’illicéité (Fribourg, Éditions universitaires, 1955) [160] ff. 42 M Puech, L’illicéité dans la responsabilité civile extracontractuelle (Paris, LGDJ, 1973) [31] ff.

The Definition of Civil Fault  87 grey area’43 in which harmful behaviour, which the legislator has not, or at least, not yet, foreseen, can be sanctioned. This area is relatively uncertain: ‘it is not known in advance whether acts that fall within it will be declared lawful or unlawful’.44 It  is for the court to decide. The general duty of care or diligence is therefore intrinsically endowed with a nature that is vague and indeterminate, providing more of a framework. The courts would gradually sketch out its component parts: a duty of loyalty, a duty of precaution,45 and so forth. But what should be their methodological approach? The Projet de réforme says nothing about it, and leaves the door open for any technique. The definition may therefore seem reassuring, but its appropriateness is open to question.

B.  A Definition which may be Misleading By choosing to use a list to define the action of persons giving rise to liability, the Chancellerie highlights two kinds of fault: fault consisting of ‘a violation of a legislative duty’, sometimes referred to as ‘fault by contravention’, and fault consisting in ‘a violation of the general duty of care or diligence’. It is questionable, however, whether this distinction is useful, either in teaching or in practice.

(i)  A Dichotomy which is not Useful in Teaching? The purpose of definitions is to provide clarification, and the latter plays a major role in legal education.46 It remains to be seen whether article 1242 of the Projet is useful in teaching, and in particular whether it conveys more clarity than confusion. At first glance, the answer is in its favour: it is normal to separate fault which consists in a violation of legislation from fault which consists in a violation of the general duty of care and diligence, and this seems to show that this presentation will be useful in didactic terms. However, its appropriateness may be questioned at two levels. In terms of sources, a rigorous approach would have contrasted a violation of legislative duties with a violation of extra-legislative duties, or a violation of ­legislative duties with a violation of duties developed by the case law, or even customary duties.47 As it stands, the provision establishes an asymmetry. Fault by contravention is as narrow as fault by carelessness is wide. Indeed, the latter 43 F Guisan, ‘La protection de la personnalité et le boycott commercial’ in Études de droit commercial, Mélanges en l’honneur de Carl Wieland (Basel, Helbing & Lichtenhahn, 1934) 149, 160. 44 Ibid. 45 For a different view, see Bacache (n 6) 23: ‘by referring only to the general duties of care or ­diligence, the provision excludes other duties which the courts have recognised for two centuries, whose violation constitutes fault. In addition to the duties of good faith, loyalty, safety and respect for the rights of others, we can add the very recent duty of precaution’. 46 L-M Schmit, Les définitions en droit privé (Paris, LGDJ, 2017) [474]ff. 47 For this approach, see Viney, Jourdain and Carval (n 34); le Tourneau (n 22).

88  Marie Dugué seems liable to encompass everything. No doubt this explains why the definition here does not mention abuse of right, even though this has been recognised as fault in French law, for does not an abuse of one’s rights also constitute a breach of the general duty of care or diligence? Carelessness has the potential to devour everything, and one ends up wondering whether a violation of legislation itself should not equally have been included within it. English law avoids all this, by making the tort of negligence and breach of statutory duty different torts, each with their own rules. Thus, while a violation of legislation may be an indication of negligence (in the sense of a lack of reasonable care), it does not necessarily characterise it.48 This approach is a priori the more rigorous: taken in their general and usual sense, carelessness and negligence imply a lack of attention and care, so that they cannot be excluded systematically from legislative provisions, especially those that define obligations as being of result and therefore require the attainment of a certain objective without reference to the efforts of the contractual debtor. But it certainly seems that this obstacle has been overcome in French law:49 the boundary between the two types of fault has indeed been blurred. Thus, there are many legislative provisions which are imprecise and do nothing more than impose a general duty of care or diligence, so that the characterisation of fault cannot be made without evaluating the defendant’s behaviour by reference to an external standard. The analysis is similarly extremely difficult when the legal provision that is alleged to have been infringed establishes a right for an individual (un droit subjectif), since it is then necessary to undertake a balance of the interests involved to ascertain whether or not the right has in fact been violated.50 Conversely, an assessment of a failure in a general duty of care and diligence is often based on a finding that a rule imposing a particular result has not been complied with:51 in this case the rule need not be legislative (otherwise the detour through the general duty of care and diligence would be superfluous) but sporting, technical, ethical – or even one created by case law. In this way, there is no difficulty in holding that it is negligence to cause a climber to fall;52 as well as (in the older law) to deliver a defective product.53 In terms of proof of fault, if there is to be any distinction, it should be between fault comprised in a violation of a defined duty and fault comprised in a violation of an undefined duty, rather than between fault by contravention, and fault by carelessness or negligence. However, if such a structure is not to be adopted,

48 See Whittaker (n 10) 190–91. 49 ibid, esp 160. 50 The conclusion that there has been a violation of the right to respect of a person’s private life, as established by art 9 Cc, is based on it having been checked that the infringement of privacy is not justified by the public’s right to information. 51 The same observation can also be made of English law, where we see that in certain areas a lack of care is assessed so narrowly that it amounts to strict liability: Giliker (n 11) [5-004]. 52 Cass civ (2) 18 May 2000, Bull civ II no 85, JCP G 2000 I 280 no 11 obs G Viney, Gaz Pal 21–23 April 2002, 15 note A Bolze. 53 See eg Cass civ (1) 17 January 1995, Bull civ I no 43, RTD civ 1995, 631 obs P Jourdain.

The Definition of Civil Fault  89 it would be preferable to abandon the dichotomy altogether, and to follow the path taken by the Principles of European Tort Law: fault is the violation of the required standard of conduct,54 it being specified that ‘[r]ules which prescribe or forbid certain conduct have to be considered when establishing [this standard]’.55 This would have the advantage of giving the court some room for manoeuvre, rather than (as in the present provision proposed by the Projet de réforme) requiring it to find that there is fault where a legislative provision is clearly violated which is certainly not without difficulty in practice.

(ii)  A Dichotomy which is Unhelpful in Practice? The assimilation of a breach of a legislative requirement to fault, as is done by article 1242 of the Projet de réforme, is likely to compel the courts to hold that there is an action giving rise to liability (fait générateur) whenever it is clear that the person has contravened a provision which has the force of legislation or of regulation. However, although it is generally claimed that such a contravention ipso facto constitutes a ‘civil fault’,56 the case law does not seem so clear. Actually, it is not uncommon for courts to infer fault from violation of legislation, but they do sometimes find a contravention without drawing any conclusion in relation to liability. The cases are rare, certainly, but they do exist. For example, in one case, the Cour de cassation stated that ‘the mere fact that a person conducts an affair with a married man does not constitute such a fault as to engage her liability towards his wife’. This solution applies where the mistress is herself married, when she violates in consequence article 212 Cc, which establishes a duty of fidelity between spouses.57 In the same way, the Cour de cassation has had occasion to affirm that a victim of fraud could not rely on the non-observance by a bank of certain provisions of the Monetary and Financial Code in order to claim damages against the financial institution.58 Breach of legislation does not always seem to constitute fault. Even if the definition of fault in the Projet is accepted, it would clearly be useful to clarify it a little.

II.  Should the Legislative Definition have Said More? If the reform called for a definition of fault, should not the Chancellerie have used it to shape this notion more clearly? Would it not have been useful to specify how 54 Art 4:101. 55 Art 4:102(3). 56 See eg Viney, Jourdain and Carval (n 34) [448]; le Tourneau (n 22) [2211.04]; Bacache-Gibeili (n 23) [144]. 57 Cass civ (2) 4 May 2000, JCP G 2000 II 10356 comm T Garé, RTD civ 2000, 810 obs J Hauser. 58 Cass com 28 April 2004, Bull civ IV no 72, JCP G 2004 II 10105 note C Cutajar, D 2004, 1380 obs V Avena-Robardet, RD banc fin 2004, 273 note F Boucard.

90  Marie Dugué fault is to be assessed? Would not it have been appropriate to say something about other traditional classifications of fault? These are the questions we try to answer in the following paragraphs.

A.  How Fault is to be Assessed Article 1242 of the Projet leaves it a mystery as to how the infringement of the general duty of care or diligence is to be determined. It does not seem to attach any importance to the purpose of the rules that have been violated; nor does it say anything about the reference model that is to be used.

(i)  The Purpose of the Rules that have been Infringed In English law, negligence is an undeniably relative concept in the sense that, as is often emphasised, ‘it does not exist in the air’: it can thus be found only in relation to particular persons, and particular harm. It is the concept of ‘duty of care’ that gives expression to the ‘relativity’ of this form of fault: the fact that the defendant did not act as a reasonable person would have done is not necessarily a ground of liability, even where it has caused loss; it must first be shown that the defendant had a duty to act with care in favour of the claimant. The requirement of a duty of care may seem totally incongruous to a French lawyer to whom it seems pointless, as in French law the duty of care and diligence is ‘general’, as the Projet de réforme makes clear – that is to say, it is owed by ­everyone, at all times, in all places, and for the benefit of all. This leaves the only question as being whether it has been breached.59 The English approach, however, is not without interest; the duty of care has been seen as a technique for controlling the extent of civil liability:60 it is the means by which English law determines the scope of interests that are protected against negligence (‘scope of duty’).61 In this respect, in practice, decisions on duty of care reflect a clear reluctance to compensate intangible harm, whether purely intangible loss (notably, ‘moral harm’ such as mental distress) or (and in particular) purely economic loss.62 Pure economic loss rarely satisfies the criteria for the recognition of a duty of care, because in addition to foreseeability of the damage and proximity between the parties, it is

59 See Borghetti (n 14) 284 ff, regretting that fault is ‘an increasingly abstract and disembodied idea’. 60 J Fleming, ‘Remoteness and Duty: The Control Devices in Liability for Negligence’ (1953) 31 Can Bar Rev 471; and on the origins of ‘duty’, see PH Winfield, ‘Duty in Tortious Negligence’. 61 See esp Lord Bridge in Caparo Industries plc v Dickman [1990] 2 AC 605, 627: ‘It is never sufficient to ask simply whether A owes B a duty of care. It is always necessary to determine the scope of the duty by reference to the kind of damage which A must take care to save B harmless.’ 62 By contrast, ‘in general where a person’s positive conduct leads to physical harm in another person, the courts recognise an existence of the duty of care’: S Whittaker (n 10) 181.

The Definition of Civil Fault  91 also necessary that it appears ‘fair, just and reasonable’ to require the defendant to compensate them.63 Compensation for this category of loss on the basis of negligence often requires an ‘assumption of responsibility’:64 the defendant must have assumed the risk that his act will cause harm to another.65 It therefore remains exceptional for an award to cover pure economic loss. It is not surprising that this restrictive approach extends also to the tort known as breach of statutory duty. Where a violation of a legislative provision is relied on, the claimant must first show that the legislature intended that the breach could be the basis of an action for damages,66 and that the provision was intended to protect him against the harm in fact suffered.67 Such reasoning is known in France under the name of the ‘theory of Aquilian relativity’: but would it not be worth introducing it into our law? It seems that the answer should be yes, where a specific rule has been infringed, since this approach makes it possible for certain claims for compensation to be rejected where they are based, almost ‘accidentally’, on the breach of a norm whose intended scope was quite different. Moreover, French courts are not entirely hostile to the ‘theory of Aquilian relativity’ in such a case. Although it is true that they do not explicitly refer to it, it seems that the theory provides a good explanation of those cases in which the courts refuse to find fault in the simple violation of a rule, even if it is contained in legislation or a regulation. The explanation lies in the scope of the norm that has been breached: the victim’s claim generally seems to succeed where the rule was intended to protect him.68 It would therefore have

63 On this requirement, see T Weir, An Introduction to Tort Law, 2nd edn (Oxford, Oxford University Press, 2006) 41ff; Giliker (n 11) [2-015]. This test, referred to as the ‘Caparo test’ may, however, be out of favour: J Goudkamp, ‘A Revolution in Duty of Care?’ (2015) 131 LQR 519, discussing Michael v Chief Constable of South Wales [2015] UKSC 2, [2015] AC 1732. 64 Whittaker (n 10) 161. 65 On this, see the seminal decision in Hedley Byrne & Co v Heller and Partners Ltd [1964] AC 465. For more recent cases, see eg Henderson v Merrett Syndicates Ltd [1995] 2 AC 145 and Lejonvarn v Burgess [2017] EWCA Civ 254, [2017] BLR 277. 66 This is not straightforward: see eg Enterprise and Regulatory Reform Act 2013, s 69, which provides that breach of health and safety regulations introduced by the Health and Safety at Work Act 1974 shall not generally give rise to an action for damages, even though it is in the context of accidents at work that the tort of breach of statutory duty traditionally operated. 67 The idea that the duty was intended to protect the claimant may be seen in Knapp v Railway Executive [1949] 2 All ER 508 (CA). There, the gate of a railway crossing, which was closed against vehicles on the highway, but was not securely fixed as required of the railway company by a statute, was struck by a car with the result that it swung in front of a train which hit it, injuring the engine driver. It was held that the engine driver had no right of action against the railway company, because the statute was intended for the benefit of users of the highway. An illustration of the requirement that the duty must relate to the harm suffered by the claimant can be found in Gorris v Scott (1874) LR 9 Ex 125. There, a statute, which provided that areas on board vessels used for the transport of animals should be equipped with pens, was not complied with, and sheep carried on a vessel were washed overboard and lost during a voyage. Their owner’s action for damages was rejected on the basis that the purpose of the statutory provision was to prevent the spread of contagious diseases amongst the animals, rather than to preserve them in case of storm. 68 Cass com 28 April 2004 (n 58) is quite clear in this respect. See also, in relation to ethical rules, M Dugué, L’intérêt protégé en droit de la responsabilité civile (Paris, LGDJ, 2019) [86].

92  Marie Dugué been useful for the Projet to have included this nuance within its definition of fault; it would not have been difficult to provide that, where the ‘requirement’ has a purpose that can be clearly determined, this must be taken into account in the assessment of fault. On the other hand, it is right that the Projet retained the ‘general’ aspect of the duty of care and diligence. This is not because French law, which is imbued with a less liberal philosophy than that which prevails across the Channel, has no need of a means of circumscribing the extent of liability, a role which causation is generally said to perform unofficially.69 The reason is rather that to say that a person is sometimes allowed to be negligent would not only collide head-on with French tradition, it would also have been superfluous. This is because a court may already take into account the nature of the damage that has been caused, as well as the relationship between the parties in determining whether the defendant’s conduct has departed from the reference model: it is sufficient to ask how a reasonable person would have acted, bearing in mind these factors.70 Therefore, any ‘relativisation’ of the general duty of care and diligence on the model of English law risks appearing redundant.71 As it stands, however, the extent of this risk remains difficult to assess, as the outline of the reasonable person remains unclear.

(ii)  Delineating the Reference Standard As already pointed out, an assessment of a violation of the general duty of care or diligence traditionally calls for a process of comparison between the conduct of the person whose action is to be assessed and the conduct of a reasonable person in similar circumstances. The standard is intended to be objective in both French law and English law, but in both systems the same questions arise as to the circumstances that should be taken into account. An assessment made in the abstract allows the balancing of various elements, but identifying them is not easy. In this regard, the thinking is perhaps less fully developed on the French side, and the Projet de réforme seems to be content with this as it gives no indication of the method to be followed. Some things are clear. It seems to be accepted in the current law that the model for comparison should not be endowed with the defendant’s shortcomings: things such as clumsiness or thoughtlessness could not be taken into account without reducing his fault to zero, at the expense of his victim. Equally, we know that some

69 See eg P Malaurie, L Aynès and P Stoffel-Munck, Les obligations, 9th edn (Paris, LGDJ, 2017) [96]. 70 This is noted by Whittaker (n 10) 160: ‘fault defines the scope of the tort as well as founding the basis of liability’. 71 This position has also sometimes been disregarded in English law: issues relating to the duty of care and to breach of duty tend to be conflated, to such an extent that some have advocated the ‘deconstruction’ of the former: D Nolan, ‘Deconstructing the Duty of Care’ (2013) 129 LQR 559; J Goudkamp, ‘Breach of Duty: A Disappearing Element of the Action in Negligence?’ (2017) 75 CLJ 12.

The Definition of Civil Fault  93 of the defendant’s particular positive attributes should be taken into account,72 such as his profession. Beyond that, however, there are grey areas. Some questions do not receive a clear answer. For example, of what relevance is the defendant’s age? English law seems to favour an adaptation of the required standard of care;73 paradoxically, in French law, where the question of the capacity of small children to commit fault has given rise to passionate debate, the answer remains unclear. The case law is not settled,74 and the Projet is careful not to take sides either way. Furthermore, some other questions seem always to be avoided. The English books discuss in some detail the criteria that should play a role in the characterisation of a breach of duty: customary practices, as well as the foreseeability of the damage, its severity, the magnitude of the risk run, the cost of measures necessary to avoid it or even the social utility of the defendant’s conduct; but the French books have much less to say.75 French scholars usually prefer instead to give examples of fault, rather than to identify the justifications that may explain the finding of fault. This should not really be surprising. The decisions of the Cour de cassation are sparsely justified, and it is never clear what considerations led the judges of our highest court to find that there was or was not fault. This does not, however, prevent our speculating about it. Notably, it seems neither probable nor desirable that the degree of care required should be the same where the risk is of bodily injury as opposed to purely economic loss. It can be seen that the infliction of personal injury on another person almost always is found to have constituted fault, whereas the causing of economic loss to another does not always do so,76 as a recent decision seems to confirm.77 In this case, a man had played a football lottery (‘Loto Foot’), making predictions on 14 matches. Only one of these predictions had proved wrong: a goal was scored, while the gambler had bet on a draw. He argued in relation to the referee that the goal that was scored was offside, and therefore constituted fault, and the gambler sought reparation of his financial loss

72 N Dejean de la Bâtie, Appréciation in abstracto et appréciation in concreto en droit civil français (Paris, LGDJ, 1965) [46]: ‘l’homme raisonnable met à profit … les aptitudes et les connaissances particulières qu’il possède. On comprend donc facilement que ces aptitudes spéciales puissent ­ contribuer à rendre fautive la conduite de celui qui n’en tire pas tout le parti désirable’ (‘a reasonable man makes use of … the particular skills and knowledge that he possesses. It is therefore easy to understand that these special skills can contribute to a finding of fault in the case of one who does not use them as fully as he ought’). 73 Mullin v Richards [1998] 1 WLR 1304 (CA). 74 See Viney, Jourdain and Carval (n 34) [467]. 75 See Whittaker (n 25), noting that, unlike in the common law, there is no clear guide as to the different factors the courts should take into account and evaluate in order to justify their decision, nor is any attention paid to a cost/benefit analysis in economic or social terms. See also F Viney, ‘La personne raisonnable’ (Thesis, University of Paris I, 2013) [771]: ‘French lawyers have … nothing to fear in developing more explicit reasoning in relation to the question of the assessment of fault.’ 76 See esp B Starck, Essai d’une théorie générale de la responsabilité civile considérée en sa double ­fonction de garantie et de peine privée (Paris, L Rodstein, 1947) 85; M Fabre-Magnan, Droit des obligations, 2 – Responsabilité civile et quasi-contrats, 3rd edn (Paris, PUF, 2013) 93. 77 Cass civ (2) 14 June 2018, D 2018, 1784 note J-S Borghetti, RTD civ 2018, 908 obs P Jourdain.

94  Marie Dugué both from the referee and the football club in relation to what he would have won if the goal had not been recognised. Unsurprisingly, the Cour de cassation approved the decision of the Court of Appeal which rejected the claim. However, the wording of the decision gives us something to think about, as it stated that ‘only an action intended deliberately to undermine the risk inherent in a sporting bet is of a nature to engage liability in a player … with respect to the gambler’, as the court thereby accepted that fault committed against an opponent, such as a dangerous tackle which might give the injured player a right to reparation, would not constitute fault against the gambler even where the action thwarted his predictions. In French law, fault is therefore not such an absolute notion as it seems. While it does indicate what is unlawful and what is not, it also plays a part in establishing a differentiated protection of interests. In the end, fault therefore fulfils in French law the two functions assigned in English law to duty of care and breach of duty. The question therefore arises whether it would not have been helpful to give a greater ‘visibility’ to this dual function in the Projet, and, more broadly, to set out the criteria which influence courts in the process of weighing up the degree of misconduct. They would not have to be exhaustive: it would have been possible to set them out in the same way as the Principles of European Tort Law.78 Similarly, it would have been useful for the Projet to press the analysis a little further and propose a more precise classification of the kinds of fault that can occur.

B.  Classification of Fault As it stands, article 1242 of the Projet merely distinguishes fault by ‘contravention’, and fault by ‘carelessness’. But it could very well have set out other kinds of fault, which are just as common:79 fault by commission, contrasted with fault by omission; and minor fault, contrasted with serious fault.

(i)  Fault by Commission; Fault by Omission Article 1242 does not say much about the way in which fault should be assessed; and it does not indicate whether it is appropriate to proceed in the same way in a case of an omission on the part of the defendant, and of his action. Are there no reasons to provide a special treatment of omissions, other than as regards cases where an omission contravenes a legislative requirement, or is motivated by an intention to injure? 78 Art 4:102: ‘The required standard of conduct is that of the reasonable person in the circumstances, and depends, in particular, on the nature and value of the protected interest involved, the d ­ angerousness of the activity, the expertise to be expected of a person carrying it on, the foreseeability of the damage, the relationship of proximity or special reliance between those involved, as well as the availability and the costs of precautionary or alternative methods.’ 79 On this, see P Brun, ‘Premiers regards sur l’avant-projet de réforme du droit de la responsabilité civile’ RLDC 2016 no 140 [6].

The Definition of Civil Fault  95 This is the approach taken by English law, in which the distinction between ‘pure omission’ or ‘non-feasance’ and ‘action’ or ‘misfeasance’ is very well developed and reflects a reluctance to impose duties to act on people except where legislation has made provision to this effect. In particular, there is no general duty to rescue others.80 The reasons for this cautious approach are well known. The imposition of such a duty would seriously compromise the personal freedom in society, it would have no economic justification, and it would often be arbitrary in terms of causation.81 In this latter respect, it has been noted that although the occurrence of harm implies that there were causes of such a nature as to give rise to it, it also presupposes ‘the absence of events to disturb the causal process’,82 and these are so numerous that it would be absurd to ‘stigmatise’ just one of them, one person’s failure to intervene.83 French law has not failed to take notice of this controversy.84 At first, it also refused to condemn omissions, before revising its position when the omission could be linked to an action, and more generally to the creation of a risk. However, further developments occurred, so that today it is traditionally taught that fault can be found entirely in an omission, even a pure omission,85 and it goes without saying that there is no immunity for a person who fails to rescue a person in distress where he could do so without any danger to himself.86 Does this mean that the question is only a ‘cultural matter’?87 This is doubtful. Indeed, some writers note that a failure to act continues to receive special treatment in French law, and a certain reserve can be detected in recent cases, ‘which makes one hesitate to conclude that it has become routine to recognise fault by omission’.88 Should we regret that the Projet de réforme has not clarified things here? My answer is a qualified one. It is not clear that fault by omission and fault by commission are assessed in fundamentally different ways: a court must determine whether a duty has been breached, and this may lead it to consider what a reasonable individual would have done. It is the intensity of the analysis which

80 Weir (n 63) 1: ‘you may ignore an infant drowning in a pond unless it is your infant or your pond or you are the lifeguard’. 81 These arguments appear clearly in the opinion of Lord Hoffmann in Stovin v Wise [1996] AC 923, 943. 82 N Dejean de la Batie, obs on Cass civ (2) 29 January 1969, JCP G 1969 II 16027. 83 R Teisseire, ‘Essai d’une théorie générale sur le fondement de la responsabilité’ (Thesis, University of Aix, 1901) 192. 84 See esp J Carbonnier, ‘Le silence et la gloire’ D 1951 chron 119. 85 P Jourdain, Les principes de la responsabilité civile, 7th edn (Paris, Dalloz, 2007) 49. 86 Such behaviour, if voluntary, would even give rise to a criminal penalty under art 223-6 CP. 87 A-F Delvilani, ‘La responsabilité civile pour omission ou abstention’ (Thesis, University of Grenoble, 1978) 303 (un ‘sujet de civilisation’). 88 Brun (n 3) [314]. The example referred to is Cass civ (1) 18 April 2000, Bull civ I no 117, RCA 2000 comm 213 (H Groutel) in which a person who had slipped on an icy pavement on which neither sand nor salt had been spread sued the occupant of the adjacent building. The Cour de cassation disapproved the decision below which had found fault on the part of the occupier: ‘by so deciding, without considering (as it had been requested to do) what legislative or regulatory provision required such steps, the Cour d’appel had failed to give a legislative basis for its decision’.

96  Marie Dugué the courts undertake that varies: it certainly seems difficult to deny that omissions are not viewed with the same rigour as acts of commission. Indeed, legislation recognises this, for example, when it treats a lie more seriously than fraudulent silence.89 There is every reason to believe that the factors discussed above – including the nature of the harm, the ‘cost’ and even the dangerousness of the measures necessary to avoid it, and the nature of the relationship between the victim and the defendant – are decisive here,90 because it is ‘more difficult to define duties to act than obligations not to act’,91 and so some guidance is welcome. Again, this could have provided an incentive to highlight these factors in the legislation. Perhaps more problematic, though, is the Projet’s lack of reference to the seriousness of fault.

(ii)  The Seriousness of the Fault The Code as it currently exists contrasts delicts with quasi-delicts, ie intentional with non-intentional fault. Breaking with this logic, article 1242 of the Projet chooses to bury all reference to the seriousness of the conduct committed by the person potentially liable. Other reform proposals, however, had set about this differently; the Avant-projet Terré, for example, in article 5, stated that ‘fault consists in committing an unlawful action, voluntarily or by negligence’, an action being unlawful ‘when it contravenes a rule of conduct imposed by legislation or by the general duty of care and diligence’.92 In this provision, the Avant-projet Terré intended to emphasise that fault has a function in imputing responsibility beyond its normative role of saying what is and is not permissible: it attributes the prohibited behaviour to a person who is designated as being the one with a duty to make reparation.93 However, this formulation had its own drawbacks: indeed, it verged on tautology, in that the violation of the general duty of care and diligence is traditionally regarded as indicative of negligence. Its use would therefore have required some (too) subtle adjustments.94 This does not mean, however, that the silence of article 1242 on this point is completely unproblematic. Indeed, it supports the widespread idea that on principle the seriousness of the behaviour is irrelevant at the stage of deciding whether civil liability should or should not be imposed. Intentional or even merely voluntary fault will of course 89 See esp art 1137 Cc, which provides that ‘it is not fraud for a party not to reveal to the other contracting party his assessment of the value of the act of performance’, but does not say that a lie of this kind cannot be sanctioned. 90 In the case of fraudulent silence, again, the case law suggests that the existence of a particular relationship between the parties could give rise to a duty to inform about value: see esp Cass com 27 February 1996 (arrêt Vilgrain) JCP G 1996 II 22665 note J Ghestin, D 1996, 618 note P Malaurie. 91 Bacache-Gibeili (n 23) [165]. 92 Quoted above, n 35. 93 Some other legal systems, such as German law, make this clear by limiting fault to the latter role, and addressing the former by a different concept, referred to as ‘unlawfulness’: C Bloch, ‘Définition de la faute’ in Terré (n 10) 105–06. 94 ibid 110.

The Definition of Civil Fault  97 be sanctioned, but it does not require separate treatment. Unless it consists of a ­violation of a legislative provision, the intentional nature of the fault may be taken into account as part of the breach of the general duty of care and diligence. However, there is no doubt that the finding of fault is easier where there is an intention to cause the harm, and sometimes it even depends on the existence of such a state of mind. The case of Loto Foot, already mentioned, shows this clearly: the weaker the interest that is harmed, the more culpable the conduct must be for liability to be imposed. It would therefore have been useful to make clear that the seriousness of the actor’s behaviour is one of the criteria that play a role in the assessment of fault. Indeed, English law traditionally takes this approach. Thus, the economic torts, which are directed at the reparation of purely economic loss, often require proof of serious misconduct; intention to cause the claimant to act in a particular way is generally required (in inducing breach of contract and deceit) or even intention to cause harm (in intimidation).95 Comparison between the two legal systems illustrates once again how subtle and complex the notion of ‘fault’ is, but the French legislative provisions persist in obscuring this. Our purposes are not easy to achieve, but it would be possible to broaden the definition of the conduct giving rise to liability so as to lessen the impression that its characterisation is only a matter of judicial intuition (sometimes bordering on legal uncertainty), without in any way diminishing the courts’ room for manoeuvre.96

95 Weir (n 63) 199. 96 A Touffait and A Tunc, ‘Pour une motivation plus explicite des décisions de justice notamment celles de la Cour de cassation’ RTD civ 1974, 32, [10], noting that the style of French judgments ‘allows propositions to be presented as findings of fact which conceal difficult legal issues’, such as fault.

98

6 Crime, Breach of Legislative Duties and Fault MATTHEW DYSON

I. Introduction The Projet de réforme de la responsabilité civile provides an excellent opportunity to assess the concepts and reasoning engaged by the law of tort. One particularly important aspect of this is what unity underpins two important concepts, ­‘legislative duty’ and ‘fault’, and in particular, what unity can be found across tort law and criminal law and how they interact? Is ‘fault’ the same thing, doing the same work, across tort and crime? These are important questions: the process of seeking answers to them reveals a lot about what tort law, and the wider legal system it is part of, does and how it is understood by the legal actors within it. Questions like these also highlight hierarchies and processes of raising and resolving uncertainties and ambiguities. The first step towards understanding French tort law’s position on crime, breach of legislative duties and fault will be to set out the key ways tort law and criminal law connect (section II). Attention will then be turned to what we can learn about fault (section III), and how legislative duties are integrated across tort and crime (section IV). From there, we can assess what insights we have gained about unity in fault and in legislative duties (section V), and how and why any unity has ­developed (section VI). This kind of inquiry also helps to put French law itself in context, as briefly noted at the end of the chapter (section VII). As will be seen, it is an outlier compared to other European legal traditions. For at least four reasons it is exceptional, since it displays: (i) a unilateral and extensive integration of legislative duties into tort law; (ii) extensive integration of civil law into criminal procedure accompanied by some procedural safeguards against abusive claims; (iii) apparent freedom from inhibition in prioritising compensating through the tort system those who have suffered harm; and (iv) a focus on shorter, simpler and arguably more opaque formal tort rules, even if the current Projet de réforme represents some slight steps towards greater formal enunciation of the relevant rules.

100  Matthew Dyson

II.  Interfaces Between Tort and Crime Tort law and criminal law interact on at least five levels: the institutions engaged by them, the reasoning of legal actors, substantive content, procedural aspects, and the resolutions available to legal disputes.1 All five will be introduced briefly, but the focus of this chapter will be on two issues of substance: duties and fault. First, through what institutions does a legal system construct a relationship between tort and crime? For most systems, criminal law is most closely associated with public law: involving state power, but expressing a punitive rather than constitutional character. By contrast, tort is uniformly seen as part of the law of obligations, private and civil law. Tort law can still adjust for context, such as nuances in actions against public authorities, but it remains tort law. This is certainly the case for English law. In France, criminal law is in a distinctive position compared to England. First, the general law (ordre judiciaire), including tort and crime, is contrasted with administrative law, for the purposes of the organisation of French legal actions and the courts that hear them. That is, French administrative law’s own distinctiveness downplays the distinctiveness we perceive elsewhere, particularly between tort and crime. Secondly, whatever the precise relationship, France does distinguish between ‘criminal’ and ‘civil’. They are significantly separated from each other: separate courts, separate competences, separate rules of evidence and procedure and, often, separate purposes or focuses. Identifying the bases for such separation is difficult, and beyond the scope of this chapter.2 It tends to feature a combination of moral claims, public or private concern, penalties contrasted with compensation, and/or some positivist categorisation, particularly for criminal law. The reality appears to be that whatever normative claims are made, in practice courts and legal actors are able to determine whether a particular dispute is civil, and its resolution governed accordingly by either civil procedure or criminal procedure as the appropriate mechanism. French law, like most other legal systems, acknowledges that there are links from tort/crime out to other areas of law, such as contract, family law and property law. There are also links into newer divisions of the law, such as medical law and competition law. Thirdly, the tort/ crime classification is also context- and purpose-dependent. An interesting example is French legal academia: owing to pressures from other disciplines, criminal law is classified as part of private law for posts in French universities, with ‘public law’ being reserved for constitutional and administrative law, and a few places each year for legal history. The next level, legal reasoning, concerns how legal actors think about the law they are dealing with. This includes the prevalence of particular forms of

1 See further, M Dyson, ‘Tortious Apples and Criminal Oranges’ in M Dyson (ed), Comparing Tort and Crime (Cambridge, Cambridge University Press, 2015). 2 See eg M Dyson, ‘Overlap, Separation and Hybridity Across Crime and Tort’ in M Dyson and B Vogel (eds), The Limits of Criminal Law (Cambridge, Intersentia, 2018).

Crime, Breach of Legislative Duties and Fault  101 r­easoning, or processes, like the greater role for statutory interpretation in criminal law than in tort law. The relative weight of certain principles may vary between tort and crime. It might be wondered, for instance, whether tort lawyers value fairness more, while criminal lawyers value legal certainty more. One thing is certain, criminal lawyers tend to require greater specificity in their legal rules than tort lawyers do in theirs. Many other interesting questions arise here, such as whether England’s lack of a prescription period for most criminal offences offends legal certainty, while at the same time English law will not countenance a possible alternative, trials in absentia, while many other systems on the continent will. However, in practice the force of this ‘legal certainty’, howsoever phrased, is often diluted. For instance, in France, while criminal offences must, in theory, be precisely defined, some of them do resemble the general provisions found in articles 1240 and 1241 of the Code civil (former articles 1382 and 1383). Legal reasoning will also include the self-perception of the system and its actors, something addressed in section V of this chapter. Turning to substantive law, the third level, there are obvious questions about whether the same tests for liability or for a defence exist in tort and crime. It might be thought that these comparisons have no need to be made, since a court will only need to answer one and the other is simply not relevant. That is largely taken to be the position in England. However, as discussed below, French law has extensive links between civil and criminal law which mean a given court might indeed be faced with incongruities between the two areas of law. Procedural connections between tort and crime are the most obvious in legal practice, and the most intriguing. Even if a legal system has not carefully considered how tort and crime interact on a normative level, there have to be procedural mechanisms to handle issues that arise in practice. The interplay between procedure and the remedies available to resolve disputes shape the fourth and fifth level of interaction between tort law and criminal law. There are six basic issues concerning the remedial framework provided by crime and tort that a legal system must address.3 First, French law prohibits parallel proceedings, instead suspending the civil claim while the prosecution is ongoing: le criminal tient le civil en l’état.4 English law permits them unless, on application by a party to the proceedings, a judge decides the interests of justice do not support them, in which case the civil case will be suspended. Secondly, is a conviction admissible and/or conclusive in regards to facts it must have been founded on, or even as regards liability, in a later civil court? This is a particularly relevant issue, as discussed below, since the more similar the substantive issues the prosecution had to prove for the conviction are to the substantive 3 See eg M Dyson, ‘The State’s Obligation to Provide a Coherent System of Remedies across Crime and Tort’ in A du Bois Pedain et al (eds), Criminal Law and the Authority of the State (Oxford, Hart Publishing, 2017). 4 Art 4(2) CPC.

102  Matthew Dyson issues in the tort claim, the greater utility the conviction would be to a later civil claim. Thus, procedure is driving the effect on substantive issues. Two obvious examples are the content of the relevant duty and the form of fault required. Responses range from the independence of the two actions, to the conviction being evidence, being persuasive, or even, being binding on a later civil court. It is also possible for different contexts to be relevant; for instance, in England, the conviction is binding only in later defamation actions, while being admissible only for all other actions.5 In France, the criminal judgment has the effect of a res ­judicata: the later civil court cannot contradict a decision by a criminal judge on the key aspects of the criminal court’s decision. The leading, though not the first, case was in 1855, and concerned whether the defendant had taken appropriate care to avoid an accident involving his horse-drawn cart.6 The earlier criminal court had found that the defendant had used all proper care, and so a later civil court could not find differently. This means that a civil judge cannot contradict a decision made by the criminal judge on the elements noted in the criminal decision. Those key elements are essentially the facts and the legal consequences of them: that the defendant did, or definitely did not, do the acts complained of, any accompanying fault, and that any causal link between the defendant’s act and the harm to the victim, of intentional or negligent misconduct.7 By contrast with England, which, outside defamation actions, only accepts the earlier conviction as evidence of the facts upon which it must have been founded, French courts see the criminal determination as being on the same legal questions, as well as the same factual ones. This is a practical expression of part of the unity of civil and criminal fault in France. Thirdly, a legal system has to decide whether to permit a person to derive an advantage from whichever of civil or criminal law offers procedural advantages. For example, criminal law typically has greater procedural protections for the defendant than civil law, which might permit more extensive discovery of evidence or inferences to be drawn from evidence that criminal law cannot.8 Fourthly, French law has one of the oldest and most complex regimes for victims of crime to pursue compensation in a modern criminal justice system. From the eighteenth century, the partie civile, a civil party, has been able to claim within the criminal proceedings.9 Article 1 CPP sets out that while judges, prosecutors, and civil servants are entrusted with prosecuting crimes, the injured party can also do so: in particular, the victim can bring about a prosecution. The overriding purpose is efficiency and ensuring compensation, but a further role is to empower the victim to oversee the state’s prosecutorial practice, tempered by consequences for abuse of that power. Those who have personally suffered damage (art 2 CPP), whether material, bodily or moral (art 3 CPP), can bring a claim as partie civile.

5 Civil 6 Cass 7 ibid. 8 See 9 See

Evidence Act 1968, s 13. Cf 11. civ 7 March 1855, D 1855, I, 81 (arrêt Quertier).

eg S Whittaker, Liability for Products (Oxford, Oxford University Press, 2005) chs 14–15. eg Code des délits et des peines du 3 Brumaire An IV (25 October 1795) art 6.

Crime, Breach of Legislative Duties and Fault  103 So too, in certain situations, can associations acting for the public good, such as those combating racism and discrimination.10 (The civil claim can also be brought separately from the criminal prosecution, before a civil court, but it must be after the prosecution has run its course.) The amount of compensation is set by the court under articles 371 and 464 CPP, having regard to equity and the financial situation of the defendant. The rules applied are theoretically civil law norms, irrespective of the fact that they are applied by a criminal judge. This is particularly evident in those cases where a criminal court can award damages even though it does not convict. Since 1992 it has been possible to award damages where the facts show a civil wrong even though no crime was committed, under the application of articles 372 and 470-1 CP.11 Since 2000, this has had a particular meaning for nonintentional harm, as there the civil party or his insurer can request compensation, but where ‘third parties bearing liability should be joined in the proceeding’ the criminal court must refer the matter to a competent civil court. That civil court will then examine the matter under a simplified procedure.12 It is also vital to understand that there are strong limits to the partie civile. Most significantly, article 392-1 CPP gives the Correctional court the ability to fine up to €15,000 for an abusive or dilatory prosecution forced through the partie civile; for that reason, the partie civile who claims without using legal aid must lodge security in order to begin the claim. Indeed, to bring the claim in the first place he must lodge a deposit to cover any possible fine following an acquittal (unless the civil party is entitled to receive legal aid). In addition, a criminal court which acquits a defendant may also hear a claim for damages for an abusive use of the action civile under article 472 CPP.13 By contrast, English law gives no power to a victim to force a prosecution, only to request a review by the Crown Prosecution Service if the victim believes a prosecution should have been brought. A private prosecution might be brought, but it has a high chance of being taken over by the Crown Prosecution Service and either discontinued, or in any case, run as they see fit. Every criminal court must consider whether to award a compensation order to the victim, but it is a separate, hybrid order, not the same as a tort law judgment against a defendant. Most significantly, the value of a compensation order is limited by the means of the defendant, whereas a tort law claim is not. The compensation order’s limit is to prevent the state’s coercive power being unfairly used for compensatory ends, which are beyond the justifications for the use of such power. Correspondingly, there are no significant risks for the victim in seeking a compensation order and, perhaps

10 Arts 2-1 to 2-23 CPP. 11 See eg V Dervieux, ‘The French System’ in M Delmas-Marty and JR Spencer (eds), European ­Criminal Procedures (Cambridge, Cambridge University Press, 2002) 227. 12 Art 470-1 CPP. 13 A powerful example, based on an unfavourable review of a book on European Union law, is JR Spencer, ‘Libel Tourist Ordered to Pay 8,000 Euros’ (2011) 70 CLJ 317.

104  Matthew Dyson surprisingly, at present there are even discretionary powers to pay the costs of even an unsuccessful private prosecutor.14 Fifthly, whether a particular remedy is available in both tort and crime will be of significance to the parties, especially the victim in a criminal case. For example, the partie civile in France has access to compensation without the limit of the means of the defendant, a limit that applies in England; that said, the resulting judgment in France is a civil order, so is not enforced more effectively by the criminal justice system, on behalf of the victim, as is the case in England for a compensation order. Another significant difference is the range of remedies: the partie civile only has access to claims for damages and claims with respect to the other civil consequences of the offence, actions à fins civiles, cannot be brought within the criminal court. This is founded on the unity of the act: it is both tortious and criminal, and compensating for it is therefore neither an unacceptable extension of the criminal court’s jurisdiction, not practically burdensome. This unity does not apply beyond the compensatory consequences of the act, such as to the criminal court restoring property. By contrast, in England, there is no civil right to the restoration of specific property, and damages are instead the remedy typically granted; but by contrast the criminal courts do order the restoration of property on a daily basis.15 Sixthly, and finally, assuming any possible defendant has limited means, the effect of criminal and civil remedies will be to reduce the same ‘pot’, and which of these two has priority might be important. It has been obvious for centuries that a potential civil defendant who is in prison is not normally in a position to earn money to satisfy a civil claim. The problem is even more acute where, as in England, the criminal court is considering imposing fines, imprisonment, confiscation of property related to criminal conduct and compensation. Rules of priority are vital, and will represent choices between the relative importance of the victim and the state. Another example is that the French criminal justice system formally permits the reduction, deferral, or abstention from punishment if the defendant compensates the victim before an order to do so.16 Many other states almost certainly have informal mechanisms to do so, whether within general discretions exercised by prosecutors or judges, hybrid remedies, or even in the p ­ ost-sentencing phase through assessing parole or early-release conditions. Finally, such a payment may result in no claim being made where the victim must claim or even where the civil part of the criminal proceedings is settled.

14 See, generally, M Dyson and P Jarvis, ‘Remedies of the Criminal Courts’ in G Virgo and S Worthington (eds), Commercial Remedies: Resolving Controversies (Cambridge, Cambridge University Press, 2017). 15 See eg M Dyson and S Green, ‘The Properties of the Law: Restoring Personal Property through Crime and Tort’ in M Dyson (ed), Unravelling Tort and Crime (Cambridge, Cambridge University Press, 2014). 16 V Malabat and V Wester-Ouisse, ‘The Quest for Balance between Tort and Crime in French Law’ in M Dyson (ed), Comparing Tort and Crime (Cambridge, Cambridge University Press, 2015) ch 3, 118ff, in particular noting mediation (art 41-1 CPP), composition (art 41-2 CPP), sanction-réparation (art 131-8-1 CP) and exemption (art 132-59 CP).

Crime, Breach of Legislative Duties and Fault  105 From understanding where tort and crime can fit together, we might now turn to ask what work legislative duties and fault are doing within the legal system.

III.  The Work that Legislative Duty and Fault Do Within the Legal System A.  Six Basic Issues Within Tort Law From a comparative’s eye view, we might identify six basic questions within a tort claim: attribution, capacity, wrongfulness and illegality, fault, causation and damage/harm/loss. This is a reconstruction in some cases: many legal systems, including the French and the English, do not necessarily separate out these six issues into separate and discrete questions or components of tort liability. The ­issues of duty and fault are most obviously relevant to the third and fourth of these: wrongfulness and illegality, and fault. It is certainly not the case that every tort claim requires each of these six issues be addressed expressly within the conditions for liability and any associated defences. Obvious examples are torts where fault is not required, forms of liability where nothing unlawful was done by the defendant (such as liability for others in the absence of fault).17 There are also greyer areas, where it is unclear that the issues are being directly address or omitted, such as loosening tests for attributability or capacity, redefining what constitutes harm/ damage/loss or how causation is established. However, the issues seem to underlie that whatever tort law does, and the absence of one or more in any instance is regarded as needing explanation.

(i) Attribution Typically, where D commits a prohibited act or omission, its result is attributed to D. The same is true where, though D does not act, D was under a duty to act and failed to do so: the omission can be attributed to D. Typically, D’s own acts or, where there is an obligation to act, D’s omission, the prohibited result can be attributed to D. However, tort law also recognises other forms of attribution, particularly joint liability and the liability for others, particularly vicarious liability. This aspect of liability is typically not controversial in practice.

(ii) Capacity In England, all persons are generally able to sue and be sued. Historically there were a number of exceptions to this for classes of people but for the most part

17 See

ch 8 below, p 147.

106  Matthew Dyson these have been removed.18 The same is true in France, which saw a similar trend in the mid-twentieth century to reduce the relevance of capacity. However, since in France, actions against the state were already dealt with under administrative law, there were fewer specific exclusions remaining already so the focus was more on removing the remaining ground of non-liability: insanity.19 In addition, the focus was more expressly on ensuring compensation, rather than other grounds for reconsidering such defendants. This is clearly not the case in criminal law, where a defendant might be outside the scope of the criminal law by being too young, or insane. There might also be difficulties of remedy, if not liability, if the defendant is a legal person rather than a natural person. English law has a number of offences and modes of committing offences which can be committed by companies, though French criminal law does not. In tort, this element of liability is typically not controversial in practice either.

(iii) Unlawfulness/Wrongfulness What is it that makes civil wrongs, wrongs? The most common understanding of ‘wrongful’ is that the particular conduct is in a breach of a duty owed to the ­claimant.20 Unlawfulness is a wider category, indicating some reason why the conduct is prohibited. Wrongfulness expresses the grounds for why liability should exist, whereas unlawfulness merely states the fact of the liability being imposed. In fact, tort laws do not appear to have a single element, present in every claim, which captures the idea of unlawfulness or wrongfulness. Typically, in modern law we ask what duty the defendant breached, or what right or interest the ­defendant infringed, but this question might or might not be explicit. In fact, even early in English law, when the forms of action of claim dominated legal thinking, the action on the case (which became the law of negligence) was originally based upon the doing of an unlawful act, ie, ‘by the common custom of the realm’ the conduct of the defendant was regarded by the law as wrongful.21 So too did Roman law differentiate between harm and wrongful harm, as noted by Lord Rodger in the House of Lords in 2005: ‘Harm which constitutes a “wrong” in the contemplation of the law must, of course, be remedied. But the world is full of harm for which the law furnishes no remedy.’22 Many legal systems make the 18 Eg Law Reform (Married Women and Tortfeasors) Act 1935, the Crown Proceedings Act 1947, the State Immunity Act 1978. 19 Which legislation did in 1968: art 414-3 Cc. See G Viney, P Jourdain and S Carval, Traité de droit civil, Les conditions de la responsabilité civile, 4th edn (Paris, LGDJ, 2013) [578]–[593-1]. 20 Typically, this is a relational duty not to infringe another’s relevant right, such as the right to personal integrity: WN Hohfeld, ‘Some Fundamental Legal Conceptions as Applied in Judicial Reasoning’ (1913) 23 Yale Law Journal 16, 32. See generally eg H Koziol (ed), Unification of Tort Law: Wrongfulness, Principles of European Tort Law (The Hague/London/Boston, Kluwer Law I­nternational, 1998) vol 3. 21 WS Holdsworth, A History of English Law, 1st edn (London, Methuen & Co, 1925) vol VIII, 449–50. 22 D v East Berkshire Community Health NHS Trust [2005] 2 AC 373 [100].

Crime, Breach of Legislative Duties and Fault  107 unlawful/wrongful requirement for liability express in their civil codes. A famous example is the German enunciation of protected interests in § 823(1) BGB, complemented by a prohibition against intentionally causing harm to another in § 826. Another is the Italian civil code, which tracks a German influence on the earlier French provision.23 That said, the reality in England, and the practical effect of wide categories of protected interests or rights, is that unlawfulness/ wrongfulness is only rarely a live issue in practice. By contrast, criminal law sometimes uses the act itself, for example, deliberate killing, and sometimes uses the fault element, to make an otherwise lawful act, unlawful. For example, in theft, where it is dishonestly intending the other person to be deprived of the property permanently which distinguishes theft from so many everyday touchings of property. The relevant wrong might also be obvious in criminal law, though it is also sometimes elusive, such as in conspiracy (an agreement that a crime take place in the future) and assault (causing another to apprehend immediate physical force against him or her).

(iv) Fault One of the most enduring justifications for civil liability is that the defendant was at fault. Modern legal systems have moved away from always requiring fault, though many still retain elements of fault, such as through defences proving the lack of fault, even in otherwise strict liability claims. What is particularly important is that fault has shifted its focus: whereas in the past intention played a significant role in some tort claims, it is largely irrelevant in modern tort law. Fault has become almost synonymous with negligence, with the failure to take the care warranted by the situation. This is slightly less clear in France, where la faute has a much more flexible meaning depending on the context. Where liability is established without proof of fault, the justification must come from the inherent unlawfulness or wrongfulness of the conduct, not from the fault with which it was committed.

(v) Causation Causation between the defendant’s wrongful conduct and the harm complained of is a necessary element of all civil claims for appreciable harm. There are some tort claims which do not require harm, or, require a form of harm which is not material or physical, and the nature of the defendant’s causative role is more loosely understood. By contrast, criminal law is more open to conduct offences which do not require causation, like attempt or conspiracy. 23 Art 2043: ‘Risarcimento per fatto illecito. Qualunque fatto doloso o colposo, che cagiona ad altri un danno ingiusto, obbliga colui che ha commesso il fatto a risarcire il danno.’ This might be rendered as ‘Any intentional or negligent conduct, which causes unlawful harm to another, obliges the person by whose conduct it was caused to make good the damage.’ It would also be possible to refer more expressly to ‘…which interferes with a protected interest…’, if a less direct translation were used, but in either case, reference to unlawfulness is clear.

108  Matthew Dyson

(vi) Harm/Damage/Loss The legal system tends to recognise certain harms as actionable, and not others. Merely being unhappy with an outcome does not normally give rise to an a­ ctionable claim. Legal systems are typically more restrictive about economic losses than they are about personal injury or property damage.

B.  French Law’s Basic Questions Formally, French legal doctrine would normally see something like three categories, although six basic issues can nonetheless be seen at work within these categories: (i) The first category is represented by fait générateur, an unlawful action that might result from the defendant’s faute, a thing under his custody, or a person for whom he is responsible, and which is imputed to the defendant personally. (ii) The second category consists of dommage and, typically, préjudice, where the former relates to the harm inflicted on the claimant while the latter to the loss consequential on that harm. (iii) The third category is lien de causalité, a causal link between the event and the harm. The obvious point is that at the first stage, the fait générateur covers a lot of different elements. It includes unlawfulness/wrongfulness as well as fault, though that is not immediately apparent from the term itself. It also ties into the dommage and préjudice. The element of unlawfulness/wrongfulness can perhaps best be understood as providing the answer to a question French tort law often does not choose to ask: what is it about this conduct, other than the harm that resulted, which requires that the law respond. It is therefore not surprising that when it is asked, the question most commonly comes out as a requirement for faute. Faute for this purpose includes an objective aspect of unlawfulness as well as subjective aspects, to the extent they are still required, attribution (expressed as imputation) and capacity.24 While the project defines préjudice as the injury to a lawful interest, which interests are lawful and therefore protected is unclear: in theory, not just the human body, corporeal and incorporeal property should be protected, but any interest whatsoever,25 provided that it is not outright illegal.26 In practice, however, French law, like English law, does not normally try to resolve these interests into any

24 Viney, Jourdain and Carval (n 19) [442]–[592-2]. 25 ibid 159–63. See generally, Viney, Jourdain and Carval (n 19) [450]–[482-1]. 26 Recent case law does not offer many examples of such illegal interests; see however Cass civ 2, 24 January 2002, no 99-16576, regarding revenues derived from undeclared work as not being worthy of compensation.

Crime, Breach of Legislative Duties and Fault  109 finer detail, and there can certainly be problems over which interests are actually sufficient to deserve protection. For example, there was even a period, from 1950 to 1970, where the criminal chamber of the Cour de cassation refused to admit civil actions brought by private persons as partie civile, when the rule violated was regarded as protecting society as a whole, not the individual specifically.27 Even now, it is not entirely clear how much French law would seek to vindicate underlying rights without there being an overt compensatory rationale. It might be that such a rationale, a general provision like article 9 Cc on the right to respect for a private life, in a moral harm or similar, can normally be found. By contrast, English law has recently held that a civil claim could proceed in order to establish the truth of a wrongful act by the defendant and his agents against the deceased, even though the defendant had admitted negligence and settled that part of the action for a sum which included all possible damages.28 Finally, French law faces the difficulty that if wrongfulness is conceived of as part of fault, then strict liability torts, torts without fault, seem to lack wrongfulness. Combined with French law’s comparatively wide categories of strict liability for harm done by things of which the defendant is the gardien, liability for others and other special legislative liabilities such as are imposed for motorvehicle accidents by the Loi Badinter,29 the operational space for fault, and thus, for wrongfulness might in fact be narrow. However, the limits on such strict liability might in fact demonstrate that there remains some role for wrongfulness. For example, one might see the effect of a wrongfulness requirement or the importance of relative fault where even strict liability claims are reduced under comparative negligence, or liability existing only for the passive and not abnormal role of a thing.30 We now turn to fault as the nexus of liability that is expressly acknowledged by French law.

IV.  The Unity of Fault While it may seem that faute has had a timeless meaning, it has not. Indeed, it should not be forgotten that even if the words of the tort provisions in the Code civil have remained the same since 1804, their meaning has not.

27 G Viney, Traité de droit civil, Introduction à la responsabilité, 3rd edn (Paris, LGDJ, 2008) [93]– [93-1]. It should be noted, however, that this restrictive position was formally based on art 2 CPC governing actions civiles, and not on the Code civil provisions on liability. With thanks to Marie Dugué for suggesting this point. 28 Ashley v Chief Constable of Sussex Police [2008] 1 AC 962. 29 Loi no 85-677 du 5 juillet 1985 tendant à l’amélioration de la situation des victimes d’accidents de la circulation et à l’accélération des procédures d’indemnisation. 30 See eg O Moréteau, ‘Basic Questions of Tort Law from a French Perspective’ in H Koziol (ed), Basic Questions of Tort Law from a Comparative Perspective (Wien, Jan Sramek, 2015) [1/56]–[1/163].

110  Matthew Dyson As Halpérin has shown,31 the code provisions were originally perceived to be part of the Roman tradition of Domat (1689)32 and Pothier (1768),33 in that article 1382 Cc concerning fault was a delictual provision, and article 1383 Cc concerning negligence was thought to represent the somewhat amorphous collection of quasidelict (which was not, in Roman times, united by negligence as a fault standard at all). Halpérin cites a lawyer, Garat, writing in 1874–75, as saying that the law cannot be neutral between the victim of the damage, who suffers, and the tortfeasor, who errs. Halpérin argues that this was a sentence used by the codifiers to explain article 1382 Cc. It was probably the same for the judges and lawyers of the first half of the nineteenth century, using their training in Roman law to ground the new code’s provisions. It was only in the second half of the nineteenth century that the exegetic movement took hold, building on cases after the code, and the intellectual materials provided by Sourdat and Dalloz.34 The most relevant consequences of that movement for present purposes were the generalisation of delictual liability, away from specific delicts or even of categories of obligation like delict and quasi-delict, and the substantive expansions of the terms fait, to include any wrongful act or omission, and faute, to almost be a negative test: where the defendant had not been authorised to do the harmful act, he was at fault.35 It would be a fault concept without the Roman, or glossed, concepts of culpa lata and culpa levissima: all faults, however small, led to full compensation. This wide and unified (here meaning ‘non-stratified’) understanding of fault was put under pressure in the nineteenth century in two key ways.36 First, changing risks in society, particularly through the industrial revolution, led to the growth of strict liability and the reduction in fault’s ambit, even fault as widely interpreted as it had begun to be; this went hand in hand with the growth of insurance, and preceded the more recent move to socialise certain classes of risk, bearing ultimate fruit in the wide range of state compensation schemes existing in France.37 Secondly, the procedural connections between criminal law and tort law began to be felt more keenly,38 and there was thus a need to understand how civil fault

31 See particularly, JL Halpérin, ‘French Doctrinal Writing’ in N Jansen (ed), The Development and Making of Legal Doctrine (Cambridge, Cambridge University Press, 2010) 73–85. 32 J Domat, Les lois civiles dans leur ordre naturel (Paris, Coignard, 1689); See eg O Descamps, Les origines de la responsabilité pour faute personnelle dans le Code civil de 1804 (Paris, LGDJ, 2005) 428–29. 33 RJ Pothier, Traité des obligations selon les règles tant du for de la conscience, que du for extérieur (Paris, 1768). 34 A Sourdat, Traité géneral de la responsabilité (Paris, Cosse, 1852); MD Dalloz, Répertoire de législation de doctrine et de jurisprudence (Paris, Bureau de la Jurisprudence Générale, 1858) esp vol 39, 293–456. 35 See eg Toullier, Le droit civil français (Paris, Warée oncle et Warée fils aîné, 1824) vol XI, 156–58. 36 For reasons of space, the relationship with contract law will not be discussed here. 37 Viney (n 27) [18]–[32-1], and particularly Viney’s doctoral thesis on the effect of insurance: G Viney, Le déclin de la responsabilité individuelle (Paris, LGDJ, 1965). 38 It is not entirely clear why the civil/criminal links exerted more pressure from the end of the nineteenth century.

Crime, Breach of Legislative Duties and Fault  111 related to criminal fault.39 The Code of Criminal Procedure of 1808, in article 3, connected proceedings through the partie civile. It was quickly obvious that criminal law and tort law would have further requirements than fault, with criminal law countenancing crimes without the victim that tort law would require, while criminal law might well find merely negligent infliction of some losses to be insufficient while tort law would not.40 However, the presence of civil claims in criminal courts made it particularly obvious that some of the components of liability appeared to have the same names, were doing the same work, and that divergence in their meaning would jeopardise the effectiveness of the procedural connections. This potential contrast became even starker as the nineteenth century ended and the twentieth century began, a time of increased industrial and road traffic accidents. The response was that the Cour de cassation in 1912 declared the unity of civil and criminal fault, allied with the precedence for a criminal prosecution over parallel civil claims.41 The case concerned an accident in a lift, where the wrongdoer operated the lift without securing the safety rails and despite being unauthorised to do so. The Cour de cassation overturned the Cour d’appel of Lyon, which had found there was no criminal liability for negligent homicide or negligent injury but there was civil liability.42 The court went further, holding that the prescription period for the civil claim would, like the criminal action it was running parallel to, be three years (the events having happened 10 years earlier).43 The case appears neatly to encapsulate how offences of negligence in particular were being found by first instance and appeal courts to not be made out, while in the background the civil claim would in fact also be blocked because of the binding effect of the criminal decision on the civil, ensured by the doctrine of la chose jugée (which is the broad equivalent of the English res judicata). The sense that criminal liability based on negligence could not be as wide as the tortious liability, and that there was therefore more than one meaning of ‘negligence’, and thus ‘fault’, was removed, and the wide civil standard imposed across the civil and criminal law. The slightest

39 For the interactions generally, see Viney (n 27) [67]–[155-1]. 40 Sourdat (n 34) 5; LVLJ Larombière, Théorie et pratique des obligations (Paris, G Pedone-Lauriel, 1885) vol V, 684–85. 41 Cass crim 18 December 1912, D 1915, I, 17. See also P Bonfils, L’action civile, essai sur la nature juridique d’une institution (Aix-en-Provence, Presses universitaires d’Aix-Marseille, 2000). The evolution of fault in Spanish doctrine also began at a similar time: for detail see M Casals and A Ruda, ‘The Development of Legal Doctrine on Fault in Spanish Tort Law’ in Nils Jansen (ed), The Development and Making of Legal Doctrine (Cambridge, Cambridge University Press, 2010) 195–96. 42 Cass crim 18 December 1912, D 1915, I, 17: ‘articles 319 and 320 of the Criminal Code punish by means of criminal penalties anyone who, through their mistake, imprudence, inadvertence, negligence, or violation of regulations, has unintentionally committed a homicide or caused personal injuries, with the light character of the fault committed having the effect only of mitigating the measure of the penalty’ (‘articles 319 et 320 du Code pénal punissent de peines correctionnelles quiconque, par maladresse, imprudence, inattention, négligence ou inobservation des règlements, a commis involontairement un homicide ou causé des blessures, sans que la légèreté de la faute commise puisse avoir d’autre effet que celui d’atténuer la peine encourue’). 43 It also involved issues slightly outside our present concerns because the case involved claimants and defendants who were actually the employers of the victim and wrongdoer.

112  Matthew Dyson fault would suffice to constitute negligence, and it would be up to whatever other components of the relevant criminal offence, or its prosecution, to separate out those who deserved punishment from those who did not. This formal unity of fault survived for almost a century, until 2000. The law of 10 July 2000 changed the doctrine of chose jugée on a procedural level within article 4-1 CPP, but that tied in with changes in article 121-3 CP in respect of non-intentional crimes. The unity of civil and criminal fault had, it was thought, led to complications for certain types of people, predominantly mayors who might indirectly be the cause of harm, this being especially important, given French law’s wide rules of causation. There was a view that the politically important and numerous town mayors were being prosecuted, at times driven by parties civiles, for accidents and disasters linked in some way to the mayors’ failure to implement health and safety rules.44 So, since 2000, criminal liability for indirectly and non-intentionally caused harms requires one of two higher degrees of fault: délibérée, meaning a clearly deliberate breach of a protective rule, or caractérisée, misconduct creating a serious risk to another which the defendant must have known about. Thus, civil and criminal fault are now not unified in the unintentional harms at least: all criminal faults will be civil faults, but not all civil faults are c­ riminal.45 This has happened only where the causative link is weaker, ­particularly to remove those on the periphery of criminal liability, who, it must be noted, would ­probably not be in the frame for criminal liability, and perhaps not even civil (save as employers) liability in many other legal systems. It further reinforces the links between the key issues in tort law, but most importantly, the particularly powerful links between the three elements of French tort law. In this case, the wide rules of causation put pressure on the unity of civil and criminal fault, and that unity ultimately gave way. The alignment and unity of fault have so far focused on negligence and its criminal analogue, imprudence. The role of intention has declined in tort, and even in criminal law,46 save perhaps for certain paradigmatically serious crimes like murder.47 The Projet de réforme makes no express reference to this unity of fault, just as the doctrinal discussions about fault, delict, quasi-delict and other notions have not shown on the face of the code. The assumption must be that the unity of civil and criminal fault will continue in its present form, and in particular (as discussed below) that article 1242 of the Projet de réforme will enshrine the rule that the violation of a criminal duty under statute constitutes a civil fault. The pressure for changes to the concept of fault have, in both cases, come from criminal law

44 See eg JR Spencer and M-A Brajeux, ‘Criminal Liability for Negligence – a Lesson from Across the Channel?’ (2010) 59 ICLQ 1, 9–14. 45 See eg B de Lamy, ‘Responsabilité civile, responsabilité pénale: impossible divorce, mais liaison dangereuse’ RDA 2013, 52, 53–56. 46 See Malabat and Wester-Ouisse (n 16) 95. 47 Art 221-1 CP.

Crime, Breach of Legislative Duties and Fault  113 being a vehicle for a civil claim: that vehicle was more effective at compensating where criminal law was as wide as civil law. Interestingly it does not appear to have been the case that there was intellectual difficulty or the risk of delay and error in having more than one meaning for similar fault concepts. Rather, the impetus began as compensatory effectiveness, and concluded through a push-back against over-extensive liability. That push-back is a particularly important example of the feedback mechanisms in a legal system kicking in and altering the path of legal development. It is all the more important to see such a limit and the response, since French law already appears to go a lot further towards compensation than other systems. It is also fundamental to appreciate that this unity of fault is premised on the clarity of what fault must have meant in those circumstances: that whatever the civil and criminal definitions of fault were, they were ascertainable and indeed, ascertained. In both civil and criminal law fault terms like négligence, imprudence and indeed faute left a great deal of scope to the lower court judges. It is here that the content of the fault, and its grounding in some form of unlawfulness component, come together, which will be the focus of the next section of the chapter.

V.  Integrating Duties Across Tort and Crime A.  Unlawful Acts A simple way to integrate norms from outside the purely civil sphere is to use a different building block than the norm itself, such as integrating ‘unlawful acts’. Dutch law’s key definition of an unlawful act is as, amongst other things, ‘an act or omission in violation of a duty imposed by written law’ and all criminal laws must be in written form (in addition, breaching social standards of due care also generates liability).48 While neither the French code’s current or proposed tort provisions refer to the ‘unlawful act’ phrasing, it can still be discerned playing a role in French law. For one thing, doctrinal writers from the nineteenth century have expressly reasoned along these lines, apparently as a free translation of the Roman ‘damnum injuria datum’.49 As Halpérin explains: There was a real consensus in the French doctrine of the nineteenth century about the idea that fault should be an act “contra legem” to merit compensation. The problem – not truly envisaged by doctrinal writing – is that many civil faults are not punished by a statute law: the sphere of civil liability is larger than that of penal law. Doctrinal writing was thus playing with words: ‘unlawful act’ is not the same as “unlawful offence”.



48 Art

6:162, Burgerlijk Wetboek (Dutch civil code). (n 31) 80–81.

49 Halpérin

114  Matthew Dyson The unlawful act included, for instance, infringements of customs or practice rules; this was the case with doctors and surgeons, whose liability was considered as delictual and not contractual. Subjective elements, rooted in moral concepts, were thus inserted in the apparently objective notion of the unlawful act.50

It will be immediately apparent that what makes the act unlawful might be tightly bound up with notions of fault, and with notions of individual rights (that liberty ceases when one harms others). Yet, according to Halpérin, even together they cannot entirely explain the case law and its development. The unlawful act concept provided a hook within tort law from which to hang unlawfulness and fault, whether imported from another area of law or not necessarily clearly of one area or another.51

B.  General Integrative Techniques The Projet de réforme expressly states that it is ‘a fault to violate a legislative requirement’ (article 1242).52 This enshrines the rule long recognised by courts and authors, that any legislative duty, especially one which sets a penalty for its breach, must be obeyed, or otherwise there will be a fault.53 This might be vulgarised into the idea that fault is predominantly about what a reasonable person would do, and a reasonable person obeys statutory obligations. The difficulty is that statutory obligations need not actually require any fault, neither intention nor negligence, and it hardly makes sense to say that a reasonable person will do everything, including care which is unreasonable, in order to achieve a result set in a statutory duty. French law applies no overt limits on what duties are thus imported into French tort law. That is an unusual approach. The issue will turn on what legislative duties are created and which of those apply to individuals, the breach of which causes harm that French tort law recognises and compensates for. It may be that any necessary limiting is left to other components of liability, like causation. A famous example of a narrower general integrative technique is § 823(2) BGB. First, § 823(1) requires anyone who, intentionally or negligently, unlawfully injures the protected interests of German law, life, body, health, freedom, property or another right, to pay compensation to that person. German law further adds, in § 823(2), that the same duty to compensate is owed by someone who breaches a norm intended to protect another, so long as it is done with fault. For clarity, ‘fault’ there is a reference to the general standard of the responsibility of persons subject

50 ibid. 51 The integrative techniques in contract law, whether through public policy, illegality or otherwise, are sadly beyond the scope of this chapter. 52 The parallel questions of integrating other standards, such as the level or expressions of care set by professional bodies, or industry practice, is regrettably outside the scope of this chapter. 53 Viney, Jourdain and Carval (n 19) [448].

Crime, Breach of Legislative Duties and Fault  115 to an obligation in § 276, that is, intention or negligence. A protective norm, or Schütznorm, will be found in many criminal offences which might not otherwise be fitted within § 823(1) easily, but, more importantly, it means that once a criminal offence involving such a norm has been shown, the core of civil liability has already been established as well and the potential claimant has already had most of her work done for her, only requiring fault to be proven if it had not already been established as part of the norm’s breach. French law, by contrast, has no additional requirement to prove fault, since the breach of the relevant legislative requirement is itself fault. By contrast, Spain makes its integration even more obvious: ­article 109(1) of the Spanish Criminal Code provides that: The carrying out of an act described by the law as delito obliges reparation, according to law, for the damage and losses caused by it.54

That is, the law turns on damage being caused by what happens to be a criminal offence. English law still tends to treat each case on its facts. Criminal offences and torts each have their own different substantive components, and it is only recently that there has been discussion about how they relate to each other. The default belief in the last 150 years has been that it would cause confusion for one branch of the law to employ the substantive law of the other.55 Therefore, a first key point is that there is no general rule on fault; that said, criminal law tends to focus on intention and recklessness in its traditional core, though it does use negligence in some offences. In addition, there is a specific nominate tort integrating legislative duties from outside tort law: the tort of breach of statutory duty. Secondly, some torts can have the same source as crimes, even though they are then read through a new lens, again as is the case under the separate tort of breach of statutory duty. This is a means of creating a civil claim for harm caused by the breach of a duty in a statute where the statute does not itself say whether there is civil liability but which does impose a duty, the breach of which has caused harm. The claimant has to show that, despite the absence of an express reference to creating civil liability in the statute, that is what Parliament intended; it is a factor against civil liability that the statute does impose a criminal penalty since it can be argued that the criminal penalty appearing alone shows that such penalty is all Parliament wished to supply as an enforcement mechanism.56 54 Delito means any crime punished by law. See also art 100 LECrim. See generally, L Bachmaier Winter, C Gómez-Jara Diez and A Ruda-González, ‘Blurred Borders in Spanish Tort and Crime’ in M Dyson (ed), Comparing Tort and Crime (Cambridge, Cambridge University Press, 2015); I González Pacanowska, ‘The Development of Traffic Liability in Spain’ in Wolfgang Ernst (ed), The Development of Traffic Liability (Cambridge, Cambridge University Press, 2010) 151, 153–58. 55 Eg on automatism: Mansfield v Weetabix [1998] 1 WLR 1263, 1266, 1268–79; on the civil law of ownership and theft: Bentley v Vilmont (1887) 12 App Cas 471, 477; cf R v Hinks [2001] 2 AC 241, 263–70. 56 Carroll v Barclay (Andrew) & Sons Ltd [1948] AC 477, 489–90, 493; Biddle v Truvox Engineering Co [1952] 1 KB 101, 103; Cutler v Wandsworth Stadium [1949] AC 398 and see also Goodhart (1946) 62 LQR 316, 317 (note).

116  Matthew Dyson There are legal systems which employ both forms, unlawful act, and a general integrative technique. For example, the Dutch Civil Code, article 6.162 says: 1. A person who commits a tortious act (unlawful act) against another person that can be attributed to him, must repair the damage that this other person has suffered as a result thereof. 2. As a tortious act is regarded a violation of someone else’s right (entitlement) and an act or omission in violation of a duty imposed by law or of what according to unwritten law has to be regarded as proper social conduct, always as far as there was no justification for this behaviour.57

And, in a clear mirror of the German provision (§ 823(2) BGB discussed above, though with an express requirement of fault), and to some extent, the English ­position, article 6.163 adds: There is no obligation to repair the damage on the ground of a tortious act if the violated standard of behaviour does not intend to offer protection against damage as suffered by the injured person.

It should also not be forgotten that there are other integrative techniques than the two noted above, and these might look beyond just duties. First, there are some torts and crimes which share the same requirements: examples in England include public nuisance58 and harassment;59 perhaps the most notable example in France is defamation, under a special law of 1881.60 Secondly, at the other end of the liability process, a determination by a criminal court that there cannot be liability might bind a civil court. This is connected to the idea that the duty would be the same across tort and crime, but focuses not on a duty expressly, but on some other component of liability, such as whether there is a justification, or no wrong. English law has not adopted that position at the most common point where it might arise, self-defence, though some other legal systems have.61 Other possible areas, particularly where criminal law might recognise a defence, are areas where tort law does not (for example, duress, insanity) and/or where the scope of the defence is so uncertain as not to be easily compared with tort law (such as the proto-defence of necessity in criminal law,62 which is arguably not a free-standing defence outside certain medical situations, at least not yet). Thirdly and finally, instead of addressing the same component directly, we might consider the other side of the coin. That is, the method used in the tort of breach of statutory duty might be applied almost in reverse, through a defence where a tort claim would be incompatible with a rule of criminal law. Put more generally, the defence of illegality, or ex turpi causa non oritur actio, which prevents the exercise 57 Taken from www.dutchcivillaw.com/legislation/dcctitle6633.htm on 3 February 2019. 58 See eg JR Spencer, ‘Public Nuisance – A Critical Examination’ (1989) 48(1) CLJ 55; R v ­Rimmington, R v Goldstein [2005] 3 WLR 982. 59 Protection from Harassment Act 1997, ss 1–3. 60 Art 29 of the law of 29 July 1881. 61 Ashley v Chief Constable of Sussex Police [2008] 1 AC 962; cf art 65 Codigo procesal penal de Brazil. 62 Re A (Conjoined twins) [2001] 2 WLR 480.

Crime, Breach of Legislative Duties and Fault  117 of a right in the context of wrongful conduct,63 is relatively new to English tort law, even though it is well known to other areas of law such as contract. Simply happening to be doing something unlawful will not prevent a claim, but the more connected the claim is to that illegality, the more likely it is that the claim will be rejected.64 It is not clear what French law would do in this situation, though there does not appear to be a dedicated defence, other than some form of public policy bar. Equally, it should not be forgotten that one area of law might rely on the other not just to close off liability, but to substantiate it. There may be times when criminal law relies on substantive rules of the civil law:65 a classic example is the definitions of property and who has a right to possess it.66

VI.  How Unity and Disunity Develop It might be wondered how the unity and disunity of a legal system might develop. One interesting question is whether legal systems go so far as to think of duties as being neutral, independent of the area of law they are best known in, enforced by or created into. That is, a particular prohibition does not create a ‘criminal duty’, but a duty created and enforced by the criminal law. If the same duty is applied within another area of law, is a duty being applied, or a ‘criminal duty’ being applied? Legal actors not surprisingly looking at a duty will almost immediately look for its source, authority, and the rules which are used to interpret it. These factors make complete neutrality unlikely. It seems that French law goes far further than English law, for when tort law looks at a duty, the key characteristic is its legislative character, not that it was a duty originating in, for instance, criminal law. That said, the practical reasons why the examination of the duty arise, in particular the partie civile, mean that the criminal law origins of the duty and its effects seem not to be absent entirely. Neutrality might be unlikely, but portability, or transferability of concepts across criminal to civil law is certainly something French law has achieved and which the reforms would enshrine.

A.  Abstract Compared to Practical Development The French Code civil was a landmark achievement in the rational and systematic treatment of a corpus of law, and the intellectual work underpinning it continues to support even the reform proposals now being considered. Combined particularly 63 CR Symmons, ‘Ex Turpi Causa in English Tort Law’ (1981) 44 MLR 555; J Goudkamp, ‘The Defence of Illegality: Gray v Thames Trains Ltd’ (2009) 17 Torts Law Journal 1. 64 See eg Revill v Newbery [1996] 2 WLR 239, Vellino v Chief Constable of Greater Manchester [2002] 1 WLR 218, Gray v Thames Trains [2009] 1 AC 1339, Patel v Mirza [2016] UKSC 42. 65 Eg JC Smith, ‘Civil Law Concepts in the Criminal Law’ [1972B] CLJ 197. 66 English law has made some particularly startling departures, eg R v Hinks [2001] 2 AC 241.

118  Matthew Dyson with the Code of Criminal Procedure, it linked civil and criminal law in the ways already described, but it did not seal every possible connection. The abstract and theoretical thinking that went into the Codes clearly did not engage with all the practical needs of legal actors governed by the Codes. Most of the key theoretical and conceptual work for the modern law appears to have developed through the courts working through the practical implications of cases being brought. In a sense, the infrastructure for examining the relationship between French civil and criminal law was put in place by the Codes, but the actual work has been done by courts, and by commentators. This is not an unusual position. English law has developed in a similar way, though the structures channelling disputes were never as systematised as French law.

B.  Conceptual Understanding of Fault and Duty How unified must a legal system be to merit being called a system at all? That question cannot be answered here, but it is interesting to note that the term ‘unity’ is used to mean different things, and is rarely used in its most expansive sense. Two models might be taken. First, the German conception of ‘unity’ is more thoroughly worked out than most others. It is nonetheless a particular use of the word. It does not mean substantive unity, or resolutive unity, which might be the most obvious meanings. Instead: According to this principle the entirety of the legal norms forms a consistent system… The principle of unity requires that each legal norm is interpreted in such a way that the consistency of the legal system is preserved.67

This has been interpreted to mean that unity demands consistency, not sameness.68 The second model is exemplified by the French approach, at least in the relationship between civil and criminal (but not civil and administrative) faults. That conception of unity was more literal, in that the objects in question must be the same in their entirety. That all-encompassing unity tended to flow from the purposes those objects were being put to, most notably the compensatory function of tort law and the space made for it to operate within criminal law. At the same time, that unity has a far more restricted scope: each instance of unity has been restricted to one concept at a time, rather than describing the system as a whole. Both conceptions therefore come to a similar practical place, though perhaps they express different aspirations.

67 P Hellwege and P Wittig, ‘Delictual and Criminal Liability in Germany’ in M Dyson (ed), C ­ omparing Tort and Crime (Cambridge, Cambridge University Press, 2015) 123. 68 ibid, text preceding fn 4.

Crime, Breach of Legislative Duties and Fault  119

VII.  Distinctive Characteristics of French Tort Law and What we Learn from them As noted at the beginning of this chapter, French tort law is distinct in at least four ways: (1) a unilateral and extensive integration of legislative duties and fault into tort law; (2) a widespread integration of civil law into criminal procedure accompanied by some procedural safeguards against abusive claims; (3) an apparent freedom from inhibition in prioritising compensating through the tort system those who have suffered harm; and (4) a focus on shorter, simpler and arguably more opaque formal tort rules, even if the current Projet de réforme represents some slight steps towards greater formal enunciation of the relevant rules. The simpler but more opaque rules are particularly evident in the requirement of a fait générateur, where the doctrinal requirement for fault might even be thought of as the composite ‘reason why the defendant should pay the claimant’, without the specific threads of unlawfulness/wrongfulness and fault being evident. These four characteristics reinforce each other. The first two, extensive substantive integration of duties and fault, and widespread integration of procedure, mean that there are even more opportunities for, or perhaps pressure towards, compensating all harms and using the more open-textured provisions of the Code civil to do so. Some of the interesting questions for comparatists are about how this system functions without the limits that other systems seem to think they need. Put another way, where are the unifying forces that keep the doctrine and function of the system together and prevent tort liability from becoming too great a burden on potential defendants and, through insurance, on all insureds? Some examples of checks and balances preventing the links between civil and criminal law becoming too burdensome have already been noted: the feedback on the unity of civil and criminal fault for indirect and negligently caused harm, and the risks that a partie civile runs when s/he forces a prosecution (but not when merely joining one that is already taking place). Some of the doctrinal work that is done by fault and causation in French law is remarkable, seemingly both having little solid structure but managing to drive all of tort liability. That balance of strength without structure, force with flexibility, is perhaps the greater enigma of French tort law. In other words, French law illustrates the different ways to balance stability and legal development within a legal system: an internal structure like those covering crime/tort, legislative duties and fault within a legal system can both reduce and encourage substantive legal solutions; but how it does so is mediated through the culture of the legal actors within the system. It seems likely that there is something in the legal training and culture of French tort lawyers that they can handle the material and develop it over time, something incredibly impressive to those outside the system who might feel that they barely understand it.

120

part iii Liability without Fault

122

7 The Role of Liability without Fault JONAS KNETSCH

Liability without fault, or strict liability (responsabilité sans faute or ­Gefährdungshaftung)1 is one of the features of the law of liability which exists in all legal systems, but with which lawyers from different legal traditions associate strikingly different ideas. If asked about this notion, a lawyer trained in a French university would probably think of the famous decision in the arrêt Jand’heur which formally approved a general principle of liability for the actions of things2 or the Loi Badinter governing traffic accidents.3 His German counterpart would probably start by sifting through the areas in which there exists special legislation, such as regarding trains, electricity, motor-traffic or the environment, but he would probably also mention the ceilings on liability and would insist on the exceptional nature of these regimes in relation to the general law.4 As for an English lawyer, he would perhaps have a similar reaction to a German lawyer and would emphasise the peculiarity of strict liability,5 while having in mind the famous case of Rylands v Fletcher in which the House of Lords delivered a fundamental decision as regards the origins of the law governing private nuisance and environmental liability.6 Three legal systems and as many visions of liability without fault!

1 To avoid repetition, this chapter uses without distinction the expressions ‘liability without fault’ (or responsabilité sans faute) and ‘strict liability’ (or responsabilité objective). 2 Ch réun 13 February 1930, rapp Le Marc’hadour, concl Matter S 1930, 1, 121 note P Esmein, DP 1930, 1, 57 note G Ripert. 3 Loi no 85-677 of 5 July 1985 tendant à l’amélioration de la situation des victimes d’accidents de la circulation et à l’accélération des procédures d’indemnisation. 4 E Deutsch, ‘Methode und Konzept der Gefährdungshaftung’ Zeitschrift für Versicherungsrecht 1971, 1. 5 See eg M Lunney, D Nolan and K Oliphant, Tort Law – Text and Materials, 6th edn (Oxford, Oxford University Press, 2017) 991 (‘exceptional in the modern law’). See also S Deakin, A Johnston and B Markesinis, Markesinis and Deakin’s Tort Law, 7th edn (Oxford, Oxford University Press, 2013) 503, and WE Peel and J Goudkamp, Winfield & Jolowicz on Tort, 19th edn (London, Sweet & Maxwell, 2014) [1-044] (‘limited in their practical importance or of rather haphazard application’). 6 On the importance and impact of the decision in Rylands v Fletcher, see J Murphy, ‘The Merits of Rylands v Fletcher’ (2004) 24 OJLS 643.

124  Jonas Knetsch The picture which emerges from this first overview of the subject is all the more disconcerting when one notes that the few comparative studies which have been dedicated to the topic have identified profound convergences in the national laws studied.7 Indeed, throughout the world there are in one form or another regimes of civil liability which dispense the claimant from proof of any failure of conduct in the defendant. These studies make clear, moreover, that while the regimes of strict liability may have various extents and different names, they can be classed into three broad categories according to the context to which they apply: liabilities for the actions of other people,8 liabilities for the actions of dangerous activities or things9 and liabilities for the action of encroaching on another person’s land.10 If it is possible to find illustrations of these three models of liability in each of the three systems with which this chapter is concerned (German, English, and French), their importance for civil liability more generally without doubt varies from one country to another. In French law, the integration of liability without fault within the system of civil liability is most advanced, this no longer even being seen as an exceptional basis of liability as compared to liability for fault. By contrast, German law and English law share a same vision of Gefährdungshaftung and strict liability as a collection of occasional derogations from the idea according to which one is liable only for conduct where one is at fault, English law showing itself as even more restrictive in this respect than German law.11 It is also at the level of the scope of application of the cases of liability without fault that the comparison between the French and German legal systems12 gives rise to a striking contrast.13 While German law is characterised by regimes of strict liability whose scope is restricted and by the prohibition of their application by analogy to similar harmful circumstances,14 French law recognises not merely a

7 See especially B Koch and H Koziol (eds), Unification of Tort Law: Strict Liability (The Hague/ London/Boston, Kluwer Law International, 2001); F Werro and V Palmer (eds), The Boundaries of Strict Liability in European Tort Law (Durham NC/Bern/Bruxelles, Carolina Academic Press/Stämpfli Verlag AG/Bruylant, 2004). Compare with the more descriptive work of C Oertel, Objektive Haftung in Europa (Tübingen, Mohr Siebeck, 2010). 8 In particular, liability of an employer, of parents for their minor children and for persons who undertake the care of vulnerable adults, notably those with a serious disability. 9 Here can be cited, in chronological order, liability for the actions of animals and of dilapidated buildings, liability for the action of defective products and liability for the action of dangerous installations. 10 This category would include civil liability for nuisance (troubles de voisinage) and, more recently, liability for environmental harm. 11 F Werro, V Palmer and A-C Hahn, ‘Strict Liability in European Tort Law: An Introduction’, in Werro and Palmer (n 7) 3, esp 4. 12 In the present study I will focus on these two legal systems and will consider English law only occasionally. 13 G Wagner, ‘Grundstrukturen des Europäischen Deliktsrechts‘ in R Zimmermann (ed), Grundstrukturen des Europäischen Deliktsrechts (Baden-Baden, Nomos, 2003) 189, esp 275ff. See also Oertel (n 7) 311ff. 14 The German Federal Court of Justice has left no room for doubt on this question. See BGH, 25 January 1971, III ZR 208/68, BGHZ 55, 229. On this point, see also H Kötz and G Wagner, ­Deliktsrecht, 13th edn (Munich, Franz Vahlen, 2016) [509]–[514].

The Role of Liability without Fault  125 general principle of liability for the actions of things, based on a very generous interpretation of the former article 1384(1) of the Code civil, but also the idea that there may be cases of liability for the actions of other people other than those accepted formally by the Code.15 Curiously, French and German legal scholars debate or have debated the development of their national systems in a way which allows a rapprochement between their two contrasting models, as if the two conceptions had had a mutual attraction. Indeed, while in the 1980s certain German writers advocated the introduction of a general legislative provision (a ‘general clause’) on liability for the creation of a risk,16 part of French legal scholarship has showed itself resolutely hostile to the principle of liability for the action of things,17 suggesting instead that it should be replaced by liability for the actions of dangerous installations.18 This diversity in unity is even more clear when one turns to the delicate task of describing the very concept of liability without fault. In all legal systems, jurists appear to have to face the same difficulties in identifying the contours of the concept of strict liability. Often, as in French law, these difficulties are obscured by a binary presentation of the law governing liability, often in a highly developed way. By contrasting the principle of liability for fault with other regimes of liability founded on risk, legal textbooks on the law of obligations give the impression of a sharp delineation between the two categories of actions which give rise to civil liability (faits générateurs de responsabilité civile), whose theoretical foundations are diametrically opposed.19 The success of this analysis of the law in this binary way probably has its advantages in terms of the exposition of the law, the different cases of liability without fault being able to be discussed as alternative grounds of liability to the original ground of liability for fault.

15 This even makes some writers say that the courts have discovered a ‘general liability for the actions of other people’: see in particular P le Tourneau (ed), Droit de la responsabilité et des contrats, 11th edn (Paris, Dalloz, 2017) [2232.00]ff. For a more nuanced analysis see C Rade, ‘Faut-il reconnaître l’existence d’une responsabilité générale du fait d’autrui?’ in Responsabilité civile et assurances. Études offertes à Hubert Groutel (Paris, LexisNexis, 2006) 376. See also below, ch 8, pp 152–153. 16 For a comparative treatment of these proposals (formulated notably by E Deutsch, H Kötz and H Weitnauer), see E Deutsch, Allgemeines Haftungsrecht, 2nd edn (Cologne, Carl Heymanns, 1996) [694]ff. 17 J-S Borghetti, ‘La responsabilité du fait des choses, un régime qui a fait son temps’ RTD civ 2010, 1. Cf P Brun, ‘De l’intemporalité du principe de responsabilité du fait des choses’ RTD civ 2010, 487. See also I Maria, ‘La responsabilité générale du fait des choses à l’épreuve du contentieux judiciaire’ RLDC 2011, no 78, 62. 18 Art 23 Avant-projet Terré (for a presentation of this text, see J-S Borghetti, ‘Des principaux délits spéciaux’ in F Terré, Pour une réforme du droit de la responsabilité (Paris, Dalloz, 2011) 176ff). Cf also art 1362 Avant-projet Catala which suggests a liability for the action of dangerous activities in addition to the general liability for the actions of things. A proposal to this effect had also been put forward for Belgian law by G Schamps, La mise en danger: un concept fondateur d’un principe général de responsabilité. Analyse de droit comparé (Bruxelles/Paris, Bruylant/LGDJ, 1998) 843ff. 19 See eg F Terré, P Simler, Y Lequette and F Chénedé, Droit civil, Les obligations, 12th edn (Paris, Dalloz, 2019) [905]ff. This presentation is even more systematic as regards public liability where the dichotomy between liability for fault and without fault constitutes a fundamental structural element.

126  Jonas Knetsch Such a division of the facts which give rise to liability into two, clearly ­ elineated categories is not, moreover, the prerogative of French teaching m d ­ anuals. It is also very much present in the studies which concern the philosophical foundations of civil liability. In one of the foundational works on the question in German legal scholarship,20 Esser emphasised the duality (Zweispurigkeit) of civil liability by underlining that, unlike liability for fault which is based on considerations of commutative justice, liability without fault rests on distributive justice as ­understood by Aristotelian philosophy.21 At the risk of over-simplification, the thought here can be summarised by explaining the relationship between commutative justice and liability for fault by the fact that the latter takes effect irrespective of the nature of the person in question by sanctioning behaviour which no one may adopt towards other people. Inversely, liability without fault takes into account the nature of the person to be held liable and so would put into effect a form of distributive justive. The purpose of the law here is not to redress a wrong that has been committed against another person, but rather to attribute to someone the burden of bearing the cost of a harm on the ground of their own position (the manufacturer of a defective product, the keeper of a vehicle, etc).22 However, the presentation of the regimes of liability as a juridical edifice resting on two great pillars, liability for fault and liability without fault, is undermined by further analysis to the point of losing its power to convince. At the level of legal philosophy, the work of Jansen has shown in a convincing way that the ideas of distributive and commutative justice hardly lend themselves to an apposite justification of such a distinction and that considerations of morality and of policy which underlie liability without fault are much more nuanced than had earlier been thought.23 In particular, according to Jansen: the categories of distributive and commutative justice are not mutually exclusive [so that] there is no difference in this respect between regimes based on fault and on strict liability.24

20 J Esser, ‘Die Zweispurigkeit des Haftpflichtrechts’ Juristen-Zeitung 1953, 129. See also by the same author, Grundlagen und Entwicklung der Gefährdungshaftung (Munich, CH Beck, 1943, reprinted in 1967 (the stock of the first edition of this work was destroyed during the Second World War before it was distributed, and its importance in legal scholarship became recognised only after its republication in 1967)). 21 On the ambiguities in the reception of the ideas of distributive and commutative justice by legal scholars in the twentieth century, see N Jansen, Die Struktur des Haftungsrechts (Tübingen, Mohr Siebeck, 2003) 78. 22 On all this see, in German, C-W Canaris, ‘Die Gefährdungshaftung im Lichte der neueren ­Rechtsentwicklung’ Juristische Blätter 1995, 2, esp 15ff (the author deduces from this that the principle of fault is characterised by a ‘superiority from the point of legal ethics’ (rechtsethische Überlegenheit) in relation to liability without fault). 23 Jansen (n 21) 89ff. 24 Jansen (n 21) 118: ‘Die Gerechtigkeitskategorien der Verteilung und des Ausgleichs [schließen] einander nicht wechselseitig aus: … Insoweit besteht kein Unterschied zwischen den verschuldensabhängigen Tatbeständen des Deliktsrechts und der strikten Haftung’.

The Role of Liability without Fault  127 Above all it is at the level of technical law that the binary presentation of the regimes of liability fails as a result of its over-simplification. The requirements governing the actions giving rise to liability as well as their proof are characterised by such a considerable variety that one should instead speak of a sliding scale with liability for fault at one extreme and liability without fault at the other. There are a multitude of regimes in between which do not perfectly belong to one or the other of these categories, to an extent that the idea of a ‘continuum’, identified so clearly in English legal scholarship,25 seems very much more appropriate. A few examples provide convincing support for this point of view. What should we think about the liability for presumed fault in the person who undertakes the clinical testing of medicines,26 or the professional or business liabilities for the failure to perform a ‘re-enforced safety obligation to take care’ (une obligation de sécurité de moyens renforcée) or an ‘attenuated obligation of accomplishment’ (une obligation de résultat atténué)?27 How should we classify the civil liability of bankers and notaries, which has all the appearances of a delictual liability for fault, but for which the required standard of behaviour – the famous Sorgfaltsbegriff of German law, or ‘standard of care’ of English law – is so strict that, in practice, the regime could be mistaken for being one of strict liability?28 Finally, what should we say about liability for defective products which is based on a defect of safety: is it genuinely a liability without fault, given that the producer can escape liability, at least in theory, by showing that the defect of safety could not have been discovered at the moment of the putting into circulation of the product, having regard to the state of scientific and technical knowledge at the time?29 25 P Cane, ‘Fault and Strict Liability for Harm in Tort Law’ in W Swadling and G Jones (eds), The Search of Principle: Essays in Honour of Lord Goff of Chieveley (Oxford, Oxford University Press, 1999) 171, esp 172; V Palmer, ‘A General Theory of the Inner Structure of Strict Liability: Common Law, Civil Law, and Comparative Law’ (1988) 62 Tul L Rev 1303, esp 1304ff ; idem, ‘Trois principes de la responsabilité sans faute’ RIDC 1987, 825, esp 828. See also I Englard, The Philosophy of Tort Law (Aldershot, Dartmouth Publishing Co, 1993) 21ff. 26 Art L. 1121-10 Code de la santé publique. On this regime, see C Grellier, ‘Recherche médicale et indemnisation’ Médécine & droit 2011, no 106, 41. 27 For an outline of the different circumstances in which these apply, see in particular G Viney, P Jourdain and S Carval, Traité de droit civil, Les conditions de la responsabilité, 4th edn (Paris, LGDJ, 2013) [499]ff. 28 On the liability of bankers and notaries for neglect of their duties of information and warning, see le Tourneau (n 15) [3324.61]ff and [3331.64]. In the case of medical liability for fault based on art L 1142-1 Code de la santé publique, one could come to the same conclusion. See M Bacache-Gibeili, Les obligations. La responsabilité civile extracontractuelle, 3rd edn (Paris, Economica, 2016) [860]ff (the author emphasises also the existence of ‘virtual faults’, deduced directly from the existence of a harm, which bear only the name of a fault). 29 The significance of the development risks defence on the legal nature of the liability regime ­stemming from the Directive of 25 July 1998 has hardly been discussed in the French literature. However, see P Oudot, ‘Le piège communautaire de la responsabilité du fait des produits d ­ éfectueux’ Dr & patrim 2003, no 111, 40, esp 45ff as well as J-S Borghetti, La responsabilité du fait des produits. Étude de droit comparé (Paris, LGDJ, 2004) [671]. German legal scholarship is more abundant in this respect. See HC Taschner, ‘Produkthaftung – Noch einmal: Verschuldenshaftung oder vom Verschulden unabhängige Haftung?’ Zeitschrift für Europäisches Privatrecht 2012, 560, which includes many other references on this issue.

128  Jonas Knetsch It is therefore hardly surprising that the difficulties in delineating the scope of application of strict liability become equally apparent in any attempts to define the two categories of liability. While we can agree on the idea that liability for fault concerns situations in which a person has not behaved as he or she ought to have done in a particular situation, it is much more difficult to find a similar definition for liability without fault.30 In general, the latter is presented as a default category including all the regimes of liability which are not based on any lack of care, covering therefore cases of accidental harm which would be dealt with by liability for fault only with difficulty.31 If some of the regimes of liability without fault can be linked to the notion of risk by saying that a person who has created a particular risk or was better placed to control it must bear the consequences of the event taking place,32 it must nevertheless be said that such an explanation does not justify all the cases in which either the legislator or the courts impose strict liability, notably in the area of nuisance (troubles anormaux de voisinage).33 Given that all three laws face these issues, it is surprising to note that liability without fault is perceived so differently. The explanation could lie in the relative lack of interest which the comparative study of liability without fault has attracted until now in private law scholarship in France, England and Germany. Indeed, apart from rare articles or monographs, until very recently the subject does not seem to have received the attention which it needs.34 In order to help fill this gap, this chapter will return to two commonplace views which structure the teaching of liability without fault in France and to assess their relevance in the light of what we can learn from scholarly work outside France. I will thus look in turn at the link between liability without fault and industrialisation, which is considered to be the historical origin of this form of liability, and then spend some time considering to what extent the various liabilities without fault have been integrated into the French system of liability, which appears to assimilate the case of strict liability to other actions which give rise to liability (faits générateurs) like any other.

30 On this difficulty, see Werro, Palmer and Hahn (n 11) 9ff as well as PS Atiyah, Accidents, ­Compensation and the Law (London, Weidenfeld & Nicolson, 1970) 156. 31 Werro, Palmer and Hahn (n 11) 10. See also Palmer (n 25) 1305. 32 This is the essential basis of the theory of risk as developed by French writers of the end of the 19th century to justify the relaxation of the French system which was dominated until then by the dogma of fault. See notably P Brun, Responsabilité civile extracontractuelle, 5th edn (Paris, LexisNexis, 2018) [154]–[157]. 33 Unless one considers that the mere geographical proximity of two areas of land creates a specific risk of a nuisance. On this point see the historical and comparative study of J Gordley, ‘Immissionsschutz, Nuisance and Troubles de voisinage in Comparative and Historical Perspective’ ZEuP 1998, 13. 34 Other than the works of Werro and Palmer as well as of Koch and Koziol, above n 7, see also C von Bar, Gemeineuropäisches Deliktsrecht, t 2 (CH Beck, 1999) [306]ff and, more recently, F Werro and E Büyüksagis, ‘The Bounds between Negligence and Strict Liability’ in M Bussani and AJ Sebok (eds), Comparative Tort Law. Global Perspectives (Cheltenham, Edward Elgar Publishing, 2005) 201ff; E Büyüksagis and W van Boom, ‘Strict Liability in Contemporary European Codifications: Torn between Objects, Activities, and Their Risks’ (2013) 44 Georgetown Journal of International Law 609.

The Role of Liability without Fault  129

I.  Mechanisation and Liability without Fault: The Deconstruction of a Myth In French law faculties, it is customary to present liability without fault as the offspring of the industrial revolution.35 The arrêt Teffaine of 1896 is usually cited as being the starting-point of the gradual formal recognition of a general principle of liability for the actions of things, which was justified by the advent of mechanisation and the multiplication of work accidents which it generated.36 The relationship between industrialisation and liability without fault is, moreover, also clear in the legislative activity of the period and, in particular, in the loi of 6 April 1898 on work accidents. Adopted following a particularly long and agitated parliamentary debate, this legislation imposed on employers a strict liability towards their employees for accidents of which they were the victim during the performance of their work. These two events taking place one after the other, together of course with the famous arrêt Jand’heur, made liability without fault a kind of acquis social (established body of social law), obtained as a reaction to the appearance of new sources of accidents. It is true that the importance of the arrêt Teffaine in 1896 and the adoption of legislation governing accidents at work in 1898 should not be underestimated. These developments definitely combined at the same period with legislative intervention and a movement within the Cour de cassation which sought to free itself so as to allow it to tilt the French system of liability, which was essentially based on fault, towards a system instead moving towards strict liability. But while it is undeniable that the rise of liability without fault is explicable by reference to changes taking place in French society during the second half of the nineteenth century, it is not accurate to say that this involved a new idea and that the law of civil liability of the Napoleonic period was exclusively concerned with liability for fault. By focusing on the history of liability, it becomes clear that the feature of strict liability existed in the ancien droit as well as in the Code civil of 1804 itself and that it was imposed only after an intense ideological debate.

A.  Liability without Fault before the Napoleonic Codifications There is a provision in the Napoleonic Code which has remained unchanged until our own time and which bears witness to the antiquity of the idea of liability 35 See eg G Viney, Traité de droit civil, Introduction à la responsabilité, 3rd edn (Paris, LGDJ, 2008) [24] (‘the first legislative provision to have officially created a liability without fault is the great loi of 9 April 1898 which governed the law of work accidents’). To the same effect see Brun (n 32) [154] as well as J Flour, J-L Aubert, É Savaux, Les obligations, 2: Le fait juridique, 14th edn (Paris, Sirey, 2011) 79. Compare, however, with J Carbonnier, Droit civil, 2: Les biens. Les obligations (Paris, PUF, 2004) [1119]–[1120] (the author terms this development correctly as an ‘(apparent) return to the very ancient past’, as we will later explain). 36 Cass civ 16 June 1896, S 1897, 1, 17 note P Esmein.

130  Jonas Knetsch without fault. This is article 1386 (now article 1244) Cc concerning liability for the action of ruinous buildings. Somewhat fallen into scholarly oblivion,37 this article provides that: The owner of a building is liable for the harm caused by its ruin, where the latter occurred as a result of a lack of maintenance or defect in construction.38

At first sight this provision could be thought to duplicate article 1382 (now article 1240) Cc and its general principle of liability for fault. However, what allows us to classify this rule of liability as falling within the category of liability without fault is the fact that the defect in construction or lack of maintenance do not necessarily involve the actions of the present owner, the first having probably existed since the building was constructed and the second being able to result from neglect by a tenant who bears a share in the maintenance of the property. It is therefore understandable that, behind its innocuous appearance, article 1386 of the Code civil of 1804 already expressed the idea according to which a building owner can be liable without having committed any fault.39 This observation contrasts with the position adopted by French legal scholarship which places the origins of liability without fault in the socio-economic upheavals of the end of the nineteenth century, and not in the roots, whether Romanist or Germanic, of the Code civil.40 However, the liability without fault of a building’s owner finds its origin in Roman law and, in particular, in the cautio damni infecti, a compulsory guarantee which the owner of a building had to provide to his neighbours in order to cover the risk of harm resulting from its falling into dilapidation,41 a liability which legal historians have recognised since the end of the nineteenth century bore the characteristics of a regime of strict liability.42 Indeed, there existed in Roman law a number of regimes of liability without fault which, while they applied only occasionally and still fell under the aegis of criminal responsibility, completely dismantle the idea that strict liability was a product of the modern or even contemporary legislator. Whether we are concerned 37 See, however, the profound study by V Depadt-Sebag, La justification du maintien de l’article 1386 du Code civil (Paris, LGDJ, 2000). 38 ‘Le propriétaire d’un bâtiment est responsable du dommage causé par sa ruine, lorsqu’elle est arrivée par une suite du défaut d’entretien ou par le vice de sa construction’. 39 On the embarrassment of the drafters of the Code civil in relation to the peculiarity of art 1386 in a system of liability dominated by the idea of fault, see Depadt-Sebag (n 37) [156] with further bibliography. 40 See the authors cited above n 35. 41 On this law see, in French, J Mazodier, La cautio damni infecti (Paris, A Giard/H Jouve, 1890) and, more recently, Depadt-Sebag (n 37) [9]ff and D Deroussin, Histoire du droit des obligations, 2nd edn (Paris, Economica, 2012) [781]–[782]. In German, see O Wendt, Lehrbuch der Pandekten (Jena, G Fischer, 1888) § 285 at 672ff. 42 This guarantee could indeed be undertaken even outside cases where the harm could have been prevented if careful maintenance had been effected. For an example of a scholar who emphasises the strict nature of this regime see, in German, J Unger, Handeln auf eigene Gefahr. Ein Beitrag zur Lehre vom Schadensersatz, 2nd edn (Jena, G Fischer, 1893) 75ff (‘dem Bereich des subjektiven Verschuldens entrückt‘) and H von Burckhard, Die cautio damni infecti (Erlangen, Palm & Enke, 1875) 7 (‘von einer Verschuldung oder einem Surrogat derselben kann hier keine Rede sein‘).

The Role of Liability without Fault  131 with liability for the action of animals (actio de pauperie and actio de pastu), the action of the throwing out of solid things or the spilling of liquids from buildings (actio de deiectis vel effusis), the action of using another person’s building materials (actio de tigno juncto) or the action relating to mining or water-works, the examples can serve very well to illustrate the antiquity of liability without fault43 and, correlatively, the relative modernity of the primacy which is generally attributed to liability for fault.44 Decried by some nineteenth-century Roman legal scholars as mere curiosities45 and unwelcome, given the elevation of fault (culpa) as the basis of all civil liability,46 these cases of strict liability were nevertheless not without practical importance, as is evidenced in particular by their vitality in those legal systems in which the sources of Roman law are still directly applicable, as in the law of South Africa.47 The importance of liability without fault was even greater in the Germanic laws, which were, however, characterised by a relative confusion of the civil and criminal aspects of liability.48 According to legal historians, from the beginning of the n ­ ineteenth century, proof of fault was hardly required in order to characterise a person’s action as unlawful. The legal concept of strict liability (Erfolgshaftung) is found most prominently in the Saxon Mirror (Sachsenspiegel), a medieval codification which was very influential across the whole of central and northern continental Europe.49 Whether in relation to liability for the actions of animals50 or as regards certain cases of liability of masters for the actions of their h ­ ousehold servants,51 here we can discern the elements which are the precursors of the modern c­ onception of liability without fault. 43 On all these see R Bienenfeld, Die Haftungen ohne Verschulden (Berlin, J Springer, 1933) 13ff (see also the several bibliographical references cited at p 45ff, fnn 5–19). 44 On this point see the leading study by O Descamps, Les origines de la responsabilité pour faute personnelle dans le Code civil de 1804 (Paris, LGDJ, 2005). 45 On the attempts to reinterpret these regimes as cases of liability for fault under the Lex Aquilia, notably by Jhering in his work Fault in Roman Private Law (R von Jhering, Das Schuldmoment im römischen Privatrecht (Giessen, E Roth, 1867)), see Bienenfeld (n 43) 46. 46 To this effect see Esser (n 20) 47 (‘die Pandektisten und Romanisten wußten (mit diesen H ­ aftungen) nichts anzufangen‘). 47 See notably M Loubser and R Midgley (eds), The Law of Delict in South Africa, 2nd edn (Oxford, Oxford University Press, 2012) 375ff. For a recent case in which the scope of the actio de pauperie was discussed in relation to an ostrich bite see Van der Westhuizen v Burger 2018 (2) SA 87 (SCA). 48 On this characteristic of laws of Germanic origin see notably H Brunner, ‘Sippe und Wergeld nach niederdeutschen Rechten‘, Zeitschrift der Savigny-Stiftung für Rechtsgeschichte. Germanistische Abteilung 1882, 1ff. Cf, for the French ancien droit, Descamps (n 44) 227ff. 49 See notably F Hepp, Die Zurechnung auf dem Gebiete des Civilrechts (Tübingen, CS Osiander, 1838) 118ff (the author emphasises the contrast between Roman law and the Germanic laws) as well as the earlier R Maurenbrecher, Iuris Germanici atque praesertim Speculi Saxonici de culpa doctrinam (Düsseldorf, Schaub, 1827). More recently see W Ludewig, Die Gründe der Schadensersatzpflicht aus unerlaubten Handlungen in den Volksrechten (Marburg, Elwert, 1925). 50 The fundamental study is by H Isay, ‘Die Verantwortlichkeit des Eigenthümers für seine Tiere‘, Jherings Jahrbücher für die Dogmatik des bürgerlichen Rechts 1898, 209, esp 232ff. 51 O Hammer, Die Lehre vom Schadensersatze nach dem Sachsenspiegel und den verwandten Rechtsquellen (Breslau, W Koebner, 1885) 79ff, esp 82ff. (strict liability concerned in particular harm caused by servants in serfdom and castle retainers).

132  Jonas Knetsch This recollection of the ancient sources of strict liability is not merely i­mportant solely from the angle of the history of legal ideas, but also throws a new light on the position in German law. Indeed, the Germanic conception of liability without fault resurfaced at the time of the preparatory works of the BGB and especially in the writings of von Gierke. Von Gierke was an indefatigable defender of a socialist conception of civil law, one which invoked the Germanic origins of liability without fault in order to extol the virtues of the relaxation of the dogma of culpa and, in this way, the introduction of his famous ‘drop of socialist oil’ (Tropfen sozialistischen Öls) into the BGB.52 And von Gierke expressed regret that liability without fault, ‘no matter whether it has deep roots in German legal thought, cannot be reconciled with the “principles” of our future German codification’ before remarking bitterly that the latter was ‘more Romanist than Roman law itself ’.53 Since then, the pre-eminent place which nineteenth-century civil law ­scholarship attributed to fault as the basis of civil liability has sat rather uncomfortably with the Romanist and Germanic sources on which the drafters of the French Code civil and the German BGB very largely relied. By a curious chance of history, von Gierke’s invocation of a ‘Germanic vision’ of liability detached from fault ­coincided with the debate in France about compensation for the victims of accidents at work whose outcome is well-known: the confirmation of liability without fault by the Cour de cassation in its case law in the arrêt Teffaine and by the loi of 6 April 1898, as the second pillar of French law’s system of civil liability.

B.  The Industrial Revolution and the Spectre of Liability without Fault One cannot overemphasise the importance of accidents at work for the history of civil liability in Europe,54 and, more broadly, for the social history of our ­continent.55 In France, it is the adoption of the loi of 9 April 1898 ‘concerning liability for accidents of which workers are the victims during their work’ which constituted the culminating point of a political debate which has remained exceptional as much for its length as for its intensity.56 This legislation was the result of a clash between

52 On this point see Esser (n 20) 45. 53 O Gierke, Der Entwurf eines bürgerlichen Gesetzbuches und das deutsche Recht (Berlin, Duncker & Humblot, 1889) 261. 54 In English, see B McMahon, ‘The Reactions of Tortious Liability to Industrial Revolution: A Comparison’ (1968) 3 Irish Jurist 18, 284. Cf also for US law D Gifford, ‘Technological Triggers to Tort Revolutions: Steam Locomotives, Autonomous Vehicles, and Accident Compensation’ (2018) Journal of Tort Law, vol 11, no 1. 55 Within the French literature see the fundamental study F Ewald, La naissance de l’État providence (Paris, Grasset & Fasquelle, 1986). 56 See the very numerous references cited by G Aubin and J Bouveresse, Introduction historique au droit du travail (Paris, PUF, 1995) [315]. For a more nuanced analysis, see however G Aubin, ‘La loi du 9 avril 1898, rupture ou continuité ?’ Droit social 1998, 635.

The Role of Liability without Fault  133 conservative and socialist currents and marked a break with the liberal logic which made liability for fault the driving force of all progress, individual or collective. With the encouragement of case law of the Cour de cassation in favour of the erosion of fault as the exclusive basis of duties of reparation, P ­ arliament made a major contribution to the revival of the idea of strict liability the germ for which, as we have just seen, was already present in the Code of 1804. An analysis of the travaux préparatoires for the legislation of 1898 bears witness to the strength of the forces of resistance against the formal recognition of liability in employers devoid of any link with any kind of negligence or intentional fault. It was only after 18 years of discussion that the elected representatives succeeded in agreeing on a regime of liability which imposed on employers on the occurrence of an accident at work a duty of reparation detached from any investigation into the possible negligence on the part of employers or workers. In return for this quasi-automatic liability, however, the legislation of 1898 provided only for fixed-rate compensation, a principle still in force in the regime governing work accidents and work-related illness (called the ‘AT/MP regime’).57 Beyond its interest for the study of French society at the end of the nineteenth century, the origins of the legislation of 1898 also illustrate an ideological charge in the law of liability. While it has been written that this branch of the law is essentially technical and escapes the influence of interest groups and does not have a role for the better distribution of wealth,58 it is difficult to accept such a neutral vision of civil liability. Indeed, more than any other branch of civil law, civil liability appears to be the result of a subtle game of political balances, marked by the opposition of divergent interest groups. To return to the example of the debate on the compensation of the victims of accidents at work, this very much belonged to thinking about the appropriateness of the redistribution of wealth within a society shaken by industrialisation. In reading the position statements adopted in the Assemblée nationale and the Sénat, one can easily see the fear of some members of Parliament faced with the idea of liability without fault, itself considered a serious infringement of freedom of enterprise and a first step towards the socialisation of French industry.59 With regard to the idealogical flavour of the debates which preceded the adoption of the legislation of 1898, it could be tempting to see in liability without fault an instrument of legal policy which is at once for the benefit of the very poorest 57 Art L 451-1 Code de sécurité sociale. On this feature of the current system and it being anachronistic, see P Morvan, Droit de la protection sociale, 8th edn (Paris, LexisNexis, 2017) [171]ff. 58 See eg W Landes and R Posner, The Economic Structure of Tort Law (Cambridge/London, Harvard University Press, 1987) 15 (‘it does not seem plausible to suppose either that an interest group would organize to seek redistribution through the accident-law system … or that some existing group – the poor, blacks, retail druggists – would place accident law high on the agenda for legislative action … the scope for systematic wealth redistribution in regard to accidents … does seem much more limited than in regard to tariffs, price regulation, taxation, wages, the development of public land, and other familiar areas of specific interest legislation’). 59 See E Tarbouriech, La responsabilité des accidents dont les ouvriers sont victimes dans leur travail (Paris, Giard & Brière, 1896) [127] and [131] with several quotations from parliamentary debates. Cf R Ogorek, Untersuchungen zur Entwicklung der Gefährdungshaftung im 19. Jahrhundert (Cologne, Böhlau-Verlag, 1975) 98ff.

134  Jonas Knetsch and to the detriment of the very richest people. However, other examples show that the political issues surrounding liability without fault are very subtle indeed, as can clearly be seen in the debate on the question of the liability for trains in Prussia in the 1830s.60 Indeed, the liability without fault of carriers which was introduced by legislation of this date, benefited passengers coming from a relatively comfortable class in society who had been exposed to the dangers stemming from a technology which had not yet been mastered, but it also benefited farmers who feared that sparks coming from the trains would cause devasting fires on their land.61 The important place in the parliamentary preparatory work given to the question of compensation for farmers shows clearly the influence of the owners of large estates in the Prussian Council of State (Preußischer Staatsrat) and, more generally, the diversity of the political interests in play.62 These reflections on the political context of the creation of liability without fault in certain areas of activity does not, however, explain why the French and German systems accord such a very different importance to strict liability. The example of the legislation of 3 November 1838 which introduced very early a liability without fault in the context of railway accidents and the avant-garde character of ­Bismarckian legislation on insurance against accidents at work contrasts strikingly with the modest place which liability without fault occupies in German legal thought. Inversely, the ubiquity of liability without fault in French law can be shocking when contrasted with the supremacy of fault in the system of liability in the Napoleonic code and the lateness of the French debate on the fate of injured workers. In the absence of a deeper historical and comparative study of this question, we can only sketch out the main tracks of reflection which may explain the ­curious contrast between the two legal systems. First, the structure of the French Code civil lent itself particularly to its system of civil liability becoming progressively stricter. By their brevity and their generality, articles 1382 to 1386 of the Code of 1804 provided a fertile soil for the development of the law of liability by the courts, while paragraphs 823 et seq of the BGB, adopted at a time when industrialisation was already in full swing, left little room for the creative role of judges.63 60 The law was promulgated on 3 November 1838 and named Gesetz über die Eisenbahn-Unternehmungen (Preußische Gesetzessammlung 1838, 505). 61 On the political context of this law, see, in German, T Baums, ‘Die Einführung der Gefährdungshaftung durch F.C. von Savigny‘, Zeitschrift der Savigny-Stiftung für Rechtsgeschichte. Germanistische Abteilung 1987, 277 and, in English, J Kleeberg, ‘From Strict Liability to Workers’ Compensation: The Prussian Railroad Law, the German Liability Act, and the Introduction of Bismarck’s Accident ­Insurance in Germany’ (2003) 36 NYU J Int’l L & Pol 53, esp 66ff. 62 In this sense, see also Jansen (n 21) 109, mentioning other references as well. For a less nuanced view which emphasises the role of liability without fault as an instrument for the redistribution of wealth see, however, M Adams, Ökonomische Analyse der Gefährdungs- und Verschuldenshaftung (Heidelberg, R v Decker, 1985) 111ff. 63 Some of the general notions used by the text of § 823(1) BGB (‘sonstiges Recht [another right]’) and of § 826 BGB (‘vorsätzliche sittenwidrige Schädigung [the intentional infringement of good morals]’),

The Role of Liability without Fault  135 Since this time, the ambition of the Cour de cassation to adapt French law as inherited from Domat and Pothier to new social circumstances was able to be expressed much more liberally in France than in Germany. To give an important place to liability without fault was therefore a way in which elements of social policy could be introduced into the civil law which had generally been earmarked by the legislator for inaction.64 Another reason is without doubt even more decisive. In the first half of the twentieth century, the great majority of German jurists were under the i­ntellectual influence of the Pandectist School and its renewed vision of Roman law. As a result, legal scholarship was, much more so than in France, attached to the ­principle of liability for fault (Culpa-Prinzip) which the Pandectists had worked so hard to distil from the Roman legal texts, while passing over in silence the cases of strict liability which has already been mentioned. As a result, liability without fault could not find a home except outside the newly codified law. Even nowadays, German law is characterised by the almost total absence of cases of extra-contractual liability without fault,65 which, if one looks at legal commentaries, appears to be a source of embarrassment to no one. In this respect it is quite revealing that in current usage the expression Deliktsrecht is reserved for regimes of liability based on fault, and another form of words, Recht der unerlaubten Handlungen (literally, the law of ‘unlawful actions’) is used to describe the area as a whole including therefore cases of liability without fault.66 This last terminological point underlines the contrast between the French and German systems at another level, that is, the degree of integration of liabilities without fault in the system of civil liability.

II.  The Degree of Integration of Liabilities without Fault Within the System of Liability From the French point of view, the different regimes of liability for the actions of things and of other people appear to be integrated without difficulty within the

which could have allowed the relaxation of the system of delictual liability, have always been interpreted strictly. See on this point, in French, M Fromont and J Knetsch, Droit privé allemand, 2nd edn (Paris, LGDJ, 2017) [365], [373]–[375]. 64 H Mazeaud, L Mazeaud and A Tunc, Traité théorique et pratique de la responsabilité civile délictuelle et contractuelle, 6th edn (Paris, Montchrestien, 1965) vol 1, [85]ff; Carbonnier (n 35) [1176] at 2361 (the author emphasises the importance of the nomination of Josserand to the Civil Chamber of the Cour de cassation). See also the more nuanced analysis by R Savatier, Traité de la responsabilité civile en droit français, 2nd edn (Paris, LGDJ, 1951) [274]ff. 65 The only exception being that of liability for the actions of pet animals (Luxustiere), established in § 833 BGB. On this point, see Fromont and Knetsch (n 63) [385]–[386]. 66 See J Hager, commentary in Staudinger BGB. §§ 823 (A-D) (Berlin, Sellier/de Gruyter, 2017), Vorbemerkung zu §§ 823 ff, no 25; K Larenz and CW Canaris, Lehrbuch des Schuldrechts, t. 2/2, 13th edn (Munich, CH Beck, 1994) 600 (‘für die Haftung aus Delikt oder unerlaubter Handlung i.S. der §§ 823 ff. BGB ist die Anknüpfung an das Verschuldensprinzip konstitutiv’ (emphasis added)).

136  Jonas Knetsch general theory of liability so that they are subject to the same rules governing harm and causation as well as to the same effects of liability as are defined by the general law of liability for fault. This apparent absence of special treatment separates it strikingly from the perception of liability without fault in other legal systems such as German law. Indeed, German writers are unanimous in underlining its ­‘exceptional character’ (Ausnahmecharakter)67 and emphasise the derogations which distinguish it from the general law of civil liability.68 Despite a tendency towards a certain rapprochement with the general law,69 liability without fault continues to be presented as a model which is diametrically opposed to the model of liability for fault.70 In the light of these insights from German law, we need to assess the degree of integration of liability without fault in French law and to investigate whether there are at least some traces of the special treatment which characterise it in other legal systems. In order to do so, we will examine in turn the two aspects which, according to German writers, justify a separate treatment, that is to say, the extent of the reparation and the assessment of causation.

A.  The Limitation of Reparation versus the Principle of Full Reparation The difference in concept of liability without fault in the two different legal systems is made clear first and foremost at the level of the effects of liability. In principle, it is well-known that once the conditions of liability are established, the claimant can demand that he be restored to the position which he would have been in if the harmful action had not taken place, with neither gain nor loss. French legal scholars classify this return to the status quo ante as the principle of full reparation, as the claimant has the right to reparation for all the repercussions of the harmful action (‘all the loss, but nothing but the loss’).71 Moreover, the notion of 67 Deutsch (n 16) [642]; D Medicus, ‘Gefährdungshaftung im Zivilrecht‘ Juristische Ausbildung (Jura) 1996, 561 (‘Verschuldenshaftung als Regel‘); J Esser and HL Weyers, Schuldrecht, t. 2/2: Besonderer Teil. Gesetzliche Schuldverhältnisse, 8th edn (Cologne, CF Müller, 2000) 266ff; G Wagner, commentary in Münchener Kommentar zum BGB, t. 5: Schuldrecht. Besonderer Teil III, 7th edn (Munich, CH Beck, 2017) Vor § 823, [17]. 68 See M Fuchs, W Pauker and A Baumgärtner, Delikts- und Schadensersatzrecht, 9th edn (Berlin/ Heidelberg, Springer, 2017) 284ff. 69 On the reform legislation of 2002 which extended reparation for non-pecuniary damages to ­liability without fault see below. 70 For a key study see U Bälz, ‘Ersatz oder Ausgleich? Zum Standort der Gefährdungshaftung im Licht der neuesten Gesetzgebung’ JuristenZeitung 1992, 57 (the author defends the thesis that liability without fault should possess a regime fundamentally different from liability for fault, the first belonging rather to the law of unjustified enrichment whereas the second serves to protect the safety of people and property). On this see also von Bar (n 34) [306] and fn 5. 71 On this principle see Brun (n 32) [598]ff; C Coutant-Lapalus, Le principe de réparation intégrale en droit privé (Aix-en-Provence, PUAM, 2002) as well as, more recently, M Denimal, ‘La réparation intégrale du préjudice corporel: réalités et perspectives’ (Thesis, Lille 2, 2016).

The Role of Liability without Fault  137 full ­reparation appears so firmly entrenched that it does not necessarily cross the mind of a French jurist that it does not apply self-evidently in other legal systems.72 A glance at German law soon puts into perspective the claimed universal significance of the idea that full reparation is the necessary result of civil liability. If the expression Totalreparation is known there,73 it is understood rather in terms of the method of assessment of damages. Under the theory of Differenzhypothese, ‘patrimonial loss’ consists of ‘the difference between the sum which represents a person’s patrimony at a given moment and the sum which that patrimony would have represented at the same moment had it not been for the harmful action’.74 As a result, the amount which is adequate to ‘make reparation for the bad thing [and] make it appear to be no more than a dream’, to quote the famous words of ­Carbonnier, involves a comparison between two ‘patrimonial’ situations, one actual and one fictitious.75 Having regard to what has just been said, one could be tempted to accept that there is a certain equivalence between the French principle of full reparation and the German Differenzhypothese, as they lead more or less to the same result, that is to say, to a full compensation of all heads of loss. However, it is in the area of extra-patrimonial losses that this apparent harmony between French and German law is strikingly lost. As is well-known, German law has a narrower conception of compensation for ‘non-material loss’ (Immaterialschaden), granting reparation of it by way of damages only in situations specially provided for by legislation.76 And it was exactly the regimes of liability without fault from which, until 2002, full compensation of the victim including losses of an extra-patrimonial nature were excluded. Traditionally, this exclusion was justified by the idea that considerations of distributive justice could concern only direct losses of a patrimonial nature77 and, in the absence of fault, damages cannot perform their proper function of

72 The French conception of full reparation is notably incompatible with a system of civil liability which rests on a principle of listing the interests to be protected and which excludes, for example, pure economic loss. Cf, however, the comparative law study by P Pierre and F Leduc, ‘La conception g­ énérale de la réparation intégrale’ in P Pierre and F Leduc (eds), La réparation intégrale en Europe: études comparatives des droits nationaux (Bruxelles, Larcier, 2012) 31, esp 32 (having set out that ‘all countries respond positively’ to the question of the existence of a principle of full reparation, the authors admits that ‘the recognition of the principle of full reparation is more or less straightforward depending on the legal system’). 73 See H Lange and G Schiemann, Schadensersatz (Tübingen, Mohr, 2003) 9ff as well as the report (untitled) in French of U Magnus in Pierre and Leduc (n 72) 135ff. 74 Very generally adopted by German legal scholarship, this definition is derived from F Mommsen, Beiträge zum Obligationenrecht II: Zur Lehre vom Interesse (Braunschweig, CH Schwetschke, 1855) 3 (the translation is borrowed from O Berg, La protection des intérêts incorporels en droit de la réparation des dommages (Bruxelles/Paris, Bruylant/LGDJ, 2006) [67]). 75 Carbonnier (n 35) 2253. 76 On § 253 BGB, see Fromont and Knetsch (n 63) [421]. For a comparative analysis, see also G Wagner, ‘Ersatz immaterieller Schäden: Bestandsaufnahme und europäische Perspektiven’ ­JuristenZeitung 2004, 319. 77 Esser (n 20) 108ff.

138  Jonas Knetsch making amends in the case of ‘moral loss’.78 However, modern legal scholarship has challenged this legislation unremittingly, both at the level of its theoretical justification79 and at the level of its practical consequences which were adverse to the organisation of justice.80 When the law of civil liability was reformed, the legislator therefore decided to allow the monetary compensation of ‘immaterial’ loss, regardless of the legal basis of the claim for damages.81 It would not be correct to think that since the reform of 2002, liability without fault in the German conception has completely entered the fold of the general theory of the effects of liability and that, on basing their claims on a regime of strict liability, claimants can obtain a truly full reparation. If one can now claim ‘a fair compensation in money’ by way of reparation of one’s ‘immaterial’ loss in the case of ‘damage to the body, to health, to personal liberty or to sexual self-­ determination’,82 account must be taken of the fact that there is a ceiling of liability attached to the vast majority of regimes of liability without fault.83 According to some authors, the existence of these ceilings is justified by considerations which are linked to the insurability of the risk of liability, their amount being generally the result of a negotiation between insurance, public authorities and legal actors likely to incur civil liability.84 However, if it is understandable that insurance companies are reluctant to take on a new risk which is difficult to assess with established actuarial techniques, it is quite a different matter as regards risks that have existed for decades, even for more than a century. The truth is that the ceilings of liability are intended rather to reduce the cost of the risk of liability and so favour a category of economic actors than to produce an economically viable regime of liability. From this perspective the link between liability without fault and ceilings of liability put forward by German legal scholarship is hardly

78 Larenz and Canaris (n 66) 603. 79 See in particular H Kötz, ‘Zur Reform der Schmerzensgeldhaftung‘ in Festschrift für Ernst von Caemmerer zum 70. Geburtstag (Tübingen, Mohr, 1978) 389, as well as J Köndgen, Haftpflichtfunktionen und Immaterialschaden (Berlin, Duncker & Humblot, 1976) 124ff. 80 Köndgen (n 79) 136ff as well as H Kötz, ‘Gefährdungshaftung‘ in Bundesministerium der Justiz (ed), Gutachten und Vorschläge zur Überarbeitung des Schuldrechts (Bonn, Bundesanzeiger, 1980) vol 2, 1779, esp 1824ff. 81 On this aspect of the reform, see C Katzenmeier, ‘Die Neuregelung des Anspruchs auf Schmerzensgeld‘ JuristenZeitung 2002, 1029. 82 § 253(2) BGB: ‘Ist wegen einer Verletzung des Körpers, der Gesundheit, der Freiheit oder der sexuellen Selbstbestimmung Schadensersatz zu leisten, kann auch wegen des Schadens, der nicht Vermögensschaden ist, eine billige Entschädigung in Geld gefordert werden‘. 83 Only liability for the actions of animals and the legislative regime of liability for water pollution have no ceiling for compensation. On this point, see Medicus (n 67) 561, esp 565. For an overview of the ceilings applicable to the different regimes of strict liability see L Jaeger and J Luckey, ‘Das Zweite Schadensersatzrechtsänderungsgesetz. Ein Überblick über das neue Recht‘ Monatsschrift für Deutsches Recht 2002, 1168, esp 1174. 84 On the genesis of of the Convention of Bruxelles of 29 November 1969 on civil liability for oil pollution damage (‘Convention CLC’), see J Knetsch, Le droit de la responsabilité et les fonds d’indemnisation (Paris, LGDJ, 2013) [194].

The Role of Liability without Fault  139 convincing, and this explains why many authors are sceptical or even openly opposed to a systematic limitation of compensation for strict liability.85 The fact remains, though, that the existence of a ceiling of reparation is always seen by German jurists as an essential characteristic of liability without fault, while its French equivalent sees this type of limitation, which also exists in French law,86 as a rather strange exception within the law of civil liability.87 This difference between the two systems of liability (for fault and without fault) which is made clear at the level of the effects of liability is confirmed when one turns to the conditions of liability and notably to causation.

B.  The Assessment of Causation in the Context of Liability without Fault Should causation be assessed differently depending on whether the claimant’s right to reparation is based on a liability for fault or on one of the regimes of strict liability? Here again the French conception is distinct from other legal systems which answer this question in a resolutely positive way. By drawing on the causal theories of the ‘equivalence of conditions’ and ‘adequate causation’, the French case law retains a very flexible vision of the causal link between the action giving rise to liability (fait générateur) and the harm relied on by the claimant, having always refused to take a position in the scholarly debate as to the different possible methods of its assessment.88 In the past, there were attempts by legal scholars to systematise the case law solutions on causation according to the nature of the liability relied on, relating the theory of the equivalence of conditions to liability for fault and the theory of adequate causation to liability for the actions of things,89 but these did not attract 85 See, among others, M Will, Quellen erhöhter Gefahr (Munich, CH Beck, 1980) 311ff; H Kötz, ‘Haftung für besondere Gefahr‘ Archiv für die civilistische Praxis 1970, vol 170, 1, esp 36ff; more recently, Kötz and Wagner (n 14) [540] (‘(versteckte) Subvention der begünstigten Aktivität‘). In this sense see also Deutsch (n 16) 703. 86 This is notably the case in the law of transport, where regimes of liability originating in international or European conventions provide almost systematically for ceilings of reparation. Closer to the general law, liability for defective products does not always allow full reparation. On the threshold foreseen by art 1245-1(2) Cc in the case of ‘damage to property other than the defective product itself ’, see notably Borghetti (n 29) [509]. See also see Viney, Jourdain and Carval (n 27) [505]–[519] (including other examples). 87 See notably le Tourneau (n 15) [2321.111] (‘some hypotheses’). In most textbooks on the law of obligations, the existence of these ceilings is not even mentioned in their discussions of full reparation. Cf also M Fabre-Magnan, Droit des obligations, 2: Responsabilité civile et quasi-contrats, 3rd edn (Paris, PUF, 2013) [147] (‘the rule of full reparation is scarcely contested’). 88 On this whole question see in particular Viney, Jourdain and Carval (n 27) [348], as well as Brun (n 32) [246]ff. 89 Flour, Aubert and Savaux (n 35) [162]; P Malaurie, L Aynes and P Stoffel-Munck, Droit des ­obligations, 9th edn (Paris, LGDJ, 2017) [96]; le Tourneau (n 15) [2131.54] and [2131.56]; for a more in-depth analysis, see C Grare, Recherches sur la cohérence de la responsabilité délictuelle (Paris, Dalloz, 2005) [261]ff.

140  Jonas Knetsch the adherance of the whole of legal scholarship.90 In particular, it could be noted that such an approach did not provide an explanation for the practice of the courts, which contained a number of contrary examples91 and, above all, that ‘this doctrinal position rests on the debatable proposition that the same notion [can] receive different definitions according to the [particular] action giving rise to liability which is in issue’.92 Moreover, this final argument provides support for the French view that treats cases of strict liability as actions giving rise to liability just like any others by subjecting them to the rules provided by the general law developed in the context of liability for fault. However, despite the very casuistic approach of French courts in the context of causation, in my view the assessment of causation in the context of liability without fault is indeed imbued with a certain specificity, but that this is to be seen in relation to the means of proof of causation rather than in the area of the various theories of causation. Clearly it is not by chance that presumptions of causation93 find their preferred area of application in the context of strict liability. Whether it is a matter of litigation relating to the transfusion of contaminated blood,94 to defective products,95 to the action of things,96 or even to nuclear accidents,97 the reversal of the burden of proof is applied most often where the court is deciding upon a regime of liability without fault. Although there is no systematic connection, the preponderance of presumptions of causation in the area of liability without fault makes clear two important points. On the one hand, the assessment of causation should not be disconnected from the nature of the liability relied on by the claimant, for presumptions of causation must be taken into account whether these are imposed by the legislator or, more rarely, by the courts. Contrary to the way in which this is presented by

90 As clear evidence to this effect see C Quezel-Ambrunaz, Essai sur la causalité en droit de la responsabilité civile (Paris, Dalloz, 2010) [130]. 91 This is very readily accepted by those writers who favour a distinction being drawn as to the application of theories of causation. See notably Malaurie, Aynes and Stoffel-Munck (n 89) [96] (the authors recognise that ‘the court is guided by considerations which are not solely logical, but are also utilitarian’ and that there is in the context an ‘impressionistic casuistry’). 92 Bacache-Gibeili (n 28) [500] in fine. To the same effect see also P Brun, ‘Responsabilité du fait personnel’ Répertoire de droit civil Dalloz 2015, no 118. 93 On this technique see the detailed study by Quezel-Ambrunaz (n 90) [287]ff. (the author does not, however, relate the nature of civil liability to the existence of a presumption of causation). 94 Art 102 of loi no 2002-303 of 4 March 2002 and art L 1221-14(2) Code de la santé publique (in relation to the origin in a transfusion of contamination of blood by hepatitis C). 95 According to the Cour de cassation, proof of causation can be made by means of ‘serious, detailed and concordant presumptions’. In relation to liability for defective health products, see most recently le Tourneau (n 15) [6313.61] (with many references). 96 On the presumption of an active role of the thing and its relationship to causation, see Brun (n 32) [394]–[397]. 97 Arts L 597-12 and L 597-36 Code de l’environnement: ‘A non-exhaustive list of medical conditions which, subject to proof to the contrary, are presumed to have had as their origin the [nuclear] accident is established by means of an administrative regulation depending on the radiation and contamination received and on the delay in which the condition was found.’

The Role of Liability without Fault  141 legal scholars, the strict nature of a regime of liability is very much reflected in the assessment of causation so that, at least at a theoretical level, the fait générateur forms an inseparable whole with other conditions for the imposition of liability. On the other hand, the relaxation of the requirements regarding causation shows very well the flexibility of the system of civil liability. Beyond the definition of the actions which give rise to liability, causation also plays the role of variable adjuster so as to make the regime of liability more or less restrictive, allowing the avoidance of an overly rigid dichotomy between liability for fault on the one hand and liability without fault on the other, and so integrates it into a more differentiated system.98 The way in which the legislator or the courts understand the assessment of causation therefore determines what a German writer has called a ‘liability threshold’ (Haftungsschwelle)99 which a claimant must show that he has crossed and which, in the case of fault liability, is to be found more in the area of the characterisation of the defendant’s conduct as negligent rather than in proof of causation.100 When one considers the rise in presumptions of causation in the areas where the claimant finds himself from the outset dispensed from proving fault, it may be asked whether we should reintroduce restrictions limiting the scope of liability without fault, in particular in cases in which the harm in question is an expression of the realisation of the risk whose identification led to the recognition of the regime by the legislator or by the courts. To put this another way, is it appropriate to require proof of a link of imputation between the harm and this risk? Such a link, which is called Risikozusammenhang in German, is an element which German legal scholars emphasise in the context of liability without fault101 and which is seen as related to the theory termed ‘Aquilian relativity’ (Lehre vom Schutzzweck der Norm) according to which harm is reparable only where it was caused by the infringement of an interest which the norm violated had the purpose of protecting.102 To illustrate this restriction, a decision of the German Federal Court of Justice may be cited in which the Court did not impose strict liability on the keeper of a vehicle against which a motorcyclist was thrown when it skidded.103 98 For a presentation of the different models of liability without fault depending on their ­respective requirements of causation, see Deutsch (n 16) [661]ff (the author distinguishes between ‘narrow’ liability without fault (enge Gefährdungshaftung) from ‘enlarged’ liability without fault (erweiterte Gefährdungshaftung) and from ‘liability for presumed causation’ (Kausal-Vermutungshaftung)). 99 E Deutsch, ‘Das neue System der Gefährdungshaftungen: Gefährdungshaftung, erweiterte ­Gefährdungshaftung und Kausal-Vermutungshaftung’ NJW 1992, 73. 100 Deutsch (n 16) [687] in fine. 101 Among the most recent teaching manuals, see Kötz and Wagner (n 14) [520]; Fuchs, Pauker and Baumgärtner (n 68) 285. See also, in English, F Werro, V Palmer and AC Hahn, ‘Synthesis and Survey of the Cases and Results‘ in Werro and Palmer (n 7) 387, esp 405 (the same idea is equally present in Italian and Portoguese law). 102 On this theory, see in the francophone literature H Slim, ‘Le lien de causalité: approche comparative’ RLDC 2007 Suppl to no 40, 63, esp 65ff; O Berg, ‘L’influence du droit allemand sur la responsabilité civile française’ RTD civ 2006, 53, [24]ff ; more recently, M Lacroix, ‘La relativité aquilienne en droit de la responsabilité civile – analyse comparée des systèmes germanique, canadien et québécois’ Revue de droit de McGill 2013, 425, esp 437ff (with other references mentioned in fn 46). 103 BGH, 15 June 1983, réf IVa ZR 209/81, NJW 1984, 41, note H Tschernitschek. On this, see Esser and Weyers (n 67) 269.

142  Jonas Knetsch According to the Court, the specific risk created by motorised traffic did not arise on the facts,104 as the motorcyclist could equally have been thrown against a wall or against a pedestrian.105 Whereas in German law, this restriction is generally respected by means of a strict interpretation of the actions which give rise to liability,106 the French system is more supple.107 As a function of considerations of legal policy, the link of imputation of harm to the type of risk on which the regime of liability is based is sometimes put into effect by a judicial refinement of the action which may give rise to liability,108 sometimes simply suppressed entirely as in the context of motortraffic accidents where the legislation of 1985 grants a right to reparation whenever a vehicle is involved ‘in whatever respect’ in contributing to the accident.109 In certain situations, there are also defences (specifically, causes d’exo­ nération)110 provided by the regime of liability without fault which can fulfil the same restrictive function. The possibility of destroying the causal link by showing the existence of force majeure or fault in the victim in effect allows the confinement of the regime of reparation to harm which constitute the realisation of the risk which led the legislator or the court to put them in place. Again account needs to be taken of the diversity of political choices, as are made clear, for example, in the acceptance or in the rejection of force majeure as a means of escaping liability. A comparison of the different national regimes of liability governing road accidents, but also the different special regimes within an individual legal system, make clear the very wide range of possible solutions and brings this aspect of establishing causation closer to its function as a variable mechanism of adjustment of liability.111

104 BGH (n 103): ‘Das verkehrsgerechte Parken eines Pkw auf einem Seitenstreifen außerhalb der Fahrbahn ist nicht allgemein geeignet, die Verletzung eines Mopedfahrers herbeizuführen, insbesondere nicht dessen Frontalzusammenstoß mit einem anderen Fahrzeug in Fahrbahnmitte, durch den er 20 m weit durch die Luft und gegen den parkenden Pkw geschleudert wird.’ 105 Esser and Weyers (n 67) 269. 106 This is notably the case for the term ‘exploitation’, which is used in relation to almost all the provisions providing for a strict liability rule (the conditions of liability are generally formulated in the following way: ‘If a person is killed or … injured or … a thing is damaged by the exploitation of …‘). From a comparative perspective see, on this point, Werro, Palmer and Hahn (n 11) 12. 107 Compare, however, art 23 Avant-projet Terré, according to which the liability of operators of classed installations is limited to cases where ‘it is precisely the realisation of the risk justifying the classification which has caused the damage’. On this condition, see Borghetti (n 18) 163, esp 179. 108 The clearest example is certainly the development of the notion of ‘positive action of the thing’ (fait actif de la chose), absent from the letter of art 1384(1) of the Code civil of 1804. It is also possible to mention the requirement of an abnormal conduct of the animal (fait anormal de l’animal) to trigger the liability rule envisaged in art 1243 Cc. 109 Loi no 85-677 of 5 July 1985 Loi Badinter art 2; Cass civ (2) 2 April 1997, no 95-13303. 110 On these see ch 14 below, pp 292ff. 111 See notably Deutsch (n 16) [694]ff, as well as Larenz and Canaris (n 66) 604. In the same sense see, in English, Werro, Palmer and Hahn (n 101) 405.

8 Fait d’autrui in Comparative Perspective BIRKE HÄCKER

I. Introduction All legal systems have to grapple with the question of when, why and how to impose liability on one person for harmful acts committed by another. They give different answers, depending on their historical evolution, overall conceptual framework, social and economic setting as well as the underlying value judgements and policies – such that even closely related orders tend to deviate in detail. Occasioned by the proposals for the reform of the French civil liability regime published in March 2017,1 the purpose of the present chapter is to outline and compare both the current and the envisaged French regime on liability for fait d’autrui with the relevant English and German law. The main focus will be on parental liability for children and an employer’s (typically a business’s) liability for the wrongful actions of employees, but other scenarios will also be mentioned where appropriate.

II.  Setting the Scene To set the scene and facilitate the overview, we should bear in mind some ­background information and ponder briefly a few basic issues and distinctions.

1 Ministère de la Justice, Projet de réforme de la responsabilité civile (13 March 2017), www.justice. gouv.fr/publication/Projet_de_reforme_de_la_responsabilite_civile_13032017.pdf; English translation by S Whittaker and J-S Borghetti, Reform Bill on Civil Liability (March 2017), www.textes.justice. gouv.fr/art_pix/reform_bill_on_civil_liability_march_2017.pdf (both last accessed 23 April 2019). The text of the reform proposal is cited in accordance with this English translation.

144  Birke Häcker

A.  Acts of Persons and Damage Done by Things There are special challenges to holding someone liable for the acts of another – a human being who is a free actor and herself2 a subject of the law to whom liability can (potentially) be ascribed. In this, the human actor differs from an inanimate object or an animal inflicting the damage. Nevertheless, we find engrained in the structure of the French Code civil a close connection between the two situations. The current article 1242 (former article 1384)3 Cc declares in paragraph 1: One is liable not only for the harm which one causes by one’s own action, but also for that which is caused by the action of persons for whom one is responsible, or of things which one has in one’s keeping.

Without here going into the way in which this provision was understood and developed by the courts,4 it is worth noting that the reform proposal retains this close connection, dealing in article 1243 and in articles 1245–49 of the Projet de réforme with the ‘action of things’ and with the ‘imputation of harm caused by another person’, respectively. Article 1243 provides:5 A person is liable strictly for harm caused by the action of corporeal things within his keeping. An action of a thing is presumed wherever, while moving, it comes into contact with the person or property which is harmed. In other cases, it is for the victim to prove the action of the thing by establishing either its defect or the abnormality of its position, its state or its behaviour. The keeper of a thing is the person who has the use, control and direction of the thing at the time of the action causing the harm. The owner of a thing is presumed to be its keeper.

Interestingly, a separate paragraph dealing with animals, which had been contained in the initial draft put out for consultation on 2016,6 no longer features in the proposal. Yet this is not a reversion to the traditional view (now widely ­considered outdated) that simply equates animals with ‘things’. Instead, it seems that the 2 For exposition purposes, the present chapter will treat the defendant as male and the person (child, employee, etc) for whose actions he is to be held liable as female. 3 The text of the reformed French law of obligations, which came into force on 1 October 2016 and was partially revised with effect from 1 October 2018 during the parliamentary ratification process, is cited in accordance with the English translation by J Cartwright, B Fauvarque-Cosson and S Whittaker, www.textes.justice.gouv.fr/art_pix/THE-LAW-OF-CONTRACT-2-5-16.pdf (last accessed 23 April 2019). 4 On which see below, text accompanying and following n 50. 5 This is essentially a codification of the case law which was initiated by the arrêt Teffaine, Cass civ 16 June 1896, S 1897, I, 17 note Esmein, DP 1897, I, 433 note Saleilles, later generalised by the famous decision in Jand’heur, Ch Réun 13 February 1930, S 1930, I, 121 note Esmein, DP 1930, I, 57 note Ripert, and subsequently further developed and refined by cases such as Franck, Ch Réun 2 December 1941, S 1941, I, 217 note Mazeaud, DC 1942, 25 note Ripert. 6 At the time, art 1243 of the Avant-projet contained a fifth paragraph to the effect that the provision was also applicable to the ‘acts of animals’.

Fait d’autrui in Comparative Perspective  145 explanation lies in the recent introduction of a new article 515-14 into the Code civil by virtue of which animals are expressly recognised as ‘living beings endowed with sentience’, but – subject to specific welfare provisions – nonetheless governed by the ordinary property law regime.7 Given the generalisation of liability for the ‘actions of things’, the proposal is able to dispense with any equivalent to the current article 1243 (former article 1385) Cc, according to which ‘the owner of an animal, or the person who makes use of an animal while he uses it, is liable for the harm which the animal has caused, whether the animal was in his keeping or whether it had gone astray or escaped’.

B.  The Defendant’s Own Actions: Attribution Rules The genetic connection between liability for damage caused by things and persons and the fact that the former is justified, in principle, by reference to the element of control exercised by the defendant over the object concerned (la garde), contains a first clue as to how the legal concept of ‘responsibility’ for another subject might be understood or filled out. Before looking more closely, however, at what makes the defendant ‘responsible’ for another person and her actions, we may observe that there are situations where the law ascribes a person’s conduct and/or state of mind to the defendant in such a way as to make it legally his own. We tend to find this in the realm of corporate bodies which, though subjects of the law and normatively as real as any living person, have no physical existence and thus cannot act except through human agents. Many systems here resort to a metaphor and actually describe the appointed human agent as an ‘organ’ of the ‘juristic person’. A company director, for instance, represents the company in every respect when acting in this capacity. Any wrongful act is thus the company’s own, and will trigger the appropriate liability. The German BGB8 formulates the basic principle in connection with its law governing incorporated associations, which epitomise ‘juristic persons’9 more generally: § 31 BGB Liability of Association for Organs The association is liable for any loss or damage [Schaden] inflicted on a third party by its board, a member of the board or other constitutionally appointed representative, through 7 cf the similar provision in § 90a of the German BGB, introduced in 1990. 8 All English translations of provisions from the German BGB are the author’s own. The German word Schaden has been translated as ‘loss or damage’. 9 During the nineteenth century, there was a prominent dispute over the legal nature of a ‘juristic person’, more specifically about whether the juristic person existed only by virtue of a legal fiction and thus needed to have attributed to it the acts of natural persons functioning as its agents (so-called ­Vertretertheorie of Friedrich Carl von Savigny) or whether it had a ‘real corporate personality’ and was itself capable of acting through its representatives as ‘organs’ (so-called Organtheorie of Otto von Gierke). The BGB tried to avoid taking a stand on the issue, though § 31 BGB is often read as endorsing the ‘organ theory’.

146  Birke Häcker an [unlawful] act committed in the course of carrying out the business it is ­entitled and obliged to conduct and giving rise to a liability to compensate [the third party].10

In the Projet de réforme, a similar understanding is implied in article 1242-1, following straight after the delictual general clause in article 1241: Article 1241 A person is liable for the harm caused by his fault. Article 1242 A violation of a legislative requirement or a failure in the general duty of care or ­diligence constitutes a fault. Article 1242-1 [Fault in a legal person results from fault in one of its organs or a failure in its organisation or its functioning.]11

C.  Sources of a Person’s Responsibility for Others In the biblical tale of Cain and Abel, Cain asks God why he should be his brother’s keeper.12 There is indeed a question about what makes one person responsible for another, both in the sense of protecting the other from harm and in the sense of ensuring that no harm emanates from the subject.13 German criminal law theory has coined useful categories for thinking about this conundrum in the context of omissions liability (over and above that liability which potentially arises from the minimum solidarity which members of the public owe one another under German law).14 Where a person is responsible for guarding another’s ­well-being, he is described as a Beschützergarant (‘guarantor-protector’); where he is responsible for preventing another from causing harm to third parties, he is known as ­Überwachergarant (‘guarantor-monitor’). We are here only concerned with the

10 The German provision is formulated and commonly understood as an instance of what might be described in English as a form of ‘vicarious liability’, making the association liable where one of its organs has committed a delict (eg, under § 823 BGB) or otherwise incurred liability. § 31 BGB is said to be liability-allocating, not itself liability-generating: BGH 13 January 1987, BGHZ 99, 299, 302, NJW 1987, 1193. But this should not obscure the fact that the association itself is ultimately regarded as having committed the wrongful act through its organs or representatives. 11 It appears that this provision is in square brackets in the current draft because it is not yet clear whether it will actually feature in the final proposal. The emphasis has been added. 12 Genesis IV, 1–18. 13 In respect of the latter, it is worth noting that art 1253(1) of the Projet de réforme expressly ­recognises that an act of a third party (fait du tiers) may – and will often – amount to force majeure as far as the defendant is concerned. It concerns the case whether the defendant would otherwise be liable on some ground, but the intervening act of the third party has the consequence of exonerating him. 14 Contrast § 323c of the German Criminal Code (StGB), a statutory provision ‘intended for the protection of another’ under § 823(2) BGB, with what Lord Nicholls said in Stovin v Wise [1996] ­AC 923 (HL) 931, about the absence of a general legal duty positively to protect or save others under English law.

Fait d’autrui in Comparative Perspective  147 latter situation. An alternative classification seeks to uncover the origins of the guarantor’s responsibility. It distinguishes between a responsibility based on institutional grounds (Garantenstellung qua Institution) and one acquired by organisational acts (Garantenstellung qua Organisation). While parents are institutionally responsible for their children, simply by virtue of parenthood, employers will typically be responsible for their employees because they have chosen to take them on and have then deployed them in a particular way.15 The distinction is not one between children and employees, however. A babysitter who agrees to look after a child (eg, on the basis of a contract with the parents) is ‘organisationally’ responsible for it while the supervision arrangement lasts.16 We will see that these sources of responsibility are reflected in the English, French and German private law regimes, but in different ways and to varying degrees.

D.  Two Different ‘Models’ of Extra-Contractual Liability for the Actions of Others The fact that someone may have a particular (institutional or organisational) connection with another, making him responsible for ensuring that her actions do not become a source of harm to third parties, says little about the extent of this responsibility or the circumstances under which liability for her actions will be imposed on him. Liability may be strict or fault-based, and – if the latter – with or without a reversal in the burden of proof. Moreover, there are two fundamentally different ways of conceptualising the defendant’s liability, depending on whether the primary wrongdoer’s conduct is imputed on him ‘objectively’ or ‘subjectively’.17 On the ‘objective’ model, the primary wrongdoer incurs a liability (which may itself be fault-based or strict), and the defendant is burdened with it simply on account of his relationship with and responsibility for the wrongdoer. This form of liability could also be described as ‘secondary’. On the ‘subjective’ model, by contrast, the defendant’s liability is justified by his own (culpable) failure to prevent the person for whom he is responsible from harming the claimant. The defendant is thus liable as a second18 ‘primary’ wrongdoer. The difference between these two basic approaches can be illustrated by ­reference to the English case of Rose v Plenty.19 In the days when fresh milk was still commonly delivered to front doors, the first defendant (Mr Plenty) was employed by the second defendants as a roundsman. Contrary to his employers’ express

15 The two sources of responsibility may overlap, of course, and certain acts of organisation will generate a new institutional responsibility. 16 See, eg, § 832(2) BGB. 17 This is how the distinction is helpfully described by, eg, A Albrecht, Die deliktische Haftung für fremdes Verhalten im französischen und deutschen Recht (Tübingen, Mohr Siebeck, 2013) 2. 18 Assuming the immediate actor herself is also liable. 19 Rose v Plenty [1976] 1 WLR 141 (CA).

148  Birke Häcker instructions, Mr Plenty ‘sub-contracted’ a 13-year old boy, the claimant (Leslie Rose), to do much of the running between customers’ front doors and the delivery vehicle, the milk float. One day, while sitting on the float, the claimant was injured as a result of the first defendant’s negligent driving. It was quickly determined that Mr Plenty was liable in tort to compensate him. The question for the Court of Appeal was whether the second defendants, as Mr Plenty’s employers, were also liable on a ‘vicarious’ basis. The interesting point for present purposes is that the judges tackled the question in very different ways. Lord Denning MR, with whom Scarman LJ agreed,20 took the objective approach: In every case where it is sought to make a master liable for the conduct of his servant the first question is to see whether the servant was liable. If the answer is Yes, the second question is to see whether the employer must shoulder the servant’s liability.21

By contrast, Lawton LJ opted for the more ‘subjective’ route, looking for a tort committed by the second defendants. He asked ‘What duty did the employers owe to the boy?’22 and concluded that, because they expressly prohibited passengers from travelling on the float, they did not owe him any duty of care or – we may infer – that they had at any rate not breached it.23 In other words, Lawton LJ appeared to be interested in establishing whether or not the employers themselves had committed the tort of negligence vis-à-vis Leslie Rose.24

E.  Historical Background Apart from questions of principle, the historical forebears of liability for the acts of others can also help put into comparative focus our understanding of modern systems. Classical Roman law had no all-embracing concept of ‘vicarious’ liability, but allowed a paterfamilias (father of the family) to be sued for delicts committed by his slaves and by his children while they were still in his power. This liability was 20 ibid 147. 21 ibid 144, citing his own earlier judgment in Young v Edward Box & Co Ltd [1951] 1 TLR 789 (CA) 793. 22 Rose v Plenty (n 19) 146. 23 Unfortunately, Lawton LJ’s judgment is not entirely clear. Although he said (ibid) that ‘the injured boy has failed to establish that the employers owed him any duty of care’, this is prefaced by the following reasoning: ‘Had [Mr Plenty] obeyed his instructions, he would not have had a passenger to whom he owes a duty of care. It was his disobedience which brought the injured boy into the class of persons to whom the employers vicariously owed a duty of care’ (emphasis added). Tricky as it may be, the exact construction of the primary liability envisaged by Lawton LJ does not matter much for present purposes (though see the comment in n 24 immediately below). 24 Lawton LJ is widely understood as asking whether the employers had been negligent in a legally relevant way, but it is possible that he had in mind a version of the so-called ‘master’s tort theory’ (discussed below, in the text to nn 125–26), which leads to an employer’s primary liability on the basis of the employee’s actions being attributed to him as his own, in the manner outlined in section IIB above.

Fait d’autrui in Comparative Perspective  149 known as ‘noxal’, and it was strict in the sense that it did not depend on any fault on the father’s part. Its special feature was that, if the paterfamilias did not want to pay up, he had the option of surrendering the wrongdoer to the victim. The father’s ‘risk’ was therefore effectively confined to the wrongdoer’s economic or intrinsic value.25 Yet noxal liability ceased to apply to children in the post-classical age and later lost its remaining relevance when slavery fell into disuse.26 Outside the realm of slaves and children, we have a smattering of Roman texts suggesting that a defendant debtor was liable where someone charged with this task (whether a free man or a slave) failed to perform a contractual obligation owed by the defendant to the claimant.27 No corresponding texts seem to exist in the core law of delict, particularly as far as free servants were concerned.28 Things were different, however, for the praetorian so-called quasi-delicts.29 The occupier of a house was strictly liable if anything was poured or thrown from it,30 or if something was dangerously suspended from it,31 no matter by whom. And persons pursuing the reputationally ‘dubious’32 trades of sea carrier, inn- or stablekeeper were strictly liable for any theft or property damage committed on their premises, including by employees,33 as was a tax collector for delicts committed by members of his familia (understood in classical times as anyone engaged in the tax collector’s business).34 During the Middle Ages and the early modern period, the Roman quasidelictual regime became the anchor point for an argument (first formulated by Bartolus) that anyone who assigns a task to another should be held strictly liable for delicts committed by the other in fulfilling this task.35 It made its way into Roman-Dutch Law,36 via Pothier into the French Code civil,37 and through two 25 See R Zimmermann, The Law of Obligations: Roman Foundations of the Civilian Tradition (Oxford, Clarendon Press, 1996) 1118. 26 ibid 1119. 27 ibid 1120–21, pointing out that such contractual liability was nominally fault-based, though ­occasionally with an almost impossibly high standard of care being required. 28 T Rüfner, ‘Substance of Medieval Roman Law: The Development of Private Law’ in H Pihlajamäki, MD Dubber and M Godfrey (eds), The Oxford Handbook of European Legal History (Oxford, Oxford University Press, 2018) 309, 329. 29 For details and further references see H Wicke, Respondeat Superior (Berlin, Duncker & Humblot, 2000) 80–105. 30 So-called actio de effusis vel deiectis. 31 So-called actio de posito vel suspenso. 32 Rüfner (n 28) 329. 33 There is a question about the exact relationship between the quasi-delictual liability of the nauta, caupo or stabularius for theft and property damage with the (competing) actio de recepto based on their custodia. 34 Specifically on the praetorian liability of publicani and the way that differed from ordinary noxal liability (namely in that a noxal surrender was apparently excluded), see G MacCormack, ‘Culpa in Eligendo’ (1971) 18 RIDA 525, 551; Wicke (n 29) 101–5. 35 Rüfner (n 28) 329. 36 Wicke (n 29) 110–54. 37 See R-J Pothier, Traité des obligations, selon les règles tant du for de la conscience, que du for extérieur (Paris and Orléans, Debure and Rouzeau-Montaut, 1764) vol 1, 140–41 (Pt 1, ch 1, § 2, no 121) and 618–19 (Pt 2, ch 6, § 5, no 456).

150  Birke Häcker seminal judgments of Holt CJ even into English law.38 However, the respondeat superior principle, according to which a ‘master’ had to answer for the acts of his ‘servants’, did not come to prevail everywhere. Nineteenth-century German pandectist doctrine remained firmly wedded to the libertarian proposition that there could be no liability without fault,39 and this conviction has left its indelible mark on modern German law.40 Parental liability for children evolved along an entirely different route. Once noxal liability had disappeared, there was no obvious ‘peg’ for imposing liability without any fault on the part of the parent (short of likening children to animals, for which in Roman law liability was strict).41 Instead, the principle that ‘he who could have prevented harm is liable for not doing so’,42 coupled with the early modern recognition that a father’s power over his children (patria potestas) entailed responsibilities as much as privileges to educate and control them,43 led early codifiers to enact fault-based provisions.44 The French Code civil went a step further in 1804 and reversed the burden of proof by establishing a rebuttable presumption that the father as chef de la famille (head of the family) could have prevented the child’s harmful conduct.45 This model proved very influential and was also adopted by the draftsmen of the German BGB.46

III.  The French Regime47 As is inevitable with a piece of legislation over two centuries old, provisions become outdated and the law in action is no longer the same as the law in the ­(statute) books. Following the successful modernisation in 2016 of the law of contract, the general law of obligations and the law governing the proof of ­obligations,48 during which the French delictual provisions were all renumbered w ­ ithout

38 Boson v Sandford (1689/91?) 2 Salk 440, 91 ER 382; Turberville v Stampe, aka Tuberville v Stamp (1698) 1 Ld Raym 264, 91 ER 1072. For a detailed historical analysis, see Wicke (n 29) 155–215, esp 188–96. 39 See ch 7 above, pp 123 et seq, esp at 135–36. 40 See Zimmermann (n 25) 1124–26 and the discussion of § 831 BGB in section VB–C below. 41 So-called actio de pauperie, on which see Zimmermann (n 25) 1096–1104. 42 Digest 9.2.45.pr (Paul). 43 B Kannowski, ‘§§ 831–839a, 841. Unerlaubte Handlungen: Sondertatbestände’ in M Schmoeckel, J Rückert and R Zimmermann (eds), Historisch-kritischer Kommentar zum BGB, vol III/2 (Tübingen, Mohr-Siebeck, 2013) [22]. 44 See esp Pt 4, ch 16, § 6, no 8 of the Bavarian Codex Maximilianeus of 1756 as well as Pt 1, title 6, § 57 of the Prussian General State Law (Allgemeines Landrecht) of 1794. 45 Art 1384 paras 4 and 7 Cc as originally enacted. The mother, whose liability was originally confined to cases where the husband had died, was only fully equated with the father in 1970. 46 § 832(1) BGB, discussed further in section VA below. 47 I am extremely grateful to Professor Jean-Sébastien Borghetti for his helpful comments on an earlier draft of this section. All errors are my own. 48 On which see the contributions in J Cartwright and S Whittaker (eds), The Code Napoléon ­Rewritten – French Contract Law after the 2016 Reforms (Oxford, Hart Publishing, 2017).

Fait d’autrui in Comparative Perspective  151 being changed in substance,49 the reform effort has now turned to the regime governing (mainly extra-contractual) civil liability. As part of this, the provisions dealing with liability for actions of others are being redrafted and updated. We will look first at the ­existing (‘old’) law before assessing the new proposals.

A.  The Existing (‘Old’) Law The statement introducing what is now article 1242 Cc, quoted above,50 was originally not intended as a self-standing basis of liability, but merely as a kind of preface to the more specific provisions which follow it. However, already during the late nineteenth and early twentieth centuries, the industrial revolution with its accident-prone working conditions had led French courts to imbue paragraph 1 of (what was then) article 1384 Cc with life.51 On the basis of the idea that one is responsible for the harm caused by ‘things which one has in one’s keeping’, they imposed a strict liability on whoever was the keeper (gardien) of the object in question.52 In due course, this liability proved to be a ‘monster’ which ‘had to be tamed’.53 As regards harm ‘caused by the action of persons for whom one is responsible’, guidance was for a long time provided only by the provisions which follow the general statement in the first paragraph, namely those governing specifically the liability – – –

of parents for harm caused by their ‘minor children who live with them’, but only to the extent to which they exercise parental authority;54 of masters and employers for harm caused by their servants and employees within the functions for which the former have employed the latter;55 and of teachers and artisans for harm caused by their pupils and apprentices during the time which they are under their supervision.56

As already mentioned, the liability of parents was nominally subject to the proviso that they should be able to defend themselves by proving that they could not have prevented the harmful act, ie by disproving fault.57 The exculpation defence also 49 Arts 1382 et seq Cc thus became arts 1240 et seq Cc. 50 In the text following n 3. 51 But see J Knetsch in ch 7 above, pp 129–35, who argues that it is a ‘myth’ to regard responsibility without fault as a child of the industrial revolution. 52 See especially the Teffaine and Jand’heur cases referred to in n 5 above. There was less need to expand liability for harm done by animals as this was specifically catered for by art 1385 (now art 1243) Cc: see the text following n 7 above. 53 J Bell, S Boyron and S Whittaker, Principles of French Law (Oxford, Oxford University Press, 2008) 382 (S Whittaker). 54 Art 1242(4) (former art 1384(4)) Cc. 55 Art 1242(5) (former art 1384(5)) Cc. 56 Art 1242(6) (former art 1384(6)) Cc. 57 Art 1242(7) (former art 1384(7)) Cc.

152  Birke Häcker applied to artisans,58 with teachers being subject to the ordinary burden of proof.59 For masters and employers, by way of contrast, the original theory was that fault in the form of negligently choosing (culpa in eligendo) or failing properly to supervise (culpa in vigilando) the servant or employee could be ‘irrebuttably presumed’,60 so that no opportunity for exculpation arose.

(i)  Liability for the Actions of Others under Article 1242(1) (former Article 1384(1)) of the Code Civil) In 1991, the Cour de cassation came to decide a case where a mentally disabled person had set fire to a forest while in the care of a privately owned occupational therapy centre, which was now being sued by the owner of the wood for the resulting damage.61 Unable (or unwilling) to expand the former article 1384(4) Cc beyond parents and their children,62 but faced with Conseil d’État case law allowing similar claims against public institutions,63 the Assemblé plénière took the step of generalising liability for the actions of others, much as it had previously done for harm caused by things.64 It emerged from the arrêt Blieck and the case law following it that both organisations and individuals could be liable (without exculpation defence, but subject to force majeure)65 for harm emanating from a person over whom the defendant has control (so-called garde d’autrui).66 Typically such ‘control’ arises as the result of a judicial decision or administrative order allowing the defendant authoritatively to determine the controlled person’s ‘way of life’ (mode de vie).67 58 ibid. 59 Art 1242(8) (former art 1384(8)) Cc. See also comment in n 90 below. 60 This fiction was in due course jettisoned in favour of an open recognition of strict liability: see below, text following n 79. 61 Ass plén 29 March 1991, D 1991, 324 note Larroumet, JCP 1991, II, 21673 concl Dontenwille, note Ghestin (Association des centres éducatifs du Limousin et autre c Blieck). 62 See Cass civ (2) 24 November 1976, D 1977, 595 note Larroumet, 596. 63 See esp CE 3 February 1956, D 1956, 596 note Auby, JCP 1956, II, 9608 note Lévy (Thouzellier); CE 13 July 1967 (Département de la Moselle); CE 29 April 1987, JCP 1988, II, 20920 note Pacteau (Ministre de la justice c Banque populaire de la région économique de Strasbourg). 64 Interestingly, the idea already existed before 1991 for situations involving nuisance liability between ‘neighbours’ (troubles de voisinage), where landlords had long been held strictly liable for their tenants’ behaviour: see G Viney and P Jourdain, Traité de Droit Civil: Les conditions de la responsabilité, 3rd edn (Paris, LGDJ, 2006) [948] (at 1210). I am grateful to Professor Ciara Kennefick for drawing my attention to this. For a discussion of troubles de voisinage, see further ch 11 below, pp 223 et seq. 65 This is implied by Cass crim 26 March 1997, D 1997, 496 note Jourdain (Foyer Notre-Dame des Flots), a case decided soon after the Bertrand decision (see below, n 74 and text thereto) was handed down. In it, the Cour de cassation held that ‘individuals held liable for the action of other people within the meaning of article 1384(1) Cc, cannot escape the strict liability resulting from this provision by ­proving that they did not commit any fault’ (emphasis added). 66 For a fuller account in English of the development and its implications, see eg Bell, Boyron and Whittaker (n 53) 397–99 (S Whittaker); P Giliker, Vicarious Liability in Tort: A Comparative Perspective (Cambridge, Cambridge University Press, 2010) 133–37. 67 Initial concerns that liability might also be imposed on people who undertake to look after others by contract or as part of informal social arrangements between family, friends and neighbours (eg babysitters) have proved unfounded. See below, n 98 and the accompanying text.

Fait d’autrui in Comparative Perspective  153 Controversially, however, subsequent decisions have extended the Blieck principle to a different type of situation, in which the defendant does not control a vulnerable person’s ‘way of life’, but instead organises and controls what might be rationalised as an accident-prone group activity in which ordinary members of the public take part. Under this second line of cases, liability has been imposed on sports clubs for injuries brought about by their players during competitions,68 on hunting associations for shotgun accidents caused by their members (for a while at least),69 and even on the organisers of a majorette parade in the course of which a participant was hit by another’s twirling drum stick.70 Just as with liability for things, liability for the actions of others was seen by many as having become a ‘monster’ which needed to be – and to an extent has subsequently been – reined back in.71

(ii)  Expansion of Parental Liability for the Actions of their Children In light of this development, it is perhaps unsurprising that parental liability, too, should have witnessed a significant judicial expansion.72 A trio of cases – commencing with the arrêt Fullenwarth of 1984,73 via the arrêt Bertrand of 1997,74 and culminating in the Levert decision of 200175 – established that, in the words of a confirmatory decision by the Assemblée plénière handed down in 2002: [T]he strict legal responsibility [responsabilité de plein droit] incurred by the father and mother for the damage done by their underage child residing with them does not depend on the existence of any fault [faute] on the part of the child; … consequently, 68 Cass civ (2) 22 May 1995, Bull civ 1995 II no 155, JCP 1995, II, 22550 note Mouly; Cass civ (2) 3 February 2000, Bull civ 2000 II no 26, JCP 2000, II, 10316 note Mouly. 69 TGI Cusset 29 February 1996, JCP 1997, II, 22849 note Mouly; CA Dijon 5 September 2002, JCP 2003, IV, 1413. But this line of cases has subsequently been disapproved: Cass civ (2) 11 September 2008, Bull civ 2008 II no 192, JCP 2008, II, 10184 note Mouly. 70 Cass civ (2) 12 December 2002, Bull civ 2002 II no 289, JCP 2003, I, 154 no 49 obs Viney, RTD civ 2003, 305 obs Jourdain (so-called affaire des majorettes). 71 See esp Ass plén 29 June 2007, D 2007, 2455 note François, JCP 2007, II, 10150 note Marmayou. Cf also the hunting decision mentioned in n 69 above: Cass civ (2) 11 September 2008, Bull civ 2008 II no 192, JCP 2008, II, 10184 note Mouly. It now seems settled that only a qualified form of faute on the part of the member who has caused the damage (such as gross disregard for the rules of the sport or game being played) will lead to the organising association’s strict liability under art 1242(1) (former 1384(1)) Cc. 72 For a fuller account in English see Giliker (n 66) 213–17. 73 Ass plén 9 May 1984, Bull AP 1984 no 4 (Fullenwarth). The Assemblée plénière there decided that parents could be liable for the actions of their child regardless of whether the child had the requisite level of discernment to appreciate the wrongfulness of its conduct, because the parents’ ‘liability’ was to be ‘presumed’ whenever the child committed an act which directly caused the victim to suffer harm. The Fullenwarth appeal was consolidated with the equally well-known arrêt Gabillet, where the court held that even a three-year-old may become liable as the ‘keeper’ of a stick under art 1384(1) (now art 1242(1)) Cc: Ass plén 9 May 1984, Bull AP 1984 no 1. 74 Cass civ (2) 19 February 1997, Bull civ 1997 II no 56, JCP 1997, II, 22848 concl Kessous, note Viney. This case effectively abolished the exculpation defence. 75 Cass civ (2) 10 May 2001, Bull civ 2001 II no 96, JCP 2001, II, 10613 note Mouly, underlining the fact that no faute was required in the part of the child in order for the parents to be held liable (théorie du fait causal).

154  Birke Häcker the parents’ liability is engaged by any act committed by their minor child living with them which causes damage …76

What started as a delict based on a ‘presumption of fault’ and was for a time (after Fullenwarth) conceptualised as a ‘presumption of liability’ had thus turned into a full-blown strict liability wrong, subject only to force majeure or the victim’s own contributory fault. Connected with the shift to strict liability, the Cour de c­ assation has moreover started relaxing the criterion of cohabitation (that the minor must ‘live with’ his parents) quite significantly.77 For instance, parents cannot now escape liability for a 13-year-old daughter’s actions even where she has resided with her grandparents from the age of one.78

(iii) The Strict Liability of Employers for the Actions of their Employees As regards the liability of employers (technically: commettants) for harm caused by their employees (technically: préposés),79 the original fiction that the lack of any exculpation defence could be explained by an irrebuttable presumption of culpa in eligendo vel vigilando proved unsustainable. Instead, various theories have been advanced to justify the employer’s strict liability. One is the idea that the employer creates a risk to others by availing himself of the employee’s services, while at the same time exercising authority over the employee and profiting from her ­activities (théorie du risque-profit). Another regards the imposition of strict vicarious l­iability as a way of ensuring that victims harmed by the employee are not confined to suing the latter, but may have resort to a deep-pocketed defendant who effectively ‘guarantees’ the claim (théorie de la garantie). The latter, if applied strictly, would, of course, rule out in the employment context a Levert-type case,80 where the defendant is held liable although the primary ‘wrongdoer’ herself has not fulfilled all the requisite conditions of liability. And indeed, some sort of faute on the part of the préposé has always been seen as lying at the heart of the former article 1384(5) Cc.81 Nevertheless, the Cour de cassation in

76 Ass plén 13 December 2002, D 2003, 231 note Jourdain, JCP 2003, II, 10010 note Hervio-Lelong (Minc and Poullet). 77 As Giliker (n 66) 217 explains: ‘This requirement was relevant when liability rested on the question of whether the parent had properly supervised the child, but in view of the modern interpretation of “parental authority”, this appears to have minimal importance.’ 78 Cass crim 8 February 2005, JCP 2005, II, 10049 note Steinlé-Feuerbach. 79 There are interesting questions as to what constitutes the necessary relationship between ­commettant and préposé and how far the law goes (and should go) beyond the classical or typical employment context, eg, in connection with temporary workers. For a comparative discussion see Giliker (n 66) 55–143. 80 See above, n 75. 81 See, eg, Cass civ (2) 8 April 2004, Bull civ II no 194, D 2004, 2601 note Serinet. This case concerned an injury caused by a professional (hence employed) footballer during a match. The Cour de cassation required ‘a fault characterised by the violation of the rules of the game’ on the part of the player.

Fait d’autrui in Comparative Perspective  155 the year 2000 cast doubt on the ‘accessory’ nature of the employer’s liability by ­bestowing a special ‘immunity’ upon the employee in cases where she did not commit a voluntary intentional faute (such as a crime) and stayed within the limits of the task or ‘mission’ assigned to her by the commettant.82 It may be noted at this point that French courts handle the statutory criterion of employees acting ‘within the functions’ for which they are employed (so-called connexité) quite generously,83 so that an employer could well be liable under article 1242(5) Cc even where the employee exceeds her ‘mission’ by going beyond her mandate and/or where she commits a criminal act.84

B.  The Proposal The proposed reform of the Code civil consecrates much (though not all) of the jurisprudence that has developed around article 1384/1242 Cc. But rather than dealing with the ‘imputation of harm caused by another person’ under a single provision, the relevant rules now spread out over five articles (articles 1245–49 of the Projet de réforme). In respect of these provisions, there is hardly any difference between the version that was put out for consultation in 201685 and that contained in the 2017 final draft, with only a few small linguistic amendments86 and two slight shifts of substantive emphasis (to be discussed below).87 Heading the sub-section is article 1245 of the Projet de réforme, which might be described as a sort of ‘general part’ for the specific provisions which follow. Much in the manner of the original intention behind the former article 1384(1), but ­noticeably clearer in disclaiming any substantive content of its own, article 1245(1) states: A person is liable for harm caused by another person in the cases and subject to the conditions laid down by articles 1246 to 1249.

The proposal thereby impliedly rejects the idea that there might be a general ­principle of liability for actions of others outside certain (newly) defined relationships. Tellingly in this regard, articles 1245–49 of the Projet de réforme do not 82 Ass plén 25 February 2000, Bull AP 2000 no 2, D 2000, 673 note Brun, JCP 2000, II, 10295 concl Kessous, note Billiau, RTD civ 2000, 582 obs Jourdain (Costedoat). For an assessment see Giliker (n 66) 36–38; Bell, Boyron and Whittaker (n 53) 396–97. Giliker writes (at 37): ‘The 2000 Costedoat decision would appear to mark a firm step away from the perception of employer as mere guarantor towards the notion of employer as the bearer of the social risk of employment.’ 83 In this respect, French law is perhaps closer to English than German law. For a detailed comparative assessment, especially with a view to the ‘course of employment’ criterion traditionally invoked by English lawyers, see Giliker (n 66) 145–95. 84 In such a case, the employer and the employee would then both be liable (in solidum). Nothing but the clearest ‘abuse’ by the employee of her ‘functions’ (abus de fonctions) would appear to relieve the employer of liability: Ass plén 19 May 1988, D 1988, 513 note Larroumet. 85 Avant-projet de réforme du droit de la responsabilité civile (26 April 2016). 86 Such as to art 1249(1) of the Projet de réforme. 87 In the text surrounding nn 100 and 108 below.

156  Birke Häcker feature under the sub-section entitled ‘Action giving rise to extra-contractual liability’ (Le fait générateur de responsabilité extracontractuelle),88 but have been relegated to a sub-section of their own.89 In overview, the four specific provisions governing liability deal with: – the liability for minors of persons or bodies charged with their care (article 1246 of the Projet de réforme); – the liability of persons or bodies charged with the care of adults (article 1247 of the Projet de réforme); – the liability of persons who have voluntarily assumed such supervision ­(article 1248 of the Projet de réforme); and – the liability of employers for harm caused by their employees (article 1249 of the Projet de réforme). The proposal thus encompasses Blieck-type situations without, however, adopting a ­generalised notion of garde d’autrui. On the other hand, it has dropped any express reference to the liability of teachers for their pupils or artisans for their apprentices (without this entailing any substantive change).90 The archaic language of ‘master’ and ‘servant’ has similarly disappeared from article 1249 of the Projet de réforme, which now mentions only the employer (commettant) and his employee (préposé). Interestingly, a distinct linguistic feature separates articles 1246–48 from article 1249 of the Projet de réforme. While the latter – in line with the general provision in article 1245(1) of the Projet de réforme – refers to liability for the ‘harm caused’ (dommage causé)91 by an employee, articles 1246–48 instead speak of liability for the ‘action’ (fait) of a minor and other supervised person.

(i)  Analysis of Liability under Articles 1246–48 of the Projet de réforme In article 1246 of the Projet de réforme, the words ‘father’ and ‘mother’ have been replaced by the generic term ‘parents’, and there is no longer any cohabitation requirement.92 Parental authority alone suffices as the basis of liability. Any last vestiges of the exculpation defence have been removed. Equated with the parents (for liability purposes) are a child’s ‘guardian or guardians [son ou ses tuteurs], 88 This sub-section covers only arts 1241–44 of the Projet de réforme. 89 Concerning ‘The imputation of harm caused by another person’ (L’imputation du dommage causé par autrui). 90 Since a loi of 5 April 1937, the liability of teachers has been subject to proof of faute personelle on their part (see art 1242/1384(5) Cc), so that it is no longer, in essence, a liability for fait d’autrui and can be caught by the delictual general clause. An apprentice will today typically be trained under a contract of employment, so that the artisan’s liability will be that of a commettant for his préposé. 91 In art 1245(1) of the Projet de réforme, the proposal uses the singular (dommage), while the plural form (dommages) features in art 1249(1). 92 cf art 1356 of the Avant-projet Catala and art 14 of the Avant-projet Terré.

Fait d’autrui in Comparative Perspective  157 to the extent to which they are charged with care of the minor’s person’ and ‘a ­physical or legal person charged by judicial or administrative decision with organising and controlling the minor’s way of life on a permanent basis’. In the latter case, the parents cease to be liable for the minor’s actions,93 presumably because they no longer have parental authority enabling them to influence or control the minor. A significant feature of the proposal is the fact that it appears to reverse or at any rate realign the case law culminating in the arrêt Levert.94 By virtue of the general pronouncement in article 1245(2) of the Projet de réforme, a person’s accountability for the harm caused by another always ‘rests on proof of an action of a nature to engage liability in the direct author of the harm’. In Levert and subsequent cases like it, parental liability could arise even though the child itself behaved in a way which displayed not merely age-appropriate discernment, but actually satisfied the ordinary level of care expected of a reasonable person. In other words, under the current article 1242(4) Cc, neither subjective blameworthiness (imputabilité)95 nor an objective faute on the part of the child are required to trigger her parent’s responsibility. Parental liability is based on a purely causal theory (théorie du fait causal). Under the proposal, by contrast, the mere fact that the child’s action is a ‘direct cause’ of the harm suffered by the victim is not enough. Instead, the child will have to have acted in a manner that justifies an objective96 finding of faute. Article 1247 mirrors the third variant of article 1246 of the Projet de réforme (concerning a ‘physical or legal person charged by judicial or administrative decision with organising and controlling the minor’s way of life on a permanent basis’), except that the person whose life is being controlled and organised by the defendant on a permanent basis is an adult rather than a minor. Since the primary wrongdoer is of full age, parenthood or guardianship do not feature in this ­provision. The supervisory function envisaged by article 1247 of the Projet de réforme can only be based on a judicial or administrative decision. By contrast, where a person has voluntarily undertaken the supervision by private contract and97 as part of a professional activity, whether in respect of a minor or an adult, liability is determined by article 1248 of the Projet de réforme. This provision confirms that babysitting relatives or neighbours will not be held

93 Contrary to art 1356 of the Avant-projet Catala, which preferred a model of cumulative liability. 94 See above, nn 73–76 and the accompanying text. 95 It is nowadays settled that a child can commit a faute even in the absence of subjective discernement: Ass plén 9 May 1984, Bull AP 1984 no 2 (Lemaire), a case which was consolidated with the arrêt ­Fullenwarth and the arrêt Gabillet (both n 73). 96 The proposal does not seek to reintroduce the more subjective notion of faute which pertained prior to Ass plén 9 May 1984, Bull AP 1984 nos 1–4 (cf n 95). 97 The wording of the provision is not entirely clear in this respect. There appears to be a comma missing, or there may alternatively be one comma too many. The provision as drafted in 2017 reads: ‘Les autres personnes qui, par contrat assument, à titre professionnel, une mission de surveillance d’autrui ou d’organisation et de contrôle de l’activité d’autrui, répondent du fait de la personne physique surveillée à moins qu’elles ne démontrent qu’elles n’ont pas commis de faute.’

158  Birke Häcker accountable for a child’s actions,98 since they do not act professionally. However, at the same time it marks a distinct break with the jurisprudence of the Cour de cassation which has so far rejected recognising a delictual claim under former ­article 1384(1) Cc where a professional defendant undertakes someone’s (an adult’s or a child’s) supervision by contract rather than on the basis of some judicial or administrative decision.99 Two further features are remarkable about article 1248 of the Projet de réforme. One is that while the Avant-projet of 2016 only spoke of the ‘supervision of another person’ being undertaken, the 2017 draft added the words ‘or the organisation and control of the activity of another person’ (yet apparently without making the relevant follow-on adjustment to the second part of the sentence).100 It looks as though this addition consecrates the line of cases that allows sports clubs and the like to be held liable for injuries caused by their members during accident-prone activities,101 subject only to the caveat that the defendant must have assumed responsibility for the activity’s organisation and control in a ‘professional’ capacity (à titre professionnel).102 The last notable feature concerning article 1248 of the Projet de réforme is that it still contains an exculpation defence. Defendants ‘are liable for the action of the physical person supervised unless they show that they did not commit any fault’ (emphasis added). We can only speculate about what may lie behind this adherence to the old model of former article 1384(7) Cc. Perhaps the exculpation defence, which first appeared in the 2016 Avant-projet, is intended to counter-balance the expansion of liability to cases of professional supervision undertaken by contract.103 ­Superficially, however, the only difference between articles 1246–47 and 1248 of the Projet de réforme seems to be the source of the defendant’s responsibility for the primary wrongdoer. This could be described as ‘institutional’ in the former situations and ‘organisational’ in the latter case.104 But should a b ­ usiness that undertakes to control and supervise a person really be better off than ­someone on whom the same responsibility has been imposed by the state? P ­ rofessional defendants will typically act for a fee, not gratuitously. All in all, one gets the impression that the section (particularly after its expansion in the 2017 draft to ‘activities’ under the

98 The existing case law has rejected liability, for instance, in Cass civ (2) 18 September 1996, Bull civ 1996 II no 217; Cass civ 5 February 2004, Bull civ 2004 II no 50. 99 See esp Cass crim 18 May 2004, Bull crim 2004 no 123; Cass civ 24 May 2006, Bull civ 2006 II no 136; Cass civ (1) 15 December 2011, Bull civ 2011 I no 220, D 2012, 539 note Develay, JCP 2012, 205 note Bakouche. As far as children were concerned, liability under art 1384/1242(1) and (4) Cc was regarded as being mutually exclusive. 100 Which still refers only to a liability for the ‘action of the physical person supervised’ (emphasis added). 101 See above, nn 68–71 and the accompanying text. 102 As in the past (see n 81), it seems that a line is to be drawn between ordinary lay members and employed sporting professionals, with liability for injuries caused by the former being governed by art 1248 of the Projet de réforme and as regards the latter by art 1249 of the Projet. 103 See above, text accompanying n 99. 104 See above, section IIC, text following n 12.

Fait d’autrui in Comparative Perspective  159 defendant’s control) may be trying to do too much by covering everything from life in a private nursing home to the organisation of a football game during a paid-for holiday camp.

(ii) Analysis of Liability under Article 1249 of the Projet de réforme Article 1249 of the Projet de réforme provides in paragraph 1 that ‘[a]n employer is strictly liable for harm caused by his employee’ and goes on to define the employer as ‘a person who has the power to give orders and instructions to his employee in relation to the performance of [her] functions’. The control element thus retains its central importance to the notions of commettant and préposé, with the exact nature of the parties’ legal relationship (eg, an employment contract, contrat de travail) being secondary to the commettant’s authority to give orders and the factual degree of subordination by the préposé. What matters is that there should be between the parties a bond referred to as lien de préposition (roughly translatable as ‘relationship of employment’ in the non-technical sense of ‘employment’, encompassing all manner of deployment situations).105 Paragraph 2 deals with the case where the relationship of employment is ­‘transferred’ from one commettant to another (transfert de lien du préposition), such that the new commettant acquires control over the préposé. In that situation, the ‘liability is borne by the beneficiary of the transfer’, ie the new commettant.106 Since this would appear to be self-evident for acts committed after the transfer,107 it is arguable that the provision must be understood to be (also) aimed at cases where the relevant liability has already arisen prior to the transfer – in which case it too is ‘transferred’ – or where the harm has not yet been caused, or has yet to emerge, even though the underlying fait générateur lies before the date of the transfer. The remaining two subsections of article 1249 of the Projet de réforme merit closer consideration. Paragraph 3 reads: An employer or a beneficiary of such a transfer is not liable if he proves that the employee acted outside the functions for which [she] was employed, without authorisation and for purposes alien to [her] attributions. Nor is he liable if he establishes collusion between his employee and the victim.

What is formulated by the current article 1242(5) Cc as a precondition of ­liability (namely that the culprit employees should have acted ‘within the functions’ for which  they were employed), is turned by the proposal into a defence, hence an ‘escape  route’ from liability. The burden of proof lies flatly on the 105 There is, however, (pre-Blieck) case law which suggests that this bond need not be a particularly close or long-lasting one. It can even be a so-called préposition occasionnelle: see Cass civ (2) 27 November 1991, RCA 1992 no 41. A question therefore arises as to the proper demarcation line between art 1249 and art 1248 of the Projet de réforme, in so far as the latter governs the defendant’s organisation and control of other people’s activities (see the text accompanying nn 100–02 above). 106 Already suggested by arts 17–18 of the Avant-projet Terré. 107 In fact, however, the existing case law on the matter is not as clear as it ought to be.

160  Birke Häcker affected ­commettant. Otherwise the criteria – ‘lack of authorisation’ (sans autorisation) and ‘purposes alien to [the employee’s] attributions’ (fins étrangères à ses ­attributions) – reflect those in the existing case law. In respect of the second sentence, the proposal deviates from the 2016 Avant-projet. Rather than excluding liability only in cases of ‘collusion’,108 the Avant-projet had spoken more broadly of a victim who ‘could not reasonably believe that the employee was acting on behalf of the employer’. Much litigation will probably be avoided by narrowing this to demonstrable active collusion. Paragraph 4 turns from the employer to the employee and codifies the ­‘immunity’ afforded to employees through the line of cases surrounding the arrêt Costedoat.109 It provides: An employee is not subject to any personal liability except in the case of intentional fault, or where without authorisation he acted for purposes alien to his attributions.

This impliedly acknowledges and confirms (in line with article 1245(2) of the Projet de réforme) that what is in principle required is not merely a fait causal, but a fait fautif – namely ‘an action of a nature to engage liability in the direct author of the harm’. In the very same instance, however, the employee is completely shielded from liability, provided only that she acted within the limits of her authorisation and/or the assigned task, and without committing an intentional wrong. It would be fair to say that, in most cases, the only purpose of establishing a fait fautif on the part of the employee is to impose liability on the employer. The justification for the employee’s immunity in paragraph 4 lies in the (typical) background employment context. Most legal systems today recognise that the risks and costs connected with an employer’s activities ought primarily to be borne by him, even if they materialise through the negligent conduct of an employee – think only of a truck driver whose slight carelessness causes an accident damag­ ing the valuable lorry and injuring passing pedestrians. There are therefore widespread restrictions on the circumstances under which an employer can have internal recourse to the employee for losses negligently caused, and sometimes even rules allowing employees who have been sued personally to seek an indemnity from the employer where appropriate.110 In granting the employee immunity 108 Cf art 1253(1) of the Projet de réforme, which includes amongst the general grounds exonerating a defendant from civil liability situations in which the victim has brought the harm upon himself in a manner that amounted to force majeure from the defendant’s perspective. 109 See above, n 82 and the accompanying text. 110 In German law, these are generally referred to as principles of ‘internal loss reallocation’ ­(innerbetrieblicher Schadensausgleich): see esp BAG (GS), 27 September 1994, BAGE 78, 56, NJW 1995, 210. Under French law, there appears to be less of a tradition of ‘internal’ fine-tuning (see, eg, the rejection of an indemnity claim by the employee against the employer in CA Paris, 20 October 1934, Gaz Pal 1934, 2, 895), but cf Cass soc 25 October 2005, Bull 2005 V no 299 (restricting the employer’s recourse against the employee to faute grave) as well as art L 121-12 Code des assurances. In England, the employer’s recourse against his employee, controversially accepted in Lister v Romford Ice and Cold Storage Co [1957] AC 555 (HL), has been severely limited – praeter legem – through a ‘gentlemen’s agreement’ between insurance companies: Giliker (n 66) 32–33; O Kahn-Freund, ‘Comparative Law as an Academic Subject’ (1966) 82 LQR 40, 57–58.

Fait d’autrui in Comparative Perspective  161 from suit vis-à-vis third parties, French law gives ‘external’ effect to what other systems can achieve only through internal reallocation mechanisms. Admittedly, however, this simplification comes at the price of occasionally leaving a third-party victim unable to recover from either the employee or the employer (namely where the latter is uninsured and insolvent).111 Looking at paragraphs 3 and 4 side by side, the question arises as to how much scope there is left for cases where both employer and employee are liable to the victim under article 1249 of the Projet de réforme.112 The terminology used in ­paragraph 3 to define when an employer escapes liability by showing that the employee has ‘acted outside the functions for which [she] was employed, without authorisation and for purposes alien to [her] attributions’ is almost a complete mirror image of that used in paragraph 4 to outline the conditions of the employee’s own accountability towards the victim – ‘where without authorisation [she] acted for purposes alien to [her] attributions’. The question is how much turns on the slight variation of words, especially whether paragraph 3 establishes three cumulative criteria (as suggested by previous case law)113 or whether the term ‘without authorisation and for purposes alien to [her] attributions’ merely explains when an employee exceeds her ‘functions’. Continuity would point towards the former interpretation, but if the provision came to be read in the latter sense, then mutual exclusivity beckons. It would mean that either the employer escapes liability because the employee has acted ‘outside [her] functions’ or the employee is immune because she has stayed within them. The only case where the liability of both employer and employee would then be engaged would be that of the employee acting within her ­functions, but intentionally.

IV.  The English Regime Under English law, there is a big conceptual division between, on the one hand, situations where an employer is held accountable for wrongful acts committed by an employee in the course of her employment and, on the other hand, ­situations where the liability is imposed for someone’s failure to supervise children (or indeed adults in appropriate circumstances). As regards the former, liability is ‘strict’ in the sense of not depending on any fault on the employer’s part. In the latter case, by contrast, all turns on whether the ordinary preconditions for liability in the tort of negligence are satisfied.114

111 This is why art 1359-1 of the Avant-project Catala had envisaged a subsidiary liability of the employee. 112 cf n 84 and the accompanying text. 113 See especially the seminal decision of Ass plén 19 May 1988, D 1988, 513 note Larroumet. 114 This chapter will not address other potentially relevant torts, such as private nuisance (roughly equivalent to troubles de voisinage under French law), but cf the comment in n 64 above.

162  Birke Häcker

A.  Vicarious Liability of Employers for their Employees The liability of an employer for wrongful acts committed by his employees115 is in English law described as ‘vicarious’. Just as under the original formulation in the Code civil, the language used was long that of ‘master’ and ‘servant’, thus ­emphasising the element of authority and control in the defendant’s relationship with the primary wrongdoer. According to the (now) dominant understanding,116 vicarious liability is a form of secondary liability and operates along the lines suggested by the majority in Rose v Plenty.117 It is first asked whether the ‘servant’ has committed a civil wrong against the claimant (typically a common law tort) and, if so, then in a second step it is determined whether the ‘master’ should be held liable for it in addition to the ‘servant’. This depends on whether or not the latter acted ‘within the scope’ or ‘in the course of [her] employment’ – a criterion which has been significantly ­broadened as well as fundamentally reconceptualised over the past two decades. Today, courts no longer ask (as they did in the past) whether the servant’s conduct amounted to ‘either (1) a wrongful act authorised by the master, or (2) a wrongful and unauthorised mode of doing some act authorised by the master’.118 Instead, they consider the degree and ‘closeness’ of the connection between the employee’s tort and what she was employed to do,119 and they increasingly decide the matter more openly by reference to whether it would be ‘fair and just’ to hold the employer liable.120 The traditional view121 that it is insufficient for the employer to have provided the employee with the ‘mere opportunity’ to commit a tort has been abandoned. In the recent case of Mohamud v Wm Morrison Supermarkets plc, where the employee of a petrol station run by the defendants had inflicted racist abuse and a violent physical attack on the claimant customer, Lord Toulson said: In the simplest terms, the court has to consider two matters. The first question is what functions or “field of activities” have been entrusted by the employer to the employee, or, in everyday language, what was the nature of his job. As has been emphasised in several cases, this question must be addressed broadly … Secondly, the court must decide whether there was sufficient connection between the position in which he was employed and his wrongful conduct to make it right for the employer to be held liable under the principle of social justice which goes back to Holt CJ. To try to measure the closeness of connection, as it were, on a scale of 1 to 10, would be a forlorn exercise and, what is more, it would miss the point.122 115 Lack of space prevents this chapter discussing a person’s (exceptional) liability for the conduct of independent contractors based on so-called ‘non-delegable’ duties of care. 116 See below, text accompanying nn 129–30. 117 See above, nn 20–21 and the accompanying text. 118 This is the so-called ‘Salmond test’, which was for a long time seen as the decisive touchstone of liability. See J Salmond, The Law of Torts (London, Stevens & Haynes, 1907) 83. 119 See especially Lister v Hesley Hall Ltd [2001] UKHL 22, [2002] 1 AC 215. 120 See especially Mohamud v Wm Morrison Supermarkets plc [2016] UKSC 11, [2016] AC 677. 121 Still adhered to in Lister v Hesley Hall Ltd (n 119) [28]. 122 Mohamud v Wm Morrison Supermarkets plc (n 120) [44]–[45].

Fait d’autrui in Comparative Perspective  163 In Mohamud, the Supreme Court held the employers liable. Their Lordships did ‘not consider that it is right to regard [the employee] as having metaphorically taken off his uniform the moment he stepped from behind the counter’.123 The comparative lawyer cannot help but wonder whether or not this case is one where article 1249(3) of the French Projet de réforme would have enabled the employers to prove that ‘the employee acted outside the functions for which he was employed, without authorisation and for purposes alien to his attributions’. In conceptual terms, the ‘secondary liability’ account of the English vicarious liability regime is by no means uncontroversial. The above-mentioned minority judgment in Rose v Plenty124 illustrates that there are people who regard it as a form of primary liability. However, their reasoning does not tend usually or necessarily to turn on finding any actual personal fault on the part of the employer. Instead, the so-called ‘master’s tort’ theory maintains that the tort committed is the ‘master’s’ own because the employee’s actions and attendant state of mind are attributed to him.125 It is thus neither a case of the employer shouldering the employee’s liability, nor of his having failed carefully to select, instruct, or supervise the employee, but of the employer being primarily liable because the nature of the employment relationship warrants applying the maxim that ‘he who acts through another, acts himself ’ (qui facit per alium facit per se). If, therefore, the employee negligently injures someone while driving a car in the course of her employment, the employer will also be deemed to have driven negligently. The idea is essentially the same as that underlying the attribution to corporations of the conduct of their ‘organs’.126 Although there is something to be said for the master’s tort theory (eg, that it has no problem accommodating an immunity for the employee and nevertheless holding the employer liable),127 it has fallen out of fashion.128 ­Advocates of the ‘master’s tort theory’ are today few and far between,129 and the judiciary appears to have long settled on the ‘secondary liability’ model.130

123 ibid [47]. 124 See above, nn 22–24 and the accompanying text. 125 See esp G Williams, ‘Vicarious Liability: Tort of the Master or of the Servant?’ (1956) 72 LQR 522, and from the older case law, Twine v Bean’s Express Ltd [1946] 1 All ER 202 (KB) (distinguished in Rose v Plenty (n 19)), as well as Broom v Morgan [1953] 1 QB 597 (CA). Interestingly, it was none other than Denning LJ who said in the latter case (at 609) that ‘the master’s liability for the negligence of his servant is not a vicarious liability but a liability of the master himself owing to his failure to have seen that his work was properly and carefully done.’ 126 R Stevens, Torts and Rights (Oxford, Oxford University Press, 2007) 259–60, and see above, section IIB of the present chapter. In Twine v Bean’s Express (n 125) 204, Uthwatt J observed that ‘[t]he law attributes to the employer the acts of a servant done in the course of his employment and fastens upon him responsibility for those acts’, adding that ‘the law does not attribute to the employer the liability which attaches to the servant’ (emphases added). 127 Such an immunity issue arose in Broom v Morgan (n 125); cf the envisaged regime of art 1249(3)–(4) of the Projet de réforme. 128 See the criticism of it by Giliker (n 66) 15. 129 But see Stevens (n 126) 244–45, 257–74. 130 See esp Staveley Iron and Chemical Co v Jones [1956] AC 627 (HL).

164  Birke Häcker

B.  Parental Liability for Failure to Supervise Children (and Similar Cases) By contrast with civilian systems, which tend either to hold parents strictly liable for the wrongful acts of their children (as modern French law does) or at least to subject them to a reversed burden of proof (as in the original conception of the Code civil), English law fits parental accountability into the ordinary framework of its tort of negligence. This means that parents will only be liable where they themselves have breached a duty of care towards the victim. But though people are not usually under an obligation to prevent others from harming one another under English law,131 the parents’ relationship with their child and the control they are able to exercise over the child impose on them a responsibility to take reasonable care in the child’s supervision. On this basis, a liability in the tort of negligence will arise when parents fail to take the necessary steps to supervise or restrain their child, resulting in the child inflicting loss or injury on a third party (whether or not the child herself acted culpably). The required standard of care varies with all the circumstances. Where the child’s conduct constitutes a known source of danger, the standard will be appropriately higher; the older and more self-reliant the child becomes, the less will be required of the parents.132 On the whole, English courts have shown great sympathy for the plight of overstretched parents and the degree of control they can r­ ealistically (and desirably) be expected to achieve.133 When assessing the English position, it is worth bearing in mind that a victim will in many cases have a claim against the child itself (whatever such a claim may be worth) since there is no minimum age below which a minor cannot be liable in tort. So long as the little delinquent is capable of controlling her body and appreciates the potentially deleterious impact of her actions on others,134 and – where negligence is at issue – subject to the proviso that the standard of care required will be age-appropriately lower than for adults,135 she will have to compensate the victim as any other tortfeasor would. Nevertheless, the limited recourse to parental pockets has prompted suggestions that English law should think about adopting a vicarious liability regime, either by statute (as most US states

131 See above, n 14 and the accompanying text. 132 For examples, see Giliker (n 66) 200–02. 133 In the recent case of Ellis v Kelly [2018] EWHC 2031 (QB), [2018] 4 WLR 124, for instance, a car driver’s contribution claim against the mother of an injured child failed because (at [71]) ‘holding [her] responsible would be to impose far too high a standard on an ordinary parent making ordinary decisions in the course of parenting as to how to keep her child reasonably safe while gradually being allowed more responsibilities and freedoms’. 134 As R Bagshaw, ‘Children Through Tort’ in J Fionda (ed), Legal Concepts of Childhood (Oxford, Hart Publishing, 2001) 127, 128, points out, there is no fixed age limit by reference to which courts treat children as being ‘of tender years’ and thus incapable of having the requisite tortious state of mind, with some cases suggesting that even four-year-olds can be held liable. 135 See Mullin v Richards [1998] 1 WLR 1304 (CA).

Fait d’autrui in Comparative Perspective  165 have done)136 or through a sensitive judicial extension of the existing doctrine developed in the employment context.137 Because English law as it stands applies its ordinary tort rules to parental liability for children, it has not encountered any conceptual problems in going beyond the parent–child relationship and holding other persons with super­ visory functions and powers liable for negligence on the same model. Thus, foster parents and teachers are also under a duty to take reasonable steps to ensure that their ward or pupil does not cause loss or injury to third parties,138 and it has long been settled that officers of a borstal training centre assume a corresponding responsibility towards neighbouring property owners when they take the boys out to work.139 Indeed, such has been the expansion of negligence liability on the one hand and vicarious liability on the other that the two have begun to overlap. While in the borstal case the supervising officers’ liability in negligence was seen as a precondition of the Home Office’s vicarious liability for their failure to prevent the boys escaping, the UK Supreme Court in the 2016 case of Cox v Ministry of Justice simply treated a prisoner working in the prison kitchen as being ‘akin’ to an employee of the prison service, so that vicarious liability for his actions could be imposed directly on the authorities.140

V.  The German Regime German law stands, in a sense, at the opposite end of the spectrum when compared with French and English law. Rather than imposing strict liability on both employers and parents (as modern French law and the Projet de réforme do), or just on employers (as in English law), delictual liability under the BGB remains wedded to the defendant’s fault in both situations. By and large, the text of §§ 823–53 is still that drafted by the ‘fathers of the BGB’ at the end of the nineteenth century in the spirit of their time. Although social conditions have changed significantly in the interim, it is hard to envisage German courts departing quite as openly from the wording of the Code as the French Cour de cassation did, for instance, in the 136 The relevant ‘parental responsibility laws’ often impose liability on parents on the basis of their child’s ‘wilful misconduct’ and typically include a maximum cap. See, eg, § 1714.1 of the Californian Civil Code; § 3-112 of the New York General Obligations Law; ch 231, s 85G of the Massachusetts General Laws. 137 Giliker (n 66) esp 196–97, 203–05, 223–26. 138 See, eg, Carmarthenshire CC v Lewis [1955] AC 549 (HL). 139 Home Office v Dorset Yacht Co Ltd [1970] AC 1004 (HL). 140 Cox v Ministry of Justice [2016] UKSC 10, [2016] 2 WLR 806 (hereinafter ‘Cox’). Since the prisoner’s wrongful conduct consisted in negligently dropping a bag of rice onto another member of the catering team, there was no problem about his having acted ‘in the course of ’ the activity that was ‘akin’ to employment. However, coupled with the Mohamud decision (discussed above, in the text to nn 120–22), it seems just about conceivable that the Dorset Yacht case, too, would today be decided as a direct case of vicarious liability (without the need to find any of the borstal officers negligent as an interim step).

166  Birke Häcker Bertrand ­decision.141 ­Nevertheless, we will see that judges have moulded and adjusted German private law in a way that avoids the full rigours of the fault-based regime as far as the employment context is concerned.

A.  Parental Responsibility for the Actions of their Children (and Similar Cases) § 832 BGB provides generally for the ‘liability of a person under a duty to supervise’,142 hence not only as regards parents and children: (1) A person who is, by operation of law, under a duty to supervise someone needing supervision on account of being a minor or on account of their mental or physical constitution, is liable to compensate for the loss or damage [Schaden] which the supervised person unlawfully inflicts on a third party. This liability to compensate does not arise if the supervisor has sufficiently discharged the duty to supervise or if the loss would have occurred despite all proper supervision. (2) The same responsibility is to be borne by a person who undertakes the supervision by contract.

It will be evident that the provision applies to both ‘institutional’ and ­‘organisational’ supervision duties,143 and that it does not distinguish between them as regards the fault requirement. In both cases, the exculpation defence is available to the defendant. The language used in the second sentence of § 832(1) BGB signifies to the initiated reader that it is for the ‘supervisor’ to prove a lack of causation or a sufficient degree of supervision and control. What the duty to supervise requires in a given situation is to be determined in light of all the circumstances of the case. Interestingly, to the extent that there have been calls for a reform of § 832 BGB in the parent–child context, these are not uniformly in favour of making ­liability stricter.144 Some authors worry that the reversed burden of proof undermines the privileged position which the German Constitution affords to families.145 They argue not merely that the ordinary burden of proof should be restored, but that parental liability should be confined to cases where parents have acted i­ ntentionally or with gross negligence,146 or at any rate that it should be excluded for culpa levis 141 See above, n 74 and the accompanying text. 142 For a detailed account in English of the German case law and literature relating to this provision, see G Wagner, ‘Children as Tortfeasors under German Law’ in M Martín-Casals (ed), Children in Tort Law. Part I: Children as Tortfeasors (Wien, Springer, 2006) 217, 235–63. 143 Parts of the literature insist on a contractual (rather than a merely social) undertaking, probably so as to protect neighbours, friends and family against liability in the context of occasional factual supervision: see G Spindler in HG Bamberger, H Roth and others (eds), BGB, Band 3: §§ 705–1117, [commentary] 4th edn (München, CH Beck, 2019) (hereinafter ‘Bamberger/Roth’) § 832, [13]. 144 An overview of the state of debate is provided by Albrecht (n 17) 146–53. 145 Art 6(1) of the Basic Law (Grundgesetz). See, eg, F Bernau, Die Aufsichtshaftung der Eltern nach § 832 BGB – im Wandel! (Berlin, Duncker & Humblot, 2005) 402. 146 B Großfeld and B Mund, ‘Die Haftung der Eltern nach § 832 I BGB’ [1994] FamRZ 1504.

Fait d’autrui in Comparative Perspective  167 and levissima.147 If enacted, this suggestion would take German law even beyond the current English position with its pragmatic insistence that p ­ arental n ­ egligence liability follows the ordinary rules. However, there are also authors who would have German law align more closely with the modern French regime and who want to make liability strict (albeit coupled with a compulsory system of thirdparty insurance).148 It will have been noted that § 832(1) BGB speaks of the supervised person ‘unlawfully’ inflicting loss or injury on the third party. This language is significant, for it shows that the primary wrongdoer herself need not be liable for her actions. Unlawfulness does not necessarily indicate individual culpability or accountability. Under German law, children under seven years are not ‘responsible’ (verantwortlich) – and ­therefore are not liable – for the losses they cause.149 In the case of traffic accidents, the threshold has been raised to 10 years, except where the child acted deliberately so as to injure.150 For minors over seven (or 10) years of age, their liability depends on whether they had ‘the insight and ability to appreciate their responsibility’ when committing the harm and wrongful act.151 German law, t­ herefore, has no difficulty holding parents liable where their children are not. In fact, this is a very common scenario. Moreover, given that the scope of the ­parental duty to supervise diminishes with the child’s increasing age and maturity, it is likely that parents will be able to exculpate themselves in many cases where the minor is held personally accountable for her deeds.152

B.  Primary Delictual Liability of Employers for Wrongs Committed by their Verrichtungsgehilfen By contrast with § 832 BGB, the provision which precedes it and deals with the ­liability of employers or principals (technically, Geschäftsherren, hereinafter referred

147 Bernau (n 145) 394–96, specifically for cases where the child and her parents are cohabiting (since this is the only type of relationship protected by art 6(1) of the Basic Law). 148 See, eg, E von Hippel, ‘Ruinöse Haftung von Eltern und Minderjährigen?’ [2001] FamRZ 748; F Niboyet, Die Haftung Minderjähriger und ihrer Eltern nach deutschem und französischem Deliktsrecht zwischen Dogmatik und Rechtspolitik (Berlin, Duncker & Humblot, 2001) 192–96. On the insurance aspect see below, section VI of the present chapter. 149 § 828(1) BGB. But cf § 829 BGB, which provides that the minor may nevertheless have to pay compensation ‘for reasons of equity’ (aus Billigskeitsgründen) where the victim cannot obtain redress from a supervisor (such as the parents) and where, on balance, the loss is more appropriately borne by the minor. 150 § 828(2) BGB. 151 § 828(3) BGB. The formulation indicates that the defendant minor bears the burden of proving that she lacked this insight and ability. 152 Inspired by Dutch law, some authors have even argued for a system where parents would be strictly liable up to the child’s 14th birthday (the child itself being immune), with a reversed burden of proof for damage done by minors aged 14–15 and the ordinary onus for minors aged 16–17: G Wagner in Münchener Kommentar zum Bürgerlichen Gesetzbuch, Band 5: §§ 705–831, 7th edn (München, CH Beck, 2017) (hereinafter ‘Münchener Kommentar’) § 828, [22].

168  Birke Häcker to as ‘employers’) for the acts of their employees (technically, V ­ errichtungsgehilfen) is easily one of the most contentious in the entire German Civil Code. Headed ‘liability for employees’, § 831 BGB reads: (1) A person who employs another to perform a task is obliged to provide compensation for the loss or damage [Schaden] which the other unlawfully inflicts on a third party in the execution of this task. The liability to compensate does not arise if the principal has exercised all due care in selecting the person employed to perform the task and, to the extent that he has to provide apparatus and equipment or that he has to supervise the execution of the task, in making the provision or ­supervision, or if the loss or damage would also have occurred despite the exercise of all due care. (2) The same responsibility is to be borne by a person who, by contract, undertakes to perform on the principal’s behalf one of the functions mentioned in the second sentence of sub-section 1 [ie, selection, provision, supervision, instruction, etc].

The question whether an employer’s liability should be strict or fault-based was already a bone of contention when the BGB was being drafted.153 It has been speculated that a misleading account of article 1384 Cc by the editor of the preparatory draft, Franz von Kübel,154 may initially have created the incorrect impression that French law allowed an employer to exculpate himself in the same way as a father.155 This misunderstanding was soon rectified, however.156 When the policy issue was later expressly debated during the legislative process, the majority of the so-called ‘second commission’ rejected the ‘foreign’ French approach in the interest of protecting the budding German industry as well as small agricultural businesses, but at least agreed to reverse the burden of proof.157 As a result, an employer can escape liability for the wrongful act committed by an employee in connection with the entrusted task158 (only) if he proves that he exercised all due care in selecting, instructing

153 A historical overview is provided by HH Seiler, ‘Die deliktische Gehilfenhaftung in historischer Sicht’ (1967) 22 JZ 525; C Förster in Bamberger/Roth (n 143) § 831, [1]–[2.4]; Wagner in Münchener Kommentar (n 152) § 831, [1]; Kannowski (n 43) [5]–[6]. 154 Wagner in Münchener Kommentar (n 152) § 831, [1], referring to W Schubert (ed), Die Vorlagen der Redaktoren für die erste Kommission zur Ausarbeitung des Entwurfs eines Bürgerlichen G ­ esetzbuchs: Recht der Schuldverhältnisse, Teil 1 – Allgemeiner Teil (Berlin, de Gruyter, 1980) 699–700. See also §§ 711–12 of the first draft of the BGB. 155 The misunderstanding was probably promoted by the version of the Code civil promulgated for the Grand Duchy of Baden in 1810, which differed from the original 1804 Code Napoléon in that it contained an exculpation defence for all the persons dealt with in the provision. 156 See Motive zu dem Entwurfe eines Bürgerlichen Gesetzbuches für das Deutsche Reich ­(Berlin/ Leipzig, Guttentag, 1888) vol 2, 736, where French law’s strict liability regime for commettants is clearly noted. 157 A Achilles, H Gebhard and P Spahn (eds), Protokolle der Kommission für die zweite Lesung des Entwurfs des Bürgerlichen Gesetzbuchs (Berlin, Guttentag, 1898), vol 2, 603. See also § 754 of the second draft of the BGB. 158 German lawyers traditionally distinguish between unlawful conduct which the employee commits in the course of performing the assigned task (in Ausführung der Verrichtung) from that for which the employer’s task provided the employee with no more than the occasion (bei Gelegenheit der Verrichtung).

Fait d’autrui in Comparative Perspective  169 and supervising the employee.159 The fact that l­iability is notionally based on culpa in eligendo vel instruendo vel vigilando/custodiendo shows that this is a case of the employer being held liable for his own delict, not vicariously for the employee’s. § 831 BGB poses particular problems where the employer is a large business or other organisation based on hierarchical structures. In this case, the business owner or its top managers will rarely have anything to do with choosing or controlling employees on the third level or further down the pyramid. Although persons to whom these responsibilities are delegated could theoretically be (but are not in practice)160 made liable under § 831(2) BGB for a failure to prevent unlawful acts committed by their subordinates, the owner or the top-level team, ie those who typically count as a corporation’s ‘organs’ under § 31 BGB,161 will merely have to show that the delegates themselves were carefully chosen and instructed. This so-called ‘decentralised’ form of exculpation (dezentralisierter Entlastungsbeweis) has the potential to prevent employers being liable even if there has been a failure, within their sphere, in the selection and supervision of the employee from whom the wrongdoing emanated.

C.  Strategies for Plugging the Gap Left by § 831 BGB There is now widespread consensus that the conceptual framework of § 831 BGB is flawed and rests on a legislative ‘mistake’, but concrete attempts to reform the provision and introduce a French- or English-style strict liability model have failed.162 It may be that there would be greater impetus for change if the consequences of this flaw were more noticeable and dramatic. However, in practice, German law achieves results which are not at all far off the French or English positions. This is because courts have gone out of their way to circumvent the unwelcome exculpation defence by creatively refashioning surrounding areas of law, without having to go against the express wording of § 831 BGB. They have, in short, undermined its spirit and gist, while nevertheless staying true to its form. Three main strategies have been developed over the years to plug the gap(s) left by § 831 BGB.

(i)  Liability under § 823(1) BGB for Organisational Failure The first focuses on the organisational structure of large businesses.163 Even if those at the top of a command pyramid can demonstrate that they have ­delegated all 159 The historical legislator intended that the employer should be under a duty to exercise care either in the selection or in the instruction of his employees, but the Imperial Supreme Court (Reichsgericht) soon required him to show diligence in both cumulatively: Wagner in Münchener Kommentar (n 152) § 831, [35]; Förster in Bamberger/Roth (n 143) § 831, [52]–[53]. 160 The provision of § 831(2) BGB is not made serviceable for the ‘internal’ delegation of duties: BGH 14 May 1974, NJW 1974, 1371, 1372; Förster in Bamberger/Roth (n 143) § 831, [60]–[61]. 161 See above, text accompanying n 10. 162 See Wagner in Münchener Kommentar (n 152) § 831, [3]–[5]. 163 Seminal in this respect: RG 20 November 1902, RGZ 53, 53.

170  Birke Häcker personnel matters to reliable and competent individuals lower down the ­hierarchy, a company may still be liable if its overall organisation and structure falls short of good business practice. The main reproach is then not any specific culpa in eligendo vel vigilando at the highest level, but that the company (acting through its ‘organs’ in the case of a corporation) failed to devise and implement a suitable system of internal control and accountability, making it possible for a lower-level employee to become a source of danger to outsiders. Such organisational responsibilities are treated as non-delegable.164 Accordingly, the touchstone of liability is the delictual general clause in § 823(1) BGB rather than § 831 BGB.

(ii) Broad Reading of Liability for ‘Representatives’ under § 31 BGB The second strategy adopted by the courts consists of an expansion of § 31 BGB. It will be recalled that this provision concerns the attribution to a corporation of unlawful acts committed by one of its ‘organs’. Although § 31 BGB speaks only of board members and ‘constitutionally appointed representatives’, judges have interpreted the provision to apply to anyone entrusted by the corporation with managing significant aspects of the business independently and therefore ­‘representing’ the corporation in a non-technical sense.165 Otherwise, so the reasoning goes, the corporation could escape the strict liability regime of § 31 BGB and benefit from the exculpation defence in § 831 BGB merely by dint of the way it chooses to organise its internal decision-making processes and responsibilities.166

(iii) Expansion of Contractual Liability and Application of § 278 BGB Lastly, and perhaps most importantly, German courts have fundamentally reconceptualised the ‘boundary’ between contractual and extra-contractual liability. In the absence of anything resembling the French non-cumul rule, they frequently resort to contractual claims where competing delictual claims are potentially barred by the exculpation defence. What to an English or French lawyer might look like a straightforward case of vicarious liability for an employee’s wrong, German lawyers will most likely tackle first from the perspective of contract law.167 The reason lies in § 278 BGB, a provision featuring within the ‘general part’ of the law of obligations. § 278 BGB defines a debtor’s ‘responsibility for third parties’ 164 See, eg, RG 29 April 1926, RGZ 113, 293, 297; BGH 30 May 1978, VersR 1978, 722, 723. 165 See, eg, BGH 30 October 1967, BGHZ 49, 19, NJW 1968, 391; BGH 12 July 1977, NJW 1977, 2259. 166 There is a tricky question about the relationship between the broad reading of § 31 BGB and a straightforward case of organisational fault giving rise to a direct liability under § 823 BGB. A company could be said to have breached its organisational duties (thus generating a claim under § 823 BGB) where it fails to ensure that the responsibility for choosing and supervising employees in respect of whose acts it is ‘only’ liable under § 831 BGB (ie, in respect of whom the exculpation defence is available) lies with representatives for whom it is strictly accountable in accordance with § 31 BGB. 167 German lawyers are trained to analyse cases by reference to a strict order of (potentially competing) claims, and here contract features before delict.

Fait d’autrui in Comparative Perspective  171 and stipulates that he ‘is answerable for fault on the part of his legal ­representative as  well as of persons whom he engages to perform his obligation [towards the ­creditor] to the same extent as for his own fault’.168 Within the framework of the BGB, § 278 BGB amounts to a limited recognition of truly vicarious liability by the historical legislator.169 It cuts off any possibility of a debtor who fails to render a performance as owed exculpating himself (under what is now § 280(1), 2nd sentence, BGB) merely by proving that he was not personally at fault. Where he has chosen to discharge the obligation through another (so-called Erfüllungsgehilfe), be it an employee or an ­independent contractor, the latter’s fault is imputed to him as if it were his own. Courts soon realised that § 278 BGB might prove useful as a means of circumventing the unwelcome § 831 BGB if only they managed to (re)classify the victim’s claim as a contractual rather than a delictual one.170 This is relatively straightforward where one contracting party sustains loss or injury through the other’s employee, for contracts can be said to give rise to mutual obligations to respect each other’s rights and interests, in addition to generating positive performance obligations.171 So a building company will be liable in contract for any damage negligently caused to the client’s property by its employees (or indeed even by an independent sub-contractor) during the building process, whether or not it has failed in any way carefully to select, train or ­supervise the employee (or the independent sub-contractor). Where the injured claimant is not a party to the contract, judges have felt able to draw on the principles applying to contracts concluded for the benefit of third parties172 to argue that certain contracts are intended to extend their ‘protective effect’ towards third parties for whose well-being the creditor is responsible.173 This allowed, for instance, a temporary housemaid, whose employer had commissioned the defendant company to carry out works on the gas meter, to sue the defendant for injuries sustained when the defendant’s chief installer negligently left a screw unfixed.174 But what if the parties have not yet contracted when the wrongful injury occurs? A case where a customer in a shop was injured by a falling roll of linoleum carpet, which had been negligently stowed by an employee, became the judicial 168 § 278, 1st sentence, BGB. Emphasis added. 169 Zimmermann (n 25) 1125. The genesis of § 278 BGB is traced by MJ Schermaier, ‘§§ 276–278. Verantwortlichkeit des Schuldners’ in M Schmoeckel, J Rückert and R Zimmermann (eds), Historischkritischer Kommentar zum BGB, vol III/1 (Tübingen, Mohr-Siebeck, 2007) esp [71]–[75], [82]. 170 For an overview see, eg, Schermaier (n 169) [101]–[107]. 171 The doctrine of positive Vertragsverletzung (abbreviated pVV) was first developed praeter legem soon after the enactment of the BGB by H Staub, Die positiven Vertragsverletzungen und ihre Rechtsfolgen (Berlin, Guttentag, 1902). It was eventually codified in the 2002 reform of the law of obligations and is now to be found in §§ 280(1), 241(2) BGB. 172 See §§ 328 et seq BGB. 173 So-called ‘contract with protective effect vis-à-vis third parties’ (Vertrag mit Schutzwirkung für Dritte). 174 RG 10 February 1930, RGZ 127, 218.

172  Birke Häcker starting point of the modern German doctrine of culpa in contrahendo.175 Liability for culpa in contrahendo was characterised as ‘similar to contract’ on the basis that the parties’ mutual rights and duties arose from their pre-contractual relationship. The resulting vast expansion of liability can perhaps best be seen from the so-called ‘lettuce leaf ’ case.176 A 14-year-old girl accompanied her mother to a supermarket and injured herself when she slipped on a vegetable leaf lying on the ground. The Federal Supreme Court combined the doctrine of culpa in contrahendo with the principles governing contracts with ‘protective effects’ vis-à-vis third parties and held that the supermarket owed the girl an obligation akin to a contractual one to take all necessary steps to ensure her safety on the premises. Even if, therefore, the supermarket had been in a position to prove, under § 831 BGB, that the employees who were responsible for cleaning and checking the floor were carefully selected, well-instructed and supervised, this would not have provided a defence. To escape liability, the supermarket would have had to show, in accordance with § 278 BGB, that the employees had properly discharged their cleaning duties. Yet this it was unable to do. The upshot of all this is that the ‘structural defect’ in § 831 BGB – the lack of a genuine vicarious liability in the realm of delict – has led German contract law to become inflated. It now covers vast fields, from the pre-contractual stage to the phase after the contract has been performed (post contractum finitum), and it lends a helping hand to many people who on an English or French understanding have nothing to do with the contract. The system has adjusted and today works well, but the story of § 831 BGB may serve as a reminder of how supposedly ‘small’ decisions in one part of the law can cause large ripples in another.

VI.  The Insurance Environment Instead of providing a summary by way of conclusion, this chapter closes by ­placing some of the questions raised in a broader comparative perspective. As in other areas of the law of tort or delict, the insurance context often plays a central role when one person’s liability for the harmful actions of another are at issue. Indeed, the ministerial statement which accompanied the presentation of the Projet de réforme when it was published in March 2017 expressly acknowledged that the liability regime was ‘inseparably’ tied up with the insurance market.177

175 RG 7 December 1911, RGZ 78, 238. The doctrine is now codified: see §§ 280(1), 311(2) BGB. 176 BGH 28 January 1976, BGHZ 66, 51, NJW 1976, 712, MDR 1976, 570. 177 J-J Urvoas, ‘Présentation du projet de réforme du droit de la responsabilité civile’ (13 March 2017), www.presse.justice.gouv.fr/archives-discours-10093/archives-des-discours-de-2017-12856/ projet-de-reforme-de-la-responsabilite-civile-29780.html (last accessed 23 April 2019): ‘tout système de responsabilité civile est aujourd’hui indissociable du mécanisme de l’assurance’. This statement – though general in nature – was made specifically in connection with liability for fait d’autrui.

Fait d’autrui in Comparative Perspective  173 The interrelationship between the substantive law of tort or delict and the insurance environment is (potentially, at least) a two-way street. On the one hand, the availability of insurance may influence the evolution of substantive law; on the other hand, the existing liability regime will have repercussions for the types of insurance products that are available. What follows should be seen not so much as a comprehensive and systematic analysis of the interaction between liability and insurance, but as merely flagging two themes for the reader’s special attention.

A.  (How Far) Does Liability Depend on the Existence or Availability of Insurance? First, there is a question of whether and to what extent the existence or availability of (third-party)178 insurance ought to determine a defendant’s liability. As far as the courts are concerned, prima facie, the answer should be: not at all. Since a defendant insures against liability, it would lead to a vicious logical circle to maintain that his liability depends on whether or not insurance will ultimately cover it. Nevertheless, judges in France have long been amenable to taking the insurance situation into account. It is widely recognised, for instance, that the expansion of parental liability for their children, which culminated in the Cour de cassation’s Levert decision,179 would not have happened but for the fact that that almost all parents are insured against such liability: French courts are all the more willing to extend the ambits of tort law knowing (or at least thinking) that potential defendants are likely to be insured … Tort law and liability insurance thus reinforce each other: the breadth of tort law in France creates a strong incentive to take out liability insurance and the generalisation of liability insurance makes further extensions of tort law more acceptable.180

By contrast, German and English courts have traditionally rejected insurancerelated reasoning. German lawyers apply a so-called ‘principle of separation’ in this context (versicherungsrechtliches Trennungsprinzip): neither the existence nor the extent of liability may be influenced by the parties’ insurance situation.181 As far as English law is concerned, Lord Bridge once stated in no uncertain terms: At common law the circumstance that a defendant is contractually indemnified by a third party against a particular legal liability can have no relevance whatever to the measure of that liability.182 178 Third-party insurance concerns situations where the insured person is liable to someone else, while first-party insurance provides cover where the insured person is himself or herself injured or sustains a loss. The focus of the present section is on third-party insurance. 179 See above, n 75 and the accompanying text. 180 J-S Borghetti, ‘The Culture of Tort Law in France’ (2012) 3 Journal of European Tort Law (JETL) 158, 166–67 (footnote omitted). 181 For a clear recent statement see, eg, BGH 27 October 2009, NJW 2010, 537. 182 Hunt v Severs [1994] 2 AC 350 (HL) 363.

174  Birke Häcker Of the three systems, German law is arguably strictest in segregating the liability issue (the first question) from that of who bears the cost of shouldering it (the second question) and the loss-spreading effects of insurance (a mere background factor).183 In England, however, judicial attitudes appear to have significantly softened over the years. An early example is the well-known flirtation by Lord Denning MR in Nettleship v Weston – a road traffic accident case – with the idea that the existence of insurance could be a liability-enhancing factor.184 More recently, courts have begun to consider the availability of insurance in connection with their development of vicarious liability.185 In a 2016 Supreme Court decision, Lord Reed explained that the parties’ relative financial strength and insurance situation were not usually of independent significance, but could have some relevance: It is, of course, true that where an individual is employed under a contract of employment, his employer is likely to have a deeper pocket, and can in any event be expected to have insured against vicarious liability. Neither of these, however, is a principled justification for imposing vicarious liability. The mere possession of wealth is not in itself any ground for imposing liability. As for insurance, employers insure themselves because they are liable: they are not liable because they have insured themselves. On the other hand, given the infinite variety of circumstances in which the question of vicarious liability might arise, it cannot be ruled out that there might be circumstances in which the absence or unavailability of insurance, or other means of meeting a potential ­liability, might be a relevant consideration.186

In Ellis v Kelly, a parental liability case decided in 2018, the judge openly ­maintained that because ‘parents are not reasonably able to secure insurance to guard against the risk of claims arising out of their parenting generally’, it was ­inappropriate to hold them to a high standard of care, thereby incentivising them to be ‘over-cautious’.187 All in all, therefore, the English judicial approach – in terms of reasoning, though not necessarily in terms of result – seems to be increasingly approximating that adopted by French courts.

183 Yet even German law allows (compulsory) third-party insurance to be a factor in determining a defendant’s ‘equitable’ liability under § 829 BGB (cf n 149): see, eg, BGH 29 November 2016, VersR 2017, 296. This is very controversial and has given rise to much litigation over the years: Wagner (n 142) 224–26; idem (n 152) § 829, [18]–[22]. 184 Nettleship v Weston [1971] 2 QB 691 (CA) 700: ‘Morally the learner driver is not at fault; but legally she is liable to be because she is insured and the risk should fall on her.’ Looking at the whole judgment, it is arguable that Lord Denning had not so much in mind the learner driver’s cover in the individual case (which the claimant had checked before agreeing to give her driving lessons), but the broader system of compulsory car insurance. 185 See, eg, Lord Phillips in the so-called ‘Christian Brothers’ case: Various Claimants v Catholic Child Welfare Society [2012] UKSC 56, [2013] 2 AC 1, [35]: ‘There is no difficulty in identifying a number of policy reasons that usually make it fair, just and reasonable to impose vicarious liability on the employer …: (i) the employer is more likely to have the means to compensate the victim than the employee and can be expected to have insured against that liability’ (emphases added). 186 Cox (n 140) [20]. 187 Ellis v Kelly (n 133) [77]–[78]. Interestingly, in this case it was a driver’s insurance company that sought a contribution from the mother of the child who had been injured in a car accident.

Fait d’autrui in Comparative Perspective  175 When it comes to the legislator, there is – on any view – nothing problematic about taking the insurance situation into account when setting the appropriate standard or ‘level’ of liability. This regularly happens. When, for instance, the fathers of the BGB discussed whether to hold employers strictly liable for the wrongful actions of their employees,188 an additional reason given for not doing so was that only large businesses (which could afford it) were likely to take out insurance, and that the drafters lacked the legislative competence to make ­ ­insurance compulsory for all.189 Furthermore, it will be recalled that suggestions in Germany for imposing true vicarious (ie strict) liability on parents by statute are regularly coupled with ideas for introducing a compulsory third-party insurance system.190

B.  Different Insurance Products – and Different Cultures of Claiming The second way in which substantive law and the insurance environment interact concerns the influence the former has on the latter. As a result of ordinary market forces, the availability and shape of insurance products will typically be tailored towards the (perceived) need to insure against liability. We have already seen that a legislator may be driven towards imposing a requirement that certain parties have compulsory third-party insurance in order to counter-balance an existing or envisaged strict liability regime. Where this is the case, appropriate insurance products will develop almost automatically (especially where they need to be checked and approved by a regulator). However, quite irrespective of their remarkably different approaches towards the liability of parents or employers for the acts of their children or employees respectively, neither England nor France nor Germany operate a fully comprehensive191 system of compulsory third-party insurance covering potential defendants. The availability and reach of insurance products is therefore determined primarily by their historical evolution and by existing customer demand. Given that this demand will, in turn, be influenced by the (actual or perceived) risk of parents or employers being found liable for the actions of others, it is unsurprising that we should find certain differences in the national insurance landscapes. The differences are more noticeable in the realm of parental liability than that of employers.192 They are perhaps best illustrated by a concrete observation. 188 See above, text accompanying n 157. 189 Protokolle, vol 2 (n 157) 603. 190 See above, n 148 and the accompanying text. 191 Certain aspects of an employer’s vicarious liability are be covered by compulsory insurance, eg, where one employee is injured by another (in Britain under the Employers’ Liability (Compulsory Insurance) Act 1969). As regards parental liability in France see n 200 and the accompanying text. 192 It seems probable that this is because even German law, despite the fault-based starting point of § 831 BGB, arrives at a liability regime that is more or less strict through the work-arounds outlined above (in section VC).

176  Birke Häcker Germans moving to England are often surprised by the fact that the UK insurance market appears to offer no product equivalent to what in Germany is regarded as a key type of insurance, namely the so-called ‘private’ or ‘personal third-party liability insurance’ (private Haftpflichtversicherung). Although not compulsory, it is one of the insurance products that few adults in Germany would dare to go without.193 It covers insured parties against delictual liability incurred towards third parties (for non-intentional wrongs).194 One reason why demand for such private third-party insurance products is particularly great amongst German families195 is that it protects parents in situations where they are liable under § 832 BGB because they have failed in their duty to supervise a child that has caused loss to a third party. Due to the reversed burden of proof, the prospect of liability is real and not uncommon. Think only of a child visiting someone’s house and causing mischief there, or a football game in the garden getting out of hand with the ball escaping and shattering a neighbour’s large sitting-room window. Such is the expectation that the culprit’s family will pick up the cost that many insurance providers offer policies which will pay for the damage even if neither the parents nor the child could, strictly speaking, be held legally liable (eg, where the child is too young to control or appreciate the wrongfulness of her conduct and where the parents have not failed in their duty to supervise). Morally parents feel responsible to compensate the party who has suffered the loss, and they guard against it by taking out appropriate insurance. In the UK, the concept of general third-party liability insurance is virtually unknown outside the business context. What most English adults instead regard as essential is having a so-called ‘home’ or ‘household insurance’.196 This is, at its core, a form of first-party insurance.197 It typically protects the policyholder against loss or damage to his home (the building itself, its contents and sometimes other belongings), whether the loss or damage comes about accidentally, by theft or as a result of someone’s negligence. A side-effect of this setup is that most policyholders prefer to claim on their own home insurance whenever they can, rather than seeking to hold liable whoever is legally responsible for the loss or damage. In the above examples, the host or neighbour would be unlikely to pursue the mischievous child or her parents, but would expect (and probably would be expected) to refer the matter to his insurer instead. Although the insurer would then, of course, be subrogated to any existing claims vis-à-vis the culprit or her parents, it is rare for insurers to have

193 The general private third-party liability insurance sits alongside (and does not include) third-party liability insurance for road traffic accidents (Kfz-Haftpflichtversicherung), which is compulsory for anyone owning/operating a car. 194 Cover is routinely excluded for intentional conduct by the insured party. 195 A so-called ‘family third-party liability insurance’ (Familienhaftpflichtversicherung) covers not only the person who takes it out, but also his or her family members, including minor children. 196 Such home insurance policies are also available in Germany, but they are not regarded as a ‘must have’. 197 On the difference between first- and third-party insurance, see n 178 above.

Fait d’autrui in Comparative Perspective  177 recourse against them – not least because the hurdles for establishing parental ­liability are comparatively high.198 To the extent that personal liability insurance cover is available at all on the UK insurance market, it tends to be included in certain home insurance packages as an (optional) third-party add-on. The contrast between the English and the German position is particularly revealing when one compares the insurance market in France. As might be expected in light of the strict regime of parental liability,199 having private third-party ­insurance cover for damage done by children is deemed essential. Interestingly, it is traditionally part and parcel of the ‘household insurance’ ­(assurance multirisque habitation), which most individuals or families have and which is in some cases compulsory.200 As noted by Borghetti: In effect, most natural persons in France are … protected by liability insurance through their household insurance. In the event that someone does not have household insurance, they are able to purchase civil liability insurance at a very reasonable price.201

Yet even where a person is comprehensively covered against third-party liability via the household insurance, the incentives to guard against the economic risk of incurring personal liability appear to be such that it is quite common to take out an additional ‘German style’ separate civil liability policy (assurance responsabilité civile vie privée).202 A corollary of insurance being so widespread in France is, as we have seen, that it has emboldened courts to expand parental liability – but at the same time it also takes the sting out of a finding of liability.203 There is therefore nothing particularly awkward about friends, colleagues or neighbours claiming compensation from each another, as might be the case in England. Just as the substantive framework of accountability for fait d’autrui differs between the systems, so do insurance products and cultures of claiming. U ­ ltimately, loss allocation and distribution mechanisms vary between French, English and German law. The envisaged reforms of the French civil liability regime are unlikely to change that.

198 Ellis v Kelly (n 133), the case in which the judge took into account that ‘[p]arents are not ­reasonably able to secure insurance to guard against the risk of claims arising out of their parenting generally’ (see the text accompanying n 187), actually concerned a contribution claim by an insurer against an (injured) child’s mother. 199 See above, section IIIA(ii) of this chapter. 200 Tenants in rented accommodation are required to take out a household insurance. 201 Borghetti (n 180) 166 (footnotes omitted). 202 See ibid 167, fn 26, suggesting that lay people often take up offers of civil liability insurance without realising that they are already protected by their household insurance. This, of course, potentially gives rise to all sorts of double insurance problems. 203 Borghetti (n 180) 168 writes that ‘[n]othing indicates that people in France are particularly afraid of having to compensate damage that they, or those for whom they have to answer, may cause’.

178

part iv ‘Harm’

180

9 Loss and its Compensation in the Proposed New French Regime of Extra-contractual Liability DOROTA LECZYKIEWICZ

In 2016, following the reform of the Code civil provisions relating to contracts,1 the French Ministry of Justice proposed to reform provisions relating to ‘civil liability’, which in the French tradition technically encompasses both c­ ontractual and extra-contractual liability.2 In March 2017, a new version of the Projet de réforme de la responsabilité civile was published, some sections of which were to apply both to contractual and extra-contractual liability, and some of which concerned only extra-contractual liability.3 Despite that, the project was generally regarded as transforming primarily the French regime of extracontractual ­ liability, formerly known as the law of delicts and quasi-delicts (articles 1240–51 Cc, previously 1382–86 Cc). The rules surrounding the condition of loss/harm4 and the calculation of damages are where we see the greatest increase in codified rules and new solutions. The French law of extra-contractual liability as currently expressed in the Code civil possesses two distinctive characteristics. The first one is the absence of any apparent restriction on the types of harm recoverable under the Code civil rules. The second is the absence of codified rules on how damages should be calculated

1 Ordonnance of 2016. 2 See for example, G Viney, Traité de droit civil, Introduction à la responsabilité, 3rd edn (Paris, LGDJ, 2008). 3 For the text of the project, still not, as of April 2019, adopted as law, www.justice.gouv.fr/ publication/Projet_de_reforme_de_la_responsabilite_civile_13032017.pdf. 4 As I will explain below, the proposed provisions use two different terms to refer to the condition of damage – dommage and préjudice. Simon Whittaker, when preparing the English version of the Projet de reforme de la responsabilité civile, translated the term dommage as ‘harm’, and the term p­ réjudice as ‘loss’. I will follow this convention in this chapter. However, I will use the term ‘damage’ where I want to use a neutral term that refers neither to the French concepts of dommage nor to the French concept of préjudice, where both terms might, in fact, be applicable, or where the proposed provisions are ­inconsistent in the way they use both terms.

182  Dorota Leczykiewicz and what is required to realise the principle of full reparation (réparation intégrale), in itself for now a merely judicial principle. The Projet proposes the introduction of a number of mechanisms that can broadly be related to controlling actionability and recoverability of damage and to providing more guidance to courts on setting the compensatory award. The stated objectives of these new mechanisms can be gleaned from the Discours of the French Minister of Justice of 13 March 2017, presenting the second version of the project.5 In summary, they are threefold; to bring French codified civil law up-to-date by reference to the case law, provide better protection to victims, in particular of personal injury, and achieve a higher level of predictability of the law into the future.6 The last of these objectives has implications for the first. A codification of the case law will only succeed in achieving a higher level of predictability if it goes beyond the superficial layer of the jurisprudence and engages with the deeper principles and considerations motivating French judges when deciding cases. It should also internalise issues of judicial discretion and recognise the fact that the proposed codified rules will be used by the Cour de cassation to control lower courts. Related to predictability is also the issue of the persisting effect of the case law that has not been expressly codified. One way in which certainty is unlikely to be improved is by populating the ­written law, and thus equipping judges, with a large number of what English lawyers would call ‘control devices’, that is concepts or distinctions that are not based on principle but enable judges to address particular matters ad hoc, often in response to policy considerations. Bearing these observations in mind, in this chapter, I would like to discuss four issues. First, I will look at the significance of the distinction, introduced for the first time by the proposed reform, between harm (dommage) and loss ­(préjudice). Secondly, I will focus on the characteristics that préjudice needs to possess to be recoverable, some of which concern issues of proof, while others concern more policy-motivated restrictions on actionability of certain forms of damage. Thirdly, I will consider the consequences of introducing a formal distinction between patrimonial and non-patrimonial losses (préjudices patrimoniaux and préjudices extrapatrimoniaux). In this discussion, I will focus on what is expressly stated as one of the reform’s objectives – the improvement of the legal position of personal injury victims, in particular in relation to deductions that their compensation may suffer in favour of ‘third-party payers’ (social security funds and insurance companies). Fourthly, I will examine the manner in which damages would be calculated after the reform, focusing in particular on damages awarded in personal injury cases.

5 Discours de Monsieur Jean-Jacques Urvoas, garde des sceaux, ministre de la justice, Présentation du projet de réforme du droit de la responsabilité civile, Académie des Sciences morales et politiques, 13 March 2017, www.presse.justice.gouv.fr/art_pix/Discours%20-%20Pr%E9sentation%20 du%20projet%20de%20r%E9forme%20du%20droit%20de%20la%20responsabilit%E9%20civile% 2013.03.2017.pdf (accessed 6 May 2019). 6 ibid 2–3.

Loss and its Compensation in the Proposed New French Regime  183

I.  The Distinction between Dommage and Préjudice The central provision on extra-contractual liability in current French law is article 1240 Cc (previously article 1382 Cc).7 It is notorious for its brevity and vagueness. Despite that, it does define the obligation which it imposes on the person whose ‘fault’ (faute) has caused harm to another, which is the obligation to provide reparation for harm. Against the wording of article 1240 Cc, it is also clear that only conduct ‘causing harm to another’ creates that obligation.8 Article 1240 Cc does not qualify the recoverable harm in any way, from which it is frequently inferred that all types of harm are recoverable in French tort law.9 The provision uses the term dommage, not préjudice, to refer to the harm for which the defendant will be liable. Nevertheless, préjudice has been routinely substituted for dommage by ­scholars and judges, especially in the context of different kinds of harm that over time have been recognised as recoverable under the heading of dommage.10 Thus, under current law it is not possible to maintain that the distinction between dommage and préjudice concerns recoverability of damage, whereby only some items of préjudice would count as dommage and would therefore be recoverable. Instead, the use of two terms has allowed scholars and judges to have two different categorisations of damage. One of them distinguishes physical personal injury (dommage corporel) from other types of harm; dommage moral and dommage matériel (damage to property or financial assets).11 The other categorisation focuses on the 7 It reads: ‘Any human action whatsoever which causes harm to another creates an obligation in the person by whose fault it occurred to make reparation for it’ (‘Tout fait quelconque de l’homme, qui cause à autrui un dommage, oblige celui par la faute duquel il est arrivé à le réparer’). 8 I will describe the person to whom harm is caused as ‘the victim’ rather than ‘the claimant’. This is because in French tort law it is necessary to distinguish claimants from victims. Tort claims are routinely brought by parties other that the persons who have suffered harm, such as their insurers or social security funds, who subrogate the victim in their claim and whose entitlement against the compensatory award often takes priority over that of the victim (for more on this, please see below). Thus, in the context of French law, it is better to speak of the victim’s harm and the victim’s right to compensation, even if part of that compensation will actually be diverted to a third party. 9 French law may indeed be regarded as based on a tacit assumption of recoverability of all harm and is thus occasionally presented as a loss-based regime, to be contrasted with a rights-based or interestsbased regime, where only loss flowing from a violation of a right or of a legally protected interest is recoverable under the general law of extra-contractual liability. See § 823(1) BGB and Section 2, Book VI DCFR. 10 See for example, préjudice d’agrément, préjudice esthetique, préjudice commercial etc. For more on the distinction between dommage and préjudice, see G Viney, P Jourdain and S Carval, Traité de droit civil, Les conditions de la responsabilité, 4th edn (Paris, LGDJ, 2013) 3–4, and the literature cited in fn 3. After the 2016 reform, the provisions on extra-contractual liability use the term préjudice in the context of environmental harm (préjudice écologique). See art 1246 Cc and the following. 11 The term dommage corporel is not used anywhere in the Code civil. The section on liability for defective products uses the term ‘une atteinte à la personne’ (see art 1245-1 Cc, previously art 1386-2 Cc). French scholarship has been divided as to whether dommage moral belonged to the category dommage corporel or was independent from it. In favour of its independent status speaks the fact that those secondary victims who suffer emotionally because a person close to them was physically, and emotionally, injured are also regarded to have suffered dommage, that is dommage moral, even though they have not suffered dommage corporel. See F Terré, P Simler and Y Lequette, Droit civil. Les Obligations, 11th edn (Paris, Dalloz, 2013) 767–75.

184  Dorota Leczykiewicz nature of the consequences and distinguishes between those that can be directly expressed in economic terms (conséquences patrimoniales) and those that cannot (conséquences extrapatrimoniales).12 The obligation to provide reparation in article 1240 Cc has been taken to mean either an obligation to provide a reparation in kind (practically, of little significance in French, or any, tort law) or an obligation to pay compensation (damages), ie dommages et intérêts, to borrow the term used in the context of contractual liability.13 Article 1240 Cc says nothing about the (im)possibility of awarding nominal or punitive damages but the fact that the tortfeasor14 is obliged to provide ‘reparation for harm’ means that the provision can only be a legal basis for awarding compensatory damages. As for article 1242 Cc, which lays down the two other main principles of extra-contractual liability in French law (liability for the acts of others and liability for harm caused by things), its focus on the compensatory ­function is less explicit. It uses the phrase ‘is liable’ (on est responsable), not ‘is obliged to provide reparation’ (oblige à réparer), but just like article 1240  Cc it refers to harm (dommage) as the circumstance giving rise to the tortfeasor’s liability.15 The connection between the obligation resulting from article 1242 Cc and the duty defined in article 1240 Cc, while implicit, is considered to exist. Thus, in the literature it is never questioned that the secondary (remedial) obligations arising from both provisions are identical. The Projet de réforme proposes to introduce a whole array of new solutions in relation to the condition of loss/harm. The first novelty is that the reform proposes a separate section on Le préjudice réparable (sub-section 1 of Section 1, entitled Dispositions communes aux responsabilités contractuelle et extracontractuelle, located in Chapter II Les conditions de la responsabilité) preceding the regulation of both contractual and extra-contractual liability and expressly described as ‘common’ to both regimes. The first provision of sub-section 1 is article 1235, which defines what is ‘reparable’, ie for what the claimant can recover. The claimant can recover not for dommage but for préjudice, despite the fact that the tortfeasor is liable for dommage, as we can see in proposed articles 1241, 1243 and 1245. Moreover, article 1239 defines causation as a link between ‘the action attributed to the defendant and the harm’ (‘le fait imputé au défendeur et le dommage’). ­Article 1259, on the other hand, speaks of ‘full reparation of the loss’ (‘la réparation

12 Viney, Jourdain and Carval (n 10) 4. 13 The proposed provisions include a chapter (Chapter IV) on the effects of liability, which is common to contractual and extra-contractual liability. This chapter mentions two kinds of remedies: la réparation en nature and les dommages et intérêts. See arts 1260–64 of the Projet de réforme. 14 I will use the term ‘tortfeasor’ to describe someone who is liable in French law on one of the noncontractual bases of liability regulated by the French Code civil. 15 See also art 1285 of the Projet de réforme on liability for harm caused by motor-vehicles and art 1289 of the Projet de réforme on liability for harm caused by defective products, which also use the term dommage. The exception is liability for environmental harm (art 1279-1 of the Projet de réforme), where the term dommage appears only in the heading of sub-section 3 in the phrase dommage environmental. The provisions themselves use exclusively the term préjudice.

Loss and its Compensation in the Proposed New French Regime  185 intégrale du préjudice’). These expressions as well as the fact that in article 1235 préjudice is only réparable if, among other requirements, it results from dommage, would suggest that the new provisions introduce two levels of analysis. Préjudice is what is repaired or compensated; what damages will remedy. Dommage, on the other hand, operates at a higher level; it is a condition of liability and thus controls the possibility of bringing a tort action. The connection between the two levels is provided by article 1235, which requires that préjudice, to be recoverable, must result from dommage. However, being a result of un dommage is not the only condition that préjudice needs to satisfy, which would suggest that not all losses flowing from harm caused by the tortfeasor will be recoverable. If so interpreted, the new regime would depart from a view, currently followed, that all losses covered by the concept of dommage are recoverable. The second novelty is Section 2 of the proposed Chapter IV, regulating the effects of civil liability and entitled Règles particulières à la réparation des ­préjudices résultant de certaines catégories de dommages. This section overtly introduces certain special rules that apply to losses stemming from certain enumerated categories of dommage, namely dommage corporel, dommage matériel, dommage environnemental and le retard dans le paiement d’une somme argent. As I will explain later, explicit regulation of certain categories of harm acts in favour of the losses which could be connected to these categories, especially those expressly mentioned by specific provisions.16 Their recoverability seems to be settled by the proposed provisions and it is not clear to what extent recovery for them can be denied by reference to the requirements codified in the sub-section on ‘le préjudice réparable’, a point I will discuss further below. Regardless of this, it seems that Section 2 of Chapter IV will have an implicit role in controlling actionability of the enumerated types of harm and the recoverability of the losses which flow from them, functions that are currently performed simply by the concept of dommage in article 1240 Cc.

A.  Actionability of Damage in Current Law I have mentioned above that the approach currently used in French tort law is that there is no obvious distinction between the concepts of préjudice and dommage. Losses become recoverable when they are seen as covered by the concept of dommage, which is the term used by article 1240 Cc. For this reason, it is the concept of dommage, not préjudice, which the Cour de cassation uses to control

16 Provisions on dommage corporel expressly mention le déficit fonctionnel (art 1270 of the Projet de réforme), la perte de gains professionnels, la perte de revenus des proches and l’assistance d’une tierce personne (art 1272 of the Projet de réforme) as flowing from this type of harm. Provisions on dommage matériel, on the other hand, expressly mention l’atteinte à un bien corporel, with further explicit differentiation of la privation de jouissance, les pertes d’exploitation and tout autre préjudice (arts 1278–79 of the Projet de réforme).

186  Dorota Leczykiewicz actionability of certain forms of damage and to introduce requirements that the victim’s loss needs to meet to be recoverable. Loss, to constitute dommage, needs to be personnel, direct and certain.17 The requirement of the personal character of the harm traditionally meant that it was not possible to claim under article 1240 Cc compensation for collective or environmental harm, although these rules have been relaxed in the case of public interest organisations.18 However, the requirement of the personal character of the harm would not prevent secondary victims (victimes par ricochet), ie those who suffered harm only indirectly, from bringing a claim. This has been true also when the secondary victim suffered harm of a purely emotional character.19 Neither did the requirement of directness of the harm stand in the way. That requirement is generally taken to refer to the connection that needs to exist between the harm pleaded by the claimant and the fault of the defendant. The fact that the defendant had no dealings with the claimant, that their conduct affected the claimant only through the harm inflicted on the primary victim, is no bar to actionability. More generally, harm caused by non-physical means has also been actionable. One may thus wonder what the function of the requirement of directness has been. It seems that directness is best described as a ‘control device’, which in English legal tradition is understood to mean a concept that allows judges to accept or reject some claims not necessarily on any principled ground, but by responding to the special facts of a particular case or to implicit policy considerations.20 How else can one explain the fact that intervening acts of third parties do not in French law automatically break the chain between fault and harm for the purpose of its actionability? The requirement that harm be certain has been less of a ‘control device’ than directness. Instead, it has performed a number of relatively independent functions. This is the reason why it is difficult correctly to translate dommage certain into English. On the one hand, it could be translated as ‘existing damage’, on the other, as ‘highly probably damage’, or also as ‘special’ or ‘clearly visible/easily identifiable damage’. The condition of dommage certain is also taken to refer to evidential 17 Cass civ (2) 12 November 1986, Bull civ 1986 II no 164, 111. 18 Moreover, the 2016 reform introduced additional provisions in the Sub-Title II on extra-contractual liability dealing with environmental harm (Chapter III: La réparation du préjudice écologique, arts 1246–52 Cc). These provisions do not create so much a right to compensation for the benefit of a victim of a tort as an obligation to repair harm suffered by the environment. An action is not brought by an individual victim, as in the case of a standard tort action in French law, but a person who has an interest in bringing an action against the polluter. These persons are enumerated in art 1248 Cc (the state, specialised governmental agencies, local government, non-governmental organisations whose mission is the protection of the environment). Arts 1246–52 Cc thus indirectly confirm that the harm recoverable under the general regime, arts 1240–44 Cc, needs to be personal to the victim/claimant, including an association at least in the sense that the interest that has been infringed has to be ‘personal’ to them because of the goals they pursue. 19 It is so also regardless of the fact that the defendant can rely on the primary victim’s contributory fault also against the secondary victim. To that effect, see Ass plén 19 June 1981, JCP G 1982 II 19712. 20 For an English or a comparative lawyer, French private law is full of such ‘control devices’, even if French lawyers themselves do not want to admit that this is what they are.

Loss and its Compensation in the Proposed New French Regime  187 requirements.21 Thanks to this concept, in conjunction with article 9 of the French CPC, it is clear that the claimant, at least in principle, needs to prove the existence of harm. La certitude of damage indicates that the claimant has a duty to adduce evidence that some harm actually exists,22 although it remains unclear whether the claimant has to prove the existence of all items of harm for which they claim damages. A somewhat tricky point concerns the requirement of certainty in the context of future harm. The Cour de cassation already in 1932 held that future harm is actionable, and recoverable, but only when it constitutes ‘the certain and direct prolon­gation of a current state of affairs’ (‘la prolongation certaine et directe de l’état de choses actuel’).23 Given the fact that this formulation incorporates two ambiguous terms, certaine and directe, in itself it does not clarify when the victim can recover for future harm. Thus, a rich body of case law has emerged stipulating in a casuistic way in what circumstances future harm should be treated as certain and direct.24 In order to summarise that case law French scholars employ the concept of ‘normal foreseeability’ and claim that the case law shows that it is the ‘normally foreseeable consequences of the event causing harm’ that are recoverable.25 These include both ‘normally foreseeable’ future expenses of the victim, such as the cost of future care and assistance, and ‘normally foreseeable’ gains, such as future income. In both cases, what is ‘normally foreseeable’ is an evaluative matter and should invite questions about the level of probability that has to be shown before the claimant can recover for a particular item of future harm. Yet, in French law these questions are not openly discussed. Neither are they addressed in the case law, probably because French law rejects the idea of standard of proof.26 Nevertheless, French scholarship and case law have shown reluctance to endorse the idea that the requirement of certainty means that harm has to be proved with 100 per cent probability.27 The solution is thus to leave the matter, on the one hand, to the sovereign assessment of the juges du fond, and, on the other, to rely on the case law of the Cour de cassation, which gradually has been identifying the instances of recoverable future harm and the circumstances in which lower courts are obliged 21 ‘Le préjudice ne peut être réparé que s’il existe réellement: il doit être certain’: H Mazeaud, L Mazeaud, J Mazeaud and F Chabas, Leçons de droit civil. Obligations. Théorie générale. Tome II. Volume I, 9th edn (Paris, Montchrestien, 1998) 415. 22 Cass civ (2) 10 January 1990, JCP 1990 IV 95, Cass civ (1) 28 May 1991, JCP 1992 II 21864, Cass civ (1) 25 June 1996, Bull civ I no 271, Cass civ (1) 27 March 2001, JCP 20001 II 10089. 23 Cass 1 June 1932, S 1933, 1, 49, D 1932, 1, 102, Gaz Pal 1932, 2, 363. 24 See the list of cases in fn 605 at 123 in Viney, Jourdain and Carval (n 10). 25 ibid. 26 The French standard of proof is that of judicial conviction (l’intime conviction), ie what is enough to convince the judge of fact. High probability is implicitly regarded as sufficient but it is not clear if it is enough to show that something is simply more probable than not. 27 ‘The certainty required by the courts is not and cannot be absolute certainty … Such a requirement would lead to ruling out any compensation for future harm, which is never absolutely certain. Judges must be satisfied with relative certainty, with a sufficient degree of probability’. P Jourdain, RTD civ 1994, 108. See also, Cass civ (1) 3 October 1974, JCP 1975 II 18156; Cass civ (2) 5 October 1967, D 1968, 197.

188  Dorota Leczykiewicz to include in the ­compensatory award a sum which constitutes recovery for those instances of future harm.28 And so, despite the requirement of dommage certain, French law allows full recovery for loss of a profit that the victim would only have obtained in the future. That the course of events would have run in their favour needs to be established only with some level of probability, sufficient to distinguish the victim’s harm from purely hypothetical damage.29 One thing to bear in mind is that according to the case law of the Cour de cassation, compensation is to be estimated on the date of the judgment, and this is the point in time when judges assess what harm is already in existence and what losses lie in the future. Another condition which the victim’s loss has to satisfy before it becomes dommage is that of legality of damage. This is sometimes expressed as the requirement that dommage needs to be ‘lawful’/’legal’ (licite) or that dommage occurs only when the victim’s loss stems from a violation of an interest that is legitimate (le caractère légitime du dommage).30 This judge-made addition shows most vividly how in French law the list of the conditions of extra-contractual liability can be expanded without changing the text of the relevant provision. Legality as a condition of actionability can have two dimensions – a negative and a positive one. The negative dimension establishes a prohibition. Judges are not permitted to treat as dommage a loss whose existence stems from a relationship or a situation which the law does not regard as legal, or which is not recognised by law. And thus, actionability of losses suffered, as secondary victims, by ‘concubines’ had been rejected not because they could not prove harm, or because their harm was insufficiently direct, but because it stemmed from a relationship with the primary victim that was regarded as irregular, a marital relationship being the norm.31 Another example of the negative dimension of the requirement of legality can be found in the context of cases rejecting recovery for criminal proceeds and illegal gains.32

28 For example, a child claimant is entitled to damages for the consequences of their future limited physical and intellectual ability, or loss of a future educational and professional advantage. Such loss is not always dealt with under the heading of ‘the loss of a chance’ (la perte d’une chance). See for example, Cass civ (2) 9 April 2009 no 08-17.333, where loss of a future education and professional advantage was compensated under its own heading of préjudice scolaire, universitaire ou de formation. For a recent example, see Cour d’appel, Paris, Pôle 2, chambre 3, 14 May 2018 no 16/23197. 29 So, for example, the pain and suffering that the victim will suffer in the future on account of a further surgical operation are treated as ‘hypothetical’ when it is known that the victim will not agree to the procedure. Cass civ (1) 28 June 2012 no 11-19.265. 30 Terré, Simler and Lequette (n 11) 763; Viney, Jourdain and Carval (n 10) 111. 31 The view was abandoned in 1959 by Cass crim 24 February 1959, JCP 1959 II 11095, RTD civ 1959, 354. This was then confirmed by Chambre mixte on 27 February 1970, D 1970, 201, JCP 1970 II 16305, RTD civ 1970, 353. 32 Cass civ (2) 30 January 1959, Bull civ II no 116; Cass civ (2) 24 January 2002 no 99-16.576, Bull civ II no 5, JCP 2002 II 10118; Cass civ (2) 28 May 2009 no 08-16.143. However, not all forms of irregular conduct will prevent the victim from claiming compensation. Victims of traffic accidents will be able to recover for personal injury also when they travelled without a ticket or in a stolen car. See Cass civ (2) 19 February 1992, Bull civ II no 54, JCP 1993 II 22170; Cass civ (1) 17 November 1993, Bull civ I no 326, RTD civ 1994, 115.

Loss and its Compensation in the Proposed New French Regime  189 The positive dimension of legality, where the condition is treated as a requirement that the victim has to satisfy in each case by showing that their harm stems from a violation of a right or a legally protected interest,33 is not currently visible in French tort law.

B.  What Controls Actionability in the New Provisions: Dommage or Préjudice? A significant change that the proposed provisions intend to introduce is the transfer of two of the necessary characteristics of dommage in current law (its legality and certainty) to the concept of préjudice. According to article 1235 of the Projet de réforme it is préjudice that needs to be certain and stem from a violation of a lawful interest (intérêt licite).34 This would suggest that not only recoverability but also actionability of damage is now controlled by the provision overtly regulating what is reparable – article 1235, and not the provisions laying down the particular principles of liability (articles 1241, 1243 and 1245 of the Projet). But this is difficult to reconcile with the fact that article 1235 insists that one of the conditions of reparability of loss (préjudice) is that it stems from dommage. Dommage, on the other hand, is not qualified in the proposed provisions in any way. So what is the point of insisting that reparable loss needs to be a result of dommage? It cannot be ruled out that courts will use the requirement that préjudice stem from dommage to restrict recoverability of some losses satisfying the other conditions in articles 1235–38 of the Projet, but there is nothing in the provisions laying down the principles of liability, which as I have mentioned above use only the term dommage, that could tell us what criteria might apply. Another ambiguity is created by Section 2 of the proposed Chapter IV, regulating the effects of civil liability, entitled Règles particulières à la réparation des préjudices résultant de certaines catégories de dommages. This section introduces a non-exhaustive list of the types of dommage for which the tortfeasor might be liable.35 This list includes two types of harm of particular importance to a regime of extra-contractual liability: dommage corporel and dommage matériel. Dommage moral, on the other hand, is no longer given an independent status. Instead, it is redesignated as préjudice extrapatrimonial, which can result, as it

33 See § 823(1) BGB. 34 Art 1235 of the Projet de réforme states: ‘Any certain loss is reparable where it results from harm and consists of an injury to a lawful interest, whether patrimonial or extra-patrimonial’ (‘Est réparable tout préjudice certain résultant d’un dommage et consistant en la lésion d’un intérêt licite, patrimonial ou extrapatrimonial’). 35 Sub-sections 1 to 4 of Section 2 of Chapter IV.

190  Dorota Leczykiewicz seems, from both dommage corporel or dommage matériel and, in each case, following article 1235 of the Projet de réforme will be recoverable. This raises the question about secondary victims suffering purely emotionally. It seems that also in this case their préjudices extrapatrimoniaux will be regarded, as required by article 1235, to have resulted from dommage, ie the physical harm suffered by the primary victim, which would mean that the dommage that makes the victim’s losses recoverable does not have to be personal to them. One could also wonder how the other conditions of articles 1235 and 1236 will apply to the categories of losses expressly mentioned as flowing from the enumerated types of dommage,36 such as le déficit fonctionnel after ‘consolidation’,37 which according to article 1236 Cc, as future loss, should be recoverable only if it is shown to constitute ‘la prolongation certaine et directe d’un état de choses actuel’. As for dommage matériel, articles 1278–79 of the Projet de réforme use this term in a narrower meaning of physical damage to property, rather than the currently prevalent conception, which treats as dommage matériel any interference with the victim’s patrimony. Patrimony (le patrimoine) is defined as a collection of subjective rights, ‘a collection of rights and debts, current or future, in which the rights and the debts balance each other’ (‘un ensemble de droits et de charges, actuels et futurs, dans lequel les droits répondent des charges’).38 As a result, dommage matériel is to be found whenever the victim’s financial interests have been interfered with. This includes out-of-pocket expenses but also lost income or profit lost due to, for example, unfair commercial practices and loss of custom. For this reason, dommage matériel is used interchangeably with préjudice patrimonial. This practice will not be maintained if the project becomes law. This might mean that French law for the first time will expressly restrict recovery for what English lawyers would call ‘pure economic loss’ to cases where the loss stems from personal injury or physical damage to property suffered by the primary victim. However, the list of dommages included in Section 2 of Chapter IV is not exhaustive so it is possible for judges to add to it other, non-codified, types of harm, such as le trouble commercial (interference with commercial interests), a form of pure economic loss.

II.  The Necessary Characteristics of a Recoverable Préjudice I have already mentioned that the new provisions require that préjudice be certain, result from dommage and consist in a violation of a lawful interest. Given the fact that dommage has no conditions attached to it in the new provisions, it seems

36 See above n 16. 37 See art 1270 of the Projet de réforme. On the notion of ‘consolidation’, see below n 77. 38 J Ghestin, G Goubeaux and M Fabre-Magnan, Traité de droit civil, Introduction générale, 4th edn (Paris, LGDJ, 1994) 156–57. See also H Mazeaud, L Mazeaud, J Mazeaud and Chabas (n 21) 403.

Loss and its Compensation in the Proposed New French Regime  191 that only two of the four requirements previously attaching to dommage have been transposed into the proposed regime: certainty and legality. Article 1235 of the Projet de réforme makes no mention of directness or the personal character of the loss.

A.  Certainty of the Loss As to the requirement of certainty, it is not evident what intention the drafters of the proposed provisions had when including it among the conditions of recoverability. One interpretation would be that their intention was to indicate that loss has to be shown or proved, before damages will be awarded for it.39 But if this was the drafter’s intention, wouldn’t it have been simpler to say that? Perhaps it would, but such a statement would create a friction with well-settled judge-made rules according to which acts of unfair competition or violations of the right to a private life are actionable without a proof of loss.40 While it is doubtful that the drafters actually considered this potential friction, or that this was the reason why they used the vague concept of certainty in place of other, more specific requirements, their neglect will not prevent the confusion that the requirement of certainty might create among less informed (foreign) users of the new regime. To the extent that certainty is actually associated with issues of proof, it is only the case law, and not the codified provisions, that will reveal that in French law the requirement of proof is not enforced strictly. Latitude comes not only from the fact that the Cour de cassation may choose to dispense altogether with the requirement of proof in some circumstances, but also from the fact that dommage in principle falls to a large extent within what is described as ‘sovereign assessment’ of the courts of fact – the juges du fond.41 The assessment is described as ‘sovereign’ for two related reasons. The first reason is the absence of a duty to justify the finding of dommage. The starting position is that there is no duty to explain on what grounds the lower courts established dommage, unless this liberty has been

39 One should recognise that in French law the requirement of proof of harm stems much more directly from procedural than substantive law rules. According to art 31 CPC the right of action is available only to those who have a ‘legitimate interest’. It is a widely held view that the intérêt légitime can be established if the claimant shows they have suffered harm. See Cass civ (1) 19 November 1996, RTD civ 1997, 156. 40 The proposed rules include a special provision only on the other situation in which French law does not require a proof of loss, le trouble de voisinage (nuisance caused by a neighbour). Art 1244 of the Projet de réforme states that the property owner is liable de plein droit (ie strictly) for dommage resulting from the trouble de voisinage. However, the provision does not state unequivocally that it creates a presumption of loss and in what way, if at all, art 1235 of the Projet de réforme (and in particular the requirement of certainty) applies to it. 41 The French term is that of l’appréciation souveraine des juges du fond and it indicates sovereign power, ie uncontrolled discretion of lower judges in some matters, considered for this reason to be questions of fact. Issues of law are within the jurisdiction of the Cour de cassation and thus are open to correction by that court if the juges du fond get them ‘wrong’.

192  Dorota Leczykiewicz qualified by the case law of the Cour de cassation. This means that in principle courts may establish the existence of harm even in the absence of evidence. The second reason for describing the mode of establishing dommage by the courts of fact as ‘sovereign assessment’ is that the Cour de cassation has only limited possibilities of intervention. In practice, the Cour de cassation has been interfering quite frequently and extensively, holding for example that a Cour d’appel had incorrectly applied article 1240 Cc by not finding the existence of harm in particular circumstances, even though those circumstances were of a kind in which strict proof of harm was not possible, further relaxing, but only in a casuistic way, the requirement of proof.42 In my view, it was the retention, albeit by reference to préjudice, of the requirement of certainty that triggered the need to regulate separately the question of recovery for future harm in article 1236 of the Projet de réforme. Such a provision would not otherwise be necessary because article 1262 of the Projet, included in ­Chapter IV regulating the effects of liability, in any event obliges judges to take into account in the assessment of damages the reasonably foreseeable ‘evolution’ of the harm (dommage).43 Thus, it would make more sense to drop the requirement of certainty and include in article 1236 a negative rule excluding the recovery for insufficiently probable loss. But that would look too much like a standard of proof and thus create a friction with the loss of a chance provision (article 1238), a friction that English law knows all too well and therefore rejects actionability of the loss of a chance.

B.  Legality of the Loss Violation of a lawful interest as a condition of actionability is not mentioned in any of the provisions laying down the three main principles of extra-contractual liability and is instead located only in the provision dealing with the issue of reparable loss. Regulating loss before all other issues of liability was supposed to indicate that victims were becoming the focus of the new civil liability regime; that the system was oriented towards providing them with a remedy. This view of the new regime can be questioned the moment we observe that proposed article 1235 of the Projet de réforme makes legality a necessary condition of recoverability, and thus, unlike the case law of the Cour de cassation, it incorporates the positive, as well as the negative, dimension of legality. This is because there is a difference between an unwritten principle, which can be applied selectively to preclude recovery of certain losses, and an express legislative requirement with respect to which the burden of proof rests on the claimant. As a result, the wording of a­ rticle 1235 42 See, for example, Cass civ (2) 14 June 2018 no 17-18.503. 43 ‘Damages are assessed as of the day of judgment, taking into account all the circumstances which could have affected the make-up and the value of the loss since the day of the manifestation of the harm, as well as its reasonably foreseeable development.’

Loss and its Compensation in the Proposed New French Regime  193 invites judges to consider the character of the violated interest also in cases where such an inquiry is not needed. Moreover, the inclusion of legality as a necessary condition constitutes a statutory empowerment for judges to regulate recoverability by reference to policy considerations without increasing transparency of the grounds on which such a regulation should take place. As I have shown above, the requirement of legality currently has significance only in two situations. It is used to dismiss the loss of illegal gains as a reparable type of harm and to deny compensation to unworthy claimants, for example criminals bringing an action against a private person who has helped to apprehend them. A much better solution would thus be to separate these two functions and introduce two simple and precise provisions, one excluding recovery for the loss of criminal proceeds or the failure to make illegal profit, and the other empowering the courts to reject on policy grounds actions brought by certain claimants. This would constitute a much more faithful and transparent articulation of the case law rules. In particular, it would be clearer that in a particular case French judges are relying on policy considerations to reject a claim, which would subject their decision-making to more scrutiny.

III.  Préjudices Patrimoniaux and Préjudices Extrapatrimoniaux I have discussed above the distinction which the proposed provisions make between, first, certain and hypothetical losses and, second, between lawful and unlawful interests. I would now like to focus on another distinction to be found in article 1235 of the Projet de réforme, whose significance comes to the surface only if we examine the special provisions on dommage corporel. I am referring here to the distinction between préjudices patrimoniaux and préjudices extrapatrimoniaux. One may wonder why highlighting that both patrimonial and non-patrimonial (or extra-patrimonial) losses are recoverable was necessary in the light of article 1258 of the Projet, which codifies the principle of full compensation. A refusal to compensate non-patrimonial losses would constitute a breach of this principle and thus a violation of article 1258 of the Projet. However, there is one context in which the distinction is highly important. The proposed provisions for the first time set out to regulate the limits of the third-party payers’ right of recourse (le recours des tiers payeurs) by reference to the distinction between préjudices patrimoniaux and préjudices extrapatrimoniaux.

A.  Social Benefits and Third-party Payers One important point which it is necessary to bear in mind when discussing French tort law is its limited significance as a regime of personal injury compensation.

194  Dorota Leczykiewicz After the Second World War, France introduced a comprehensive system of social security, which today takes care of most victims of personal injury, including those whose harm is attributable to a tort.44 A victim bringing an action against the tortfeasor needs to disclose that they are insured.45 In 1985 article L 376-1 was added to the Code de la sécurité sociale, which confirms that a victim of a tort retains the right to bring an action against the tortfeasor even where they have already been (partially) compensated by a social security fund. Equally, the victim continues to be entitled to benefits under social security law even when they could recover from the tortfeasor.46 The same provision also laid down the right of the social security fund to reimburse itself against the damages which the tortfeasor owed to the insured, a victim of a tort (the right of indemnity, or the right of recourse, le recours).47 The part of tort damages against which the social security fund could reimburse itself was designated by the legislator as that compensating ‘injury to the physical integrity of the victim’ (‘l’atteinte à l’intégrité physique de la victime’), referred to in short as dommage corporel.48 However, not all tort damages compensating dommage corporel were available to the social security fund. Within tort damages compensating dommage corporel one had to distinguish damages compensating ‘personal harm’ (préjudice de caractère personnel), ie damages for pain and suffering, the emotional harm stemming from bodily disfigurement, or loss of enjoyment of life (‘correspondant aux souffrances physiques ou morales par elle endurées et au préjudice esthétique et d’agrément’).49 Tort damages compensating ‘personal harm’ were not available to the social security fund and could be awarded exclusively to the victim. Article 30 of the Law no 85-677 of 5 July 1985, referred to as Loi Badinter,50 which covers the right of indemnity against the tortfeasor of not only social 44 According to the Cour de cassation’s judgment, recourse to social security funds is not possible unless the victim has suffered personal injury (dommage corporel). Cass soc 17 January 1962, Bull civ 1962 no 73. 45 According to art L 376-1 Code de la sécurité sociale, in force from 1985 and not changed by the 2017 reform: ‘The person concerned or his dependants must indicate, at all stages of the procedure, that the victim of the accident had the status of a beneficiary of social insurance and to which social security funds the victim is or was affiliated for the various risks. The person concerned or his dependants must notify these funds so that they can be treated as parties to the proceedings. If they have failed to comply with any of these obligations, the substantive judgment may, within two years from the date on which the said judgment became final, be claimed null at the request of the ministère public, the social security funds concerned, or the third parties held liable, if the latter have an interest [in the nullity].’ The 2017 reform added in art L 376-1 a paragraph which obliges the insured to inform the social security fund that someone else is responsible for their harm. 46 Art L 454-1 Code de la sécurité sociale uses the same rules for accidents at work in cases where the employer or their e­ mployees were responsible for the accident. 47 I will refer to these damages as ‘tort damages’, to distinguish them from the compensation that the victim has obtained through third-party payers’ benefits and awards. 48 The wording of art L 376-1 undoubtedly accelerated the evolution of dommage corporel as a ­separate category of harm in French tort law. 49 It is believed that at the time (1985) these were the only types of emotional harm recoverable under the Code civil in the case of personal injury. 50 Loi no 85-677 du 5 juillet 1985 tendant à l’amélioration de la situation des victimes d’accidents de la circulation et à l’accélération des procédures d’indemnisation, JO 6 July 1985.

Loss and its Compensation in the Proposed New French Regime  195 security funds but also other ‘third parties’ (les tiers payeurs), such as insurance companies, provides that the third-party payer claiming indemnity against tort damages is subrogating in an action under civil law the victim. This was taken to have two consequences. First, the third-party payer’s action was not independent and all defences available against the victim were available also in proceedings brought by the third-party payer. Secondly, the damages to which the victim was entitled against the tortfeasor under civil law were to be reduced by the value of the benefits which the victim had obtained from a social security fund.51 This reduction was to affect solely the part of damages repairing ‘les atteintes à son intégrité physique’. The breadth of the third-party payer’s entitlement and the corresponding reduction in the victim’s damages meant that in practice victims of personal injury were not likely to bring civil actions under article 1382 Cc or the other Code civil provisions imposing a private law duty to provide reparation for harm. It is not an exaggeration to say that before the law was changed in 2006, third-party payers’ subrogation entitlements meant that a victim compensated by a social security fund or by an insurance company was effectively losing their right to obtain compensation from the tortfeasor in favour of the third-party payer. This impression is further intensified by the fact that the Cour de cassation had been constantly extending the range of losses that social benefits were taken to compensate. Damages compensating emotional harm (dommage moral) were not as a category regarded as ‘compensation personal to the victim’ (‘indemnité de caractère personnel’). Alas, the Cour de cassation began to separate from tort damages compensating non-physical consequences of personal injury damages compensating ‘objective’ harm (préjudice objectif), and interpreted the term ‘­préjudice de caractère personnel’ as meaning exclusively ‘subjective’ harm.52 This is visible in particular in the context of damages compensating the so-called ‘functional deficit’ (déficit fonctionnel), that is the non-economic aspects of disability, such as its effect on private and social relations, which some lower courts wanted to regard as compensating emotional harm linked to the difficulty of leading e­ very-day life after the injury. This view was rejected by the Assemblée plénière of the Cour de cassation in 2003, which used the distinction between déficit fonctionnel and préjudice d’agrément to hold that a Cour d’appel that excluded the third-party payer’s access to tort damages for ‘functional deficit’ as compensating ‘préjudice moral extrapatrimonial’ (ie ‘personal’ harm) violated article 1382 Cc and the relevant provisions of the Code de la sécurité sociale and of the Loi Badinter.53 This ruling further reduced the amount to which the victim was entitled after the third-party payer had satisfied its claims. In these circumstances, the two types of

51 Ass plén 31 October 1991 no 89-11.514, D 1993, somm 271, JCP G 1992 II 21800 (3rd case), RTD civ 1992, 129. 52 See, for example, Cass soc 5 January 1995, Bull civ 1 no 8. Another meaning in which the term préjudice objectif is used concerns environmental harm, in which case there is no individual person suffering the harm (no subject of the harm). 53 Ass plén 19 December 2003, Bull civ ass plén no 7, RCA 2004, comm 7.

196  Dorota Leczykiewicz harm expressly mentioned by article L 376-1 Code de la sécurité sociale, préjudice esthétique and préjudice d’agrément, as excluded from the third-party payers’ right of recourse grew in significance. Thus, there has been a tendency, especially among French legal scholars, to expand the category of préjudice d’agrément against other categories of emotional harm in order to insure that third-party payers would not be able to put their hands on all of the victim’s tort damages. Article L 376-1 Code de la sécurité sociale and article 31 of the Loi Badinter were revised in 2006.54 The reform introduced a rule that third-party payers are to be allocated the victim’s tort damages poste par poste (‘item by item’), which means that the social security fund in principle should have recourse only to defined parts of the damages to which the victim is entitled under civil law against the tortfeasor. These parts are identified by reference to the items of harm for which the social security fund has taken responsibility (compensated the victim) through the benefits granted to them.55 The reform also confirmed the victim’s right to bring an action against the tortfeasor every time recovery through social benefits has only been partial.56 Furthermore, all damages compensating ‘personal harm’ were excluded from the range of damages which the third-party payer could access under its subrogation right, unless they have established that the benefit they paid to the victim had in fact covered also this type of harm. Also here, however, access to the claimant’s damages was only ‘item by item’.57 Thus, if it so happened that the social security fund compensated a particular item of harm more generously than the court awarding tort damages, the fund could not use the remaining damages, damages compensating other items of harm, to recover the difference.58 In other words, the risk of overestimation for the first time started to rest on the social

54 Loi no 2006-1640, 21 December 2006. L 454-1 was not changed even though it concerned the same issue in the context of work-related personal injury. The omission was corrected in 2007 by the Cour de cassation. 55 Art L 376-1(3) Code de la sécurité sociale states that: ‘The subrogatory recourse of the third-party payers can be claimed only item by item and solely in relation to the parts of the award that correspond to losses that they [the third-party payers] have compensated, to the exclusion of harm of a personal character’. The term poste (‘item’) refers here to a portion of the victim’s damages. The items of harm compensated by the social benefits are referred to in art L 367-1 as préjudices. 56 Art 31 of the Loi Badinter after the 2006 reform and art L 376-1(1), which after the 2006 reform states that ‘When … the injury caused to the person benefiting from social security or his successors in title is attributable to a third party, that person or his successors in title shall retain against the author of the accident the right to claim reparation of the loss caused, in accordance with the rules of the general law, insofar as that loss is not remedied by application of this Book’ (‘Lorsque … la lésion dont l’assuré social ou son ayant droit est atteint est imputable à un tiers, l’assuré ou ses ayants droit conserve contre l’auteur de l’accident le droit de demander la réparation du préjudice causé, conformément aux règles du droit commun, dans la mesure où ce préjudice n’est pas réparé par application du présent livre’). 57 Art L 376-1(5): ‘However, if the third-party payer establishes that he has in fact previously paid the victim a benefit that undeniably compensates for an item of personal harm, his recourse may be exercised in respect of this item of loss.’ 58 This practice was possible before the 2006 reform, to a large extent, due to the case law of the Cour de cassation that permitted lower courts not to specify within the overall award what amount was granted for each of the accepted items of harm. This method of awarding damages was described as ‘l’évaluation globale’.

Loss and its Compensation in the Proposed New French Regime  197 security fund. Moreover, the social security fund could no longer benefit from the generosity of ordinary courts towards victims of personal injury. But the 2006 reform, while generally advantageous to victims, had also a weakness. It blurred the distinction between ‘personal’ and ‘objective’ harm. Previously, damages for ‘préjudice de caractère personnel’ were excluded from third-party payers’ right of recourse. The contentious point was thus whether tort damages awarded under a particular heading compensated only ‘personal’ harm, in which case they were in their entirety unavailable to the third-party payer, or (also) ‘objective’ harm, in which case the third-party payer could claim against them. After the 2006 reform, also damages compensating ‘personal’ harm are available to third-party payers. In order to access them, they need to show that the benefits they provided unquestionably compensated also ‘personal’ harm. The only limitation is that they can get access only to tort damages compensating the item of harm they had previously compensated, and not to the entirety of the tort damages compensating the victim’s ‘personal’ harm.59 This provision was enough to reverse some of the positive changes that the 2006 reform would otherwise bring about. In 2009 the Cour de cassation created a presumption that certain social security benefits compensated also ‘personal’ harm, enabling third-party payers routinely to access also that part of the victims’ tort damages.60 Moreover, because the new wording of article L 376-1 no longer mentions expressly préjudice d’agrément as a type of harm whose damages are excluded from the third-party payers’ right of recourse, the 2009 judgments enable third-party payers to satisfy their claims also from this part of the victim’s award. It is important to make one further observation about the categorisation of harm into ‘personal’ and ‘objective’. If we contrast the 2009 judgments with the 2003 judgment of the Cour de cassation on déficit fonctionnel we are able to see a c­ urious thing – classifying an item of harm as ‘personal’ can, depending on the context, be either beneficial or harmful to the victim’s interest. This depends on whether the courts are using this term by reference to an item of harm to be compensated by tort damages or an item of harm for which a social security fund now seeks reimbursement. Thus, any reformer of the French law of extra-contractual liability, instead of picking on the problem of classification (whether a particular type/item of harm is ‘personal’ or ‘objective’, ‘patrimonial’ or ‘non-patrimonial’) should first decide whether it wants to allow judges to control the extent of third-party payers’ recourse. If they want the issue to be regulated legislatively, they should introduce not a classification of tort damages but a list of benefits for which social security funds can recover from the victim’s damages.61 Classification of tort damages into those 59 See art L 376-1(5) Code de la sécurité sociale. This means that the exclusion of damages compensating ‘personal’ harm from the third-party payers’ right of recourse, preserved in art L 376-1(3), became merely an evidential rule. 60 See, for example, Cass civ (2) 28 May 2009 no 08-16.829; Cass civ (2) 11 June 2009 no 08-16.089. 61 This shows that it may be inappropriate to regulate jointly the right of recourse of social security funds and insurance companies. Only the former, but not the latter, can be ordered to compensate victims under specified headings.

198  Dorota Leczykiewicz against which third-party payers have and do not have recourse is necessary to delineate the outer limits of that right but in order to perform a victim-protective role it has to be supplemented by an exhaustive list of the social security benefits for which the funds can claim against the tortfeasor. Nevertheless, the overall consensus in the early 2000s was that French law should move in the direction of assisting the victims. The solution that had been proposed was to compose a list of different items of harm recoverable as consequences of personal injury. A special group was convened under the leadership of Jean-Pierre Dintilhac, a former President of the second Civil chamber of the Cour de cassation. In 2005, a year before the law regulating the right of recourse of ­third-party payers against the victim’s tort damages was revised, the group published a list, known as ‘la nomenclature Dintilhac’.62 The list covers both primary and secondary victims of personal injury, referred to as ‘direct’ and ‘indirect victims’. Compensatable items of harm are divided into ‘patrimonial’ and ‘non-patrimonial’ (préjudices patrimoniaux and préjudices extrapatrimoniaux). In the case of ­‘indirect victims’ the list distinguishes the situations where the primary victim has died from those where the primary victim has survived. Secondary victims are listed as entitled to compensation for both economic and emotional losses that they have suffered as a result of the primary victim’s personal injury, but this right exists only if they are the primary victims’ proches (‘close ones’). Despite the fact that the list has never succeeded in becoming binding law, it has made a noticeable contribution to the transparency and terminological consistency of the French case law on extra-contractual liability. However, the role of the list in assisting (or regulating) judges determining the extent of the third-party payers’ right of recourse is limited. This is because the list categorises the items of harm recoverable in personal injury cases not by reference to the distinction between ‘personal’ and ‘objective’ harm but, as mentioned above, as either ‘patrimonial’ and ‘non-patrimonial’ losses. And thus, even though the list mentions déficit fonctionnel under the heading of ‘non-patrimonial’ harm, the Cour de cassation in 2010 was able to regard it as ‘objective’ harm, this time in order to justify a more generous measure of damages for préjudice d’agrément.63 This case, and the others I have discussed earlier in this section, demonstrate definitively the futility of the distinction between ‘patrimonial’ and ‘non-patrimonial’ losses.

B.  Changes to the Regime The proposed provisions of the new regime of extra-contractual liability undoubtedly protect the interest of the victims better than article L 376-1 62 The list is available at www.justice.gouv.fr/publication/dacs/consult/20141120-projetannexe.pdf (accessed 6 May 2019). 63 The Cour de cassation condemned a Cour d’appel that awarded very modest damages for ­préjudice d’agrément on account of the fact that the victim’s personal/subjective harm had already been ­compensated under the heading of déficit fonctionnel permanent. Cass civ (2) 2 July 2010 no 09-69.119.

Loss and its Compensation in the Proposed New French Regime  199 Code de la sécurité sociale and article 31 of the Loi Badinter. Article 1276 of the Projet de réforme on the third-party payers’ right of recourse puts the victim in a privileged position vis-à-vis the social security fund or the insurance company in bringing an action or laying claim to damages which can practically be obtained from the tortfeasor. This rule, which was introduced in 2006, put a stop to a practice of giving preference to third-party payers’ interests when the defendant was not able to pay the full amount of the damages awarded against them, for example because of insolvency. A special legislative rule to this effect was required because until 2006 the Cour de cassation accepted partial recovery of the victim’s losses by giving third-party payers such a preference, even though this approach could be seen as inconsistent with the principle of full compensation. The benefit of the new wording envisaged in article 1276 of the Projet is that the link between the victim’s priority to claim against whatever assets the defendant still possesses and the principle of full compensation is expressly made. The interest of third-party payers is, however, enhanced this time at the tortfeasor’s expense because the tortfeasor cannot rely on the victim’s contributory fault to reduce the damages compensating the items of harm which were ‘repaired’ by benefits furnished by third-party payers. The second change which the proposed provisions would introduce is that damages compensating ‘non-patrimonial’ losses will always be off-limits for third-party payers.64 Thirdly, the proposed regime preserves one other innovation introduced into the Code de la sécurité sociale and the Loi Badinter in 2006. Third-party payers’ reimbursement is to take place ‘item by item’ (poste par poste). So far, this has proved to be an imperfect solution to the extent that the compensation which the victim obtains from third-party payers is not divided into the same categories of harm, or is not awarded under the same headings, as the ones used in civil law. This generates a lot of case law on what civil law items of harm particular social benefits compensate so that it can be identified to which parts of the tort damages the third-party payer can lay claim. The proposed provisions do not rectify this deficiency because article 1269 of the Projet de réforme provides only for future regulation of the postes, here better translated as ‘headings’, under which patrimonial and non-patrimonial harm stemming from dommage corporel are to be determined. I have explained above why relying on the distinction between ‘personal’ and ‘objective’ harm or between ‘patrimonial’ and ‘non-patrimonial’ harm is unlikely to produce desirable results, both in terms of legal certainty and better protection of victims, especially if the case law will continue to insist that certain types of loss or harm have a hybrid character. So the different wording of article 1276 of 64 See art 1276(1) of the Projet de réforme: ‘Benefits giving rise to recourse are to be imputed item by item solely to the compensation due by the person liable for the heads of loss for which a third party payer has taken responsibility, to the exclusion of extra-patrimonial losses.’ Note that what is excluded from the third-party payers’ right of recourse is damages compensating ‘extra-patrimonial’, and not ‘personal’, losses/harm.

200  Dorota Leczykiewicz the Projet de réforme will only produce a change if it is accompanied by a legally binding typology of losses, prescribing not only their prima facie recoverability in the case of personal injury but also their categorisation as between préjudices patrimoniaux and extrapatrimoniaux. Article 1269 of the Projet de réforme can be read as presupposing such a solution. It empowers the government to adopt ‘une nomenclature non limitative des postes de prejudices’. The word ‘nomenclature’ suggests nothing more than a list of headings under which damages for personal injury should be awarded. But it is clear that the role of the list can go much beyond that. The itemised character of the list will strengthen the judicial obligation, now expressly imposed by article 1269, to enumerate all losses for which judges wish to award compensation (evaluation poste par poste) and to indicate a concrete amount of damages against each of them.65 Moreover, article 1276 could be interpreted as imposing on judges an obligation to include in the list of recoverable losses all those which are mentioned in the list adopted by the government, as soon as their existence has been shown by the claimant. Moreover, the list may be treated as an implicit authorisation to presume the existence of certain losses in cases of personal injury. Thus, article 1276 and the list whose creation it prescribes regulate not only the form that the compensatory award should take (a list of losses which will be compensated) but also the recoverability of various consequences of dommage corporel. More important for our present purposes is the fact that, similarly to ‘la nomencalture Dintilhac’, the list adopted by the government is likely to divide the recoverable consequences of personal injury into items of ‘patrimonial’ and ‘nonpatrimonial harm’. Thus, it will be the list itself which will determine the character of a particular type of recoverable loss. Damages compensating non-patrimonial losses, so labelled by the government’s list, should thus be treated as excluded from the range of damages to which third-party payers can get access, even if in the light of the Cour de cassation’s pre-existing case law, certain social benefits, such as la rente, compensate also ‘non-patrimonial’ items of harm.

IV.  The Principle of Réparation Intégrale and the Calculation of Damages Another ‘unwritten’ element of the current regime of civil liability in French law is the principle of réparation intégrale, which defines the content of the defendant’s

65 Art 1269 of the Projet de réforme states: ‘Patrimonial and extra-patrimonial losses resulting from personal injury shall be determined item by item following a non-exclusive terminology of items of loss fixed by decree after consulting the Conseil d’Etat’ (‘Les préjudices patrimoniaux et extrapatrimoniaux résultant d’un dommage corporel sont déterminés, poste par poste, suivant une nomenclature non limitative des postes de préjudices fixée par décret en Conseil d’Etat’).

Loss and its Compensation in the Proposed New French Regime  201 duty to provide reparation for the victim’s harm.66 This principle is often translated into English as ‘full compensation’. The name of the principle comes from the Latin term restitutio in integrum, which originally meant the bringing of the parties back to the position in which they had been before the contract, and thus referred to a restitutionary remedy. Over time, restitutio in integrum has come to mean the bringing of the party to the position in which they would have been had the contract been performed or the tort had not been committed.67 Thus, the focus of the principle of réparation intégrale would seem to be on the victim. The object is to ensure that the victim can claim not only sustained losses (damnum ­emergens) but also foreseeable future expenses and lost profits (lucrum cessans),68 and that the victim is compensated for both physical and non-physical harm.69 Moreover, the victim’s entitlements should be determined on the date of the judgment by taking into account all known elements of the harm.70 They remain unaffected by the fact that putting the victim in the position in which they would have been had the damage-creating event had not taken place would entail excessive costs for the defendant, disproportionate to the magnitude of their culpability.71 However, over time, under the influence of the principle prohibiting ­enrichissement sans cause (unjustified enrichment), the principle of réparation intégrale started to acquire a negative dimension, protecting not only the victim but also the defendant. And thus, it became to be interpreted as a principle of ‘exact’ compensation, under which courts needed to ensure that the victim was not ‘enriched’ by the compensatory award.72 This created problems, especially in cases of property damage where, for example, repairs of damaged chattels were carried out by using new or more modern parts.73 Moreover, victims began to be

66 The principle is sometimes regarded as a development of the rule codified in art 1231-2 Cc ­(previously art 1149 Cc), according to which, ‘In general, damages due to the creditor are for [both] the loss that he has incurred and the gain of which he has been deprived, with the following exceptions and qualifications’ (‘Les dommages et intérêts dus au créancier sont, en général, de la perte qu’il a faite et du gain dont il a été privé, sauf les exceptions et modifications ci-après’). See G Viney, P Jourdain and S Carval, Traité de droit civil, Les effets de la responsabilité, 4th edn (Paris, LGDJ, 2017) 154. 67 See Cass civ (2) 20 December 1966, D 1967, 169, Bull civ II no 640; Cass civ (2) 8 April 1970 no 68-13.969. 68 Cass civ (2) 25 February 1987 no 85-16.777. 69 This is true also where the victim is in a vegetative state. See Cass civ (2) 22 February 1995 no 92-18.731. 70 Cass civ (2) 3 March 1982 no 80-15.366. 71 For example, the tortfeasor will be asked to pay the cost of purchase and adaptation of a dwelling for a victim who as a result of their personal injury will have to use a wheelchair and cannot stay in their current house. Cass civ (2) 9 October 1996 no 94-19.763. A victim of property damage is generally entitled to the cost of replacement and not just to the diminution in the value of the property, unless the property had not been used by the victim for a long time before the damage-creating event. Cass civ (2) 14 January 1999, Bull civ II no 14, RTD civ 1999, 142. 72 Cass civ (2) 6 January 1988 no 86-16.192. 73 Here the Cour de cassation ruled that the victim’s damages should not be reduced on account of this ‘enrichment’. See Cass civ (2) 7 December 2006, Bull civ II no 348; Cass civ (2) 11 June 2009 no 08-16.507.

202  Dorota Leczykiewicz denied the right to demand repair where replacement was less expensive.74 From about 2003, a new wording started to appear in judgments of the Cour de cassation concerning non-contractual liability. The principle of réparation intégrale began to be defined as requiring a compensatory award that creates neither loss nor profit for the victim (‘ni perte ni profit pour la victime’).75 This allowed the Cour de cassation to exclude certain losses from the victim’s damages on the ground that the same item of harm has already been compensated by benefits that the victim had obtained from other sources.76 In summary, it can be said that the function of the principle of réparation intégrale has been three-fold. On the one hand, it has enabled the Cour de ­cassation to quash judgments which did not recognise and thus did not compensate a certain item of harm, which the Cour de cassation felt should in principle be recoverable. This applied especially to different forms of dommage moral. On the other hand, the principle has enabled the Cour de cassation to set the measure of damages for different circumstances and introduce more detailed rules as to how particular losses should be assessed. Thirdly, it has allowed the Cour de cassation to regulate the amount of damages available to victims who have been compensated by parties other than the tortfeasor, in particular social security funds. Article 1259 of the Projet de réforme would codify the principle of ­réparation intégrale. However, much more significant seems to be article 1258 of the Projet, which defines as the object of la réparation, taking the form either of une réparation en nature (‘reparation in kind’) or of dommages et intérêts, restoring the victim, as far as possible, to the position in which they would have been if the h ­ arm-causing event had not happened. It also adds that the reparation may not result in the victim incurring any loss, and thus the provision requires full compensation of the victim’s losses. The same provision also precludes the victim from profiting from the tort, which would in principle prohibit over-compensation. There are some obvious inconsistencies between the principle of exact compensation and allowing réparation en nature, which I do not have the space to discuss here. What is worth highlighting in this context is that the desire to offer the victim exactly the combined value of their losses may sit uneasily with the ambition to regulate more densely not only the headings under which awards will be made but also, in the case of some types of losses, how their evaluation or assessment should take place.

74 Cass civ (2) 7 December 1978, Bull civ II no 260; Cass civ (2) 9 July 1981, Bull civ II no 156; Cass civ (2) 18 November 2010 no 09-17.301. 75 See Ass plén 19 December 2003 no 02-14.783, JCP 2004 II 10008 and then, the next time in a tort case, Cass civ (3) 12 January 2010 no 08-19.224. The phrase had been used by the Criminal chamber of the Cour de cassation since the early 1990s. See Cass crim 28 June 1990 no 89-84.650. See also the provision of art 1258 of the Projet de réforme: ‘The aim of reparation is to replace the victim as much as is possible in the situation in which he would have been if the harmful action had not taken place. It must cause him neither a loss nor an advantage’. 76 Ass plén 19 December 2003 no 02-14.783, JCP G 2004 II 10008.

Loss and its Compensation in the Proposed New French Regime  203 As for the manner in which dommage corporel should be compensated, according to the general provision on damages, applying to all types of harm, damages should be evaluated separately for each chef de préjudice (article 1262 of the Projet de réforme). This excludes the possibility of awarding a general sum (general damages), not divided into individual awards linked to the specified types of harm and the losses flowing from them. In the case of dommage corporel, the special rules go further because article 1269 of the Projet obliges judges to use the (non-exhaustive) list of patrimonial and non-patrimonial losses to be found in a decree of the government. The provision says that this list should be used to ‘determine’ les préjudices which, as mentioned above, could be interpreted as meaning that the list regulates only the headings under which the different losses suffered by victims of personal injury should be established. However, it is probably expected that judges will use the same ‘headings’ (postes) to evaluate and award damages. The evaluation of certain préjudices stemming from personal injury is regulated in the proposed provisions of the Code civil separately. And so article 1270 of the Projet deals with ‘functional deficit’ before consolidation,77 which according to the proposed rules should be evaluated with the use of a single but tentative medical scale, to be adopted by an executive act. With regard to non-patrimonial losses, the government has been entrusted with the task of selecting the items that can be evaluated with the aid of a ‘repository’ of damages awards (un référentiel indicatif d’indemnisation). This repository will collect information about judgments of the Cours d’appel awarding damages for non-patrimonial harm in cases of personal injury caused by traffic accidents.78 This means that the proposed provisions envisage not only a way in which judges would be regulated more closely in how they assess damages in cases of personal injury but also a way of keeping a record of their decisions and identifying any emerging trend. This is likely to restrict the autonomy of lower courts and encourage some consistency across the compensatory awards. But I wonder if these provisions do not constitute to a much greater extent the curbing of the power of the Cour de cassation to define the headings under which compensation is to be awarded, to create new headings to expand the scope of liability, to define the nature of the harm compensated under a particular heading, and to exploit the principle of réparation intégrale to increase or decrease the damages awarded to the claimant. This, much more than the codification of the pre-existing case law, would contribute to greater predictability of outcomes in tort cases and could implement a more victim-friendly policy into French tort law.

77 ‘Consolidation’ describes a moment when the victim’s personal injury becomes stable, ie the harm has stopped evolving and can be assessed decisively and in full. Whether ‘consolidation’ has occurred is determined by medical experts. Once the victim’s condition is considered stable it is possible to determine which of their harms were only temporary and which will be permanent. Thus, when the victim decides to bring an action before their harms have ‘consolidated’, the damages awarded will be only provisional. 78 It is not clear why information is to be gathered only about traffic accidents and not about other events causing personal injury for which courts award damages.

204  Dorota Leczykiewicz

V. Conclusion The authors of the reform consider it a significant weakness of the current French regime of extra-contractual liability that so much of it is only found in the case law. The law is perceived as lacking transparency, violating the rule of law, and being confusing. The remedy seemed self-evident – codification of judge-made rules. If judge-made rules were translated into legislated provisions, French tort law would instantly seem less esoteric and would regain the admiration it once had, the drafters seem to have believed. Moreover, it would be seen as progressive, encompassing rules on liability for environmental harm and offering increased protection for victims of personal injury. I am not sure if these ambitions have been successfully realised in the proposed provisions on civil liability. As to deficiencies in transparency, in my view this problem has less to do with the fact that the current law is based on case law and more with the rudimentary character of the judgments that lay down new rules, especially those of the Cour de cassation. A change on this front would require a shift in the j­ udicial culture that no substantive provisions can by themselves instigate. The Projet de reforme is thus mainly reactive. The reformers may hope that by introducing a denser web of rules, they will reduce the scope for judicial creativity and increase legal certainty and transparency. But the reform cannot be regarded as well-conceived if there is a lack of clarity about such a fundamental issue as what makes the victim’s harm actionable and how the requirement that le préjudice stem from un dommage will be used to regulate recoverability of the victim’s losses. Transparency is not enhanced when seemingly neutral provisions setting out and regulating certain types of dommage in fact introduce a presumption of recoverability in favour of particular losses, and when provisions on le préjudice exclude some losses from the range of what is recoverable by reference to such open-ended concepts as certainty and legality of the violated interest. It is clear that when it comes to the assessment of damages, the proposed rules try to eliminate both the arbitrariness and the resulting inconsistencies plaguing personal injury cases by binding judges to lists and scales. The power of the Cour de cassation is also circumscribed, although it will be interesting to observe how that court will treat the lists adopted by the executive. The proposed rules on the calculation of the compensatory award are thus likely to increase convergence among lower courts’ judgments, and increase the protection their rulings currently enjoy from review by the Cour de cassation. However, the question of whether a particular unmentioned type of harm or an unlisted item of loss are actionable and recoverable will continue to be regulated by the casuistic case law of the Cour de cassation.

10 The Concepts of ‘Harm’ in the French and Italian Laws of Civil Liability PIETRO SIRENA

I.  The French Distinction between Dommage and Préjudice The Projet de réforme de la responsabilité civile of March 2017 seeks to provide a certain recognition of the distinction between dommage (‘harm’) and préjudice (‘loss’), which may as a result have a major impact on the future development of French tort law.1 At present, the two terms dommage and préjudice,2 which correspond to those of damnum and praeiudicium in Roman law, seem to be used interchangeably by French courts and most scholars.3 Since the 1950s, however, there has been a strand in legal scholarship that seeks to differentiate between them sharply when addressing issues of civil liability.4 This approach is intended not only to attain an allegedly higher level of precision and appropriateness in legal language but also, and more importantly, to regulate, if not to restrain, the relentless and sometimes chaotic growth of compensation for harm that has tended to characterise French law.5 On the other hand, there is a widespread view which, while not denying the soundness of the distinction, plays down its value, considering that it is merely a matter of terminology, or style of reasoning, and could prove unnecessary, as

1 Most of the debate is thoroughly discussed in J-S Borghetti, ‘Les intérêts protégés et l’étendu des préjudices réparables en droit de la responsabilité civile extra-contractuelle’ in Études offertes à ­Geneviève Viney (Paris, LGDJ, 2008) 145. 2 For a comparative assessment of such concepts, see T Weir, ‘La notion de dommage en responsabilité civile’ in Pierre Legrand Jr (ed), Common law d’un siècle l’autre (Cowansville, Yvon Blais, 1992) 1. 3 See M Bacache, ‘La recodification des principes classiques – Articles 1235 à 1238; 1241 à 1249; 1253 à 1256’ JCP G Suppl to nos 30–35 (25 July 2016) 20; F Leduc, ‘Faut-il distinguer le dommage et le préjudice?: point de vue privatiste’ RCA 2010, 3, dossier no 3. 4 See G Viney, P Jourdain and S Carval, Traité de droit civil, Les conditions de la responsabilité civile, 4th edn (Paris, LGDJ, 2013) [246-1]. 5 P le Tourneau, Droit de la responsabilité et des contrats, 11th edn (Paris, Dalloz, 2017) [1304]ff.

206  Pietro Sirena well as criticising it on the ground that it has no basis in the provisions of the Code civil.6 In fact, in making provision for the compensation of harm, article 1240 of the Code civil as presently in force mentions only dommage and does not require proof of any préjudice as a further occurrence; it is, therefore, dommage as such which is recoverable, insofar as it is ‘personal, direct and certain’. This detracts from the assumption that the purpose of civil liability must be that of providing compensation for the losses sustained by the victim and that any award of damages must not result in a gain for her.7 On the contrary, according to article 1235 of the Projet de réforme, what is recoverable by the victim is the préjudice certain resulting from the dommage which he suffers. After the reform, dommage would therefore still be required for civil liability to be imposed, as well as, pursuant to article 1239 of the Projet, a causal relationship between this harm and ‘the action attributed to the defendant’. Yet in addition a ‘préjudice’ would also be required in order to trigger civil liability stricto sensu, that is, an award of damages by way of compensation. As it lacks any proposed legislative definition, dommage may prospectively be identified in any ‘harm’ caused by the tortfeasor to someone else. By contrast, ‘préjudice’ is expressly defined by article 1235 of the Projet de réforme as ‘an injury to a lawful interest’.8 In this way, the Projet does not envisage merely a shift in terminology (from compensation of a dommage to that of a préjudice) nor a neutral change from one concept to another. Instead, it lays down an additional requirement for the imposition of civil liability beyond those provided for by the Code civil as now in force. Indeed, at a conceptual level, dommage would contrast with préjudice, the latter resulting logically (if not chronologically) from the former; in other words, a dommage would not be recoverable in itself but only insofar as it is logically (if not chronologically) followed by a préjudice.9 Furthermore, dommage would be treated as a matter of fact (quaestio facti), whose assessment by courts of first instance and of appeal (together, the juges du fond) could not be scrutinised by the Cour de cassation, except where their decision-making is devoid of a proper justification; by contrast, the assessment of

6 Viney, Jourdain and Carval (n 4) [246-1]. 7 See notably P Malaurie, L Aynès and P Stoffel-Munck, Droit des obligations, 10th edn (Paris, LGDJ, 2018) 143ff. According to art 1258 of the Projet de réforme, ‘La réparation a pour objet de remplacer la victime autant qu’il est possible dans la situation où elle se serait trouvée si le fait dommageable n’avait pas eu lieu. Il ne doit en résulter pour elle ni perte ni profit’. 8 ‘[L]a lésion d’un intérêt licite’. For a comparative overview, see C von Bar, The Common European Law of Torts. Damage and Damages, Liability for and without Personal Misconduct, Causality, and Defences (Oxford, Clarendon Press, 2000) vol 2, 1; H Koziol, Basic Questions of Tort Law from a Germanic Perspective (Wien, Jan Sramek Verlag, 2015) 17ff; C van Dam, European Tort Law (Oxford, Oxford University Press, 2013) 167ff. 9 However, the strictness of such a proposition may be qualified, as it is agreed that a person claiming damages would not bear the burden of proof in relation both to dommage and to préjudice. To this effect see Borghetti (n 1) 150 fn 17 and 154 fn 34, who claims, moreover, that the proof of dommage might provide a ground for presuming a resulting préjudice.

The Concepts of ‘Harm’ in the French and Italian Laws of Civil Liability  207 préjudice would be treated as a matter of law (quaestio iuris)10 on which the Cour de cassation is the final arbiter and thereby competent to decide which losses are or are not recoverable (unless as regards matters covered by EU law, where this role is played by the Court of Justice of the EU,11 as in the case of compensation of ‘non-material damage resulting from the non-performance or improper ­performance of the services constituting a package holiday’).12

II.  The Italian Distinction between Danno-evento and Danno-conseguenza Article 2043 of Italian civil code (Codice civile, or Cod civ) sets the basic rule of civil liability, stipulating that: Any action which deliberately or negligently causes an unlawful harm to another person, obliges the person who committed that action to compensate the harm.13

This provision of the Codice civile is centred on the concept of danno ingiusto (‘unlawful harm’),14 which stands apparently both as a factual requirement of the delict and as harm to be compensated.15 However, since the 1960s, Italian scholars have sought to distinguish two notions of danno and to differentiate between them in their terminology: on the one hand, there is the unlawful harm inflicted on the claimant (referred to as danno-evento); on the other hand, there is the loss that she (or other victims, such as her relatives) sustained as a result (referred to as danno-conseguenza).16 Following this distinction, under article 2043 Cod civ the occurrence of danno-evento, ie an unlawful harm, is seen as a requirement of establishing a delict as such, whereas danno-conseguenza, ie a consequential

10 Borghetti (n 1) 153. 11 D Leczykiewicz, ‘Compensatory remedies in EU law: The Relationship between EU Law and National Law’ in P Giliker (ed), Research Handbook on EU Tort Law (Cheltenham, Edward Elgar Publishing, 2017) 63. 12 Case C-168/00 Simone Leitner v TUI Deutschland GmbH & Co KG [2002] ECR I-2631. See A Johnston, ‘“Spillovers” from EU Law into National Law: (Un)intended Consequences for Private Law Relationships’ in D Leczykiewicz and S Weatherill (eds), The Involvement of EU Law in Private Law Relationships (Oxford, Hart Publishing, 2015) 362. 13 ‘Qualunque fatto doloso o colposo, che cagiona ad altri un danno ingiusto, obbliga colui che ha commesso il fatto a risarcire il danno’. On this see PG Monateri, ‘La responsabilità civile’ in R Sacco (ed), Trattato di diritto civile (Torino, UTET, 1998) 195ff; M Franzoni, ‘Fatti illeciti. Art. 2043, 2­ 056-2059’ in F Galgano (ed), Commentario del Codice Civile Scialoja-Branca (Bologna, Zanichelli-Il Foro Italiano, 2004) 173ff. 14 C Castronovo, Responsabilità civile (Milano, Giuffrè, 2018) 129ff. For a general and in-depth overview of doctrinal contentions on the topic, see L Buonanno, ‘Rilievi civilistici in tema di ­ ­responsabilità del legislatore da atto normativo’ (2016) 5 Jus civile 369. 15 S Patti, ‘Danno patrimoniale’ in FD Busnelli and S Patti, Danno e responsabilità civile (Torino, Giappichelli, 1997) 3. 16 For references, see CM Bianca, Diritto civile. La responsabilità, 2nd edn (Milano, Giuffrè, 2012) vol 5, 584ff.

208  Pietro Sirena loss, is a further requirement of civil liability stricto sensu, that is, compensation by way of damages. To award compensation, there are therefore two requirements: first, that the defendant inflicted an ‘unlawful harm’ on the victim (danno-evento) either intentionally or negligently; and, secondly, that a consequential loss ­(danno-conseguenza) resulted from such an ‘unlawful harm’. Only on satisfaction of both conditions (danno-evento and danno-conseguenza) is compensation awarded. On the other hand, where only the first requirement (danno-evento) is met, other remedies (for example, self-help or injunction) may possibly be granted to the victim of the delict, but civil liability stricto sensu, ie compensation, does not (at least yet) arise. The basis of this restrictive approach is generally traced to the fundamental function of civil liability, which is considered not as serving the purposes of punishment of the wrongdoers or the deterrence of wrongs, but rather compensation of losses actually incurred by the victims of the wrongs. In other words, it is generally assumed that liability arises from a wrong only insofar as there is evidence that it actually entails a loss sustained by the victim. The legal basis of this restriction is allegedly found in article 1223 Cod civ, which stipulates that: ‘Compensation of harm for non-performance or delay shall encompass both the actual loss and the loss of profits, provided that they constitute an immediate and direct consequence’. Notwithstanding some minor amendments, this provision is a literal translation of former article 1151 Cc (corresponding to article 1231-4 Cc in force).17 Furthermore, the distinction between danno-evento and ­danno-conseguenza has led Italian law to differentiate between two types of causal nexus,18 ­mirroring to some extent the German approach to causation.19 To constitute a tort, d ­ anno-evento must be caused to the victim by the defendant’s conduct. This first causal nexus is depicted as ‘factual’ (or ‘natural’) and is termed in German law haftungsbegründende Kausalität, ie the causation which initiates civil liability. If this causal nexus is lacking, the court will conclude that no delict was committed at all and this will end the dispute. If, instead, the claimant manages to prove that the defendant’s intentional or negligent conduct caused the danno-evento, she will have then to demonstrate a further causal link between this danno-evento and d ­ anno-conseguenza, that is between the unlawful harm and her compensatable loss.20 This second nexus of 17 The wording and the position of art 1223 Cod civ (and a similar consideration applies to the French provision cited in the text above) clearly indicate that it relates to contractual liability. Nevertheless, art 2056, 1st paragraph, Cod civ expands the scope of art 1223 Cod civ to extra-contractual liability, as it does for other similar provisions regarding contractual liability. In this way, a single regime of compensation is provided which, apart from some minor variations, is unified for both contractual and extra-contractual liability. 18 The ground-breaking work was done by G Gorla, ‘Sulla cosiddetta causalità giuridica: “fatto dannoso e conseguenze”’ (1951) Rivista di diritto commerciale, I, 405ff. For an overall discussion of the topic, see recently A Belfiore, ‘Il binomio “causalità giuridica-causalità materiale” e i criteri di determinazione del danno da risarcire (artt. 1223 e 2056 cc)’ (2017) Europa e diritto privato 117ff. 19 See notably K Larenz and C-W Canaris, Lehrbuch des Schuldrechts, II/2, Besonderer Teil, 13th edn (München, CH Beck, 1994) § 75 I 2 e. 20 For an in-depth comparative discussion, see S Steel, Proof of Causation in Tort Law (Cambridge, Cambridge University Press, 2015) 15ff.

The Concepts of ‘Harm’ in the French and Italian Laws of Civil Liability  209 causality is depicted as ‘juridical’ and constitutes a necessary condition of liability. For this reason, under German law this second causal nexus is referred to as the haftungsausfüllende Kausalität, ie the causation which completes civil liability.

III.  The Italian Doctrines of Danno Biologico and of Danno Esistenziale Article 2059 Cod civ provides that ‘non-patrimonial harms are to be compensated solely in those cases where it has been so determined by legislation’.21 The wording of this provision basically corresponds to § 253, Abs 1, BGB. Compensation for this kind of harm is granted by article 185 of the Italian Criminal Code in cases of criminal offences. Apart from this situation and a few others across Italian legislation, the wording of article 2059 Cod civ would unmistakably lead to the rejection of any other claim for compensation of nonpatrimonial losses. However, the determination of courts and scholars to override this restriction generated one of the most interesting developments of the I­talian law of tort, namely the so-called danno biologico (literally, ‘biological harm’),22 later on followed by the danno esistenziale (literally, ‘existential harm’, but better translated as ‘amenity harm’).23 From the 1980s, in fact, Italian courts,24 with the support of a part of legal scholarship, began to react against a strict application of article 2059 Cod civ to personal and psychological injury.25 The issue arose especially as regards car crashes, where it is generally hard to prove the negligence of the drivers involved, with the result that none of them is generally convicted of a criminal offence. Even though in cases of accidents the fault of car drivers is presumed by article 2054 Cod civ, and even though motor vehicle insurance is compulsory, article 2059 Cod civ apparently barred any claim for compensation for personal and psychological injuries and, by doing so, it threatened the right to health enshrined in article 32 of the Italian Constitutional Charter. As a result, article 2059 Cod civ was brought 21 ‘Il danno non patrimoniale deve esserre risarcito solo nei casi determinati dalla legge’. See generally Bianca (n 16) 189ff. 22 Castronovo (n 14) 157ff; FD Busnelli, ‘La parabola della responsabilità civile’ in Busnelli and Patti (n 15) 119. 23 P Ziviz, Il danno non patrimoniale. Evoluzione del sistema risarcitorio (Milano, Giuffrè, 2011). 24 Ground breaking Trib Genova 25 May 1974 [1975] Giurisprudenza italiana II 54; Trib Genova 30 May 1974 [1975] Responsabilità civile e previdenza 416; Trib Pisa 10 March 1979 [1979] Responsabilità civile e previdenza 356; Trib Pisa 4 April 1981 [1982] Responsabilità civile e previdenza 783. The jurisprudence of the Tribunals of Genova and of Pisa was later on upheld by Cass civ (3) 6 June 1981 no 3675 [1981] Foro it I 1884; Cass civ (2) 6 April 1983 no 2396 [1983] Responsabilità civile e previdenza 760; Cass civ (2) 14 April 1984 no 2422 [1984] Responsabilità civile e previdenza 333; Cass civ (3) 11 February 1985 no 1130 [1986] Giurisprudenza italiana I 786. 25 Ground breaking R Scognamiglio, ‘Il danno morale (contributo alla teoria del danno extracontrattuale)’ (1957) Rivista di diritto civile, I, 227ff; idem, ‘Appunti sulla nozione di danno’ (1969) Rivista trimestrale di diritto e di procedura civile 464.

210  Pietro Sirena before the Constitutional Court, it being alleged that it was ­detrimental to a fundamental human right. While the Constitutional Court resolutely refrained from annulling this legislative provision and dismissed the claim, in doing so it nonetheless accepted two major propositions that radically changed Italian law.26 First, the Constitutional Court referred to article 2059 Cod civ as dealing not with a type of harm (danno-evento) but with a type of loss (­ danno-conseguenza), consisting of any kind of pain and suffering, emotional distress, etc (danno morale (soggettivo)). Secondly, and conversely, personal and psychological injuries should be classified not as a type of loss (­danno-conseguenza) but as a type of harm (danno-evento), therefore amounting to an ‘unlawful harm’ in the sense of article 2043 Cod civ. While this interpretation was deemed ­necessary at the time to pave the way for compensation for danno biologico, it was subsequently abandoned by the Corte costituzionale (as well as by the Corte di cassazione), which between 1994 and 2003 reconceptualised danno biologico as a danno-conseguenza (a type of loss, not a type of harm).27 This change took place because the earlier approach to danno biologico as a type of harm apparently endangered the compensatory function of Italian tort law while enhancing its punitive dimension, insofar as compensation by means of damages was granted to the victim irrespectively of any consequential loss she might have incurred. In this way, civil liability could appear to serve the purpose more of punishing the tortfeasor for having committed a wrong rather than the purpose of putting ‘the injured party in the same condition he would have been in had he not sustained the wrong’.28 Committed as they were to preserving the compensatory nature of civil liability and to barring the award of punitive damages,29 the courts had to reinterpret danno biologico as a danno-conseguenza. A brand new doctrine was thus established, whereby, in cases not covered by a statute, personal (and psychological) injuries are recoverable under Italian law because the constitutional status of the right to health commands compensation, despite the terms of article 2059 Cod civ.30 In other words, this provision of the Codice civile must be interpreted in

26 Corte Cost 14 April 1986 no 184 [1986] Nuova giurisprudenza civile commentata I 534 note G Alpa; [1986] Foro it I 2976 note PG Monateri; [1986] Responsabilità civile e previdenza 533 note G Scalfi. 27 Corte Cost 27 October 1994 no 372 [1995] Nuova giurisprudenza civile commentata I 406 note P Ziviz; [1995] Giurisprudenza italiana I 406 note A Jannarelli; [1995] Foro it I 3297 note G Ponzanelli. 28 Livingston v Rawyards Coal Co (1880) 5 UKHL App Case 25. 29 See the essays collected in P Sirena (ed), La funzione deterrente della responsabilità civile. Alla luce delle riforme straniere e dei Principles of European Tort Law (Milano, Giuffrè, 2012); A Zoppini and M Maugeri (eds), Funzioni del diritto privato e tecniche di regolazione del mercato (Bologna, il Mulino, 2010). More recently, see among others FD Busnelli, ‘Deterrenza, responsabilità civile, fatto illecito, danni punitivi’ (2009) Europa e diritto privato 2009, 909, as well as F Quarta, Risarcimento e sanzione nell’illecito civile (Napoli/Roma, Edizioni Scientifiche Italiane, 2013). 30 FD Busnelli, ‘Problemi di inquadramento sistematico del danno alla persona’ in Busnelli and Patti (n 15) 41.

The Concepts of ‘Harm’ in the French and Italian Laws of Civil Liability  211 conformity with the Constitutional Charter, and particularly with its article 32,31 as advocated by the doctrine of the so-called mittelbare Drittwirkung of constitutional provisions (enshrining human rights).32 In the light of these developments, the Italian notion of danno biologico now corresponds to the French notion of préjudice/déficit fonctionnel or préjudice physiologique and poses similar questions of assessment as are addressed by article 1270 of the Projet de réforme. In the 1990s, a similar development occurred in relation to injuries that do not affect the victim’s health as such, but do impair her ability to undertake sporting, artistic or leisure activities. It was thus acknowledged that a so-called danno esistenziale (‘amenity loss’) is recoverable, which to some extent corresponds to the French préjudice d’agrément as defined by the Assemblée plénière of the Cour de cassation in its judgment of 19 December 2003.33 To start with, the award of compensation for this type of harm was based by first instance courts on the violation of an alleged right to happiness in individuals;34 inevitably, this conceptual framework induced them to uphold claims which were sometimes negligible (non-pecuniary damages for a bad haircut, for loss of leisure, etc).35 Later on, the Italian Court of cassation ‘normalised’ compensation for danno esistenziale in the sense that, following a position adopted by legal scholars,36 it is now awarded only provided that the non-pecuniary loss alleged (and proved) by the claimant is due to the violation of a human right protected by the Italian Constitution.37 It is settled law that, following the general rule in article 2697(1) Cod civ, the burden of proof of any loss to be compensated shall lie with the claimant and so no normative, or in re ipsa, damages are awarded for the mere commission of the wrong itself.38 However, where there is evidence that a human right enjoying 31 Corte Cost 11 July 2003 no 233 [2004] Nuova Giurisprudenza civile commentata I 13 note A Thiene; [2003] Danno e responsabilità 939 note G Ponzanelli; [2003] Corriere giuridico 1017 note M Franzoni; [2003] Giurisprudenza italiana 1777 note P Cendon and P Ziviz. 32 In Italian literature, see lately P Femia (ed), Drittwirkung: principi costituzionali e rapporti tra privati: un percorso nella dottrina tedesca (Napoli, Edizioni Scientifiche Italiane, 2018); E Navarretta (ed), Effettività e Drittwirkung, vols 1-2 (Torino, Giappichelli, 2018). 33 Ass plén 19 december 2003 (2003) 8 Bull AP 21, JCP 2004 II 1008 note P Jourdain and I 163, n 32ff note G Viney; D 2004, 161 note Y Lambert-Faivre; RCA chr 2004, 9 note H Groutel; LPA 2004, [73]-[74] note by M-F Feuerbach-Steinle; RTD civ 2004, 300 note P Jourdain. 34 For references, see P Cendon and P Ziviz (eds), Il danno esistenziale: una nuova categoria della responsabilità civile (Milano, Giuffrè, 2000); P Cendon (ed), Trattato breve dei nuovi danni. Il ­risarcimento del danno esistenziale: aspetti civili, penali, medico-legali, processuali (Padova, CEDAM, 2001). 35 G Ponzanelli, ‘Le “pericolose” frontiere della responsabilità civile: il caso dei danni da black-out elettrico’ (2006) Danno e responsabilità 54. 36 Above all E Navarretta, Diritti inviolabili e risarcimento del danno (Torino, Giappichelli, 1996). 37 Cass civ, sez un, 11 November 2008 nos 26972-26973-26975 [2009] Giustizia civile I 913 note M Rossetti; [2009] Foro it I 120ff notes A Palmieri, R Pardolesi, R Simone, G Ponzanelli, E Navarretta; [2009] Nuova giurisprudenza civile commentata I 102 notes E Bargelli and F Di Marzio; [2008] Diritto e giurisprudenza 526 note S Patti. For a wider account, see C Scognamiglio, ‘Il sistema del danno patrimoniale dopo le decisioni delle Sezioni Unite’ (2015) Responsabilità civile e previdenza 261; for comparative law analysis, see G Christandl, ‘Das italienische Nichtvermögensschadensrecht nach 2008 – eine Lektion für Europa?’ (2011) ZEuP 392. 38 Lastly Cass civ (1) 8 January 2019 no 207, (2019) 6 Guida al dir 60.

212  Pietro Sirena constitutional status has been infringed, this is deemed by Italian courts to provide a suitable basis for a presumption of a corresponding non-pecuniary loss, in the absence of proof to the contrary. Even where the occurrence of a loss has been proven, a claimant seeking compensation must still prove its quantum, unless this proof is impossible or excessively difficult; in fact, article 1226 Cod civ provides that ‘if damages cannot be determined in their exact amount, they shall be assessed by the judge on an equitable basis’. By their nature, this is mostly the case as regards non-pecuniary losses, since they are not measurable in an objective way (with the exception of personal and, possibly, psychological injuries, for which an expert evaluation is suitable). As a result, non-pecuniary losses are mostly assessed by the courts on an equitable basis, pursuant to article 1226 Cod civ.39

IV.  The Concept of ‘Harm’ in the French and Italian Laws of Civil Liability Unlike its French counterpart, the Italian law of civil liability depends on whether the harm caused is ‘unlawful’, a requirement which features prominently in the basic rule stipulated by article 2043 Cod civ, as earlier noted.40 In fact, in Italian law a tort is committed if and only if a danno ingiusto (‘unlawful harm’) is caused to someone else.41 Italian scholars tend to remark on the novelty and the originality of this legislative formula, which is often depicted as a sort of third way between the French approach to liability for fault under a ‘general clause’ and the German approach to it as the infringement of one of the rights enumerated in a legislative list, often termed a numerus clausus.42 At any rate, until the late 1970s, Italian courts identified the unlawfulness of damage in the infringement of ownership (dominium), or of a real right to possess someone else’s property (ius in re aliena, as in the case of usufruct, servitude, or easement). However, even as early as the 1960s, some scholars had advocated the expansion of the range of danno ingiusto beyond such limits.43 And in a pivotal decision in 1971, the Court of cassation 39 E Navarretta (ed), Il danno non patrimoniale: principi, regole e tabelle per la liquidazione (Milano, Giuffrè, 2010); ead (ed), I danni non patrimoniali: lineamenti sistematici e guida alla liquidazione (Milano, Giuffrè 2004); G Ponzanelli, ‘Il quantum del risarcimento del danno non patrimoniale’ (2005) Danno e responsabilità 126; V Ricciuto and V Zeno-Zencovich, Il danno da mass-media: elementi per la valutazione e criteri di liquidazione (Padova, CEDAM, 1990). 40 Above text to fn 13. 41 See also above, section II. 42 R Sacco and P Rossi, Introduzione al diritto comparato, 6th edn (Torino, UTET, 2015) 93ff; FD Busnelli, ‘Frontiere europee della responsabilità civile’ in Busnelli and Patti (n 15) 159. For an account of the wider influence of the Italian model of civil liability in other jurisdictions, see M Infantino, ‘Diffusing Law Softly: Insights into European Travels of Italian Tort Law’ (2015) 6 JETL 260. 43 P Schlesinger, ‘La ingiustizia del danno nell’illecito civile’ (1960) Jus 344ff; R Sacco, ‘L’ingiustizia di cui all’art. 2043 cc’ (1960) Foro padano I 1420ff; S Rodotà, II problema della responsabilità civile (­Giuffrè, Milano, 1964).

The Concepts of ‘Harm’ in the French and Italian Laws of Civil Liability  213 held that the killing of a professional football player in a car accident constituted a danno ingiusto for his club, with the result that the latter was entitled to compensation for the loss of the future services of the deceased player.44 This meant that the violation of a right in personam (here, the club’s right under its contract with the player) could satisfy the unlawfulness requirement,45 provided that such a violation was caused by the defendant’s fault.46 Until the late 1990s, a typical case in which the requirement of unlawfulness was seen as not satisfied was where harm was caused by a public official or a public body in the course of carrying out an administrative task involving the issuing of an authorisation, permission or similar enabling provisions to citizens, as in the case of public tenders and other procedures of public procurement, state exams, or permissions to build. The traditional approach of Italian law here was that a citizen does not have a right to such an authorisation or permission, but only a legitimate interest (interesse legittimo) in preventing officials and public bodies from abusing their discretionary powers; therefore, she could not recover for any losses consequential on their refusal of such an authorisation or permission. However, in 1989 the European Directive on public supply and public work contracts compelled Italian law to award compensation of this type47 and eventually the Court of cassation changed its position on the requirement of unlawfulness in the administrative context in order to make good any harm deliberately or negligently caused by a public ­official or a public body in the course of carrying out any administrative task. This change in case law was marked by the judgment of the Court of cassation of 22 July 1999 no 500,48 where it was decided that there is a danno ingiusto whenever there is an impairment of legally protected interests, whatever their formal categorisation (most notably, regardless of whether they are classified in terms of interesse ­legittimo or diritto soggettivo).49 This formula proved flexible enough to cover harm in relation to which Italian courts had already awarded compensation – as happened since 1994 in the ‘widow cases’ between partners who were not married50 – to allow

44 Cass civ, sez un, 26 January 1971 no 174 [1971] Giurisprudenza italiana I 680 note G Visintini, [1971] Foro it 342, 1284 notes by AC Jemolo and by FD Busnelli (so called ‘Meroni case’). 45 On which see FD Busnelli, La lesione del diritto di credito da parte di terzi (Milano, Giuffrè, 1964). 46 For an argument that liability here should be limited to cases of the murder of dependants, workers and partners, as well as to any intentional torts, see Bianca (n 16) 608. 47 Council Directive 89/665/EEC of 21 December 1989 on the coordination of the laws, regulations and administrative provisions of review procedures to the award of public supply and public work contracts. 48 Cass civ, sez un, 22 July 1999 no 500. 49 ibid [9]: ‘Una volta stabilito che la normativa sulla responsabilità aquiliana ha funzione di ­riparazione del “danno ingiusto”, e che è ingiusto il danno che l’ordinamento non può tollerare che rimanga a carico della vittima, ma che va trasferito sull’autore del fatto, in quanto lesivo di interessi giuridicamente rilevanti, quale che sia la loro qualificazione formale, ed in particolare senza che assuma rilievo determinante la loro qualificazione in termini di diritto soggettivo, risulta superata in radice, per il venir meno del suo presupposto formale, la tesi che nega la risarcibilità degli interessi legittimi quale corollario della tradizionale lettura dell’art. 2043 c.c.’. See FD Busnelli, ‘Lesione di interessi legittimi: dal “muro di sbarramento” alla “rete di contenimento”’ in Busnelli and Patti (n 15) 85. 50 Cass civ (3) 28 March 1994 no 2988 [1994] Giustizia civile I 1849.

214  Pietro Sirena a gradual broadening of the scope of recoverable losses, and to adapt the law of civil liability to interests in the meanwhile acknowledged in the social context. In this way, courts have been induced to sort and to classify these developments rationally and to refrain from idiosyncratic or bizarre case-by-case judgments (no piecemeal approach). A similar notion of harm (there termed ‘damage’) is to be found in the basic rule on civil liability provided by article VI-1:101(1) DCFR, which reads: A person who suffers legally relevant damage has a right to reparation from a person who caused the damage intentionally or negligently or is otherwise accountable for the causation of the damage.51

Similarly, article 2:101 PETL states that ‘[d]amages require material or immaterial harm to a legally protected interest’.52 By contrast, the French law of civil liability has traditionally abstained from recognising any sort of limitation or restriction on compensable losses and French courts have consistently adopted and then preserved a very broad range of harms (dommages) and losses (préjudices) triggering civil liability.53 One of the rare occasions where this tendency was resisted can be found in the so-called ‘widow cases’ decided by French courts between the 1930s and the 1960s (and on occasion until 1999),54 in which actions brought by ‘concubines’ (as they were called at the time) claiming compensation for losses caused by the death of their partners were regularly dismissed.55 The reason for this case law was that, not being married to the victim at the time of his death, the concubine was deemed not to have a ‘legitimate interest worthy of legal protection’ (intérêt légitime juridiquement protégé).56 Except for this very particular case law (which was abandoned by the Cour de cassation in the early 1970s),57 the attitude of French law to compensation for losses has always been extremely liberal.58 However, it is apparent that the Projet de 51 See the comment in C von Bar and E Clive (eds), Principles, Definitions and Model Rules of ­European Contract Law. Draft Common Frame of Reference (DCFR). Full Edition, 4 (Munich, Sellier, 2009) 3083ff, as well as C von Bar, ‘The Notion of Damage’ in AS Hartkamp and others (eds), Towards a European Civil Code, 4th edn (Alphen aan den Rijn, Kluwer Law International, 2011) 387ff. For a critical discussion of this provision and its terminology, see S Whittaker, The ‘Draft Common Frame of Reference’. An Assessment, 2008, Assessment commissioned by the UK Ministry of Justice, 112ff, who points out that the DCFR deploys the (numerous) concepts of ‘loss’, ‘damage’ and ‘injury’, but does not always define them (as is the case of ‘injury’); their mutual relations are, therefore, doomed to remain obscure and possibly tainted with inconsistency. 52 See the comment by H Koziol, in European Group on Tort Law (ed), Principles of European Tort Law. Text and Commentary (Wien/New York, Springer, 2005) 27ff. 53 Borghetti (n 1) 145ff; Viney, Jourdain and Carval (n 4) [250]ff. 54 For the cases themselves see Viney, Jourdain and Carval (n 4) [272]. The case law referred to in the text was finally abandoned after the statute of 15 November 1999 that legitimised the relation between non-married partners, even if of the same sex: see arts 515-1 and 515-8 Cc. 55 Viney, Jourdain and Carval (n 4) [248]. 56 For references and comments on these cases, see Weir (n 2) 10ff, fn 26. 57 Chambre mixte 27 February 1970, D 1970, 201 note D Combaldieu; JCP 1970 II 16305 notes R Lindon and P Parlange. 58 O Moréteau, ‘Basic Questions of Tort Law from a French Perspective’ in H Koziol (ed), Basic ­Questions of Tort Law from a Comparative Perspective (Wien, Jan Sramek Verlag, 2015) 3.

The Concepts of ‘Harm’ in the French and Italian Laws of Civil Liability  215 réforme intends to counter this attitude and to mitigate its effects,59 particularly in the field of personal injuries (dommage corporel). To this end, the Projet provides for the standardisation of heads of losses (postes de préjudices), which are to be fixed by the government (articles 1269 and 1271 of the Projet). It also provides for the regulation of the assessment of personal and psychological injuries, or better, of the resulting préjudice fonctionnel (article 1270 of the Projet). Therefore, it may be said that both French law and Italian law seek to limit civil liability, but the techniques which they deploy are different. Italian law seeks to select recoverable harms at the primary level of the requirements of the tort itself (in French terms, it restricts the range of dommages), whereas French law seeks to select recoverable losses at the secondary level of civil liability stricto sensu (in French terms, it restricts the range of préjudices). This difference results in both convergences and divergences between the two legal systems. Convergences may be found as regards a legislative (or quasi-legislative) regulation of the assessment of danno biologico/préjudice functionnel; divergences mainly concern the attitudes towards ‘pure economic loss’, on one side, and ‘pure non-economic losses’, on the other, as I will now explain.

V.  The Doubtful Compensation of ‘Pure Economic Losses’ and of ‘Pure Non-economic Losses’ Even were the Projet de réforme to be enacted, it would seem that compensation for ‘pure economic losses’ (préjudices purement économiques) is not barred under French law, whereas under Italian law these types of losses are generally irrecoverable, as they do not satisfy the test of unlawfulness of harm.60 This characteristic of the French law of civil liability is peculiar and probably unique in the European context, since German law has set its face decidedly against the compensation of such losses, as has English law as well, if perhaps to a lesser extent.61 For the same reasons, French law runs the risks of opening the gate to compensation of ‘pure non-economic losses’, thus again departing from the stance of most national jurisdictions across Europe. Paradigmatic in this context are those tort claims which may be depicted as arising from ‘wrongful birth’, on the one hand, and from ‘wrongful death’, on the other. The issue of recovery in respect of ‘wrongful life’ was finally addressed by the French law of civil liability in 2000 in the famous affaire Perruche,62 though the 59 See notably P Giliker, ‘Codifying Tort Law: Lessons from the Proposal for Reform of the French Civil Code’ (2008) 57 ICLQ 561. 60 Castronovo (n 14) 299ff. 61 See the essays collected in M Bussani and VV Palmer (eds), Pure Economic Losses in Europe (Cambridge, Cambridge University Press, 2003). 62 Ass plén 17 November 2000, Bull AP 2000 no 9, D 2001, 332 obs by D Mazeaud and P Jourdain, JCP 2001 II 10438 obs J Saint-Rose and F Chabas.

216  Pietro Sirena Cour de cassation had earlier delivered two judgments on the question in 1996.63 In the affaire Perruche itself, the parents had not been informed of the foetus’s abnormality during the pregnancy and so the doctors were held liable to compensate the non-patrimonial losses suffered by the impaired newborn child. The judgments rendered by the Cour de cassation seem consistent with French law given that it does not require any unlawful factor of harm for the award of damages, in particular given its absence from the provisions of the Code civil governing civil liability themselves. Similar decisions were made by the French Cour de cassation in 2001, but in 2002 the legislator decided to intervene and enacted legislation overriding this case law and providing that ‘no one may rely on a loss arising from the sole fact of birth’.64 This is itself interesting, as the fact that the French legislator was compelled to take such a step shows that, under the general rules on civil liability, claims of this sort had to be upheld by the courts, as indeed they were. By contrast, however controversial its judgments may have been, the Italian Court of cassation was able to dismiss similar actions on the sole basis of the Codice Civile, since a claim which fails the test of unlawfulness imposed by article 2043 Cod civ (danno ingiusto) determines the outcome of the case.65 In fact, a Plenary Session of the Corte di cassazione in 2015 dismissed an action of this kind, on the ground that there is no right to be born healthy and, therefore, the ‘unlawfulness’ requirement was not met.66 Moreover, it was further held that there was no causal relationship between the claimant’s impairment and the fact that the doctors had not informed the claimant’s parents of it during pregnancy.67 Similar questions are raised by cases of ‘wrongful death’ in claims arising from fatal accidents. From the viewpoint of Italian law, causing the death of another person does not in itself give rise to a non-patrimonial loss sustained by the victim (danno ­tanatologico),68 so that no claim for such compensation is transmitted on death to her heirs.69 Therefore, the deceased’s heirs are not entitled to claim for their ‘pure non-economic loss’ (pretium mortis), whereas, pursuant to article 2059 Cod civ, her dependants are 63 Viney, Jourdain and Carval (n 4) [249-6]. 64 Loi no 2002-303 of 4 March 2002 regarding patients’ rights, art 1er-I, alinéa 1er (‘Nul ne peut se prévaloir d’un préjudice du seul fait de la naissance’) later codified as art L 114-5 of the Code de l’action sociale et des familles. 65 To this effect, see FD Busnelli, ‘Verso una giurisprudenza che si fa dottrina. Considerazioni in margine al revirement della Cassazione sul danno da c.d. nascita malformata’ (2013) Rivista di diritto civile 1519. 66 Cass civ, sez un, 22 December 2015 no 25767 [2016] Responsabilità civile e previdenza 152, note M Gorgoni; [2016] Foro it I 494 note C Bona. Previously, however, a similar claim for damages had been upheld by Cass civ (3) 2 October 2014 no 16754 [2013] Responsabilità civile e previdenza 124 note M Gorgoni; for a wider account, see FP Patti, ‘Danno da nascita e danno da morte: due sentenze a confronto’ (2014) Responsabilità civile e previdenza 764. 67 For a comparative discussion of the topic, see Weir (n 2) 27ff. 68 For a comparative discussion of the topic, see FP Patti, ‘Danno da morte, coscienza sociale e ­risarcimento per i congiunti: verso una riforma del BGB?’ (2017) Rivista critica del diritto privato 39. 69 This traditional ruling was challenged in 2014, when a judgment of the 3rd panel of the Court of cassation, which is competent for civil liability, ruled that danno tanatologico is recoverable: Cass civ (3) 23 January 2014 no 1361 [2014] Responsabilità civile e previdenza 492 note CM Bianca; [2014]

The Concepts of ‘Harm’ in the French and Italian Laws of Civil Liability  217 entitled to claim compensation for their bereavement loss (pretium doloris) unless it is proved that at the time of the death they did not have any significant tie with the deceased. In addition, the deceased’s dependants are entitled to get compensation for their economic loss, provided that it is both claimed and proved. On a different note, Italian law awards compensation for danno biologico terminale or danno catastrofale70 which, although very differently described, consists of a curtailment of life expectance due to the injuries suffered by the victim.71 The Italian Constitutional Court approved the reasoning behind this position in 1994, in the course of giving a judgment whose author was one of the most famous ­Italian scholars of private law of the twentieth century, Luigi Mengoni.72 Returning to French law, it seems difficult to ascertain whether claims for compensation for ‘wrongful death’ should be upheld by the courts. The question is answered in the negative by a large majority of scholars, but the reasons justifying this position are not very clear. Most point out that such damages would not benefit the victim but only his or her heirs73 and this argument seems to be based more on ethics or morality than on law. At any rate, it does not explain the reasons why the requirements of civil liability would not in fact be met here.

VI.  The Compensation of ‘Harms’ in the Absence of Consequential ‘Loss’ On occasion, settled case law recognises that damages may be awarded for the primary ‘harm’ inflicted by a tort irrespective of any evidence of a consequential loss. Paradigmatic here is the French case law relating to breach of a duty of ­information, particularly in the field of medical liability, and to acts of unfair competition, where it is often affirmed that the mere ‘trouble’ of which the tort consists is enough for recovery.74 For this purpose, the courts usually resort to a presumption Foro it I 719 note A Palmieri and R Pardolesi; for a wider account, see A Galasso, ‘Il danno tanatologico’ (2014) Nuova Giurisprudenza civile commentata II 257; FP Patti (n 66) 764. The contrast thus raised within the jurisprudence of Italian Court of cassation was however stopped by the plenum (Sezioni Unite) of the Court, which reaffirmed the negative answer: Cass civ, sez un, 22 July 2015 no 15350 [2015] Responsabilità civile e previdenza 1530 note M Bona; [2015] Foro italiano I 2682 note A Palmieri and R Pardolesi; for a wider account, see FD Busnelli, ‘Tanto tuonò, che … non piovve – Le sezioni unite sigillano il “sistema”’ (2015) Corriere giuridico 1208; E Navarretta, ‘La “vera” giustizia e il “giusto” responso delle S.U. sul danno tanatologico iure hereditario’ (2015) Responsabilità civile e previdenza 1416; C Scognamiglio, ‘Il danno tanatologico e le funzioni della responsabilità civile’ (2015) ­Responsabilità civile e previdenza 1430. This last judgment of the Court of cassation was recently confirmed by Cass civ (3) 13 February 2019 no 4146; in the meanwhile, however, courts of first instance have occasionally proved inclined to award damages in respect of the death of the tort’s victim. 70 Recently Cass civ (3) 23 October 2018 no 26727 [2019] Foro it I 114. 71 For other jurisdictions (including France) on this topic, see Weir (n 2) 16ff. 72 Corte Cost 27 October 1994 no 372 (n 27). 73 Viney, Jourdain and Carval (n 4) [265-5]. 74 The notion of trouble, which is to be found in art 809(1) CPC is contended in the context of private law; see Caroline Guillemain, Le trouble en droit privé (Presses Universitaires d’Aix-Marseille, 2000). See also ch 18 below, pp 382; Cass civ (3) 5 July 2019 no 18056.

218  Pietro Sirena of the existence of préjudice from the sole fact of proof of dommage caused by the defendant’s fault, but it is clear that this is little more than a fictitious expedient, used to disguise the real reasons actually underlying the judgments. This attitude is mostly concerned to stay within the compensatory function of civil liability and to conceal any deterrence-based rationale. These judicial decisions, however, are better explained if it is plainly admitted that it is the harm caused by the tortfeasor’s fault which is in itself being compensated, rather than a consequential loss implied or presumed on the basis of such a harm. Breach of information duties has raised an instructive line of French cases in the field of medical liability.75 Since a key judgment in 2010,76 the Cour de cassation has applied a formula according to which ‘the failure of a doctor to discharge the duty of information he owes to his patient, causes the person to whom this information is owed a loss that under [former] article 1382 [now article 1240 Cc] the court cannot leave without reparation’.77 It is true that more recently the Cour de cassation has been inclined to restrict civil liability of this kind, as it has started to hold that liability for breach of a duty of medical information requires compensation of harm only if the risk which thereby remained unknown to the patient in fact materialised to her detriment; nonetheless, it is important to note that compensation is still awarded not only where the patient would have avoided the risk if duly informed by her doctor, but also where receipt of the information would have allowed her to have better prepared to deal with the risk.78 All in all, therefore, while the Cour de cassation has taken a more stringent approach to the requirement of a primary ‘harm’ as well as to the requirement that this harm is causally connected with the doctor’s conduct, it does not actually require a consequential loss in order to award damages. In the field of unfair competition, it is settled law accepted by the Cour de cassation that acts of unfair competition necessarily imply the occurrence of a loss (‘les faits de concurrence déloyale générateurs d’un trouble commercial impliquent nécessairement l’existence d’un préjudice’).79 A similar rule has been applied to cases of any breach of a negative obligation, ie an obligation not to do something (obligation de ne pas faire). In this respect, although apparently concerned with the award of damages irrespectively of any letter of formal notice (mise en demeure), the former article 1145 Cc, which stipulated that ‘if the obligation is not to do, a person who contravenes it owes damages by the sole fact of the contravention’, was interpreted as though it provided that ‘a person who contravenes an obligation not to do owes damages by the sole fact of the contravention, irrespective of whether or not the claimant suffers a consequential loss’.80

75 Regarding the scope of such a duty see Malaurie, Aynès and ­Stoffel-Munck (n 7) 324. 76 Cass civ (1) 3 July 2010 no 09-13591, D 2010, 1522, JCP G 2010, 788. 77 In recent times, see Cass civ (1) 12 July 2012 no 11-18.327, D 2012, 1610, Revue de droit sanitaire et social 2012, 757. 78 Cass civ (1) 14 November 2018 no 17-27980, 17-28529. 79 Cass civ, Commercial Chamber 9 January 2019 no 17-18.350. 80 Cass civ, Commercial Chamber 9 January 2019 no 17-20.526.

The Concepts of ‘Harm’ in the French and Italian Laws of Civil Liability  219 Even if traditionally civil liability has not been much developed in the field of violations of competition law in the sense of antitrust law (les pratiques anticoncurrentielles),81 it is to be noticed that, pursuant to article 17(2) of the 2014 Directive on the private enforcement of competition law,82 ‘it shall be presumed that cartel infringements caused harm’, even if ‘the infringer shall have the right to rebut that presumption’.83 Similar remarks may be made in relation to compensation of personal and psychological injuries (dommage corporel),84 or loss of consortium, which, albeit commonly depicted as consequential losses, are actually compensated in themselves. This is also the case of any kind of unlawful discrimination, this being first recognised in the field of labour law.85 Displaying a similar attitude, the French Cour de cassation eventually came to hold that ‘under article 9 of the Code civil, a mere finding of an infringement of a person’s private life gives rise to a right to reparation’.86 By contrast, consequential losses are to be found in pain and suffering, emotional distress, etc (danno morale (soggettivo)/dommages moraux/Schmerzensgeld), presumably as a result of personal or psychological injuries. However, the Italian case law tends to reject compensation of pain and suffering, emotional distress, etc (danno morale (soggettivo)), provided that personal and psychological injuries, or the loss of consortium, etc have been compensated (danno biologico and danno esistenziale).87

VII. Conclusion The comparative analysis which has been outlined above suggests that the distinction between dommage/danno-evento on the one hand and préjudice/­­dannoconseguenza on the other is unmistakably useful in the context of civil liability and, 81 Viney, Jourdain and Carval (n 4) [247-3]. 82 Directive 2014/104/EU of the European Parliament and the Council of 26 November 2014 on certain rules governing actions for damages under national law for infringement of the competition law provisions of the Member States and of the European Union. 83 In Italian literature, see E Camilleri, ‘Il risarcimento del danno per violazione del diritto della concorrenza: ambito di applicazione e valutazione del danno’ (2018) Nuove leggi civili commentate 143; F Mezzanotte, ‘Sul private enforcement del diritto antitrust (in vista del recepimento della Dir. 2014/104/UE)’ (2015) I Contratti 1077; id, ‘Il trasferimento del sovrapprezzo anticoncorrenziale’ (2018) Nuove leggi civili commentate 215ff. 84 J Knetsch, ‘Le traitement préférentiel du dommage corporel’ JCP G Suppl to nos 30–35 (25 July 2016) 9. 85 Cass civ (3) 17 January 2018 no 901, (2018) Danno e responsabilità 453 note G Ponzanelli. Cass civ 17 January 2018 (1) no 901 (2018) Foro it I 911, where it is stated, albeit obiter, that a danno ­esistenziale can be acknowledged in case of stalking, even if the victim experienced no bodily injuries but just detrimental ‘changes in her habits’ as a consequence of such a wrong. 86 Cass civ (1) 5 November 1996, Bull civ I no 378 (‘selon l’article 9 du Code civil, la seule constatation de l’atteinte à la vie privée ouvre droit à réparation’), JCP G 1997 II 22805 note J Ravanas, JCP G 1997 I 4205 no 1 note G Viney, RTD civ 1997, 632 note J Hauser. 87 Cass civ, sez un, 11 November 2008 nos 26972-26973-26975 (n 37); Cass civ (3) 23 January 2014 no 1361 (n 69). Yet, this contention is increasingly challenged by the latest judgements issued by the Corte di Cassazione, which seems now ready to couple danno biologico/danno esistenziale and danno morale soggettivo: Cass civ (3) 27 March 2018 no 7513; cass civ (3) 31 May 2018 no 13770.

220  Pietro Sirena to some extent, may even be essential. However, such a distinction should not lead us to conclude that only consequential losses (préjudice/danno-conseguenza) and not also the primary harm (dommage/danno-evento) are recoverable. In fact, such a view would not reflect the present state of the French and Italian laws of civil liability, either in terms of legislation or in terms of case law; moreover, and perhaps more importantly, it would reduce the ability of compensation to protect some of the most fundamental rights from being infringed by a wrongdoer. The question of consequential losses as recoverable damages has been discussed in the course of the development of French law with the aim of expanding the scope of compensation of harm to consequences beyond those foreseeable by the tortfeasor. The source of article 1231-4 Cc as enacted in 2016 and now in force may be traced back to a couple of paragraphs of Pothier’s Treatise on obligations,88 where he raises the question of whether ‘other harms that I have suffered which are a more remote and more indirect consequence of the deceit of my debtor’ are ­recoverable.89 Pothier’s aim was, therefore, to expand the scope of compensation to consequences beyond those envisaged by the tortfeasor, as in the case where, for example, being aware of the illness of a head of cattle, a seller conveys it to a buyer and as a result the rest of the latter’s livestock is infected and eventually her farm falls into ruin. The seller is liable for the loss of the buyer’s other livestock,90 but is she liable for the ruin of the latter’s farm as well? Pothier’s response was that: one must not include within the damages which a debtor is bound to pay on the ground of his deceit, those which are not only a remote consequence, but which are not a necessary consequence and could have other causes.91

This principle proved to be a wise device to deal with issues of risks arising from a tort and to balance them with the requirements of fault and causal connection,92 which are intended to protect the tortfeasor from bearing a limitless liability. Given this, Pothier did not dispute the proposition that, as long as it was attributable to the negligence of the person who committed the tort, the ­ primary harm was to be compensated as such, whether or not it resulted in any 88 The same applies to art 1223 Cod civ, as elucidated by G Visintini, Trattato breve della responsabilità civile, 3rd edn (Padova, CEDAM, 2005) 680ff. 89 RJ Pothier, Traité des obligations (Paris, Dalloz, 2011) [167] (‘autre dommages que j’ai soufferts, qui sont une suite plus éloignée et plus indirecte du dol de mon dêbiteur’). 90 ibid [166] (‘on ne doit pas comprendre dans les dommages et intérêts dont un débiteur est tenu pour raison de son dol, ceux qui non seulement n’en sont qu’une suite éloignée, mais qui n’en sont pas une suite nécessaire, et qui peuvent avoir d’autres causes’). 91 ibid [167]. 92 For a discussion of the topic, see the essays collected in M Dyson (ed), Regulating Risk through Private Law (Cambridge/Antwerp/Portland, Intersentia, 2018); among them, see particularly those by V Wester-Ouisse, S Taylor and D Fairgrieve, ‘Risk and French Private Law’, 55ff and of N Coggiola and B Gardella Tedeschi, ‘Risk and Italian Private Law’, 113ff; see also M Infantino and E ­Zervogianni, The European Way to Causation in M Infantino and E Zervogianni (eds), Causation in European Tort Law (Cambridge, Cambridge University Press, 2017) 85ff. For a specific issue, see J-S Borghetti, ‘Litigation on hepatitis B vaccination and demyelinating diseases in France: breaking through scientific u ­ ncertainty?’ in M Martín-Casals and DM Papayannis (eds), Uncertain Causation in Tort Law (Cambridge, Cambridge University Press, 2016) 11ff.

The Concepts of ‘Harm’ in the French and Italian Laws of Civil Liability  221 c­ onsequential loss. This harm is to be compensated as such because and insofar as it has been caused by the wrongdoer at fault, whereas any consequential loss is not to be attributed to her negligence and for this very reason is recoverable only where it amounts to a necessary and strict consequence of the primary harm inflicted on the tort’s victim. Article 1235 of the Projet de réforme, which provides that ‘[a]ny certain loss is reparable where it results from harm’ etc, should be read in the sense that compensation for the harm caused by the defendant at fault (article 1241 of the Projet) is to be accompanied by, rather than replaced by, compensation for consequential losses, which, independently of any fault, result from the harm (dommage) inflicted on the victim of the tort. In other words, the rationale of article 1235 of the Projet is to be found in the fact that it expands the scope of compensation to consequential losses (even if subject to some strict requirements) and not in the fact that it excludes compensation of the primary harm. The development in some jurisdictions of a tort of ‘intrusion on seclusion’ in the context of privacy shows that the primary harm, namely the violation of someone else’s privacy irrespectively of any disclosure or exploitation of personal information,93 is to be compensated as such, whether a loss results from this wrongful behaviour or whether it instead remains a mere case of ‘snooping’.94 Similar considerations apply to discriminatory conduct, particularly as prohibited by EU law.95 The constitutional status demanded by such rights may be enough to determine the outcome of cases of this kind, thus creating a special law allowing the recovery of damages for the violation of human rights.96 Damages in this situation have been tentatively conceptualised as ‘vindicatory damages’,97 in the sense that, though they are free of any purpose of punishment or deterrence, they serve solely to vindicate a (human) right. Overall, therefore, increasingly heed must be taken of developments in these new areas of both theory and practice of civil liability in order to avoid them being encapsulated in niches of their own, and instead to ensure that they are fully ­integrated into general tort law. 93 For references to new case law in common law jurisdictions, see the essays collected in JNE ­Varuhas and NA Moreham (eds), Remedies for Breach of Privacy (Oxford, Hart Publishing, 2018). 94 E Chamberlain, ‘Snooping: How Should Damages be Assessed for Harmless Breaches of Privacy?’ in K Barker, K Fairweather and R Grantham (eds), Private Law in the 21st Century (Oxford, Hart Publishing, 2017) 389ff; JNE Varuhas, ‘Varieties of Damages for Breach of Privacy’, in Varuhas and Moreham (n 93) 55; R Stevens, ‘Damages for Wrongdoing in the Absence of Loss’ in Varuhas and Moreham (n 93) 97; NA Moreham, ‘Compensating for Loss of Dignity and Autonomy’ in Varuhas and Moreham (n 93) 125; E Descheemaeker, ‘Claimant-Focused Damages in the Law of Privacy’ in Varuhas and Moreham (n 93) 143. 95 For an overall picture, see U Belavusau and K Henrad (eds), EU Anti-Discrimination Law Beyond Gender (Oxford, Hart Publishing, 2019); J Mulder, Eu Non-Discrimination Law in the Courts. Approaches to Sex and Sexualties Discrimination in EU Law (Oxford, Hart Publishing, 2017). 96 For an in-depth analysis of the topic, see JNE Varuhas, Damages and Human Rights (Oxford, Hart Publishing, 2016) 13ff. For a number of national reports on the topic, see E Bagińska, Damages for Violation of Human Rights. A Comparative Study of Domestic Legal Systems (Cham, Springer, 2016); for a comparison among them, see, ‘Damages for Violation of Human Rights: A Comparative Analysis’ in E Bagińska ibid 443. 97 On this category, see J Edelman, ‘Vindicatory Damages’ in Barker, Fairweather and Grantham (n 94) 343ff; JNE Varuhas (n 96) 36ff.

222

11 Nuisance and Coming to the Nuisance The Porous Boundary between Torts and Servitudes in England and France CIARA KENNEFICK

Searching the French Code civil for equivalents to the English tort of nuisance is a delightfully disorientating experience for a common lawyer. There is nothing quite like it in any of the provisions of Book III of the Code civil which are devoted to ‘extracontractual liability’. The iconic text on liability for fault in ­article 1240 is out of contention, as liability in nuisance is, and always has been, strict in English law.1 Many instances of nuisance could conceivably come within the provision on strict liability for ‘things’ in article 1242(1),2 but it has never functioned as an equivalent to the tort of nuisance. Two factors may have inhibited such a development: until 1928, only personal property came within this rule and liability was not strict until 1930.3 Then one finds the spread of fire, a familiar instance of nuisance for a common lawyer,4 in articles 1242(2) and 1242(3), but this rule is restricted to these particular facts and, moreover, is based on fault.5 Articles 1246 to 1252 on ‘ecological harm’, which were added to the Code civil in 2016,6 are a final possibility. However, they simply recognise a new head of 1 Read v Lyons [1947] AC 156, 183; Transco v Stockport [2004] 2 AC 1, 15. There are two exceptions. First, the liability of occupiers and landlords in nuisance for the acts of third parties and their tenants respectively is not strict: Sedleigh-Denfield v O’Callaghan [1940] AC 880; Cocking v Eacott [2016] QB 1080. See too the text from n 28 to n 29 below. Secondly, an occupier is not strictly liable in nuisance for acts of nature; indeed, in contrast to the general law of negligence, the standard of care here is subjective: Goldman v Hargrave [1967] 1 AC 645; Leakey v National Trust [1980] QB 485. 2 F Terré, P Simler, Y Lequette and F Chénedé, Droit civil, Les obligations, 12th edn (Paris, Dalloz, 2019) 1062. 3 Req 6 March 1928, D 1928 1, 97; Cass civ 13 February 1930, D 1930, 1, 57. 4 The rule in Rylands v Fletcher (1868) LR 3 HL 330 has frequently been invoked in such cases but claims based on this rule will now succeed more rarely following the restrictive decision of Gore v Stannard (t/a Wyvern Tyres) [2014] QB 1. The rule in Rylands v Fletcher is treated as a ‘subspecies of nuisance’: Transco (n 1) 10. 5 However, before these provisions were added to the Code civil in 1922, liability for fire was essentially strict in French law: F Terré, P Simler and Y Lequette, Droit civil, Les obligations, 11th edn (Paris, Dalloz, 2013) 805–06. 6 Loi no 2016-1087 of 8 August 2016, art 4.

224  Ciara Kennefick damage and do not, therefore, constitute a separate basis of liability; furthermore, article 1248 specifies that such a claim can be made by only a limited number of claimants. Retreating to Book II of the Code civil, one finds the elegant abstractions on ownership in article 544, which subsequently splinters into a dazzle of detailed ‘servitudes’ dealing with many aspects of relations between neighbouring landowners in Chapter 2 of Title II. Wells, cesspits, hearths, forges, kilns, furnaces, stables and heaps of salt and corrosive substances are the subject of merely one provision, article 674. Specific minimum or maximum distances are frequently prescribed7 and, puzzlingly, this numerical information is presented in both metric and ­imperial forms in some instances.8 These provisions in Book II, which were enacted in 1804, essentially mirror pre-Revolutionary customary law on these matters and thus bear the distinct hue of that bygone era.9 As such, this portion of the Code civil is, unsurprisingly, alien territory for the common lawyer. There is, however, a civil wrong in French law which is truly an equivalent to the tort of nuisance.10 Unusually for French law, though, this basis of liability is not to be found anywhere in the Code civil, and indeed its principal features have been developed almost exclusively by judges rather than legislators. The foundations can be traced to the case of Derosne v Puzin, which generated two seminal ­decisions of the Cour de cassation in 1844 and 1849.11 As the unsuccessful defendant in this case justifiably pointed out in the first episode of this litigation, there was ‘no obstacle’ to his behaviour in the Code civil.12 Nonetheless, he was held liable for the noise and smoke emanating from his factory which interfered with the claimants’ enjoyment of land on the basis of a doctrine which, though the terminology varies and notably includes ‘nuisance’, is generally described as ­troubles de voisinage (neighbourhood disturbances). These are exciting times for the law of nuisance in both French and English law. In the last decade, there has been a flurry of proposals on reforming the Code civil. As one of the principal objectives of these endeavours is to ensure that the Code civil better reflects the content of French private law, the i­ntegration

7 See arts 657 and 663 and arts 676 to 679 Cc. 8 See arts 676 and 677 Cc. It is likely that this inconsistency can be explained by the fact that units of measurement were in a state of flux in France from 1793 (when the first draft Code civil was produced) to 1804 (when the Code civil was promulgated). See A Favre, Les origines du système métrique (Paris, PUF, 1931) 131–224. 9 See eg R-J Pothier, Coutumes des duché, bailliage et prévôté d’Orléans et ressort d’iceux (Paris, Debure, 1780) 389–407. 10 There is also a similar rule in French administrative law: R Chapus, Droit administratif général, 15th edn (Paris, Montchrestien, 2001) vol 2, 665–73; N Foulquier, Droit administratif des biens, 4th edn (Paris, LexisNexis, 2018) 776, 788–807. See too the text from n 108 to n 113 below. 11 Cass civ 27 November 1844, D 1845, 1, 13; Req 20 February 1849, D 1849, 1, 148 (the defendant’s surname is spelled Desrone in this report). 12 D 1845, 1, 13, 15.

Nuisance and Coming to the Nuisance  225 of the stray doctrine of troubles de voisinage is, naturally, recommended in each of the four  proposals which are devoted to civil liability and property.13 ­ Nonetheless, and indeed as we shall see, the appropriate location of this body of law in the Code civil is a controversial question to which these studies provide different answers. The law of nuisance is in an equally mutable state in England. In the seminal decision of Coventry v Lawrence in 2014, the UK Supreme Court fundamentally altered some aspects of nuisance and engaged in detailed discussions of an obiter nature about the way in which other features of this tort ought to develop.14 The focus of this comparative study is threefold. Section I examines the nature and scope of the tort of nuisance and troubles de voisinage. This section demonstrates that despite striking similarities, liability under troubles de voisinage is significantly more extensive than that under the tort of nuisance. Section II explores the concept of coming to the nuisance, the extent to which the defendant’s prior use of land is treated as a reason for restricting his liability in both systems. In this section, we see that while coming to the nuisance has been partially embraced in English law, it exists in a much more powerful form in French law. Section III shows that coming to the nuisance is undermined by a serious problem of principle: it is, in essence, a servitude, but it lacks the indispensable limitations of such property rights. This important finding has different implications in each system. As discussed in the preceding paragraph, the law of nuisance is in a state of flux in both England and France: this auspicious opportunity to address the problems raised by the introduction of versions of coming to the nuisance ought to be seized in both systems.

I.  The Tort of Nuisance and Troubles de Voisinage: Similarities and Differences The English tort of nuisance and the French doctrine of troubles de voisinage have numerous structural features in common; these similarities are especially remarkable as there is no evidence that they are the result of the transplantation of legal ideas from one system to the other. Liability is strict in both the tort of nuisance and troubles de voisinage.15 Furthermore, they deal with strikingly similar types

13 Avant-projet Catala (art 1361); Association Henri Capitant, Propositions de l’Association Henri Capitant pour une réforme du droit des biens (Paris, Lexis Nexis, 2009) (Avant-projet Capitant) (arts 629 to 630); Avant-projet Terré (art 24); the current Projet de réforme de la responsabilité civile (art 1244). 14 Coventry v Lawrence [2014] AC 822. All references to ‘nuisance’ simpliciter in English law denote the tort of private nuisance only and do not extend to the tort of public nuisance. 15 See n 1 for English law. G Viney and P Jourdain, Traité de droit civil, Les conditions de la responsabilité, 3rd edn (Paris, LDGJ, 2006) 1205–07. ‘The Cour de cassation in the end gave up making futile and artificial references to fault’: ibid, 1214.

226  Ciara Kennefick of interference such as noise,16 smells,17 and even airborne balls from sporting events.18 There are also commonalities in the types of damage for which a remedy is available: the focus in both French and English law is on the interference with the enjoyment or use of land but physical damage to land also comes with the scope of the rules in each system.19 An additional point of convergence is the relevance of the locality within which the wrong is committed: the standard by which the defendant is judged is relative rather than absolute.20 Finally, regulations of a public law nature are of secondary importance in both French and English law. Compliance with such regulations does not, on its own, exonerate the defendant from civil liability in either system.21 However, in English law, the existence of planning permission may influence the assessment of the character of the area22 and the question whether an injunction should be granted23 while, in French law, injunctions cannot be granted if they would encroach on the decision of an administrative court on account of the separation of powers.24 There are, naturally, differences too between the tort of nuisance and troubles de voisinage. When these differences are considered collectively, it is clear that a defendant is more exposed to liability in French law than in English law. While the tort of nuisance is regarded as a means of protecting the claimant’s property rights, troubles de voisinage, is generally considered to be a limitation on the property rights of the defendant: thus, the former is a tort to land and the latter is a wrong by land. These differing starting points influence the rules on who can sue and who can be sued. In the leading case of Hunter v Canary Wharf, a majority of the House of Lords affirmed, somewhat controversially, that only claimants who have an interest in land can invoke the tort of nuisance.25 In contrast, mere occupiers have been permitted to invoke troubles de voisinage.26

16 See eg Cass civ (2) 13 January 2005 no 04-12623 and Coventry (n 14) both of which related to the racing of motor vehicles. 17 See eg St Helen’s Smelting Co v Tipping (1865) 11 HL Cas 642 and Cass civ 18 February 1907, D 1907, 385 (Dupont c Lecante) both of which concerned smells from factories. 18 Cass civ (2) 10 June 2004 no 03-10434, Bull 2004 II no 291 (golf balls); Miller v Jackson [1977] QB 966 (cricket balls). 19 The claimants succeeded in their claims with respect to both types of damage in the cases on cricket balls and golf balls in England and France respectively: Miller (n 18); Cass civ (2) 10 June 2004 (n 18). 20 Viney and Jourdain (n 15) 1220–21. This is true only of claims relating to interference with enjoyment in English law: St Helen’s. As Simpson notes, this rule is principally the result of the headnote in that case rather than ‘the three sloppy opinions delivered by the Law Lords’: AWB Simpson, Leading Cases in the Common Law (Oxford, Oxford University Press, 1995) 184. 21 Viney and Jourdain (n 15) 1217–18; Coventry (n 14). Nonetheless, in French law such compliance plays a significant role in the rule on coming to the nuisance, which is discussed in detail in section IIB. 22 Coventry (n 14). 23 This was one of the fundamental changes introduced in Coventry (n 14). 24 Viney and Jourdain (n 15) 1217. In contrast to the position in England, French administrative law is a distinct body of law which has been developed by a separate system of administrative courts. 25 Hunter v Canary Wharf [1997] AC 655. 26 Cass civ (2) 17 March 2005 no 04-11279, Bull 2005 II no 73.

Nuisance and Coming to the Nuisance  227 Similarly, claimant landlords out of possession are treated more generously in French law than in English law. While in English law, such a claimant cannot obtain a remedy for nuisances of an impermanent nature even if they reduce the value of the land, there are no similar obstacles in French law precisely because this reduction in value is considered to constitute sufficient harm.27 The wider scope of troubles de voisinage is equally evident in the approach to the question of who may be sued. A limited but fairly loose class of ‘neighbours’ may be liable for troubles de voisinage, whereas the wrongdoer is liable in English law regardless of whether he is a neighbour.28 Nonetheless, liability is still,­ ultimately, less extensive in English law. In this system, where a nuisance has been committed by a tenant, the landlord (who, by definition, does not have exclusive possession) is liable only if he authorised the tort, and the occupier of the relevant land is liable only if he adopted or continued the nuisance committed by another.29 In contrast, the liability of all ‘neighbours’ for nuisances originating on their land is strict in these circumstances in French law.30 Indeed, there is an intriguing form of what a common lawyer would consider to be vicarious liability here as it has been held that such ‘neighbours’ are strictly liable for each other: they may subsequently sue the wrongdoer amongst them to recover the sum which they have been ordered to pay to the victim.31 Three further differences demonstrate that liability under troubles de voisinage is more extensive than that under the tort of nuisance. First, notwithstanding older authorities which are more equivocal on this point, damages for personal injuries are not considered to be available under the tort of nuisance.32 In contrast, damages for personal injuries33 and indeed for all other types of harm can be recovered in French law.34 The position of French law here is rather unsurprising given that, unlike English law, no sharp distinctions are drawn between the types of harm in the rest of the law of civil liability. Secondly, it is significant that in French law, 27 Mott v Shoolbred (1875) LR 20 Eq 22 (obstructing the claimant’s right to light by erecting a building sufficed for this purpose, unlike blocking up the street with vehicles); Lecante (n 17). 28 Cass civ (3) 22 June 2005 no 03-20068, Bull 2005 III no 136; MA Jones (ed), Clerk & Lindsell on Torts, 22nd edn (London, Sweet & Maxwell, 2018) 1420. 29 Cocking (n 1); Sedleigh-Denfield (n 1). 30 Cass civ (3) 17 April 1996 no 94-15876, Bull III no 108 (where the formulation is particularly blunt). As Viney and Jourdain observe, this is an instance of ‘liability for others’ which predates the Blieck case, the seminal authority on this civil wrong: (n 15) 1210. 31 Cass civ (3) 22 June 2005 no 03-20068 and no 29001 (n 28). Builders were held liable in this case in their capacity as voisins occasionels (neighbours by chance), an understandably controversial development in a system which, unlike English law, considers nuisance to be a wrong by land rather than to land. Unsurprisingly, builders have been implicitly excluded from the list of neighbours who may be liable under troubles de voisinage in all four of the relevant law reform proposals: Avant-projet Catala (art 1361); Avant-projet Capitant (art 630); Avant-projet Terré (art 24); Projet de réforme (art 1244). 32 Hunter (n 25) 706. Damage to personal property and lost profits may, however, be recovered as consequential loss: ibid. 33 Though there are not many cases in which such damages were recovered. See eg Cass civ (2) 22 October 1964, D 1965, 344 (the claimants’ health was damaged by grains of ricin emanating from a nearby oil factory). 34 Viney and Jourdain (n 15) 1216.

228  Ciara Kennefick the defendant’s use of land may constitute a trouble de voisinage only because of the claimant’s particular circumstances.35 While French law is ‘far from consistent on this point … [it] is always willing to accept such claims if the claimant is acting in a business capacity’.36 In contrast, this is precisely the context in which it was held that this factor precludes a successful claim in nuisance in English law.37 The final and most striking difference concerns the types of interference which suffice for claims under troubles de voisinage and which have no counterparts in English law. In recognising claims for privacy and views, French law goes far beyond English law in this respect. While articles 675 to 680 Cc are intended to protect owners from the prying eyes of neighbours, by placing various restrictions on windows (principally), a defendant may be liable under trouble de voisinage even if he complies with these requirements. In one such instance, liability was imposed for putting the claimant in a position where he would be ‘constantly exposed to the gaze’ of the defendant.38 In contrast, the contention that there was ‘an action for opening a window to disturb the plaintiff ’s privacy’ was rejected in English law as early as 1811.39 Interference with views is also a significant point of divergence between French and English law. Such claims have been accepted on the basis of t­roubles de ­voisinage notwithstanding the absence of any relevant servitude since the 1970s.40 In subsequent decades, defendants have been liable for disturbing views by constructing a building with 10 floors and a 30-metre chimney,41 replacing a vineyard with 72 houses,42 installing an electricity generator and removing trees43 and even storing old machines, vehicles and scrap materials44 close to the boundary with the claimant’s land. The reasoning of the Cour de cassation in these cases reveals the startlingly broad nature of liability under troubles de voisinage: it was held that ‘degradation of the landscape and the urban environment’, ‘visual disturbance’ and even ‘aesthetic annoyance’ justify a remedy.45 The development of this aspect of trouble de voisinage understandably led one commentator to question why anyone would now need to purchase servitudes to restrain neighbours from blocking views.46 In contrast, such claims are inconceivable in English law in the absence of an easement of light or a relevant restrictive covenant. Indeed, a claim

35 ibid 1223–24. 36 ibid. 37 Robinson v Kilvert (1889) 41 ChD 88. 38 Cass civ (3) 7 February 2007 no 05-21405. See too Cass civ (3) 12 October 2005 no 03-19759. 39 Chandler v Thompson (1811) 3 Camp 80. 40 This observation is based on the cases discussed in G Liet-Veaux, note to Cass civ (3) 26 January 1993, Gaz Pal 1993, 1, 277, 278–79. 41 Cass civ (3) 26 January 1993 (n 40). 42 Cass civ (3) 9 May 2001 no 99-16260. 43 Cass civ (2) 23 October 2003 no 02-16303, Bull 2003 II no 318. 44 Cass civ (2) 24 February 2005 no 04-10362, Bull 2005 II no 50. 45 Cass civ (3) 9 May 2001 (n 42); Cass civ (2) 23 October 2003 (n 43); Cass civ (2) 24 February 2005 (n 44). 46 G Liet-Veaux, note to Cass civ (3) 26 January 1993 (n 40) 279.

Nuisance and Coming to the Nuisance  229 for interference with a ‘prospect’ was rejected as far back as 1610 in the ­seminal decision in Aldred’s Case.47 English law perhaps comes closest to the French­ position in its acceptance of some claims based on activities which ‘force themselves on the sense of sight’ only.48 However, this is a very exceptional and limited category of cases essentially concerning prostitution and the operation of a sex shop.49

II.  Coming to the Nuisance: The Significance of the Defendant’s Prior Use of Land Might a claimant successfully invoke the tort of nuisance or trouble de voisinage if he came to the affected land after the defendant had begun to use his land in the way of which the claimant complains? The answers given to this question followed a similar trajectory in both French and English law across time. The concept of coming to the nuisance,50 which depends on a positive response to this question, spent over one century in the wilderness in French law and was there for almost double that time in English law before being embraced in different forms in both systems.

A.  English Law: Rejection to (Limited and Badly Disguised) Reception The rejection of coming to the nuisance for ‘over 180 years’51 in English law was based on two landmark decisions of the nineteenth century. The concept appears for the first time in case law, already clothed in its current terminology, in Bliss v Hall.52 In this case before the Court of Common Pleas, the claimant sought compensation for smells emanating from his neighbour’s chandlery, which had

47 (1610) 9 Coke Reports, 57b, 58b. 48 Thompson-Schwab v Costaki [1956] 1 WLR 335, 339 (Lord Evershed MR). 49 ibid and Laws v Florinplace Ltd [1981] 1 All ER 659 respectively. As Lord Goff noted ‘[s]uch cases must … be relatively rare’: Hunter (n 25) 686. 50 The term defence is eschewed in this chapter for two reasons. First, while coming to the nuisance is commonly described as a defence in English law, its French equivalent (exonération) is only one of many terms which are used in this context in French law. Secondly, the question whether a rule is a defence is considered to be of theoretical significance in the common law world but it is not necessary to engage with these debates here. Nonetheless, it is notable that a servitude acquired by prescription has been classified as a denial rather than a defence: J Goudkamp, Tort Law Defences (Oxford, Hart Publishing, 2013) 113. Goudkamp does not discuss coming to the nuisance, but it ought not to be described as a defence in either French or English law if the arguments advanced in this chapter are correct. 51 Coventry (n 14) 840. 52 Bliss v Hall (1838) 4 Bing 183.

230  Ciara Kennefick been operating for three years before the arrival of the claimant. Counsel for the defendant argued that ‘it was [the claimant’s] own election to approach so near’53 or, as it was formulated in the more neglected report of this case, ‘the complainant came to the nuisance’.54 In rejecting this argument, all four judges relied exclusively on property law: it was held that no servitude permitting the nuisance had been acquired. Similarly, property law is the sole focus of the judgments in Sturges v Bridgman, the second judicial milestone in the story of coming to the nuisance in English law.55 In this case, the defendant sought to escape liability by arguing that the noise from machines in the kitchen which he used for his confectionary business disturbed the claimant only because the latter had recently changed the use of his land: the claimant doctor initiated his action in nuisance having built in his garden a consultation room abutting the party wall, behind which lay the defendant’s machines. The judges at first instance and in the Court of Appeal directed their minds only to the question of whether the defendant could have acquired a servitude to make such noise. Thus, in both Bliss and Sturges, the judicial analysis of the concept of coming to the nuisance is exclusively proprietary: no separate doctrine in the law of torts was conceivable. Similarly, in 1839, Gale, the leading authority on easements, began his discussion on ‘Doctrine of coming to the nuisance exploded’ in the first edition of his famous treatise by declaring that ‘[i]t is difficult to see on what principle this doctrine could have been supported … and the recent decisions of the Court of Common Pleas upon this point appear to have restored the law to an accordance with the general principles of easements’.56 Therefore, the principal nineteenth-century sources perceived that there really is only one way of taking into account the relevance of the defendant’s prior use of land and that is in the law of property. It is true that this view is ‘assumed’57 rather than expressly articulated and defended but, as we shall demonstrate in section III, it is justified by the fact that any form of coming to the nuisance in the law of torts essentially dissolves into a servitude. The concept of coming to the nuisance resurfaced in the twentieth century to a puzzlingly favourable reception. In Miller v Jackson, a village cricket club argued that it should not be liable in nuisance for, as Cumming-Bruce LJ quipped, exposing the claimants ‘to cricket balls occasionally falling like thunderbolts from the heavens’, on the ground that the claimants’ house had been built long after the club had begun playing on the adjacent field.58 Lord Denning MR, in a vivid and even emotive judgment, held that the defendant was not liable. In characteristically glib style, he purported to reject Sturges on the bizarre ground that it ‘turned 53 ibid 185. 54 (1839) 5 Scott 500, 503. 55 Sturges v Bridgman (1879) 11 ChD 852. 56 CJ Gale, A Treatise on the Law of Easements (London, Sweet, 1839) 276. 57 Coventry (n 14) 839 (Lord Neuberger commenting on Sturges (n 55)). See too Geoffrey Lane LJ who uses the term ‘assumption’ in this respect in his discussion of the same case in Miller (n 18) 987. 58 Miller (n 18) 988.

Nuisance and Coming to the Nuisance  231 on the old law about easements and prescriptions, and so forth’, thereby ­implying that this part of the law of property had somehow disappeared or was now otiose.59 Lord Denning MR was, however, in a minority of one on this point as the other two judges, correctly, held that they were constrained by precedent to uphold the claim. Geoffrey Lane LJ, was, though, particularly reluctant to reject the defendants’ argument: he stated that ‘[i]t did not seem just’ to hold someone liable in such circumstances and he lamented the fact that ‘the question [was] not open’.60 The most recent episode in the story of the concept of coming to the nuisance in English law is, unquestionably, the most momentous. In the seminal case of Coventry v Lawrence in 2014, the UK Supreme Court adopted a version of coming to the nuisance, albeit guardedly and only in dicta.61 The arguments advanced by Lord Neuberger, the only Law Lord who addressed this point, are seriously weakened by his failure to consider fully the perspective of property law. First, he invokes Blackstone’s apparent approval of the concept of coming to the nuisance where the latter states that a claimant cannot complain of foul air if he comes to live next to a tannery, as ‘the nuisance was of [his] own seeking’.62 However, the context of this passage demonstrates that Blackstone was actually discussing the law of property: the immediately prior example concerns an easement of light and the section in which this discussion appears begins by stating that ‘the benefit of the elements, the light, the air, and the water, can … be appropriated by­ occupancy’.63 Lord Neuberger also relies on R v Cross, in which the concept of coming to the nuisance was accepted in the distinct but related tort of public nuisance.64 Yet the different approach to this question in the tort of public nuisance is firmly grounded in logic: as servitudes to commit a public nuisance are an impossibility,65 coming to the nuisance does not conflict with the law of property in that field. The defendant in Coventry argued that he should not be liable for the noise emanating from his racing track on the ground that his business had been in ­operation at the time when the claimant had bought a house nearby. The favourable treatment of coming to the nuisance in the High Court in Watson v Croft ­Promo-Sport,66 a case on almost identical facts four years earlier, may have ­emboldened the defendant to invoke this concept in Coventry. At the end of a detailed analysis of the materials, Lord Neuberger ultimately adopted a version of coming to the nuisance which was limited to situations where the nuisance 59 ibid 981. 60 ibid 986. 61 Coventry (n 14). 62 W Blackstone, Commentaries on the Laws of England (Oxford, Clarendon Press, 1766) vol 2, 403. 63 ibid 402 (my emphasis). 64 R v Cross (1826) 2 C&P 483. 65 R v Cross (1812) 3 Camp 224, 227. 66 Watson v Croft Promo-Sport [2008] EWHC 759, [67]–[69]. There was no appeal on this point: [2009] EWCA 15.

232  Ciara Kennefick i­nterfered with ‘the senses’ only and was the result of an ‘alteration in the claimant’s property after the activity in question [had] started’.67 Given the second condition, this rule was of no relevance on the facts of Coventry; importantly, therefore, it did not form part of the ratio in this case. The adoption of this limited form of coming to the nuisance was intended to achieve two objectives. First, the ‘concerns expressed by Lord Denning MR in Miller’ relating to the perceived injustice of the claimant’s ability abruptly to convert a defendant’s legal use of land into a tort clearly weighed on Lord ­Neuberger’s mind.68 However, he succeeded in addressing these concerns only partially as in that case, Lord Denning MR and indeed Geoffrey Lane LJ did not purport to confine coming to the nuisance to interferences which are ‘offensive to the senses’: after all, it would also have applied to the physical damage caused to the claimant’s house.69 Lord Neuberger also clearly sought to avoid a conflict with property law: he stated that as ‘nuisance is a property-based tort … the right to allege a nuisance should, as it were, run with the land’.70 The reasons for which this attempt was futile are examined in section III. One final point of significance which emerges from this important development in Coventry is that Lord Neuberger integrated coming to the nuisance into English law while appearing to maintain the position that English law still rejects this concept. He sought to achieve this feat by subtracting the case of ‘an alteration in the claimant’s property after the activity in question has started’ from the concept of coming to the nuisance, stating, unconvincingly, that it ‘raises a rather different point’.71 It is not clear why Lord Neuberger went to such lengths to underplay the innovative aspect of his speech; his approach may, though, be related to the fact that he thought that his reformulation of the law on coming to the nuisance would not have led to different outcomes in the previous authorities on this point.72

B.  French Law: Rejection to Extensive Embrace The concept of coming to the nuisance has experienced a similar fate of rejection and reception in French law. The version which has been adopted in French law is so sweeping in its scope that it almost swallows up the rule of trouble de voisinage; the courts, though, have refrained from extending it even further. In 1855, just over a decade after the Cour de cassation recognised trouble de voisinage as an independent ground of liability,73 Demolombe, the author of the

67 Coventry 68 ibid.

69 Miller

(n 14) 841.

(n 18) 981, 986. (n 14) 841.

70 Coventry 71 ibid. 72 ibid.

73 Derosne

v Puzin (n 11).

Nuisance and Coming to the Nuisance  233 epic 31 volumes of the Cours de Code Napoléon, introduced French law to the idea of coming to the nuisance.74 He argued that a defendant ought not to be liable if the claimant ‘of his own free will subjected himself to the nuisance’ though he warned against interpreting ‘this right of prior occupation’ too broadly, given that it was based only on ‘fairness and necessity’.75 He then applied this rule to the common case, then and now, of a factory which had been in operation before the construction of a building on nearby land.76 Such a rule was, at that time, entirely novel. On examination, none of the pre-Revolutionary sources to which­ Demolombe refers, but tellingly does not cite, contain anything which would sustain his broad formulation: they deal with very specific issues only, such as the priority given to a neighbour whose well was built first when applying the rules on the necessary distances, which varied from place to place, between wells and privies.77 The concept of coming to the nuisance was expressly rejected by both scholars and judges in the following century.78 A leading treatise in 1905 engaged directly with Demolombe’s hypothetical case: the authors, Baudry-Lacantinerie and ­Chauveau, argued that such a claimant ought to be entitled to damages, as the harm which he has suffered is the inability to build on his land due to the proximity of the factory.79 Coming to the nuisance was then forcefully rejected by the Cour de cassation two years later in Dupont v Lecante, its inaugural decision on this point.80 In this case, the claimant had built two houses on land which she had purchased near a tile factory. Smoke from the factory’s kiln was a source of ‘intolerable discomfort’ for her tenants, who duly left, at which point she invoked troubles de voisinage against the owner of the factory for ‘depreciating … the rental value’ of her land.81 Overturning the decision of the Cour d’appel of Lyon, the Cour de cassation held that it was irrelevant that the claimant had come to the nuisance. The reasons underpinning its position are not explained in its typically laconic judgment, but in his note on this decision, Ripert provides a compelling justification which mirrors that of Baudry-Lacantinerie and Chauveau which we have just examined: the defendant caused the damage when he began using his land in the way of which the claimant complained, but the damage became apparent only when the claimant sought to use the land in a particular way.82 The rejection of the concept of coming to the nuisance in Dupont v Lecante is of­ 74 C Demolombe, Traité des servitudes ou services fonciers (Paris, Durand and Hachette, 1855) vol 2, 157–59. 75 ibid 158, 159. 76 ibid 158. 77 ibid. 78 In 1907, Ripert noted that Demolombe’s ‘theory had not been widely accepted’: G Ripert, note to Cass civ 18 February 1907 (n 17) 386. 79 G Baudry-Lacantanerie and M Chauveau, Traité théorique et pratique de droit civil: des biens, 3rd edn (Paris, Larose and Tenin, 1905) 169. 80 Cass civ 18 February 1907 (n 17) 386. 81 ibid. 82 G Ripert, note to Cass civ 18 February 1907 (n 17) 386.

234  Ciara Kennefick particular significance given the facts of the case: the claimant had purchased the second plot, which was even closer to the factory, two years after the first and, as the Cour d’appel of Lyon noted, ‘it is physically impossible that … she did not notice … the smoke coming from the kiln’ while supervising the construction on the first plot.83 This position was consistently applied in subsequent cases.84 One apparent exception though is the decision of the Cour de cassation in 1968 concerning noise from jet engines at Nice airport,85 the culmination of years of litigation which, unusually for the law of civil liability, had captured imaginations beyond the closed circle of jurists.86 It was held that the claimant was ‘at fault’ for the harm which he had suffered, as the airport predated the construction of his two apartment blocks which were adjacent to the runway. Arguably, this decision is not actually inconsistent with the previous case law. First, it was article L 141-2 of the Code de l’aviation civile, rather than the judge-made troubles de voisinage, which applied on the facts.87 Moreover, the Cour de cassation applied an implausibly and unprecedentedly wide understanding of ‘fault’, given that the jet engines had appeared in the airport only after the construction of the claimant’s apartment blocks. The anomalous character of the decision in this case is further underlined by the fact that just over two months earlier, a different chamber of the Cour de cassation reached the very opposite conclusion in a case where the claimant had built a house adjacent to the wall of a noisy factory, an act which appears to be far more deserving of the label ‘fault’ than that which occurred in the case of Nice airport.88 1976 was a watershed in the story of coming to the nuisance in French law. In that year, Parliament introduced a rule which provided that a defendant would not be liable for nuisances caused by ‘industrial, agricultural, commercial or a­ rtisanal’ uses of land if they predated the claimant’s request for planning permission to construct a building.89 It was initially proposed by two députés, Marc Masson and Bertrand Denis, on 14 April 1976 as an amendment to the text which would become the Code de l’urbanisme.90 Originally, only the defendant’s agricultural use of land was targeted because, as one senator openly acknowledged, this provision was introduced at the behest of ‘farmers who were infuriated’ by the claims against them which had been initiated by ‘urbanites’.91 Vivid portraits of the sounds and 83 Cass civ 18 February 1907 (n 17) 387. 84 See eg Req 26 June 1929, S 1930, 1, 19; Cass civ 4 December 1935, S 1936, 1, 208; Cass civ 30 November 1961, S 1962, 131; Cass civ (2) 22 October 1964 (n 33). 85 Cass civ (2) 8 May 1968, D 1968, 609. 86 F Derrida, note on TGI of Nice 9 December 1964, D 1965, 221, 222. 87 Therefore, the defendants were the relevant airlines (as ‘the owner[s] of the aircraft’) rather than the airport, and the claim thus fell outside the jurisdiction of the administrative courts. See n 24. 88 Cass civ (1) 20 February 1968, D 1968, 350. 89 Art 70 of the loi no 76-1285 of 31 December 1976 which was incorporated in art L 421-9 of the Code de l’urbanisme. 90 JO Ass nat, 2nd session, 14 April 1976, 1754. 91 JO Sén, 21 May 1976, 1255 (M Chauty).

Nuisance and Coming to the Nuisance  235 smells of French rural life decorated the debates in both chambers: references to singing guinea fowl, cooing wood pigeons, Angelus bells and manure and slurry culminated in the apocryphal anecdote, repeatedly invoked in the debates in the Assemblée nationale, of a Parisian engineer who invoked troubles de ­voisinage because he was being woken up too early by noisy geese on a farm near his ­weekend retreat in the Dordogne. The Assemblée nationale and the Sénat were sharply divided on this provision. Several reasons were provided in the Sénat to justify its opposition, but the most salient claim for our purposes seems to be that such a rule was thought to be unnecessary as courts had already rejected claims on the ground of coming to the nuisance. Sénateur Pillet said that in such cases, the claimant was considered to have been at ‘fault’; he claimed that this was ‘a long-standing ­principle’.92 Yet as the foregoing paragraphs in this section demonstrate, the reverse was true. The Assemblée nationale was outraged by the outright rejection of its proposal, which may explain why the deputés responded by extending its scope to cover ‘artisanal, industrial and commercial uses’ too on the recommendation of Député Eugène Claudius-Petit.93 Nonetheless, the focus of the provision was still primarily on urbanites coming to live in the countryside; the industries and artisans which the deputés described in their examples were mainly rural.94 In December 1976, after several lengthy debates on coming to the nuisance, one deputé dismissed as a ‘soundbite’ the commitment made by the government ‘to consider going even further’ at a later date, but that is indeed what happened four years later,95 when it was decided to treat purchases and leases of buildings as equivalent to requests for planning permission to construct a building.96 Therefore, since 1980, a claimant is considered to have come to the nuisance if he requests planning permission to construct a building or if he purchases or leases a building. As we shall see in section III, this amendment removed any doubts as to whether this version of coming to the nuisance was in essence a servitude. This was also the moment when the rule on coming to the nuisance was transferred to article L 112-16 of the Code de la construction et de l’habitation, where it still resides today. The final piece was added in 2003 when Parliament extended the protection provided by article L 112-16 to ‘aeronautical’ uses of land.97 The objective was to enshrine expressly in legislation the case law of the Cour de cassation in this particular area, which, as we have seen, was already distinctive,98 thereby clarifying the law and avoiding ‘futile litigation’.99 This amendment was influenced by an older but unsuccessful attempt to extend article L 112-16 to

92 JO

Sén, 13 December 1976, 4421, 4423. Ass nat, 1st session, 23 November 1976, 8586–88. 94 ibid 8587. 95 JO Ass nat, 18 December 1976, 9681. 96 The loi no 80-502 of 4 July 1980. 97 Art 72 of the loi no 2003-590 of 2 July 2003. 98 See the text from n 84 to n 89. 99 B Laurent, Rapport fait au nom de la commission des lois, 18 November 1992, 25. 93 JO

236  Ciara Kennefick aeronautical uses but also, intriguingly, to roads, tourism, culture, and sport.100 Therefore, the French version of coming to the nuisance could have been even more sweeping than it is now. Many legal scholars have treated article L 112-16 with hostility, as it effectively provides ‘a right to cause harm to others’,101 but this view is certainly not unanimous; article L 112-16 was commended in a recent study for providing some degree of ‘legal certainty’ to a defendant facing the ‘random’ ways in which troubles de voisinage can be applied in practice.102 Neither perspective, however, appears to explain the application of this provision in the courts. While article L 112-16 has generally been interpreted in a way which is generous to the claimant, it is not possible, as one commentator has observed, to say that this is the result of ‘a ­deliberate choice on the part of the judges’.103 An analysis of three features of this rule suffices to demonstrate the notably narrow way in which it is applied in practice. First, the courts have applied rather exactingly the requirement, introduced in 1980, that the use of land must comply with relevant public law regulations. The most extreme instance is perhaps the case of the golf club near Montpellier in which the Cour de cassation refused to apply article L 112-16, even though it was factually impossible for the defendant to fulfil this condition, as no public law regulations governing this particular use of land existed.104 Furthermore, article L 112-16 has been held not to apply if there are no buildings on the land, as there are then ‘no occupants of a building’ as the text strictly requires.105 Finally, the protection provided by article L 112-16 extends only to the uses to which the defendant put his land before the claimant came to the nuisance. The significance of the status quo ante has constituted a necessary condition since the very inception of this rule in 1976. An especially picturesque application of this rule is found in the decision not to apply article L 112-16 in a case where a lowland farmer had replaced the original bell on his leading cow with ‘a large and powerful bell generally used in the mountains’ after the claimant had acquired her house nearby; consequently, the intolerable noise could not be justified by the status quo ante.106 Significantly, this feature of the rule had been a source of contention during the parliamentary debates on article L 112-16’s predecessor in 1976. The Assemblée Nationale had sought to protect defendants even if their uses of land had altered or become more i­ntensive, 100 ibid 20–21, 25–27. 101 J-L Bergel, M Bruschi and S Cimamonti, Traité de droit civil, Les biens (Paris, LGDJ, 2000) 136. See too Viney and Jourdain (n 15) 1221–22; J Carbonnier, Droit civil 3, Les biens, 19th edn (Paris, PUF, 2000) 282. 102 P Stoffel-Munck, ‘La théorie des troubles du voisinage à l’épreuve du principe de précaution: ­observations sur le cas des antennes relais’: D 2009, 2817, 2819. 103 N Namiech, ‘Vingt ans d’interprétation restrictive de l’article L.112-16 du Code de la construction et de l’habitation’ Revue juridique de l’environnement 2001, 47, 56. 104 Cass civ (2) 10 June 2004 (n 18). 105 Cass civ (3) 3 June 1987 no 85-14221. 106 Cour d’appel of Pau, 29 September 2003. See too Cass civ (2) 7 November 1990 no 89-16241, Bull 1990 II no 225, 115; Cass civ (3) 18 January 2005 no 03-18914.

Nuisance and Coming to the Nuisance  237 but the Sénat had not supported this position and it had been excluded from the provision in the compromise which ultimately emerged.107 Perhaps unsurprisingly, the députés made a renewed attempt to alter the law on this point during the parliamentary debates on the amendments made in 1980 but the sénateurs rejected the proposal on the ground that ‘extravagance of all kinds is objectionable’.108 Indeed, troubles de voisinage would, undoubtedly, be almost entirely defunct in practice if article L 112-16 were not restricted to the status quo ante. When assessing the merits of article L 112-16, a glance at French administrative law, an exercise which is often described as internal comparative law, is instructive.109 The Conseil d’Etat developed its own autonomous rule on coming to the nuisance before the advent of article L 112-16’s predecessor in civil law in 1976. It is, therefore, unsurprising that it has held that article L 112-16 does not apply in administrative law.110 Unlike its civil law counterpart, the version of coming to the nuisance in administrative law is not restricted to specific uses of land. However, similarly to civil law, this administrative law rule does not apply if the defendant has gone beyond the status quo ante, the limit for which is the time when the claimant came to the nuisance.111 Finding a version of coming to the nuisance in administrative law ought to lead to a reconsideration of whether it is indeed an appropriate feature of civil law. During the parliamentary debates in 1976 on article L 112-16’s predecessor, Sénateur Michel Chauty, explained that his opposition to this provision was based partly on the fact that ‘we cannot grant an exceptional right to some and withhold it from others’.112 This astute observation could be used to justify a divergence between civil and administrative law on coming to the nuisance. Protecting defendants who are acting in a public capacity in this way in litigation on nuisance is consistent with the special powers of the administration in many other fields. Therefore, ‘exceptional rights’, to borrow Sénateur Chauty’s phrase, are the very essence of administrative law where they are understandable on the ground that they are considered to be necessary for the effective promotion of the public weal but they are anomalous and much more difficult to defend in the sphere of civil law.

III.  Coming to the Nuisance as a Servitude Neither French nor English law draws a sharp distinction between property and torts in the context of nuisance. The close relationship between these two fields 107 JO Ass nat (n 93) 8587–8588; JO Ass nat (n 95) 9681. 108 M Sordel, Rapport fait au nom de la commission des affaires économiques et du plan, 23 April 1980, 47; M Rudloff, Avis fait au nom de la commission des lois, 22 April 1980, 26. 109 See n 24. 110 CE 25 November 1988 no 74009. 111 CE 20 November 1992 no 84223. 112 JO Sén (n 91) 1255.

238  Ciara Kennefick of law manifests itself, nonetheless, in different ways in each system. In English law, the very function of the tort of nuisance is to protect interests in land and while it is frequently joined by the tort of trespass in this endeavour, it is, notably for our purposes, the only tort which is capable of protecting servitudes. In French law, both fields claim troubles de voisinage as their own; no trivial conflation in a civil law system where the coherent classification of rules within codes is at the heart of the rule of law. An examination of this civil wrong can be found in treatises on property law in addition to those devoted to the law of obligations though, strikingly, it is almost entirely absent from one of the leading tomes on the law obligations.113 Similarly, both Book II and Book III of the Code civil, on property and obligations respectively, are suitors for troubles de voisinage in the four proposals for the reform of French private law which have emerged since the beginning of this century.114 The central claim in this section is that one aspect of this entanglement of ­property and torts is common to both systems: the forms of coming to the nuisance which have been developed in French and English law are, in essence, proprietary rather than personal. Honoré devised a philosophically sophisticated framework for considering the distinction between personal and proprietary rights; it is ­especially illuminating here because the reflections are derived from ‘hints found both in continental and … Anglo-American writers’.115 As he explains, the core characteristic of a proprietary right is immunity from divestment on the alienation of property.116 Therefore, the critical question for the purposes of this chapter is whether the successors in title of the person who originally came to the nuisance are precluded from successfully invoking the tort of nuisance or troubles de ­voisinage; alternatively, in the language of servitudes, it is necessary to ask whether the burden is binding on the servient land.117 It is contended that this is already the position in French law, and it may be true of English law too. While this point has not yet been expressly acknowledged, there is ample and consistent evidence that French law effectively regards the concept of coming to the nuisance as a servitude. In Dupont v Lecante, the seminal decision in which coming to the nuisance was decisively rejected, it is notable that the claimant successfully argued that the use of her land had not been ‘restricted … by any servitude’ which would permit the defendant’s actions.118 Furthermore and similar

113 Terré, Simler, Lequette and Chénedé (n 2) 1062 (where it is mentioned in passing) and 1182 (where it appears in a footnote). 114 The Avant-projet Capitant, on property law, is in favour of incorporating troubles de ­voisinage into Book II whereas the other three proposals mentioned in n 13 recommend putting this doctrine into Book III. 115 AM Honoré, ‘Rights of Exclusion and Immunities against Divesting’ (1960) 34 Tul L Rev 453, 461–68, esp 467. 116 ibid 461–68. 117 The question of whether the benefit runs with the dominant land is also important but it is ­unnecessary to examine it for the purposes of this chapter. 118 Cass civ 18 February 1907 (n 17) 387.

Nuisance and Coming to the Nuisance  239 to the older authorities in English law,119 the Cour de cassation placed servitudes at the centre of its reasoning in the rare cases in which it actually explained the reason for which invoking coming to the nuisance was to no avail before the advent of article L 112-16 and its predecessor. In 1968, for example, the sole justification for dismissing the defendant’s argument in this respect was that ‘the fact the factory had been built beforehand … does not impose a servitude on the­ neighbouring land’.120 In none of these cases, however, was the claim for troubles de voisinage initiated by the successor in title to the person who originally came to the nuisance, the litmus test for a proprietary right. By expressly providing that one comes to the nuisance simply by purchasing or leasing a building, article L 116-12 specifically allows such a scenario and as such, it cements the status of coming to the nuisance in French law as unquestionably proprietary.121 Indeed, in most of the cases on article L 112-16, the claimant came to the servient land by purchase or lease from the person who had originally come to the nuisance.122 It is likely that article L 112-16’s predecessor had this effect too even though it did not refer expressly to purchases and leases of the land. Certainly, that was the intention of the drafters. The language of servitudes is conspicuous in the parliamentary debates: it was said in the Sénat that ‘rural life has its servitudes’, and there was a revealing mixture of genres in one deputé’s reference to ‘continuous and discontinuous nuisances’, adjectives which are exclusive to servitudes, the term to which he then reverted in the very next sentence.123 Furthermore, several hypothetical cases evoked by the legislators envisaged the application of this provision to claims initiated by those who had purchased and leased buildings the planning permission for which had been requested by their predecessor in title. It is likely that the English version of coming to the nuisance is also, in essence, a servitude. As explained in the foregoing section, for almost two centuries English judges rejected the idea that coming to nuisance could be regarded as anything other than a servitude. The attempts since then to conjure a separate rule in the law of torts appear to have obscured the extent to which servitudes are simply inevitable in this area of law. In Miller, the Court of Appeal, consciously or unconsciously, treated coming to the nuisance as a servitude. First, as the claimants had, in fact, purchased their house from a developer,124 an application of coming to the nuisance to their claim necessarily assumed the existence of a burden which could bind successors in title to the party who had originally come 119 See section IIB. 120 Cass civ (1) 20 February 1968, D 1968, 350; see too Cass civ (2) 29 June 1977, D 1978, IR, 35. 121 This provision specifies that the purchase or lease must be concluded by acte authentique. An acte authentique is a formal document which can be drawn up only by certain persons, notably n ­ otaries. See art 1369 Cc. 122 Cass civ (2) 10 June 2004 (n 18) is one exception. In this case, the claimant built a house on her land. 123 Arts 688 and 690 to 692 Cc; JO Sén (n 91) 1255; JO Ass nat (n 93) 8586. See too n 140. 124 Lord Denning MR pointedly excoriated the developer as ‘thoughtless and selfish’: Miller (n 18) 978.

240  Ciara Kennefick to the nuisance. Indeed, the detailed discussion of the knowledge of the ­claimants at the time at which they purchased the house would not have made sense if it was a case involving simply a change in the use of land by an existing owner. Similarly, it is notable that in a more recent case in which Simon J indicated that he would have applied coming to the nuisance had he not been bound by ­previous authorities to the contrary, the claimants were the successors in title to the party who had built the houses.125 Finally, one aspect of the version of coming to the nuisance which was developed in Coventry, replicates a restriction on the exercise of easements: the defendant’s use of land must cause ‘no greater nuisance than when the claimant first carried out the building or changed the use’.126 Indeed, Lord Neuberger mentions this feature of the law on easements earlier in his speech when addressing the issue of prescription.127 The principal question, though, is whether the version of coming to the nuisance which was outlined by the UK Supreme Court in the landmark case of Coventry in 2014 also purports to bind transferees of the person who came to the nuisance by altering the use of his land. Lord Neuberger does not directly address this question in his speech, but logic suggests that it ought to be answered in the positive. His starting point is that ‘nuisance is a property-based tort, so … the right to allege a nuisance should, as it were, run with the land’.128 Yet as it was suggested that such a claim should not succeed when the claimant changed the use of his land, the subtraction of the ‘right to allege a nuisance’ from the claimant’s bundle of property rights, to borrow a famous metaphor, is also clearly envisaged. Significantly, two factors suggest that such subtractions ‘run with the land’ too and thus bind claimants’ successors in title. First, Lord Neuberger emphasises that the critical factor is the use of the claimant’s land and not his ‘identity’.129 If that is true of the operation of the claim in nuisance, it must also be true of the operation of the rule on coming to the nuisance as the opposite conclusion would lead to an absurdity: the transferee would be able to make the very claim which was unavailable to his predecessor in title. The burden would, therefore, disappear arbitrarily on the transfer of the land, removing the very advantages which Lord Neuberger sought to achieve by allowing defendants to invoke coming to the nuisance. Indeed, in 1848, restrictive covenants, a category of servitudes which is ‘an extension in equity of the doctrine of negative easements’,130 were elevated from personal to proprietary rights for similar reasons.131 Furthermore, the view that coming to the nuisance is proprietary rather than merely personal is justified by the fact that Lord Neuberger approved the outcome in Miller and, as discussed in the foregoing 125 Watson (n 66) [69]. 126 Coventry (n 14) 842. Of course, this requirement is also similar to that in French law, which is discussed in the text from n 105 to n 109. 127 Coventry (n 14) 838. 128 ibid 841. 129 ibid. 130 London & South Western Railway Co v Gomm (1882) 20 Ch D 562, 583. 131 Tulk v Moxhay (1848) 2 Ph 774, 777–78.

Nuisance and Coming to the Nuisance  241 paragraph, the land in that case had been altered not by the ­claimants but by their predecessors in title.132 Coming to the nuisance not only behaves like a servitude; it transcends the limits placed on servitudes by the law of property. Compared to servitudes, the versions of coming to the nuisance in both French and English law are remarkably wide in two respects: the first relates to types of nuisance and the second concerns the element of time. Both the French and English versions of coming to the nuisance embrace a wider range of nuisances than those which can be legalised by servitudes. The rule on coming to the nuisance which was developed in Coventry applies only to nuisances ‘which affect the senses’133 yet the scope of easements is even narrower.134 The most notable restriction is the need to show that the easement will accommodate the dominant land. Until very recently, the right had to be one ‘of utility and benefit and not one of mere recreation and amusement’,135 which might explain why Lord Denning MR stated in Miller that ‘there is no such easement known to the law as a right to hit cricket balls into your neighbour’s land’.136 However, in 2018, the Supreme Court charted a more liberal course in the important case of Regency Villas v Diamond Resorts, by holding that the requirement of accommodation could be satisfied notwithstanding the fact that a particular use of the servient land was purely recreational.137 In French law, there is also a notable gap between the types of nuisance which fall within article L 112-16 and those which could constitute the subject matter of a servitude. While no type of nuisance falls outside the scope of article L 112-16, servitudes are a much more limited category. In principle, as declared in article 686 Cc, ‘landowners … can create any servitudes which they see fit’, but similar to the requirement of accommodation in English law, articles 637 and 687 require servitudes to be for ‘the use’ of the dominant land.138 There is a distinct paucity of materials in French law on the question of which nuisances can be legalised by servitudes but the cases in which the Cour de cassation expressly rejected the idea of coming to the nuisance on the ground that there was no servitude plainly indicate that some overlap is conceivable.139 Furthermore, in his note 132 It is unclear, though, whether Lord Neuberger directed his mind to this important factual distinction. 133 Coventry (n 14) 841. 134 Easements only are considered here as there is no overlap between nuisance and profits à prendre and as only the exceptional and limited category of nuisances mentioned in the text from n 47 to n 49 could conceivably come within restrictive covenants. 135 Re Ellenborough Park [1956] Ch 131, 177. 136 Miller (n 18) 978. 137 Regency Villas v Diamond Resorts [2018] 3 WLR 1603. 138 The similarities between French and English law on this point are not coincidental: Roman law has shaped the law of servitudes in both systems. Indeed, the law of servitudes has been described as ‘perhaps the most Roman part of English law’: B Nicholas, An Introduction to Roman Law, 2nd edn (Oxford, Clarendon Press, 1975) 148. 139 See the text from n 118 to n 120. In one notable case (also discussed in ch 15 below, p 317, the Cour de cassation rejected the argument that a right deriving from a contractual clause in which the holder

242  Ciara Kennefick on Dupont v Lecante, Ripert assumes that there could be a servitude ­allowing the defendant to emit smoke onto the claimant’s land140 and, reverting to an older oracle, it is notable that Domat states that emanations such as ‘smoke or unpleasant smells like those from tanneries and dyers, and other types of disturbances which one neighbour can cause another, must be endured, if a servitude has been established’,141 thus indicating that such servitudes are indeed possible. An analysis of the acquisition of servitudes by prescription is also instructive, as the role of time demonstrates especially plainly the worrying ease with which burdens can be created in both systems by the rules on coming to the nuisance. In French law, article 690 Cc provides that the prescription period for servitudes is 30 years. In contrast, any period of priority in time suffices for article L 112-16. Furthermore, unlike coming to the nuisance, prescription is limited to ‘continuous and apparent’ servitudes only;142 thus important servitudes such as the right to emit noise and smoke cannot be acquired in this way. It is also significant that prescription applies only between owners of land; lessees in French law are, therefore, excluded as they are considered to hold property rights rather than ownership.143 In contrast, article L 112-16 expressly includes the lease of a building. A very similar pattern is apparent in English law. While any period of priority in time suffices in the case of coming to the nuisance, the prescription period for easements is generally 20 years. In addition, prescription may not turn all types of nuisance into an easement: while the Supreme Court has confirmed that of a droit de superficie in land above the famous Creusot mine had agreed not to claim damages for loss caused by the ordinary operation of the mine, was merely personal: Cass civ 12 December 1899, D 1900, 1, 361. The successor in title to the original holder of the droit de superficie was, therefore, precluded from claiming damages for lost profits. However, it is notable that the court eschewed the term servitude, describing the effect of the clause instead as ‘a renunciation of one of the attributes of the droit de superficie … and … thus a restriction of this right for the benefit of the owner [of the mine]’: D 1900, 1, 361, 363. Such apparently deliberate ambiguity may be the result of the fact that it was then unclear if a droit de superficie qualified as ownership or was simply a property right: J-E Labbé, note to Cass 27 April 1891, S 91, 1, 369, 369–70. Indeed, there are still diverging views on this question: F Terré and P Simler, Droit civil, Les biens (Paris, Dalloz, 2018) 843. If a droit de superficie does not constitute ownership, it is, in theory, not capable of being subject to a servitude. See n 143 on the distinction between ownership and property rights in French law. 140 However, such a servitude could not have been created on the facts as it was ‘neither continuous nor apparent’ and thus fell outside the scope of the doctrine of destination du père de famille in arts 692 to 694 Cc: Cass civ 18 February 1907 (n 17) 386. This method of implying servitudes greatly influenced the development of the English law, where the imported terms ‘continuous and discontinuous’ still cause difficulties today: Wood v Waddington [2015] EWCA Civ 538, [14]–[21]. 141 J. Domat, Les Lois Civiles dans leur ordre naturel (Paris, Coiguard, 1689) 420. 142 Arts 690 and 691 Cc. This position, which is strikingly different to usucapio in Roman law, seems to have been intended to be a compromise between the law in the pays de droit écrit and that in the pays de droit coutumier: P-A Fenet, Recueil complet des travaux préparatoires du Code civil (Paris, Marchand du Breuil, 1827) vol XI, 325–27. 143 This distinction between ownership and property rights is a fundamental idea in French property law. Property rights are conceived as particular fragments which are detached from ownership which, as art 544 Cc proclaims, is the ‘right to enjoy and dispose of things in the most absolute way’. A servitude, therefore, is a property right and cannot constitute ownership.

Nuisance and Coming to the Nuisance  243 certain nuisances such as noises and smells are not excluded in principle from prescription, it may be difficult in practice to prove that they have blossomed into servitudes.144 Furthermore, apart from excluding damage other than that which interferes with ‘the senses’,145 coming to the nuisance does not appear to be restricted in quite the same way. Finally, prescription operates in favour of and against freehold land only,146 the most absolute form of ownership known to the common law. Significantly, Jessel MR highlighted this limitation in Sturges: ‘it appears that in fact both the Plaintiff and the Defendant are lessees under the Duke of Portland … On what theory of law I am to presume a grant as against the Duke of Portland’s lessee I do not know’.147

IV. Conclusions In essence, comparative law is akin to geometry: it projects legal systems onto the planes of others. The transposition of one system into the coordinates of the other illuminates properties of both systems which were previously overlooked. There are three principal findings of this kind here. First, while there are many striking parallels between the tort of nuisance and troubles de voisinage, liability under the latter is quite extensive in comparison to the former. Secondly, versions of coming to the nuisance, which restrict liability, emerged in both systems following a long period in which it was held that only a servitude would suffice. The French variety, article L 112-16, is conspicuously wider than its English equivalent. Viewed in isolation, this finding might not appear to be peculiar: the system in which defendants are exposed to more liability is the one in which one finds the wider version of coming to the nuisance. Yet article L 112-16 is surprising: however narrowly it has been interpreted by the courts, such a sweeping restriction of liability fits poorly with the general, and indeed pronounced, propensity of the French law of torts to be ‘victim-oriented’.148 A third parallel concerns the (terminological, conceptual and doctrinal) entanglement of torts and servitudes in both systems. It  reveals a problem common to both systems: coming to the nuisance is, in essence, proprietary and when examined closely, it is clear that it is even more invasive than a servitude. Unlike geometry, comparative law can take a normative turn: it is not limited to presenting in novel ways the immutable features of the past and the present. The findings in this chapter indicate that certain changes are desirable. In English law, it is contended that Lord Neuberger’s obiter introduction of a form of coming to the nuisance in Coventry should not be followed: as we saw in 144 Coventry (n 14) 837. 145 See the text from n 66 to n 67. 146 EH Burn and J Cartwright, Cheshire and Burn’s Modern Law of Real Property, 18th edn (Oxford, Oxford University Press, 2011) 670. 147 Sturges (n 55) 855. 148 J-S Borghetti, ‘The Culture of Tort Law in France’ (2012) 3 JETL 158.

244  Ciara Kennefick section III, this doctrine encroaches in an unprincipled way on easements. This innovation in the law of torts is even less justified now that some significant restrictions on the law of easements have been removed.149 Such an argument is not very radical. As section II demonstrates, the adoption of coming to the nuisance is a very recent development in English law. Only a revolutionary change in the law of property, such as the abolition of prescription which was recently considered and rejected by the Law Commission, would justify the retention of coming to the nuisance.150 In French law, the case in favour of abolishing article L 112-16 is different but equally compelling.151 The content of article L 112-16 is unjustifiable: as we have seen, only certain activities benefit from the special treatment bestowed by this provision, and parliamentary debates suggest that the choices were made rather capriciously by political fiat rather than by legal logic. In contrast, the versions of coming to the nuisance in English law and in French administrative law are, at least, shaped by a general principle rather than arbitrary sectional interests. Why  should business uses of land fall under the protective wings of article L 112-16? Why are domestic and other uses less deserving? Furthermore, as explained in section III, article L 112-16 is, like its English counterpart, essentially proprietary. Therefore, if article L 112-16 is preserved, it ought to take its place alongside the other servitudes enumerated in Book II of the Code civil,152 as recommended by one of the recent law reform proposals.153 The location of rules is a matter of great importance in a codified system which, by definition, is founded on the logical and coherent organisation of rules. The two alternatives which appear in the other reform proposals are especially unsatisfactory in this respect. Inserting article L 112-16 in Book III of the Code civil, which deals with obligations, would obscure the proprietary nature of this provision.154 On the other hand, leaving article L 112-16 to languish in the Code de la construction et de l’habitation while troubles de voisinage alone is triumphantly enshrined in the Code civil would unjustifiably sever the two sides of this civil wrong and, thus, result in the presentation of a very misleading picture of its real scope in the document which is considered to be the centrepiece of French law.155 149 See above text from n 134 to n 137 and from n 143 to n 144. 150 Law Commission, ‘Making Land Work: Easements, Covenants and Profits à Prendre’ (Law Com No 327, 2011) 42–44. 151 The constitutional avenue is closed as it has been held that art L 112-16 is not inconsistent with the Constitution: Conseil Constitutionnel, 8 April 2001 no 2011-116. The court considered a very limited number of points but a challenge based on other foundations (such as those developed in this chapter) is now completely foreclosed as the court’s sweeping declaration that L 112-16 is consistent with the Constitution precludes all future challenges: FG Trébulle, ‘Le Conseil constitutionnel, l’environnement et la responsabilité: entre vigilance environnementale et pré-occupation’ (2011) RDI 369. 152 See the text from n 6 to n 10. 153 Avant-projet Capitant (art 630). See too n 114. 154 The Avant-projet Terré (art 24) recommends this option. 155 By omitting art L 112-16, both the Avant-projet Catala (art 1361) and the current Projet de réforme de la responsabilité civile (art 1244) appear to favour this approach.

part v Causation

246

12 Liability for Alternative Causation and for the Loss of a Chance NUNO MANUEL PINTO OLIVEIRA

I.  Introduction: Liability for Uncertain, (Merely) Partial Causation under French, Austrian and German Law1 My starting point is a comparison of articles 1238 and 1240 of the Projet de réforme de la responsabilité civile, which govern partial causation, with §§ 1301 and 1302 of the Austrian Civil Code, § 830 of the German Civil Code and § 1294 of the Austrian reform project of 2006–07. Article 1240 of the Projet, on liability for alternative causation, deals with concurrence of two events which would trigger liability (faits générateurs de responsabilité),2 whereas article 1238, on liability for the loss of a chance, deals with cases of concurrence of an event which would trigger liability and an event which would not (chance or coincidence).3 1 In this chapter, I have provided English translations for the German texts. In doing so, I have drawn with gratitude on the following translations: M Infantino and E Zervogianni (eds), Causation in European Tort Law (Cambridge, Cambridge University Press, 2017) xxviii, which itself refers to K Oliphant and B Steiniger, European Tort Law, Basic Texts (Vienna, Sramek Verlag, 2011) and for the BGB, BS Markesinis and H Unberath, The German Law of Torts. A Comparative Treatise, 4th edn (Oxford, Hart Publishing, 2002) and the official translation, available at the site of the German Federal Ministry of Justice, www.gesetze-im-internet.de/englisch_bgb/. On liability for uncertain, especially alternative, causation in comparative law, see generally HLA Hart and T Honoré, Causation in the Law 2nd edn (Oxford, Clarendon Press, 1985) 249–253; G Brüggemeier, Haftungsrecht: Struktur, Prinzipien, Schutzbereich. Ein Beitrag zur Europäisierung des Privatrechts (Berlin/Heidelberg/New York, Springer, 2006) 188–92; H Koziol, Basic Questions of Tort Law from a Germanic Perspective (Wien, Jan Sramek, 2012) 140–62; S Steel, Proof of Causation in Tort Law (Cambridge, Cambridge University Press, 2015); S Steel, ‘Justifying Exceptions to Proof of Causation in Tort Law’ (2015) 78 MLR 729; EJ Weinrib, ‘Causal Uncertainty’ (2016) 36 OJLS 135; NMP Oliveira, ‘Em tema de causalidade alternativa’ in AC Guedes and NMP Oliveira (eds), Colóquio de direito civil de Santo Tirso. O Código Civil 50 anos depois: balanço e perspectivas (Coimbra, Livraria Almedina, 2017) 513ff. 2 The Digest of European Tort Law refers to ‘damage caused by less than all possibly harmful events outside the victim’s sphere’: B Winiger, H Koziol, BA Koch, R Zimmermann (eds), Digest of European Tort Law, Essential Cases on Natural Causation (Wien/New York, Springer, 2007) vol I, 4. 3 In spite of the fact that art 1238 of the Projet clearly qualifies loss of a chance as a matter of (reparable) loss, I suppose that the loss of a mere chance may qualify as harm (for instance, in classic cases of

248  Nuno Manuel Pinto Oliveira In Austria, § 1301 of the Civil Code states that ‘[m]ultiple tortfeasors may become liable for unlawfully inflicted damage by contributing thereto jointly, in a direct or indirect manner, by instigating, threatening, ordering, helping, concealing or the like, or by omitting to comply with a special duty to prevent the harm’. § 1302 distinguishes between intentional and non-intentional, ie merely negligent, torts. As regards intentional torts, § 1302 states: ‘[i]f … the damage was caused intentionally … all are liable for one’, whereas as regards non-intentional, negligent torts, § 1301 states: ‘if the damage was caused negligently, and the contributions to it can be determined, each participant is liable only for the part of the damage caused through his negligence. If, however, … the contributions of each to the damage cannot be ascertained, all are liable for one’. In Germany, the first sentence of § 830(1) BGB reads: ‘[i]f several persons have caused damage by an unlawful act committed in common, each is responsible for the damage’ and the second sentence reads: ‘[t]he same rule applies if it cannot be discovered which of several participants has actually caused the damage’. In the Austrian project to reform civil liability,4 the provisions on alternative causation in §§ 1301 and 1303 were to be replaced by a single provision in § 1294. In spite of the fact that this project seems to have been abandoned, it reflected an influential view on civil liability, drawing on Wilburg’s concept of a flexible system,5 and a certain view on uncertain causation, drawing on, for example, Bydlinski’s position.6 The first sentence of § 1294(2) of the Austrian project states

medical liability, where there is perte d’une chance de guérison ou de survie). In fact, if liability for alternative causation were to be about harm, about the causes of harm, and liability for the loss of a chance were to be about (reparable) loss, ie about the consequences of harm, the scope of the doctrine of the loss of a chance would be rather limited – it could not apply to cases where no causal relationship could be established between the action attributed to the defendant and the harm sustained by the claimant. 4 For a general presentation of the Austrian project, see for instance I Griss, G Kathrein and H Koziol (eds), Entwurf eines neuen österreichischen Schadensersatsrechts (Wien/New York, Springer, 2006); H Koziol, ‘Die grossen Züge des österreichischen Entwurfs – Dargestellt an Hand der Regelungen betreffend die geschützte Interessen, die Kausalität, die Sorgfaltswidrigkeit und das Verschulden’ in B Winiger (ed), La responsabilité civile européenne de demain. Projets de révision nationaux et principes européens/Europäisches Haftungsrecht morgen. Nationale Revisionsentwürfe und europäische Haftungsprinzipien (Bruxelles/Genève, Bruylant/Schulthess, 2008) 3–24, and, for a translation into English, see H Koziol, ‘Draft proposal submitted by the working group set up by the Federal Ministry of Justice for a new Austrian Law of Damages’ in Koziol (n 1) 325–43. 5 On the concept of a ‘flexible system’, see, for instance, Koziol (n 1) 14–16; F Bydlinski, ‘Bewegliches System und juristische Methodenlehre’ in F Bydlinski, H Kreijci, B Schilcher and V Steininger (eds), Das Bewegliche System im geltenden und künftigen Recht (Wien/New York, Springer, 1986) 21–42; H Koziol, ‘Das bewegliche System. Die goldene Mitte für Gesetzgebung und Dogmatik’ Austrian Law Journal, 3-2017, http://alj.uni-graz.at/index.php/alj/article/view/111; P Ancel, ‘Rôles respectifs de la loi et du juge dans les projets européens’ in Le droit français de la responsabilité civile confronté aux projets européens d’harmonisation (Paris, IJRS, 2012) 33–57; J Lete and M Carballo, ‘Rôles respectifs de la loi et du juge. Rapport de synthèse’ in Le droit français de la responsabilité civile confronté aux projets européens d’harmonisation, 59–64. 6 See F Bydlinski, ‘Mittäterschaft im Schadensersatzrecht’ 156 Archiv für die civilistische Praxis (1956) 410; F Bydlinski, Probleme der Schadensverursachung nach deutschem und österreischischem Recht (Stuttgart, Enke, 1964). On the influence of Bydlinski’s view in Germany, see K Larenz and CW Canaris, Lehrbuch des Schuldrechts, 13rd edn (Munich, CH Beck, 1994) vol II, 564–80; G Wagner, ‘Proportionalhaftung für ärztliche Behandlungsfehler de lege lata’ in G Müller, E Osterloh and T Stein

Liability for Alternative Causation and for the Loss of a Chance  249 that ‘[d]amage can be attributed to a person if he caused it or the causative event was otherwise within his sphere’, and the second sentence explains that the rule in the first sentence applies to cases where there are two alternative, causal events: ‘if only one or the other of the events could have caused the damage … then the damage is to be apportioned according to the weight of the respective grounds for imputation and the likelihood of causation’.7 In this comparison of the French Projet with Austrian and German law, I would like to challenge the assumption that liability for alternative causation and liability for loss of a chance are two completely different, fully independent problems. First, as to the assumption that the two problems are completely different, I would suggest that they are comparable, similar problems as in both cases we are dealing with uncertainty.8 The (so-called) conceptual ‘shift from causation to [reparable loss]’9 could not eliminate substantial continuity between (i) uncertainty about causation of the final harm and (ii) uncertainty about the relationship between intermediate and final harm. In fact, uncertainty about causation of the final harm frequently results from uncertainty about the relationship between intermediate harm, consisting of the loss of a chance, and the final harm, c­ onsisting of the loss of the advantage the chance would have procured.10 (eds), Festschrift für Günter Hirsch (Munich, CH Beck, 2008) 453–67; G Wagner, ‘Proportionalhaftung bei mehreren möglichen Schadensursachen‘ in T Eger, C Ott, J Bigus and G von Wangenheim (eds), Internationalisierung des Rechts und seine ökonomische Analysis/Internationalization of the Law and Its Economis Analysis. Festschrift für Hans-Bernd Scháfer zum 65. Geburtstag (Frankfurt-am-Main, Gabler Edition Wissenschaft, 2008) 193–206; G Wagner, ‘Neue Perspektiven im Schadensersatzrecht – Kommerzialisierung, Strafschadensersatz, Kollektivschaden’ in Verhandlungen des 66. Deutschen ­Juristentages Stuttgart, 2006, vol I – Gutachten (Munich, CH Beck, 2006) 53ff; G Wagner, ‘The Law of Torts in the Draft Common Frame of Reference’ (2009) 24–26; C Seyfert, Mass Toxic Torts (Berlin, Duncker & Humblot, 2004); R Wilhelmi, Risikoschutz durch Privatrecht. Eine Untersuchungen zue ­negatorischen und deliktischen Haftung unter besonderer Berücksichtigung von Umweltschäden ­(Tübingen, Mohr Siebeck, 2009). 7 For a commentary on § 1294 of the Austrian reform project, see F Bydlinski, ‘Die Verursachung im Entwurf eines neuen Schadensersatzrecht’ in Griss, Kathrein and Koziol (n 4) 37–45; H Koziol, ‘Die grossen Züge’ (n 4) 15–16; JF Sinde Monteiro, ‘Définition de la causalité dans les projets nationaux’ in Le droit français de la responsabilité civile confronté aux projets européens d’harmonisation (Paris, IRJS Editions, 2012) 369–89, esp 380 and 383–85. 8 See, for instance, C van Dam, European Tort Law, 2nd edn (Oxford, Oxford University Press, 2013) 337–42; N Jansen, ‘The Idea of a Lost Chance’ (1999) 19 OJLS 271; A Beever, Rediscovering the Law of Negligence (Oxford, Hart Publishing, 2007) 498ff and 503 (‘so called “loss of a chance” is in fact a form of causal uncertainty’); T Kadner-Graziano, ‘La “perte d’une chance” en droit privé européen: “tout ou rien” ou réparation partielle du dommage en cas de causalité incertaine’ in C Chappuis and B Winiger (eds), Les causes du dommage: journée de la responsabilité civile 2006 (Genève, Schulthess, 2007) 218–48; T Kadner-Graziano, ‘Loss of a Chance in European Private Law: “All or Nothing” or Partial Liability in Cases of Uncertain Causation’ (2008) ERPL 1009; Weinrib (n 1) 135–64, at 159 (‘the idea of loss of a chance seems to apply whenever factual causation is known with less than complete certainty’). 9 See O Moréteau, ‘Basic Questions of Tort Law from a French Perspective’ in H Koziol (ed), Basic Questions of Tort Law from a Comparative Perspective (Wien, Jan Sramek, 2015) 1–95, at 55. In spite of the fact that Moréteau refers to a ‘shift from causation to damage’, I suppose that, in keeping with the conceptual framework of the French Projet, as translated by Whittaker and Borghetti, ‘damage’ should be replaced with ‘harm’ or ‘loss’, perhaps with ‘reparable harm’ or ‘reparable loss’. On the distinction between the two, see above, ch 9. 10 See, for instance, F G’sell-Macrez, ‘Recherches sur la notion de causalité’ (Thesis, Université de Paris I – Panthéon-Sorbonne, 2005) 493ff.

250  Nuno Manuel Pinto Oliveira Secondly, as to the assumption that the two problems are fully independent, I would suggest that comparable problems should be resolved by consistent principles and rules. Article 1240 of the Projet refers to two elements as being relevant to establishing liability for alternative causation, ie to the degree of coordination amongst tortfeasors (‘acting in concert’) and to the nature of the claimant’s loss (personal injury (dommage corporel)). Article 1238 of the Projet refers to a third element, relevant to establishing liability for the loss of a chance, ie to the likelihood of actual causation of the final harm. In the event that liability has been established, article 1240 renders each merely potential, alternative tortfeasor liable for the whole of the loss and article 1238 renders each merely potential tortfeasor liable for a part of the final, definitive loss, proportional to the likelihood of his contribution.

II.  Concurrence between Two Events which would Trigger Liability The concurrence of (at least) two events which trigger liability, as alternative causes, means that there are (at least) two persons who may be held liable for the damage.11 In the hypothesis of actual causation, ie of actual contribution to the damage, French, Austrian and German civil law provide for a rule of collective liability. In French law, all tortfeasors are liable in full. Even though the classification of their obligation is open to discussion and even though it is unclear whether actual tortfeasors owe a joint and several obligation (an obligation solidaire, or an obligatio in solidum),12 it is clear that the victim may claim compensation from any debtor

11 Viney thus prefers the formula ‘alternative liability’ (responsabilité alternative) to alternative causation and defines alternative liability as referring to situations where ‘the harmful action can only have been committed by one of the persons involved’ (see G Viney, ‘L’espoir d’une recodification du droit de la responsabilité civile’ D 2016, no 24, 1378–85). 12 In accordance with the French Code civil, as amended by the Ordonnance of 2016, joint and several obligations have both primary and secondary, ancillary effects. The primary effects of joint and several obligations result from art 1313 Cc: firstly, ‘[t]he joint and several nature of an obligation amongst debtors imposes on each of them an obligation for the whole of the debt’ and, secondly, ‘[s]atisfaction by one of them [ie by one of the debtors] discharges them all as regards the creditor’. The secondary, ancillary effects of joint and several obligations now result from art 1314 Cc: ‘A claim for interest made against a single joint and several debtor causes interest to run against all’ (for the translation of arts 1313 and 1314 Cc, see J Cartwright, B Fauvarque-Cosson and S Whittaker, The Law of Contract, the General Regime of Obligations, and the Proof of Obligations. The new provisions of the Code civil created by Ordonnance n° 2016-131 of 10 February 2016 and including revisions made to the text by Loi no 2018-287 of 20 April 2018 translated into English, www.textes.justice.gouv.fr/art_pix/Translationrevised2018final.pdf. The practical significance of the distinction between joint and several obligations, or obligations solidaires, and obligations in solidum, presupposes the distinction between primary and secondary effects of joint and several liability. In what concerns primary effects, there is no difference between obligations solidaires and obligations in solidum. In both obligations solidaires and obligations in solidum, the creditor may require satisfaction from any joint and several debtor he may choose.

Liability for Alternative Causation and for the Loss of a Chance  251 he may choose.13 In Austrian and German law, all tortfeasors are liable jointly and severally for the whole loss. In the hypothesis of a merely potential causation, ie of merely potential contribution to the damage, French, Austrian and German law feature corresponding rules of collective liability. The Austrian and German rules apply directly, for alternative causation is within the scope of both § 1301 of the Austrian Civil Code and § 830(1) BGB. The French rule on liability in solidum applies indirectly, by means of analogy. Irrespective of their legislative starting-points, both French and German case law found a similarity between multiple actual tortfeasors and multiple potential tortfeasors. French courts started requiring intentional coordination – multiple potential tortfeasors should be held jointly and severally liable if, but only if, they had acted in concert. German courts started requiring slightly less than intentional coordination – multiple tortfeasors should be held jointly and severally liable if, but only if, they had acted in a similar way.14 Over the course of the decades, it was progressively realised that the requirement of coordination among multiple, potential tortfeasors was too restrictive. In fact, it would exclude liability, for instance, in cases relating to blood transfusions or defective pharmaceuticals, such as diethylstilbestrol and thalidomide.15 In France, there is long-standing case law on liability for collective torts, where the unidentified author of harm ‘fait partie d’un ensemble déterminé de personnes potentiellement responsables’.16 The Avant-projet Catala and the Avant-projet Terré summarised the criteria developed by the courts.17 The Avant-project Catala of 2005 stated that ‘[w]here harm is caused by an unascertained member of a group, all its identified members are answerable for it jointly and severally, except that

In what concerns secondary, ancillary effects, obligations solidaires have both primary and secondary effects, while obligations in solidum do not. Obligations in solidum have primary effects, and primary effects alone. It follows that an act which interrupts the running of the limitation period with regard to one debtor in solidum does not interrupt the running of the limitation period against the other, or the others, and that a claim for interest made against a single debtor in solidum does not cause interest to run against all. On this area of the law, see also below, ch 13 pp 266–267. 13 In keeping with art 1265 of the Projet de réforme, all persons who have actually contributed to the causation of the damage are jointly and severally liable: ‘[w]here one or more persons are liable for the same harm, they are jointly and severally liable to make reparation for it to the victim’. 14 See, for instance, Brüggemeier (n 1) 960. 15 The case of diethylstilbestrol (DES) counts as a good example. DES was a drug put into the market by multiple manufacturers, which was used by pregnant women in the 1950s, 1960s and 1970s. The most important side-effect of the drug, cervical cancer, was only apparent 20 to 30 years after birth. Therefore, it was practically impossible for the claimant to determine which manufacturer exposed her to the risk of cervical cancer (see eg C van Dam (n 8) 332–34; C Quézel-Ambrunaz, ‘La fiction de la causalité alternative, Fondement et perspectives de la jurisprudence “Distilbène”’ D 2010, 1162; T Thiede, ‘Defective Pharmaceuticals and Indeterminate Tortfeasors: A German Law Perspective on DES-Daughters Scenarios’ (2013) ERPL 617). 16 See, for instance, G Viney, P Jourdain and S Carval, Traité de droit civil, Les conditions de la ­responsabilité, 4th edn (Paris, LGDJ, 2013) 294 ff. 17 ibid 307–09.

252  Nuno Manuel Pinto Oliveira any one of them may escape liability by showing that his own actions were not implicated’,18 and the later Avant-projet Terré of 2011, stated that ‘[w]here harm is caused by an unascertained member of a group of persons acting in concert, each of them is answerable for all of it, except that he may escape liability by showing that his own actions were not implicated’.19 The concept of persons acting in concert was kept both by the Avant-projet de loi of 2016 and by the current Projet de réforme. In the wake of the Cour de cassation’s rulings of 2009 on the diethylstilbestrol cases,20 the Avant-projet de loi proposed to go beyond collectively committed torts, so as to apply to independently committed torts ‘for similar reasons’ (pour des motifs similaires).21 This was criticised on the grounds that it went too far, for collective, common liability would be alien to French law and to French legal culture,22 but conversely also on the grounds that it did not go far enough, for French law should not require potential tortfeasors to act in concert or to act on similar grounds or for similar reasons.23 The Projet de réforme takes account of both criticisms. Whereas the Avant-projet de loi hesitated to limit the special rule to liability for personal injury (dommage corporel), article 1240 of the Projet de réforme provides that deviations from the rule requiring actual causation to be proved by the claimant should be limited to cases of personal injury.24 Whereas the Avant-projet proposed to extend liability from coordinated actions to actions based upon similar reasons, the Projet proposes to extend liability from actions based upon similar reasons to similar actions. It requires that all potential ­tortfeasors exercise similar activities. In Germany, the first sentence of § 830(1) BGB is clear. It requires intentional coordination, stating that ‘[i]f several persons have caused damage by an ­unlawful act committed in common, each is responsible for the damage’. The second sentence of this provision is, however, a trifle unclear and even obscure, providing that ‘[t]he same rule applies if it cannot be discovered which of several participants has actually caused the damage’. The problem is that ‘the same rule’ (das Gleiche gilt, literally, ‘the same applies’) may be interpreted in two different ways. On the 18 Art 1348. 19 Art 12. 20 See Cass civ (1) 24 September 2009 nos 878 and 880, www.courdecassation.fr/jurisprudence2/ premiere_chambre_civile_568/arrets_n_13645.html. 21 Art 1240 of the 2016 avant-projet de loi. 22 See M Fabre-Magnan, Les obligations, (Paris, PUF, 2004) 747: ‘this form of collective liability where innocent persons are found liable is alien to our legal culture’. 23 See Viney (n 11) 1380: ‘[w]hen it is established that the harm is necessarily due to the action of one of the persons against whom claims for liability have been brought, without it being possible to determine which [one has caused the harm], each is reputed to have caused it and it is for her to prove that she is not the author of the damage’. 24 In limiting the scope of liability for alternative causation for personal injury, the Projet de réform conforms to the criticism by M Mekki, ‘Le projet de réforme du droit de la responsabilité civile: maintenir, renforcer et enrichir les fonctions de la responsabilité civile’, Gaz Pal, vol 136, no 3, 2016, 1512–14: ‘[i]f a derogation is justified, it is because of the seriousness of the harm caused. This presumption should therefore be confined to cases of losses resulting from personal injury.’

Liability for Alternative Causation and for the Loss of a Chance  253 one hand, it may be interpreted so as to refer both to the requirements and the effects of tortious liability. On the other hand, it may be interpreted so as to refer only to the effects of tortious liability. In keeping with the first reading, the second sentence in § 830(1) BGB would require potential tortfeasors to be participants in an ‘act committed in common’; in keeping with the second reading, it would not include such a requirement and could apply to potential tortfeasors acting independently. German courts initially opted for the first reading.25 Accordingly, the scope of the rule should be limited to cases where subjective or objective requirements were met. Firstly, joint and several liability would depend on a subjective requirement, and that subjective requirement would consist of mutual knowledge and, secondly, it would depend on an objective requirement, and that objective requirement would consist of objective connection, in space and time. In keeping with the subjective requirement multiple potential tortfeasors should be held liable if, but only if, each of them knew or should have known about the actions of the others. In keeping with the objective requirement multiple potential tortfeasors should be held liable if, but only if, their actions were to be considered parts of a common process. The Federal Supreme Court used to require the so-called Gleichartigkeit der Gefährdung,26 ie the similarity of the ways of endangering legally protected interests. From the beginning of the 1980s, German case law evolved from the first to the second reading of the second sentence of § 830(1) BGB. ­Accordingly, the scope of the rule was not to be limited to cases where subjective and/or objective requirements are met; therefore, the second sentence was to apply to non-­coordinated, independent potential tortfeasors.27 In Austria, § 1301 of the Civil Code applies to common joint-tortfeasors (Mittäter) whereas § 1302 applies to independent tortfeasors (Nebentäter): common joint tortfeasors act together; independent tortfeasors act separately. Furthermore, common joint tortfeasors always act intentionally; independent tortfeasors may act either intentionally or non-intentionally, ie negligently. In accordance with § 1301, the liability of common joint tortfeasors is always full liability. Case law refers to a presumption that the acts of each tortfeasor are psychologically determined or, at least, psychologically influenced by the acts of

25 See, for instance, G Planck, commentary to § 830, in Bürgerlichen Gesetzbuch, vol II – Recht der Schuldverhältnisse, 3rd edn (Berlin, J Guttentag, 1907) 999–1101; C Crome, System der Bürgerlichen Rechts, vol II – Recht der Schuldverhältnisse (Tübingen/Leipzig, Mohr Siebeck, 1902) 1069; R Saleilles, commentary to § 830, in C Bufnoir, P Cazelles, J Challamel, J Drioux, F Gény, P Hamel, H Lévy Ulmann and R Saleilles, Code civil allemand, traduit et annoté, vol II – Livre II (suite). – Section VII (art. 433 à 853). Livre III (art. 854 à 1296) (Paris, Imprimerie Nationale, 1906) 398: ‘provided that there are several persons involved in a common action from which or from the course of which the harmful action resulted’. 26 See Brüggemeier (n 1) 189. 27 See, for instance, A Staudinger, commentary to § 830, in R Schulze (ed), Bürgerliches Gesetzbuch. Handkommentar, 5th edn (Baden-Baden, Nomos, 2007) 1108.

254  Nuno Manuel Pinto Oliveira the others (psychische Kausalität).28 In accordance with § 1302, the liability of independent tortfeasors may be either full or partial, depending on the nature of each tortfeasor’s fault. In cases of intentional wilful misconduct, each tortfeasor is liable for the whole damage while in cases of non-intentional, merely negligent misconduct, tortfeasors whose contributions could be ascertained are liable for a part of the damage, proportional to their contribution. Tortfeasors whose contributions could not be ascertained are liable for the whole. The Austrian reform project of 2007 confirmed the interaction between the elements of causation, fault and damage within the framework of a flexible system of tort law. The first sentence of § 1294(2) of the Austrian project provided that ‘[d]amage can be attributed to a person if he caused it or if the causative event was otherwise within his sphere’. The second sentence extended the rule to cases where there are two events ‘highly likely to have caused the damage’, even though ‘only one or the other could have caused [it]’. If all potential tortfeasors acted together, a rebuttable presumption applies: ‘it is presumed that each of them caused the entire damage’. If, however, potential tortfeasors acted separately, no presumption applies: ‘damage is to be apportioned according to the weight of the respective grounds for imputation and the likelihood of causation’. It followed from § 1294 that the greater the degree of the defendant’s fault and the greater the danger arising from the source kept within his sphere of influence and interest, the more justified would be liability for merely potential, alternative causation. Furthermore, it followed from § 1294 that the weight of the grounds for imputation is to be compared with the weight of causation, that is to say, with the likelihood of actual causation. As regards coordinated tortfeasors who were acting together, § 1294(2) provides for full liability. Each potential tortfeasor is liable for the whole – for the entire damage. As regards non-coordinated independent tortfeasors, who were acting separately, § 1294(2) provides for partial liability – each potential tortfeasor is liable for his share, and his share is to be proportional to his degree of fault and to the likelihood of causation.

III.  Comparative Remarks on Article 1240 of the Projet de réforme Instead of a flexible provision, similar, for example, to § 1294(2) of the Austrian reform project, article 1240 of the French Projet de réforme establishes a rigid rule, excluding liability for any losses arising from damage to physical property, from purely economic harm, and from purely ‘moral’ harm.

28 E Karner, commentary to § 1301, in H Koziol, P Bydlinksi and R Bollenberger (eds), Kurzkommentar zum AGBG, 2nd edn (Wien/New York, Springer, 2007) 1459.

Liability for Alternative Causation and for the Loss of a Chance  255 It has been pointed out that the expression dommage corporel (personal injury) implies a limitation, and indeed a double limitation, on the scope of the rule on alternative causation. In the first place, article 1240 is meant to apply exclusively to non-­contractual, tortious liability. It follows from article 1233-1 of the Projet that in principle, ‘[l]osses resulting from personal injury are subject to reparation on the basis of the rules of extra-contractual liability, even if they are caused in the course of performance of a contract’. Secondly, within tortious liability, article 1240 is meant to apply exclusively to pecuniary and non-pecuniary losses arising from personal injury.29 Therefore, it precludes compensation of losses arising from physical harm (dommage matériel), from purely economic harm (dommage économique), or from purely ‘moral’ harm, such as mental distress, independent of any personal injury (dommage moral). Instead of a comprehensive provision, similar to § 830(1) BGB, and instead of a flexible provision, similar to § 1294 of the Austrian reform project, article 1240 of the Projet de réforme establishes a rigid rule, precluding liability of non-coordinated, independent tortfeasors acting in a dissimilar, or in a not-so-similar way. If, for instance, the claimant proves that he was the victim of a defective drug, put onto the market by one of two manufacturers, it is submitted that each of them would be held liable; if however the claimant proves that he was the victim of one of two defective drugs, put onto the market by two manufacturers, it is submitted that that none of them would be held liable under article 1240, as they did not exercise similar activities. Finally, instead of distinguishing full liability of coordinated potential tortfeasors and partial proportional liability of non-coordinated, independent potential tortfeasors, article 1240 establishes a single comprehensive rule determining full liability.

IV.  Concurrence between an Event which would Trigger Liability and an Event which would not German and Austrian law conceptualise both cases covered by articles 1240 and 1238 of the Projet de réforme as hypotheses of uncertainty about causation – they assume that the defendant’s (mis)conduct is not proven to be a conditio sine qua non, or even a necessary event of a sufficient set,30 for there are at least two 29 For a critique, see E Lemaire and GM Veldt, ‘La causalité alternative dans l’avant-projet de réforme de la responsabilité civile. Regards comparatifs franco-néerlandais’ in Regards comparatistes sur la réforme de la responsabilité civile. Le rapprochement des responsabilités contractuelle et délictuelle dans l’avant-projet de réforme, abordé sous l’angle du droit comparé (Paris/Amsterdam, Université Paris II Panthéon-Assas/Universeit van Leiden, 2017) 28–37. 30 On the definition of cause as a necessary element of a sufficient set of conditions, see Hart and Honoré (n 1); RW Wright, ‘Causation in Tort Law’, (1985) 73 Cal L Rev 1735; T Honoré, Responsibility

256  Nuno Manuel Pinto Oliveira a­lternative, mutually exclusive causal explanations for the harm the claimant sustained. By contrast, French law conceptualises cases covered by article 1240 as hypotheses of uncertainty and cases covered by article 1238 as hypotheses of certainty about causation: even though for the latter there is uncertainty as to the causation of the final, definitive harm, the defendant’s (mis)conduct is proven to be conditio sine qua non, or a necessary element of a sufficient set of conditions of the claimant’s intermediate harm consisting of the loss of a chance. In France, the doctrine of the loss of a chance reconceptualises reparable loss, ‘thus removing all obstacles of causation in the sense of a conditio sine qua non’.31 In general terms, the case for liability for the loss of a chance is frequent in three types of situations – liability of legal professionals, liability of medical professionals, and competitions ‘in fields such as politics, business, sports and others’.32 The doctrine of the loss of a chance claims that the client’s interest in a chance of winning a case is different, albeit not fully independent, from the interest in winning the case itself.33 With regard to the liability of medical professionals, it claims that the interest of a patient in a chance of cure is different, albeit not fully independent, from his or her interest in a cure itself, and that the interest in a chance of survival is different from the interest in survival itself.34 Even if there is uncertainty as to the causation of the final, definitive harm, there is certainty as to the causation of the intermediate harm, as to the causation of the loss of the chance of cure, of survival, of winning a case and certainty about the causation of the intermediate harm is sufficient to compensate the claimant. The shift in conceptualising loss of a chance from causation to (reparable) loss has both theoretical and practical reasons: from a theoretical point of view, conceptualising the loss of a chance as a distinct reparable loss allows French tort law to apply old, established principles about causation,35 proof of causation36 and (full) reparation of the loss;37 from a practical point of view, it allows French tort and Fault (Oxford, Hart Publishing, 1999) 94–120; T Honoré, ‘Causation in the Law’ in The Stanford Encyclopedia of Philosophy (Winter 2010 edn), http://plato.stanford.edu/archives/win2010/entries/ causation-law/; I Puppe and RW Wright, ‘Causation in the Law: Philosophy, Doctrine and Practice’ in M Infantino and E Zervogianni (eds), Causation in European Tort Law (Cambridge, Cambridge University Press, 2017) 17–59. 31 Koziol (n 1) 153. 32 See Kadner-Graziano, ‘Loss of a Chance in European Private Law’ (n 8) 1015. 33 On the liability of legal professionals for the loss of a chance under French law, see for instance Viney and Jourdain, (n 16) 142–44ff. 34 On the liability of medical professionals for the loss of a chance under French law, see for instance Viney and Jourdain, (n 16) 287–94; L Khoury, Uncertain Causation in Medical Liability (Oxford, Hart Publishing, 2006) 93–142; E Hyslop, European Causation in Tort Law. A Comparative Study With Emphasis on Medical Law in the United Kingdom, Germany and France, and Luxemborg (Luxembourg, University of Luxembourg, 2015). 35 See art 1239(1) of the Projet: ‘Liability supposes the existence of a causal relationship between the action attributed to the defendant and the harm’. 36 Art 1239(2) of the Projet – assuming that the claimant bears the burden of proving causation – provides that ‘[a] causal relationship may be established by any means of proof ’. 37 See art 1258 of the Projet: ‘The aim of reparation is to replace the victim as much as is possible in the situation in which he would have been if the harmful action had not taken place. It must cause him neither a loss nor an advantage.’

Liability for Alternative Causation and for the Loss of a Chance  257 law to harmonise old, established principles with new, innovative, results – for instance, proportional liability.38 In Germany, the rule is that there is no liability for alternative causation in cases of concurrence between an event which would trigger liability and an event which would not. § 823(1) BGB, the basic norm of German tort law, provides an a priori limitation on non-contractual or tortious liability to losses arising from the infringement of absolute rights, such as life, body, health, freedom, or property, thus precluding (mere) chances from being protected interests,39 and § 830(1) BGB provides a limitation on tortious liability for alternative causation to cases of concurrence of two events which would trigger liability (where ‘it cannot be discovered which of several participants [in the unlawful act] has actually caused the damage’).40 In the area of medical liability, however, German law allows for an exceptional relaxation of the no-liability rule, by reversing the burden of proof.41 § 630h(5) BGB, on the so-called Treatment contract, provides that: [i]f gross malpractice has been committed, and if this is susceptible as a matter of principle to cause an injury to life, limb or health of the nature which in fact took place, it is to be presumed that the malpractice was the cause of this injury.

The defendant has to prove that his grossly negligent misconduct was not the cause of the claimant’s loss: ‘he created a (grossly) unjustified risk to the patient, and this suffices for liability even though in many cases the injury did not stem from that risk, but would have occurred anyway’.42 In spite of this recent relaxation of the no liability rule by § 630h(5) BGB, in general German law consistently applies an all-or-nothing approach and allows full liability for merely potential, alternative causation only in particular specific situations.

38 Moréteau (n 9) 55: ‘As unorthodox as things may look from a theoretical point of view, it serves very pragmatic purposes.’ 39 Chances, for instance chances of cure or survival, could not be counted as independent interests: ‘only bodily integrity, health and life are seen as protected legal interests which must be compensated in cases of infringement’, ‘not the chance to become healthier’: Koziol (n 1) 155. Koziol goes even further and qualifies the loss of a chance to become healthier as pure economic loss (ibid). It follows that, in Germany, arguments in favour of liability for the loss of a chance always were stronger in contract than in tort (see eg G Mäsch, Chance und Schaden. Zur Dienstleiisterhaftung beim unaufklärbaren Kausalverläufen (Tübingen, Mohr Siebeck, 2004). 40 German case law consistently holds that it cannot apply to cases of concurrence of an event which would trigger liability and an event which would not, ie it cannot apply (i) where the loss may have been caused by chance or coincidence (Zufall); (ii) where the conduct of at least one potential tortfeasor is not culpable, ie neither intentional nor negligent; (iii) where the conduct of at least one potential tortfeasor is not unlawful, as when he may invoke a defence; and (iv) where harm may have been caused by the victim himself (see Staudinger (n 27) 1109). 41 See, for instance, Staudinger (n 27) 1109; H Lange, Schadensersatz, 2nd edn (Tübingen, Mohr Siebeck 1990) 162–64; C Katzenmeier, Arzthaftung (Tübingen, Mohr Siebeck, 2001) 439–70; G ­Spindler, ‘­Kausalität im Zivil- und Witschaftsrecht’ (2008) 208 Archiv für die civilistische Praxis 283; A ­Stremitzer, ‘Haftung bei Unsicherheit des hypothetischen Kausalitätsverlaufs. ­Berechnungsmethoden am Beispiel der Arzthaftung’ (2008) 208 Archiv für die civilistische Praxis 676, 676–79. 42 M Stauch, The Law of Medical Negligence in England and Germany (Oxford, Hart Publishing, 2008) 157.

258  Nuno Manuel Pinto Oliveira In Austria, alternative causation replaced the doctrine of the loss of chance. § 1304 of the Austrian Civil Code provides that, ‘[i]f, in a case of [loss], there is also fault on the part of the person harmed, he has to bear the loss proportionately with the injurer; and, if the proportion cannot be determined, in equal shares’.43 Bydlinski convincingly argued that the rule on partial proportional liability in § 1304 could apply by analogy to cases where the contribution of the claimant is merely potential and to cases where there is no contribution of the claimant whatsoever, neither actual, nor potential. It could apply to cases where the contribution of the claimant is merely potential, ie where the loss may have been caused by ‘fault on the part of the person harmed’, and it could apply to cases where there is no contribution of the claimant whatsoever, ie where the loss may have been caused by chance or by coincidence (Zufall).44 In spite of criticisms,45 the Austrian Supreme Court (OGH) ruled in favour of applying § 1304 to all cases of alternative causation – liability for merely potential, alternative causation in cases of concurrence of an event which would trigger liability and an event which would not is partial proportional liability.46 The Austrian project to reform civil liability consolidates partial, proportional liability for alternative causation. § 1294(2) dealt both with the concurrence of two events which would trigger liability and with concurrence of an event which would trigger liability and an event which would not. In doing so, § 1294(2) established three basic, fundamental rules. In the first place, the question whether the defendant should be held liable should depend on two elements, on the likelihood of actual causation and on the weight of the grounds for imputation, such as fault and risk. Secondly, if the defendant were to be held liable, the claimant should be awarded partial, proportional compensation. Thirdly, the extent to which the claimant should be awarded compensation depends on the two elements relevant to establishing whether the defendant should be held liable: ‘the damage is to be apportioned according to the weight of the respective grounds for imputation and the likelihood of causation’. Even if § 1295 of the Civil Code, which forms the basic norm of Austrian tort law, does not a priori confine tortious liability to losses arising from the infringement of absolute rights,47 the doctrine of the loss of a chance has been rejected on

43 For the translation of § 1304 of the Austrian Civil Code, see Infantino and Zervogianni (n 30) xxviii (with ‘damage’ instead of ‘loss’). 44 See Bydlinski (n 6). 45 See, for instance, R Welser, ‘Zur solidarischen Scahdenshaftung bei ungeklärter Verursachung im deutschen Recht’ (1968) 9 Zeitschrift für die Rechtsvergleichung 38; M-S Häusler, Haftung ohne Kausalitätsnachweis (Vienna, University of Vienna, 2010); M-S Häusler, ‘Halber Schadensersatz zur “halben” Kausalitätsbeweis?’ (2012) Ecolex. Zeitschrift für Wirtschaftsrecht 17. 46 For further developments, see Koziol (n 1) 146–52. 47 In spite of the fact that § 1295 of the Austrian Civil Code is formally closer to French law, it is substantially closer to German law, for the general clause contained in § 1295 is consistently thought to be compatible with a differentiated, hierarchical system of legally protected interests (see for instance E Karner, commentary to § 1295, in Koziol, Bydlinksi and Bollenberger (n 28) 1434–45).

Liability for Alternative Causation and for the Loss of a Chance  259 two grounds. Interests in chances (eg chances of cure, chances of survival) should not be considered separately from interests in the final, definitive goods (eg life); furthermore, losses arising from the infringement of interests in (mere) chances could not be considered separately from losses arising from the infringement of interests in the advantages ‘which [those chances] would have procured if [they] had been realised’. It has been argued, for instance, that ‘if [health and the chance to become healthier] were really two independent legal goods, then a double claim for compensation would hardly be avoidable’.48

V.  Comparative Remarks on Article 1238 of the Projet de réforme The assumption that there are interests in mere chances is relatively u ­ nproblematic. Jansen accurately points out that the ‘[l]aw is not … concerned with a philosophically true picture of the world – whatever that may be’, but ‘is rather concerned with the world as it is perceived by human beings’, and that human beings apparently perceive the world in a way that renders chances valuable: ‘[p]eople apparently think that chances are important things to be legally protected’.49 The claim that interests in mere chances are different from interests in the advantages ‘which [those chances] would have procured if [they] had been realised’ is, however, highly problematic. In both civil law and common law jurisdictions, it has been argued that ‘[a]lmost any claim for loss of an outcome could be reformulated as a claim for loss of a chance of that outcome’.50 ‘The law could not recognise the independent existence of both [the right to the chance of being freed from an injury and the right to be free from the ultimate injury]’,51 for both rights would refer to ‘different stages of the maturation of the same unreasonable risk’.52 If ‘the two approaches [could not] sensibly live together’,53 the fact that ‘tort law has already chosen to concern itself with the ultimate injuries’54 would preclude that it would concern itself with intermediate injuries: ‘the positing of a right to the chance of avoiding injuries [would] create intolerable tensions’.55 Irrespective of the highly abstract, theoretical problem of whether interests in mere chances are different from interests in the advantages ‘which [those chances] would have procured’, the results of Austrian law turn out to be comparable with



48 Koziol

(n 1) 158. (n 8) at 281. 50 Gregg v Scott [2005] UKHL 2, [2005] AC 176, 224 (Baroness Hale). 51 ibid 225 (Baroness Hale). 52 ibid. 53 ibid. 54 Weinrib (n 1) 160. 55 ibid. 49 Jansen

260  Nuno Manuel Pinto Oliveira the results of French law, with the (possible) difference that French law refers only to the likelihood of actual causation and that Austrian law refers both to the likelihood of causation and to the weight of the grounds for imputation. A practical approach would therefore be that the appropriateness of legal norms on concurrence of an event which would trigger liability and an event which would not (chance or coincidence) should depend on two criteria: (i) on whether they openly address all ‘normatively relevant aspects’ of the situation;56 and (ii) on whether they are consistent with the rules on concurrence of two events which would trigger liability, especially with the rules on concurrence of two events which are imputable to non-coordinated, independent potential tortfeasors. The first criterion suggests that the rule on liability for the loss of a chance could require the court to take into account, for instance, the nature of the defendant’s fault and the nature of the claimant’s harm; the second criterion reinforces the suggestion that it could require the court to take into account the fact that the claimant sustained personal injury as opposed to some other kind of harm. It has been demonstrated that French judges refer to moral, ethical and practical arguments (arguments d’équité57 and arguments d’utilité58) and that, among moral and ethical arguments the nature of the defendant’s fault and the nature of the claimant’s harm have a significant weight in establishing causation.59

VI.  Conclusion – An Interaction of Common Factors? As regards cases of concurrence of two events which would trigger liability, comparison between France, Austria and Germany suggests that four elements are relevant to establishing liability: the degree of coordination among potential tortfeasors, the nature of the defendant’s fault, the nature of the claimant’s rights and interests, and the likelihood of actual causation. Irrespective of whether one is concerned with external or only with internal relations, both the Austrian and the French reform projects suggest that the higher the probability of actual causation, the more justified is liability for merely potential, alternative causation. With regard to cases of concurrence of an event which would trigger l­iability and an event which would not, such as coincidence (Zufall), our comparison suggests there are only two relevant elements, the nature of the defendant’s fault and the likelihood of causation. German case law distinguishes intentional and non-intentional or negligent torts; the Austrian reform project requires the court to balance fault and risk with the likelihood of causation and the French Projet de réforme requires the court to consider the likelihood of causation, ie whether the favourable eventuality, which disappeared, was present and certain. 56 Jansen (n 8) 296. 57 See G’sell-Macrez (n 10) 643–52. 58 ibid 652–77. 59 ibid 645: ‘The nature of the defendant’s fault, its seriousness, and its intentional character constitute obvious arguments when it is necessary to mitigate the uncertainty of the causal connection.’

Liability for Alternative Causation and for the Loss of a Chance  261 In spite of the fact that there are common, constant elements across the three jurisdictions when it comes to establishing liability, in my view they have different ‘strengths’ and different ‘weights’. It follows that the differences between France, Austria and Germany are differences in the interaction of common factors. If one of the elements is absent, or if it is present in a ‘weaker’ form, its absence or relative weakness may be compensated by the presence of other elements in a ‘stronger’ form. Furthermore, I would submit that different ‘strengths’ and different ‘weights’ of common, constant elements are relevant both in determining (i) whether the defendant should be held liable and, if it should, (ii) to what extent. If liability is partial, then it should be proportional to the likelihood of damage, to the nature of the defendant’s fault and/or to the nature of the claimant’s rights and interests.

262

13 ‘Solidary’ Liability and the Channelling of Liability CARLOS GÓMEZ LIGÜERRE*

I. From Obligatio in Solidum to Joint and Several Liability Cases involving multiple tortfeasors deserve a specific rule that grants victims a protection that is not provided by the general liability rule usually designed for single tortfeasors. Two main legal regimes are applicable: joint and several liability, and non-joint (or several) liability.1 Under the first regime, the damage is fully attributable to each and every one of the tortfeasors, and the victim is entitled to ask for recovery from any of them. Under the second regime, every tortfeasor is only liable for a share of the damage, which corresponds to his own contribution to the harm suffered by the victim – provided, of course, that such a share can be identified. In fact, the problems in identifying individual shares of liability in cases involving several tortfeasors, as well as the special protection given to victims in such cases, have driven most European legal systems to apply the joint and several liability solution to cases where several tortfeasors were responsible for the damage. Under French law, however, a rigid principle according to which joint and several liability (arising from obligations solidaires) can arise only by legislation or by agreement has barred the application of this regime in tort law cases. As a result, * The author thanks the participants of the colloquium on ‘French Civil Liability in Comparative Perspective’ which took place in Oxford on 26–28 September 2018 for their comments, as well as the editors of this volume, Professor Simon Whittaker and Professor Jean-Sébastien Borghetti, for their comments and suggestions on a previous draft. Any errors are, of course, my own. 1 The terms ‘joint and several liability’, ‘solidary obligation’ or ‘solidarity’ are used here as equivalent and as an English translation of the French term ‘obligation solidaire’ as regulated under arts 1310–19 Cc and defined by art 131(1) Cc: ‘Where the obligation is joint and several amongst creditors each of them may require and receive satisfaction of the right in full. Satisfaction made in favour of one, who must account to the others, discharges the debtor as regards them all. The debtor may satisfy any of the joint and several creditors as long as he has not been sued by one of them.’ (‘La solidarité entre créanciers permet à chacun d’eux d’exiger et de recevoir le paiement de toute la créance. Le paiement fait à l’un d’eux, qui en doit compte aux autres, libère le débiteur à l’égard de tous. Le débiteur peut payer l’un ou l’autre des créanciers solidaires tant qu’il n’est pas poursuivi par l’un d’eux.’)

264  Carlos Gómez Ligüerre French courts were forced to develop a specific rule, referred to as obligatio in solidum, applicable to cases of multiple tortfeasors’ liability. The solution resembles the general joint and several liability regime, since it makes any tortfeasor liable in full for the damage suffered by the victim. However, unlike the joint and several liability regime, which is established beforehand by legislation or by agreement, the obligatio in solidum is, as mentioned above, a judicial creation. As well as introducing it, French courts have also shaped its legal regime: it works like joint and several liability, since it allows the victim to recover from any liable tortfeasor but, differently from the joint and several liability among several tortfeasors, only some effects of the solidarity are applicable to the liable actors in solidum, since the so-called ‘secondary effects’ of the solidary obligation do not apply. In the current French law of civil liability, the obligatio in solidum solution lives together with cases in which the joint and several liability regime is statutorily applied to specific tortfeasors (eg liability of parents or liability of manufacturers for defective products). The Projet de réforme de la responsabilité civile aims to reduce the complexity of the current French civil liability system.2 The Projet incorporates a general rule for multiple tortfeasors in its article 1265, according to which: Where one or more persons are liable for the same harm, they are jointly and severally liable to make reparation for it to the victim. If all or certain of them have committed a fault, they make contribution to each other in proportion to the seriousness and the causal role of the action giving rise to liability which is attributable to them. If none of them has committed a fault, they make contribution in proportion to the causal role of the action giving rise to liability which is attributable to them, or, by way of default rule, in equal parts.3

Article 1265 of the Projet therefore abandons the distinction between ­obligation solidaire and obligatio in solidum, by applying across the board the general regime of the former to cases involving multiple tortfeasors. If the rule is eventually adopted, it will cover one of the most troublesome gaps in the French law of civil liability.

A.  The General Rule: Solidarity Cannot be Presumed The current French Code civil does not contain a specific legal rule for the case in which the damage should be attributed to several liable tortfeasors. The former 2 For an overview of the previous proposals for the modernisation of the French civil liability rules, see Y Buffelan-Lanore and V Larribau-Terneyre, Droit Civil, Les Obligations, 16th edn (Paris, Sirey, 2018) 737–44. 3 ‘Lorsque plusieurs personnes sont responsables d’un même dommage, elles sont solidairement tenues à réparation envers la victime. Si toutes ou certaines d’entre elles ont commis une faute, elles contribuent entre elles à proportion de la gravité et du rôle causal du fait générateur qui leur est imputable. Si aucune d’elles n’a commis de faute, elles contribuent à proportion du rôle causal du fait générateur qui leur est imputable, ou à défaut par parts égales.’

‘Solidary’ Liability and the Channelling of Liability  265 article 1382 Cc set out, like the current article 1240 Cc, that causing harm leads to a claim for damages (‘Any human action whatsoever which causes harm to another creates an obligation in the person by whose fault it occurred to make reparation for it’). This rule does not address the problems attached to multiple liable tortfeasors. General rules governing the law of obligations contained in Title IV of the Code civil as reformed in 2016 did not help to resolve the problem, not least as their natural subject-matter was (and is) contractual obligations rather than the obligations de réparer arising from extra-contractual liability. On the one hand, article 1309 Cc provided a general pro-tanto rule where many debtors or creditors were joined by the same obligation: An obligation binding multiple creditors or debtors is divided by operation of law as between them. A division takes place also as between their successors, even if the obligation is joint and several. In default of specific regulation by legislation or by the contract, the division takes place in equal parts.4

The division, though, was not practicable in cases in which the participation of each and every potentially liable tortfeasor could not be individually i­ dentified. On the other hand, according to article 1310 Cc, joint and several liability (­solidarity) can never be presumed and can be imposed only either by legislation or contract.5 There is no room for solidarity in the absence of legislative provision or agreement. Therefore, in the absence of a specific legislative provision, where two or more tortfeasors have caused the same damage, each of them can only be liable in respect of his identified respective share of responsibility. The burden of doing so is on the victim and it can lead to the claimant losing his claim, especially because the usual causation complexities in cases involving several tortfeasors make the identification of individual shares of liability among the potentially liable tortfeasors either impossible or very costly. To avoid such troubles, the current obligatio in solidum doctrine makes all tortfeasors liable in full.

B.  Two Specific Cases of Solidarity Among Tortfeasors Although there is no general legislative provision in current French tort law on solidarity among tortfeasors, such a legal regime applies in some specific cases. Article 1242 Cc makes parents jointly and severally liable for the accidents caused by their minor children (‘To the extent to which they exercise parental authority, a father and mother are jointly and severally liable for harm caused

4 ‘L’obligation qui lie plusieurs créanciers ou débiteurs se divise de plein droit entre eux. La division a lieu également entre leurs successeurs, l’obligation fût-elle solidaire. Si elle n’est pas réglée autrement par la loi ou par le contrat, la division a lieu par parts égales.’ 5 ‘La solidarité est légale ou conventionnelle; elle ne se présume pas.’

266  Carlos Gómez Ligüerre by their minor children who live with them’).6 According to article 1245-7 Cc, the liability of manufacturers of different parts of the same defective product is also deemed to be joint and several (‘In the case of harm caused by a defect in a product incorporated into another product, the producer of the component part and the person who effected its incorporation are jointly and severally liable’).7 In both cases, solidarity among persons who are liable is expressly provided by the legislation. It confirms the principle according to which the joint and several liability regime can be applied only as a result of a legislative provision or by an agreement between the creditor and several debtors. Both cases reinforce the rule according to which non-contractual solidarity can be imposed only by legislation.

C.  The Discovery of the Obligatio in Solidum French case law found a solution outside the rigidities of the black letter law. The French Cour de cassation coined the category of obligatio in solidum to refer to the legal nature of the obligation linking several tortfeasors to the same victim when all of them could be deemed liable for the same damage.8 According to such a way of becoming liable, liability can be imposed on any of the defendants for the whole harm.9 This solution constitutes a tertium genus between the general joint and several liability rule governed by articles 1310–1319 Cc and the individual liability rule set out by article 1240 Cc.10 Unlike the general legislative joint and several liability regime, only some effects of the solidarity principle are applicable to the persons liable in solidum. Specifically, only those usually referred to as the ‘primary’ effects of solidarity, ie mainly, the unity of the obligation shared by the persons liable 6 ‘Le père et la mère, en tant qu’ils exercent l’autorité parentale, sont solidairement responsables du dommage causé par leurs enfants mineurs habitant avec eux.’ 7 ‘En cas de dommage causé par le défaut d’un produit incorporé dans un autre, le producteur de la partie composante et celui qui a réalisé l’incorporation sont solidairement responsables.’ 8 As it was defined by the Court: ‘each of the co-authors of the same harm, resulting from their respective faults, is liable in solidum to make reparation in respect of the whole harm, each of these faults having contributed to its occurrence’ (‘chacun des coauteurs d’un même dommage, conséquence de leurs fautes respectives, doit être condamné in solidum à la réparation de l’entier dommage, chacun des fautes ayant concouru à le causer tout entier’) (Cass civ 4 December 1939 DC 1941.125). The evolution of French case law in the years after the adoption of the in solidum solution is well explained by J Vincent, ‘L’extension en jurisprudence de la notion de solidarité passive’ RTD civ 1939, 601. 9 See L Mazeaud, ‘Obligation in solidum et solidarité entre codébiteurs délictuels’ RCLJ 1930, 149. 10 P Delebecque, F-J Pansier, Droit des obligations, 7th edn (Paris, LexisNexis, 2016) 264:‘The ­obligatio in solidum arose in the area of civil liability, including tort, to circumvent the principle that solidarity is not presumed and must be expressly stipulated by a contract, or provided by legislation. This rule seemed to exclude solidarity from the field of tort. In order to remedy the negative effects of this lack of protection of the victims, the judges in the nineteenth century recognised the solidarity of the co-authors of harm by – contra legem – resorting to article 1202 of the Code civil. In a second phase, abandoning their reliance on article 1202 of the Code civil, the Court of cassation established a ­principle of “full compensation”.’

‘Solidary’ Liability and the Channelling of Liability  267 (article 1313 Cc) and the possibility for any of the debtors to set up against the creditor the objections that any other solidary debtor could himself set up if he compelled to the payment of the debt (régime des exceptions, article 1315 Cc). The so-called ‘secondary’ effects of solidarity11 do not apply in the particular situation of obligatio in solidum. The secondary effects are mostly those related to the interruption of the legislative prescription period (article 2245 Cc) and the ­application of the late payments regime (article 1314 Cc). Unlike the individual liability rule, the in solidum doctrine allows the victim to claim against the potentially liable tortfeasors without having to prove their respective shares of liability.12 The protection of the victim was – and still is – the justification of this case law-made rule. The obligatio in solidum aims to make easier the claim for damages to victims of accidents caused by several tortfeasors. Despite criticism13 and the problems explaining it to foreign lawyers,14 the obligatio in solidum doctrine has endured in French tort law and has been the way of allowing victims to recover from accidents involving multiple tortfeasors.15

11 The secondary effects of the joint and several liability regime are considered by French legal scholars as a consequence of the agency relationship that exists between joint and several debtors united as such by agreement or as a consequence of a legislative provision. See B Fages, Droit des obligations, 7th edn (Paris, LGDJ, 2017) 416–17: ‘These effects are so named because they do not flow clearly from the mechanism of solidarity. They are classically explained by the existence of a mutual agreement among solidary debtors, who would have given each other a mandate to act on behalf of the others in their relationship with the creditor, an idea that today is no longer very convincing.’ 12 B Starck, H Roland, L Boyer, Obligations. Responsabilité délictuelle, 4th edn (Paris, Librairie de la Cour de Cassation, 1991) 533: ‘[t]he position of the courts is now perfectly clear and unanimous: each of the co-authors of the same harm is liable to compensate the whole harm, without it being necessary to take into account the shares of responsibility which the juges du fond determined among the liable co-authors, an apportionment which affects only the reciprocal relationship of the co-authors among themselves, but in no way the extent of their obligations towards the claimant. However, the juges du fond have an obligation to determine the share of each co-author in the total amount of compensation.’ 13 See G Viney, P Jourdain and S Carval, Traité de droit civil, Les conditions de la responsabilité, 4th edn (Paris, LGDJ, 2013) 297ff; F Chabas, ‘Remarques sur l’obligation in solidum’ RTD civ 1967, 310; B Starck, ‘La pluralité des causes de dommage et la responsabilité civile’ JCP 1970, no 2339. 14 See M Ferid and HJ Sonnenberger, Das Französische Zivilrecht, Band 2, 2. Auflage, Verlagsgesellschaft Recht und Wirtschaft, 1986, 462: ‘The liability of several tortfeasors is not regulated as for example under German law according to paragraph 840 BGB simply as joint and several liability. Joint and several liability (under French understanding) only applies in criminal cases when there is an offence or crime. Otherwise there is an obligatio in solidum. The victim can have a claim against each tortfeasor for the whole damage. Only in exceptional cases the apportioning of liability comes into consideration when the damage is divisible. A special case is liability of several tortfeasors in cases of unclear (alternative) causation, which should however be solved not on the liability level, but as a causation problem.’ 15 The solution has been very influential in jurisdictions close to the French tradition. This is true of Spain, whose Supreme Court follows the same pattern as the French case law when deciding cases against several tortfeasors. Like the Code civil, the Spanish Civil Code does not contain a specific rule on liability in cases of multiple tortfeasors. The Spanish Civil Code also regulates joint and several liability as a legislative or negotiated regime, barring its application in non-legislative and non-­contractual cases. Spanish case law found in the French approach an appropriate solution for granting victims a right to claim against several liable tortfeasors. Spanish case law uses the term ‘obligatio in solidum’ as well as other terms specifically used in such scenarios, like ‘solidaridad impropia’ or ‘solidaridad imperfecta’. These expressions seek to distinguish the specific tort law solution from the general joint and several liability regime. See, among others, JR León Alonso, La categoría de la obligación in solidum

268  Carlos Gómez Ligüerre

D.  The Proposal of a True Solidarity Regime If the proposed article 1265 of the Projet enters into force, the obligatio in s­ olidum doctrine will come to an end. The Code civil will then have a specific rule on the matter and solidarity will apply in this field by operation of law in full accordance with the origins of joint and several liability as they are identified by article 1310 Cc. The main consequence of the adoption of the new rule would be that multiple tortfeasors would be jointly and severally liable for the harm caused to their victim according to a legislative provision of general application. No longer being only a judge-made solution, the joint and several liability doctrine – ie a true solidarity regime – will become one of the distinctive features of French tort law, fully recognisable and foreseeable by potential defendants in its application, within and beyond the French legal system. As a result, the joint and several liability regime will be fully applicable, with all its consequences, as in the context of parental liability for the damage caused by their minor children pursuant to article 1242 Cc.16 Hence, there will no longer be any reason to distinguish between primary and secondary effects in the solidarity among multiple tortfeasors.17 All persons liable will be linked by a fully joint and several obligation. This will imply not only the unity of the common duty and the common effectiveness of the régime des exceptions, but also the existence of a common regime of interruption of the prescription periods and of the accrual of interest owed for late payment. Article 1265 of the Projet de réforme is the only legislative provision of sub-section 3 (L’incidence de la pluralité de responsables) belonging to Section 1 (Principes) of Chapter IV (Les effets de la responsabilité). The Projet places the solidary liability rule after those devoted to reparation in kind (La réparation en nature) in articles 1260 and 1261, and to the assessment of damages (Les dommages

(Publicaciones de la Universidad de Sevilla, 1978); I Díez de Lezcano, La no presunción de la solidaridad en las obligaciones (Marcial Pons, 1997); C Gómez Ligüerre, Solidaridad y derecho de daños. Los límites de la responsabilidad colectiva (Civitas, 2007); MA Esteve Pardo, La solidaridad impropia de deudores (Marcial Pons, 2014). 16 See Viney, Jourdain and Carval (n 13) 1215: ‘As to the scope of the solidarity imposed by this ­provision, it is determined by article 1200ff of the Civil Code, for it is not a simple obligatio in solidum, but a full solidarity.’ 17 See F Terré, P Simler, Y Lequette and F Chénedé, Droit civil, Les obligations, 12th edn (Paris, Dalloz, 2019) 1468: ‘(Secondary effects). In this way we name certain consequences attached by the law to solidarity, which do not derive naturally from its definition, but which are deduced, with varying degrees of obviousness, from the very concept of solidarity. By their nature, they benefit sometimes the creditor, sometimes the co-debtors. Various texts of the Code civil establish that a notice to perform [mise en demeure] addressed to one of the joint and several co-debtors is effective against the others; that such a notice will, in particular, start the running of interest for late payment and put the risk of the thing on all co-debtors; that the interruption of the prescription period against one debtor, ­resulting in particular from proceedings started against him or from his recognition of the debt, constitutes ­interruption of the prescription period against all; that a decisive oath [serment décisoire] required for the benefit of one person liable benefits the others, but if it has been required for all, it benefits only those who required it.’

‘Solidary’ Liability and the Channelling of Liability  269 et intérêts) in articles 1262 to 1264. After the rule on multiple tortfeasors, the Projet deals in Section 1 also with injunctions (La cessation de l’illicite) in article 1266, and with civil penalties (L’amende civile) that can complement damages in some special instances of intentional misconduct, according to the proposed article 1266-1. The structural position of article 1265 shows that the drafters of the Projet intended to raise the solution to the multiple tortfeasors’ cases to the status of a principle in the new scheme envisioned by the Projet. The rule applies to all cases of civil liability, both contractual and extra-contractual. The rule does not require that all persons liable are bound by the same fait générateur, but by the même dommage. As a result, it can apply to persons whose liability is imposed on different legal grounds. The new rule would be an instance of solidarité légale and it will exclude the application of article 1309 Cc (‘An obligation binding multiple creditors or debtors is divided by operation of law as between them’)18 for cases involving several wrongdoers, intentional as well as negligent, while saving the victim the trouble of apportioning liability among them.

II.  A Common Trend in Comparative Civil Law All legal systems have to deal with the problem of multiple tortfeasors. Most ­European tort law systems have adopted a joint and several liability solution for such cases. In most European jurisdictions victims of accidents caused by several tortfeasors are entitled to claim damages against them as solidary debtors.

A.  A Codified Rule The adoption of a general joint and several liability rule would include French tort law in the long list of European legal systems that provide a similar rule for multiple tortfeasors’ cases. This is the case in Germany, probably the best example among the codified European solutions. Paragraph 830 of the German Civil Code (BGB) makes liable all persons contributing to some extent to the damage caused to the victim: (1) If more than one person has caused damage by a jointly committed tort, then each of them is responsible for the damage. The same applies if it cannot be established which of several persons involved caused the damage by his act. (2) Instigators and accessories are equivalent to joint tortfeasors.19 18 ‘L’obligation qui lie plusieurs créanciers ou débiteurs se divise de plein droit entre eux.’ 19 ‘The translations of the BGB are taken from the official English translation, available at the site of the German Federal Ministry of Justice, at www.gesetze-im-internet.de/englisch_bgb/. (1) Haben mehrere durch eine gemeinschaftlich begangene unerlaubte Handlung einen Schaden verursacht, so ist jeder für den Schaden verantwortlich. Das Gleiche gilt, wenn sich nicht ermitteln lässt, wer von mehreren Beteiligten den Schaden durch seine Handlung verursacht hat. (2) Anstifter und Gehilfen stehen Mittätern gleich.’ See also BS Markesinis, The German Law of Torts. A Comparative Introduction,

270  Carlos Gómez Ligüerre All tortfeasors will be jointly and severally liable under paragraph 840 BGB: (1) If more than one person is responsible for damage arising from a tort, then they are jointly and severally liable. (2) If besides the person who is obliged to make compensation for damage caused by another person under sections 831 and 832 the other person is also responsible for the damage, then in their internal relationship the other is obliged alone, and in the case specified in section 829 the person with a duty of supervision is obliged alone. (3) If besides the person who is obliged to make compensation for damage under sections 833 to 838 a third party is responsible, then the third party is solely obliged in their internal relationship.20

Article 2055 of the Italian Civil Code also applies a solidary regime to the several liable tortfeasors: If the act or omission causing damage can be attributed to more than one person, all are jointly and severally liable to make reparation of the harm. The person who has compensated the damage has recourse against each of the others in proportion to the degree of fault of each and to the consequences arising therefrom. In case of doubt, the degree of fault attributable to each is presumed to be equal.21

It is the same in Portugal, whose Civil Code provides in article 497 that: 1. 2.

Where several persons are liable to the same harm, they are jointly and severally liable. Contribution among joint and several tortfeasors is made to each other in proportion of their respective faults as well as the consequences of them, although all faults will be presumed to be equal.22

3rd edn (Oxford, Clarendon Press, 1994) 905: ‘§ 830 BGB deals with the important practical problem raised by torts committed by several persons, their liability to the victim and their rights and duties against each other … This is the case where the parties commit a breach of a joint duty or act in pursuance of a common design. If this element of common design is satisfied it is irrelevant how significant is the contribution of each party in causal terms. Even mental participation in the form of helping out with the planning of the act or merely providing friendly support will suffice. § 830.II BGB treats instigators and accomplices in the same way.’ 20 ‘(1) Sind für den aus einer unerlaubten Handlung entstehenden Schaden mehrere nebeneinander verantwortlich, so haften sie als Gesamtschuldner. (2) Ist neben demjenigen, welcher nach den §§ 831, 832 zum Ersatz des von einem anderen verursachten Schadens verpflichtet ist, auch der andere für den Schaden verantwortlich, so ist in ihrem Verhältnis zueinander der andere allein, im Falle des § 829 der Aufsichtspflichtige allein verpflichtet. (3) Ist neben demjenigen, welcher nach den §§ 833 bis 838 zum Ersatz des Schadens verpflichtet ist, ein Dritter für den Schaden verantwortlich, so ist in ihrem Verhältnis zueinander der Dritte allein verpflichtet.’ 21 ‘Se il fatto dannoso è imputabile a più persone, tutte sono obbligate in solido al risarcimento del danno. Colui che ha risarcito il danno ha regresso contro ciascuno degli altri, nella misura determinata dalla gravità della rispettiva colpa e dall‘entità delle conseguenze che ne sono derivate. Nel dubbio, le singole colpe si presumono uguali.’ 22 ‘1. Se forem várias as pessoas responsáveis pelos danos, é solidária a sua responsabilidade. 2. O direito de regresso entre os responsáveis existe na medida das respetivas culpas e das consequências que delas advieram, presumindo-se iguais as culpas das pessoas responsáveis.’

‘Solidary’ Liability and the Channelling of Liability  271 More recently, article 6:102 of the Dutch Civil Code has also opted for a joint and several liability solution:23 1.

2.

When two or more persons are individually liable for the same damage, then they are jointly and severally liable for it. In order to assess what each of them has to contribute by virtue of Article 6:10 on account of their internal relationship with each other, the damage is imputed to them in accordance with Article 6:101, unless a different imputation results from law or a juridical act. When the damage is caused as well by circumstances which are attributable to the injured person himself, then Article 6:101 is applicable to the obligation of each of the liable persons meant in the previous paragraph to compensate the damage to the injured person, on the understanding that the injured person may, overall, not claim more of each of the liable persons than he could if only one of them would have been liable as a result of the circumstances on which their liability is based. When it is not possible to recover a contribution in full from one of the persons with an internal obligation to contribute in the damages, then the court may order, upon the request of one of these persons, that in the application of Article 6:13 the unrecovered contribution shall be imputed also over the injured person.

The Austrian Civil Code also adopts the joint and several liability rule in all cases in which the accident has been caused by the conduct of several tortfeasors. Article 1301 of the Austrian Civil Code sets forth that: Several persons may be held responsible for unlawfully caused harm whereas they contribute to the harm while acting in concert, in a direct or indirect manner, by inducement, threatening, commanding, helping, hiding and the like; or, even by o ­ mission of a specific obligation to prevent the harm.24

Belonging to a different tradition, in common law systems joint and several liability has been the usual outcome in cases of agency, vicarious liability and concerted action,25 as well as in cases of single, indivisible injury caused by several tortfeasors.26 23 Unofficial translation available at the ‘Dutch Civil Law’ website, www.dutchcivillaw.com/civilcodebook066.htm. 24 ‘Für einen widerrechtlich zugefügten Schaden können mehrere Personen verantwortlich werden, indem sie gemeinschaftlich, unmittelbarer oder mittelbarer Weise, durch Verleiten, Drohen, Befehlen, Helfen, Verhehlen u. dgl.; oder, auch nur durch Unterlassung der besonderen Verbindlichkeit, das Übel zu verhindern, dazu beigetragen haben’. See Helmut Koziol, Österreichisches Haftpflichtrecht. ­Allgemeiner Teil, 3. Auflage (Wien, Manz Verlag, 1997) 460: ‘According to paragraph 1301 the joint and several liability of several tortfeasors requires that all of them have contributed to the damage. However, the prevailing understanding assumes that intentional tortfeasors are always jointly and severally liable. The causal role of each individual tortfeasor is not examined separately, since it is presumed that each tortfeasors is a conditio sine qua non for the damage.’ 25 S Deakin, A Johnston and B Markesinis, Markesinis and Deakin’s Tort Law, 7th edn (Oxford, Oxford University Press, 2013) 850: ‘Tortfeasors are “joint” in cases of express authorisation or instigation: principal and agent; vicarious liability; the liability of an employer and an independent contractor; the liability of tortfeasors who act in breach of a joint duty; and tortfeasors who act in pursuance of a common design … In joint liability each tortfeasor is liable for the full amount of the claimant’s loss but there is only one tort and so the cause of action against each one is the same and is supported by the same evidence.’ See also, T Weir, An Introduction to Tort Law, 2nd edn (Oxford, Clarendon Press, 2006) 103–22. 26 E Peel, J Goudkamp, Winfield and Jolowicz on Tort, 19th edn (London, Sweet & Maxwell, 2018) 679: ‘Where two or more people by their independent breaches of duty to the claimant cause him to

272  Carlos Gómez Ligüerre

B.  The Rule as Proposed by the PETL and the DCFR The adoption of a joint and several liability regime for cases involving several tortfeasors has been also recommended by the soft-law proposals that aim to harmonise or at least set a common principle for tort law in Europe. Joint and several liability is the solution proposed by article 9:101 of the Principles of European Tort Law: (1) Liability is solidary where the whole or a distinct part of the damage suffered by the victim is attributable to two or more persons. Liability is solidary where: a) b) c)

a person knowingly participates in or instigates or encourages wrongdoing by others which causes damage to the victim; or one person’s independent behaviour or activity causes damage to the victim and the same damage is also attributable to another person; or a person is responsible for damage caused by an auxiliary in circumstances where the auxiliary is also liable.

(2) Where persons are subject to solidary liability, the victim may claim full compensation from any one or more of them, provided that the victim may not recover more than the full amount of the damage suffered by him. (3) Damage is the same damage for the purposes of paragraph (1)(b) above when there is no reasonable basis for attributing only part of it to each of a number of persons liable to the victim. For this purpose it is for the person asserting that the damage is not the same to show that it is not. Where there is such a basis, liability is several, that is to say, each person is liable to the victim only for the part of the damage attributable to him.

Solidarity is also the legal regime recommended by article VI-6:105 DCFR, according to which: Where several persons are liable for the same legally relevant damage, they are liable solidarily.27

III.  Conditions and Effects of the Proposed Solidarity Rule Like the legislation and soft-law proposals outlined above, article 1265 of the Projet de réforme clearly advocates the adoption of a joint and several suffer distinct injuries, no special rules are required, of each tortfeasors is liable for the damage which he caused and only for that damage. Where, however, two or more breaches of duty by different persons cause the claimant to suffer a single, indivisible injury the position is more complicated. The law in such a case is that the claimant is entitled to sue all or any of them for the full amount of his loss, and each is said to be jointly and severally liable for it.’ 27 The DCFR defines ‘solidarity’ in III-4:102, which states: ‘[a]n obligation is solidary when each debtor is bound to perform the obligation in full and the creditor may require performance of any of them until full performance has been received’.

‘Solidary’ Liability and the Channelling of Liability  273 liability regime.28 The dynamics of the joint and several liability regime always entails distinguishing the relationship between several solidary debtors and the creditor from the relationship between the solidary debtors themselves. Thus, two different aspects are always present in a solidary obligation. The so-called ‘external relationship’ deals with the legal position of the creditor (the victim in our case) in relation to several persons liable. It concerns the victim’s claim to be compensated, and the procedural requirements of the claim. The main peculiarity of the joint and several liability regime is that it entitles the victim to claim full recovery of the damage from any solidary tortfeasor, independently of his real share of responsibility. Every solidary tortfeasor becomes full debtor in respect of the whole damage. The so-called ‘internal relationship’ looks at the pattern of relationships that exist between several debtors (the multiple tortfeasors in our case). It deals with the respective amounts owed as between them once the victim has been compensated. A solidary debtor that has paid the damages in full (or a larger amount of damages than those attributable to his involvement in the case) is entitled to ask for the respective contributions from other tortfeasors. Such claims for contribution (recours en contribution) should reflect the share of liability of each of them.

A.  Liability for the ‘Same Harm’ In the rule set forth in the first paragraph of article 1265 of the Projet, the “same harm” (un même dommage) requirement justifies the imposition of solidarity among several

28 The Projet is more explicit and clearer than other proposals made in previous projects of modernisation of French tort law. Art 11 of the Avant-projet Terré proposed the adoption of a rule with the following wording: ‘Unless otherwise provided, where several persons have caused the same damage, each of them is liable for the whole. If they all have committed a fault, they contribute among themselves in proportion to the gravity of their respective faults. If none of them has committed a fault, they contribute equally. If only some of them have committed a fault, they alone bear the final charge of the damage.’ (‘Sauf disposition contraire, ceux qui ont causé un même dommage en répondent chacun pour le tout. S’ils ont tous commit une faute, ils contribuent entre eux à proportion de la gravité de leurs fautes respectives. Si aucun d’eux n’a commis une faute, ils contribuent par parts égales. Si certains seulement d’entre eux ont commit une faute, ils supportent seuls la charge définitive du dommage.’). For a critique of this rule, see J-S Borghetti, ‘De la causalité’ in F Terré (ed), Pour une réforme du droit de la responsabilité civile (Paris, Dalloz, 2011) 146: ‘Article 11 of the [Avant-]projet enshrines the rules of positive law in case of multiple authors of damage. The fundamental principle that each co-author responds for the whole damage caused to the victim is affirmed. The text does not explicitly state whether their liability is joint and several or in solidum; however, it must be interpreted as creating a new case of joint and several liability, thus making recourse to the concept of liability in solidum useless. The article also states that when the co-authors of damage have all committed a fault, they contribute each in proportion to the severity of their respective faults. The solution usually applied by the courts is therefore confirmed. At the same time, the other solution sometimes adopted in this respect, which consists in making the contribution of each co-author dependent upon the causal role of his fault, is discarded. The remainder of the article is also in conformity with the positive law, which did not appear to call for any reform.’

274  Carlos Gómez Ligüerre tortfeasors: Where one or more persons are liable for the same harm, they are jointly and severally liable to make reparation for it to the victim.29

If this rule were to be adopted, joint tortfeasors would be deemed to be solidary debtors and subjected to the legal regime provided by the current article 1313 Cc, according to which: The joint and several nature of an obligation amongst debtors imposes on each of them an obligation for the whole of the debt. Satisfaction by one of them discharges them all as regards the creditor. The creditor may require satisfaction from any joint and several debtor he may choose. An action brought against one of the joint and several debtors does not prevent the creditor from bringing similar actions against the others.30

The proposed legal regime depends on the causation of the harm attributable to all tortfeasors. According to article 1265 of the Projet, only those liable for ­causing the same harm to the victim can be considered jointly and severally liable. The Projet’s approach follows the solution foreseen by the European national ­legislative provisions mentioned in the previous section. Despite being essential for the application of the rule provided by article 1265, the concept of the ‘same harm’ is not defined by the Projet. In this context, it should be understood that the ‘same harm’ does not differ from the harm caused by a single tortfeasor except for the fact that it has been caused by several actors. Under the Projet jointly and severally liable tortfeasors are those causing a single and indivisible harm to the victim. For that reason the victim is entitled to ask for the full compensation from each tortfeasor.31 Since the harm is the ‘same’, it cannot be apportioned between the tortfeasors and all of them have to be liable in full for all losses consequential on the harm. In this respect, article 1265 of the Projet refers to the même dommage and not to the préjudices (losses) arising from it. The proposed rule does not take into account the distinction between harm (dommage) and loss (préjudice) outlined by article 1235 of Projet (‘Any certain loss is reparable where it results from harm and consists of an injury to a lawful interest, whether patrimonial or extra-patrimonial’).32 Hence, each and every tortfeasor will be liable in full for all losses resulting from the ‘same harm’ attributable to all of them. 29 ‘Lorsque plusieurs personnes sont responsables d’un même dommage, elles sont solidairement tenues à réparation envers la victime.’ 30 ‘La solidarité entre les débiteurs oblige chacun d’eux à toute la dette. Le paiement fait par l’un d’eux les libère tous envers le créancier. Le créancier peut demander le paiement au débiteur solidaire de son choix. Les poursuites exercées contre l’un des débiteurs solidaires n’empêchent pas le créancier d’en exercer de pareilles contre les autres.’ 31 The definition of ‘same damage’ proposed by art 9:101 PETL may help to understand the rule in the Projet: ‘(3) Damage is the same damage … when there is no reasonable basis for attributing only part of it to each of a number of persons liable to the victim. For this purpose it is for the person asserting that the damage is not the same to show that it is not. Where there is such a basis, liability is several, that is to say, each person is liable to the victim only for the part of the damage attributable to him.’ 32 ‘Est réparable tout préjudice certain résultant d’un dommage et consistant en la lésion d’un intérêt licite, patrimonial ou extrapatrimonial.’

‘Solidary’ Liability and the Channelling of Liability  275

B.  The Contribution Claim between Tortfeasors The contribution between solidary debtors is the second step in the life cycle of the solidary obligation. Contribution seeks to apportion liability among the ­tortfeasors. The alleged unfairness of making each and every tortfeasor liable in full in the external relationship is balanced with the contribution between them reflecting the internal aspect of the solidary relationship. Liability in full in the external relationship turns into proportionate liability in the internal relationship between tortfeasors. The ease of functioning of the ‘external relationship’ solution contrasts with the complexity of apportioning liability in the ‘internal relationship’. Where solidarity has a contractual origin, the solidary debtors have had the opportunity to agree on the extent of their respective shares in responsibility (and therefore liability) in the internal relationship or on the scope of a contractual right to an indemnity. In this case, the rule provided by article 1317 Cc, according to which ‘[a]s between themselves, the contribution of each joint and several debtor is limited to his own share’,33 will follow what was agreed between the solidary debtors. If an equivalent agreement existed between jointly and severally liable tortfeasors, their individual liability would be decided accordingly, at least to the extent that the harm to the victim was caused unintentionally. However, where the tortfeasors did not agree upon the internal distribution of their respective liability, a legal rule has to fill this gap.34 In assessing the importance of the actions or omissions of each tortfeasor, article 1265(2) of the Projet de réforme determines the importance of the acts or omissions of each tortfeasor in a way that resembles the task that the victim would carry out absent the joint and several liability regime. The apportionment rule in article 1265 follows the rules established by French case law when configuring the contribution claim between the tortfeasors liable in solidum.35 The rule seeks a fair result and resembles the relevance of the fault principle typical of the civil law tradition. If the tortfeasors are at fault, they will contribute to each other according to the seriousness (­gravité) of their respective faults, combined with the causal impact of the breach of their respective duties. If they are not at fault, their contribution will depend on

33 ‘Entre eux, les codébiteurs solidaires ne contribuent à la dette que chacun pour sa part.’ 34 See A Hontebeyrie, Le fondement de l’obligation solidaire en droit privé français (Paris, Economica, 2004) 94–100. 35 See Fages (n 11) 383–84: ‘According to which criterion must the cost of compensation be distributed between multiple defendants? The answer varies depending on the circumstances. In the case of defendants who have all committed a fault, we assess the contribution of each one by considering the gravity of his respective fault, so that the apportionment is not necessarily egalitarian and may even, when the fault of one defendant is fully absorbed by the fault of the other, lead the latter to have to bear entirely the cost of compensation. As an extension of this solution, which reflects a certain preeminence of moral considerations, the case-law considers that a person who has been found liable without fault on the basis of a strict liability rule may have full recourse against the person to whom a fault has been imputed; the latter, conversely, is left without any possibility of recourse. When all the defendants have been found strictly liable, compensation must paid in equal shares.’

276  Carlos Gómez Ligüerre the respective causal roles of their actions, as long as these can be ascertained; ­otherwise, liability will be apportioned in equal shares among the tortfeasors.36 The rule distinguishes between the tortfeasors at fault and those not at fault. At first sight, the rule resembles the distinction between negligence (liability for fault) and strict liability (liability without fault). However, article 1265(2) of the Projet does not refer to the liability of those tortfeasors who caused the damage without fault as a liability de plein droit, an expression used by the Projet to refer to strict liability regimes. This is the case, for instance, of article 1285 (liability for traffic accidents) and of article 1289 (liability for defective products) of the Projet. Hence, it seems that the rule on contribution looks at the actual fault of each tortfeasor rather than to the standard of liability (ie negligence or strict liability) that makes any of them liable for the harm caused to the victim.37 The application of article 1265 of the Projet to the contribution claims among tortfeasors will surely be problematic. It restricts the discretion that the French courts currently enjoy when deciding on the effects of the obligatio in solidum and exposes them to some practical difficulties when apportioning liability. First, distinguishing between fault and causation is no easy task, especially in cases involving several tortfeasors in which the respective faults may have different causal effects. This is the case of concurrent causation scenarios, in which the combined effect of the respective actions of the tortfeasors may render the task of distinguishing between individual faults and individual causation roles just impossible. An example would be the environmental harm caused by a combination of hazardous wastes: in such a case, the seriousness (gravité) of each and every tortfeasor’s fault, as well as its causal role, depends upon the harm already caused by the toxic wastes spilt by the other tortfeasors. A second difficulty arises in situations where the harm has been caused by some tortfeasors who were at fault together with other tortfeasors who were not at fault. The impact of the fault and causation of the former have to be distinguished from the exclusively causal role played by the latter, who will contribute in equal shares among themselves if their relative causal roles cannot be identified. In such cases, the arithmetic of the internal relationship would reach a high degree of complexity, since the share of internal liability attributable to each group of tortfeasors would differ. A third problem relates to the relevance of the subjective fault, which takes into account moral considerations that have already been abandoned. Article 1241 Cc 36 These rules will prevail over the discretion of the court, which is subject to the guidance provided by art 1265(2) of the Projet. The solution differs from other options which are much more deferential towards judicial discretion, as in the case the English Civil Liability (Contribution) Act 1978, s 2(1) according to which ‘the amount of the contribution recoverable from any person shall be such as may be found by the court to be just and equitable having regard to the extent of that person’s responsibility for the damage in question’. See Winfield and Jolowicz on Tort (n 26) 679–84; and MA Jones (ed), Clerk & Lindsell on Torts, 22nd edn, (London, Sweet & Maxwell, 2018) 292–97. 37 Then, even if all or some of the joint tortfeasors are strictly liable but committed the wrong with fault, their fault (ie its seriousness and its causal relevance) will be taken into account when assessing the contribution owed by them.

‘Solidary’ Liability and the Channelling of Liability  277 (former article 1383 Cc)38 has been understood as the basis for adopting an objective definition of fault which disregards moral or personal factors. The objective character of fault for the purposes of the imposition of liability is stressed in the Projet de réforme, whose article 1242 defines fault as the breach of a legislative duty or as acting below the general duty of care or diligence. So, while the seriousness of a person’s fault is irrelevant to liability generally, it becomes relevant to the amount of contribution which that person will pay. It would be probably easier to administer a rule providing the courts with a greater degree of discretion in apportioning liability, combined with a clear and easily applicable default rule, like the one proposed by article III- 4:106 DCFR: (1) As between themselves, solidary debtors are liable in equal shares. (2) If two or more debtors have solidary liability for the same damage, their share of liability as between themselves is equal unless different shares of liability are more appropriate having regard to all the circumstances of the case and in particular to fault or the extent to which a source of danger for which one of them was responsible contributed to the occurrence or extent of the damage.

Article 1265 of the Projet says nothing about the internal effects of the insolvency of any of the tortfeasors. Where this is the case, the rule provided by the existing article 1317(3) Cc would apply and so the share of any insolvent tortfeasor would be distributed among the other liable and solvent tortfeasors in proportion to their respective liability.39 Article 1317 Cc does not take into account fault in order to distinguish between the tortfeasors as article 1265 of the Projet instead would provide. The doubt is therefore if the share of an insolvent tortfeasor has to be borne only by the others belonging to the same class (ie those at fault or those without fault) or if it should be distributed among all solvent tortfeasors (independently of the way in which their liability is apportioned). It seems that the most appropriate solution is to deal with all solvent tortfeasors in the same way and so make all of them responsible for the share of the insolvent one. In the end, all of them are liable for the same harm and the rule in article 1317 Cc seeks to avoid problems arising from insolvency. Its purpose has nothing to do with the fairness goals underlying article 1265(2) of the Projet.

IV.  The Questionable Distinction between the Harm Caused by an Undetermined Member of a Group, and the Joint and Several Liability Case The Projet distinguishes its general rule on solidarity between tortfeasors from a specific rule governing cases of personal injury caused by an undetermined 38 ‘Everyone is liable for harm which he has caused not only by his action, but also by his failure to act or his lack of care.’ 39 ‘Si l’un d’eux est insolvable, sa part se répartit, par contribution, entre les codébiteurs solvables, y compris celui qui a fait le paiement et celui qui a bénéficié d’une remise de solidarité.’

278  Carlos Gómez Ligüerre (personne indéterminée) member of a group. This case is governed by the rule proposed in article 1240 of the Projet, according to which: Where personal injury is caused by an undetermined person among two or more identified persons acting in concert or exercising a similar activity, each person is liable for the whole, unless he shows that he could not have caused it. The persons so liable make contribution to each other in proportion to the probability which each had caused the harm.40

This provision assumes, first, that such a group can be identified and, secondly, that the group is not endowed with a legal form (such as a company), as otherwise the rules governing the liability of legal entities would apply. The rule in article 1240 of the Projet does not distinguish between the cases in which there was a common intention to cause the harm from those in which the harm was a negligent result of the activity conducted in common and so the legal solution which it provides applies to both. It should be noted, though, that the rule does not mention the fault of the members of the group as a measure to assess their individual liability or as a requirement of their liability. In the situation foreseen by article 1240 of the Projet the victim has suffered a harm caused by one (or more) of the members of a group, and it is not possible to know who among them was the wrongdoer. The legal solution is, then, to make each and every member of the group responsible in full (chacune en répond pour le tout). This legal solution is, though, based on a rebuttable presumption, since any member of the group is allowed to escape liability if he shows that he could not have caused the damage (sauf à démontrer qu’elle ne peut l’avoir causé). The burden of proof lies on the potentially liable member of the group, not on the victim, who is protected by the legal presumption. Article 1240 of the Projet also provides how the liability imposed on members of the group should be distributed between them, each being liable in full towards the victim. Once the victim has recovered, the group member who has paid damages in full can ask for contribution from every other group member in proportion to the probability with which each has caused the harm. Interestingly, the provision refers in this case to probability (‘la probabilité que chacun ait causé le dommage’), rather than the causal role of each member of the group on the assumption that the members’ individual causal roles cannot be established: this assessment solves the individual causation uncertainty. Although the article does not deal with this, if the individual probabilities cannot be ascertained, the only suitable solution would be apportioning the liability in equal shares among all liable members of the group. Of course, where the members’ individual causal roles can be established, then liability should be determined accordingly without making all group members liable in full under article 1240 at all. 40 ‘Lorsqu’un dommage corporel est causé par une personne indéterminée parmi des personnes identifiées agissant de concert ou exerçant une activité similaire, chacune en répond pour le tout, sauf à démontrer qu’elle ne peut l’avoir causé. Les responsables contribuent alors entre eux à proportion de la probabilité que chacun ait causé le dommage.’

‘Solidary’ Liability and the Channelling of Liability  279 This special rule certainly differs from the joint and several liability regime envisaged by article 1265 of the Projet, under which all tortfeasors are deemed jointly and severally liable with the result that each of them is liable in full in the external relationship, while internally their liabilities are apportioned according to the rules set out by article 1265(2) of the Projet. By contrast, since article 1240 of the Projet deals with situations where personal injury is caused by an undetermined member of a group and it is not clear whether the other members have causally contributed to it as well, the draft provision gives the members of the group the chance to disprove causation and so escape liability. From the perspective of defendants, there is a significant difference between the regime proposed by article 1240 of the Projet and the joint and several liability regime. Under the former, a member of the group who did not cause the harm is entitled to prove that he acted correctly, either towards the victim or in the internal relationship with the liable wrongdoers, and so avoid his own liability – provided, of course, that such proof is possible where all members of the group were engaged in a similar activity or acted in concert.41 Such an exoneration should not be (and is not) possible under the joint and several liability regime, as it is premised on the assumption that each tortfeasor has contributed to the harm for which they are all liable, even though their shares of liability may be different. From the victim’s perspective things look different. The joint and several ­liability solution protects the victim of a harm caused by several tortfeasors; the rule in article 1240 of the Projet protects the victim of a harm caused by an undetermined member of the group of persons whose common activity caused it. In both cases, though, the victim is entitled to ask for full compensation from any member of the group or from any of the jointly and severally liable tortfeasors. In neither situation is the victim expected to establish the respective shares of liability (in the joint and several liability case) or the individual causal link between the harm and the activity of each potentially liable tortfeasor (in the group liability case). Thus, from the victim’s standpoint, there are no relevant differences between the two legal regimes. Although the case of joint tortfeasors and the case of harm caused by an undetermined member of a group can be distinguished in theory, their differences may not be so clear in their application in specific cases and courts may indeed face problems in choosing the rule applicable. Article 1240 of the Projet would apply to cases in which the harm has been the result of a concerted action or an a­ ctivity conducted in common by several persons. Joint and several liability, as set out by 41 See P Brun, Responsabilité civile extracontractuelle, 5th edn (Paris, LexisNexis, 2018) 181–82: ‘[b]ut on the basis of a scientifically more rigorous reasoning, it has also sometimes been considered that in such circumstances the harm must be attributed to a combination of facts that cannot be separated from each other.’ In the same vein, P Malaurie, L Aynes and P Stoffel-Munck, Droit civil, Droit des obligations, 10th edn (Paris, LGDJ, 2018) 124: ‘[s]econdly, because if it was simply a matter of proving who caused the harm, the requirement of a concerted action would not be justified. In reality, only the “communitarian mentality” of the defendants justifies all of them being found liable despite the absence of evidence against each of them’.

280  Carlos Gómez Ligüerre article 1265 of the Projet, would apply to cases in which several tortfeasors are liable for the same harm. It may, though, be difficult to distinguish between those cases in which one of several persons acting in concert or conducting a common activity caused the harm and those in which the harm can be attributable to all of them. Since the victim is entitled to ask for full compensation from any person potentially liable in both situations, differences between the two rules would vanish from the victim’s perspective. In addition, article 1240 of Projet only applies to personal injuries. Such a restriction would force the courts to apply article 1265 of the Projet to cases of property damage and economic loss caused by an (undetermined) member of the group, in order to grant to the victim of these kinds of harm the protection given in cases of harm caused by several (or any among several) tortfeasors. For all these reasons, it seems better to avoid the distinction proposed by articles 1240 and 1265 of the Projet and to deal with the case of harm caused by an undetermined member of a group (acting in concert or conducting a similar activity) simply as a case of joint and several liability. That is, to include the cases governed by article 1240 under article 1265.42

V.  The Drawbacks of a General Solidarity Rule in the Law of Torts Article 1265 of the Projet de réforme seeks to state a general solidarity rule for multiple tortfeasors’ cases. The rule, if finally enacted, would solve the practical

42 Following, for instance, the example of § 830 BGB. After setting forth joint and several liability for the multiple tortfeasors’ case, the German rule provides that ‘[t]he same applies if it cannot be established which of several persons involved caused the damage by his act’. The possibility to distinguish the liability of a member of the group from the general joint and several liability case was also discussed in the DCFR and the drafters eventually decided not to include a rule equivalent to art 1240 of the Projet. Instead they proposed only the rule contained in art IV-4:102 DCFR: ‘A person who participates with, instigates or materially assists another in causing legally relevant damage is to be regarded as causing that damage.’ The rule must be read in conjunction with art VI-6:105 (Solidary liability), under which participants, instigators and accessories are solidarily liable with the (principal) wrongdoer. As is explained by the Study Group on a European Civil Code, Research Group on EC Private Law (Acquis Group), edited by C von Bar and E Clive, Principles, Definitions and Model Rules of European Private Law. Draft Common Frame of Reference, Full edition (Munich, Sellier, 2009) vol 4, 3594: ‘During the preparation of this Book, intense discussion took place on the issue of whether a further rule should be adopted in addition to this Article. Under this mooted rule, where a member of a group intentionally causes a third party a legally relevant damage, other members of the group would be liable for the damage in so far as the risk of intentional occurrence of damage of that type was foreseeable and those members should have abstained from participating in the group. It was also discussed whether this liability should rest only upon those who were present at the scene of the wrong and whether their liability should only come into play in a subsidiary fashion. This suggestion was, however, rejected as being too far-reaching. The main objection was that such a rule, which would forego the requirement of collaboration and would be based solely upon “membership” in a group, would not be capable of being into harmony with the freedom to demonstrate.’

‘Solidary’ Liability and the Channelling of Liability  281 problems caused by the current obligatio in solidum doctrine. The rule of joint and several liability for multiple tortfeasors would entail, though, some problems that have been identified in some jurisdictions with a more developed tradition in the field.43 Under the joint and several liability regime each tortfeasor bears the risk of insolvency of one or more of the co-defendants. While the risk of insolvency of the defendant is borne by the victim of an accident caused by a single t­ortfeasor, where several persons are jointly and severally liable, the insolvency of any one or more of them will be borne by the other co-defendants. This effect is the most valuable result of the application of joint and several liability for the victim of harm attributable to several tortfeasors, as well as, on a general note, for any creditor of several debtors. Solidarity therefore strengthens the claim for damages and makes the victim’s redress easier and more likely to be achieved. Indeed, in this light, the victim of several tortfeasors enjoys a privilege, both procedural and substantive. Joint and several liability allows the victim to receive full compensation even though one or several co-defendants become insolvent. Under the joint and several liability regime it is possible that some of the co-defendants have to pay a larger share of the damages than the share which is attributable to their own fault or their own causal role in the accident. Conversely, some of the co-defendants may, in fact, pay nothing or less than the amount which they should have paid in law given their fault or their causal role in relation to the damage. As a result, solidarity breaches the principle according to which a tortfeasor has to face the consequences of his own wrongful acts. The absence of any relationship between the real share of a person’s responsibility and the final amount of damages paid by each tortfeasor therefore causes problems in terms of the deterrent function of the law of civil liability but, ironically, joint and several liability becomes an attractive legal regime for potentially insolvent tortfeasors. Anyone who expects to be legally subject to a regime of joint and several liability knows that his liability is covered by other liable tortfeasors. Hence, solidarity discourages the adoption of precautions to avoid the harm that are higher than the liability which is likely to be imposed. Since the likely burden of liability for an insolvent tortfeasor is zero, his investment in avoiding harm will also amount to zero. Joint and several liability incentivises under-deterrence among insolvent actors. In addition, the distortion between the liability to be borne as a matter of law and the liability actually borne within the internal relationship between ­tortfeasors affects negatively the insurability of the persons potentially involved in cases attracting joint and several liability. The respective insurers may well be able to assess accurately the expected liability of their own insured in order to determine the level of the premium charged. But they are not able to anticipate the insolvency

43 See, for instance, American Law Institute, Restatement of the Law (Third) Torts: Apportionment of Liability, American Law Institute Publishers, 2000.

282  Carlos Gómez Ligüerre of other solidary tortfeasors even though this may increase the actual liability to be borne by their insured. For all these reasons, it would be preferable for the Projet de réforme to provide for the joint and several liability mechanism as a default rule to be applicable only in cases in which the individual liability of each tortfeasor cannot be identified, rather than setting a general rule that makes all of them jointly and severally liable in all cases.

part vi Defences

284

14 Defences to Tortious and Contractual Liability in French Law SANDY STEEL*

This chapter is in two sections. Section I examines general issues surrounding defences in French tort and contract law, comparing the position to that in English law. Section IA considers the meaning of the term ‘defence’, and identifies a number of concepts to which the term refers. The focus of the chapter is upon facts which defeat liability even if the elements of a rule are satisfied (‘the externalist definition’) and, to the extent they differ, facts which defeat or restrict liability, which are for the defendant to prove (‘the proof-based definition’). Section IB shows that both English law and French law contain defences in these senses. Section IC considers whether the concept of a defence is recognised as part of the law or as an organising idea in doctrinal understanding of the law in each system. Section II of the chapter examines some particular defences and compares their scope and rationales in each system.

I.  General Issues Surrounding Defences A.  What is a Defence? In the common law of tort and contract, the term ‘defence’ is used to mean a ­variety of different things, often without the precise meaning being clear. It can mean at least the following:1 (1) any fact (proof of) the existence of which entails that no liability arises;2 (2) a fact (proof of) the existence of which entails that no liability arises, although the elements of a cause of action/liability-generating rule are satisfied; * With thanks to Jean-Sébastien Borghetti, Simon Whittaker and participants at the conference, for very useful comments. Any errors are mine. 1 For a similar list, see J Goudkamp, Tort Law Defences (Oxford, Hart Publishing, 2016) 1–5. 2 Strictly: ‘which entails that no liability arises with respect to a particular cause of action’. ‘Fact’ should be taken to refer to ‘fact or set of facts’.

286  Sandy Steel (3) a fact (proof of) the existence of which entails that no liability arises, although the elements of a cause of action/liability-generating rule are satisfied, for reasons which bear upon the substantive merits of the case; (4) a fact (proof of) the existence of which entails that an established liability is reduced or restricted in some way; and (5) a fact which releases the defendant from liability, or reduces their liability, the burden of proof of the existence or non-existence of which is upon the defendant. According to definition (1), the plea that the defendant did not breach a duty of care is a ‘defence’: the absence of a breach of a duty of care entails that no liability arises in the tort of negligence. Definition (1) therefore includes what are also referred to as ‘denials’: a fact whose absence entails that a cause of action is not satisfied. This usage is similar to that found in the French Code of Civil Procedure, where a défense au fond is defined as: ‘any plea which may reveal, after examination of the substantive law, the claim of the opponent to be unjustified’.3 Definition (2) refers to a subset of the facts referred to by definition (1). It refers to those facts which defeat liability despite the elements of a cause of action being possibly satisfied. A limitation of actions defence has this form. If the defendant argues that the claim is barred because the limitation period has expired, this is nonetheless consistent with the elements of the cause of action being satisfied. Call this the ‘externalist definition’, since it defines a defence as a fact that is external to the elements of the cause of action or liability-generating rule. Definition (3) refers to a subset of the facts referred to by definition (2). Where the defendant raises a limitation defence, his argument is not that his conduct, which would otherwise have given rise to liability, was in some way rationally defensible, for instance, that it was justified. Rather his argument is that the claim should fail for reasons which have no bearing upon the merits of the case. According to definition (3), then, a limitation of actions rule is not a defence. The distinction between definitions (3) and (1)/(2) is recognised in the French Code of Civil Procedure. A substantive defence (défense au fond) is distinguished both from an exception de procédure (arts 73–121 CPC) and a fin de ­non-recevoir (arts 122–26 CPC), all falling within the category of ‘moyens de défense’. An exception de procédure refers to procedural irregularities or jurisdictional lack of competence barring the hearing of the claim, while a fin de non-recevoir concerns the absence or loss of the right to sue due, for instance, to prescription. Call definition (3) the ‘merit-based definition’ of defences. Definition (4) refers to rules which restrict the scope of the remedy or range of remedies available where liability is established, but do not defeat liability entirely. For instance, contributory negligence is routinely referred to as a defence in English law, but is not under definitions (1)–(3).



3 Art

71 CPC (author’s translation).

Defences to Tortious and Contractual Liability in French Law  287 Definition (5) refers simply to who has the burden of proof in relation to a factual issue. If the defendant must prove the absence of fault under a particular rule, then the absence of fault is a ‘defence’ according to this definition. Typically, the classification of a fact as a defence in the sense of definition (2) entails that the burden of proof is on the defendant in respect of that fact. Thus there is a significant overlap between definitions (2) and (5). Call this the ‘proof-based definition’ of defences. This chapter will primarily be concerned with defences in the externalist definition (2) and proof-based definition (5) senses. Why? First, the externalist and proof-based conceptions have some claim to being the central or, at least, most widely employed notions of defence.4 Second, if we use defence to mean definition (1), then we are essentially referring to the entirety of tort and contract law. If definition (1) includes a ‘fact’ such as the absence of breach, or the absence of damage, then any absent element of any cause of action can be described as a defence. ­Similarly, if definition (4) is employed, then any remedy-restricting fact, such as a break in the chain of causation, will be counted as a ‘defence’. It is not possible to compare the entirety of English and French tort and contract laws in this chapter: a more restrictive definition is needed. This is subject to the caveat that, if a fact is counted as an externalist defence in English law, but is a denial in French law (or vice versa), then it merits inclusion here. Furthermore, the focus will be upon facts which satisfy both the merit-based and proof-based definitions. This is, again, partly for reasons of space, but also for the reason that such facts are likely to be within the core of what is counted as a defence.5

B.  Do French and English Law Recognise Defences? English tort law recognises defences on any of definitions (2)–(5) identified above. For instance, self-defence would fall within each definition except definition (4).6 The terminology of ‘defence’ is less prominent in contract, but it is clear that, for instance, duress, misrepresentation and frustration would satisfy at least one of these definitions, namely, the proof-based conception of defence (5).7 It is more difficult to identify merit-based defences (3) in contract law – that is fact(s) which defeat liability in virtue of a feature of the defendant’s conduct, despite the cause 4 See, on the pervasiveness of the proof-based definition, E Descheemaeker, ‘Tort Law Defences: A Defence of Conventionalism’ (2014) 77 MLR 493, 499; for the externalist conception, see K Barker, ‘What Is a Contractual Defence (and Does It Matter)?’ in A Dyson, J Goudkamp and F Wilmot-Smith (eds), Defences in Contract (Oxford, Hart Publishing, 2017) 17, 28 (describing it as ‘popular within the academic community’). 5 For a conception of ‘defence’ which combines the merit-based and proof-based definitions, see L Duarte d’Almeida, ‘Defining “Defences”’ in A Dyson, J Goudkamp and F Wilmot-Smith (eds), Defences in Tort (Oxford, Hart Publishing, 2015). 6 For an extensive list of existing defences in tort satisfying definition (3): Goudkamp (n 1) 135. 7 See A Dyson, J Goudkamp and F Wilmot-Smith, ‘Thinking in Terms of Contract Defences’ in Dyson, Goudkamp and Wilmot-Smith (n 4) 2–6.

288  Sandy Steel of action being satisfied – necessity, self-defence and statutory authority are rarely discussed in contract law works. Strikingly, the core Code civil provisions on extra-contractual tortious liability make no reference to a merit-based defence (3).8 Indeed, there are no such defences to liability under article 1240 Cc. Liability-defeaters such as self-defence, necessity and statutory authority are folded into the concept of faute: where one of these is made out, the defendant’s conduct is not fautif and so one of the elements of the rule in article 1240 Cc is not satisfied. Similarly, barring the claim in virtue of the claimant’s illegal conduct is primarily relevant, if at all, to the existence of dommage, which requires interference with a legitimate interest.9 So this defence (2) in English law is also merged into the elements of liability in French law. Contributory fault (faute de la victime) is recognised, despite the absence of clear provision for it in the Code: it is an example of (4) and (5). The legal nature of certain facts which defeat liability under the strict liability for the act of things under article 1242 Cc is not entirely clear. For instance, the defendant may rely upon self-defence.10 However, self-defence is normally conceptualised as a fact which precludes ‘fault’.11 Since fault is not an element of liability under article 1242, the plea cannot function as a defeater of fault. There are at least two possibilities: either it involves a denial of an implicit unlawfulness or wrongfulness requirement under article 1242, and is thus not a defence in sense (3), or it is extraneous to an established liability under article 1242, and is thus a defence in sense (2). At any rate, there are a number of liability-defeaters under article 1242 for which the defendant bears the burden of proof and thus are defences in sense (5) – for instance, the burden of proof for force majeure rests upon the defendant.12 The contract Code provisions refer at various points to the idea of an ‘­exception’.13 For instance, article 1219 Cc is part of a subsection entitled Exception d’inéxecu­ tion. This article provides that a person is entitled to withhold performance of their contractual obligation in the event of the other party’s non-performance, if that non-performance is sufficiently serious. Other provisions state defences in the proof-based sense of (5), even if they are not explicitly termed ‘exceptions’. Thus, a person will be liable in damages for breach of contract unless they prove that the breach was due to force majeure.14 As in English law, vitiating factors such

8 Arts 1240–1244 Cc. Art 1245-11 Cc, on liability for defective products, refers to a ‘cause d’exonération’. For other references to ‘exceptions’ in the general part on obligations, see art 1315 Cc (defence (exception) of co-debtors available to each debtor, unless personal to the debtor); art 1324 Cc, (debtor’s defences against assignee of debt); art 1328 (defences of substituted debtor); art 1336 Cc; art 1346-5 Cc (defences opposable against subrogated creditor); art 1386 Cc. 9 See below pp 305–307. 10 See below p 291. 11 See below p 291. 12 A Bénabent, Droit Civil, Les Obligations, 12th edn (Paris, LGDJ, 2010) [622]. 13 See further art 1182 Cc; art 1216 Cc (concerning ‘exceptions’). 14 Art 1231-1 Cc. Similarly, art 1151 Cc referring to a party’s ability to ‘faire obstacle à’ an action for nullity of a contract, without using the word ‘exception’.

Defences to Tortious and Contractual Liability in French Law  289 as duress and mistake are integral to whether a valid agreement exists, as opposed to defences in sense (2) or (3). However, the burden of proof for these issues is on the defendant. The primary contrast in relation to defences between French and English law, then, is between their tort laws: in French law, there are fewer defences in sense (2) because these are merged within the concept of faute or dommage. Three explanations of this may be given. First, the greater Roman law influence: in the classical Roman law of delict, matters such as self-defence went to whether a damnum was iniuria datum: the issue of fault in the sense of culpability and fault in the sense of wrongfulness were built into a single concept.15 In German law, too, these issues are not explicitly mentioned in the tort provisions of the BGB, and are part of the issue of unlawfulness (Rechtswidrigkeit). Second, the distinction between elements of the ‘cause of action’ and elements external to the cause of action is more difficult to draw in systems that are not explicitly focused on the concept of a wrong (putting aside questions of the burden of proof). Wrongs call for justification (or excuse). If what the defendant has done is not legally classified as pro tanto wrongful, then the idea of justification has no room to operate. French law clearly has implicit wrongfulness ideas, but these are less central to the structuring of the modern law. Relatedly, there appears to be no direct translation of ‘cause of action’ in French, so it is not straightforward to describe certain elements of liability as external to something else (a cause of action). Third, English law was conceptualised as a system of actions until relatively recently. In a system structured around the procedure for obtaining a courtordered remedy, there is a natural focus on what must be pleaded by whom and when.16 This perhaps helps to explain why certain elements of liability came to be ­considered external to the ‘cause of action’.

C.  The Concept of a Defence in French and English Law A legal system could have defences and not recognise the concept of a defence, or not consider ‘defences’ to be an important legal category. This is not the case in French law. This is shown in at least two ways. First, as noted earlier, the Code of Civil Procedure recognises a general category of défenses, which isolates meritbased liability-defeaters from others.17 Second, doctrinal writers group together

15 See E Descheemaeker, The Division of Wrongs: a Historical Comparative Study (Oxford, Oxford University Press, 2009) 171. 16 For the view that pleading is essential to the notion of ‘defence’, see R Stevens, ‘Should Contributory Fault be Analogue or Digital?’ in A Dyson, J Goudkamp and F Wilmot-Smith (n 4) 247, 250. 17 See above, p 286.

290  Sandy Steel certain liability-defeaters in one overarching category – causes d’irresponsabilité – and have elaborated taxonomies of this category. The overarching feature of this category – although this itself is often not considered explicitly to be the important unifying feature of this category – is that the burden of proof is on the defendant in respect of these liability-defeaters. Although merit-based defences in English law tend to be collapsed into elements of the ‘cause of action’ in French law, the merit-based conception of defence is nonetheless implicitly recognised. Consider Jourdain’s reason for holding that consent is not a justification (fait justificatif): ‘it acts much less as a justification than as a circumstance which prevents ab initio the characterisation of the harmful event as fautif’.18 The distinction being drawn here seems to be similar to that between an element of liability which prevents there being a wrong, and one which justifies an established wrong.19 In English law, the concept of defence in sense (2) is clearly recognised, though a clear distinction between defences in sense (2) and defences in sense (5) is really only present in more abstract doctrinal works.20 Tort textbooks include sections on ‘defences’, almost always in relation to specific torts rather than across torts as a whole. The underpinning idea is usually an unarticulated amalgam of defences in sense (2) or sense (5). There is rarely a need to distinguish between sense (2) and sense (5), since all defences in sense (5) are also defences in sense (2). However, not all defences in sense (2) are defences in sense (5). Goudkamp gives the e­ xample of limitation of actions: the claimant bears the burden of proof that their claim is not beyond the relevant time period.21

D.  The Taxonomy of Defences Certain liability-defeaters are typically grouped together by doctrinal writers as causes d’irresponsabilité.22 This category is, in turn, typically subdivided between faits justificatifs (justifications) and causes d’irresponsabilité subjectives (subjective agent-focused factors which defeat liability).

18 P Jourdain, ‘Droit à réparation – Responsabilité fondée sur la faute – Fait Justificatifs’, JCl Civil Code, arts 1382–1386, Fasc 121-20, [67]: ‘il agit bien moins comme un fait justificatif que comme une circonstance qui empêche ab initio de reconnaître au fait prétendument dommageable un caractère fautif ’. 19 See also the revealing title to the faits justificatifs section in G Viney, P Jourdain and S Carval, Traité de droit civil, Les conditions de la responsabilité civile, 4th edn (Paris, LGDJ, 2013) 644: ‘Les circonstances permettant de justifier la transgression d’une norme juridiquement obligatoire.’ 20 See generally Goudkamp (n 1) ch 1. 21 ibid 4. 22 See P Jourdain, ‘Droit à réparation – Responsabilité fondée sur la faute – Imputabilité’, JCl Responsabilité civile et assurances, Fasc 121-10, [1].

Defences to Tortious and Contractual Liability in French Law  291

(i)  Faits Justificatifs The faits justificatifs category includes legitimate defence (légitime défense), necessity (état de necessité), legal authority or permission (ordre de la loi; permission de la loi), and order of a legitimate authority (commandement de l’autorité légitime).23 The unity of faits justificatifs is usually said to be that they are all objective circumstances which prevent an act from being classified as fautif.24 They are ‘objective’ in the sense that they alter the normal legal quality of the act in question, rather than bearing upon the (subjective) imputability of the act to the agent. The standard definition implies, then, that faits justificatifs are denials of fault, rather than pleas which are external to the notion of fault. However, this is not entirely correct, as there can be a fait justificatif to liabilities not based upon fault.25 Perhaps in recognition of this Jourdain states that a fait justificatif removes the unlawfulness (illicéité) of an act, though this would require one to grant that strict liability under, eg article 1242 Cc, has an implicit unlawfulness requirement.26 The admitted ‘justifications’ in French law are similar to those found in English law. But there is no generally applied taxonomy of defences in English doctrinal writing, such as the French one between ‘justifications’ and ‘subjective l­iability defeaters’.27 This is perhaps a reflection of the relative neglect of the subject of defences in tort law. As a consequence of this, too, it is not always clear how ­justificatory defences operate. This is likely to vary, however, across torts. In the tort of negligence, for instance, (justified) self-defence would preclude a finding of breach of duty. In trespass to the person, self-defence appears to be extraneous to the cause of action: the tort is fully established, but justified nonetheless.28

(ii)  Causes d’Irresponsabilité Subjectives This category refers to factors negativing the ‘imputability’ of the act to the defendant’s agency, such as the defendant’s insanity or infancy, thereby relieving the defendant of liability. This seems to be an empty set in French tort law: there are no subjective agency-based factors which defeat tort liability.29 In criminal law, this category would include matters such as insanity. It is clear that the insanity of the 23 ibid [1]. 24 S Hocquet-Berg, ‘Synthèse – Faute delictuelle’ JCl Responsabilité civile et assurances, 18 November 2018, [31]. 25 Cass civ (2) 10 June 1970, D 1970, 691; Jourdain (n 18) [43] (on the former art 1384 Cc). 26 For the view that it does, see O Moréteau, ‘Basic Questions of Tort Law from a French Perspective’ in H Koziol (ed), Basic Questions of Tort Law from a Comparative Perspective (Wien, Jan Sramek, 2015) [1/161]–[1/165]. 27 However, recent work (Goudkamp (n 1)) has led to the re-organisation of a doctrinal text: see E Peel and J Goudkamp (eds), Winfield & Jolowicz on Tort, 19th edn (London, Sweet & Maxwell, 2014) ch 26. 28 Goudkamp (n 1) 107. 29 Jourdain (n 22) [2].

292  Sandy Steel defendant does not bar their civil liability under article 1240ff of the Code civil, nor even the possibility of their being at fault.30 However, although factors concerning the reasoning ability of the defendant, which negate imputability in the criminal law, are irrelevant, factors which negate ‘physical’ imputability defeat fault-based personal liabilities. Thus, if a person’s leg entirely unforeseeably strikes another person due to an uncontrollable muscle spasm, this will prevent liability under a liability rule where liability is contingent upon one’s own act (eg article 1240 Cc).31 In relation to the tort of negligence, English law and French law reach similar results here. A person may be liable in this tort for acts done during a schizophrenic episode; only a condition which ‘entirely’ eliminates responsibility, such as full loss of consciousness, would relieve of liability.32 It is not clear, however, that a person who is so relieved benefits from a ‘defence’ in English law, unless the first expansive notion of defence is adopted.33 An implicit requirement of liability in the tort of negligence is that one is engaged in conduct – one is ‘acting’. Suddenly falling unconscious is not something one does. Similarly, in French law, it might be questioned whether, under article 1240 Cc, fait de l’homme already requires attributable conduct.

(iii)  Causes d’Exonération Causes d’exonération are usually distinguished from faits justificatifs, but partially overlap with causes d’irresponsabilité subjectives. Thus each of the recent reform proposals distinguishes causes d’exonération from causes d’exclusion, the latter mirroring the category of faits justificatifs.34 The central instances of this category are force majeure and contributory fault. Sometimes, acceptation des risques and consentement are also included here.35 One explanation of this distinction is that causes d’exonération are denials, while faits justificatifs are defences in the sense of (3). This works for force majeure: where this plea is accepted, it will either be the case that the defendant is not at fault, or the causal link between the defendant’s conduct, or the causal link between the thing or person for whom the defendant is responsible, and the claimant’s damage, has been broken (or never existed).36 But it does not account for faute de la victime, which is not a denial (except where it acts as force majeure). Moreover, there is a general problem with this way of drawing the distinction: because faits justificatifs operate to negate fault, it is not straightforward to treat them as other than denials. 30 Art 489-2 Cc. 31 See Cass civ (2) 4 February 1981, Bull civ II no 21, D 1983, 1 note P Gaudrat, JCP G 1981 II 19656 (no liability for injury inflicted from fall caused by heart attack). See further G Viney, ‘Réflexion sur l’article 489-2 du Code civil’ RTD civ 1970, 262. 32 Dunnage v Randall [2016] QB 639. 33 See above, p 285. 34 Arts 1349–1351-1 of the Avant-projet Catala (exonération); art 46 of the Avant-projet Terré; arts 1253–1256 of the Projet de réforme. See also Bénabent (n 12) [621]ff. 35 Bénabent (n 12) [621]. 36 It is true that ‘factual’ or sine qua non causation will not have been disproven, but attributable or legal causation will have been.

Defences to Tortious and Contractual Liability in French Law  293 Another possibility is that causes d’exonération are distinguished from faits justificatifs by their rationale, rather than their conceptual form. The former might be termed ‘responsibility-based’ defences: they release from or reduce liability on the basis of the reduced responsibility the defendant bears for the outcome. Where the victim’s fault is partially causative of their damage, this reduces the ­defendant’s responsibility (at least if the defendant’s conduct is not especially culpable in relation to the claimant’s damage). But the fit is imperfect. To the extent that consent is treated within this category, as it is in the 2017 reform proposal, this understanding is misleading.37 The fact a person that consents to an act does not show that they are responsible for its occurrence. Suppose that I permit you to destroy my goods, but you would have destroyed my goods regardless of my permission. My permission has no bearing upon your action, so your action cannot be said to be my responsibility. Nonetheless, the fact that I permitted you to destroy my goods could relieve you of liability. This shows that consent operates to defeat liability independently of its bearing on the victim’s responsibility for the damage caused. It seems, then, that causes exonératoires, as a category, lacks unity either of form or rational justification, albeit, by and large, it could be said to be concerned with reduced responsibility for the outcome. The practical import of the classification of these rules as causes exonératoires in French law is that the burden of proof for their existence is placed upon the defendant.38 English doctrinal writers typically classify contributory fault as a defence, though a minority treat it as part of the law of damages, in virtue of the fact that it does not defeat liability, but reduces the amount of compensatory damages.39 There is no general defence of force majeure applicable across tort and contract, though an unforeseeable, uncontrollable event40 for which the defendant was not responsible could entail that an element of liability (for instance, breach of duty in negligence) is not satisfied.41 In practice, since force majeure will exclude faute, it is not truly a ‘defence’ to fault-based liability in French law either.42

(iv)  Other Defences? Illegality, exclusion or limitation of liability, and limitation of actions do not figure amongst faits justificatifs, causes d’irresponsabilité subjectives, or causes ­exonératoires, albeit these are typically classed as defences in English law. The exclusion of illegality, exclusion of liability, and limitation of actions, from the category ‘defences’ could arguably be justified on the basis that exclusion of liability, illegality, and limitation of actions, do not bear upon the substantive merits of the case. The fact that the limitation period has expired does bear upon the 37 Art 1257-1 of the Projet de réforme. 38 See eg Bénabent (n 12) [610]–[611] and [621]–[623]. 39 See NJ McBride and R Bagshaw, Tort Law, 6th edn (Harlow, Pearson Education, 2018) 737. 40 The classic conditions for force majeure are imprevisibilité and irresistibilité: Viney, Jourdain and Carval (n 19) 332–33. 41 ‘Act of God’ is termed a defence to the strict liability rule in Rylands v Fletcher (1868) LR 3 HL 330. 42 See Viney, Jourdain and Carval (n 19) 356.

294  Sandy Steel parties’ substantive entitlements (their primary and secondary rights). There is some support for the view that illegality operates in this way in French law, and limitation of actions is mentioned in the moyens de défenses section of the Code of Civil Procedure.43

(v)  Comparison with Common Law Taxonomy The only systematic, in depth, taxonomic study of defences in English tort law is Goudkamp’s Tort Law Defences. Goudkamp understands the concept of a defence in sense (2) and proposes that the existing law can be profitably divided into two sub-categories: (a) justifications and (b) public policy defences. This taxonomy, perhaps unsurprisingly, cannot be easily transposed to French law. First of all, as noted, faits justificatifs are not defences in sense (2). There are no justificatory defences in sense (2) in French tort law. If we adopted defences in sense (5), then the taxonomy captures faits justificatifs, and ‘policy’ defences such as illegality and limitation of actions, but it misses the other rules for which the defendant bears the burden of proof (eg force majeure) which are normally classified together.

II.  Specific Defences This section examines a selection of particular defences in each system. The discussion is structured around ‘justificatory defences’, ‘excusatory defences’, ‘responsibility-based defences’, ‘consent’ and ‘public policy defences’. This structure has been adopted to facilitate comparison of the two systems by reference to important normative issues raised by the law on defences. Particular defences are selected (a comprehensive analysis would require at least a book) in order to illustrate general themes, such as the role of the criminal law, the extent to which justified or excused conduct attracts liability, the scope of consent, the extent to which responsibility for outcomes is shared between victims and injurers, and the categorisation of a liability-defeating argument as internal to the claim or as an ‘external’ defence.

A. Justifications The terminology, concept and conditions of application of faits justificatifs are largely a product of the criminal law and doctrine.44 The admitted faits ­justificatifs 43 See below, p 305. On exclusion of liability, see P Delebecque, ‘RÉGIME DE LA RÉPARATION.— Modalité de la reparation.—Règles particulières à la responsabilité contractuelle.—Conventions relatives à la responsabilité’, JCl Notarial Répertoire, V Responsabilité civile, Fasc 210, [5]. 44 See Jourdain (n 18) [4]: ‘Construite par les criminalistes, la notion de “fait justificatif ” est issue du droit pénal’.

Defences to Tortious and Contractual Liability in French Law  295 in civil law, absent from the Code civil itself, are drawn largely from the ­provisions of the Code pénal.45 The recent reform proposals each would incorporate an explicit reference to the relevant provisions of the Code pénal.46 There are additional faits justificatifs recognised in special regimes, such as the exceptio veritatis, or parliamentary privilege in the law on the press.47 Strikingly, the general position is that, if a person has a valid justificatory defence in respect of an act under the Code pénal, then that act will not give rise to civil liability in virtue of it not being fautif.48 This is said to follow from the general ‘primacy’ given to the criminal law over civil law and the traditional principle of the unity of criminal and civil fault.49 As in English law, the criminal law of defences has received much greater doctrinal attention and theorisation than in tort.50

(i) Self-Defence; Légitime Défense51 The existence of légitime défense as defined by the criminal law excludes civil fault in respect of the damaging act which amounted to legitimate defence.52 ­Jourdain describes it as ‘natural’ that an act which is justified for the purposes of the criminal law be justified for the purposes of civil liability.53 This depends upon what the law is holding when it considers a person ‘justified’. Outside of the law, one acts justifiably when one acts for an undefeated reason.54 If, when the criminal law holds that person is justified, it is saying this – that the person acted for an undefeated reason – then it would indeed seem ‘natural’ that this conduct could not amount to fault in the civil law. If fault amounts to a failure to behave as a reasonable person, how could acting for an undefeated reason amount to a failure to behave as a reasonable person? In English law, when a person is held not to be guilty of an offence in the criminal law on grounds of self-defence, the law need not, however, be holding that this person acted for an undefeated reason. It suffices for the purposes of self-defence in English criminal law that the person honestly believed that they were responding to a threat, and their response was necessary and proportionate, given

45 Arts 122-4 to 122-7 CP. 46 Art 1352 of the Avant-projet Catala; art 45 of the Avant-projet Terré; art 1257 of the Projet de réforme. 47 See Jourdain (n 18) [17]–[18]. 48 ibid [5]. 49 ibid [5]. See ch 6 above, pp 109–113. 50 There appear to be few theses on faits justificatifs in civil law. The two usually mentioned are: V Bergeret, ‘La notion de fait justificatif en matière de responsabilité pénale et son introduction en matière civile’ (Thesis, University of Grenoble, 1946); J Dingome, ‘Le fait justificatif en matière de responsabilité civile’ (Thesis, University of Paris I, 1986). 51 See generally Viney, Jourdain and Carval (n 19) [563]ff. 52 See Jourdain (n 18) [43] citing, inter alia, Cass crim 31 May 1972, Bull crim 1972 no 184, Gaz Pal 1972, 2, 633, tracing back to Cass crim 19 December 1817, S 1818, 1, 170. 53 Jourdain (n 18) [5]. 54 See J Gardner, ‘The Many Faces of the Reasonable Person’ (2015) 131 LQR 563, 565.

296  Sandy Steel their beliefs about the threat.55 It may be that the person’s action was objectively unjustified, and even unjustifiable by reference to the beliefs that the defendant ought to have had if they were reasonable. It is consistent, then, for English law to adopt the position it does: that a valid plea of self-defence in the criminal law need not preclude a finding of trespass, or negligence, in tort.56 French criminal law, by contrast, adopts a more objective approach to selfdefence. It requires not merely an honest belief in the existence of the threat, but a reasonable one.57 Since the criminal standard is objective, there is thus less room for divergence from the objective standard of faute in civil law. However, it still seems open to doubt whether the reasonableness of a belief needs to be assessed in the same way in criminal law and tort law, given their different functions.58 Even if one might dispute the logic of excluding the possibility of civil fault where the criminal defence of self-defence is made out, it is likely that in most cases where a person acts in valid self-defence according to French criminal law, the claimant will bear significant responsibility for the damage. This perhaps helps to explain why self-defence extends to exclude strict liability under article 1242 Cc.59 However, this is not wholly convincing since there could be (perhaps rare) cases where the defendant has a reasonable belief that the claimant is attacking him, and yet the claimant is not posing a threat at all. In such cases, the claimant bears no responsibility at all for their damage (unless they have culpably created the impression that they are about to attack). It is not obvious why such cases should not fall within article 1242 Cc if the self-defence occurs through the use of a thing.60

(ii) Necessity; Etat de Necessité Under certain conditions, necessity excludes civil faute.61 Thus if the only, or only reasonable means for A to avoid a greater evil is to inflict a lesser evil that would normally be wrongful, then, if A is not at fault for this choice having to be made, A is permitted to inflict the lesser evil. However, in contrast to self-defence, it seems that necessity does not exclude the compensatory liability of the person who inflicts harm in order to prevent greater harm under non-fault-based tort liabilities, such as article 1242 Cc.62

55 J Horder, Ashworth’s Principles of Criminal Law, 8th edn (Oxford, Oxford University Press, 2016) 142. 56 See Ashley v Chief Constable of Sussex Police [2008] UKHL 25. 57 This seems to be the majority view: C Mascala, ‘Faits justificatifs. – État de nécessité’, JCl Pénal Code, art 122-7, Fasc 20, [46]–[47]. 58 For arguments against assimilation of civil and criminal fault in this context: Viney, Jourdain and Carval (n 19) 646. 59 See Cass civ (2) 22 April 1992, Bull civ 1992 II no 127, D 1992, 353 note J-F Burgelin, RCA 1992, comm 257, RTD civ 1992, 768 obs P Jourdain. 60 Unless, again, there is an unlawfulness requirement under art 1242 Cc, as noted above, n 26. 61 Viney, Jourdain and Carval (n 19) [566]ff. 62 See eg Cass civ (2) 26 January 1994, RTD civ 1994, 864 obs P Jourdain.

Defences to Tortious and Contractual Liability in French Law  297 So the victim of a person who acts in justified, reasonably mistaken, self-defence may have no claim in tort, but the victim of an act of necessity may do so, if the conditions of article 1242 are satisfied. It is not clear what justifies this distinction. If A inflicts harm on B in order to prevent much greater harm to A or C, and B does not have a claim under a strict liability rule, B may still be entitled to a remedy against A or C in quasi-contract.63 The Projet de réforme states that a fait justificatif removes liability for a ‘fait dommageable’ which entails that necessity should also be a defence to a contractual claim for damages.64 In English law, necessity is a tort defence in the sense of definition (3). Some textbooks draw a distinction between ‘public’ and ‘private’ necessity, arguing that only the former is a defence.65 Private necessity is roughly where a person acts in a way that is normally wrongful in order to prevent greater harm to himself; public necessity is where the greater harm avoided is to other persons. The decision of the Minnesota Supreme Court in Vincent v Lake Erie Transportation Co is usually adduced in support of the proposition that there is no defence of private necessity.66 In this case, the defendant shipmaster fastened his ship to the plaintiff ’s dock during a storm in order to prevent the ship from drifting off to sea. This caused damage to the dock. The Court held that the defendant was liable to pay compensation for the damage caused, holding that the defendant was permitted to damage the dock in order to preserve its property in an emergency, but that compensation must be paid. By contrast, where the important interests of others are at stake, (public) necessity has been held to be a defence.67 The English courts do not, however, explicitly draw the distinction between public and private necessity. Necessity is almost never discussed as a defence in contract. As a matter of principle, if property rights, as protected by tort law, may be overridden by public necessity, contractual rights also ought to be subject to this override.

B.  Excusatory Defences An excuse can be defined as a plea that the defendant ought to be relieved of liability not because they acted permissibly, but because they were not (fully) blameworthy for what they did.68 There appear to be no excuses in this sense in French tort law. If insanity is not a defence, then it would be odd if other facts which defeat blameworthiness were relevant.69 If the plea of force majeure operates as a denial of causation or fault, then it, too, is not an excusatory defence. 63 Viney, Jourdain and Carval (n 19) [571]. 64 Art 1257 of the Projet de réforme. 65 See Peel and Goudkamp (n 27) 796. 66 Vincent v Lake Erie Transportation Co 124 NW 221 (Minnesota, 1910). 67 Southport Corporation v Esso Petroleum [1953] 2 All ER 1204, reversed by the Court of Appeal, but upheld by the House of Lords: [1956] 2 AC 218. 68 For discussion of distinctions between justifications and excuses, see J Goudkamp (n 1) 85–87. 69 See above, pp 291–292.

298  Sandy Steel English tort law also appears not to recognise excusatory defences.70 There are rules which ameliorate the legal position of those who have reduced moral blameworthiness. For instance, the standard of care is modified by reference to a person’s youth.71 This is not a ‘defence’, however. None of the defences recognised by French and English contract law is ­probably best characterised as an excuse. Consider, for instance, the automatic discharge of contractual obligations due to a frustrating event in English law. Where a party successfully relies upon the doctrine of frustration, the legal conclusion is not that they breached their contractual obligation but are relieved of liability due to the absence of moral blame. The conclusion is that their obligation ceased to exist upon the occurrence of the frustrating event: the obligation is discharged, not breached.72 Similarly, the absence of fraud and duress are generally considered to be conditions for the validity of a contractual obligation in both systems, rather than distinct defences.73

C.  Responsibility-Based Defences A defence which reduces or eliminates liability in virtue of the attenuated responsibility of the defendant (or the shared responsibility of the victim) in respect of the outcome generating liability can be termed a ‘responsibility-based’ defence. This section focuses upon one such defence: the contributory fault of the victim. Both English and French law recognise that the victim’s own fault, where it causally contributes to their damage, may reduce their entitlement to damages. The law is conveniently considered by reference to these four issues: (a) the legal source and rationale of the defence; (b) the scope of the defence; (c) the assessment and effect of victim fault; (d) reform.

(i)  Source and Rationale Articles 1240–1245 Cc contain no explicit reference to contributory fault except in the incorporated provisions on liability for defective products.74 Nor is the rule mentioned in the contract provisions. Nonetheless, the rule applied by the courts is that the victim’s contributory fault may reduce partially or fully damages in both tort and contract.75 In English law, the triggering conditions of the defence are 70 For a defence of this claim, see Goudkamp (n 1) 85–97, and J Goldberg, ‘Inexcusable Wrongs’ (2015) 103 Cal L Rev 467. 71 Mullins v Richards [1998] 1 WLR 1304. 72 See generally E McKendrick, ‘Frustration: Automatic Discharge of Both Parties?’ in Dyson, ­Goudkamp and Wilmot-Smith (n 4). 73 Arts 1128 and 1130 Cc. 74 Art 1245-12 Cc. 75 P Jourdain, ‘Droit à réparation. – Lien de causalité. – Pluralité des causes du dommage’, JCl Civil Code, arts 1382–1386, Fasc 162, [56], [67] (contract).

Defences to Tortious and Contractual Liability in French Law  299 determined primarily by the common law, while the Law Reform (Contributory Negligence) Act 1945 requires the court to apportion liability where the defence is applicable.76 As to rationale, French authors have proposed at least two different theories. One is that contributory fault operates as a justified sanction of the claimant’s fault.77 This theory does not fit well with the fact that fault can in principle be established against the insane and against young children. Furthermore, if it operates as a sanction, it is a rather arbitrary one: the quantum of punishment is significantly affected by the size of the loss suffered. A different theory is that the victim is a joint cause of the damage whose responsabilité is only triggered by their fault.78 On this view, the victim is in a similar position to a jointly liable person against whom a contribution claim may be brought. The only peculiarity is that the jointly ‘liable’ person is the victim. A recognised difficulty with this view is that the victim’s fault need not involve a genuine liability to another person: the victim’s failure to take care for their own safety is not necessarily a generator of liability to the defendant or another person.79 This last point has led, in the English context, Robert Stevens to criticise the existence of the defence.80 His argument is that the claimant’s failure to take reasonable care for their own safety is not a breach of a legal duty owed to the defendant, and thus should be left out of consideration, except to the extent that the claimant’s unreasonable conduct breaks the chain of causation. In short, since the claimant does not owe it to the defendant to take reasonable care of itself, the fact that the claimant has not done so is not the defendant’s concern.81 A major difficulty with this argument is that, if the claimant’s unreasonable conduct is justifiably taken into account in determining whether there is a break in the chain of causation (as Stevens accepts),82 and may thereby relieve the defendant entirely of liability, then it is not clear why it cannot be taken into account, where it is less serious, in order partially to reduce the defendant’s liability. The argument for this cannot be that the claimant’s fault is none of the defendant’s business: this is inconsistent with taking into account the claimant’s fault as a matter of causation. A possible argument is that causation is binary: either the defendant caused the outcome or they did not. Once the claimant’s fault is

76 The ‘fault’ of the claimant which gives rise to the defence is conduct which ‘would apart from this Act give rise to the defence of contributory negligence’: Law Reform (Contributory Negligence) Act 1945, s 4. 77 B Starck, H Roland and L Boyer, Obligations I, Responsabilité délictuelle, 4th edn (Paris, LITEC, 1991) [1291]; B Puill, ‘Gravité ou causalité de la faute de la victime en responsabilité civile’ D 1984, chron 58. 78 F Chabas, L’influence de la pluralité des causes sur le droit à réparation (Paris, LGDJ, 1967) [47]. 79 The difference is recognised in L Josserand, ‘La responsabilité envers soi-même’ DH 1934, chron 73. 80 Stevens (n 16). 81 ibid 253: ‘the risks I run in relation to my own interests are nobody’s concern but mine’. 82 ibid 255.

300  Sandy Steel s­ erious enough, this breaks the chain of causation, such that the defendant is not causally responsible at all; conversely, if the claimant’s fault is not at that threshold of seriousness, then the defendant remains fully responsible. However, this argument would require us to accept that (legal) causation – or responsibility for an outcome – cannot come in degrees.83

(ii) Scope Both systems recognise contributory negligence as a generally applicable liabilityreducing rule, subject to exceptions. In English law, the defence does not apply to trespass to the person,84 the tort of deceit,85 or claims for conversion under the Torts (Interference with Goods) Act 1977.86 Lord Hoffmann explained that the basis of this exclusion of the defence in relation to deceit is ‘moral disapproval of fraud’.87 The need to signal disapproval of highly culpable conduct does not explain, however, the exclusion of the defence in conversion, which can be committed innocently, and in trespass to the person, which can also be committed without moral culpability. The defence only applies to claims for breach of contract where the contractual duty breached gives rise to a concurrent liability in tort.88 This has the odd consequence that a negligent breach of contract may be treated more favourably than a breach of a strict contractual duty because the person in breach can then rely upon the concurrent liability in tort to raise the defence; the current position in relation to breach of contract has been cogently criticised on this ground.89 In French law, there are two situations where the rule does not apply. First, it does not apply to intentional wrongful interference with goods resulting in a profit to the wrongdoer. Where the defendant had forged the signature of one of the claimant company’s directors to cash eight cheques on the company’s account between November 2000 and December 2001, the Cour d’appel had reduced the company’s damages by half on the ground that it ought to have verified its 2000 accounts, noticed the forgery, and prevented the later cheques from being drawn. The Cour de cassation overturned this.90 To allow the reduction would have been to permit the wrongdoer to retain some of their fraudulently acquired gains.91

83 For an argument that causation does come in degrees, A Kaiserman, ‘Partial Liability’ (2017) 23 Legal Theory 1. 84 Co-operative Group (CWS) Ltd v Pritchard [2012] QB 320. 85 Standard Chartered Bank v Pakistan National Shipping Corp (nos 2 and 4) [2003] 1 AC 959. 86 Torts (Interference with Goods) Act 1977, s 11. 87 Standard Chartered Bank [2003] 1 AC 959, 968. 88 Forsikringaktieselskapet Vesta v Butcher [1989] AC 852 (CA). 89 See Stevens (n 16) 262–63; J O’Sullivan, ‘Contributory Negligence and Strict Contractual Obligations Revisited’ in Dyson, Goudkamp and Wilmot-Smith (n 4). 90 Cass civ (2) 19 November 2009 no 08-19.380. 91 ibid. For a decision allowing contributory fault against an intentional wrong to goods where no gain was made by the wrongdoer: Cass crim 19 March 2014 no 12-87.416.

Defences to Tortious and Contractual Liability in French Law  301 However, it is clear that the mere fact that the wrong is intentional is not a bar to the application of the doctrine of contributory fault.92 Nonetheless, it could have a significant impact on the size of the reduction.93 Secondly, under the special strict liability regime applicable to traffic accidents, the defence is partially excluded. Where the victim is not a driver, and the injury is a personal injury (atteinte à la personne), the victim’s damages will be reduced only if their fault is ‘inexcusable’ and the sole cause of the damage.94 Within this category, there are also protected categories of victim – those aged less than 16 or more than 70, and those with a recognised disability of greater than 80 per cent reduced capability.95 In all cases, however, if the victim has ‘freely sought out’ the damage, this bars their claim.96 Although these rules – apart from those on protected victims – might be considered simply to apply the general law on causation – if the victim’s fault is the exclusive cause of their damage, then it follows that the defendant is not responsible for it – the defendant is only entitled to this causation argument if the claimant’s fault was inexcusable. Thus, in theory, the regime is more favourable to victims.

(iii)  Assessment of Victim Fault In French law, the same principles are said to apply to the determination of victim fault as apply to the determination of defendant fault.97 This is notionally consistent with the idea that the victim is partially ‘liable’ for its own damage and the contribution claim between the defendant and the victim coalesces into the claim for damages against the defendant.98 In English law, it is generally assumed that the same principles apply.99 Once the false idea that the victim is ‘liable’ for its own damage is set aside, however, it is not clear that the same principles ought to apply.100

(iv)  Assessment of Shares In both systems, judges enjoy discretion to determine the size of the reduction. Section 1(1) of the Law Reform (Contributory Negligence) Act 1945 provides that damages ‘shall be reduced to such extent as the court thinks just and equitable

92 Cass crim 23 September 2014 no 13-83-357. 93 Jourdain (n 75) [65]. 94 Art 3(1) loi no 85-677 du 5 juillet 1985. 95 Art 3(2) loi no 85-677 du 5 juillet 1985. 96 Art 3(3) loi no 85-677 du 5 juillet 1985. 97 Jourdain (n 75) [58]: ‘les règles appliquées sont parfaitement symétriques’. 98 See above, p 299. 99 See J Goudkamp, ‘Rethinking Contributory Negligence’ in S Pitel, J Neyers and E Chamberlain (eds), Challenging Orthodoxy in Tort Law (Oxford, Hart Publishing, 2013) 325. 100 ibid 325–27.

302  Sandy Steel having regard to the claimant’s share in the responsibility for the damage’. This has been interpreted as requiring the courts to consider the relative blameworthiness of the parties and the relative causal potency of their conduct.101 The Act itself does not prescribe specific reductions or other limits on the discretion to reduce damages. However the courts have developed some fixed reductions in particular factual situations – for instance, a 25 per cent reduction applies to a failure to wear a seat belt, if this would have entirely avoided the claimant’s damage.102 In French law, the same considerations – relative gravity of the respective faults in cases of fault-based liabilities and degree of causal contribution – appear to apply.103

(v) Reform Article 1254 of the Projet de réforme would give recognition to the defence within the Code as follows: A failure by the victim in his contractual obligations, his own fault or that of a person for whom he is responsible, provide a partial exoneration where they contributed to the occurrence of the harm.104

Two observations may be made. First, the article distinguishes breach of a contractual obligation from faute, and includes the former within the defence. This seems to imply that breach of a strict contractual obligation could reduce the entitlement to damages. If so, it is not obvious why strict contractual obligations should be singled out for special treatment – why not any situation where the victim is in breach of a strict obligation? Second, the second paragraph of article 1254 of the Projet adds that only gross fault (faute lourde) can lead to a partial reduction in cases of bodily injury. It is not clear whether this should be taken as the only situation (apart from the special regimes dealt with elsewhere in the Projet de réforme) where the defence does not apply. The cases involving fraudulently acquired gains are not explicitly considered.105 Subject to these points, the Projet de réforme would improve the position of victims in two ways.106 First, it excludes the defence in cases of simple fault contributing to bodily injury.107 This rule is explicitly extended to driver victims in the context of traffic accidents.108 Only if the victim’s fault is inexcusable does 101 Stapley v Gypsum Mines Ltd [1953] AC 663, 682. 102 Froom v Butcher [1976] QB 286. See further J Goudkamp and D Nolan, Contributory Negligence: Principles and Practice (Oxford, Oxford University Press, 2018). 103 Jourdain (n 75) [65]. 104 ‘Le manquement de la victime à ses obligations contractuelles, sa faute ou celle d’une personne dont elle doit répondre sont partiellement exonératoires lorsqu’ils ont contribué à la réalisation du dommage.’ 105 See above 14. 106 One category of victims who would be disadvantaged by the Projet de réforme are victims par richochet: see art 1256 of the Projet. 107 Art 1254 of the Projet. 108 Art 1287 of the Projet. On the other hand, the rule does not apply to product liability: art 1299 of the Projet.

Defences to Tortious and Contractual Liability in French Law  303 it reduce or exclude liability in the traffic accident context.109 Second, the faute of a victim lacking capacity to reason (privé de discernement) does not reduce their entitlement to damages unless it amounts to force majeure.110 The Projet de réforme would, then, generally enhance the protection of victims and more sharply distinguish the ‘responsibility’ of victims from that of injurers.

D.  Consent and Exclusion of Liability (i)  Consentement, Volenti Non Fit Injuria, Acceptation des Risques Both systems recognise that the victim’s consent to an infringement that would normally constitute a wrong or liability-generating event sometimes precludes the existence of the wrong or liability-generating event. One way in which this is recognised is through the plea of ‘consentement’ or, in English law, consent. In French law, consent is generally described as not an independent fait ­justificatif, mirroring the general position in the criminal law.111 However, in certain categories of case, it clearly precludes the existence of liability. In ­relation to interferences with property, privacy and related rights, consent prevents ­liability.112 In relation to interferences with the body, the general rule is stated to be that consent does not bar liability subject to exceptions. Cases falling within the general rule are, for example, duels and euthanasia.113 The exceptions are not clearly defined, but include therapeutic medical treatment, organ donation, biomedical research.114 Aesthetic surgery is permissible with consent so long as the risks are not disproportionate to the expected benefits.115 Jourdain goes as far as to say that consent defeats liability in relation to interferences which are relatively minor or useful.116 The Projet de réforme in article 1257-1 reflects the existing law in relation to proprietary rights: ‘Equally, there is no room for liability where an action causing harm prejudices a right or an interest over which the victim has a power of disposal if the latter has consented to it.’117 Use of the word disposer, which implies an alienable right, seems to preclude the application of this article to bodily interferences. If so, the Projet de réforme oddly has nothing (explicit) to say about consent to bodily interference in tort.

109 ibid. 110 Art 1255 of the Projet. 111 Jourdain (n 18) [64]. Sometimes the reason for considering consentement not to be a true fait ­justificatif is that it does not justify an established faute. See above n 19. 112 Jourdain (n 18) [65]–[66]. 113 ibid [68]. 114 ibid [69]. 115 ibid [73]. 116 ibid [69]: ‘légères ou utiles’. 117 ‘Ne donne pas non plus lieu à responsabilité le fait dommageable portant atteinte à un droit ou à un intérêt dont la victime pouvait disposer, si celle-ci y a consenti.’

304  Sandy Steel In English law, the general assumption, albeit there is a paucity of authority to vindicate this assumption, is that consent may defeat liability in relation to any (otherwise) tortious interference.118 This contrasts with the position in English criminal law, where the limits of consent are more tightly policed; consent to, for instance, certain levels of bodily harm is not valid unless it falls within an exceptional category.119 Perhaps the general French tendency to follow the criminal law in relation to faits justificatifs has influenced the structure of the law here. In French law, there is also the doctrine of acceptation des risques (‘acceptance of risk’). The conceptual relationship with consentement is not entirely clear. For instance, injuries inflicted during a boxing event fall within acceptation des risques, but could presumably also fall within consentement.120 The primary context in which acceptation des risques has been applied is sports and dangerous games.121 There are three main implications in this context. First, the decision to play the sport or game does not preclude strict liability arising under article 1242 Cc; the participants in the game are not taken to have given up their right to sue for injury caused by a thing under another’s control.122 Second, the voluntary nature of the activity can, however, influence what constitutes fault within the activity; generally a higher level of fault seems to be insisted upon.123 Third, there is an overlap with acceptation des risques and faute de la victime (this also applies beyond the sporting context). Sometimes the acceptance of the risk may itself be ‘faute’ and lead only to a partial reduction in damages. The Projet de réforme says nothing explicit on acceptation des risques. One explanation of this may be that its only role is to influence the interpretation of faute and faute de la victime: it is not an autonomous defence. Perhaps it is implicitly to that extent retained in references to those concepts in the Projet de réforme.124 Just as French law contains consentement and acceptation des risques, English law also has two strands to the notion of volenti non fit injuria, although these are not generally disaggregated in the doctrinal literature. One strand is transactional: consent in this sense is something that is given to another person in respect of an act. The other strand is choice-based: here the victim does not communicate or attempt to communicate their permission to another person, but simply freely and knowingly chooses to take a risk. Both strands are likely to defeat liability, but they

118 See eg Peel and Goudkamp (n 27) 786. 119 R v Brown [1994] 1 AC 212, 231 (exceptions including: surgery, male circumcision, tatooing, ear-piercing, boxing). 120 CA Douai 3 December 1912, DP 1913, 2, 198; TGI Paris 26 June 1973, D 1974, 185. 121 Jourdain (n 18) [86]. 122 Cass civ (2) 4 November 2010 no 09-65.947, JurisData no 2010-020692, Bull civ 2010 II no 176. 123 Jourdain (n 18) [89]. 124 In US tort law, arguments about ‘assumption of the risk’ have increasingly been reallocated to other defences: see K Simons, ‘Reflections on Assumption of Risk’ (2002) 50 UCLA L Rev 481, 482.

Defences to Tortious and Contractual Liability in French Law  305 do so in different ways. A free choice to expose oneself to risk is likely to break the chain of causation. By contrast, consent acts independently to defeat liability and not merely through its effect on causation.125

(ii)  Exclusion and Limitation of Liability126 In English law, where a person validly consents to an otherwise wrongful interference, no wrong is committed. Where a person relies upon the defence of exclusion of liability, this is not so: a contractually agreed exclusion or limitation of liability leaves untouched the wrong, but regulates the remedial consequence of that wrong.127 In short, consent goes to primary rights, exclusion of liability goes to remedial rights.128 This analysis is also accepted by French authors.129 Contractually agreed exclusions of tort liability are not valid under the current law, but are valid under certain conditions for contractual liability. The Projet de réforme would allow exclusion of non-fault-based liability, except where it concerns bodily injury.130 There is no necessary inconsistency in allowing consent to operate in relation to fault-based tortious liability but not permitting the exclusion of faultbased liability. As noted, consent operates on primary rights, precluding the existence of a wrong, while exclusion relates to remedial rights.

E.  Public Policy Defences A public policy defence can be said to be one whose justification is not based upon the justifiability of the defendant’s conduct or the responsibility of the defendant for the outcome. This section focuses upon one putative public policy defence: the illegal conduct of the claimant. In French law, the fact that damage arose from an illegal act by the victim may influence the ability to obtain compensation in at least two ways. First, the fact that the damage arose from illegal conduct, or compensation would serve to award a person the revenue they would have made from an illegal activity, could serve, procedurally, to bar a right to have the claim heard, without necessarily affecting the underlying primary rights and duties. Article 31 CPC states that ‘the right

125 See, drawing attention to the need for an agreement for volenti, and contrasting this with the more expansive state of the law, Peel and Goudkamp (n 27) 789. 126 See in detail ch 15. 127 Stevens (n 16) 249. 128 Occasionally, however, volenti non fit injuria is said to require an agreement to waive one’s right to sue. See Nettleship v Weston [1971] 2 QB 691, 702 (‘waive any claim for damages’). This is incorrect: consent relates to primary rights, not rights to sue or to obtain damages. 129 See above n 43. 130 Arts 1281–1283 of the Projet.

306  Sandy Steel of action is available to all those who have a legitimate interest in the success or dismissal of a claim’.131 This provision has been used to reject claims based upon the fact that the remedy would restore a person’s illegal earnings, for instance.132 Despite this, there are cases which claim that the maxim nemo auditur propriam turpitudinem auditur has no application in civil liability.133 The procedural bar does, however, provide a route by which illegality can prevent a claim for damages. Second, the illegal origin of the damage may affect the underlying substantive rights. A condition of the validity of a contract, for instance, is that it has a contenu licite.134 Similarly, in tort, sometimes it is said that only damage caused to a legitimate interest is compensatable.135 Nonetheless, it is usually held that the fact that the claimant’s damage arose through its own illegal conduct is irrelevant even to this issue.136 But the support adduced for this general rule is usually rather unconvincing. It consists of cases where the claimant’s illegal conduct is not in a causal connection with the damage suffered.137 Furthermore, the cases where the illegal conduct has been held to bar a claim are ones in which there is a clear causal connection or where permitting the claim would undermine the purpose of the breached prohibition.138 Illegal conduct of the victim is conceptualised as a defence in sense (2)  in English law139 and may bar a claim in tort or contract where (a) upholding the claim would be inconsistent with the aims of the prohibition rendering the victim’s conduct illegal, or would allow the victim to profit from its illegal conduct, (b) there are no stronger policy arguments for allowing the claim, and (c) rejecting the claim would not be disproportionate.140 The primary justification given for the defence is consistency: in some cases, an award of compensation would involve the law of tort undermining the criminal law by, for instance, providing compensation in respect of a punishment imposed by the criminal law.141 131 ‘L’action est ouverte à tous ceux qui ont un intérêt légitime au succès ou au rejet d’une pretention.’ (Translation available at www.legifrance.gouv.fr/Traductions/en-English/Legifrance-translations). 132 Cass civ (2) 30 January 1959, Bull civ 1959 II no 116; Cass civ (2) 18 November 1959, Bull civ 1959 II no 754. 133 Eg Cass civ (1) 14 December 1982, Bull civ 1982 I no 355, RTD civ 1983, 342 obs G Durry; Cass civ (2) 19 February 1992, JCP G 1993, II, 22170 obs G Casile-Hugues; Cass civ (1) 17 November 1993, Bull civ 1993, I, no 326, RTD civ 1994, 115 obs P Jourdain; Cass civ (1) 22 June 2004, Bull civ 2004 I no 182. 134 Art 1128 Cc. See also art 1162 Cc. 135 See art 1235 of the Projet de réforme. See ch 9 above. 136 See eg S Retif, ‘Droit à réparation. – Conditions de la responsabilité délictuelle. – Le dommage. – Caractères du dommage réparable’, JCl Civil Code, arts 1382–1386, Fasc 101, [105]; Bénabent (n 12) [680]. 137 See Cass civ (2) 2 February 1994, RCA 1994, comm 176, concerning bodily injury suffered by a person negotiating the sale of drugs. 138 Cass com 30 November 1999 no 97-15978, unpublished. 139 See the multiple references to ‘defence’ in the leading decision, Patel v Mirza [2016] UKSC 32. These could be understood as references to ‘defences’ in sense (1), but that is not reflective of some judges’ understanding. See eg Patel [233] (Lord Sumption), where the defence is considered a rule of procedure akin to the position in French law. 140 Patel v Mirza [2016] UKSC 42 (Lord Toulson). 141 Ibid [99]–[101] (Lord Toulson).

Defences to Tortious and Contractual Liability in French Law  307

III. Conclusion Both French and English tort and contract law recognise defences on both the externalist and proof-based understandings of that concept. Furthermore, both systems recognise these, or closely related, concepts in their self-­understanding, either in the law of procedure or in the taxonomy of the substantive law. The primary practical significance of defences in both systems is the allocation of the burden of proof. Two main contrasts emerge from the analysis. First, the French rules on defences are typically formulated with considerable generality: the rules on justification in theory apply across tort, contract and crime, without ­modification. No such generality is evident in English law. Second, particularly in relation to tort, French law is marked by a relative absence of defences in the externalist, merit-based, sense.

308

15 Contracts Concerning Civil Liability ZOÉ JACQUEMIN Absolute freedom makes a mockery of justice. Absolute justice negates liberty. Each concept is fruitful only if it is limited by the other. Albert Camus, L’Homme révolté1

I. Introduction Contract and civil liability are, to borrow a metaphor of Lord Hailsham, as dissimilar as oil and vinegar.2 Inexorably drawn to categories, French academics put them into the same class of condiments: the law of obligations.3 However, it is also recognised that contract and civil liability are logically distinct.4 Contract is based on the will theory; it is the instrument of freedom.5 Civil liability – especially

1 ‘La liberté absolue raille la justice. La justice absolue nie la liberté. Pour être fécondes, les deux notions doivent trouver l’une dans l’autre leurs limites’: ‘L’Homme révolté’ in A Camus, Essais (Paris, Gallimard, 1965) 407, p 694. 2 ‘[T]he thinking underlying the two systems is as incompatible as oil and vinegar, the one based on what the plaintiff ought to receive, the other based on what [the jury] think the defendant ought to pay’: Broome v Cassel & Co Ltd [1972] AC 1027, 1077 (with respect to the distinction between exemplary and compensatory damages). 3 Quite apart from French law and the countries which have been influenced by the Code civil, the category of the law of obligations also exists in the Germanic systems: it is called Schuldrecht in Germany and Austria, and Obligationrecht in Switzerland. 4 In this chapter, where reference is made to ‘liability’ it refers to ‘civil liability’. ‘Civil liability’ here refers both to contractual and extra-contractual liability following the pattern in the Projet de réforme, even though la responsabilité civile often refers only to extra-contractual liability in French discussions. 5 A reform proposal which was published by the Ministry of Justice in 2008 had proposed to elevate freedom of contract to the status of a ‘core principle’ of the law of contract, in first place ahead of legal certainty and good faith. According to a comparative study which was undertaken in the member states of the European Union and published in the same year, freedom of contract (liberté contractuelle in France, Privatautonomie in Germany) also constitutes the core principle of the law of contract at a European level: Association Henri Capitant des Amis de la Culture Juridique Française, Société de Législation Comparée, Principes contractuels communs (Paris, SLC, 2008).

310  Zoé Jacquemin extra-contractual liability – is founded on the different notion of restoring the status quo ante; it is an instrument of justice.6 Is it possible then to use the freedom provided by contract to alter the justice created by the rules of extracontractual liability? For a long time, this question was answered in the negative in French law.7 The Code civil permits the parties to use their freedom of contract to modify the consequences of non-performance (by inserting clauses on contractual liability)8 but it does not provide any way of managing the consequences of extra-­ contractual liability.9 This position is about to change. Echoing the law in numerous European jurisdictions (Belgium, Italy, Portugal, Germany, Greece, Switzerland and Turkey),10 the Projet de réforme de la responsabilité civile, which was published on 13 March 2017,11 recommends recognising as valid clauses which limit or restrict civil liability. Even though the practical effects of this recognition will be limited,12 it is one of the most significant alterations of principle in this proposal. In the 2016 version of the reform proposals, the provisions on modifying civil liability by contract were located in Chapter 4, Section 3 on the effects of liability.13 By contrast, in the Projet de réforme of 2017, contract terms concerning civil liability have their very own chapter: Chapter 5. The difference between the two versions is not superficial: the title of the category itself has been altered from ‘Contracts concerning reparation of a loss’ to ‘contract terms concerning liability’. The first question is, then, whether the relevant contractual modification (limitation and exclusion clauses and penalty clauses) affects liability itself or merely the reparation which is due as a consequence of liability. While Chapter 5, Section 2 acknowledges that a penalty clause creates ‘a penalty’, article 1284 of the Projet provides that the sum is, nonetheless, owed by way of ‘reparation’.14 However, there is no reference to ‘reparation’ in articles 1281 to 1283 in the section on ‘contract terms excluding or limiting liability’. It is likely, therefore, that this change is the result of the fact that its drafters considered that contract terms of this kind exclude or limit liability and not (merely) reparation. As such, an exclusion clause could,

6 According to Carbonnier, civil liability performs a ‘function of interpersonal justice’: J Carbonnier, Droit civil, t 4, Les obligations, 22nd edn (Paris, PUF, 2000) [1114] in fine; ‘contractual justice’ is also evident in the principle of good faith in French law. See Principes contractuels communs (n 5). 7 See below pp 314–316. 8 Clauses limiting or excluding liability, penalty clauses, agreed damages clauses. 9 The author of a leading thesis on this topic explains that it was not considered worthy of discussion when the Code civil was drafted. He cites the following extract from a speech given before the corps législatif in 1804: ‘It is easy to see why … we thought that it was inconceivable that such a question would arise nowadays and you will agree with the decision not to insult our century by dealing with it’: J Abras, L’aménagement conventionnel anticipé de la responsabilité extra-contractuelle (PUAM, 2008) 9. 10 See G Viney, P Jourdain and S Carval, Traité de droit civil, Les effets de la responsabilité, 4th edn (Paris, LGDJ, 2017) [355] and the references cited therein. 11 Arts 1281 and 1283 of the Projet. 12 See below, p 313 for examples. 13 Avant-projet de réforme du droit de la responsabilité civile (24 April 2016). 14 This implicitly suggests that penalty clauses presuppose the existence of harm and not simply non-performance of the subject-matter of the obligation (the prestation).

Contracts Concerning Civil Liability  311 for example, block not only reparation but also a claim for an injunction to stop unlawful acts, a remedy which has been established in article 1266 of the Projet.15 Replacing contrats (‘contracts’) with clauses (‘contract terms’) is the second change of a terminological nature. Indeed, neither penalty clauses nor clauses excluding or limiting reparation constitute contracts. Clauses are not defined by the Code civil but the provisions which use this term16 clearly indicate that they form part of a contract; a clause constitutes one of its ‘particular provisions’.17 Clauses are integrated within a contract; they do not constitute the contract.18 Thus, the new terminology seems, at first glance, to be more logical. The final change of significance with respect to form is that articles 1282 and 1283 have switched places. It does appear to be more sensible to deal first with the more common case of clauses restricting liability in a contractual context before proceeding to the rarer case of contracts relating to extra-contractual liability. At first sight, the topic of Les contrats relatifs à la responsabilité (Contracts concerning civil liability) which I was asked to consider by the organisers of the workshop which led to the present book, seems to be covered by Chapter 5 of the Projet de réforme on ‘contract terms concerning liability’ which includes two sections: the first is devoted to ‘clauses excluding and limiting liability’ (articles 1281 to 1283) and the second concerns ‘penalty clauses’ (article 1284). There is, however, the difference earlier noted between the term clause (contract term) and contract. Is it possible that, momentarily distracted, the organisers of the colloquium failed to notice that one should, from now on, speak of contract terms rather than contracts? This question must be answered in the negative. The decision to entrust me with the topic of contracts which concern liability was deliberate. I shall, therefore, consciously restrict myself to such contracts. A further fact leads me to focus on contracts which concern liability rather than contract terms which concern liability: while the Projet de réforme recommends changes with respect to the former, it simply affirms the current law on the latter. Article 1282 of the Projet provides that contract terms which exclude or limit contractual liability are of no effect in cases of gross negligence or dishonesty (faute lourde ou dolosive); this provision would enshrine in the Code civil the rule which is currently applied by the courts.19 In addition, article 1282 of the Projet states that contract terms which ‘deprive the debtor’s essential obligation of its substance’ are to be disregarded; this position replicates article 1170 Cc.20 The provision on penalty clauses

15 It could not, however, exclude the civil penalty set out in art 1266-1 as the operation of this provision presupposes fault of a kind which would in any case constitute an obstacle to the application of the clause (see arts 1282 and 1283). 16 Arts 1119, 1170, 1171, 1184, 1189, 1191 and 1192 Cc. 17 Association Henri Capitant des amis de la culture juridique Française, G Cornu (ed), Vocabulaire juridique, 10th edn (Paris, PUF, 2014). 18 Except, as we shall see, contracts concerning civil liability. 19 It is, nonetheless, an important development. 20 This provision, which was introduced by the Ordonnance of 2016, codifies a rule developed by the Cour de cassation in what are known as the Chronopost (Cass com 22 October 1996 no 93-18632,

312  Zoé Jacquemin is not novel either. Article 1284 of the Projet reproduces article 1231-5 Cc21 which was introduced by the Ordonnance of 2016 save for the fact that a ‘sum by way of damages’ has become a ‘penalty by way of reparation’. There is nothing of note in this change. An English lawyer might be tempted to interpret it as the transformation of a rule allowing liquidated damages into one which permits penalties, but this is not the case. Penalty clauses with a deterrent effect are already valid in French law and the reform of the law of contract in 2016 has actually diminished their special status and led to their absorption by agreed damages clauses.22 Moreover, the term ‘penalty’ (pénalité) was already present in article 1231-5 Cc, which supports the idea that the only goal of this terminological change was to harmonise the vocabulary rather than to make any change of substance. As such, therefore, the Projet de réforme does not make any changes to contract terms concerning liability. However, its contribution to the law on contracts concerning liability is substantial. What does a ‘contract concerning liability’ (contrat relatif à la responsabilité) mean? Since the Ordonnance of 2016 came into force,23 a contract in French law is defined by article 1101 Cc as ‘a concordance of wills of two or more persons intended to create, modify, transfer or extinguish obligations’. Civil liability is not defined by law but it concerns ‘all duties to answer in a civil forum for the loss which one has caused to another, that is to make amends’.24 Contracts concerning civil liability correspond, therefore, to the concordance of wills of two or more persons intended to modify or even extinguish their legal obligation to provide reparation for harm caused to another person.25 Paradoxically, such a contract is conceivable only in the domain of extra-contractual liability. Indeed, if the parties to a contract agree to modify the consequences of their breach, they will do so by a contract term rather than by a separate contract. A second contract modifying any contractual liability which may arise from non-performance of a first contract would, in fact, constitute a liability clause which has been ‘exiled’ to a separate contract, and it would come within the rules on such clauses. A contract concerning liability can, therefore, be only an agreement which modifies the consequences of extra-contractual liability. Is such a contract conceivable? The situations in Bull 1996 IV no 261, 223) and Faurecia 2 cases (Cass com 29 June 2010 no 09-11841, Bull 2010 IV no 115, RDC 2010, 4, 1253 obs O Deshayes, 1220 obs Y-M Laithier, JCP E 2010, 1790 comm P Stoffel-Munck). 21 The former is intended to replace the latter, though art 1170 would not be removed by art 1282, if the reform proposal is promulgated in its current form. 22 See too D Mazeaud, ‘Les conventions portant sur la responsabilité’ RDC 2007, 149, [14] (discussing the Avant-projet Catala which treated penalty clauses and agreed damages clauses as a single type of clause, not to be distinguished from each other). 23 It came into force on 1 October 2016, but gained its full effect only on its ratification by Parliament by loi no 2018-287 of 20 April 2018 which ratified the Ordonnance and came into force on 1 October 2018. This loi also made about twenty changes to the text of the Ordonnance. 24 Vocabulaire juridique (n 17). 25 Drawing on the full definition of contract in art 1101 Cc, contracts transferring the duty to provide reparation are also conceivable. The Projet de réforme does not envisage such contracts but it is arguable that, from this perspective, civil liability insurance is a ‘contract modifying liability’.

Contracts Concerning Civil Liability  313 which it could be used are not numerous but they certainly exist. Such contracts presuppose that the future defendant and the future claimant know each other and foresee the possibility that the former might injure the latter. It could concern a case, for example, of two neighbours who agree that certain nuisances (noise, aesthetic disturbances (nuisances visuelles), pollution etc) will be tolerated. Or two people who find themselves together on the same premises or land without any pre-existing contract, might wish to manage any damage which may be caused by machines, persons or animals: for example, cattle grazing in the same field,26 people working on the same building site or ships operating in the same port.27 Similarly, cohabitants might decide to exclude their respective liability for harm caused to personal property by one party’s pet or the other’s children from a previous relationship. Parties negotiating a contract could also be tempted to establish rules on reparation in the event that one party abruptly breaks off negotiations.28 The post-contractual period could also be the subject of a contract modifying liability. Rival undertakings may seek to modify any civil liability which may be imposed on them as a result of any future acts of unfair competition on the part of one or both of them.29 Finally, there are many instances of altruistic acts or rescues where liability is extra-contractual in the absence of a contract.30 The phrase ‘contracts concerning liability’ calls to mind other legal institutions which ought to be excluded from our discussion of the Projet de réforme. Compromise or settlement agreements (transactions) are certainly contracts which concern, or which could concern, liability as they are contracts ‘in which the parties, by reciprocal concessions, terminate a dispute which has arisen or avert a future one’.31 Nevertheless, unlike contracts which fall within the Projet de réforme, compromise agreements alter civil liability ex post facto, that is once the event which gives rise to liability has occurred. They allow the parties to reach agreement on liability which has ‘arisen’ (even though the dispute has yet ‘to arise’) and are, thus, more akin to alternative ways of resolving disputes.32 The Projet de réforme leaves intact the rules on compromise agreements, which incidentally go beyond civil liability; as such, these rules will be put to one side here.33 Liability insurance contracts are also contracts concerning liability34 but they will also be

26 On this point, see p 317 below. 27 Mazeaud (n 22). 28 ibid. 29 Abras (n 9) 13. 30 ibid; Abras provides several examples: free driving lessons, advice on horsemanship, or permission to visit a factory. 31 Vocabulaire juridique (n 17). 32 Significantly, when the loi no 2016-1547 of 18 November 2016 on the modernisation of justice in the twenty-first century changed the definition of compromise agreements, it did so in a section entitled ‘encouraging alternative dispute resolution’. 33 For a discussion of compromise agreements relating to civil liability, see Viney, Jourdain and Carval (n 10) [444]–[498]; P le Tourneau (ed), Droit de la responsabilité et des contrats – Régimes d’indemnisation 2018–2019, 11th edn (Paris, Dalloz, 2017) [412.03]. 34 See too n 25 above.

314  Zoé Jacquemin excluded from this chapter as they are outside the scope of the Projet de réforme. Finally, the following discussion focuses only on French law, though I hope that differences and similarities with English law will become apparent. The contracts which I will examine here are those in which two or more parties agree in advance to limit or exclude35 the extra-contractual liability of one or more of them.36 Articles 1281 and 1283 of the Projet de réforme are devoted to contracts concerning liability. It is immediately obvious that the principal difference of substance between the reform proposals of 2016 and the Projet de réforme of 2017 is the removal of the rule which made the enforceability of contracts of this kind contingent on unequivocal acceptance of the clause.37 More generally, the Projet de réforme is remarkably vague on the legal regime applicable to contracts concerning liability, though the gaps can be filled, at least partially, by drawing on the general law of contract (le droit commun des contrats). Articles 1281 and 1283 are clearly founded on contractual liberalism and therefore freedom of contract. At the same time, however, this liberalism is limited by another requirement, the need to protect the parties to the contract. The drafters of the Projet de réforme appear to have followed Camus’s advice: combine freedom and justice so that each can be limited by the other. The search for a balance between freedom and the protection of the parties to the contract is evident both in the express recognition of the lawfulness of contracts concerning liability and in the legal regime which applies to these contracts. I will examine each of these in turn.

II.  The Lawfulness of Contracts Concerning Liability The formal recognition of the lawfulness and therefore the validity of contracts concerning liability is one of the most significant changes recommended by the Projet de réforme. This general rule is not, however, absolute, as while it is based on the parties’ freedom of contract, the need to protect the parties leads to certain restrictions.

A.  Contractual Freedom (i)  Contracts Concerning Liability and Public Policy Until the recent reform of the law of contracts, rules on contract terms limiting or excluding liability did not appear in the Code civil. The courts were left to deal 35 For a discussion of the ways in which contracts can modify reparation otherwise than by ­restrictions or exclusions of civil liability, see below, p 326. 36 Three monographs dedicated to this topic have recently been published in France: Abras (n 9); M Leveneur-Azémar, Étude sur les clauses limitatives ou exonératoires de responsabilité (Paris, LGDJ, 2017); H Conte, ‘Volonté et responsabilité civile’ (Thesis, Toulouse, 2017). 37 See the discussion below, pp 322–323.

Contracts Concerning Civil Liability  315 with these types of terms and, at least as regards contractual liability, they decided that they were valid as early as the end of the nineteenth century.38 This led to the question whether terms of this kind, or more precisely contracts of this kind, could also modify extra-contractual liability, which the Cour de cassation answered in the negative.39 In a leading judgment delivered in 1955, it justified this position on the ground that delictual and quasi-delictual liability were matters of public policy (ordre public): ‘contract terms exonerating or restricting liability in a delictual context are a nullity as articles 1382 and 1383 of the Code civil are a matter of public policy and they cannot be disabled in advance by agreement’.40 Former articles 1382 and 1383 Cc, which are now found in articles 1240 and 1242 Cc,41 are considered to be mandatory rules, from which parties are not allowed to derogate by their ‘individual wills’ on the basis that they are so fundamental that they concern public policy (they are d’ordre public) within the meaning of article 6 Cc.42 There are, however, some flaws in this approach. The decision to classify the rules on extra-contractual liability as d’ordre public can be explained by the fact that they are ‘based on the desire to reconcile freedom of action with the safety of ­individuals and groups’ and that ‘this balance is so essential to all aspects of social life that it is neither legitimate nor appropriate to allow it to be modified’.43 Here we see once again the idea that the law of civil liability fulfils a function of ‘interpersonal justice’, to borrow the expression used by Carbonnier.44 Nevertheless, it is not clear that this ‘fair balance’ which the law of civil liability establishes ought to be elevated to the status of ordre public. As discussed above, we do, after all, allow agreements on reparation after the event: if extra-contractual liability really were d’ordre public, compromise agreements relating to liability would not be valid.45 A comparison with contract terms which limit or exclude contractual liability also weakens the argument that reparation should in all cases exactly reflect the harm suffered (the so-called principle of ‘equivalency of harm and reparation’) as a matter of public policy. Why should the rule be ‘supplementary’ (and therefore optional) as regards contract liability but mandatory as regards extra-contractual liability? This public policy argument has been challenged on several occasions and there is no need to repeat the points here.46 By recognising the validity of contracts c­ oncerning 38 See in particular Cass civ 24 January 1874, DP 1876, I, 133. 39 See eg Cass civ 3 January 1933, DH 1933, 133; Cass civ 18 July 1934, D 1935, 1, 38 note Roger. 40 Cass civ (2) 17 February 1955, D 1956, 17 note P Esmein, JCP 1955, II, 8951 note R Rodière; this case also appears in H Capitant, F Chénédé, Y Lequette and F Terré, Les grands arrêts de la ­jurisprudence, 13th edn (Paris, Dalloz, 2017) [185]. 41 Since the Ordonnance of 2016, mentioned above in n 23, came into force. 42 This provision states that ‘it is not possible to derogate by individual agreement from legislation which is a matter of public policy and public morality’. 43 Viney, Jourdain and Carval (n 10) [355]. According to these commentators, this argument of ‘ordre public’ would not be sufficient to reject limitation clauses generally, but could lead to the rejection of exclusion clauses, in that such clauses affect the ground of liability itself. 44 See n 6 above. 45 See the references cited above in n 33. 46 The mere fact that the rules on concurrent liability (la règle de non-cumul) allow the exclusion of delictual liability between the parties from the moment that a contract has been concluded suffices to

316  Zoé Jacquemin extra-contractual liability, the drafters of the Projet de réforme seem to have heeded these criticisms: they have clearly, if implicitly, discarded this longstanding rationale.

(ii)  Affirming the Validity of Contracts Concerning Liability Overturning the traditional position, article 1281(1) of the Projet de réforme provides that ‘[c]ontract terms whose object or effect is to exclude or to limit liability are in principle valid, in contractual as well as extra-contractual matters’. The formal recognition that contracts concerning extra-contractual liability are lawful and valid is one of the most significant features of the reform proposal. Nonetheless, this change did not come as a surprise as the principle of validity had already been affirmed by the Avant-projet Catala,47 by the private member’s bill proposed by Sénateur Laurent Béteille in 2010,48 by the Avant-projet Terré of 201149 as well as, of course, by the reform proposal of 201650 which was the predecessor to the Projet de réforme of 2017. The terms of article 1281 in the reform proposal of 2016 were, however, slightly different: unlike article 1281(1) of the Projet de réforme of 2017, it did not contain the words ‘or effect’. It is not clear what the drafters of the Projet de réforme sought to achieve with this phrase. A contract term which negates the existence of an obligation may have an exclusionary ‘effect’ even though it is directed at the obligation rather than at liability. However, commentators generally distinguish between exclusion clauses and clauses which negate the existence of obligations (clauses de non-obligation). On the other hand, perhaps the objective of this periphrasis is to allow parties to modify the rules on prescription: for e­ xample, clauses which shorten these periods would also have an exclusionary effect. Regardless of the answer to this question, article 1181 of the Projet de réforme has affirmed, in the most absolute terms, the validity of clauses the consequences, and not just the intention, of which lead to the exclusion or limitation of civil liability. The repudiation by the courts of the validity of contracts concerning liability is not as clear as it appears at first sight. There were some cases in which the Cour de cassation stated, or rather, implied that agreements concerning extracontractual liability could be valid, two cases being usually cited. The first involved an owner of land who operated a mine and had alienated the portion above the ground (the superficie). As was the usual practice in such instances, the contract of

undermine the argument that this liability is a matter of public policy. See J-S Borghetti, ‘La responsabilité du fait des choses, un régime qui a fait son temps’ RTD civ 2010, 1, [60] and ch 3 above, p 48 n 41. 47 P Catala (ed), Projet de réforme du droit des obligations et de la prescription, La Documentation Française, 2006, arts 1382 to 1382-4. 48 Proposition de loi portant réforme de la responsabilité civile no 657, 9 July 2010, arts 1386-33 to 1386-37. 49 F Terré (ed), Pour une réforme du droit de la responsabilité civile (Paris, Dalloz, 2011) art 48. 50 Arts 1281 and 1282.

Contracts Concerning Civil Liability  317 sale contained a clause under which the purchaser promised not to seek damages for any harm which the ordinary operation of the mine may cause above ground. In its decision of 1899, the Cour de cassation held that this clause could be invoked against the transferee from the original purchaser of the superficie.51 It is notable that the court relied principally on property law in coming to this conclusion rather than on the law itself governing the contract terms, but the fact remains that a clause excluding extra-contractual liability was successfully relied on before the Cour de cassation.52 The second case concerned horses which had been grazing in the same field. Their owners agreed that neither one of them would be liable to the other for any harm which their horses might cause to those belonging to the other. In this case, there was no pre-existing contractual relationship between the parties, but simply a ‘factual proximity’ of their horses, which could possibly lead to harm. This is why in case such harm would have occurred, liability, if not excluded, would have been of an extra-contractual nature. This case then raised precisely the issue of the enforceability of contracts concerning extracontractual liability. The claimant’s horse was injured by another horse belonging to the defendant; the defendant was not permitted to rely on the clause, not on the ground that the clause was invalid, but rather on the ground that by putting a sick mare into the field, he had clearly put the claimant’s horse at risk. By holding that the owner of the sick mare could not rely on the exclusion clause because he had ‘­undermined the parties’ equality with respect to risk’, the Cour de cassation implicitly accepted that contract terms excluding former article 1385 Cc53 were, as a general rule, valid.54 By permitting contracts concerning contractual and extra-contractual liability, the Projet de réforme places a good deal of emphasis on the freedom of the parties, but this permission is limited.

B. Limitations (i)  Excluding Personal Injuries Having set out the general rule that contracts concerning civil liability are valid, the drafters of the Projet de réforme immediately make an exception for personal injuries. Article 1281-2 thus states that ‘[h]owever, in the case of personal injury, liability cannot be limited or excluded by contract’. The idea that the entirety of the law of civil liability consists of mandatory rules is, as we have seen, questionable 51 Cass civ 12 December 1899, S 1901, 1, 497 note Tissier, D 1900, 1, 361 note Geny. 52 This case highlights the fact that the topic of contracts concerning civil liability is connected to the issue of a contracting party’s liability to a third party; on this point see also below, p 324. 53 This provision stated that ‘[t]he owner of an animal or a person who is using during the period of this use is liable for the harm caused by the animal regardless of whether it is in his keeping, has been lost or has escaped’; this provision is rarely used in modern times. See now art 1243 Cc (as inserted by Ordonnance of 2016). 54 Req 16 November 1931, DH 1931, 1, 555.

318  Zoé Jacquemin but it is clear that reparation for personal injuries is certainly an essential feature of our legal system which could be classified as a matter of public policy.55 Every previous reform proposal expressly provided that the rule on the lawfulness of contracts concerning liability did not extend to personal injuries.56 More generally, the Projet de réforme establishes a hierarchy of types of harm with personal injuries at the top of the pyramid. This reform proposal recommends for the first time that a series of ‘specific rules governing the reparation for losses resulting from personal injury’ be added to the Code civil. The attempt to make this topic coherent would be seriously undermined if parties were empowered to circumvent these rules. Hence, the parties’ freedom cannot be extended to personal injuries as the need for justice requires that this type of harm be protected more strongly than others. French law already excludes personal injuries from the parties’ sphere of influence. Article 1245-14 Cc for example, provides that ‘contract terms which seek to exclude or limit liability for defective products are forbidden and deemed not written’. However, this article also provides that where the parties are not consumers, there is an exception for ‘harm caused to property which is not used by the victim mainly for his own private use or consumption’. In other words, in the field of product liability, exclusion and limitation clauses are entirely forbidden and to be disregarded only in the case of personal injuries. The doctrine of assumption of risk, applied to sporting activities, provides a further example of the distinction between personal injuries and damage to property in relation to the parties’ exercise of their contractual autonomy. The idea underlying this doctrine is that the participants are taken57 to have accepted the risks which are usually associated with the sport and as such, if harm is suffered, they are not entitled to reparation. The courts applied this doctrine for a time to liability for the action of things58 but the Cour de cassation ultimately abandoned it in a leading case in 2010.59 Nonetheless, by excluding reparation for physical harm to property which is caused to a participant by a thing, article L 321-3-1 of the Code du sport indirectly preserves this doctrine.60 The distinction between damage to property, liability 55 See Viney, Jourdain and Carval (n 10) [341]–[342]. 56 See nn 47 to 50. 57 The doctrine of assumption of risk could be understood as a contract relating to liability but its automatic application without searching for actual consent on the part of the parties (nor even their awareness of the risk), does not argue in favour of a contractual analysis. On the link between assumption of risk and defences to liability, see J-S Borghetti, ‘Le comportement de la victime’ in L-F Pignarre (ed), La réforme du droit de la responsabilité, Actes du colloque du 25 novembre 2016 (Faculté de droit et de science politique de Montpellier, 2017) 75, 80 ff. 58 It was restricted to liability for the actions of things as the courts require ‘a breach of the rules of the game’ in claims based on liability for personal action, as well as liability for the actions of other people. Such a condition is incompatible with the doctrine of assumption of risk; the latter applies only to usual risks which are, by definition, risks arising in a game which is conducted in accordance with the rules. 59 Cass civ (2) 4 November 2010 no 09-65947, Bull civ II no 176. 60 Art L 321-3-1 Code du sport: ‘During the playing of sport, a sporting competition or training for a sporting competition in a place reserved permanently or temporarily for this purpose, participants are not liable for physical harm caused to other participants by things which are in their keeping within the meaning of article 1242(1) of the Code civil’.

Contracts Concerning Civil Liability  319 for which may be modified, and personal injuries, liability for which may not be modified, is, thus, already a feature of French law. Indeed it is surprising that the Cour de cassation has never had the occasion (or perhaps the desire) to lay down a general rule prohibiting contract terms excluding or limiting contractual liability for personal injuries, even though claims for personal injuries are becoming more prominent in this context.61 The Projet de réforme seeks to fill this gap by affirming that neither contractual nor extra-contractual liability for personal injuries may be excluded or limited by contract.

(ii)  Excluding Liability for Fault While agreements relating to contractual and extra-contractual liability are equally the subject of article 1281 of the Projet de réforme, the subsequent articles cover contract terms which limit or exclude contractual and extra-contractual liability respectively. Article 1283 provides that ‘[i]n extra-contractual matters, a person cannot exclude or limit his liability for fault’. Fault in contractual liability is also governed by article 1282, but in a very different way, as it provides that in this context fault precludes the application of a term modifying liability only in the case of deliberate misconduct or gross negligence (faute dolosive ou lourde), a position which is consistent with existing case law.62 Therefore, it is necessary to establish fault of a certain seriousness in order to preclude the application of a contract term excluding or limiting contractual liability. By contrast, any kind of fault will preclude the application of a clause excluding or limiting extra-contractual liability. Indeed, technically, and again in contrast to article 1282, fault prevents a contract modifying extra-contractual liability from even being validly concluded: a person simply ‘is not able’ to exclude or limit extra-contractual liability for fault. On the other hand, in the contractual context, a person is able to exclude or limit his liability, but if the debtor seeks to rely on the term in question, its protection will be denied him where he has deliberately failed to perform (dol) or if his negligence demonstrates his inability to fulfil his contractual obligations (faute lourde). Clearly, therefore, in the contractual context, it is the behaviour of the debtor subsequent to the conclusion of the contract which deprives him of the benefit of the term modifying contractual liability. By contrast, in the extra-contractual context, the very lawfulness of any term having either as its purpose or effect the restriction liability for fault is denied: the key criterion is not the behaviour of the debtor of the obligation but the basis of liability. In other words, a contract modifying extra-contractual liability can be valid only where it can affect a person’s

61 See, in particular, the analysis of the case law in Viney, Jourdain and Carval (n 10) [342]. 62 This outcome also results from art 1231-3 Cc which provides that ‘[a] debtor is bound only to damages which were either foreseen or which could have been foreseen at the time of conclusion of the contract, except where non-performance was due to a gross or dishonest fault’. See too former art 1150 Cc.

320  Zoé Jacquemin liability for others,63 for the action of things, for defective products,64 for neighbourhood disturbances and, possibly, for motor-vehicle accidents. Reverting to the example which was discussed earlier of animals grazing in the same field, the owners of the animals may exclude or limit their liability as the keeper of a thing by contract, but not their liability for fault.65 Therefore, if an animal belonging to one owner was harmed as a result of the fault of the other, the latter will not be able to rely on a contract modifying liability in a claim based on article 1240 Cc which is brought against him by the former.66 However, the general rule that contracts relating to extra-contractual liability for fault are invalid is problematic for several reasons. First, it is not clear that the exclusion of liability for fault is justified in theory. The underlying idea seems to be that while the entirety of the law of extra-contractual liability is not a matter of public policy, liability for fault certainly is of this nature. It is true that liability for fault has acquired the status of a ‘constitutional value’ and this does give this way of thinking some credence,67 but elevating liability for fault to the status of being a matter of public policy is, arguably, a step which has been taken too hastily. And while this approach is consistent with the case law of the Cour de cassation, in the rare cases in which it has considered the validity of contracts modifying extracontractual liability, the claims were actually based on strict liability.68 Indeed, several commentators have argued in favour of treating strict liability and liability for fault differently in this respect: in their view, contracts modifying liability should apply to claims based on strict liability but the will of the parties should be disregarded in cases of liability for fault.69 However, at a practical level, as other commentators have noted,70 denying the possibility of excluding liability for fault essentially deprives contracts of this kind of almost all their usefulness. Indeed,

63 Even though this ground of liability still presupposes fault on the part of the person for whom the defendant is liable. See art 1245 of the Projet de réforme. 64 Art 1299-1 of the reform proposal of 2016 provided that ‘contract terms which are intended to exclude or limit liability for defective products are prohibited and deemed not written’. However, this provision did not appear in the Projet de réforme. 65 See above, p 317. Under the current law, this liability would be based on art 1243 Cc which concerns liability of the keeper of animals. Under the Projet de réforme of 2017, art 1243, which concerns liability for the action of things, would be applicable. 66 However, the defendant could rely on the clause if the action were brought on the ground of ­liability for the action of things – and this would indeed be the case as regards the sick mare, since the ‘thing’ (the mare) was the instrument of the harm. By contrast, if art 1283 of the Projet had adopted the same formulation as art 1282 of the clause being deprived of ‘any effect’ in the case of a deliberate or gross misconduct, the defendant would not have been able to invoke any liability, including liability for things. This relies on the different points of view of arts 1282 and 1283: fault within the meaning of art 1282 refers to the conduct of the party whereas in art 1283 it refers to the ground of liability. 67 Conseil constitutionnel, décision 82-144 DC of 22 October 1982, loi relative au développement des institutions représentatives du personnel. 68 See above pp 316–317. 69 See in particular, B Starck, ‘Observations sur le régime juridique des clauses de non-responsabilité ou limitatives de responsabilité’ D 1974, chron 157. 70 A Feray-Laurent and E Pouget, ‘Les clauses encadrant la réparation’ in L-F Pignarre (ed), La réforme du droit de la responsabilité (Presses de la Faculté de Montpellier, 2017) 277, 305.

Contracts Concerning Civil Liability  321 the personal liability set out in articles 1240 and 1241 Cc (former articles 1382 and 1383) is the subject of many of the examples which were discussed above in order to highlight the use to which contracts modifying civil liability could be put. For instance, if this exclusion of liability for fault is enacted, parties negotiating a contract would not be able to limit their liability for abruptly breaking off ­negotiations71 and rival undertakings would not be able to modify the liability which they risk incurring for unfair competition.72 Arguably, it would have been better to have excluded from any possible contractual modification liability for intentional acts (la responsabilité délictuelle in its traditional sense), while allowing it as regards liability for non-intentional acts (la responsabilité quasi-délictuelle).73 Having established the general rule that contracts relating to liability are valid, we need to consider the proper regime of rules by which they are governed. The drafters of the Projet de réforme refrained from engaging in this exercise. By considering a series of questions, the answers to which come principally from the general law of contract, we can tentatively outline the regime which applies to contracts modifying liability.

III.  The Rules on Contracts Modifying Civil Liability The main difference between the reform proposal of 2016 and the Projet de réforme is that, unlike the latter, the former made the enforceability of contracts relating to liability contingent on unequivocal acceptance and the decision to abandon this rule leaves a yawning gap in the legal regime governing contracts relating to liability. As there are no specific provisions on consent nor on the content or terms of the contract, it is necessary to look elsewhere in order to build up a picture of the rules which apply to contracts relating to liability.

A.  The Freedom to Consent (i)  Abandoning the Requirement of Unequivocal Acceptance The rejection of the rule which set more exacting standards of knowledge and consent as to the restriction of liability is the most notable difference between the reform proposal of 2016 and that of 2017. In the contractual context, article 1283 of the reform proposal of 2016 provided that clauses excluding or limiting contractual liability ‘are of no effect if they were not brought to the attention of the party against whom they are invoked before the conclusion of the contract’.



71 As

suggested by Mazeaud (n 22). though the basis of this liability is open to debate. 73 See le Tourneau (n 33) [2335.24]. 72 Even

322  Zoé Jacquemin Similarly, in the extra-contractual context, article 128274 of the reform proposal of 2016 stated that ‘in regimes of liability without fault, a contract takes effect only if the person who invokes it proves that the victim accepted it in an unequivocal way’. The enforceability (and, thus, no longer the validity) of contracts modifying liability was, therefore, contingent on notification prior to the conclusion of the contract where the liability was contractual and on unequivocal acceptance where the liability was extra-contractual. None of these restrictions appear in the Projet de réforme. It is easy to see why the rule on prior notification was abandoned in the field of contractual liability. This condition was intended to protect parties who had agreed to clauses relating to contractual liability which were set out in general conditions and had not been specifically brought to their attention. However, it is no longer necessary to protect parties in this way as article 1119 Cc provides that ‘[g]eneral conditions put forward by one party have no effect on the other party unless they have been brought to the latter’s attention and that party has accepted them’. A separate rule requiring that an exclusion or limitation clause must be brought to the attention of the person affected appears to be redundant. There are in fact two possibilities: either the clause was ‘negotiable’,75 in which case the person against whom it is invoked was in a position to reject it (article 1110 Cc);76 or, alternatively, the clause may be part of ‘a collection of non-negotiable terms which are determined in advance by one of the parties’ (article 1110 Cc),77 in which case, the clause (or the contract terms en bloc if they are general conditions) ought at least to have been brought to the attention of the other party and accepted by him (article 1119 Cc). The drafters of the Projet de réforme found it equally unnecessary to set out a specific rule requiring contracts relating to extra-contractual liability to be accepted unequivocally. They must have thought either that contract law already provided sufficient protection or that is was futile to try to ensure that the parties to this particular type of contract were fully consenting. It is hard to imagine that the drafters of the Projet deliberately refrained from protecting the parties’ consent to contracts which seem, at least at first sight, to be significantly imbalanced78 as they include a clause excluding or limiting the liability of one party to the detriment of the other. It is more likely that they considered that the existing rules of the law of contract sufficiently equipped parties who wished to exclude or limit their liability. Apart from the provisions in the Code civil specifically on

74 As noted above, arts 1282 and 1283 of the reform proposal of 2016 correspond to arts 1283 and 1282 respectively of the Projet de réforme. 75 These terms replaced the phrase ‘may be freely negotiated’: loi no 2018-287 du 20 April 2018 which ratified the Ordonnance of 2016, art 2. 76 Art 1110(1) Cc (in the version which came into force on 1 October 2018) states that ‘[a] bespoke contract [contrat de gré à gré] is one whose stipulations are negotiable by the parties’. 77 Art 1110(2) Cc (in the version which came into force on 1 October 2018) provides that ‘standard form contract [contrat d’adhésion] is one which comprises a collection of non-negotiable terms which are determined in advance by one of the parties’. 78 However, this initial impression needs to be qualified, see below pp 324–326.

Contracts Concerning Civil Liability  323 ‘contract terms’ (les clauses) which have just been discussed, the abandonment of the rule requiring unequivocal acceptance may be compensated for by drawing on the rules on the formation of contract. Article 1113 Cc provides that ‘a contract is formed by the meeting of an offer and an acceptance by which the parties demonstrate their will to be bound’. Acceptance is, therefore, entirely necessary because in its absence, there will be no binding agreement. Furthermore, a lack of equivocation is inherent in the very definition of acceptance.79 Thus, the decision to reject a specific rule on acceptance in the Projet de réforme appears to have more of a superficial rather than a substantive significance.

(ii)  Method of Agreement In the absence of an indication to the contrary, a contract relating to civil liability is a consensual contract. Article 1109 Cc provides that ‘a contract is consensual where it is formed by the mere exchange of consents, in whatever way they may be expressed’. How are the offer and the acceptance to be expressed in the context of a contract relating to extra-contractual liability? In particular, is it possible that an exclusion of liability may appear only on a notice? Once again, we need to draw on the general law of contract to answer these questions. Could a notice which states, for example, that entry to a property is prohibited and that the occupier is absolved of all liability for property damage be interpreted as an offer of a contract excluding or limiting liability?80 As article 1114 Cc tells us, an ‘offer, whether made to a particular person or to persons generally, contains the essential elements of the envisaged contract, and expresses the will of the offeror to be bound in case of acceptance’. It is reasonable to think that the party who puts up such a notice intends to enter into a contract with the rambler who ventures onto his land and the former unquestionably intends the exclusion or limitation of liability to have legal effect. It is, therefore, possible that such a notice could constitute an offer. Equally, by venturing onto the land, the reckless rambler could be thought to have implicitly accepted the offer. According to article 1113 Cc, tacit acceptance may be found in a person’s ‘declaration or unequivocal conduct’. The rambler’s behaviour can be described as unequivocal where, fully aware of the notice, he still goes onto the land. A decision of the Cour de cassation in 1970 came to a conclusion which is consistent with this position: a clause limiting liability which formed part of a set of general conditions of sale was enforced even though these conditions had not been given to the victim but had been simply displayed in the defendant’s premises.81 With respect to extra-contractual liability, it is notable that Belgian law allows the operators of car parks to erect signs at their entrances which exclude

79 The art 1118(1) Cc provides that ‘[a]n acceptance is the manifestation of the will of the offeree to be bound on the terms of the offer’. 80 It is not possible to exclude personal injuries, as earlier explained: above, pp 317–319. 81 Cass civ (1) 3 June 1970 no 67-12789.

324  Zoé Jacquemin their liability for damage to or theft of vehicles.82 Nonetheless, there remains the problem of proving the existence of the offer and, especially, the acceptance. Since the issue of the method by which extra-contractual liability can be modified by contract is not addressed in the Projet de réforme, is it conceivable that such a contract could be ‘integrated’ into another contract? In particular, groups of contracts and contractual chains come to mind. Article 1234 of the Projet expressly permits a party to a contract to enforce an exclusion or limitation clause against a third party who is affected by contractual non-performance.83 The basis of this effect of the clause on a third party could be found in a contract concluded impliedly with that third party, although the path to this conclusion may be a difficult one as the third party may not even be aware of the clause. By contrast, article 1172-2 of the Avant-projet Catala provides that the effect of clauses excluding or limiting liability extends to parties to other contracts in the group of contracts ‘provided that those parties were aware of them at the time of their own contractual undertakings and that they made no reservation in this respect’. In similar circumstances, one could treat the absence of objection as a tacit acceptance so as to hold that the contract was concluded. This reasoning could also apply to ­article 1234 of the Projet de réforme. However, it is less relevant in that context as article 1234 already leads to a similar conclusion: a clause limiting or excluding liability can be enforced against a third party without having to take the roundabout way of establishing that a contract has been tacitly concluded between one of the parties to the contract and a third party.84

B.  Restrictions as to the Content of the Contract The absence of a provision on the consent necessary for contracts relating to ­liability gives free rein to the ‘autonomy of will’ of the contracting parties. The content of the contract is, however, subject to several limitations which act as a counterweight to this contractual freedom.

(i)  A Significant Imbalance In particular, there are a number of controls on the ‘balance’ of the contract. We have defined contracts relating to liability, but we have not said whether we

82 See H Conte (n 36) [384] and the references cited therein. 83 Art 1234(2) of the Projet provides that ‘a third party who has a legitimate interest in the proper performance of the contract can equally invoke, on the basis of contractual liability, a contractual failing where the latter has caused him harm. The conditions and limitations on this liability which are applicable in the relations between the contracting parties may be set up against him. Any contract term which limits the contractual liability of a party to the contract in relation to a third party is deemed not written.’ 84 On this point, see ch 4 above, pp 62–64.

Contracts Concerning Civil Liability  325 are concerned with gratuitous or onerous contracts.85 The preceding discussion may have given the impression that contracts relating to liability are unbalanced contracts as they include a clause limiting or excluding liability to the benefit of one party and to the detriment of the other. However, one of the parties may receive a benefit in exchange for the clause excluding or limiting liability. For example, if a business which is causing pollution wishes the owner of a neighbouring house to sign a document excluding the liability of the former for neighbourhood disturbances, it is likely that the latter would do so only on payment of a sum of money. Most of the time, the exclusion of liability comes at a price. The exclusion of liability could be mutual: the parties could agree not to sue each other. This is what happens, for example, where cattle belonging to different owners share the same field. Equally, the party who benefits from the clause limiting or excluding civil liability could provide monetary compensation (potentially in the form of a periodic sum) to the other. However, if the contract is onerous, the courts may intervene to ensure that it is not unbalanced. Article 1169 Cc provides that ‘[a]n onerous contract is a nullity where, at the moment of its formation, what is agreed in return for the benefit of the person undertaking an obligation is illusory or derisory’. This provision encourages the parties to the contract modifying civil liability to ensure that there is a fair balance between the party’s protection from liability and the recompense proposed. In a similar vein, a contract relating to liability, while not significantly unbalanced when it is concluded, may become so if what was initially given in exchange subsequently becomes derisory. In these circumstances, is the party which had accepted the clause modifying civil liability able to escape from the contract? At the time of making a contract relating to extra-contractual liability, its parties generally do not know if or when harm will be suffered. The parties may decide to let the contract continue for as long as possible in order to achieve the desired objective and so become contracts of indefinite duration. The Ordonnance of 2016 introduced rules on the duration of contracts: article 1210 Cc now provides that ‘[p]erpetual undertakings are prohibited. Either contracting party may put an end to such an undertaking under the conditions provided for contracts of indefinite duration’. Article 1211 Cc adds that ‘[w]here a contract is concluded for an indefinite duration, each party may put an end to it at any time, subject to respecting any period of notice provided by the contract or, in its absence, a reasonable notice’. The prohibition on perpetual agreements and the introduction of a way of escaping indefinite contracts could be useful for contracts relating to civil liability. Another rule introduced by the Ordonnance of 2016 protects a party of a contract which subsequently become seriously unbalanced. According to article 1995 Cc, if a change of circumstances that was unforeseeable at the time of the conclusion

85 Art 1107 Cc states that ‘[a] contract is onerous where each of the parties receives a benefit from the other in return for what he provides. It is gratuitous where one of the parties provides a benefit to the other without expecting or receiving anything in return’.

326  Zoé Jacquemin of the contract renders performance excessively onerous for a party who had not accepted the risk of such a change, that party may ask the other contracting party to renegotiate the contract and, in case of a failure of renegociation, that party may be granted a right to terminate the contract. This rule could be used in the context of contracts relating to civil liability concluded for a long period of time.

(ii) The Effects of Contracts Relating to Liability As regards the effects of contracts relating to liability, it should be noted that the Projet de réforme does not recommend treating exclusion and limitation clauses differently, though several commentators argue the contrary on the basis that the two clauses are different in nature.86 Furthermore, while the Projet opens the way to agreements limiting extra-contractual liability, it does not envisage that penalty clauses could apply to extra-contractual liability. Is this a possibility which is likely to arise? Penalty clauses are intended to guarantee the performance of contractual obligations. However, a penalty clause could be used to guarantee the performance of a legislative obligation, such as the obligation not to engage in anti-competitive or unfair practices. Rival undertakings could thereby modify the consequences flowing from any future breaches of competition law on their part. This would, moreover, have the advantage of solving the problem of the assessment of damages to be awarded in the context. Without needing to go as far as advocating the formal recognition of ‘extra-contractual penalty clauses’,87 it is unfortunate that the Projet de réforme considered only clauses which exclude or limit civil liability. In fact, contracts concluded between two (or more) persons concerning their civil liability could have consequences which go beyond those envisaged by the Projet. What about ‘threshold of liability clauses’ (clauses de franchise de responsabilité) for instance? Where animals are put to graze in the same field, their owners may agree not to hold each other liable as long as the level of harm does not exceed a certain value, say €500. This would constitute a contract relating to extra-contractual liability, but it does not appear in the Projet de réforme. It is a pity that contracts relating to extra-­contractual liability were not given their very own set of provisions articles (in a section 3, for example). It would then have been possible to set out specific rules on penalty clauses and all the other methods of modifying civil liability by agreement.

86 Viney, Jourdain and Carval (n 10) [356]; Leveneur-Azémar (n 36). 87 For a discussion of this point (by an author who is clearly in favour of this approach): O Gout, ‘Les conventions sur la réparation’ JCP G 2016 Suppl to nos 30-35 (25 July 2016) 47, in fine.

part vii Liability Beyond Damages

328

16 Comparative Reflections on Punishment in Tort Law MARCO CAPPELLETTI

I. Introduction While it is generally accepted that punishment constitutes one of the fundamental functions of the criminal law, its role is far more controversial in private law and, more particularly, in the law of torts.1 Common law systems have traditionally included punitive remedies for defendants who have adopted particularly reprehensible courses of action; most notably in the form of punitive damages in the United States, and exemplary damages in England. Despite receiving formal recognition in the nineteenth century in both jurisdictions, these remedies remain controversial. Points of disagreement relate to several issues, including the functions which should or should not be assigned to tort law, its relationship with the criminal law, the perceived need to preserve tort law’s coherence and, ultimately, the question of what tort law is for.2 Similar concerns also characterise discussions which have taken place on continental Europe over the past few decades on the role of punishment in tort law.3 The traditional position, which sees compensation as the exclusive concern of tort law and punishment as a goal alien to it, has been called into question and important developments are taking place. In this respect, particularly significant is the current French Projet de réforme de la responsabilité civile. While this reform project is mostly conservative, for it essentially seeks to restate and put into legislation the law as developed by national courts over the

1 The term ‘punishment’ is often used to encompass both ‘retribution’ and ‘deterrence’, where retribution is concerned with ensuring that wrongdoers get their just desert while deterrence refers to the goal of influencing future conduct by providing appropriate incentives. At other times, punishment is used as a synonym for retribution and thus as distinct from deterrence. Unless stated otherwise, in this chapter the term ‘punishment’ includes both retribution and deterrence. 2 See sections II and III below. 3 For a comparative overview see H Koziol and V Wilcox (eds), Punitive Damages: Common Law and Civil Law Perspectives, Tort and Insurance Law Series (Wien/New York, Springer, 2009) vol 25.

330  Marco Cappelletti last century, one of the main innovations it proposes is the introduction in the Code civil of a form of civil punishment (amende civile) targeted at wrongs committed with a view to profit (faute lucrative). Whether or not it will eventually receive legislative approval, this attempt to equip French law with an overtly punitive weapon marks an important development in the French attitude towards the role of punishment in tort law.4 Against this backdrop, this chapter explores and compares the different treatment of the idea of punishing civil wrongs in the US, English, and French laws of torts. This inquiry shows that punishment can be understood in two different ways. One approach sees it instrumentally as a device geared towards the achievement of societal goals such as deterrence or retribution in the interests of society. On a second approach, punishment is seen as rooted in ideas of interpersonal justice in that it promotes private interests by, for example, empowering the victim of a wrong to take her revenge against the tortfeasor. While both the instrumental and interpersonal conceptions of punishment appear in each of the three tort systems considered, they do so to differing extents and they are seen more or less favourably depending on the way in which tort law is conceived of. Consequently, the treatment reserved to punitive measures varies depending on which conception of punishment, instrumental or interpersonal, is embraced and on how this conception relates to broader ways of seeing tort law. Finally, a word on the scope of this chapter. Tort law can react to wrongdoings in a variety of ways. Depending on the type of tort involved, different remedies may be available, including compensatory, disgorgement, punitive (or exemplary) damages and other civil penalties. This chapter does not focus on either compensatory or disgorgement damages, for it is disputed whether or not, and the extent to which, they fulfil punitive functions. I rather focus on remedies or other measures whose punitive nature cannot be called into question, for their gist can be safely identified in retribution, deterrence, or both. Consequently, I consider US punitive damages, English exemplary damages, and the French amende civile as envisaged by article 1266-1 of the Projet de réforme.

II.  Punitive Damages in the United States: Instrumentalism and Interpersonal Justice as Conflicting Approaches to Punishment and Tort Law One of the most distinctive features of the law of torts in the United States is the availability of punitive damages. In most states the courts or juries have the power

4 A wide variety of terms have been and are used to refer to the same area of French law, namely ‘delictual liability’, ‘Aquilian liability’, ‘civil liability’, and ‘extra-contractual liability’. For reasons of convenience, however, I use the term ‘tort law’.

Comparative Reflections on Punishment in Tort Law   331 to award punitive damages whenever the defendant has harmed the claimant by acting in a particularly reprehensible way. The defendant may have caused harm maliciously or intentionally, or he may have shown such a reckless or wanton disregard for the claimant’s interests that a punitive measure seems appropriate. While punitive damages are a well-established tort remedy and their desirability is rarely called into question in the US context, their purpose is controversial, and contrasting theories suggest very different ways of understanding them. This theoretical uncertainty is due to a polarisation between ‘instrumentalist’ and ‘non-instrumentalist’ approaches to tort law that characterises contemporary legal thought in the United States.5 Instrumentalists typically see tort law as a tool for the realisation of socio-economic objectives, such as deterrence, loss-spreading, or wealth maximisation. This vision of the tort system assesses rules and doctrines by measuring how good they are in achieving these goals and by considering the effects they produce. By contrast, non-instrumentalists maintain that tort law is concerned with notions of interpersonal justice between private parties and that their reciprocal rights and duties should be the primary, if not the exclusive, focus of tort rules and doctrines. This diversity of approaches leads to different views in relation to virtually every aspect of tort law and punitive damages are no exception.6 As I will explain, instrumentalist and non-instrumentalist approaches attribute different meanings to the notion of punishment and they consequently ascribe different functions to punitive damages. These differences find expression also in the practical life of punitive damages which, in most states, present a combination of features that reflect both the instrumental and interpersonal conception of punishment (and of tort law more generally).7

A.  Punitive Damages as an Instrument of Social Policy Several features of the positive law as well as the majority view among courts and scholars suggest that punitive damages constitute a valuable instrument in the pursuit of a variety of goals such as criminal-like punishment, efficient deterrence, or societal compensation. The criminal-like function of punitive damages is emphasised in several judicial decisions. For example, the US Supreme Court has stated that punitive damages ‘serve the same purposes as criminal penalties’8 and that they ‘may be properly

5 RA Posner, ‘Instrumental and Noninstrumental Theories of Tort Law’ (2013) 88 Ind LJ 469. 6 M Cappelletti, ‘Punitive Damages and the Public/Private Distinction: A Comparison between the United States and Italy’ (2015) 32 Ariz J Int’l & Comp L 799. 7 See BC Zipursky, ‘Palsgraf, Punitive Damages, and Preemption’ (2012) 125 Harv L Rev 1757, 1777ff; Cappelletti (n 6) 803–21. 8 State Farm Mut Auto Ins Co v Campbell, 538 US 408, 417 (2003).

332  Marco Cappelletti imposed to further a State’s legitimate interests in punishing unlawful conduct and deterring its repetition’.9 This criminal-like conception of punitive damages explains well why defendants who face claims for punitive damages are afforded procedural protections which are unavailable in claims for compensatory damages. For example, when seeking an award of punitive damages, the claimant must prove the seriousness of the defendant’s conduct according to a standard of proof that, in many states, is more burdensome than the ‘preponderance of evidence’ standard of civil trials: more than 30 states require ‘clear and convincing’ evidence, and one state (Colorado) requires proof ‘beyond a[ny] reasonable doubt’.10 Moreover, several states now require that the evidence necessary to trigger punitive damages and/or the quantification of the punitive award be dealt with in separate phases of trial.11 In this way, the defendant’s wealth and other potentially aggravating circumstances relevant to the award of punitive damages do not contaminate the compensatory phase of the trial and thus the risk of inflated compensatory damages is avoided.12 A criminal-like understanding of punitive damages can be also identified in academic theories that emphasise the retributive rationale of this tort remedy. For example, one view argues that punitive damages are particularly desirable in relation to powerful economic actors.13 Large corporations frequently manage to escape criminal sanctions both because egregious wrongdoings are often difficult to detect and prosecute and because of certain features and rigidities of the criminal process which are ‘friendly’ to white-collar wrongdoers.14 Punitive damages cure the inadequacies of the criminal law by providing private parties and their attorneys with a powerful incentive to do the investigative work and bring the wrongdoer to justice. In this respect, private parties act as private attorney generals and ‘assume … law enforcement functions’.15 As a result, wrongdoers are publicly punished for their conduct and deterred from treating legal norms and the cost of liability as mere taxes on their activities.16 Punitive damages are also seen instrumentally in theories committed to economic efficiency. A variety of models of punitive damages have been elaborated to suggest how deterrence of antisocial behaviour may be best achieved.17 9 BMW of N Am, Inc v Gore 517 US 559, 568 (1996). See also § 908 of the Restatement (Second) of Torts (1979). 10 AJ Sebok, ‘Punitive Damages in the United States’ in Koziol and Wilcox (n 3) 155, [74]. See Colo Rev Stat § 13-25-127(2) (2001). 11 ML Rustad, ‘The Closing of Punitive Damages’ Iron Cage’ (2005) 38 Loy LA L Rev 1297, 1321–24. 12 ibid. 13 M Galanter and D Luban, ‘Poetic Justice: Punitive Damages and Legal Pluralism’ (1993) 42 Am U L Rev 1393. 14 ibid 1440–45. 15 ibid 1445–46. The authors add that, since punitive damages are sought by private parties rather than by the state and since the personal liberty of the defendant is not at stake in tort litigation, punitive damages can be awarded without the safeguards of the criminal process (at 1454–58). 16 ibid 1425–28, 1430–35. 17 eg D Haddock et al, ‘An Ordinary Economic Rationale for Extraordinary Legal Sanctions’ (1990) 78 Cal L Rev 1; AM Polinsky and S Shavell, ‘Punitive Damages: An Economic Analysis’ (1998) 111 Harv L Rev 869; KN Hylton, ‘Punitive Damages and the Economic Theory of Penalties’ (1998) 87 Geo LJ 421.

Comparative Reflections on Punishment in Tort Law   333 What drives these theories is not the goal of retribution for blameworthy ­wrongdoings, but rather the adoption of legal rules that provide appropriate incentives to achieve the desired level of deterrence. For example, the ‘multiplier model’ suggests that punitive measures are desirable in situations where the wrongdoer has a chance of escaping liability; in this situation, achieving optimal deterrence requires that the defendant is ordered to pay a sum calculated by multiplying compensatory damages ‘by the reciprocal of the probability that the injurer will be found liable when he ought to be’.18 By contrast, the gain-elimination model suggests that, in situations where the defendant’s gain is less than or equal to the claimant’s loss, punitive damages should not seek optimal deterrence, but rather the complete deterrence of harmful conduct.19 This means that the defendant’s gain must be divided by the probability of liability; the resulting figure constitutes the minimum punitive award.20 Finally, a further expression of the instrumental conception of punitive damages sees this tort remedy as pursuing, in limited circumstances, ‘societal ­compensation’.21 This theory pays attention to the existence, in several states, of split-recovery schemes under which part of the punitive award accrues to the ­claimant and the remainder is instead allocated to some state- or court-­administered funds.22 In situations where the wrongdoer harms multiple victims, most of whom will not bring an action in tort, split-recovery schemes can turn punitive damages into socially compensatory damages, for they compensate society as a whole for the widespread harm that the defendant caused through his m ­ isconduct.23 ­Furthermore, understood in this way, punitive damages do not pursue retribution but they rather promote optimal deterrence by forcing the defendant to internalise in full the social costs of his harmful conduct.24 The picture which emerges so far suggests that elements in the positive law as well as the views of many leading legal scholars reflect a markedly instrumental understanding of punitive damages geared towards the pursuit of societal goals. This way of seeing punitive damages is part of a broader effort to interpret tort law as an instrument of social regulation and to turn tort rules and doctrines primarily 18 Polinsky and Shavell (n 17) 889. This approach assumes that the benefits of the harmful activity to the wrongdoer are to be ‘included in the calculation of social welfare’ (at 875). If this is not the case, as it happens when the wrongdoer derives pleasure from the sufferings of the victim, punitive damages are appropriate even if the wrongdoer had no chance of escaping liability (at 918ff). For a ­judicial i­ teration of this multiplier model see Mathias v Accor Economy Lodging 347 F 3d 672, 677 (7th Cir 2003) (Judge Posner). 19 Hylton (n 17). 20 ibid 423. For a judicial application of this model, see Jacque v Steenberg Homes, Inc 563 NW 2d 154, 165 (Wis 1997). 21 CM Sharkey, ‘Punitive Damages as Societal Damages’ (2003) 113 Yale LJ 347, 351. 22 ibid 372ff. 23 ibid 389ff. The author does not exclude that punitive damages may be grounded on retributive rationales in cases involving malice (at 387). 24 Sharkey (n 21) 354. See also Ciraolo v City of New York 216 F 3d 236, 245 (2nd Cir 2000) (Judge Calabresi); G Calabresi, ‘The Complexity of Torts – The Case of Punitive Damages’ in SM Madden (ed), Exploring Tort Law (Cambridge, Cambridge University Press, 2005) 333, 339–40.

334  Marco Cappelletti into mechanisms for the achievement of policy goals rather than for the accomplishment of some sort of interpersonal justice between claimant and defendant.25 In an intellectual climate where instrumentalism is predominant, the tendency to attribute societal goals to punitive damages appears as a natural development.

B.  Punitive Damages as Private Retribution (or Revenge) As a reaction to instrumentalism, a movement has emerged which prioritises considerations of interpersonal justice and marginalises or rejects policy concerns as valid justifications for the imposition of liability in tort.26 For example, a defendant may legitimately be held liable because she negligently breached a duty of care owed to the claimant, for such an obligation to repair would be readily justifiable on grounds of interpersonal morality.27 By contrast, a defendant should not be held liable simply because, if compared to the claimant, she is the cheapest cost-avoider according to some efficiency-based reasoning.28 Since theories of interpersonal justice are reluctant to accept policy considerations as legitimate reasons for imposing tort liability, the instrumental conception of punitive damages is problematic in that it grounds this tort remedy on collective goals such as criminal-like punishment, efficient deterrence, or societal compensation. Thus, if the tort system is seen as based on notions of interpersonal justice, punitive damages must be either rejected full stop, or reconceptualised in interpersonal terms. In most of its variants, the theory of corrective justice denies any role to punishment in tort law and thus rejects punitive damages.29 Based on notions of correlativity, corrective justice seeks to restore the balance between two parties after a transactional injustice has taken place. This result must be achieved while complying with two guiding principles. First, the injustice must be reversed only for reasons which are internal to the relationship between claimant and defendant (structural correlativity). Secondly, in terms of the remedies made available by the law, corrective justice requires that the claimant be put in the position in which she would have been but for the harmful event (content-related correlativity), thus not allowing more-than-compensatory damages.30 Neither structural correlativity 25 The literature showing this attitude is immense and in continuous expansion. Among the most influential and illustrative contributions, see R Coase, ‘The Problem of Social Cost’ (1960) JL & Econ 1; G Calabresi, The Costs of Accidents: A Legal and Economic Analysis (New Haven, Yale University Press, 1970); RA Posner, ‘A Theory of Negligence’ (1972) 1 J Legal Stud 29. 26 eg GP Fletcher, ‘Fairness and Utility in Tort Theory’ (1972) 85 Harv L Rev 537; RA Epstein, ‘A Theory of Strict Liability’ (1973) 2 J Leg Stud 151. 27 J Coleman, ‘The Practice of Corrective Justice’ in DG Owen (ed), Philosophical Foundations of Tort Law (Oxford, Clarendon Press, 1995) 53, 70. 28 For the concept of cheapest cost-avoider, see Calabresi (n 25). 29 eg EJ Weinrib, ‘Corrective Justice in a Nutshell’ (2002) 52 U Toronto LJ 349. Cf RW Wright, ‘Right, Justice and Tort Law’ in DG Owen (ed), Philosophical Foundations of Tort Law (Oxford, Clarendon Press, 1995) 159, 174–76. 30 EJ Weinrib, ‘Civil Recourse and Corrective Justice’ (2011) 39 Fla St U L Rev 273, 290.

Comparative Reflections on Punishment in Tort Law   335 nor content-related correlativity make any room for retribution or deterrence, and thus for punitive damages, in tort law. The demands of structural correlativity are violated because both retribution and deterrence represent ‘one-sided’ justifications that ‘focus not relationally on the parties … but unilaterally on the defendant’ for the imposition of liability; equally frustrated is content-related correlativity, for ‘[p]unitive damages do not restore to plaintiffs what is rightfully theirs, but instead give them a windfall’.31 Other theories of interpersonal justice have a broader, more flexible s­ tructure which makes room for punishment in tort law by reconceptualising punitive damages in interpersonal terms. The most influential theory of this sort is that of civil recourse.32 According to this theory, victims of wrongs are empowered by the state to recover in tort from the wrongdoer.33 Unlike corrective justice, which sees more-than-compensatory damages as an aberration, the theory of civil recourse makes room for punitive damages. Wrongs that are committed with malice or reckless disregard for the claimant’s rights may justify a remedy that entitles the victim to be punitive,34 or to seek revenge.35 In this respect, punitive damages simply constitute an additional avenue of civil recourse that the state provides to the victims of wrongs. Agreeing with the conclusion that punitive awards constitute ‘a form of legally sanctioned private revenge’,36 another well-known interpersonal theory of punitive damages justifies punishment in tort on the basis of the distinction between wrongs to society and wrongs to private parties.37 Wrongs done to society are the target of criminal punishment, whereas wrongs done to private parties are the target of private punishment (and hence of punitive damages). These two types of punishment have distinct goals: criminal punishment pursues the state’s interest in retribution and deterrence whereas punitive damages pursue the individual’s interest in obtaining revenge.38

31 ibid. See also D Braun, ‘The Risky Interplay of Tort and Criminal Law: Punitive Damages’ (2013) 11 Cardozo Pub Law, Policy & Ethics J 449, 467–69. 32 BC Zipursky, ‘A Theory of Punitive Damages’ (2005) 84 Tex L Rev 105. The author admits that the need to account for punitive damages is one of the ‘considerations motivating the development of [his] civil recourse view’ (at 156, fn 245). 33 ibid 151. See also JCP Goldberg, ‘Tort Law for Federalists (and the Rest of Us): Private Law in Disguise’ (2004) 28 Harv JL & Pub Pol’y 3, 7. 34 ibid 157. 35 AJ Sebok, ‘Punitive Damages: From Myth to Theory’ (2006) 92 Iowa L Rev 957. See also Clark v Cantrell, 332 SC 433, 442, 504 SE 2d 605, 610 (SC App 1998), where the Court of Appeals of South Carolina observed that, in addition to advancing the state’s interest in punishment and deterrence, punitive damages also vindicate private rights and that ‘[v]indication denotes punishment and revenge’ (added emphasis). On the minor differences between Sebok’s and Zipursky’s theories, both rooted in the theory of civil recourse, see Sebok, ibid 1025–29; Zipursky (n 32) 147. 36 T Colby, ‘Clearing the Smoke from Philip Morris v. Williams: The Past, Present, and Future of ­Punitive Damages’ (2008) 118 Yale LJ 392, 422, fn 125. Colby makes clear, however, that while his vision of punitive damages is close to theories of interpersonal justice, he is also ‘open to the proposition that tort law can sometimes be employed to serve solely public, regulatory aims’ (ibid 423, fn 125). 37 ibid 424–27. 38 ibid 436–40.

336  Marco Cappelletti In presenting punitive damages as essentially interpersonal and private, these theories are consonant with several features of the law of punitive damages:39 first, in US law, judges or juries cannot award punitive damages unless the claimant specifically requests them; secondly, in several states the punitive award still accrues to the claimant and not to some public fund; thirdly, there can be no punitive damages if the claimant does not get compensatory damages (even if only nominal), which means that there must have been some sort of violation of the claimant’s rights; fourthly, judges and juries have broad discretion in deciding whether or not to award punitive damages, even if the defendant’s conduct was in fact particularly reprehensible. Finally, the US Supreme Court has established that punitive damages cannot be awarded to punish a defendant for harm to individuals not involved in the litigation.40 Although state courts might not follow this rule,41 the Supreme Court’s holding emphasises that punitive damages should punish the wrongful infliction of harm on a particular private party, the claimant, and not on other members of society (no matter how numerous or how badly harmed). An interpersonal understanding of punishment fits all these features of the positive law better than criminal-like or other instrumental understandings of punitive damages and thus it strengthens those theories of tort law which seek to account for this legal domain in terms of interpersonal justice. In conclusion, in the United States there are two distinct and incompatible ways of dealing with the idea of punishing civil wrongs. On the one hand, the instrumental conception of punishment sees punitive damages as pursuing societal goals such as criminal-like punishment, efficient deterrence, or societal compensation. On the other hand, the interpersonal conception of punishment considers punitive damages as pertaining to the interaction between claimant and defendant and as empowering the former to seek private retribution or revenge against the latter. This conflict of approaches does not remain at an abstract level, for the law of punitive damages itself reflects this tension by including features that are consistent with the instrumental approach to punishment and features that are instead consistent with the interpersonal approach. Ultimately, the contrast between the instrumental and interpersonal conceptions of punishment and of punitive damages can be traced back to a wider clash between different ways of theorising tort law. The instrumental conception must be placed in the context of a more general conception, still prevalent today, of tort law as an instrument of social policy. The interpersonal conception of punishment, by contrast, is consistent with broader theories that see tort law as focusing on the reciprocal rights and duties of the parties. 39 Zipursky (n 7) 1784–85. 40 Philip Morris US v Williams 549 US 346 (2007). 41 In Philip Morris US v Williams, the US Supreme Court remanded the case to the Oregon Supreme Court to ascertain whether jury instructions made clear to the jurors that punishment had not to be inflicted for harm to non-parties. The Oregon Supreme Court, however, affirmed its prior decision. Philip Morris again petitioned the US Supreme Court, which eventually decided to dismiss the petition and thus in effect allowed the initial punitive award to stand.

Comparative Reflections on Punishment in Tort Law   337

III.  Exemplary Damages in English Tort Law … Or when Instrumentalism Clashes with Interpersonal Justice While in the United States punitive damages constitute a general remedy ­available whenever the defendant’s conduct is judged to possess a certain degree of reprehensibility, English exemplary (or punitive) damages are far more limited in their scope of application. Since the 1964 decision of the House of Lords in Rookes v Barnard, exemplary damages can be awarded in only three categories of cases: when there is an ‘oppressive, arbitrary, or unconstitutional action by the servants of the government’ (first category); when the ‘defendant’s conduct has been calculated by him to make a profit’ (second category); and whenever there is express statutory authorisation (third category).42 Lord Devlin’s categorical approach has been widely criticised by later courts and scholars on the ground that it constitutes an unprincipled confinement of a tort remedy.43 For example, defendants who seek to make a profit out of their wrongdoings are liable for exemplary damages under the second category, whereas defendants whose conduct is driven by malicious motives are not. Such a difference of treatment between gain-seeking and malicious conduct does not seem based on any rational justification.44 The reason why the law of exemplary damages lies in this peculiar state is that, on the one hand, both the courts and the legislator do not intend to expand this remedy’s scope of application and that, on the other hand, its abolition would amount to a ‘complete disregard’ of judicial and legislative authorities.45 As the ensuing discussion seeks to show, the reasons for the circumspection surrounding exemplary damages and for their failure to flourish relate to the tension between the perceived nature of exemplary damages and the prevalent way of theorising about English tort law. On the one hand, exemplary damages are widely seen as a form of criminal-like punishment, and the very few efforts to reconstruct them as a form of private retribution have so far failed to attract any meaningful support. On the other hand, English tort law is largely seen as a system of interpersonal justice which focuses on the binary relationship between claimant and defendant. This is a poor fit, for the instrumental conception of exemplary damages has the effect of importing into tort law societal considerations and concerns that are difficult to reconcile with an interpersonal vision of tort law.

42 Rookes v Barnard [1964] AC 1129, 1226–27 (Lord Devlin). 43 eg A Burrows, Remedies for Torts and Breach of Contract, 3rd edn (Oxford, Oxford University Press, 2004) 411, 424; J Goudkamp, ‘Exemplary Damages’ in G Virgo and S Worthington (eds), Commercial Remedies: Resolving Controversies (Cambridge, Cambridge University Press, 2017) 318, 330–31. 44 Kuddus v Chief Constable of Leicestershire [2002] 2 AC 122 [67] (Lord Nicholls). 45 Rookes (n 42) 1226. See also Burrows (n 43) 424.

338  Marco Cappelletti

A.  Exemplary Damages, An Example of Overt Instrumentalism Several elements in the positive law of exemplary damages as well as most academic and judicial discussions on this tort remedy are premised on the notion that exemplary damages constitute a sanction imposed in the furtherance of societal goals. The most commonly rehearsed justification for allowing exemplary damages is that the criminal law is sometimes unable to achieve fully its punitive goals.46 Exemplary damages can remedy the deficiencies of the criminal law by punishing wrongdoings which would otherwise go scot-free.47 On this view, tort law and criminal law play complementary roles in promoting the community’s interest in punishing antisocial behaviour.48 For example, exemplary damages are useful in relation to gain-seeking torts (Lord Devlin’s second category) because ‘[i]t is only if there is a prospect that the damages may exceed the defendant’s gain that the social purpose of this category is achieved – to teach a wrong-doer that tort does not pay’.49 Or, in relation to oppressive, arbitrary, or unconstitutional actions by the servants of government (Lord Devlin’s first category), exemplary damages are valuable in that they ‘uphold and vindicate the rule of law’ and ‘deter such actions in future’.50 This instrumental conception of exemplary damages is also reflected in numerous elements of the positive law. First, exemplary damages must be assessed by considering the defendant’s wealth, for ‘[i]f you make an award which might badly hurt the ordinary man in the street it might be laughable to a large company with very great means’.51 Secondly, the jury should award exemplary damages only if ‘compensation is inadequate to punish [the defendant] for his outrageous conduct, to mark their disapproval of such conduct and to deter him from repeating it’.52 Thirdly, if the defendant has already been punished by a criminal or other sanction, exemplary damages should not, in principle, be available.53 Fourthly, if multiple claimants are involved, the penalty imposed on the defendant must be divided equally among the claimants.54 46 Law Commission, Aggravated, Exemplary and Restitutionary Damages (Law Com No 247, 1997) [1.15]. 47 NJ McBride, ‘Punitive Damages’ in P Birks (ed), Wrongs and Remedies in the Twenty-first Century (Oxford, Clarendon Press, 1996) 175, 192–93; A Reed, ‘Exemplary damages: A Persuasive Argument for their Retention as a Mechanism of Retributive Justice’ (1996) CJQ 130; Burrows (n 43) 427–28. 48 Broome v Cassell & Co [1972] AC 1027, 1114 (Lord Wilberforce). 49 ibid 1130 (Lord Diplock) (emphasis added). 50 Kuddus (n 44) [79] (Lord Hutton). 51 John v MGN [1997] QB 586, 625. 52 Rookes (n 42) 1228 (Lord Devlin). 53 Archer v Brown [1985] QB 401, 402. For an overview of this issue, see J Edelman, McGregor on Damages, 20th edn (London, Sweet & Maxwell, 2018) [13–042]. Besides this double jeopardy (or non bis in idem) guarantee, it is noteworthy that, despite not officially embracing a heightened standard of proof and thus formally sticking to the ‘preponderance of evidence’ rule, there seems to be no case where an English court awarded exemplary damages when there was any doubt as to the defendant’s reprehensible conduct: see NJ McBride and R Bagshaw, Tort Law, 6th edn (Harlow, Pearson Education, 2018) 766. 54 Riches v News Group Newspapers [1986] QB 256, 288.

Comparative Reflections on Punishment in Tort Law   339 While most English legal actors entertain a profoundly instrumental view of exemplary damages, it is also possible to identify a private, interpersonal conception of this remedy. On this account, exemplary damages must not seek to achieve ‘deterrence of future wrongdoing’.55 Rather, exemplary damages can only pursue retribution, but in a very specific and narrow sense. Retribution must be understood as private retribution, ie as giving the defendant his just desert for violating a duty owed to the claimant.56 Similarly, other English scholars justify exemplary damages partly on the basis of the US theory of civil recourse, arguing that through exemplary damages the claimant is provided with an avenue of redress to satisfy ‘his natural desire to see the defendant pay for treating her [sic] in such an outrageous fashion’.57 This interpersonal approach can comfortably explain several features of the positive law: it is only the claimant who, upon infringement of his rights, can request exemplary damages; if these are granted, it is only he who can receive the award; if multiple defendants are involved, the claimant can request only one punitive sum and thus he cannot bring cumulative claims against all the defendants; insurance for exemplary damages is permissible; finally, a defendant can be ordered to pay exemplary damages in relation to a tort for which he is held vicariously liable.58 Nevertheless, contrary to the situation in the United States where theories of punitive damages based on ideas of interpersonal justice have become very influential, similar approaches to punishment and exemplary damages have not gained general acceptance in England. The reason for this may well lie in the uncompromising assumption that ‘punishment demands a state-individual relationship: it is not something for one individual to exact from another’.59 Conceiving of punishment as private retribution, or revenge between private parties, is a move that the vast majority of English legal actors is not (yet?) prepared to make. The intimate connection between punishment and the criminal law seems so entrenched in the minds of judges and legal scholars that any idea of making room for private revenge in tort law struggles to receive support.

B.  Tort Law as Interpersonal Justice: The Poor Fit with Exemplary Damages Understanding exemplary damages instrumentally does not in itself mean that their existence should be considered an anomalous feature of the law of torts. 55 R Stevens, Tort and Rights (Oxford, Oxford University Press, 2007) 85. 56 ibid. 57 McBride and Bagshaw (n 53) 767 and fn 57. Significantly, the authors also mantain, in an instrumentalist fashion, that exemplary damages are desirable in that they vindicate the rule of law and thus provide incentives to obey the law itself (ibid). See also A v Bottrill [2003] UKPC 44, [2003] 1 AC 449 at [29]: besides retribution and deterrence, exemplary damages ‘also serve as an emphatic vindication of the plaintiff ’s rights’ (Lord Nicholls). 58 Stevens (n 55) 86–87. 59 A Burrows, ‘The Scope of Exemplary Damages’ (1993) 109 LQR 358, 361.

340  Marco Cappelletti In this respect, the situation characterising the United States is instructive. As noted above, American punitive damages are widely understood as an instrument aimed at punishing antisocial behaviour in the interest of society, and yet they are not considered an incongruous feature of the tort system. This is due to the overwhelmingly instrumental approach that characterises contemporary tort law and theory in the United States. Since tort rules and doctrines are seen as pursuing a variety of societal goals, there is nothing inappropriate in having a corner of tort law that seeks to regulate conduct through a punitive remedy. By contrast, in England there is little willingness to see tort liability in instrumental terms and considerable emphasis is instead placed on ideas of interpersonal justice. On this approach, exclusive or primary relevance should be accorded to the reciprocal rights and duties of the parties whereas policy concerns such as loss spreading, wealth maximisation or, more significantly for present purposes, ­criminal-like retribution and deterrence should be irrelevant or, at most, marginal.60 In this framework, tort law is largely treated as either ­‘structurally’ inadequate to host an instrumental conception of exemplary damages or as ‘functionally’ incompatible with them.61 The ‘structural’ critique of exemplary damages is premised on the idea that tort law possesses a correlative structure which allows liability to be imposed only for reasons relating to the binary relationship between the parties.62 This means that pursuing the community’s interest in punishing antisocial behaviour frustrates tort law’s structural correlativity by breaking the claimant/defendant nexus:63 deterrence and criminal-like retribution are one-sided justifications of liability, for they both focus on the defendant alone while ignoring the particular claimant’s position; at the same time, to the extent that the defendant receives his just desert in the interest of society and that he and other members of society are deterred from adopting the same conduct in the future, exemplary damages achieve societal goals rather than interpersonal justice. A significant and related aspect is the clash between exemplary damages and content-related correlativity which, as seen above, expresses the idea that damages should restore the claimant to her status quo ante.64 As is well known, compensatory damages represent the most important example of this equivalence. Since compensatory damages are supposed to cure in full the injustice that took place between defendant and claimant, there is no interpersonal justification for deviating from the ordinary measure of damages.65 By contrast, exemplary damages enrich the claimant for reasons that

60 For an overview of the interpersonal approach to private law, see D Nolan and A Robertson (eds), Rights and Private Law (Oxford, Hart Publishing, 2012). 61 See P Cane, The Anatomy of Tort Law (Oxford, Hart Publishing, 1997) 116ff; McBride (n 47). 62 A Beever, Rediscovering the Law of Negligence (Oxford, Hart Publishing, 2007) 46. 63 A Beever, ‘The Structure of Aggravated and Exemplary Damages’ (2003) 23 OJLS 87, 105ff. 64 Livingstone v Rawyards Coal Co (1880) 5 App Case 25, 39 (Lord Blackburn). 65 Beever (n 63) 106–08.

Comparative Reflections on Punishment in Tort Law   341 are independent of her interaction with the defendant and that, as such, are at odds with the interpersonal nature of the tort system.66 In the light of all this, ‘punishment is foreign to the structure of private law’ and exemplary damages ought to be abolished to preserve tort law’s internal coherence.67 By contrast, the ‘functional’ critique of exemplary damages focuses on the distinct roles of tort law and criminal law.68 This critique, supported by several writers and judges, suggests that exemplary damages are anomalous because, through them, what should belong to the criminal law, namely retribution and deterrence, spills over into tort law, and an unacceptable blurring of functions ensues.69 On this view, the fundamental function of tort liability and damages is ‘to compensate the claimant for a wrong done to him’,70 which means ‘to place the claimant as nearly as possible in the same position as he would have been in if the tort had not been committed’.71 Without more, these statements do not reveal the goals of tort law; they simply point out what tort law does (it compensates) and what it should not pursue (punishment).72 Significantly, a commitment to compensation can derive from a variety of approaches and can be reconciled with several different objectives. For example, in economic theory compensation is generally appropriate to achieve optimal deterrence; imposing a greater or lower level of liability would result in over- and under-deterrence, respectively.73 Still different is the significance of compensation in a tort system inspired by the value of social solidarity, where it is the goal of loss spreading that fuels the imposition of liability.74 Both approaches are forward-looking and instrumental, in that they pursue societal goals such as optimal deterrence or loss-spreading and see tort liability and compensation as means to those ends. By contrast, English tort law

66 ibid 107. 67 ibid 105. 68 Rookes (n 42) 1221 (Lord Devlin); Broome (n 48) 1086 (Lord Reid); Ashley v Chief Constable of Sussex Police [2008] UKHL 25, [2008] 1 AC 962 at [17]–[18] (Lord Scott), [76] (Lord Carswell); Burrows (n 59) 361; A Burrows ‘Reforming Exemplary Damages: Expansion or Abolition?’ in Birks (n 47) 153, 158–59; H McGregor, McGregor on Damages, 19th edn (London, Sweet & Maxwell, 2014) 454, 455, 458; Department of Constitutional Affairs, The Law of Damages, Consultation Paper, CP 9/07 of 4 May 1997, [198]. 69 This position has been increasingly challenged. See eg Broome (n 48) 1114 (Lord Wilberforce); Kuddus (n 44) [63] (Lord Nicholls), [75] (Lord Hutton); GS Pipe, ‘Exemplary Damages after ­Camelford’ (1994) 57 MLR 91, 96; McBride (n 47) 195; Cane (n 61) 118; Burrows (n 43) 426; J Edelman, ‘In Defence of Exemplary Damages’ in CEF Rickett (ed), Justifying Private Law Remedies (Oxford, Hart Publishing, 2008) 225, 234–35; JNE Varuhas, ‘Exemplary damages: “public law” functions, mens rea and quantum’ (2011) 70 CLJ 284. 70 Kuddus (n 44) [95] (Lord Scott). See also AG v Blake [2001] 1 AC 268, 278 (Lord Nicholls). 71 One Step (Support) Ltd v Morris-Garner [2018] UKSC 20, [2018] 2WLR 1353 at [25], [31] (Lord Reed). 72 See the remarks of D Nolan, ‘Causation and the Goals of Tort Law’ in A Robertson and HW Tang (eds), The Goals of Private Law (Oxford, Hart Publishing, 2009) 165, 166 on the meaning of compensation. 73 Polinsky and Shavell (n 17) 873. See also R Cooter and T Ulen, Law & Economics, 6th edn (Pearson, 2012) 253. 74 See text to nn 79 and 99 below.

342  Marco Cappelletti is widely portrayed as ‘essentially backward-looking’75 and ‘driven, primarily or exclusively,’76 by the reciprocal rights and duties that the claimant and defendant hold against each other, with the consequence that policy considerations should be either rejected or treated as ancillary to interpersonal morality in determining tort liability.77 In this framework, instrumental concerns such as optimal deterrence or loss spreading are not valid justifications for compensation; it is instead the purpose of doing justice between the particular claimant and the particular defendant that should drive both the decision to impose an obligation to pay damages and the quantum of that obligation.78 Characterised in this way, the functional critique of exemplary damages collapses into, or at least becomes very similar to, the structural critique and thus conflicts with any approach that considers punishment a legitimate goal, and exemplary damages an appropriate remedy, of the tort system. It will be seen, therefore, that the English position as to the role of punishment in tort law is similar in one respect but markedly different in others to what was observed in relation to the United States. Like American punitive damages, English exemplary damages are widely understood in instrumental terms and administered in a way that turns them into devices for the attainment of goals such as retribution and deterrence of antisocial behaviour. However, while in the United States several sustained efforts have been made to conceptualise punishment and punitive damages in interpersonal terms, the same does not hold true as far as English law is concerned. The vertical, state/individual understanding of punishment greatly reduces the possibility of transforming exemplary damages into a form of interpersonal justice in which the victim of an egregious wrong can, if so she elects, exact private retribution or revenge from the tortfeasor. Furthermore, the polarisation in legal thought that characterises tort law in the United States cannot be detected in England. English tort law is predominantly seen as a system of interpersonal justice where tort obligations should depend first and foremost on interpersonal justifications and only limited, if any, relevance should be accorded to policy concerns such as retribution or deterrence. As a result, the treatment reserved to the idea of punishing civil wrongs is characterised by a poor fit between the instrumental conception of punishment and the tendency to theorise tort law in interpersonal terms. This tension is reflected in the infelicitous state of the law of exemplary damages and in the perception that they constitute an anomalous or exceptional feature of English tort law. 75 P Cane, ‘Justice and Justifications for Tort Liability’ (1982) OJLS 30, 44. 76 D Nolan and A Robertson, ‘Rights and Private law’ in D Nolan and A Robertson (eds), Rights and Private Law (Oxford, Hart Publishing, 2012) 1. 77 Nolan and Robertson (n 76) 1–2, 7, 21–23. To be sure, instrumentalist approaches may well rely on terms such as rights and duties, but these are used as mere ‘[fictions or] devices “for signifying a condition for a claim to arise and the person in whose favor it arises”’: ibid 3, quoting P Jaffey, ‘Liabilities in Private Law’ (2008) 14 Legal Theory 233, 244. 78 See Burrows (n 68) 156–57, who, in criticising exemplary damages, wrote that while ‘[t]he case for compensation rests on the uncontroversial idea of corrective justice: of restoring the plaintiff to his or her status quo ante … [t]he difficult problem revolves around finding a justification for pure punishment’.

Comparative Reflections on Punishment in Tort Law   343

IV.  Punishment in French Tort Law: Is Instrumentalism the Whole Story? Contrary to the American and English laws of torts, French tort law has not yet recognised a general remedy serving retributive or deterrent functions. Inspired by the value of social solidarity, the paramount goal of French tort law since the late nineteenth century has consisted of compensating the victims of accidents while diluting the costs of liability through a variety of spreading mechanisms.79 This focus on compensation and loss-spreading as the key concerns of tort law has monopolised the attention of French legal actors for most of the twentieth century, with little consideration given to the idea of punishing civil wrongs.80 However, since the mid-1990s the debate as to whether tort law should make some room for retribution and deterrence has greatly intensified, with several French academics discussing at length the role of punishment and the possible implications of its explicit recognition in the tort system.81 This debate has paved the way towards important reform projects that, in different ways, have sought to develop the punitive ambitions of French tort law. Most notably, in the context of a broader attempt to reform the law of obligations, in 2005 the Avant-projet Catala proposed the introduction of punitive damages for situations where the wrongdoer committed ‘a manifestly deliberate fault’.82 The solution envisaged also provided for a split-recovery scheme according to which the punitive award would accrue to the claimant but, at the court’s discretion, a portion of it may be allocated to the Public Treasury.83 Met with broad scepticism, this reform project failed to gain governmental support and it was never s­ ubmitted

79 J-S Borghetti, ‘The Culture of Tort Law in France’ (2012) 3 JETL 158. 80 cf B Starck, Essai d’une théorie générale de la responsabilité civile considérée en sa double fonction de garantie et de peine privée (Paris, L Rodstein, 1947) 396, who suggested that tort law may be usefully employed as a punitive weapon auxiliary to the criminal law. 81 See eg S Carval, La responsabilité civile dans sa fonction de peine privée (Paris, LGDJ, 1995); G Viney (ed), ‘Faut-il moraliser le droit français de la réparation du dommage?’ LPA 2002, 232; A Jault, La notion de peine privée (Paris, LGDJ, 2005); R Saint-Esteben, ‘Pour ou contre les dommages et intérêts punitifs’ LPA 2005, no 14, 53; S Carval, ‘Vers l’introduction en droit français de dommages-intérêts punitifs?’ RDC 2006, 822; M Chagny, ‘La notion de dommages et intérêts punitifs et ses répercussions sur le droit de la concurrence’ JCP 2006, I, 149, no 11; Y-M Laithier, ‘Le prononcé de dommages-intérêts punitifs en cas de faute lucrative’ in N Ferrier and A Pélissier (eds), L’entreprise face aux évolutions de la responsabilité civile (Paris, Economica, 2012) 117; S Carval, ‘L’amende civile’ in Avant-projet de loi portant réforme de la responsabilité civile. Observations et propositions de modifications, JCP G supplément au no 30-35 25 juillet 2016, 42; E Dreyer, ‘L’amende civile concurrente de l’amende pénale?’ JCP E 2017, no 25, 1344. 82 Draft art 1371 of the Code civil, as envisaged by the Avant-projet Catala. For a comparative discussion of this draft provision, see S Rowan, ‘Comparative Observations on the Introduction of Punitive Damages in French Law’ in J Cartwright, S Vogenauer, and S Whittaker (eds), Reforming the French Law of Obligations, 325. 83 This type of split-recovery scheme has been adopted in some US jurisdictions such as Alaska (ALASKA STAT §09.17.020(j)), Georgia (GA CODE ANN §1-12-5.1(e)(2)), and Utah (UTAH CODE ANN §78-18-1(3)).

344  Marco Cappelletti to Parliament for approval.84 A new attempt to provide French tort law with an overtly punitive measure is currently being made. Abandoned the idea of adopting punitive damages, the drafters of the Projet de réforme propose to introduce in article 1266-1 a form of punishment (amende civile) that aims to counteract the commission of fautes lucratives.85 As will be seen, the French conception of punishment is markedly instrumental in that it sees punitive measures as serving societal goals such as deterrence of antisocial behaviour. This approach is consistent with the profoundly instrumental way in which French legal actors understand tort law, that is as a socialised system for the allocation and spreading of losses. By contrast, theories of interpersonal justice do not play a significant role in the discussions on the structure and functions of tort law. Nevertheless, it is submitted that ideas linked to interpersonal justice can be identified in the debate on the role of punishment in tort and that these ideas considerably reduce the types of punitive measures that can be adopted without creating intolerable frictions with important principles of French law. To illustrate these points, the discussion will mainly focus on article 1266-1 of the Projet de réforme, as this provision reflects the most recent developments of French thinking in relation to the idea of punishing civil wrongs, as well as on the wider debate on the role of punishment in French tort law.

A. The Amende Civile (and Other Punitive Measures) as Instruments of Social Policy Despite the abundant use of the term peine privée (which may be translated as ‘private penalty’), the French understanding of punishment in tort law could hardly be more public. If punishment were understood in private, interpersonal terms, punitive measures should protect the claimant’s interests vis-à-vis the defendant’s conduct, rather than seek to pursue the interest of society in deterring antisocial conduct. This would entail a series of corollaries at the level of the positive law: for example, the punitive order should only be granted at the specific request of the claimant and the award should accrue in its entirety to her.86 However, if proper consideration is given to the debate on the role of punitive measures in tort law as well as to the rationale and content of article 1266-1 of the Projet de réforme, the ‘publicness’ (ie the instrumental character) of the French conception of punishment emerges in several respects. 84 J-S Borghetti, ‘La réforme du droit de la responsabilité civile en France’ LPA 2014, no 52, 16. 85 The amende civile is not a new concept, for it is already available in French law by way of ­piecemeal legislative measures. An example is art L 442-4 of the Code de commerce, which punishes unfair practices by providing that a public prosecutor or the Ministry of Economy can request that the wrongdoer be condemned to pay a civil penalty which may amount to millions of euros. 86 Jault (n 81) [358]–[412], esp [400].

Comparative Reflections on Punishment in Tort Law   345 To begin with, the adoption of punitive measures is supported on the ground that they are needed to supplement the criminal law or to fill the gaps created by a process of depenalisation which demotes certain wrongs from crimes to torts.87 The inadequacies and limits of the criminal law have been more acutely felt in relation to the fautes lucratives, ie deliberate gain-seeking torts, which may be committed in contexts such as unfair competition, environmental law, or the violation of personality rights.88 The choice itself to limit the scope of punitive measures to certain categories of wrongdoings rather than making them available whenever the defendant’s conduct infringed the claimant’s right is consistent with an instrumental approach to punishment. Furthermore, in assessing the different types of punitive measures, leading scholars have often argued that punitive damages may well be the best solution because the claimant would receive at least part of the punitive award. This would be desirable because the claimant is provided with an incentive to sue and is properly remunerated for the ‘service rendered to the community’ by bringing the defendant to justice.89 This sort of reasoning reflects a profoundly instrumental way of understanding punishment and punitive measures, for claimants are viewed as private attorney generals and tort law itself as a suitable instrument for punishing antisocial behaviour. Similarly, those who oppose the adoption of punitive remedies in tort calibrate their arguments based on the assumption of a criminal-like understanding of punishment. For example, it is often argued that tort law is only about compensation and that it should leave retribution and deterrence to the criminal law.90 ­Relatedly, if tort law were to perform punitive functions, it would u ­ ndesirably expose defendants to criminal-like sanctions without affording them the protections of the criminal process.91 Again, there is a strong inclination to see punishment as a tool for regulating conduct and advancing societal interests. But there is probably no better illustration of this than article 1266-1 of the current Projet de réforme. In proposing the introduction of the amende civile, article 1266-1 first provides that In extra-contractual matters, where the author of the harm has deliberately committed a fault with the view to making a gain or to saving money, a court may, at the request of

87 Carval (n 81) [215]; G Viney, ‘Rapport de sinthèse’ in La responsabilité civile à l’aube du XXIe siècle, Bilan prospectif, RCA no 6-bis, juin 2001, Hors-série, 82, 86; M Chagny, Droit de la concurrence et droit commun des obligations (Paris, LGDJ, 2004) [692]. 88 See G Viney, P Jourdain and S Carval, Traité de droit civil, Les effets de la responsabilité, 4th edn (Paris, LGDJ, 2017) [13]; P Jourdain, ‘Rapport introductif ’ in Viney (n 81) 3, [10]; D Fasquelle, ‘L’existence de fautes lucratives en droit français’ in Viney (n 81) 27, 28–29. 89 Carval (n 81) [319]–[323]; Chagny (n 87) [692]. 90 See eg Y Lambert-Faivre, ‘L’éthique de la responsabilité’ RTD civ 1998, 1, 19; M Behar-Touchais, ‘L’amende civile est-elle un substitut satisfaisant à l’absence de dommages et intérêts punitifs?’ LPA 2002, no 232, 36 who, while in principle not against the amende civile, argues that with it French law would have ‘le droit penal sans le droit penal’ ([34]). 91 C Jauffret-Spinosi, ‘Les dommages-intérêts punitifs dans les systèmes de droit étrangers’ in Viney (n 81) 8; Behar-Touchais (n 90).

346  Marco Cappelletti the victim or the ministère public and by specially justified decision, condemn him to the payment of a civil penalty.92

The draft provision then specifies the criteria for assessing the amende civile, which must be ‘proportionate to the seriousness of the fault committed, to the ability to pay of the author of the harm, and to any profits which he may have made from it’. Next, article 1266-1 sets a cap that the amende civile cannot exceed and it then provides that the amende ‘is allocated to the financing of a compensation fund related to the nature of the harm suffered or, if not, to the public Treasury’. It is finally established that the amende is not insurable. Both the rationale and content of article 1266-1 show the remarkable extent to which French legal actors are wedded to an instrumental understanding of punishment in tort law. To begin with, the French Ministry of Justice has identified the rationale of the amende civile in the deterrence of deliberate wrongdoings that qualify as gain-seeking torts.93 In this respect, article 1266-1 reflects the legal system’s condemnation for a cost/benefit analysis that treats potential harms to third parties and the subsequent costs of liability in tort as a cost of operation of one’s own activity. Such a blameworthy course of action requires the ‘moralising’ intervention of the law and, since gain-seeking torts may well not qualify as crimes, tort law can step in and fill the gap. But given that compensatory damages are likely to under-deter this type of conduct, a more-than-compensatory sanction is in order to induce potential defendants to refrain from exposing others to the risk of harm for reasons of mere financial convenience.94 The content of article 1266-1 also reflects in full an instrumental approach to punishment. As noted above, an interpersonal understanding of punishment would accord a central role to the claimant by providing that the punitive award can be granted only at her request,95 and that it must accrue in its entirety to her.96 By contrast, article 1266-1 marginalises the role of the claimant vis-à-vis the amende civile. It is true that the claimant is given the power to request that the defendant is condemned to pay a civil penalty, but it is unlikely that she would do so. Providing the necessary evidence to trigger the amende civile entails an effort and an increase in the attorney’s fees that are not cost-justified by what 92 As noted in n 19 of the translation provided in the appendix of this book, ‘[t]he ministère public … is a particular category of magistrat (broadly, a member of the judiciary) whose role in civil matters is to join proceedings (and sometimes initiate them) and submit oral or written arguments to the ‘sitting’ judges … as a matter of the public interest’. 93 See the statement of Jean-Jacques Urvoas, Minister of Justice, concerning the reform project on 13 March 2017, http://discours.vie-publique.fr/notices/173000657.html. It is noteworthy that the scope of application of art 1266-1 of the Projet de réforme is essentially the same as that of Lord Devlin’s second category. By contrast, art 1371 of the Avant-projet Catala was much wider in its scope of a­ pplication, as it referred to any ‘manifestly deliberate fault’. 94 The same need for a tort punitive response to fautes lucratives is echoed in the writings of French scholars. See eg G Viney, ‘Rapport de sinthèse’ in Viney (n 81) 66. 95 Jault (n 81) [275]–[277]. 96 ibid [311]–[316].

Comparative Reflections on Punishment in Tort Law   347 the claimant can get, which is nothing. The practical result of this is that in most cases it would be for the ministère public to join the tort dispute and ask the judge to order the payment of a civil penalty. Given the ministère public’s functions in criminal trials, most notably prosecuting crimes and punishing their authors in the interest of society, it is natural to consider his involvement in the operation of article 1266-1 as a strong indicator of an instrumental understanding of punishment. It may even be argued that the ministère public’s role in the tort process is the mirror image of the partie civile’s position in criminal trials: as the latter, in her capacity as victim of a crime, joins and ‘exploits’ the criminal process to promote her interest in receiving compensation, under article 1266-1 the ministère public joins and ‘exploits’ the civil process to promote the state’s interest in punishing wrongs committed with a view to profit. Furthermore, including the defendant’s wealth among the criteria for assessing the amende civile shows a commitment to deterrence which is clearly consonant with instrumentalism. The already conspicuous societal gradient of article 1266-1 is further increased by the choice of the beneficiaries of the amende. By allocating the civil penalty to compensation funds, article 1266-1 promotes the value of social solidarity. Together with mechanisms of social security, direct private insurance, and the tort system, compensation funds represent one of the pillars of the French system of social protection against the risks of harm associated with the machine age. More particularly, compensation funds constitute the collective sources of compensation which protect the victims of accidents in situations where compensation through the tort process or other separate insurance mechanisms is unavailable.97 In this respect, by financing funds that shelter otherwise unprotected victims of accidents, article 1266-1 of the Projet de réforme further reinforces the ethos of social solidarity that imbues French tort law and brings tort liability one step further away from any theory of interpersonal justice.98

B.  The Impact of the Principle of Full Compensation: A Legacy of Corrective Justice The French conception of punishment as well as the societal character of article 1266-1 of the Projet de réforme are fully consistent with the broader instrumental approach that permeates the theory and practice of French tort law. The increase in the number and severity of accidents in the second half of the nineteenth century,

97 A Favre Rochex and G Courtieu, Fonds d’indemnisation et de garantie (Paris, LGDJ, 2003). 98 On the emergence of the principle of social solidarity see J-L Halperin, Histoire du droit privé depuis 1804 (Paris, PUF, 2001) [124]–[129]. For a brief overview of the interplay among social security, direct private insurance, the tort system, and compensation funds see G Viney, Traité de droit civil, Introduction à la responsabilité, 3rd edn (Paris, LGDJ, 2008) [27]–[32].

348  Marco Cappelletti coupled with a thriving insurance market, turned French tort law from a system of individual responsibility into a socialised system for the allocation of the costs of accidents, based on the assumption that a loss hurts less if it is spread across many than if it is left concentrated on the unlucky victim.99 This pro-victim culture has weakened the claimant/defendant nexus by making tort liability depend on societal goals such as the spreading of losses, rather than on justifications ­pertaining to the interaction between claimant and defendant. Furthermore, insistence on the goal of protecting the victims of accidents has translated into a heavy reliance on strict liability, with the consequence that the role of fault as one of the conditions for the imposition of liability has been marginalised.100 As the fault paradigm lost importance, the deterrent effects of tort liability shrank. This decline of the ‘moralising’ function of tort law against blameworthy conduct has been judged negatively by several academics since the mid-1990s and, as a reaction to it, powerful voices have been raised in support of a restoration of the punitive dimension of French tort law.101 As discussed in the previous section, the effort to enhance the punitive potential of extra-contractual liability is infused with instrumentalism and it focuses on the attainment of societal goals such as deterrence of antisocial behaviour. This does not mean, however, that ideas of interpersonal justice have completely disappeared from the law of extra-contractual liability or, more importantly for present purposes, that they do not bear at all on the relationship between punishment and tort law. In this respect, the enduring influence of the principle of full compensation (principe de la réparation intégrale), visible in most discussions on the role of punishment in tort, is revealing. This principle is so firmly entrenched in French civil liability that it is very often taken for granted and seen as not requiring any justification;102 very importantly, however, several French scholars trace it back to the theory of corrective justice.103 As seen above, corrective justice is based on the notions of structural correlativity and content-related correlativity. By relying on the principle of full compensation as one of the fundamental tenets governing civil liability, French law adheres to content-related correlativity.104 The principle of full compensation requires that

99 See Viney (n 98) [18]–[26]; idem, ‘De la responsabilité personnelle à la répartition des risques’ Arch phil dr 1977, 5. 100 Carval (n 81) [1]–[14], esp [7], [14]. 101 ibid [1]–[14]; Jourdain (n 88); M Fabre-Magnan, Droit des obligations. 2, Responsabilité civile et quasi-contrats, 3rd edn (Paris, PUF, 2015) 53–62, 424. 102 Viney, Jourdain and Carval (n 88) [117]. 103 See eg C Coutant-Lapalus, Le principe de la réparation intégrale en droit privée (Paris, PUF, 2002) [79]–[86]; P Kayser, ‘La Justice selon Aristote’ RRJ 1996, [11]; Jourdain (n 88) [3]; P Malaurie, L Aynès and P Stoffel-Munck, Droit des obligations, 8th edn (Paris, LGDJ, 2016) [238]. 104 By contrast, it is difficult to reconcile structural correlativity with the way in which most French legal actors currently understand tort law. As seen above (text to nn 99–100), tort liability is often imposed

Comparative Reflections on Punishment in Tort Law   349 the victim of a tort must receive a sum that equals the loss suffered, nothing less (tout le dommage), but also nothing more (rien que le dommage).105 Thus, it appears that, cloaked behind the principle of full compensation, content-related correlativity informs the remedial dimension of French tort law and impacts on the type of punitive measures that may be adopted in French law. Indeed, the principle of full compensation is probably the most significant reason why the amende civile has attracted wider support than punitive damages as the most appropriate response to torts committed with a view to profit. Punitive damages, the solution envisaged in article 1371 of the Avant-projet Catala, are deemed inadmissible because they accrue, in whole or in part, to the claimant and thus result in what is often described as an undeserved windfall.106 This would constitute a patent violation of the principle of full compensation. To avoid this, it is often suggested that the punitive award should be allocated in its entirety to third parties and that, for this reason, the amende civile is to be preferred over punitive damages.107 Article 1266-1 of the Projet de réforme is consistent with this approach because, as seen above, it provides that the amende civile must be paid to compensation funds related to the nature of the harm suffered or, if unavailable, to the Public Treasury. By doing so, the amende civile testifies to the continuing vitality of the principle of full compensation or, in other words, it perpetuates content-related correlativity as deriving from the theory of corrective justice. To conclude, the way French legal actors treat the idea of punishing civil wrongs reflects a profoundly instrumental conception of punishment, which is fully consistent with their theorising about the functions of tort law. The French thinking on the role of punishment in tort law as well as the work of the drafters of the Projet de réforme highlight a willingness to deter certain types of intentional wrongdoing. Nevertheless, the principle of full compensation, which may be best seen as a legacy of the influential theory of corrective justice, reduces the options available when devising a punitive response against tortfeasors. In this respect, ideas of interpersonal justice still exert considerable influence on the tort system and affect the prospect of reforming the law.

not on the basis of backward-looking considerations pertaining to the interaction between defendant and claimant, but rather on the basis of broader, forward-looking justifications such as, most notably, loss-spreading. 105 See eg Cass civ (1) 30 May 1995, JCP 1995 IV 1810; Cass civ (3) 8 July 2009, D 2009, 2036, note Y Rouquet. The principle of full compensation has been increasingly challenged in recent times; for an overview of the critiques levelled against it, see Viney, Jourdain, and Carval (n 88) [118]. 106 Y Lambert-Faivre (n 90) 18–19; Fasquelle (n 88) 34; R Saint-Esteben (n 81) 58; D Fasquelle, ‘La sanction de la concurrence déloyale et du parasitisme économique et le Rapport Catala’ D 2005, 2666; E Dreyer, ‘La faute lucrative des médias, prétexte à une réflexion sur la peine privée’ JCP G 2008, I, 22, 24–25; P Brun, Responsabilité civile extracontractuelle, 4th edn (Paris, LexisNexis, 2016) [12]–[14]. See also J-S Borghetti, ‘L’avant-projet de réforme de la responsabilité civile. Vue d’ensemble de l’avantprojet’ D 2016, 1386, [21]. 107 See eg Jourdain (n 88) [11]; Behar-Touchais (n 90) [15]–[23].

350  Marco Cappelletti

V. Conclusions The treatment of the idea of punishing civil wrongs in the United States, England, and France varies depending on the conception of punishment that legal actors within each jurisdiction embrace and on how this conception relates to broader ways of seeing and theorising tort law. In the United States, the debate as to whether tort law should punish is characterised by a strong clash of views on the most appropriate conception of punishment and of punitive damages. This conflict is part of a broader contrast between instrumentalist and non-instrumentalist approaches in tort theory. Instrumentalism, still predominant today, sees punitive damages as a means to societal ends, whereas non-instrumentalism considers punitive damages a form of interpersonal justice that allows the victim of a reprehensible wrongdoing to be punitive against the tortfeasor. Preference for the instrumental or interpersonal conception of punitive damages generally reflects a wider view about the role and purposes of tort law. Given that punitive damages can be reconciled with both instrumentalism and non-instrumentalism, it is not surprising that discussions on punitive damages often focus on the best way of interpreting this remedy rather than on assessing its desirability. The treatment reserved to the idea of punishing civil wrongs is rather different in both England and France. In English law there seems to be a disconnection between the way in which most legal actors view the tort system and the dominant conception of punishment. On the one hand, tort law is widely seen as a system of interpersonal justice, in which tort obligations must be justified with reference to the relationship between claimant and defendant. In this context, it is somewhat surprising to see that very few attempts have been made to support an interpersonal conception of exemplary damages which could cohere with the predominant understanding of tort law and that such attempts have so far failed to attract meaningful support. Indeed, and on the other hand, punishment is mostly understood instrumentally, and exemplary damages are depicted as a device that, at times, may usefully attain societally desirable goals such as teaching wrongdoers that tort does not pay. Clashing with an interpersonal vision of English law, exemplary damages are treated as anomalous or as requiring careful confinement and instrumental justifications of liability such as deterrence or retribution struggle to establish themselves as fully legitimate and respectable goals of tort law. The view of French legal actors relating to the role of punishment in tort law differs again. The French approach to the issue of whether tort law should be equipped with a general punitive measure and how this measure should work reflects a markedly instrumental conception of punishment. Resembling the English position, punishment is overwhelmingly seen as serving societal goals. This approach is not surprising, considering that, similarly to what was observed in relation to the United States, tort law is conceived of in a markedly instrumental way (even though for very different purposes). Coherently with this picture, broader theories of interpersonal justice are largely absent from the debate on the nature and functions of French tort law. However, there

Comparative Reflections on Punishment in Tort Law   351 is an important qualification here, for a potent legacy of the theory of corrective justice – the principle of full compensation – still plays a key role in the French law of damages, and has the effect of limiting in a substantial way the options available when crafting tort ­punitive measures. As a result, French law naturally shuns the Anglo-American model of punitive (or exemplary) damages. Since the role of punishment and the characteristics of punitive measures depend to a large extent on the way of seeing tort law and its functions, the three legal systems discussed in this chapter are likely to keep treating the idea of punishing civil wrongs in markedly different ways and, should article 1266-1 of the Projet de réforme become part of the Code civil, they would move further apart.

352

17 Unjustified Enrichment and Civil Liability MÉLODIE COMBOT

At first glance, there is nothing in the Projet de réforme de la responsabilité civile which invites an analysis of the relationship between unjustified (or unjust) enrichment and civil liability.1 Indeed, the Projet does not contain any reference to unjustified enrichment or, to put it in a more French way, quasi-contracts.2 By contrast, the drafters of the reform project have devoted substantial effort to reframing the relationship between contractual and extra-contractual liability.3 This imbalance is not surprising, and is reflected in the work of legal scholars: while there is an abundant literature that considers the relationship between contractual and ‘tortious’ liability, there is little work on the relationship between unjustified enrichment and civil liability. Even more unusual is trying to compare the French law of unjustified enrichment with the law of civil liability in another legal system, and this is exactly what this chapter aims to do by comparing French and English law (with occasional references made to German law too). Unlike the notion of contract, which is understood in a broadly similar way across different European jurisdictions, civil liability and unjustified enrichment may vary considerably in their scope and meaning. Although it is quite usual to distinguish the civil law model of unjustified enrichment from the common

1 The terminology may vary between English and French law. While French law is more inclined to use ‘unjustified enrichment’ rather than ‘unjust enrichment’, English law is not particularly familiar with the term ‘civil liability’. At this stage, ‘unjust enrichment’ and ‘unjustified enrichment’ are deemed equivalent and they cover in both jurisdictions cases which are neither contractual nor ’tortious‘ and which are explained by the idea of reversing an enrichment to the claimant. I will use ‘civil liability’ (in French responsabilité civile) to encompass contractual and extra-contractual liability (responsabilité contractuelle and extracontractuelle) and in the English context to encompass both contractual liability (ie liability for breach of contract) and liability in tort. 2 The 2016 reform of the Code civil, concerning contract law, did not remove this Roman category, which is still seen by a large majority of scholars as underpinned by the principle of reversing unjustified enrichment. 3 See ch 2, ch 3 and ch 4.

354  Mélodie Combot law model of unjust enrichment,4 not too much emphasis should be placed on this distinction between civil law and common law. As this chapter will seek to show, the French law of unjustified enrichment presents some real peculiarities compared to other systems, especially English and German law. Indeed, so many differences exist between the French and German laws of unjustified enrichment that the latter seems closer to English law than to French law.5 This is largely due to the broad scope of French civil liability6 and to the inclusion of the concept of unjustified enrichment within the category of quasi-contracts. The combination of these two factors brings unjustified enrichment so conceptually close to civil liability that it leads one to wonder whether there is any separate French law of unjustified enrichment at all. It is also worth noting that in French law the term enrichissement injustifié possesses two distinct meanings: first, in a broader sense it can refer to the general principle underpinning every quasi-contract, this being the idea that ‘no one shall enrich himself unjustly at the expense of another’7 or, secondly, in a narrower sense, it can be used more restrictively to refer only to one of the three quasi-contracts, ie ‘the action to reverse an unjustified enrichment’ (enrichissement injustifié).8 At this stage, it is helpful to present briefly the framework of the French law of unjustified enrichment, which is actually quite complex. Unlike English and German law, and despite there being a general action, there is no unified concept of unjustified enrichment but rather a collection of actions which are classified within the category of quasi-contract. The current state of the law is the consequence of the initial structure of the Code civil, which recognised only two quasi-contracts in 1804: negotiorum gestio (gestion d’affaires or the benevolent intervention in another person’s affairs) and recovery of undue payments (répétition de l’indu). At first, the lack of a general disposition to reverse an unjustified enrichment led the French courts to resort to negotiorum gestio in order to grant an action to a claimant who performed a service for the defendant but without any

4 The term ‘unjustified’ is thus related to the absence of basis approach, which means that it is prima facie sufficient to establish that there is not (or is no longer) an obligation which provides a foundation for the transfer that enriched the defendant. By contrast, the term ‘unjust’ refers to the unjust factors approach. In this approach the claimant must establish a positive ground for such mistake, duress, or failure of consideration if she wishes to obtain any restitution; ie the mere fact that there is no justification for the original transfer is in itself insufficient. 5 In this sense, G Dannemann, The German Law of Unjustified Enrichment and Restitution, A Comparative Introduction (Oxford, Oxford University Press, 2009) 162ff. 6 See C Von Bar, The Common European Law of Torts. The Core Areas of Tort Law, its Approximation in Europe, and its Accommodation in the Legal System (Oxford, Clarendon Press, 1998) vol 1 [520], where the author explains that the narrow scope of tort law in Germany is one of the reasons for the larger scope of unjust enrichment. 7 The Code civil lists three quasi-contracts: negotiorum gestio (benevolent intervention in another’s affairs, arts 1301 to 1301-5 Cc), répétition de l’indu (arts 1302 to 1302-3 Cc) and enrichissement injustifié (arts 1303 to 1303-4 Cc). However, the category of quasi-contracts is not limited to these three instances: see the Report to the President of the Republic on Ordonnance of 2016 on Sub-title 3 Other sources of obligations. 8 Arts 1303 to 1303-4 Cc.

Unjustified Enrichment and Civil Liability  355 altruistic intention, which was referred to as ‘abnormal’ negotiorum gestio.9 This situation was quite unsatisfactory, and the Cour de cassation eventually recognised a general action, then called action de in rem verso or enrichissement sans cause, in the famous arrêt Boudier in 1892.10 The action d’enrichissement sans cause was then seen as the embodiment of the general idea that ‘no one should enrich himself at the expense of another’. As répétition de l’indu was also seen as a specific kind of the same idea, many scholars concluded that the two original quasi-contracts were two specific kinds of enrichissement sans cause, so that every quasi-contract was finally founded upon the principle of reversing an unjustified enrichment.11 The 2016 reform did not change this structure, its only effect being to codify the general action, and to change its name from enrichissement sans cause to enrichissement injustifié.12 Although this portrait of French unjustified enrichment is not uncontroversial in the light of the case law, it will be assumed – as a starting point – that all quasi-contracts aim to reverse an unjustified enrichment. As a consequence, my analysis will be limited to the three quasi-contracts regulated by the Code civil, ie enrichissement injustifié, negotiorum gestio, and répétition de l’indu. It is worth emphasising that despite the common affirmation that every quasi-contract is founded upon the principle of unjustified enrichment, this statement has no practical consequence and there are almost no connections between the three quasi-contracts; there is no common provision currently included in the Code civil governing all three. The only consequence is that enrichissement injustifé has been added alongside the two original quasi-contracts, negotiorum gestio and répétition de l’indu, without giving any consideration to the question whether all these actions form a coherent whole. Another problematic aspect of French law – compared to other jurisdictions, especially English law – is the unclear connection between the new rules governing les restitutions13 and unjustified enrichment. Apart from the cross-reference to the quasi-contract répétition de l’indu,14 which is one specific ground of restitutions among others, restitutions are generally not based on the principle of unjustified enrichment and are seen as multi-causal, that is, they rest on a number of distinct grounds. Thus, for example, even where a contract is terminated r­etroactively 9 This explains why negotiorum gestio is seen in French law as a particular case of unjustified ­enrichment, the recognition of the action d’enrichissement sans cause has just reversed the situation, which means that the enrichissement sans cause is no longer a special case of negotiorum gestio (‘abnormal’ negotiorum gestio), but the general action. Reciprocally, negotiorum gestio became seen as a specific kind of enrichissement sans cause. 10 Req 15 June 1892, DP 92.1.596, S 93.1.281 note Labbé. 11 See notably J Carbonnier, Droit civil, Les biens et les obligations, 1st edn (Paris, PUF, 1957) vol 2 [198], 693, and M Douchy, La notion de quasi-contrat en droit positif français (Aix-Marseille, Economica, 1997). 12 This relabelling is explained by the removal of the concept of cause from the Code civil by the 2016 reform. 13 Arts 1352 to 1352-9 Cc. 14 See art 1302-3 Cc concerning undue payment: ‘Restitution is subject to the rules set by articles 1352 to 1352-9’.

356  Mélodie Combot for non-performance, consequential restitutions are usually seen purely as a consequence of the termination, without any link with unjustified enrichment. Thus, in order to avoid any confusion, in this chapter I will use the term ‘unjustified enrichment’ to refer to the source of obligation founded on the general principle that ‘no one shall enrich himself unjustly at the expense of another’. I will keep the French terminology of enrichissement injustifié for the particular quasi-contract of the same name, and by contrast, in the context of English law I will use the term ‘unjust enrichment’ to describe the body of law which requires the reversal of an enrichment made at the expense of the claimant, as this is the terminological choice adopted by the vast majority of legal scholars.15 Because of these differences in scope and meaning concerning unjustified enrichment, it is first necessary to set the scene and present the main differences and characteristics of French unjustified enrichment law, compared to English and German law. It will be then possible to analyse and question the independence of unjustified enrichment from civil liability and the difficulties of finding the proper borderline between the two.

I.  The Main Characteristics of Unjustified Enrichment in French Law Even if French law recognised a general action founded upon the general principle of reversing an unjustified enrichment quite a long time ago,16 this area of law remains at an embryonic stage. It is particularly striking to see that there is not a single book pertaining to unjustified enrichment in French law, whereas there are numerous monographs or treatises on this subject in England. Unjustified enrichment in French law is not really seen as a genuine source of obligation, unlike contract or delict, because its goal is only to fill the gap where neither of the two main sources can be used. Besides, the relationship of the concept of unjustified enrichment to the quasi-contract enrichissement injustifié is far from clear17 and the new provisions of the Code civil are not very helpful. Although the name of this action – enrichissement injustifié – encourages us to understand it as a general action, the Code treats it as one of several quasi-contracts and does not provide any general rule to be applied to every quasi-contract. This action has moreover a very marginal scope. This is explained by the fear of some scholars and judges that this action might be a tool which can ‘blow up’ the entire structure of the law,18 15 See A Burrows (ed), A Restatement of the English Law of Unjust Enrichment (Oxford, Oxford University Press, 2012); C Mitchell, P Mitchell, S Watterson (eds), Goff & Jones: The Law of Unjust Enrichment, 9th edn, (London, Sweet & Maxwell, 2016); P Birks, Unjust Enrichment, 2nd edn, (Oxford, Clarendon, 2005). 16 In the famous arrêt Boudier (n 10) in 1892, as mentioned above, p 355. 17 For further analysis see E Descheemaeker, ‘Quasi-contrats et enrichissement injustifié en droit français’ RTD civ 2013, 1. 18 The expression is from J Flour, ‘Pot-pourri autour d’un arrêt’, Défrenois, 1975, n°30854, p 186.

Unjustified Enrichment and Civil Liability  357 because of the very vague meaning of enrichissement injustifié which may encourage judges to reach a decision according to considerations of fairness. This action is thus granted cautiously by the courts, and a specific requirement was framed to prevent the opening of the floodgates to a deluge of claims: the idea that the action for enrichissement injustifié is ‘subsidiary’ in the sense that it can be brought only in the absence of another basis of claiming. This is now expressed in article 1303-3 Cc, which states that: The impoverished person has no action on this basis [ie enrichissement injustifié] where another action is open to him or is legally barred, as in the case of prescription.

The general idea behind the subsidiarity requirement is that the action should not be granted if the result of its application conflicts with the effects of another rule, as can be seen from article 1303-3 itself, which gives the example of prescription. If the other action available is barred by prescription, the claimant cannot circumvent it and choose instead to claim on the basis of the action of enrichissement injustifié. In practice, enrichissement injustifié covers two types of cases: indirect enrichment and service cases where the claimant never intended to act for the benefit of the defendant. More specifically, this action is used notably in situations involving cohabitational property disputes and cases where the creditor wants to circumvent his debtor’s insolvency by suing a third party who ultimately benefits from the creditor’s initial action, as was the case in the arrêt Boudier itself.19 As some commentators have observed, these situations are not really about the principle of unjustified enrichment, but the action which bears this name is used as an instrument to achieve particular goals: the proportional distribution of assets in cohabitational property disputes,20 and a guarantee for the creditor against his debtor’s insolvency in indirect enrichment claims. The two other quasi-contracts, negotiorum gestio and répétition de l’indu, do not attract much attention from French scholars. Their practical areas of application are also quite small, even if they are more frequently applied than enrichissement injustifié. Negotiorum gestio can only apply where the intervention of the agent is useful to the principal and where the agent acts with an altruistic intention, whereas répétition de l’indu is mostly used in cases where the claimant transfers mistakenly goods or money to the defendant.21 In addition, this area of law is surrounded by the two main sources of obligations, contract and extra-contractual liability, which are understood in French law in a very comprehensive way, especially in the case of the latter. 19 A merchant (Boudier) sells manure to a tenant, who is later revealed as insolvent and unable to pay his rent and the price of the manure. The lease is terminated and, as a settlement, the tenant leaves the crops to the landlord. Boudier then decides to sue the landlord, because he has benefited from the initial supply of the manure. 20 See F Chénedé, ‘Pour un affinement de la théorie des quasi-contrats au service de la liquidation patrimoniale du concubinage’ D 2010, 718. 21 This remains true even though establishing a mistake is no longer a condition of répétition de l’indu (art 1302 Cc), except in the case of the payment of another’s debt where it is necessary to establish that the transfer was done mistakenly or with duress (art 1302-2 Cc).

358  Mélodie Combot It is worth noticing that neither the German nor the English legal systems possess a general ‘clause’ or legal ground of extra-contractual liability, whereas French law has two general provisions, namely liability for fault22 and liability for the deeds of things.23 This difference impacts on the scope of unjustified enrichment, which is much narrower in French law as compared to English or German law. However, the extensive scope of the two main sources of obligations is not the only explanation for the small ambit of unjustified enrichment. One of the main substantial differences of the French model compared to the English and German models is the importance of loss, since it is now a common assertion that every quasi-contract tends to compensate the claimant’s loss rather than reversing the defendant’s gain.24 This preponderance of loss is largely due to the place of negotiorum gestio within the French unjustified enrichment/quasi-contract category and its role in the emergence of the general action, given that, as earlier noted, enrichissement sans cause was first seen as an ‘abnormal’ negotiorum gestio.25 In fact, it is highly questionable to analyse negotiorum gestio as a particular example of unjustified enrichment, since the agent can only recover the reimbursement of his expenses and reparation of loss sustained during the performance of his activity, and does not have any right to the actual profit made by the principal.26 In contrast, in German law negotiorum gestio is treated separately from unjustified enrichment and even if there are links between these two bases of recovery, the former is not seen as a special kind of unjustified enrichment but as a different source of obligation. ­Actually, it is possible to affirm that enrichissement injustifié remains a specific case of negotiorum gestio, as it was before the arrêt Boudier. Despite its name, this action does not intend to reverse an unjustified enrichment, but to compensate a loss. However, as the claimant never intended to be useful to the defendant, he deserves less protection than the agent, and thus cannot claim full compensation for his loss, the indemnity being limited to the extent of the defendant’s enrichment.27

II.  The Distinction between Civil Liability and Unjustified Enrichment: Is Unjustified Enrichment an Independent Category? After this brief analysis of the French law of unjustified enrichment, it is possible to question the appropriateness of this category and to ask whether it should



22 Art

1240 Cc; art 1241 of the Projet. 1242(1) Cc; art 1243 of the Projet. 24 See notably F Chénedé, ‘Charles Toullier, le quasi-contrat’ RDC 2011, no 1, 3005ff. 25 See above, pp 354–355. 26 Art 1301-2(2) Cc. 27 This statement, however, is very theoretical: see below pp 364–365. 23 Art

Unjustified Enrichment and Civil Liability  359 perhaps be incorporated into the category of civil liability. Indeed, if we look at article 1258 of the Projet de réforme, which sets out a first principle governing the effects of liability, it is sufficiently wide to include in the scope of civil liability most of the cases where the goal is to reverse a gain. It defines the aim of reparation as being ‘to replace the victim as much as is possible in the situation in which he would have been if the harmful action had not taken place’.28 It depends, of course, on what is meant by ‘harmful action’, but if the defendant makes a profit from his harmful action and if this profit belonged or should have belonged to the claimant, it may be argued that the ‘correct’ reparation would be to reverse the gain – corresponding to a loss of profit for the claimant – and thus place the claimant in the position in which he would have been but for the harmful action. In this way, civil liability could include all claims which seek to restore the status quo ante. This position, which favours a very inclusive approach to extra-contractual liability, has found support among scholars in both France and England and is clearly based on very broad notions of loss and compensation.29 It is true that, in some way, in cases of unjustified enrichment the claimant sustains a loss, since he has been deprived of a gain, and that it may make sense to speak in terms of compensating that loss. However, such conceptions of loss and compensation would be too broad to be of any use and, unsurprisingly, a solution of this kind has never been adopted in either French or English law. Presently, each legal system distinguishes between civil liability and unjustified enrichment as distinct sources of obligation, essentially because of the irreducible special features of the latter. And even where some of the conditions of civil liability apply to unjustified enrichment, a closer examination of the case law reveals that these conditions are often applied differently and that their use is often purely instrumental. It is, however, sometimes argued, especially in French law, that rules of tortious liability are broad enough that they can apply to all extra-contractual matters including unjustified enrichment and, more extensively, quasi-contracts – at least where no special rules exist.30 This is particularly noticeable with the use of two notions transplanted from civil l­ iability: préjudice, through the notion of loss or impoverishment,31 and fault. Their use varies in degree across the two jurisdictions under consideration and their meanings and roles diverge from those of their civil liability equivalents. To these I now turn.

28 Translation into English is provided in the Appendix of this book. 29 For French law, see eg M Planiol, ‘Classification des sources des obligations’ RCLJ 1904, 220, who proposed to see in every quasi-contract a quasi-delict. For an account of these theories in English law, see J Edelman, ‘Gain-Based Damages and Compensation’ in A Burrows and Lord Rodger of Earlsferry (eds), Mapping the Law, Essays in Memory of Peter Birks (Oxford, Oxford University Press, 2006) 141ff. 30 G Viney, Traité de droit civil, Introduction à la responsabilité, 3rd edn (Paris, LGDJ, 2008) [191], 528. 31 This does not mean that loss or impoverishment (more commonly used in French law) are synonymous with préjudice: while a loss in unjustified enrichment can be considered a préjudice according to the rules of civil liability, the converse is not true, for not all types of compensable préjudice in civil liability can be considered losses according to the rules of unjustified enrichment.

360  Mélodie Combot

A.  The Role of Loss/Impoverishment (i)  Different Positions When it comes to unjustified enrichment, the condition of loss is far from ­obvious. The notion of loss is the subject-matter of debate and controversy not only in French law but also in English law and, for a while, even in German law (where it was eventually abandoned).32 Notwithstanding the attention devoted to the notion of loss in these three legal systems, it is possible to list at least three distinct views on its relevance in the law of unjustified enrichment. First, on one view, loss should not be a requirement of unjustified enrichment, as it would be confusing and it would not represent a relevant consideration to determine if the enrichment was attributable to the claimant. This is the current position of German law and it is a solution that has found support in some English judicial opinions as well as in Burrow’s Restatement of the English Law of Unjust Enrichment.33 On a second view, the claimant must establish a loss, but such a loss should not be taken into account in assessing the quantum of the restitutionary ­recovery, which is only limited by the extent of the defendant’s enrichment. For some ­scholars, this is the current position of English law,34 an interpretation that seems to have been confirmed by a very recent decision of the UK Supreme Court.35 It is also possible to understand in this way répétition de l’indu in French law, for while the claimant’s ‘losses’ are the goods or money he transferred to the defendant, restitution is not limited to them (ie to what was lost), and can include interest, the value of use or of enjoyment, and all the ‘fruits’ of the thing transferred (article 1352-3 Cc). A third view maintains that the claimant’s loss does not merely represent the initial connection between the claimant and the enrichment, but also constitutes a relevant consideration in determining the measure of the restitutionary recovery. Seen in this way, loss is as important as the enrichment condition, if not more so. This is the current position of the French action based on enrichissement injustifié, where both loss and enrichment provide a cap on recovery, the ‘double-cap’ rule.36

32 G Ripert and M Teisseire, ‘Essai d’une théorie de l’enrichissement sans cause’ RTD civ 1904, 727ff; T Krebs, ‘The Fallacy of “Restitution for Wrongs”’ in Burrows and Lord Rodger of Earlsferry (n 29) 380–81. 33 Kleinwort Benson Ltd v Birmingham City Council [1996] 3 WLR 1139, [1997] QB 380; Sempra Metals Ltd v IRC [2007] UKHL 34, [2008] 1 AC 561; Burrows, A Restatement of the English Law of Unjust Enrichment (n 15) 44–48 and see A Burrows, The Law of Restitution, 3rd edn (Oxford, Oxford University Press, 2011) 45. 34 E Ball, Enrichment at the Claimant’s Expense, Attribution Rules in Unjust Enrichment (Oxford, Hart Publishing, 2016) 78ff. 35 Investment Trust Companies v Revenue and Customs Commissioners [2017] UKSC 29, [2018] AC 275. 36 Art 1303 Cc, according to which ‘[a]part from the situation of management of another’s affairs and undue payment, a person who benefits from an unjustified enrichment to the detriment of another

Unjustified Enrichment and Civil Liability  361 This diversity of approaches as to the relevance of loss in the context of claims for unjustified enrichment demonstrates a first difference with its civil liability counterpart, préjudice. Whereas préjudice is one of the core conditions of civil liability, loss is not always required in unjustified enrichment and even when it is, it is not necessarily a measure of restitution. This flexibility as regards the role of loss reveals that loss is not an indispensable condition and that it is often a sign of the deficiencies of the law of unjustified enrichment, because loss undermines its coherence and prevents unjustified enrichment from fulfilling its real goal, ie the restitution of a benefit. Legal systems where the study of unjustified enrichment is more advanced often minimise or even abandon this condition, which often deprives this category of all its potential and blurs considerably the borderline with civil liability. By contrast, where the law of unjustified enrichment is less structured, for example by not being subject to stringent conditions, the courts frequently resort to the notion of loss to narrow down its scope.

(ii) Different Meanings The way loss is understood in unjust(ified) enrichment claims depends on the generality of the action. When the action is too general, meaning that it could in principle encompass a very broad range of situations, the loss condition will be used to prevent the opening of the floodgates. Therefore, while the courts may be using the language of loss, the real focus is not so much on the actual loss suffered by the claimant but rather on the consequences that restitution would have in the particular situation. A comparison between English and French law is instructive on this point and it reveals that loss is alien to the logic of unjustified enrichment. It has earlier been emphasised that one reason explaining the importance of loss in the French law of unjustified enrichment is that enrichissement injustifié was built around negotiorum gestio. However, this is not the only reason. If loss is sometimes praised as a requirement for claims based on unjustified enrichment, it is mainly because of its supposed ability to confine their scope, especially in two contexts: indirect enrichment and incidental benefit. The former context involves a situation where the defendant received the benefit from a third party rather than directly from the claimant.37 The latter refers to subsequent benefits which may arise from the claimant’s action and may benefit other persons. For instance, the renovation of a house in a poor state of repair will certainly improve the value of the owner’s land, and it may also improve the value of that of his neighbours. This is why loss does not cover the same thing in every case, and depends on the initial scope of the action and other possible safeguards. person must indemnify the person who is thereby made the poorer to an amount equal to the lesser of the two values of the enrichment and the impoverishment’. Such an approach is also praised by some English scholars. See C Mitchell, P Mitchell, S Watterson (n 15) [6-106]ff. 37 For an example, see the facts of the arrêt Boudier (n 10).

362  Mélodie Combot In the English law of unjust enrichment, the notion of loss seems to be ­embodied in the condition that enrichment is made ‘at the expense’ of the claimant, often interpreted as requiring that the claimant suffered a loss.38 Unlike French law, loss does not enjoy the same status as the notion of enrichment, appearing therefore to be quite secondary. Furthermore, in one recent UK Supreme Court decision,39 although Lord Reed referred to loss he immediately added that it was not to be understood in the same sense as in the law of damages.40 This type of statement is extremely rare in French law, where the difference between loss/impoverishment in the action of unjustified enrichment and préjudice in civil liability is not emphasised. In English law, this condition serves only to highlight the need of a subtraction from the claimant for two main purposes: to deny in principle any claim against a third party who received the benefit from a person other than the claimant and to prevent any restitution when the benefit is incidental.41 But the very word ‘loss’ is highly ambiguous and may falsely suggest that a correspondence between loss and gain is required,42 which is not the case in English law. Moreover, for indirect enrichment, loss can be a useful limit only if the causation between loss and enrichment is understood in a narrow way, which means that loss in itself is insufficient to deal with this particular set of cases and explains why English lawyers develop special rules on this subject. As for the incidental benefits cases, as Lord Reed recognised, useful limits are already provided by the requirement of an unjust factor, which will prevent the claimant from claiming restitution of that gain. In the house renovation example,43 the claimant cannot, for instance, demonstrate that he made a mistake, or that he acted under the pressure of his neighbours. In addition, according to the way in which loss is understood, it is often possible to find a loss, notably if we consider that the claimant provides a service for his neighbours and is thus deprived of an opportunity to gain a remuneration for that service. In French law, the notion of loss varies according to the type of quasi-contract involved and the need to narrow down its scope. There is no particular problem with répétition de l’indu, since its legislative provisions themselves require a direct transfer from the solvens to the accipiens,44 and because, until recently, this type of claim was not used where the claimant had provided a service.45 If this action is to be used in services cases, particularly in situations of incidental benefits, it 38 See notably C Mitchell, P Mitchell, S Watterson, (n 15) [6-106]ff. 39 Investment Trust Companies (n 35). A customer complained about a mistaken overpayment of VAT, but he brought his action directly against the Revenue instead of suing his supplier. 40 ibid [45]. 41 These reasons are expressly stated by Lord Reed in Investment Trust Companies (n 35) at [51] and [52]. 42 A Burrows, ‘“At the Expense of the Claimant”: A Fresh Look’ (2018) 25 Restitution Law Review. 43 Above, p 361. 44 See art 1302-1 Cc: ‘A person who receives by mistake or knowingly something which is not owed to him must restore it to the person from whom he unduly received it’. 45 The new provisions pertaining to restitution and applicable to répétition de l’indu expressly accept the possibility of restitution for services (art 1352-8 Cc).

Unjustified Enrichment and Civil Liability  363 will certainly be useful to find another restriction. But this does not mean that it is necessary to require a loss on the part of the claimant; it would be more appropriate to refine the notion of enrichment, which is currently too broad, although surprisingly few have proposed this solution.46 It is also interesting to note that this condition is not mentioned in the provisions of the Code civil because it is totally useless and redundant since restitution focuses on what the accipiens received, even if it is possible to say that the solvens loses something. By contrast, the so-called actio contraria of negotiorum gestio is loss-based,47 since the possible enrichment of the principal is not taken into account in assessing the indemnity owed to the agent.48 The conditions of the usefulness of the agent’s action and his altruistic intention are thus required to select losses eligible to compensation. The requirement of an altruistic intention explains the narrower scope of loss, as compared to préjudice in civil liability, because all losses consisting of missed opportunities to profit are irrecoverable in negotiorum gestio (which is consistent with the rule that the agent cannot claim any remuneration).49 This point has been intensely discussed in connection with situations where a professional intervenes in another’s affairs, for instance a doctor who rescues an injured person. If the doctor uses some of his own materials, for example, pharmaceuticals, he can surely ask for a reimbursement, but he cannot claim any fee. The law allows him to avoid being worse-off but, unlike civil liability, the rule is not ‘full compensation’, and this means that only the loss suffered because of the ­performance of his activity can be compensated.50 Nevertheless, the effect of ­negotiorum gestio, at least in relation to the agent’s action against the principal, is close to a civil liability claim, as both are loss-based. This is not, a priori, problematic since negotiorum gestio has a limited function, that is, protecting a benevolent intervener and, as a result, there is less overlap with civil liability.

(iii)  The Particular Case of the Action for Enrichissement Injustifié: The Unproductive Restriction of Impoverishment This action deserves a specific treatment because it best embodies the ‘fallacy’ of impoverishment/loss, the bad reasons justifying this requirement, and its 46 In French law, it seems that no one supported this solution, while in English law some have considered doing so: see notably J Beatson, The Use and Abuse of Unjust Enrichment (Oxford, Clarendon Press, 1991) 29ff; C Webb, Reason and Restitution (Oxford, Oxford University Press, 2016) 99ff. 47 This is the action that the agent can bring against the principal for the recovery of expenses or losses sustained during his activity (art 1301-2(2) Cc). 48 Remarkably, the Code civil refers to an indemnity rather than restitution (art 1301-2 Cc), which is also the case for the action of unjustified enrichment (art 1303-4 Cc), but not for undue payment where restitutions are granted to the claimant (art 1302 Cc). 49 For exceptions, see R Bout, ‘Quasi-contrats. – Gestion d’affaires. – Effets’, JCl Civil Code, arts 1372–1375, 2012, [60]ff. 50 The agent should not only be deprived of the right to claim a fee for his intervention, but he should also be unable to obtain compensation for any opportunity he missed because of his intervention. For example, if doctor A missed an operation planned well in advance because he was rescuing the

364  Mélodie Combot serious effect on the concept of unjustified enrichment and on its relationship with civil liability. For the same reason as mentioned above, impoverishment appears to be a useless restriction in indirect enrichment cases and incidental benefits cases, where other restricting conditions are necessary. As a result, in the case of ­enrichissement injustifié, the principal reason to reject a claim for unjustified enrichment is the presence of a justification for the defendant’s enrichment. In indirect enrichment cases, the main justification is almost always found in a contract, usually concluded between the first person who receives the profit from the claimant and the defendant. As for incidental benefits cases, French judges have elaborated a special justification, namely the claimant’s personal interest.51 Where the claimant’s conduct is solely justified by his own personal interest (most notably a desire to profit), he cannot ask for the restitution of any subsequent profit arising from that conduct which benefitted third parties. However, impoverishment becomes a real restriction when it comes to the assessment of the indemnity, especially where the claimant provides a service to the defendant leading to an end-product.52 In these circumstances, French law is particularly reluctant to accept that the claimant has a right over the end-product, which would be the case if enrichment were the only value considered. But in order to bar or to restrict such a claim, the notion of impoverishment must be interpreted in a restrictive way, and cannot encompass every loss with a monetary value. Otherwise it would be possible to argue that where the claimant improves his neighbour’s land involuntarily, he is correspondingly impoverished by not getting the monetary value of the improvements. The adoption of such a broad notion of impoverishment would not be very surprising, when compared to the notion of préjudice in civil liability, which covers a loss of future profit53 and a loss of chance to make a profit.54 But if impoverishment were understood in such a broad way, it would be of little use because it would be possible to find an impoverishment every time that the claimant’s conduct resulted in the defendant’s enrichment and to conclude that the claimant ‘lost’ this enrichment. This is why the notion of impoverishment is interpreted narrowly in French law; it encompasses only two kinds of loss, ie loss of an asset55 and loss of an opportunity to be remunerated. principal, and the disappointed patient decided to see doctor B, doctor A cannot claim the money he would have earned if he had performed the surgery as initially planned. 51 Now art 1303-2 Cc. 52 The typical example is that of a claimant who finances the construction of a building on another’s land or improves that land, but without using his own assets (otherwise there would be accessio, governed by art 555 Cc). 53 P le Tourneau (ed), Droit de la responsabilité et des contrats, 11th edn (Paris, Dalloz, 2018/2019), [2124.42]. 54 For example, judges sometimes award compensation for a loss of a chance to win a prize at a horse race, when the horse is unable to compete due to the fault of a third party (see P le Tourneau (n 53), [2123.91]). 55 This can include not only the loss of the asset itself, but also the loss of value of the asset because of depreciation since the receipt of the asset by the defendant (see below for further details).

Unjustified Enrichment and Civil Liability  365 Hence, when the enrichment received by the defendant is superior to the value of the asset transferred or to the value of the service provided by the claimant, the impoverishment will cap the indemnity. Put another way, the claimant has only a right to the very thing he transferred: the money, the asset, or the service. This narrow interpretation of impoverishment, together with a broad understanding of enrichment, leads to the paradoxical outcome that enrichissement injustifié is actually more loss-based than gain-based. This is clearly shown by the ‘pure service’ cases, ie cases where the service does not result in an end-product. In these cases, which are at the core of enrichissement injustifié, it is striking that the enrichment is deemed to be equivalent to the claimant’s impoverishment, ie to the remuneration that the claimant would have received for the service.56 As long as this enrichment was received without justification and subject to the condition of subsidiarity in article 1303-3 Cc, the defendant cannot deny that he was enriched even if he proves that the service provided was worthless to him or that he could have performed it himself at a lower price. Thus, only the lost opportunity that the claimant suffered will matter, with no attention being paid to the defendant’s position. By contrast, although a pure service can constitute an enrichment in English law, the defendant’s position is taken into account, and he can establish that he would have paid less if he had requested the service or if he had done the work himself, according to what is called ‘subjective devaluation’.57 In English law, therefore, where the claimant provides a service with no end-product, the startingpoint for the valuation of the enrichment is the objective value, ie the cost of the service in the general market. However, the defendant may ‘devalue’ this if he can establish that he would have paid less if he had requested the service or if he had done the work himself. This rule of ‘subjective devaluation’ enables the defendant to reduce the quantum of restitution to his real enrichment, ie the real cost he would have paid for the service, which can be different from the objective value. The comparison with English law of this particular case of ­enrichissement ­injustifié raises the question whether this action should be classified as a ­restitutionary claim in the sense that it is fundamentally concerned to reverse an unjustified enrichment or, even more generally, as a gain-based claim, since the defendant’s enrichment appears to be of secondary significance. Indeed, the cap on recovery set by the defendant’s enrichment (which forms one aspect of the double-cap rule) plays hardly any role in practice, for in the case of a pure service, the enrichment is deemed equal to the impoverishment. And where the

56 This was expressly recognised by the Cour de cassation in a decision where it was asserted that the collaboration of a ‘concubine’ in the work of her partner implied, in itself, an impoverishment and an enrichment. It was thus impossible for the partner to establish that he was not enriched, because if his partner had not done the work, he would have done it himself and would not have hired an employee: Cass civ 15 October 1996 no 94-20472. 57 This rule was created to overcome the difficulty posed by the subjectivity of value when the enrichment is in kind: see C Mitchell, ‘Unjust Enrichment’ in A Burrows (ed), Principles of the English Law of Obligations (Oxford, Oxford University Press, 2015) [3-27]ff. Since unjust enrichment concerns

366  Mélodie Combot service leads to an end-product, the defendant’s enrichment is often greater than the claimant’s impoverishment precisely because the impoverishment was aimed at preventing the claimant from having a right over the entire enrichment, ie the end-product, since this end-product is normally to be attributed to the owner of the asset improved, who, in the case of services, is the defendant. But even in cases where an asset is transferred, the defendant’s enrichment is sometimes irrelevant. An example can be found in a case decided by the Cour de cassation in 1988.58 A contract of sale was terminated retroactively at the request of the seller for non-payment of the price, and in addition to the restitution of the machines sold,59 the seller asked for reparation of his loss (préjudice60) resulting from depreciation of the machines, ie for the loss of value of his assets. As no fault on the part of the claimant/seller was found, and as he possessed no other action, the court decided to compensate this préjudice on the ground of enrichissement injustifié. Interestingly, there was no comparison with any corresponding enrichment, for example, the enrichment resulting from the value of the use of the machines by the defendant,61 and the defendant himself did not even try to argue that he was not enriched, or at least that he was enriched to a lesser extent than the impoverishment of the claimant. As a result, there appears to be only a single cap on recovery: the claimant’s impoverishment. The simple fact that the defendant used the machines sold justified the compensation to the extent to which the asset had depreciated. In other words, enrichment was seen only as a cause of action equivalent to an action giving rise to civil liability (fait générateur de responsabilité), such as fault or the action of things: it is not a measure of the recovery to be awarded. Overall, therefore, the true difference between recovery on the basis of unjustified enrichment and civil liability relates to the different approaches to loss, the notion of impoverishment being more restricted than the notion of préjudice, since it cannot include the loss of an opportunity to make a profit. This leads to another paradox. Thanks to the broad notion of préjudice, civil liability may be used in situations where the real aim of the claim is to reverse a benefit made by the defendant, for instance in cases of infringement of an intellectual property right or a personality right.62

the defendant’s gain and not the claimant’s loss, it is more appropriate to take into consideration the defendant’s position rather than the claimant’s one, unless the benefit received is incontrovertible. 58 Cass com 15 mars 1988 no 86-16691. 59 This action for the restitution of the machines was governed by the rules on restitution after termination of a contract, which are regarded in French law as not belonging to enrichissement ­injustifié or to quasi contrats. Only the additional action for the loss of value of the asset was founded on ­enrichissement injustifié and therefore is not, from a French perspective, an action in restitutions. On the role of r­ estitutions in French law, see above, p 355. 60 This is the word used in the decision, even if reparation was granted on the ground of e­ nrichissement injustifié. 61 The use value will rarely coincide with the depreciation of the asset, but to save appearances in this case the judges should have at least stated that the enrichment was equivalent to the impoverishment. 62 For cases where the enrichment was not transferred by the claimant see below, section III, p 371.

Unjustified Enrichment and Civil Liability  367

B.  The Role of Fault (i)  The Claimant’s Fault Another important element in the French law of unjustified enrichment is the claimant’s fault. This is very surprising, perhaps even more surprising than the relevance attributed to the notion of loss, since the main reason justifying the action of unjustified enrichment is that the claimant is entitled to recover the gain, regardless of his fault. And yet in French law the claimant’s fault is a significant factor that the court takes into account when assessing the quantum of the recovery: even though the claimant’s fault cannot bar the claim, it may lead the court to reducing the amount of whatever is to be returned to the claimant, even to ­nothing.63 There seems to be no convincing reason for the continuing vitality of this factor, which appears to be justified on the same grounds as the imposition of civil liability, namely punishment of the claimant and a ‘moralisation’ of the relevant remedy.64 The claimant’s position in unjustified enrichment is then close to that of the victim in a claim based on civil liability, for in that context the victim’s fault can also reduce compensation on the assumption that she is morally less deserving of legal protection. But what kind of fault can have this effect? The new provisions in the Code civil concerning quasi-contracts are not very helpful on this point, for they only provide for a reduction of the quantum of the recovery where the claimant commits a fault, whatever the character of the fault65 and, in particular, unlike the Projet de réforme, there is no distinction drawn between gross or dishonest fault.66 This is unfortunate because, unlike the victim in a civil liability claim who has generally played no role in the occurrence of his own harm, in claims based on a quasi-contract the claimant is often the sole person responsible for his own ‘harm’ (ie the deprivation of assets or money or the performance of an unrequested service). Furthermore, if fault is understood too broadly, it can affect almost every claim, especially if a simple mistake is equated with negligence. Before the 2016 reform of the Code civil, the case law fluctuated on this point depending on the type of quasi-contract under consideration,67 but generally speaking the effect of the claimant’s fault

63 For the evolution of fault among the different quasi-contracts, before and after the reform, see F Chénedé, Le nouveau droit des obligations et des contrats (Paris, Dalloz, 2016) [33.62]–[34.71]. 64 M Lecene-Marenaud, ‘Le rôle de la faute dans les quasi-contrats’ RTD civ 1994, 515. 65 See arts 1301-1, 1302-3, and 1303-2 Cc. 66 In certain circumstances the role of fault, either in the defendant or the claimant, may vary according to its seriousness. In the case of contractual liability, the debtor will be bound to make reparation for unforeseeable consequences of non-performance where he has committed a gross or dishonest fault (art 1251 Cc). More generally, in cases of personal injury, only the gross fault of the claimant/victim can reduce the amount of damages due by the defendant: art 1254 Cc. 67 This was particularly true for the action of enrichissement injustifié and répétition de l’indu. For the evolution of these two actions see, respectively, A-M Romani, ‘Enrichissement injustifié’ Répertoire civil Dalloz, February 2018, [227], and M Douchy-Oudot, ‘Paiment de l’indu’ Répertoire civil Dalloz, March 2018, [113].

368  Mélodie Combot varied according to the seriousness of her conduct, with the courts being inclined to limit more severely the quantum of recovery where the claimant was dishonest than where she was merely negligent. It should also be noted that the decision on this issue lies clearly within the sovereign power of assessment of the lower courts (the juges du fond) and that for this purpose the new provisions do not even mention the seriousness of fault as a possible factor to be taken into account by the court in reducing the amount due to the claimant. Instead they provide that the court may reduce the quantum of recovery, but they do not impose any obligation on the court to take into consideration the claimant’s fault.68 It is therefore possible that there will be considerable variation in what the courts decide in relation to different types of quasi-contract and even in relation to the same type of quasi-contract, as can be seen from earlier case law.69 In one case, a court may not take into account the claimant’s mere negligence, but in a very similar case another court may well do so. In other words, this assessment will be left to what the particular court feels to be fair on the facts of the case. This is tantamount to saying that the courts have a ‘margin of appreciation’ in determining the claimant’s recovery since, as earlier remarked, the claimant has often caused his own ‘harm’ and so can be easily considered to be at fault. French law’s uncertainty as to the role of fault should not be very surprising, because fault does not really fit with the idea behind unjustified enrichment, that is, the reattribution of a gain to the person entitled to it. The contrast with English law is again interesting.70 In principle, as stated in Kelly v Solari, the claimant’s fault is irrelevant to recovery of a mistaken payment ‘however careless the party paying may have been, in omitting to use due diligence to inquire into the fact’.71 The position would be different only where the claimant transferred the benefit willingly and in full awareness of the facts, or if he took a risk, because unjust enrichment only protects against a vitiated transfer of enrichment.72 Punishment, deterrence, or morality have nothing to do with unjust enrichment and, as Birks pointed out, strict liability is the only acceptable regime because it is first and foremost a question of entitlement and because the justifications for bearing a loss (under the law of civil liability) are very different from the justifications for the reallocation of a gain (under the law of unjust or unjustified enrichment),73 and the latter do not include fault. On the other hand, it is undeniable that the law of ­unjustified 68 See arts 1301-1, 1302-3 and 1303-2 Cc. 69 For example, before the 2016 reform, the claimant’s fault in the action of enrichissement injustifié (then called enrichissement sans cause) sometimes barred the action, whereas in répétition de l’indu it usually led to a reduction in the quantum of restitution; for these discrepancies in the old law, see M Lecene-Marenaud, (n 64). 70 For a comparative approach of fault in unjustified enrichment claims between French and English law, see C Kennefick, ‘La faute de l’appauvri: comparaisons anglo-françaises sur l’enrichissement injustifié’ RDC 2015, no 4, 961ff. 71 Kelly v Solari (1841) 152 ER 24, (1841) 9 M & W 54 Ex Ct at 59 (Parke B). 72 For a summary of the English position and a comparison with other jurisdictions, see J Beatson and E Schrage (eds), Cases, Materials and Texts on Unjustified Enrichment (Oxford, Hart Publishing, 2003) ch 8 (‘Interest, Fault and Risk’) 489. 73 Birks (n 15), 7–9.

Unjustified Enrichment and Civil Liability  369 enrichment can sometimes be very harsh to a defendant, who is compelled to reverse an enrichment which might well have disappeared since its receipt. Nevertheless, the claimant’s fault is not a useful tool to temper this harshness, mainly because it focuses exclusively on the claimant’s conduct and disregards the defendant’s position.

(ii)  The Defendant’s Fault Different considerations apply to the defendant’s fault, which in the context of unjustified enrichment is generally assessed in different ways depending on the defendant’s good or bad faith. Even if there is no accurate definition of good or bad faith, both French and English law look at whether the defendant knew that he was not entitled to the benefit. Both jurisdictions attach some effects to the defendant’s bad faith, but not in the same way. There are basically two possibilities, ie two different roles which may be recognised for the defendant’s fault; not surprisingly, this reflects wider differences between French and English law. A first possibility is to punish the defendant’s bad faith, meaning that he will have to reverse a greater amount than would have normally been the case. This is the position adopted by French law in the cases of enrichissement injustifié and répétition de l’indu. As to the former, the new article 1303-4 Cc provides that where the defendant is in bad faith, the double cap rule does not apply and, as a result, he must restore the greater amount as between the defendant’s enrichment and the claimant’s impoverishment. As to claims for répétition de l’indu, article 1352-2 Cc provides that where a defendant in bad faith has sold the asset received, he must restore the value of this asset rather than the price which he received for it if the former is greater than the latter. This rule may be interpreted in another way, ie not as a form of punishment of the defendant’s bad faith, but rather as a lenient solution for the benefit of defendants in good faith, for the latter will not be obliged to reverse the enrichment (the real value of the asset he received) but only what he retained from it (the price of the sale). Compared to the solution envisaged in relation to enrichissement injustifié, the solution in the case of répétition de l’indu is closer to the English position, even if the solution for répétition de l’indu is more usually interpreted as a punishment against the bad faith defendant. Another possibility is to temper the effect of a claim for unjustified enrichment for those defendants who deserve it, rather than to punish those in bad faith. English law has developed a particular defence specific to claims for unjust enrichment for this purpose: the defence of change of position. Normally, where all the conditions of an unjust enrichment are fulfilled, the defendant has to reverse all gains he unjustly received, notwithstanding his good or bad faith. But sometimes, the effect of the action can be particularly harsh for a defendant who has disposed of the enrichment wholly or in part, and can even make him worse off.74 In French law, as 74 In English law, the date used to determine whether or not the defendant is enriched, and, if so, to what extent, is the date of receipt: see notably, Burrows (n 57), [3.22].

370  Mélodie Combot regards enrichissement injustifié, article 1303-4 Cc chooses a radical solution: the relevant date in assessing the enrichment is the day of the claim, with the consequence that if the defendant disposes of all or part of the enrichment before that day, his enrichment will be decreased or even extinguished compared to what he actually received – and this even if he was in bad faith!75 All this means that the claimant alone bears the risk of ‘dis-enrichment’ even though, according to the logic of unjustified enrichment, the defendant received something that was not his own. This is why English law prefers to opt for the day of receipt rather than the day of claim as the proper time for assessment of the value of the b ­ enefit.76 But if the first solution (the date of the claim) is too harsh to the claimant, the second solution can be particularly burdensome for the defendant. To avoid this, the change of position defence allows a defendant in good faith to reduce the amount of restitutionary recovery even to nothing. To rely on this defence, good faith alone is not enough; the defendant must also establish that, had he not received the enrichment, he would have not engaged in certain courses of action which entailed ‘consumption’ of that enrichment.77 The fact that bad faith defendants cannot use the defence of change of position is not a punitive response to their being in bad faith, as good faith recipients may also be barred from relying on the defence if they fail to meet the criteria for its application, as discussed above. Clearly, punishment is not the goal of the law of unjust enrichment and it should have no part to play in it. The aim of the defence of change of position is merely to temper the potentially harsh effect of unjust enrichment claims in some circumstances. In contrast, French law does not provide any equivalent protection for defendants, notably in cases of répétition de l’indu, with the consequence that the unfortunate defendant can only hope that the claimant committed some sort of fault and that the court will on this ground reduce the quantum of the claimant’s recovery.78 The defence of change of position is an excellent illustration of the independence of recovery on the basis of unjust enrichment in English law from other legal grounds of recovery and in particular suggests that borrowings from the law of civil liability (and in particular relying on the fault of one or other party) are often 75 The same article provides that it is the greater value that will be taken into account, but does not change the date of assessment of the quantum of recovery. However, where the defendant is in bad faith he will still be obliged to compensate the claimant’s impoverishment which is likely to be the greater amount, whereas a defendant in good faith may have nothing to pay if his enrichment on the date of the claim is nil. 76 See above, n 74. It must be noted that, without apparent justification, the French position differs in relation to répétition de l’indu where it resembles the English solution. Arts 1302, 1352, and 1352-3 and, more implicitly, arts 1352-1–1352-2 Cc provide some exceptions to the solution of the day of receipt. 77 The famous example is where a person receives a large amount of money and decides to quit his job and make a world tour thanks to that money; but for the unexpected enrichment, this person would have never made these choices. However, where a person who has received money by mistake uses it to pay off an existing loan, that person cannot rely on the defence of change of position, as it is not detrimental to him to pay off a loan which would have had to be paid off at some stage anyway: Derby v Scottish Equitable Plc [2001] Pens LR 163 esp at [35]. 78 See Kennefick (n 70). Except, as noted previously, for art 1352-2 Cc, where the defendant receives an asset which is sold afterwards.

Unjustified Enrichment and Civil Liability  371 inappropriate for this ground of recovery because of the different rationales underpinning these two distinct areas of law. However, even if the independence of unjustified enrichment in French law were to be fully recognised, there would still remain some areas of overlap. We should, therefore, examine the proper boundaries between civil liability and unjustified enrichment.

III.  The Boundaries between Civil Liability and Unjustified Enrichment: Which Should be the Basis of Recovery of Unlawful Benefits? There is one particular set of circumstances which is difficult to classify as between civil liability and unjustified enrichment: which regime should apply where a benefit is unlawfully made in violation of a duty or of an obligation to which a person was bound for the benefit of another person? For example, what should happen where a party deliberately breaks her contract to make a better bargain with someone else or where a person uses another person’s image without their consent and sells it to a magazine? In order to answer these questions, it is first necessary to examine the reasons which may justify the application of the law of unjustified enrichment or instead the law of civil liability. Which should have priority in the eyes of the law as between the unlawful action of the defendant and the encroachment on the claimant’s right to the benefit in question? If it is the unlawful action of the defendant, civil liability is more likely to be appropriate. The general idea here would be the need to prevent the defendant from profiting from his unlawful conduct, and this would entail the giving up of all the profits linked to his action. If instead the general idea is not to punish or deter the defendant, but rather to reallocate a gain to the person entitled to it, then unjustified enrichment can be appropriate and more flexible in assessing the measure of recovery. Indeed, in this context, the claimant’s recovery will not necessarily take into account all the profits made by the defendant, but only that portion attributable to the claimant, as the protection of the claimant is the paramount goal here. Against this backdrop, three different approaches emerge among European jurisdictions. A first solution is to prefer unjustified enrichment, as German law does,79 on the ground that it is first a matter of attributing rights: does the law attribute this benefit (here, the profit made) exclusively to the claimant? The defendant’s unlawful action is irrelevant if it is assumed that not all unlawful behaviour triggers restitution of the subsequent benefit. Moreover, this approach ensures greater flexibility in assessing the measure of recovery than that which applies in the context of civil liability, and it prevents the claimant from receiving more than he deserves.

79 For an account of this particular branch of German unjustified enrichment law, see Dannemann (n 5) 91 ff; Krebs (n 32) 383–88.

372  Mélodie Combot This particular branch of unjustified enrichment concerns notably cases of infringement of property rights (including intellectual property) or of personality rights.80 A second solution, adopted by French law, is to treat all these claims as falling within the scope of civil liability, not only because the defendant’s fault is seen as deserving the attention of this area of law but also, and more importantly, because it will not be possible to reach a suitable result by the application of the law of unjustified enrichment whenever the benefit/profit is not taken directly from the claimant’s assets but only received by the defendant.81 For where the latter is the case, the significance of the claimant’s loss in setting a limit on the recovery prevents the practical application of the law of enrichissement injustifié.82 Other drawbacks follow from the French failure in cases of breach of a duty to distinguish between the restitution of benefits and the compensation of losses. Since the notion of préjudice is broad enough to encompass both losses and profits,83 all the cases which would naturally attract the law of unjustified enrichment are instead seen as cases where compensation is desirable and thus as belonging to the province of the law of civil liability. This attitude can lead to the distortion of the rules of civil liability as well as to a phenomenon of ‘over-restitution’. Indeed, in French law, there are almost no rules governing the assessment of damages, and the courts have in practice a discretion to decide the quantum of damages. As a result, the courts are often tempted to take into account profits made by the defendant in the valuation of the damages, even though this is hidden by the idea that these damages are still considered to be compensatory rather than restitutionary. In the worst-case scenario, this may even lead to the recovery of both restitution and compensation. An example can be found in relation to some provisions of the Intellectual Property Code, which provide that damages for breach of intellectual property rights must include not only the loss sustained but also the moral harm and the profits realised by the defendant.84 This confusion is almost inevitable if there is no proper distinction between restitution of a gain and compensation of a loss, especially if the latter is understood in a very broad way. The final and third option, embraced by English law, is more complex and still not firmly settled. The dominant view is to distinguish between restitution for unjust enrichment and restitution for wrongs. Where a wrong allows the wrongdoer to

80 German law has evolved on that point and there are still some difficulties: see notably Beatson and Schrage (n 72) 544–48. 81 The clearest example is the infringement of a patent, where a person makes profits out of an invention that is not his and without the consent of the holder of the patent. 82 Not to mention the condition of subsidiarity for the unjustified enrichment action, which will preclude the claimant to sue on that ground if a civil liability claim is also available. Répétition de l’indu will be of no use in that case since it is not the claimant who transfers the profit. 83 Above, p 364. 84 See eg art L331-1-3 C propr int. By contrast, English law rejects in principle the idea that the claimant can have both a restitutionary and compensatory right for the same wrong: see A Burrows, (n 33) 627.

Unjustified Enrichment and Civil Liability  373 make a profit, the relevant cause of action is assumed to be that wrong, and not unjust enrichment.85 Unlike French law, there is here a division between compensatory damages and ‘restitutionary’ damages, which aim to reverse the gain of the wrongdoer and not to compensate the claimant’s loss. But it is generally assumed that not all wrongs can trigger restitution, unlike compensatory damages which are always available whatever the wrong involved. However, there is currently no clear answer to the question as to which wrongs attract a restitutionary remedy. Besides, the class of wrongs which do attract a restitutionary remedy suffers from a lack of unity, especially as between cases involving a breach of contract and ones where the defendant commits a tort. In the first case, the reason to grant restitution is almost invariably to punish the party in breach and deter others from doing the same, as famously illustrated by Blake’s case.86 It is possible that restitution for breach of contract is thus a way to circumvent the impossibility to have exemplary/punitive damages on that ground.87 By contrast, where a tort is involved, a majority of cases come close to the logic of unjust enrichment and to the German doctrine of attribution.88 This approach generally concerns torts which infringe proprietary interests and has therefore been adopted in cases of trespass to land, goods, conversion, or intellectual property torts.89 It should be noted that this dichotomy between torts and breaches of contract is not absolute and that the courts have also recognised ‘negotiating damages’, mainly for breach of restrictive covenants or breach of a confidentiality agreement. This type of damages is close to the restitutionary damages granted in cases of proprietary torts. As explained in a recent decision of the UK Supreme Court, the idea is that in some circumstances a contractual right can be seen as a valuable asset, so that the claimant can argue that he was deprived of property, to be measured by the economic value of the infringed right.90 However, since it is not the claimant’s exclusive entitlement to that gain which constitutes the cause of action, it is currently very problematic to determine accurately a list of ‘restitutionary’ wrongs and it is indeed possible to find contradictory decisions on this issue.91 It is also worth noting that, in these cases, the assessment of the quantum of restitution creates difficulties because the reversal of all the profits is not always effective, and thus several ‘sub-’ remedies coexist under the heading of restitution for wrongs, as for example ‘account of profits’ or ‘money had and received’.92 85 See in particular Birks (n 15) 11–16. 86 Attorney-General v Blake [2001] 1 AC 268. 87 See J Edelman, McGregor on Damages, 20th edn (London, Sweet & Maxwell, 2018) [13-016]. 88 In that sense, see Krebs (n 32) 393–98. 89 Burrows (n 33) 643–54. 90 One Step Ltd v Morris-Garner [2018] UKSC 20, [93]. However, the reasoning in the decision is not very clear in relation to the classification of these damages, which are not seen as ‘fundamentally incompatible with the compensatory purpose of an award of contractual damages’: ibid at [91] per Lord Reed JSC. What is clear is that these damages are not granted to deprive the defendant of all the profits he made from the breach of contract, in apparent contrast with the position in Blake’s case. 91 Burrows (n 33) 659–62. 92 ibid 627.

374  Mélodie Combot The current state of English law reflects fully the ambiguity of the subject and the difficulty of including all cases of unlawful gain within the same category, whether this is unjust enrichment, civil liability, or restitution for wrongs. Where the law’s main goal is to prevent the wrongdoer from benefiting from his wrong, unjustified enrichment looks inadequate, though this does not mean that civil liability should be the path to adopt (as the criminal law may appear more appropriate). If instead the primary goal of the law is the protection of the claimant’s rights, civil liability seems a dangerous response, as it carries with it a risk of over-restitution if proper limits are not adopted or, conversely, a useless protection if the list of these restitutionary wrongs is too restricted for want of criteria to select them and to assess the measure of restitution. In this regard, German law offers a very interesting alternative and encourages a focus on the nature of the right infringed and on the proper justification for its protection. This does not mean that all problems would be resolved, as further analysis would be necessary to establish where the claimant has an exclusive right to a gain; nor does it mean that German law should be imported into French or English law. Nevertheless, this short comparison shows that, if more coherence is desired in relation to unlawful gains, it is not enough to frame the topic in terms of a binary debate between civil liability and criminal law or between reparation and punishment. There are cases which do not fit these patterns and suggest that greater attention should be paid to the special link between the claimant and the benefit or profit, rather than to the behaviour of the defendant (a criterion which seems unsuitable for the purpose of choosing the correct basis of claim). Cases involving the violation of intellectual property rights provide a clear but not isolated illustration of this type of approach, and suggest that unjustified enrichment can be a more effective answer to these sorts of cases, especially where an unauthorised person receives or takes, consciously or not, a benefit or profit that the law attributes to someone else.

IV. Conclusion Despite the silence of the Projet de réforme, interactions between unjustified enrichment and civil liability are numerous and complex. It is particularly ­striking to see the differences between English and French law, although unjustified enrichment is, at first sight, described in a very similar way. Simply comparing the law of unjustified enrichment in the two legal systems is insufficient to fully understand the reasons for these differences; in this respect, it would be desirable to include in a comparative analysis of unjustified enrichment careful consideration of nearby legal areas, especially the law of extra-contractual liability. In this way, it would be possible to identify and appreciate the reasons and essential characteristics of both areas and to map out the boundaries between the two. This is no easy task, since the category of civil liability appears to lack coherence both in English and French law, and since this lack of coherence is reflected in the law of unjustified enrichment.

Unjustified Enrichment and Civil Liability  375 In English law, no great efforts have been made to provide a coherent category of torts, or more generally of wrongs, but much energy has been spent in the past decades to establish a coherent category of unjust enrichment, with its own specific rules and conditions and without borrowing from other areas of law. Assuming that it is therefore possible to distinguish between tort law and unjust enrichment, it is nevertheless true that the eclecticism and residuary character of the former impacts on the boundaries between the two. Since tort law does not have a single purpose and its role cannot be reduced merely to the compensation of losses, why could it not deal with the restitution of gains? The problem here is that the reasons for compensating losses are not the same as the reasons for giving up a gain and a simple analysis on a tort basis has proved to be unsuccessful to select torts eligible for restitution. Applying the law of unjust enrichment would be more suitable in many cases, but to make a convincing case for such an application it is also necessary to find a reason why tort law should not deal with these cases, either wholly or in part. This requires us, in turn, to identify a specific purpose for tort which is not compatible with restitution of a gain. In French law, the distinction between civil liability and unjustified enrichment proves to be more difficult. Neither of these categories have a clear and distinct purpose. The alleged ability of civil liability to provide the model for all ­extra-contractual obligations has considerably blurred the distinction and has provoked some unfortunate transplants into the law of unjustified enrichment, which has never found a proper place in the classification of obligations, being assimilated to quasi-contracts. The inconsistencies of unjustified enrichment, in addition to the very broad scope of civil liability, have led to the consequence that civil liability is actually more easily used to reverse a gain than is unjustified enrichment. However, this new function of civil liability is not distinguished from its traditional functions: to compensate losses and sometimes to punish the defendant’s conduct. A consequence of this is that there is no difference of regime and in the assessment of damages, contrary to the English distinction between compensatory and restitutionary damages. This leads to a vicious circle and tends to strengthen the position of civil liability as a general basis for all non-contractual claims. Perhaps, this comparison between civil liability and unjustified enrichment should invite us to reconsider more generally the map of extra-contractual obligations, which are not limited to these two categories and whose boundaries could be challenged. New distinctions can possibly emerge from a deeper analysis of this still little-known area and eventually lead to a reduction in the scope of the law of civil liability, to the benefit of other extra-contractual sources of obligations.

376

18 Injunctions Requiring the Cessation of Unlawful Action PAULA GILIKER

I. Introduction The Projet de réforme de la responsabilité civile published by the French Ministry of Justice on 13 March 2017 represents an ambitious attempt to modernise the French law of tort,1 and follows in the footsteps of the contract law reforms arising from the Ordonnance of 2016.2 This chapter will examine the proposals in the Projet de réforme in a particular context – injunctive relief which requires the defendant to stop or not to undertake unlawful actions which will harm the claimant (in French, la cessation de l’illicite). This represents an innovation. While cessation de l’illicite is currently used in relation to neighbour disputes or in specific contexts such as unfair competition and infringement of intellectual property rights, the aim is to introduce it as part of the general remedial structure of tort law – the implications of which will be examined in this chapter. In particular, I will address to what extent this one provision, set out in article 1266 of the Projet, is capable of challenging how we conceptualise responsabilité civile extracontractuelle by introducing measures which do not simply seek to respond to harm caused to the victim, but which seek to prevent harm occurring in the first place. This chapter will therefore examine the theoretical and practical basis for article 1266, gaining insights from a comparison with English law. In commissioning a translation of the Projet de réforme into English, the French Ministry of Justice expressly recognised the value of such insights, stating that: 1 Responsabilité civile extracontractuelle will be translated here as tort, rather than extra-contractual or delictual responsibility, for ease of reference, but readers should note that it is not to be confused with common law tort law. The Projet de réforme uses the term responsabilité civile extracontractuelle rather than responsabilité civile délictuelle to indicate clearly that it extends to no-fault liability. 2 Ordonnance no 2016-131, 10 February 2016 (now ratified by loi no 2018-287 of 20 April 2018). See G Helleringer, ‘The Anatomy of the New French Law of Contract’ (2017) 13 ERPL 355; S Rowan, ‘The New French Law of Contract’ (2017) 66 ICLQ 805; J Cartwright and S Whittaker (eds), The Code Napoléon Rewritten. For some, however, the project is not ambitious enough: see M Fabre-Magnan, ‘Un projet à refaire’ RDC 2016, 782.

378  Paula Giliker This translation [will] allow the project to be brought to the attention of foreign legal professionals before it enters its legislative phase. Indeed … taking into account the practices and points of view of our neighbours, near or far, is essential to arrive at an attractive body of rules adapted to contemporary issues. This translation will also help, in the long term, to accompany the diffusion abroad of the new French law of obligations and to ensure its influence, especially with countries influenced by the Code Napoléon.3

My chapter will thus be given from the perspective of a voisine who, despite Brexit, I hope remains proche. I will outline the new proposals, examine parallel legal rules in the English common law and then identify a number of outstanding questions which, from the perspective of a comparative lawyer, the French legislator should consider prior to implementation of the Projet de réforme.

II. Background Injunctive relief may be found in French law in a number of legislative provisions. The most well-known procedure is that of the interlocutory référé procedure, stated in article 809 CPC.4 This provides that: The president [of the High Court] may always, even where confronted with a serious challenge,5 order in a summary procedure such protective measures or measures to restore (the parties) to (their) previous state as required, either to avoid an imminent damage or to abate a manifestly illegal nuisance.6 In cases where the existence of the obligation is not seriously challenged, he may award an interim payment to the creditor or order the mandatory performance of the obligation even where it is an obligation to do a particular thing.7 3 Ministry of Justice, ‘Projet de réforme de la responsabilité civile traduit en anglais’ (23 May 2017), translation by Giliker. See www.textes.justice.gouv.fr/textes-soumis-a-concertation-10179/projet-dereforme-de-la-responsabilite-civile-traduit-en-anglais-30553.html. For this translation by S Whittaker and J-S Borghetti, see Appendix. 4 See, generally, Y Strickler, Procédure civile, 8th edn (Bruxelles, Bruylant, 2018) [417]–[433]; C Chainais, F Ferrand, S Guinchard and L Mayer, Procédure civile: Droit interne et européen du procès civil, 34th edn (Paris, Dalloz, 2018) Pt 2, t 2, 1339, [1867]ff; M Foulon and Y Strickler, ‘Les pouvoirs du juge des référés’ Gaz Pal 25-26 May 2012, 17. 5 Inserted by decree no 87-434 of 17 June 1987 (JO 23 June 1987). 6 ‘Nuisance’ here indicates un trouble; a term extending beyond the common law concept of ‘nuisance’. This presents a dilemma for translators who, in opting for the translation ‘nuisance’, risk the misunderstanding that it is confined to troubles de voisinage. Readers should therefore note that the term is intended to extend beyond troubles de voisinage and may be better translated as ‘illegal behaviour’. Bloch, for example, in his leading treatise adopts a broad definition of trouble as ‘le fait ou l’état de fait illicite vu à travers le prisme de la perturbation qu’il est de nature à porter aux intérêts légitimes du demandeur’: C Bloch, La cessation de l’illicite, Recherche sur une fonction méconnue de la responsabilité civile extracontractuelle (Paris, Dalloz, 2008) [384]. 7 Emphasis added (translation courtesy of Légifrance with the participation of Y-A Tsegaye). In French: ‘Le président peut toujours, même en présence d’une contestation sérieuse, prescrire en référé les mesures conservatoires ou de remise en état qui s’imposent, soit pour prévenir un dommage imminent, soit pour faire cesser un trouble manifestement illicite. Dans les cas où l’existence de l’obligation

Injunctions Requiring the Cessation of Unlawful Action  379 Claims before the juge des référés thus follow a well-established procedure which is commonly praised for its speed and efficiency.8 Proceedings occur before litigation on the merits has commenced with ‘l’ordonnance de référé’ described by article 484 CPC as an interim order given at the request of one party, the other party being present or summoned, in cases where the law confers upon a judge, to whom the main issue is not referred, the power to give the necessary orders immediately. JurisClasseur, for example, notes regular use of article 809 CPC in relation to nuisance and environmental issues,9 but also in privacy cases.10 It has also been used in relation to unfair competition, attacks on freedom of expression, and employment rights.11 The term ‘illegal’ or illicite, therefore, is taken widely to include everything contrary to the law or custom.12 Recent cases include a failure to execute demolition measures ordered by a criminal judge,13 age discrimination,14 abuse of the right to strike,15 and breach of confidentiality.16 Such examples are marked by a finding by the judge of a clear, certain and unquestionable illegal act. The court responds by stating the necessary measures to restore the parties to their previous position (thereby ending the harm) or to prevent the anticipated harm, although it is not permitted to make an order if the circumstances giving rise to the trouble are at an end.17 Article 491 CPC allows the juge des référés to use the mechanism of an astreinte (a financial sanction for non-compliance with a judicial order) to ensure the decisions of the court are respected.18 Viney has commented that article 809 CPC has been the principal engine behind the development of the concept of cessation de l’illicite in private law.19 The idea of the courts acting to prevent harm, in addition to providing remedies to compensate for harm already caused, is therefore not new to French law.

n’est pas sérieusement contestable, il peut accorder une provision au créancier, ou ordonner l’exécution de l’obligation même s’il s’agit d’une obligation de faire’. 8 Strickler (n 4) [418]. 9 N Leblond, ‘Le préjudice écologique’, JCl Civil Code, arts 1382 to 1386, Fasc 112. 10 J-C Saint-Pau, ‘Jouissance des droits civils. – Droit au respect de la vie privée. – Régime. Actions’, JCl Civil Code, art 9, Fasc 20. 11 G Viney, ‘Cessation de l’illicite et responsabilité civile’ in O Cachard and X Henry (eds), Mélanges Goubeaux (Paris, LGDJ, 2009) 555; C Bloch, ‘La cessation de l’illicite’ (Thesis, Aix-Marseille III, 2006) [186]–[190]. 12 See, for example, Cass com 22 July 1986 no 84-17.521. There has been an active debate for many years whether a finding of trouble manifestement illicite is subject to the control of the Cour de cassation or rather like dommage imminent merely subject to the sovereign power of assessment of the lower court judges: Cass civ (3) 5 November 2015 no 14-18.184. It is now resolved that it is subject to the control of the Cour de cassation in that it is not simply a finding of fact: see recently Cass civ (3) 21 December 2017 no 16-25469. 13 Cass civ (3) 5 March 2014 no 13-12.540. 14 Cass soc 18 February 2014 no 13-10.294. 15 Cass soc 30 June 2015 no 14-10.764, a contrario. 16 Cass com 15 December 2015 no 14-11.500. 17 TGI Paris 9 February 1988 Gaz Pal 1988, 1, 343, note P Bertin; CA Rouen 5 May 1988 Gaz Pal 1988, 2, somm, 442. 18 See Strickler (n 4) [488]. 19 Viney (n 11) 555.

380  Paula Giliker While the 2005 Avant-projet de réforme Catala20 did not put forward prevention of harm as one of the specific purposes of liability, it did provide for preventative measures in its articles 1369-121 and 1344.22 A later working group supported by the Ministry of Justice and led by Professor François Terré which reported in 2010,23 was, however, more forthright. Article 2 of the Avant-projet Terré stated that: Independent of any reparation for harm which may have been suffered, the judge shall prescribe reasonable measures appropriate to prevent or to bring to an end the ‘trouble illicite’ to which the claimant is exposed.24

The aim, therefore, is to transform a mechanism already in existence in French law, notably through the article 809 CPC interlocutory procedure, into a procedure generally available. Influenced clearly by this recommendation, the original 2016 proposal of the Ministry of Justice, which was put out for consultation, stated in Chapter 1 as a preliminary provision (in English and in French):25 Article 1232: Independent of any reparation for loss which may have been suffered, the judge may prescribe reasonable measures appropriate to prevent or to bring to an end the ‘trouble illicite’ to which the claimant is exposed. [Only actions contravening a rule of conduct imposed by the law or by the general duty of care or diligence may give rise to such measures.]26 Article 1232: Indépendamment de la réparation du préjudice éventuellement subi, le juge peut prescrire les mesures raisonnables propres à prévenir ou faire cesser le trouble illicite auquel est exposé le demandeur. [Seuls les faits contrevenant à une règle de conduite imposée par la loi ou par le devoir général de prudence ou de diligence peuvent donner lieu à de telles mesures.]

The underlined text indicates changes from the Avant-projet Terré text. Dommage (harm) is replaced by préjudice (loss) and the word ‘shall’ becomes ‘peut/may’, giving the judge a discretion to act. Article 1232 is confined, however, to liability 20 Translated into English by John Cartwright and Simon Whittaker: www.justice.gouv.fr/art_pix/ rapportcatatla0905-anglais.pdf. 21 ‘Where harm is liable to become worse, to reoccur or to linger, a court may, at the request of the victim, order any measure appropriate to avoid these consequences, including if need be an order for discontinuation of the harmful activity in question’. 22 ‘Expenses incurred in order to prevent the imminent occurrence of harm, to avoid its getting worse, or to reduce its consequences, constitute a reparable loss as long as they were reasonably undertaken’. 23 See F Terré (ed), Pour une réforme du droit de la responsabilité civile (Paris, Dalloz, 2011). Contributors were: Cyril Bloch, Jean-Sébastien Borghetti, Clothilde Grare-Didier, Denis Mazeaud, Soraya Messaï-Bahri, Anne Outin-Adam, Philippe Remy, Pauline Remy-Corlay, Myriam Roussille, Philippe Stoffel-Munck and François Terré. 24 Translation by Giliker. In French: ‘Indépendamment de la réparation du dommage éventuellement subi, le juge prescrit les mesures raisonnables propres à prévenir ou faire cesser le trouble illicite auquel est exposé le demandeur’. 25 Ministry of Justice, ‘Consultation publique sur la réforme de la responsabilité civile’ (29 April 2016): www.justice.gouv.fr/formulaires/responsabilitecivile2016/. 26 Translation by Giliker.

Injunctions Requiring the Cessation of Unlawful Action  381 arising from breach of a rule of conduct imposed by law or by the general duty of care or diligence. Arguably, such a qualification is unnecessary when the trouble must be illicite to trigger a response, but it may be seen as an attempt to restrain the courts from interpreting this term more broadly. A significant common characteristic, however, is the prominence of both articles at the start of each draft legislative instrument, giving explicit support for the argument that preventive measures should be regarded as one of the objectives of the French law of tort.27 Such measures are also seen as bringing French law in line with other European jurisdictions. Viney remarks, for example, that similar measures exist in German law.28 Paras 935ff of the German Code of Civil Procedure (ZPO) provide for einstweilige Verfügungen (temporary injunctions) to be obtained from a court on application of a party to proceedings. These are explicitly awarded to protect the rights of the parties: ‘Injunctions regarding the subject matter of the litigation are an available remedy given the concern that a change of the status quo might frustrate the realisation of the right enjoyed by a party, or might make its realisation significantly more difficult.’29 It may be noted, however, that these provisions, like article 809, are in the Code of Civil Procedure; Germany making a clear distinction between issues of substantive law such as obligations and rights which are in the German Civil Code (BGB) and issues of procedure found in the ZPO.30 Indeed, the tort law provisions of the BGB make no specific provision for injunctive relief.31 Article 1232, however, would be in the Code civil.

III.  Injunctive Relief against Troubles Illicites in the Projet de Réforme Following consultation, the French Ministry of Justice published in March 2017 a final version of the proposals. In terms of cessation de l’illicite, the relevant

27 See C Bloch, ‘La reconnaissance de la cessation de l’illicite comme fonction autonome de la responsabilité civile’ JCP G 2016 Suppl to nos 30–35 (25 July 2016) 5. 28 G Viney, ‘Après la réforme du contrat, la necessaire réforme des textes du Code civil relatifs à la responsabilité’ JCP 2016 G doctr 99, [28]. 29 § 935 Code of Civil Procedure. Translation: www.gesetze-im-internet.de/englisch_zpo/englisch_ zpo.html#p3491. In German: ‘Einstweilige Verfügungen in Bezug auf den Streitgegenstand sind zulässig, wenn zu besorgen ist, dass durch eine Veränderung des bestehenden Zustandes die Verwirklichung des Rechts einer Partei vereitelt oder wesentlich erschwert werden könnte’. See also § 940: ‘Injunctions are also admissible for the purpose of providing for a temporary status concerning a legal relationship that is in dispute, to the extent this provision is deemed to be necessary in order to avert significant disadvantages, to prevent impending force, or for other reasons in particular in the case of long-term legal relationships.’ 30 See, in English, N Foster and S Sule, German Legal System and Laws, 3rd edn (Oxford, Oxford University Press, 2002) 122–35; A Freckmann and J Wegerich, The German Legal System (London, Sweet & Maxwell, 1999) 177. 31 BS Markesinis and H Unberath, The German Law of Torts: A Comparative Thesis, 4th edn (Oxford, Hart Publishing, 2002) 475.

382  Paula Giliker ­ rovision is now article 1266 of the Projet de réforme.32 This provides as follows p (in English and French): CHAPTER IV – THE EFFECTS OF LIABILITY – Sub-section 4. Cessation of unlawful action Article 1266: In extra-contractual matters, independently of any reparation of loss which may have been suffered, a court may prescribe reasonable measures appropriate to prevent harm or to see that an unlawful nuisance to which a claimant is exposed is stopped.33 CHAPITRE IV – LES EFFETS DE LA RESPONSABILITÉ – Sous-section 4. La cessation de l’illicite: Article 1266: En matière extracontractuelle, indépendamment de la réparation du préjudice éventuellement subi, le juge peut prescrire les mesures raisonnables propres à prévenir le dommage ou faire cesser le trouble illicite auquel est exposé le demandeur.

The 2017 provision, however, is different in a number of ways from its predecessors. First, the provision is in a sub-section of Chapter IV on les effets and no longer regarded as a preliminary provision in Chapter I. This is significant. When included in the introductory provisions, there was a clear argument that the provision stated a basic objective of the Projet de réforme: that prevention of harm should be seen as much an underlying objective of private law as compensation. This has been abandoned and it is now confined to a list of responses to wrongdoing in Chapter IV. Secondly, the wording has changed again. Article 1266 is confined to ‘extra-contractual’ matters and so will not apply to contract law. This contrasts with the general intention of Chapter IV to provide common rules for contract and tort liability as indicated in the French title which does not distinguish between contract and tortious liability: Les effets de la responsabilité. This should not, it is submitted, worry us unduly. Carval, for example, has commented that this will avoid any overlap with the remedy of l’exécution forcée in contract law which she deemed advisable.34 The drafters are therefore seeking to give litigants greater clarity. Thirdly, the attempt to specify les faits illicites which will give rise to such measures has been dropped. As stated above, the Avant-projet Terré had tried to confine the provision to breach of ‘une règle de conduite imposée par la loi ou par le devoir général de prudence ou de diligence’. Borghetti noted the potential ambiguity here – is it then confined to fault-based conduct or could it

32 Note also art 1244 of the Projet de réforme (abnormal nuisance between neighbours/les troubles anormaux de voisinage). This is confined to troubles anormaux de voisinage, however, and not all types of troubles illicites (a further indication that le trouble illicite should not be seen as confined to what a common law tort lawyer would call a ‘nuisance’). We might question, nevertheless, the relationship between these two provisions which is not articulated clearly in the Projet de réforme. Art 1244(2) expressly provides that: ‘Where a harmful activity has been authorised by an administrative means, the court may, however, award damages or order reasonable measures permitting the nuisance to be stopped.’ There will, therefore, be some interaction which will exist between arts 1244 and 1266. 33 See translation provided in the Appendix. 34 S Carval, ‘Le projet de réforme du droit de la responsabilité civile’ JCP G 2017, 401.

Injunctions Requiring the Cessation of Unlawful Action  383 extend to conduct which is not based on fault, but nevertheless harms a specifically protected right?35 The Ministry dealt with such concerns by simply deleting the text. This does suggest, therefore, that the provision is no longer confined (if it was previously) to fault-based liability. While the term illicite does operate to delimit the powers of the courts, two further absentees are the terms ‘imminent damage’ and ‘manifestly illegal nuisance’ found in article 809 CPC. It might be argued that these terms are needed to limit the powers of the courts when dealing with interim applications prior to a full hearing of the case, but their absence does raise the question of the scope of the judges’ powers. Borghetti has warned also that the current wording may give rise to a risk of judicial interference with perfectly legal (licites) activities if care is not taken to avoid this consequence.36 Nevertheless, despite such concerns, as the Ministry of Justice stated in its press release, this is an ‘évolution notable’37 and an indication that the law generally should seek to respond to the source of the harm, not simply its consequences.38 What is perhaps most significant is recognition that judicial intervention to prevent tortious behaviour should no longer be deemed confined to specific areas of law such as nuisance and privacy. Looking more closely at Chapter IV, Section 1 (Principles), introductory provisions (articles 1258 and 1259) set out the fundamental objectives of reparation in French contract and tort law; article 1259 providing that reparation may take the form of reparation in kind (réparation en nature39) or damages (dommages et intérêts), these two types of measures being able to be combined so as to ensure full reparation of the loss.40 Articles 1260 and 1261 specify that reparation in kind must be specifically appropriate to suppress, reduce or make up for the harm suffered, cannot be imposed on the victim, and must not be ordered in the case of impossibility or of manifest disproportionality between its cost for the person liable and its interest for the victim. Damages, dealt with in sub-section 2, are stated to be assessed as of the day of judgment, taking into account all the circumstances which could have affected the make-up and the value of the loss since the day of the manifestation of the harm, as well as its reasonably foreseeable development.41 What is interesting here is that article 1266 receives a separate billing in sub-section 4 (in Chapter IV, Section 1), after the discussion of reparation in kind

35 J-S Borghetti, ‘L’avant-projet de réforme de la responsabilité civile’ D 2016, 1442 on the first draft of reforms (the 2016 Avant-projet) [4]. 36 J-S Borghetti, ‘Un pas de plus vers la réforme de la responsabilité civile: présentation du projet de réforme rendu public le 13 mars 2017’ D 2017, 770, [22]. 37 Ministry of Justice, Dossier de Presse: Projet de réforme du droit de la responsabilité civile (13 March 2017). 38 ‘Il ne s’agit plus seulement de réparer le dommage, mais d’agir sur sa source’: ibid 3. 39 There is, therefore, a close link with l’exécution forcée en nature in the context of contractual nonperformance, on which see arts 1221 and 1222 Cc. 40 ‘La réparation peut prendre la forme d’une réparation en nature ou de dommages et intérêts, ces deux types de mesures pouvant se cumuler afin d’assurer la réparation intégrale du préjudice’. 41 Art 1262. See also arts 1263 to 1264.

384  Paula Giliker and damages.42 Such a provision, ambitious in terms of policy in making prevention of the occurrence of a trouble illicite an autonomous function of the French law of tort,43 increases the powers of the courts to intervene and ‘prescribe reasonable measures appropriate to prevent harm or to see that an unlawful nuisance to which a claimant is exposed is stopped.’ Thibierge argues that it should be seen as a natural evolution of tort liability which not only dissuades parties from undertaking anti-social conduct, but also seeks to anticipate and prevent serious and irreversible harm: This preventive function is therefore not new. On the other hand, what is new is that today it appears likely to take on a much larger importance: indeed it can take two forms with distinct functions and purposes: a classic deterrent function, which concerns the prevention of anti-social behaviour and which is founded on the threat of a sanction; and a new anticipatory function, which concerns the prevention of harm and which is founded on the threat of serious and irreversible harms.44

To an external observer, however, such measures may be seen as controversial, raising questions as to the relationship between cessation de l’illicite and other remedial responses. Practical concerns include questions as to the degree of proof required to trigger intervention and the degree of discretion granted to courts to assess the ‘reasonable measures appropriate to prevent harm’. Undeniably, in creating a preventive role for tort, courts are being given a greater role in policing the conduct of others and this will involve in some cases intervention before the party has acted in a tortious manner, in direct conflict with an individual’s right to self-determination. It is here, it is submitted, that comparison with an established system of injunctive relief for tortious actions causing (or likely to cause) harm is particularly useful. Article 1266 of the Projet de réforme, as we can see, leaves much to the courts’ discretion. English courts have long recognised the tension which will arise in choosing to grant injunctive relief instead of more conventional forms of reparation. In particular, the English courts have chosen to adopt a very restrictive approach towards intervention before any wrongdoing has taken place. The reasons for such concerns will be examined in more detail below.

IV.  Injunctive Relief in the English Law of Torts The common law regards injunctions as an important tool by which the court can order the defendant to stop a continuing or recurring act, or order the defendant 42 Sub-section 3 (art 1265) deals with situations where one or more persons are liable for the same harm. 43 Influential here has been the thesis of C Bloch (n 11) which sought to draw out the function played by cessation de l’illicite under the mantle of reparation in French law taking inspiration from RomanoGerman law. 44 C Thibierge, ‘Libres propos sur l’évolution du droit de la responsabilité (vers un élargissement de la fonction de la responsabilité civile)’ RTDC 1999, 561, 583 (original emphasis).

Injunctions Requiring the Cessation of Unlawful Action  385 to act in a certain way.45 The leading practitioners’ text, Clerk and Lindsell on Torts, describes the purpose of an injunction as follows: ‘to prevent the commission of torts which are threatened or anticipated or, in cases of continuing injuries, to restrain their continuance. The principle upon which such injunction is granted is that the injury to be inflicted would be of such a character that the claimant could not practically be compensated in damages.’46 They are used primarily in relation to situations where the tortious behaviour is regularly repeated and continuing, such as nuisance, or where, but for the injunction, the claimant is likely to suffer some form of harm that can never be fully redressed by an award of damages, such as breach of the right to privacy.47 Injunctions, however, as a remedy derive from the courts of equity and, as such, share the characteristic of equitable remedies of being awarded only at the discretion of the court, that is, when it is ‘just and convenient’ to do so. An injunction cannot therefore be demanded as of right, and will not be awarded where damages are an adequate remedy, or where the claimant’s conduct is such that it would not be equitable to make such an award, for example, it would be oppressive, the claimant does not have ‘clean hands’ due to misconduct or the claimant has delayed too long in taking action. Lord Neuberger in the leading case Coventry v Lawrence48 remarked that each case is likely to be fact-sensitive. In true common law fashion, therefore, courts will rely on reported cases as illustrations of circumstances in which particular judges have exercised their discretion, but refuse to give rigid guidance. We might describe this, then, as a structured discretion, whereby the courts are able to respond to the facts of each particular case, but utilising a common set of factors or elements to guide decision-making and ensure some level of consistency between courts. Three different types of injunction exist at common law: (i) prohibitory injunctions which order the defendant not to act in a certain way; (ii) m ­ andatory injunctions which order the defendant to act in a certain way; and (iii) quia timet injunctions which are injunctions granted to prevent a legal wrong before it occurs. These extend to both mandatory and prohibitory injunctions. Prohibitory injunctions are the most common injunction awarded. They deal with situations where, unless it is granted, the defendant is likely to continue acting in a tortious manner.49 Mandatory injunctions are granted more rarely.50 The courts will look at the facts of the case and consider, in particular, whether

45 See, generally, D Bean, I Parry and A Burns, Injunctions, 12th edn (London, Sweet & Maxwell, 2015); I Spry, Equitable Remedies, 9th edn (Cairns, LBC Information Services, 2013) ch 4. 46 M Jones (ed), Clerk and Lindsell on Torts, 22nd edn (London, Sweet & Maxwell, 2017) [29-01]. 47 See NJ McBride and R Bagshaw, Tort Law, 6th edn (Harlow, Pearson Education, 2018) [27.1]. 48 Coventry v Lawrence [2014] UKSC 13; [2014] AC 822, [120], also known as Lawrence v Coventry (t/a RDC Promotions), Lawrence v Fen Tigers Ltd. 49 See Coventry v Lawrence [2014] UKSC 13; [2014] AC 822, [100] and [121] (Lord Neuberger). 50 See Morris v Redland Bricks Ltd [1970] AC 652, 665 (Lord Upjohn). Contrast J Murphy, ‘Rethinking Injunctions in Tort Law’ (2007) 27 OJLS 509, who argues that injunctive relief should be granted more often.

386  Paula Giliker damages would not be a sufficient remedy. For example, the courts are generally more likely to award damages rather than an injunction where the defendant has erected a building in breach of a restrictive covenant, although inevitably much will turn on the facts of each case.51 In contrast, quia timet injunctions are regarded as an extreme remedy to be used only when the possibility of damage is high and imminent. The court will be careful to ensure that the conduct of the defendant is such that substantial damage to the claimant is almost bound to occur, and that damages are not an adequate remedy. In the leading case of Redland Bricks Ltd v Morris,52 the House of Lords held that quia timet injunctions should typically be granted in two particular types of case: (1) where the defendant has, as yet, not harmed the claimant, but is threatening and intending to do so, and if the defendant acts it will cause irreparable harm to the claimant or his or her property; and (2) where the claimant has been compensated for past damage, but alleges that the earlier actions of the defendant may lead to future causes of action. Where, therefore, in Lemos v Kennedy Leigh Development Co Ltd53 the claimants began proceedings in 1959 fearing damage to their property from the roots of trees growing on adjoining land, it was significant that the evidence showed that damage might be expected by 1962. The Court of Appeal held that there was not sufficient imminence of damage to justify an action. In Redland Bricks itself, despite a strong probability of grave damage to Morris’s land in the future, an order requiring Redland Bricks to take all necessary steps to restore the support to Morris’s land within a period of six months was overturned. It had imposed an absolutely unqualified obligation to restore support on Redland Bricks without giving them any indication of what work was to be done and hence how much it would cost. As such, it was deemed to offend a basic principle in the grant of equitable relief of this nature.54 Prohibitory and mandatory injunctions can be given provisionally prior to the final hearing (interim injunctions), or at the final hearing (final or perpetual injunctions). Interim injunctions,55 however, are considered to raise particular concerns due to the fact that the commission of a tort has yet to be determined and the full facts of the case have yet to be set out. The courts deal with this by permitting the award of an interim injunction due to the likely harm of letting the defendant’s conduct continue, but with conditions placed on its grant. For example, it is common for the court to require the claimant to give an undertaking to pay damages to the defendant for any loss suffered while the injunction is

51 Jaggard v Sawyer [1995] 1 WLR 269, 284. 52 Redland Bricks (n 50) 665–66. 53 Lemos v Kennedy Leigh Development Co Ltd (1961) 105 SJ 178; (1960) 175 EG 1199 CA. Cf Hooper v Rogers [1975] Ch 43; [1974] 3 All ER 417. 54 See Redland Bricks (n 50) 667 (Lord Upjohn). 55 See Civil Procedure Rules, r 25. Previously known as interlocutory injunctions.

Injunctions Requiring the Cessation of Unlawful Action  387 in force, should it prove to be wrongfully issued. This may be expensive in practice. For example, a claimant (C) obtains an interim injunction which prevents the defendant (D) from operating his car plant during the night. At the full hearing of C’s claim for nuisance, it is found that the area is classified as an industrial zone and the conduct does not amount to a nuisance. C therefore had no right to an interim injunction, and may find himself liable for the loss of profits experienced by D during the period of the injunction. The principles for the grant of an interim injunction for prohibitory injunctions are set out in the leading case of American Cyanamid Co v Ethicon Ltd.56 Put simply, the claimant must show a serious question to be tried and that, on the balance of convenience, such an order is warranted. It should be noted that the courts are very reluctant to impose such injunctions for claims for libel, despite American Cyanamid. If the defendant has pleaded the defences of truth, honest opinion or qualified privilege, the court will only grant such an injunction when convinced that the defence will fail. The Court of Appeal in Greene v Associated Newspapers Ltd57 confirmed that such a rule was consistent with the European Convention on Human Rights and that a lesser test would seriously weaken the effect of Article 10 ECHR. The English courts have also struggled with the question whether it should be easier to obtain an interim injunction for breach of confidence/misuse of private information.58 In Cream Holdings Ltd v Banerjee,59 the House of Lords found, having regard to section 12(3) of the Human Rights Act 1998,60 that, in most cases, an interim injunction should only be awarded where it is more likely than not that the applicant will succeed at trial, although there could be no single, rigid standard governing all applications in such cases.61 The court may also exercise at its discretion the option to award damages in addition to or in substitution for an injunction.62 Traditionally, the view has been taken that damages instead of (or ‘in lieu of ’) an injunction should only be granted in exceptional circumstances.63 In the leading case of Shelfer v City of London 56 American Cyanamid Co v Ethicon Ltd [1975] AC 396. Although there is some uncertainty here, the courts regard it as even more difficult to justify the grant of a mandatory interim injunction, due to the greater risk of injustice if it turns out to have been wrongly made: see Clerk and Lindsell (n 46) [29–30] and [29–31]. 57 Greene v Associated Newspapers Ltd [2004] EWCA Civ 1462; [2005] QB 972. 58 See PJS v News Group Newspapers Ltd [2016] UKSC 26; [2016] AC 1081. In the Greene case, the Court of Appeal distinguished Cream Holdings Ltd v Banerjee on the basis that defamation and breach of confidence raise different issues: once confidentiality is lost, it is lost for ever. 59 Cream Holdings Ltd v Banerjee [2004] UKHL 44; [2005] 1 AC 253. 60 S 12(3): ‘No such relief is to be granted so as to restrain publication before trial unless the court is satisfied that the applicant is likely to establish that publication should not be allowed.’ 61 Lord Nicholls accepted in Cream Holdings Ltd v Banerjee [2004] UKHL 44; [2005] 1 AC 253, [22], that there will be cases where the courts will be required to depart from this general approach and require a lesser degree of likelihood, for example, where the potential adverse consequences of disclosure are particularly grave or where a short-lived injunction is needed to enable the court to hear and give proper consideration to an application for interim relief pending the trial or any relevant appeal. 62 The jurisdiction was originally given under the Chancery Amendment Act 1858 (Lord Cairns’ Act). See now Senior Courts Act 1981, s 50. 63 See JA Jolowicz ‘Damages in Equity’ (1975) 34 CLJ 224.

388  Paula Giliker ­ lectric Lighting Co, AL Smith LJ famously laid down four conditions which would E lead a court to grant damages in lieu of an injunction: (i) where the injury to the claimant’s legal rights is small; (ii) where the injury is capable of being estimated in money; (iii) where it can be adequately compensated by a small money payment; and (iv) where it would be oppressive to the defendant to grant an injunction.64 The court in Shelfer was keen to resist greater use of damages in lieu of an injunction, as it would amount to a licence to commit a nuisance. More recently, however, the UK Supreme Court in Coventry v Lawrence65 argued that the time had come to signal a move away from a strict approach where an injunction would have serious consequences for third parties, such as employees of the defendant’s business, or, in the case itself, members of the public using or enjoying a speedway racing stadium. While it accepted that the claimants would still be prima facie entitled to an injunction to restrain the defendant from committing a nuisance in the future, it held that it was important not to fetter the discretion of judges. Public utility considerations will therefore be relevant. Lord Sumption added: In my view, the decision in Shelfer is out of date, and it is unfortunate that it has been followed so recently and so slavishly. It was devised for a time in which England was much less crowded, when comparatively few people owned property, when conservation was only beginning to be a public issue, and when there was no general system of statutory development control. The whole jurisprudence in this area will need one day to be reviewed in this court.66

This may be regarded as a clear message from the UK Supreme Court that courts should now be more willing to award damages in lieu of an injunction. At the very least, the court should not feel bound to grant an injunction simply because all four of the Shelfer conditions are not satisfied.67

V.  Comparing English Law and Article 1266 of the Projet de Réforme The above analysis indicates that both systems utilise injunctive relief to protect the interests of victims in the law of tort. Further, English and French law are prepared to go beyond directing a defendant to stop acting in a wrongful manner

64 Shelfer v City of London Electric Lighting Co [1895] 1 Ch 287. ML Wilde, ‘Nuisance Law and Damages in Lieu of an Injunction: Challenging the Orthodoxy of the Shelfer Criteria’ in SGA Pitel, JW Neyers and E Chamberlain (eds), Tort Law: Challenging Orthodoxy (Oxford, Hart Publishing, 2013) argues, however, that the Shelfer criteria were only intended to be a working rule and have wrongly been interpreted too narrowly. 65 Coventry v Lawrence (n 48). 66 ibid [161]. See also Wilde (n 64) who calls for a more flexible and pragmatic approach. 67 Coventry v Lawrence (n 48) [123] (Lord Neuberger).

Injunctions Requiring the Cessation of Unlawful Action  389 and will order the defendant not to commit an anticipated infringement of the claimant’s rights (in English law, quia timet injunctive relief). In contemplating the operation of article 1266 in the light of English law, two particular concerns may be identified. First, how much discretion should article 1266 grant to the courts in determining when to respond to a trouble illicite? Secondly, how should the courts deal with claims for injunctive relief in anticipation of commission of a tort? These concerns will be addressed below from the perspective of comparative law.

A.  How Much Discretion Should Article 1266 Grant to the Courts in Determining when to Respond to a Trouble Illicite? Article 1266 is drafted in general terms and this inevitably gives rise to a concern about how the discretion it grants to the courts will be exercised (an apparent concern in the 2016 consultation proposal which sought to confine it to particular actions).68 In determining, therefore, whether to intervene and what reasonable measures will be ‘appropriate’ to prevent harm or see that a trouble illicite is stopped, the court will be involved in a fact-intensive exercise and will need to consider carefully the measures it recommends. The term mesures raisonables suggests that any order will need to take into account the principle of proportionality, but other relevant factors are less clear. For the English courts, the claimant’s own conduct will be a relevant factor in determining whether to grant injunctive relief. For example, consent to the acts complained of, at least until withdrawn, will deprive the claimant of this remedy,69 as will extreme pettiness in litigation,70 or misleading the defendant and/or the court.71 The French equivalent would be reference to principles of good faith (or even bad faith), which one might assume to be applicable here. The duty on the claimant to mitigate at article 1263 of the Projet72 does also indicate that the courts can look at the conduct of both the claimant and defendant in determining les effets of liability. Would this extend to issues of fault/faute? Further, although harm generally need not be substantial, an English court will rarely award an injunction against trivial harm73 – will such pragmatic concerns similarly influence the French courts? 68 See art 1232 of the Projet de réforme, discussed above. 69 Monson v Tussauds Ltd [1894] 1 QB 671. 70 Tollemache & Cobbold Breweries Ltd v Reynolds (1983) 268 EG 52. 71 Armstrong v Sheppard & Short Ltd [1959] 2 QB 384. 72 ‘Sauf en cas de dommage corporel, les dommages et intérêts sont réduits lorsque la victime n’a pas pris les mesures sûres et raisonnables, notamment au regard de ses facultés contributives, propres à éviter l’aggravation de son préjudice’. See also art 1237 of the Projet: ‘Les dépenses exposées par le demandeur pour prévenir la réalisation imminente d’un dommage ou pour éviter son aggravation, ainsi que pour en réduire les conséquences, constituent un préjudice réparable dès lors qu’elles ont été raisonnablement engages.’ 73 A Burrows, Remedies for Torts and Breach of Contract, 3rd edn (Oxford, Oxford University Press, 2004) 517; Llandudno Urban District Council v Woods [1899] 2 Ch 705. Note, however, that art 1261 excludes réparation en nature in case of manifest disproportionality between its cost for the person liable and its interest for the victim.

390  Paula Giliker On a more theoretical level, the English courts are conscious that requiring a defendant to prevent harm or see that a nuisance is stopped will involve interference with the defendant’s freedom to act and may, for example in the context of defamation or privacy, involve a potential conflict between the reputational and private rights of the claimant and the defendant’s right to freedom of expression. The judge is therefore placed in a position of policing tortious and potentially tortious behaviour, often in circumstances where the legislator has not expressly condemned the activity in question. The response of the English court, as seen above, has been one of caution, ­reiterated by the UK Supreme Court in its most recent authoritative ruling in Coventry v Lawrence. Injunctive relief is discretionary and awarded when damages are inadequate.74 Damages awards are perceived as having distinct benefits, notably of being less intrusive into individual liberty and easier to manage. Here traditional libertarian ideals combine with common law pragmatism. The court, in granting an injunction, must ensure that the defendant knows exactly what he or she has to do or not to do.75 In a classic statement, Lord Nicholls stated that an injunction: … must be expressed in terms which are clear and certain. The injunction must define precisely what acts are prohibited. The court must ensure that the language of its order makes plain what is permitted and what is prohibited. This is a well established, soundly-based principle. A person should not be put at risk of being in contempt of court by an ambiguous prohibition, or a prohibition the scope of which is obviously open to dispute.76

The remedial response to a failure to obey an injunction – contempt of court – is also significant. It is a matter of public policy that a failure to comply with an injunction should face the threat of an order for contempt: otherwise the order of a court of law could be disregarded with impunity.77 The English law of civil contempt is remarkable in that it allows a civil court to imprison or fine the defendant, thus possessing a number of quasi-criminal characteristics.78 A recent example which received considerable press coverage involved a failure by the defendant to demolish a house built without planning permission, which led to a three-month suspended sentence, suspended on the condition that the defendant comply with all the enforcement notices.79 Here, the relationship between court orders and public regulation becomes acute. Commentators have expressed concern at this controversial role for private 74 Coventry v Lawrence (n 48) [120] (Lord Neuberger). 75 Redland Bricks (n 50) 666–67. See also OPO v Rhodes [2015] UKSC 32; [2016] AC 219, [79]: forbidding the claimant to publish ‘graphic’ material lacks sufficient clarity and certainty for injunctive relief. 76 Attorney General v Punch Ltd [2003] 1 AC 1046, [35]. 77 Attorney General v Times Newspapers Ltd [1974] AC 273, 307–09 (Lord Diplock). 78 See, generally, CJ Miller and D Perry (eds), Miller on Contempt of Court, 4th edn (Oxford, Oxford University Press, 2017) [1.16]–[1.24] and ch 12; see also Civil Procedure Rules, r 81. 79 Reigate and Banstead BC v Fidler [2015] EWHC 3863 (QB).

Injunctions Requiring the Cessation of Unlawful Action  391 law judges in granting penal sanctions and noted that it is one which marks a divide with the civil law legal tradition.80 In contrast, the nearest French equivalent of the astreinte applies indirect pressure on defendants to comply with a court order by requiring the defendant to pay a monetary penalty for the period the defendant is in default. While unequivocally punitive, it is certainly not as draconian as contempt of court in that it orders payment of a sum of money rather than a penal sanction. Nevertheless, it does bring a punitive element into a civil action without, as seen in English law, the constitutional safeguards of criminal law and the law of criminal procedure. As we have seen above, it is used in the référé interlocutory procedure81 and will, in turn, be relevant to the enforcement of orders given under article 1266. While it can be argued that the English contempt of court response, whilst initially somewhat startling in a private law court, has survived because it is used rarely in extremis, the same cannot be said of the astreinte procedure. On this basis, the broader the scope of article 1266, the greater the likelihood of the use of this latter procedure. For English law, therefore, restrictive use of injunctions permits justice, but does limit interference with personal liberty and allow the courts to use the simpler and less intrusive remedy of damages when they are deemed adequate. In Coventry v Lawrence, Lord Sumption highlighted that, with greater powers of intervention, goes the responsibility to weigh up concerns which may extend beyond the parties to litigation and involve matters of public interest and the potential infringement of third party rights: There is much to be said for the view that damages are ordinarily an adequate remedy for nuisance and that an injunction should not usually be granted in a case where it is likely that conflicting interests are engaged other than the parties’ interests. In particular, it may well be that an injunction should as a matter of principle not be granted in a case where a use of land to which objection is taken requires and has received planning permission.82

It is unlikely that a civil law court would take the same view as Lord Sumption in Coventry, but his Lordship’s comments do raise fundamental questions about the operation of article 1266. The clear message from the English courts is that too generous an application will give rise to undue interference with the rights of others (the defendant and innocent third parties). Particular difficulties are, as we will see below, considered to arise when ordering a party to act or not to act in anticipation of misconduct.

80 M Chesterman, ‘Contempt: In the Common Law, but not the Civil Law’ (1997) 46 ICLQ 521. 81 S Rowan, Remedies for Breach of Contract: A Comparative Analysis of the Protection of Performance (Oxford, Oxford University Press, 2012) 44–45; F Chabas, ‘L’astreinte en droit français’ Revista de Direito Civil, 1994, no 69, 56. 82 Coventry v Lawrence (n 48) [161].

392  Paula Giliker

B.  How Should the Courts Deal with Claims for Injunctive Relief in Anticipation of Commission of a Tort? The underlying question here is whether the French courts will, in common with English law, be less willing to grant injunctive relief in this context. A tort has yet to be committed. It is unlikely in most cases that one can be certain that it will occur. On this basis, it falls to the law to justify intervention which will interfere with the freedom of individuals with potential negative impact on innocent third parties. Redland Bricks indicates that, for English law at least, intervention can only be justified by requiring clear evidence and the high possibility of substantial damage occurring: [An injunction may be granted] where the defendant has as yet done no hurt to the plaintiff but is threatening and intending (so the plaintiff alleges) to do works which will render irreparable harm to him or his property if carried to completion … Those cases are normally, though not exclusively, concerned with negative injunctions [or] where the plaintiff has been fully recompensed both at law and in equity for the damage he has suffered but where he alleges that the earlier actions of the defendant may lead to future causes of action.83

This highlights a key concern raised by article 1266: will a high level of proof be required or does the law envisage a broader preventive role for the law of tort? In its press release of March 2017, the Ministry of Justice sought to give some assistance by means of an illustrative example.84 Article 1266 will operate where, for example: My neighbour obtains a permit to build a single-storey house. I can prove that he intends to build a two-storey house. Rather than wait until he does this, I will be able to ask the judge to forbid the construction of the extra storey.

Such an example may seem clear-cut, but, in reality, serves to underscore many of the concerns raised in this chapter. It assumes ‘I can prove’ the intention to violate the permit. Yet, in reality, what degree of proof will the courts require to anticipate that my neighbour does indeed intend to build a two-storey house when he denies such an intent? Proving intent is notoriously difficult to establish without corroborative evidence. I may, of course, have good reason to suspect a potential violation, but what if I am wrong, disrupting my neighbour’s building plans and leading him to incur costs/expenses as a result? The problem with this example is its simplicity. It assumes a certainty – an intent to violate a permit – which may not exist in practice. It further does not consider the consequences of a mistake and the potential impact on the personal liberty of the neighbour and other third parties

83 See above (n 50) 665 (Lord Upjohn). 84 Ministry of Justice, Dossier de Presse: Projet de réforme du droit de la responsabilité civile (13 March 2017).

Injunctions Requiring the Cessation of Unlawful Action  393 (eg future tenants, building suppliers) who may be negatively affected by a dispute of this nature. The transfer of article 1266 of the Projet de réforme from the opening provisions of the Avant-projet Terré/2016 Avant-projet proposals to the section on les effets does indicate that the model should be article 809 CPC rather than a more radical reinvention of tort law principle. At the very least, prevention must be balanced carefully with the other aims and objectives of the law of tort.85 While few would question the merits of preventing serious pending harm to an individual’s interests, courts will not have the benefit of hindsight or the time to engage in a broader overview of the interests involved. It would certainly be easy to justify applying article 1266 to cases where there is clear evidence of the prospect of substantial harm to the claimant’s interests. The unanswered question is, in the absence of wording delimiting the provision, will its preventive function permit intervention more generally?

VI. Conclusion The 2017 Projet de réforme is ambitious. It is not simply a matter of bringing the 1804 Code civil up-to-date to reflect the practice of the courts. As the French Ministry of Justice has stated: ‘[i]t proposes … innovative solutions to maintain the preventive function of liability in tort’.86 In this chapter, I have sought to highlight, using English law as a comparator, that however laudable the intentions of the reforms, they need to be considered in more depth. My focus has been article 1266, which states: ‘In extra-contractual matters, independently of any reparation of loss which may have been suffered, a court may prescribe reasonable measures appropriate to prevent harm or to see that an unlawful nuisance to which a claimant is exposed is stopped.’ The aim of this chapter, through a comparison with the common law, has been to move beyond the general wording of article 1266 and consider its likely operation and identify outstanding questions which need to be addressed. In extending the courts’ existing jurisdiction to extra-contractual liability generally, French law is evolving. Nevertheless, English law has advocated a cautious approach and that practical concerns must rest alongside principle. This one provision also raises fundamental questions about how we conceptualise

85 Knetsch and Jacquemin note that while the repositioning does not affect the formal recognition of actions for injunctions, it does suggest that the law of liability has been refocused on the compensation function: J Knetsch and Z Jacquemin, ‘France’ in E Karner and BC Steininger (eds), European Tort Law Yearbook 2017 (Berlin/Boston, De Gruyter, 2018) [6]. See also J Knetsch and Z Jacquemin, ‘France’ in E Karner and BC Steininger (eds), European Tort Law Yearbook 2016 (Berlin/Boston, De Gruyter, 2017) [13]–[23]. 86 Available at: www.textes.justice.gouv.fr/textes-soumis-a-concertation-10179/projet-de-reforme-dela-responsabilite-civile-traduit-en-anglais-30553.html. Translation by Giliker.

394  Paula Giliker tort law – its functions, objectives – but also its place in our society and the extent to which it permits judges (rather than legislators) to interfere with our personal freedom to act. This is not to indicate that la cessation de l’illicite should not be part of the law of tort, but that, in failing to indicate expressly in article 1266 the limits of this discretion, it will be left to French judges to determine how far they should choose to rely on this provision, in responding to unlawful actions. It is ­submitted that English law can provide comparative insights for the French M ­ inistry of Justice and that the reasons and basis for common law caution highlighted in this chapter should not be ignored.

part viii Broad Themes

396

19 The Projet de Réforme du Code Civil Belge and the Reform of the French Civil Code A Comparison of Selected Topics BERNARD DUBUISSON

I. Introduction The reform of the Belgian Civil Code, which was inherited in a direct line from the Napoleonic Code of 1804, was undertaken a little over two years ago under the leadership of the Minister of Justice, Koen Geens. The decision was made to review the provisions dedicated to obligations, proof, property and civil liability. For each of these topics, a working group was set up. The chairing of each group was entrusted to two professors, one French-speaking, the other Dutch-speaking, whose appointment was afterwards confirmed by a formal ministerial decision of 30 September 2017. At a second stage, in their turn, the two chairs identified the members of their own groups taking care as much as possible that there was a balanced representation of the linguistic communities, universities and genders. The working group dedicated to extra-contractual civil liability was chaired by my colleague, Hubert Bocken of the University of Ghent, as Dutch-speaking representative, and by myself, as French-speaking representative. The group, which was originally composed of five university professors and a judge of the Cour de cassation, started their work in June 2016. The Ministry issued no instructions limiting the drafting of the provisions,1 and the group was able to be open to any influence coming from comparative law or from the ius commune in the area. Given the multiplicity of the possible sources of inspiration and the diversity of

1 Clearly this did not mean that the Ministry’s department stayed totally outside the discussions. On certain controversial points (punitive damages, environmental loss etc), recommendations were formulated following ad hoc sessions in the presence of the Minister.

398  Bernard Dubuisson cultures within Belgium itself, no legal tradition took priority at the outset. Indeed, it was not possible in such a country as Belgium for one tradition to dominate over the other. This point was agreed within the group and did not cause the slightest difficulty. A reporter was appointed for each theme which was addressed. The reporter had to present not only the present state of Belgian law but also to cite relevant provisions coming from recent codifications or what is usually called the ‘European ius commune’ as expressed in the Principles of European Tort Law (PETL)2 or the Draft Common Frame of Reference (DCFR).3 The Explanatory Memorandum which accompanies the Projet de réforme belge reveals the sources of inspiration of the legislative provisions which are recommended. These sources vary considerably depending on the subject-matter in question, which shows the absence of any allegiance to a particular legal tradition. A first preliminary draft of the avant-projet, accompanied by an Explanatory Memorandum, was presented on 28 March 2018 to an invited audience of representatives of the judiciary, the Bar, universities and professional stakeholders (such as insurers and lawyers). The avant-projet was later published on the website of the Minister and was the object of a month’s public consultation. Comments and observations were collected and considered by the working group which in certain places decided to amend the draft. In the meanwhile, representatives of the working group met representatives of different interested ministerial departments in order to explain the broad lines of the reform and to collect their observations. The avant-projet together with its Explanatory Memorandum which had been amended following these consultations were submitted to the Ministry of Justice at the end of August 2018.4 Usually such a draft would be submitted to the Council of Ministers in the course of the month of September or October before being presented to the legislative section of the Conseil d’Etat, but political events in December decided otherwise. The fall of the government in effect put an end to any hope of seeing the projet adopted in the course of this legislature. We will, therefore, comment on the most recent version of the projet relating to extra-contractual civil liability as it was communicated to the Minister in August 2018.

2 European Group on Tort Law, Principles of European Tort Law, Text and Commentary (Wien/ NewYork, Springer, 2005); Study Group on a European Civil Code, Principles of European Law, Non Contractual Liability Arising out of Damage Caused to Another (Sellier/Bruylant/Stampfi Publishers, 2009). 3 Principles, Definitions and Model Rules of European Private Law, Draft Common Frame of Reference (DCFR) prepared by the Study Group for a European Civil Code and the Research Group on EC Private Law (Acquis Group) (Oxford, Oxford University Press, 2010), six volumes. 4 The Avant-projet may be consulted, with its Explanatory Memorandum, at http://justice.belgium. be/fr/bwcc (consulted 28 March 2019).

The Projet de Réforme du Code Civil Belge  399

II.  The Structure of the Avant-projet de Réforme In its present state, the provisions relating to extra-contractual liability are intended to be included in the Belgian Code civil5 in Book 5 (Obligations), Title 2 (The Sources of Obligations), Sub-title 2 (Juridically Significant Facts: Les faits juridiques) and Chapter 2 (Extra-contractual civil liability). If the projet on ‘Obligations’ should be adopted before the one relating to extra-contractual liability, Book 5 would reserve the necessary number of articles in this Chapter 2. The chapter relating to extra-contractual civil liability now consists of 70 articles (articles 5.141 to 5.211). In addition, Chapter 3 provides for the amendment of certain provisions scattered outside the Civil Code which we will not discuss further here. Chapter 2 follows a relatively simple plan. It is entirely dedicated to extracontractual liability and is divided into seven sections. In Section 1, which consists of introductory rules, are found general provisions relating to the relationship between actions which rest on distinct legal grounds. Until now, these relationship questions have hardly been addressed in general terms, even though they are essential given the multiplicity of the rules which can be attracted in a particular situation. In principle, as between the rules which may apply, the injured party can choose the one which will serve as the basis of his own claim as long, of course, as their own conditions of application are satisfied (articles 5.141 and 5.142). This first section also contains a new set of rules which govern the difficult question of the concurrence of contractual and extra-contractual liability (article 5.143). These are also inspired by the principle of a free choice in the claimant; we will return to these later.6 Section 2 is dedicated to the grounds of extra-contractual liability and deals for this purpose with the different actions which give rise to liability (les faits générateurs de responsabilité). It is subdivided logically according to the nature of the faits générateurs into several sub-sections dedicated respectively to liability for fault, liability for minors and for the mentally ill, liability for the action of other people and for the action of things and of animals. Section 3 concerns causation. Although the French Projet de réforme de la responsabilité civile is extremely succinct on this complex question, the Belgian avant-projet took the view that it should dedicate several provisions to it and so attempts to put forward detailed solutions, especially where the theory of the equivalence of conditions does not provide clear answers or leads to solutions which are too harsh for the injured party.7



5 In

this chapter Code civil without qualification refers to the Belgian Civil Code. below p 407. 7 On the ‘theory of the equivalence of conditions’, see Ch 7, pp 139–140. 6 See

400  Bernard Dubuisson Section 4 makes provision for harm (dommage).8 Without fundamentally questioning the existing solutions of Belgian law, the avant-projet nevertheless proposes a new vision of harm, which is both more structured and more dynamic, as it is presented as a process which starts off by the infringement of a legally protected interest which is then expressed by repercussions (economic (patrimonial harms) or non-economic (extra-patrimonial harms)) whose extent is to be assessed. Section 5 concerns the effects of liability and addresses questions regarding the reparation of harm.9 Here are found all the general principles which guide reparation (full reparation and reparation in concreto, the primacy of reparation in kind, etc). The main compensatory function of the law of liability is confirmed while any punitive function is ruled out, the working group having received advice on this subject from the Minister. However, without really derogating from the compensatory function of liability, the courts are recognised as being able in certain situations to assess the extent of harm at the level of the amount of profit unlawfully realised by the person liable so as to counter what is called ‘lucrative fault’.10 Recourse actions brought after compensation are the subject of Section 6. This deals only with recourse between two or more persons liable and not recourse actions brought by third parties who have paid the direct victim, the legal grounds of which are very diverse in Belgian law. Both in terms of their nature and their scope, the ordering of these second type of recourse actions certainly deserves to be completely reformed, but this has not been undertaken so far. The last section deals with special regimes of liability. At the moment, there are two in number. First, it concerns strict liability or liability without fault for particularly dangerous activities. This liability is new, but it is without prejudice to particular provisions which already exist. In addition, its entry into force is subjected to the adoption of a Royal Decree (un arrêté royal) listing the activities which are deemed to be abnormally dangerous. The other special regime results from the transfer into the Code civil of the loi of 1991 on liability for defective products11 which implements the European Product Liability Directive of 1985.12 On the other hand, the working group had to leave outside the scope of the reform a series of special compensation regimes whose particular characteristics did not allow them to be integrated in a coherent way within Chapter 2. Here, in particular, the compensation system for road accidents, accidents at work or travelling to work and medical accidents each immediately come to mind.13 These

8 See below p 416. 9 See below p 421. 10 See below p 422. 11 Loi du 25 février 1991 relative à la responsabilité du fait des produits défectueux, Moniteur belge 22 mars 1991. 12 Directive 85/374/EEC of 25 July 1985 concerning liability for defective products, OJ EU L 210/29 (as amended). 13 Loi of 10 April 1971 sur les accidents de travail (compensation of accidents at work and to and from work in the private sector). A compensation regime for ‘weaker road-users’ (such as pedestrians,

The Projet de Réforme du Code Civil Belge  401 systems, which provide a right to compensation for the benefit of the victim of the accident directly either against a private insurer who spreads the risk or against a compensation fund financed by public funds, do not have much in common with the regimes of liability dealt with by the avant-projet.

III.  The Relationship between Contractual and Extra-contractual Liability A.  Distinct Legal Regimes In reading the above comments, it will be understood that the view was taken from the outset that contractual and extra-contractual liability would be governed by distinct provisions, even if these provisions may be based, at least in part, on shared requirements. It should be noted that Belgian case law has not shown itself sympathetic to certain currents in French legal scholarship which advocate a rapprochement between the two orders of liability in terms of the actions which give rise to liability (liability for one’s own actions, liability for the actions of others, liability for the actions of things).14 Moreover, the working group on civil liability considered that the legal duty which is owed by everyone to act in a careful and diligent way does not correspond to the definition of an ‘obligation’ to be provided by article 1 of Book 5 of the proposed Code: An obligation is a legal relationship by virtue of which a creditor can claim before a court from a debtor the performance of the action to be performed [la prestation] and may have recourse, if necessary, to enforcement measures.15

In Chapter 2 relating to extra-contractual liaiblity for one’s own actions, it is therefore more a matter of a legislative duty imposed on a person rather than an obligation in the strict sense. This does not detract from the fact that where the conditions of extra-contractual liability are satisfied, the author of the harm has an obligation to make reparation for it, but we are then speaking about a secondary obligation. The approach of the French Projet de réforme de la responsabilité civile

cyclists and vehicle passengers) was put in place by loi of 30 March 1994, replaced by loi of 13 April 1995 and much amended. This regime was incorporated into the loi of 21 November 1989 relative à l’assurance obligatoire de la responsabilité en matière de véhicules automoteurs, art 29bis. A compensation regime for medical accidents was put in place by loi of 31 March 2010 relative à l’indemnisation des dommages résultant de soins de santé. 14 For a general analysis, see B Dubuisson, ‘Responsabilité contractuelle et responsabilité aquilienne. Comparaison n’est pas raison’ in S Stijns and P Wéry (eds) Les rapports entre les responsabilités contractuelle et extracontractuelle (Bruges, La Charte, 2010) 1–51 and the references mentioned therein. 15 ‘L’obligation est un lien de droit en vertu duquel un créancier peut exiger en justice d’un débiteur l’exécution d’une prestation et recourir, si nécessaire, aux voies d’exécution.’

402  Bernard Dubuisson on this point is very different as it intends to include the whole area of civil liability, both contractual and extra-contractual.16 It should be recognised that the separate treatment of the two orders of liability in the Belgian avant-projet is in part tied to the political choice that the two questions should be confided to distinct working groups. While coordinating meetings between them certainly took place after they had drafted their provisions in order to carry out any necessary adjustments, this was not always easy. As will be seen, this differentiated treatment did not rule out certain overlaps or cross-references. In the context of the Belgian reform, the provisions relating to contractual liability were therefore placed in a first sub-title dedicated to ‘juridical acts’ (actes juridiques), within Section 5 of Chapter 1 on contract. This Section 5 is entitled ‘Non-performance of contractual obligation and its consequences’. It is sub-divided into two sub-sections dedicated respectively to ‘Non-performance imputable to the debtor’ (articles 5.85 to 5.101) and ‘Non-performance not imputable to the debtor’ (articles 5.102 to 5.105). It should be added that this section cross-refers on a number of occasions to sub-title 6 which deals in a very general way with the non-performance of any obligation whatever its nature (articles 5.298 to 5.315). It can also be seen that, in their turn, the provisions relating to extra-contractual liability are appropriately inserted into Chapter 2 of sub-title 2 dedicated to ‘juridically significant facts’ (faits juridiques). The relationship which Chapter 2 bears with this sub-section 6 relating to the non-performance of any obligation is, as will be seen, a source of discussion. In any case, the presentation of the Belgian avant-projet raises a number of questions of internal relationship of the two kinds of liability.

(i)  ‘Extra-Contractual Faults’ which are not Covered by the Chapter Relating to Extra-Contractual Liability a.  ‘Abuse of Rights’ in the Contractual or Extra-Contractual Context The abuse of rights is dealt with by article 5.7 of the introductory title which concerns general provisions. Under the Belgian Code civil, the idea according to which an abuse of rights in the contractual context finds its source in article 1134(3) of the Code (that is to say, good faith) while the abuse of rights in the extracontractual context relies on article 1382 of the Code was gradually developed both by the courts and by legal scholars.17 Within the scheme of the reform, the abuse

16 See ch 2 above, pp 21ff. 17 For an overview on the abuse of rights and the moderating function of good faith in Belgian law see P Wéry, Droit des obligations (vol. 1), Théorie générale du contrat (Bruxelles, Larcier, 2011) 136, [112]–[116]; P Van Ommeslaghe, Droit des obligations (tome 1), Sources des obligations (Bruxelles, Bruylant, 2010) 52, [21]ff.

The Projet de Réforme du Code Civil Belge  403 of rights is now addressed by a general provision applicable to all o ­ bligations, an approach which is entirely justified. Article 5.7 makes clear that: A person commits an abuse of a right where he exercises it in a way which manifestly goes beyond the limits of the exercise of the right in question by a careful and r­ easonable person situated in the same circumstances.

This solution may clearly be defended to the extent that the prohibition of the abuse of a right is presented as a general legal principle applicable whatever the sourse of the obligation. In addition, the reference to the ‘manifest’ exceeding of the right does not fit with the traditional criteria of extra-contractual civil fault as they are set out by the new article 5.148(2) of the avant-projet belge on ­extra-contractual liability. b.  Extra-Contractual Faults Related to the Conclusion of Contracts or their ‘Opposability’ Pre-contractual liability in the case of the wrongful breaking-off of negotiations is dealt with in Chapter 1 relating to the process of formation of contracts (article 5.21), even though it is uncontested in Belgian law that the ground of this liability is found in articles 1382 and 1383 of the Code civil. A reading of article 5.21 shows clearly that the extra-contractual character of this liability is not challenged, it simply being thought more consistent to deal with this matter in the section relating to the conclusion of contracts. Otherwise, article 21 alludes to the criterion of the legitimate reliance of the future contracting party in order to justify the reparation of the loss of net benefit expected from the contract, a criterion which is also not listed by article 5.148(2) for the purposes of assessing extra-contractual fault. In Belgian law, liability of a third party accessory who assists in the breach of a contractual obligation also finds its basis in articles 1382 and 1383 of the Code civil. The Cour de cassation considers, however, that liability in a third-party accessory cannot be imposed unless the latter knew of the existence of the contract undertaken by the contractual debtor and nevertheless knowingly assisted in breach of the obligation.18 However, it is not required that the third party acts with intent to harm the person who contracts with the person with whom he deals. According to a majority of legal scholars, in this way the Cour de cassation has approved the thesis of ‘simple deceit’ (la fraude simple).19 Again, this requirement does not fit with the traditional criteria of extra-contractual fault, as in principle merely ‘minor fault’ (faute légère) is enough to justify an obligation to make reparation. Third-party complicity is therefore made the subject of a specific provision under article 5.114 in the section dedicated to the effects of contracts on third parties. Once more, on reading this article the extra-contractual character of the fault of the third party accomplice is not put in doubt. 18 Cass 22 April 1983, RCJB 1984, 359 note Y Merchiers, RW, 1983–1984, col 427 note E Dirix; Cass 28 November 2002, RGDC 2004, 402. 19 See Wéry (n 17) 618, [654]ff and the references mentioned therein.

404  Bernard Dubuisson c.  The Regulation of Contract Terms Excluding Liability Within the framework of the avant-projet de réforme relating to obligations, the rules governing terms which exclude or limit liability, whatever their scope, are provided for by article 5.92 in the sub-section relating to contractual nonperformance imputable to the debtor. This is justifiable as an exemption clause always results from agreement. Moreover, article 5.92 is careful to make clear that the contract term in question can exonerate the debtor from his extra-contractual as well as from his contractual liability. For the rest, a mention may be found in the Explanatory Memorandum according to which, in the absence of a contract term to the contrary, an exclusion clause formulated in general terms covers both contractual and extra-contractual liability. This solution can be supported by reference to the principle of the autonomy of the will, but it does not necessarily follow from the principle of the strict interpretation of contract terms which derogate from the law generally applicable. As regards extra-contractual liability, it is interesting to note that article 5.92(1)2° deems ‘not written’ any contract term which exonerates a contractual debtor from his own fault or from the fault of a person for whom he is liable where this fault leads to the death or personal injury of another person.

(ii) Contradiction and Duplication More difficult are situations where the separate treatment of contractual and extra-contractual liability may give rise to contradiction or duplication in the rules applicable, this being particularly the case as regards the actions which give rise to liability, harm and reparation. The structure of the two avant-projets can cause some uncertainty in this respect. In the avant-projet on obligations, articles 5.85 to 5.105 concern, as has been said, the non-performance of contractual obligations, and articles 5.298 to 5.315 concern the non-performance of obligations in general. At the same time, the avant-projet on extra-contractual liability also contains many provisions relating to the actions which give rise to liability, to harm and to reparation. How do all these provisions relate to each other? A priori, articles 5.85 to 5.105 set out rules specific to the non-performance of contractual obligations, and article 5.298 makes clear that its own provisions on the non-performance of obligations in general are without prejudice to these particular provisions on contractual non-performance. This leaves the question, though, of the relationship between the rules governing extra-contractual liability and those which concern the non-performance of any obligation. a.  Foreseeability of Harm Certain special features of contractual liability have been rightly preserved. Thus, the right to performance in kind of an obligation, the right to terminate the contract for non-performance, the right to price reduction or the right to suspend

The Projet de Réforme du Code Civil Belge  405 performance of one’s own obligation are set out by article 5.86 as possible sanctions arising from non-performance of contractual obligations. The provisions relating to the reparation of harm merely provide an opportunity of recalling the principle of full reparation and the modes of reparation of a harm (in kind or in monetary form). By announcing that ‘only harm which the parties were reasonably able to forsee the principle [dans son principe] at the time of the conclusion of the contract must give rise to reparation, in the absence of intentional fault on the part of the debtor’, article 5.90 retains the principle announced by article 1150 of the Code civil and so preserves a particular special feature of contractual liability. Certainly, the foreseeability of harm also forms one of the criteria of extra-contractual civil fault in the Belgian projet (article 5.158(2)), but there it is in the context of one of several other criteria. In addition, its significance is rather different as there the forseeability of harm is assessed at the time of the action giving rise to liability and not at the time of the conclusion of the contract. b. Imputability, Force Majeure etc The status of the rules contained in articles 5.298 to 5.315 in the sub-title dedicated to the non-performance of obligations is a lot less clear, as this includes shared rules or concepts which also feature in the avant-projet relating to extracontractual liability (imputability, force majeure, full reparation of harm, the duty to limit one’s own harm, reparation in kind, etc). Where they conflict, article 5.311(2) makes clear rather hastily that ‘articles 141 et seq relating to extra-contractual liability apply unless their nature or their scope are incompatible with such an application’. This means, therefore, that all the rules relating to extra-contractual liability are in principle applicable to contractual liability unless the contrary is stated or they are manifestly incompatible. It is by no means certain that this approach is enough to answer all the questions which may arise. Certainly, it can be said that the principle of compatible application would rule out the application of provisions governing the actions which give rise to liability as articles 5.299, 5.303 and 5.304 themselves determine those actions which give rise to liability as regards contractual obligations (liability for non-performance involving fault, contractual liability for another person’s actions, contractual liability for the actions of things). However, the use of the term ‘imputability’ in articles 5.85 and 5.299 of the avant-projet on obligations is problematic given that this concept has been deliberately banished from the avant-projet on extra-contractual liability. In the sub-title relating to the non-performance of obligations in general, article 5.299 makes clear that non-performance is ‘imputable’ to the debtor only if a fault can be attributed to him whether he is liable by virtue of legislation or by virtue of a juridical act. Further on, article 5.303 deals with the imputability of a fault committed by a person who helps perform one’s obligation (contractual liability for the actions of other people) while article 5.304 deals with imputability arising from use of ­defective things in performance (contractual liability for the actions of things).

406  Bernard Dubuisson It is well-known that the notion of imputability is particularly ambiguous. Sometimes we understand it to refer to the fact that liability for one’s own action involves a free and conscious will (une volonté libre et consciente), this expressing the need for a intentional element (un élément moral) as a necessary element of fault; sometimes we understand it as determining much more broadly all the conditions which are imposed for the attribution of the consequences of a harmful action to a particular person who must therefore be liable for them. Articles 5.303 and 5.304 very clearly understand the term ‘imputability’ in this second sense, as also does the French Projet de réforme de la responsabilité civile whose subsection 2 is entitled ‘The imputation of harm caused by another person’. On the other hand, article 5.299 of the Belgian avant-projet on obligations is much less clear because it does not set out the elements which constitute contractual fault. On this point, the avant-projet relating to extra-contractual liability has made an explicit choice. Without prejudice to the special rules governing the liability of minors or adults who lack mental capacity as well as the grounds of justification provided by article 5.150, no ‘moral’ or intentional element is retained as one of the constituent elements of extra-contractual liability. It remains the case that a person who commits such a fault still has the possibility of not being liable by proving the existence of a ground of justication. Time will tell whether contractual and extracontractual fault will be assessed differently on this point, although there seems to be no ground for justifying any difference here. Similarly, force majeure is given a definition by article 5.300 on the nonperformance of obligations and another definition by article 5.149 by way of the grounds of exoneration from extra-contractual liability for fault. The Explanatory Memorandum on this article explains this by pointing out that force majeure does not have the same purpose in the contractual and extra-contractual contexts as, in the latter, it does not consist of the impossibility of performing one’s obligation in the sense set out above but rather the impossibility of conforming to a rule of conduct. Beyond this, the criteria of assessment of force majeure are much the same. The wording of article 5.149(2) was revised to this effect after the public consultation. c.  The Obligation to Reduce One’s Own Harm There is another curiosity. Article 5.312 of the avant-projet on obligations gives an explicit and general recognition to the duty of a creditor of an obligation to take reasonable measures to reduce or prevent the harmful consequences of nonperformance, whereas the avant-projet on extra-contractual liability says nothing about this at all. The Explanatory Memorandum on the latter avant-projet explains this silence by noting that, according to the case law of the Belgian Cour de cassation, in the extra-contractual context there is no independent obligation for the victim to minimise his own harm, it being up to him merely to behave in a way which a reasonable and careful person would do, which, in short, is not very

The Projet de Réforme du Code Civil Belge  407 much.20 Article 5.176 of this avant-projet is content therefore to regulate who bears the costs of measures to prevent harm, as also does article 5.312(2). It remains to be seen whether article 5.312 which governs all obligations does not eventually recognise an independent obligation on the victim to minimise his own harm even in the extra-contractual context.

B.  The Coexistence and Concurrence of Liabilities (i) Coexistence By coexistence, we refer to the possibility for a third party to a contract to seek the imposition of an extra-contractual liability on a person who has failed in one of his contractual obligations (for example, a safety obligation or maintenance obligation) with the view to gaining reparation for the harm which this failure has caused him. Belgian law has not experienced the doubts and changes on this point seen in the French case law.21 It is clearly accepted that such an extra-contractual action is possible on condition that the third party/victim establishes that the contractual failure which was the source of his own harm constituted at the same time and independently of the contract, a breach of the general duty of care imposed on everyone.22 The same harmful action can therefore constitute both a contractual fault as regards the contractual creditor and an extra-contractual fault as regards a third party. This principle is so well-established that the drafters of the avantprojet did not believe that it should recognise it in any particular provision. This result does no more than confirm the general rule of liability for fault contained in article 5.146.

(ii) Concurrence a.  Extra-Contractual Actions between Parties to a Contract It would be tedious to try to summarise the developments in the case law of the Belgian Cour de cassation on the question of the concurrence of liabilities which is assuredly one of the most difficult in the law of liability. Suffice it to say here that in its most recent phase the Cour de cassation has opted for the solution of a restricted concurrence, an intermediate solution between a pure and simple ­prohibition of

20 Cass 14 May 1992, Pas 1992, I, 798; RGAR, 1994, no 12.1312; in the same sense, in France, Cass civ 19 June 2003, RTD civ 2003, 716 obs P Jourdain. 21 On the French position see ch 4 above, pp 56–58. 22 Cass 12 December 1958, Pas 1959, I, 383, RCJB 1960, 204 note G Van Hecke; Cass 11 June 1981, Pas 1981, I, 1159; Cass 14 May 1981, Pas 1981, I, 1669; Cass 21 January 1988, Pas 1988, I, 602, JT 1989, 111, RGAR 1989, no 11563, RW 1988-1989, 675 note E Dirix; Cass 29 June 1989, Rev prat soc 1989, 175; Cass 25 October 1990, RCJB 1992, 493 note RO Dalcq, RGAR 1992, no 11990; Cass 26 March 1992, JLMB 1994, 38 obs D Philippe, Pas 1994, I, 675; Cass 20 June 1997, RW 1998–1999, 435.

408  Bernard Dubuisson a choice between bringing a contractual and an extra-contractual action (thereby recognising the primacy of the contractual action) and complete freedom of choice (thereby preserving the extra-contractual action). This restricted concurrence would allow a choice but only on certain conditions. Of course, the question of an option between the two types of action cannot arise unless the conditions for each of the two liabilities are satisfied at the same time. At the moment, and after considerable uncertainty, the Cour de cassation therefore accepts concurrence or an option as long as two conditions relating respectively to fault and to the harm are satisfied.23 First, it is necessary that the harmful action constitutes not only breach of a contractual obligation, but also breach of the general duty of care which is imposed on everyone, and this assumes the existence of hybrid obligations or mixed faults which have both a contractual and an extra-contractual nature (for example, contractual safety or information obligations). It is then necessary that the harm is different from the harm which results from the contractual non-performance. The form of words used is quite obscure here, as the harm does not itself have a properly contractual or extracontractual nature. According to one interpretation, the Cour de cassation seems to have in mind a harm different from the one which results from the loss of economic benefit expected from the contract. This case law therefore suggests that there exists in contracts an impregnable part which can never give rise to an extracontractual action. This concerns the core economic exchange put into effect by the contract, its hard core, if one likes. On the other hand, if a claimant’s harm is distinct from his loss of the economic advantage expected from the contract, its reparation can be claimed by way of extra-contractual action. To the principle so laid down, the Cour de cassation makes an important exception: where the fault constitutes both a criminal offence (for example, involuntarily causing personal injury), the extra-contractual action is always possible.24 The justification for this exception remains unclear given the uncontrolled multiplication of criminal offences. The Belgian avant-projet de réforme on the question of concurrence adopts a revolutionary position which totally upsets the solutions of the Cour de cassation. It starts from a view that the rejection of extra-contractual actions between parties to a contract rests on an analysis of the implied or presumed will of the parties. The latter would not wish that what they have planned for in the contract (their prévisions) would be foiled by the bringing of an extra-contractual action by one of them which would upset their bargain. If an action claiming extra-contractual liability were available without restraint, people who have chosen to bind themselves by

23 Cass 29 September 2006 RG no C03.0502.N; see also Cass 17 March 2017 RG no C.16.0283.N. For a summary of the development of the case law of the Belgian Cour de cassation, see P Van Ommeslaghe, Droit des obligations (tome 1), Sources des obligations (Bruxelles, Bruylant, 2010) 1161, [813]. 24 Cass 26 October 1990, RCJB 1992, 497, 503 note RO Dalcq, ‘Restrictions à l’immunité de responsabilité de l’agent d’exécution’, Pas 1991, I, 216; see also Cass 28 June 1982, Pas 1982, I, 27; Cass 25 April 1983, Pas 1983, I, 958; Cass 1 June 1984, Pas 1984, I, 1202.

The Projet de Réforme du Code Civil Belge  409 contract would escape it with little difficulty by choosing this option. Under article 5.143, the avant-projet fundamentally challenges the existence of such an implied will, which in reality rests on a fiction especially since the renunciation of a right is not to be presumed. Moreover, it does not support the idea that it is necessary to protect contracts from intrusion by extra-contractual liability. Adopting exactly the opposite position adopted by article 1233 of the French Projet de réforme, article 5.143(1) as amended provides that ‘an injured party may invoke against his co-contractor the rules of extra-contractual liability as long as this possibility is not excluded either by legislation or by a contract’. First, therefore, an option is very widely available: this is not the case only where legislation or a contract explicitly forbids this option. The will to exclude the rules of extracontractual liability is not to be presumed: it must result clearly from a legislative provision or from a contract term. However, article 5.143(2) adds a qualification to this principle, stating that: However, if [the injured party] claims reparation for harm which results from the nonperformance of a contractual obligation, any special legislative provisions or contract terms specifically applicable to the obligations of the parties prevail over the rules of extra-contractual liability.

A contracting party who claims reparation of harm resulting from non-performance of a contractual obligation is therefore prevented from circumventing any special legislative provisions or contract terms which govern the parties’ obligations and which are therefore understood by the parties to apply to their own situation. These rules would apply despite any choice made by the claimant, as otherwise they would lose their point. Thirdly, the final part of article 5.143 excludes the rule of the primacy of contract where the claimant’s harm consists of personal injury. This qualification of the general rule allowing an option replaces the similar solution formerly accepted by the Cour de cassation for cases of criminal offences (a position which was difficult to understand). The idea behind this is that harm resulting from injury to a person’s physical integrity is, by its nature, a matter for extra-contractual liability, and that therefore in this area the existence of a contract cannot reduce the protection of those rules which protect victims against personal injuries. In this respect, it should be recalled that article 5.92(1)2° deems ‘not written’ contract terms which exonerate a debtor where his fault or the fault of a person for whom he is ­responsible causes death or personal injury. This change, which is inspired by Dutch and German laws, is, as will be seen, quite radical and is not without difficulty. The notion of ‘any special legislative provisions or contract terms specifically applicable to the obligations of the parties’ would be subject to interpretation. The idea is that, where a party to a contract has chosen to bring a delictual action, the general prescription rule provided for all contractual actions would not be applied as a matter of priority whereas a special prescription rule, for example governing a contract of transport, would be. The same is true of the general rules governing contractual liability such as, for

410  Bernard Dubuisson e­ xample, article 1147 of the Code civil belge which places the burden of proof of an external cause (cause étrangère) on the creditor of the obligation which has not been performed, in contrast to the specific rule applicable to contracts of graduitous deposit whose purpose is to restrict the depositee’s liability by making clear that the latter need not take the same care of the thing deposited as he would wish in his own affairs (articles 1927 and 1928 of the Code). The purpose is to avoid the general availability of an option between the two liabilities allowing the systematic circumvention of particular legal rules or contract terms. As a result, if the depositor chooses to bring an action to establish the liability of the depositee on an extra-contractual basis, articles 1927 and 1928 would nevertheless be applied.25 It should not be overlooked either that the rule of the primacy of contract laid down by article 5.143(2) could lead in particular proceedings to a mixture of the rules applicable. Indeed, this provision gives free rein to an extra-contractual action where legislation has not taken the trouble to regulate specifically the contractual relationship in issue, if the parties have not provided anything in particular in their contract or if their contract was not in writing. b.  Extra-Contractual Actions Against the ‘Agent for Performance’ of the Principal Debtor One of the results of the solution which has just been set out is that it puts an end to the case law on the immunity to civil liability of ‘agents for performance’, an area which has been much discussed in Belgium. An agent for performance is a person who is engaged by a contracting party to perform one or more of his contractual obligations, as in the case of a sub-contractor or employee. In a famous judgment in 1973, the Cour de cassation decided that an extra-contractual action brought by a creditor against an agent for performance of his own contracting partner (the agent’s principal) was itself subject to the established two conditions for permitted concurrence.26 As these conditions (as to fault and harm) were restrictive, legal 25 The Belgian Code civil contains other special rules which restrict or make more onerous the conditions of contractual liability as compared with the general law of contract. Illustrations may be found in arts 1735 and 1736 (which provide that the tenant is responsible for losses caused by a fire occurring on the premises let unless he proves that it could have taken place without his fault); art 1891 (which provides that a lender is liable for damage caused by the thing lent only if he knew of the defects and did not warn the borrower); art 1952 (which provides that a hotelier’s liability is limited to a hundred times the amount of the price per day of the accommodation); and art 1992 (which governs the liability of gratuitous agents). According to the solution advocated by the avant-projet, these rules would not be excluded even if the claimant chooses to ground his action on an extra-contractual basis. The same solution applies in German law, the BGB also containing specific rules of the same type as the ones which have just been noted: eg §§ 521, 599, 690 and 708 BGB. 26 Cass 7 December 1973, Pas 1974, I, 376, RW 1973–1974, col 1597 obs J Herbots, RGAR 1974, no 9317 obs J-L Fagnart; Cass 3 December 1976, RGAR 1978, no 9908, RW 1977-1978, col 1303 note Van Oevelen, RCJB 1978, 431 note RO Dalcq and F Glansdorff; Cass 13 September 1977, RCJB 1978, 426 note RO Dalcq and F Glansdorff, RGAR 1979, no 10041; Cass 8 April 1983, Pas 1983, I, 834, RW 1983-1984, col 163 note J Herbots; Cass 26 October 1990, Pas 1991, I, 216, RCJB 1992, 497 note RO Dalcq; Cass 13 April 1984, Pas 1984, I, 1022.

The Projet de Réforme du Code Civil Belge  411 scholars considered that the agent for performance benefited from an immunity recognised by the courts as the contractual action could not in principle be brought against him as he was a third party to the main contract and the extra-contractual action was not allowed except subject to these conditions. In this respect, the agent for performance is treated in the same way as if he were the main party to the contract. At a later stage, the Cour de cassation even extended this case law to the controlling bodies of legal persons such as companies.27 This case law, which is particular to Belgium and of which there is apparently no trace elsewhere in Europe, is based on the idea that the agent for performance of a contracting party is not really a third party to performance of the contract and that the creditor cannot therefore obtain performance by means of an extra-contractual action except subject to the two conditions applicable to the concurrence of actions. However, one of its results is that it places the creditor in great difficulties in the situation of the insolvency of the main debtor or where the latter can rely on a personal ground of exoneration, such as a limitation or exclusion clause. Once again, it is doubtful that the creditor in question has impliedly renounced the right to bring an extra-contractual action against the agent for performance of whose involvement in performance of the contract he is often unaware. The avant-projet therefore also overturns this case law for the future by subjecting any extra-contractual action against an agent for performance to the same regime as applies to such an action between contracting parties as earlier explained. Such an action is therefore in principle available in the absence of any legislative provision or contract term to the contrary. In this way agents for performance lose the protection which they have benefitted from under the caselaw of the Cour de cassation and this would allow, for example, a person who commissions a building to bring an extra-contractual claim against a building sub-contractor who is at fault where the main contractor is insolvent. Some protection nevertheless remains for agents for performance as article 5.92(2) of the avant-projet on obligations provides that ‘if the debtor has recourse to auxiliaries for the performance of the contract, the latter can rely against the principal creditor on any exclusion clause agreed between him and the debtor’.

IV.  Extra-contractual Fault A.  The Definition and Grounds of Fault The Belgian avant-projet de réforme dedicates three articles to extra-contractual fault (articles 5.146 to 5.148). The following provisions (articles 5.149 to 5.152) concern grounds of exoneration which will not be discussed here. Fault remains the basis of liability for one’s own action (le fait personnel) under the general law.

27 Cass

7 November 1997, RGDC 1998, 153.

412  Bernard Dubuisson First, article 5.146 reproduces almost identically the famous article 1382 of the Belgian Code civil: it merely suppresses the reference to the action of a man (le fait de l’homme) so as to respect gender equality and to allow the application of the provision to legal persons. The article states therefore that ‘any person who, by their fault, causes harm to another is bound to make reparation’. Since it makes no distinction between them, article 5.146 puts on the same footing positive action and omission. It does refer explicitly to the condition of unlawfulness of the harmful action as one of the constituent elements of fault as it is entirely understood that action becomes unlawful where it falls into one of the situations foreseen by article 5.147, as will later be explained.28 Nor does article 5.146 make any distinction according to the seriousness of the fault, whether it is minor, serious or intentional (faute légère, grave ou intentionnelle). Reparation is for the full harm and is therefore not proportionate to the seriousness of the fault. All this reflects the present state of Belgian law. However, unlike articles 1382 and 1383 of the Code civil, article 5.147 takes a step further by trying to define fault as a ‘failure to conform to a rule of conduct resulting from legislation (in a broad sense29) or to the general duty of care which it must be respected in social relations. A failure to abide by a norm of behaviour therefore constitutes the common denominator of all extra-contractual fault. The identification of two grounds of fault is directly inspired by the traditional definition of le fait fautif advocated by the Belgian Cour de cassation, according to which fault is: Either an act or abstention which breaks a norm of international law having direct effect in the national legal order or a norm of domestic law which imposes on the persons who are subject to it a duty to abstain or to act in a particular way. Or it is an act or an abstention which, without constituting a failure to conform to norms of this kind, consists of an error of conduct, this being assessed according to the criterion of a normally careful and prudent person placed in the circumstances.30

Article 5.147 does not recognise any other grounds of fault. As a result, the existence of fault cannot be deduced merely from the violation of an individual right (un droit subjectif, for example, a right of property, a personality right or intellectual property rights). In this situation, the ‘material aspect’ of fault certainly corresponds to the breach of the duty of respect stemming from the fact that an individual right may be relied on against third parties (which is equivalent to breach of a particular rule of conduct). The recent case law of the Cour de cassation also suggests that knowledge of another person’s individual right is not 28 Below, pp 412 and 414. 29 The French here is ‘au sens matériel du terme’ (that is, which create rules of law) which is to be distinguished from ‘au sens formel’ (that is, which are ‘legislative’ in the sense of being created by the body invested with the power to make them). 30 See notably Cass 13 May 1982, JT 1982, 772 concl Velu, RCJB 1984,19-31 obs RO Dalcq; Cass (1) 8 November 2002, Pas 2002, I, 2136; RW 2004–2005, 1259; RABG 2003, 851 note S Lust; Cass (3) 25 November 2002, Pas 2002, I, 2230.

The Projet de Réforme du Code Civil Belge  413 necessary in order to establish the existence of fault.31 In this situation, the person who has infringed the right nevertheless retains the possibility of relying on one of the grounds of exoneration provided for by articles 5.149 and 5.150, including unavoidable mistake (erreur invincible).

B.  The Constituent Elements of Fault In view of the current state of Belgian law, article 5.147 is more interesting for what it leaves out than what it says. It does not refer at all to any ‘moral’ or subjective element of fault and does not require that the breach of the rule of conduct took place freely and consciously. Until now, fault in Belgian law has assumed the existence of a free and conscious will as well as a capacity of discernment, that is to say, the capacity to foresee the harmful consequences of one’s own acts, this capacity being assessed case by case: there has been no legal age of discernment. In future, fault will be reduced to its material or objective element (breach of a rule of conduct), which would bring Belgian law closer to French law. However, the ‘moral element’ of fault has not completely disappeared from the Belgian legal landscape, but it is given expression by other juridical means which are apparently more appropriate. Articles 5.153 to 5.155 govern directly the liability of two categories of people who at the present time can be seen as lacking discernment: minors (articles 5.153 and 5.154) and persons affected by a mental problem (article 1.155). Otherwise, in a more indirect way, articles 5.149 and 5.150 identify the main causes of exoneration of liability for one’s own action, including grounds of justification. These involve the classic situations in which the person whose action is in question is the victim of unforeseeable external circumstances or who has not acted at all freely (including cases of force majeure, unavoidable mistake, necessity, legitimate defence, or the command of legitimate authority of legislation). However, under article 5.151, the burden of proof as to such a ground of justification lies in the civil context on the person who seeks to rely on it. A new article 5.153 establishes clearly that minors of less than 12 years are not liable for harm caused by their fault or by any other action otherwise giving rise to liability, and this avoids the need to consider their discernement on a case-bycase basis. On the other hand, minors of 12 years or more are liable personally for harm which they may cause, although courts have a power to reduce the amount of any compensation awarded (article 5.154). People affected by a mental problem are subject to the same regime as minors of more than 12 years (article 5.155). In their case, however, this is not new as this special regime of liability already applies to them under article 1386bis of the Belgian Code civil, introduced by legislation in 1935.32



31 Cass. 32 Loi

9 February 2017 no C.13.0143.F. of 16 April 1935 sur la réparation causée par les déments et les malades mentaux.

414  Bernard Dubuisson The position of people who loose consciousness temporarily as a result of a medical condition (for example, cardiac arrest or an epileptic event) is not explicitly covered and they therefore cannot benefit from article 5.155. This suggests that, irrespective of their loss of consciousness, they may commit a fault and so be held personally liable, as long as the conditions provided by articles 5.147 and 5.148 (breach of a rule of conduct) are satisfied and in the absence of any ground of exoneration, such as force majeure or unavoidable mistake.

C.  Criteria for the Assessment of Fault By way of continuation of article 5.147, article 5.148 addresses the question of the criteria to be used in the assessment of fault where it arises from breach of a rule of conduct. If the fault finds its origin in the contravention of legislation (in a broad sense), it is important to examine whether the rule in question requires particular behaviour. If so, fault is to be deduced from breach of this rule and in such a case, there is no place, therefore, for having any recourse to the criterion of the reasonable and prudent person nor to consider the appropriateness or the efficacity of the legal rule in question. In this way, for example, if the Road Traffic Code (Code de la circulation routière) prohibits driving on motorways at a speed above 120 kilometres per hour, the mere fact of exceeding this speedlimit constitutes fault. The same applies if the general regulation for safety at work requires workers to wear a safety harness where they work at a height of more than 2 metres.33 Moreover, article 5.148 does not require in addition any consideration of the range of people to be protected by the rule in question nor the nature of the harm which it is intended to avoid. The theory of ‘Aquilian relativity’34 is therefore not given formal expression, reflecting the fact that it is not recognised in Belgian law. Moreover, it appears difficult to avoid taking into account the nature and the significance of the rule which is broken where fault results from the contravention of legislation. However, not all rules require specific behaviour on the part of the person to whom they are addressed. Some of them require only an obligation to take care (obligation de moyens) and do no more than repeat in different terms the general duty of care imposed on everyone. Where this is the case, the criteria relating to the second ground of fault are instead applied. If there is no rule requiring specific

33 Cass 22 September 1988, Pas 1989, I, 83. 34 Or ‘Schutznormtherie’ in German law. The theory of ‘Aquilian relativity’ (la théorie de la relativité aquilienne) (named somewhat distantly after the Lex Aquilia in ancient Roman law which imposed liability for damage caused wrongfully) has been put forward as a reason for restricting the range of persons or harms for which a person should be liable for their own personal fault: see, eg, G-H Lankhorst, De relativiteit van de onrechmatige daad, (Deventer, Kluwer, 1992); D-M Philippe, ‘La théorie de la relativité aquilienne’, Mélanges offerts à R.O. Dalcq, Responsabilités et assurances (Bruxelles, Larcier, 1994) 467–86.

The Projet de Réforme du Code Civil Belge  415 behaviour or if the rule whose breach is in issue does not so require, fault is to be deduced from breach of the general duty of care which is imposed on everyone in social relations and which is therefore assessed by reference to the behaviour of a prudent and reasonable person in the same circumstances (article 5.158(2)). In this second situation, it is for the court to apply this standard of conduct which should have been followed to the circumstances of the case and to ascertain whether or not the person who caused the harm conformed to it. In this respect, there is no change from the present position. In a way which is original compared to the French Projet de réforme, article 5.158(2) continues by setting out some factors to which the court may have regard in assessing the existence of fault. The criteria which are proposed are neither exhaustive nor cumulative and should be seen as no more than a way of helping the court in its approach to decision-making. The following may be taken into consideration for this purpose: the foreseeability of the harm and of its extent, the cost or effort necessary to avoid the harm, the state of scientific and technical knowledge, professional standards or good practices as well as principles of good administration and good management. This new approach was the object of a number of criticisms during the public consultation. For example, for some people the reference to the magnitude of the cost and effort necessary to avoid the harm constituted the introduction of a ‘law and economics’ approach into the law of Belgian civil liability which is instead oriented towards the protection of victims. However, it does not seem unreasonable to rule out the existence of any fault in the head of a commercial operator where the expenses which would have to be incurred to avoid the harm are disproportionate to the likelihood and the magnitude of the foreseeable risk. Finally, we should note that the provisions which we have just discussed are applicable to public bodies (personnes morales publiques) just as much as to private bodies, such as companies. Indeed, article 5.144 of the avant-projet proclaims a principle of equal treatment of legal and physical people (personnes morales et physiques) ‘unless the contrary results from legislation or unless such an application is incompatible with the nature of the legal person’. A legal person’s liability can therefore arise directly on this basis without there necessarily being the need for any specific imputation such as is required by the ‘doctrine of corporate directors’ (la théorie de l’organe).35 Article 1242-1 of the French Projet de réforme also follows this approach.

35 This doctrine is based on the view that ‘legal persons’ (les personnes morales) can act only through their ‘organs’ (les organes), who are physical persons. These ‘organs’ are those persons who have the power to place the legal person under obligations to third parties. As a result, a fault in an ‘organ’ also becomes a fault in the legal person: art 61(1) Code des sociétés (de Belgique), which also provides that ‘the members of these organes do not contract any personal liability relating to the company’s undertakings’ (‘Les membres de ces organes ne contractent aucune responsabilité personnelle relative aux engagements de la société’).

416  Bernard Dubuisson

V. Harm A.  Infringement of a Legally Protected Interest Harm (dommage) is treated in articles 5.171 to 5.178 which form Section 4 of the avant-projet. At to the issues relating to reparation, these are addressed in the following section under the heading of the consequences of liability. As amended, article 5.171(1) provides that harm consists of the economic or non-economic repercussions of an infringement of a legally protected interest (un intérêt juridiquement protégé). Article 5.171(2) adds that its nature and extent is determined by comparing the factual situation of the victim before and after the infringement. The reference to a legally protected interest could clearly be thought surprising in a legal provision stemming from the French civil law tradition. It is certainly new when compared to the present state of Belgian law, as the Cour de cassation is usually content to recall that harm must result from the infringement of a stable and legitimate interest.36 However, this provision should not be seen as a complete reversal of the present law but rather as reflecting the concern to put in order and clarify the law. Indeed, an infringement should not be confused with harm as it can exist without necessarily resulting in any harm. For example, I may well cross land belonging to another person without authorisation (which is an infringement of that person’s right to property) without this necessarily resulting in any harm to the owner. Similarly, an infringement of my reputation (l’honneur) does not necessarily involve any ‘moral loss’ if this infringement received no publicity. The distinction proposed between infringement and harm appears to be the same as the distinction adopted by some French legal scholars between harm and loss, but this would be a mistake. The contrast betwen the infringement of an interest and harm does not lead to the consequences attached to the distinction between harm and loss in France.37 The Explanatory Memorandum to the Belgian avant-projet is completely clear on the issue: the notions of harm (dommage) and loss (préjudice) are used synonymously. Moreover, in my view, the French terminology leads to a regrettable confusion between the infringement of an interest and harm. As put forward in the Belgian avant-projet, a legally protected interest is the gateway to harm but it is not the harm itself. A legally protected interest is an

36 On all aspects of this issue see R Jafferali, ‘L’intérêt légitime à agir en réparation – Une exigence … illégitime?’ JT 2012, 253–64. 37 See the references mentioned in the following treatises: G Viney, P Jourdain and S Carval, Traité de droit civil, Les conditions de la responsabilité, 4th edn (Paris, LGDJ, 2013) 3 [246-1]; P Brun, Responsabilité civile extracontactuelle, 4th edn (Paris, LexisNexis, 2016) 119 [176]; M Bacache-Gibelli, Traité de droit civil, t. 5, Les obligations, La responsabilité civile extracontractuelle, 3rd edn (Paris, Economica, 2016) 421 [372].

The Projet de Réforme du Code Civil Belge  417 interest recognised and protected by the law (le droit).38 This is the filter which allows a claimant’s harm to enter the legal sphere. It should be noted that the rest of article 5.171(1) does not continue by setting out an exhaustive list of protected interests in the manner of paragraph 823(1) BGB. Without yielding to the temptation of a restrictive approach which would not be open to developments in legal rights, the fact remains that harm is not reparable under article 5.171(1) unless it results from the infringement of an interest judged worthy of protection (an infringement of personal integrity, health, life, ownership, one’s own image etc). This approach is justified as all harms which occur in fact (such as economic losses or physical damage to property) are not necessarily reparable in law. A thief cannot claim reparation of the harm caused by a third party to the stolen property (that is, not its owner) as the thief ’s interest is not protected in law. It is true that the issue whether an interest is legally protected may not always be easy to settle. In this respect, it is enough to recall the example of a child born disabled following a doctor’s failure to inform his or her mother of the relevant risk with the result that the mother was not able to exercise her right to an abortion in an informed way (the problem referred to as ‘wrongful life’).39 In this respect, it is important that the infringement of a legally protected interest is not confused with an infringement of an ‘individual right’ (un droit subjectif). There should be no question of returning to the ancient conception of von Ihering according to which a harm can result only from the breach of an individual right endorsed by a right of action. The infringement of a fundamental freedom clearly constitutes an infringement of a legally protected interest even though it is not a matter of an infringement of an ‘individual right’. In fact, an individual right is merely one of a number of protected interests.40 As article 5.171 does not go on to set out an exhaustive list of protected interets, it is clearly for the court to determine if, given the present state of the law, the interest in question is protected by the legal system. 38 It is worth noting that the French here refers to an interest which is juridiquement protégé and the explanation in the text refers to its recognition and protection by le droit, that is to say, the law and not more specifically by legislation (la loi). 39 Cass 14 November 2014, JT 2015, 221. In this case, the Belgian Cour de cassation refused to follow the caselaw of the French Cour de cassation in the affaire Perruche (Cass Ass plén 17 November 2000, D 2001, 316 concl J Sainte-Rose, note D Mazeaud and P Jourdain). To justify its decision, the Belgian Supreme Court relied on the idea that the reparation of harm consists of returning the victim to the situation in which he or she would have remained without the defendant’s fault. In the circumstances in question, if the doctor had committed no fault, the child would never have been born. It concluded that there was no reparable harm where reparation requires a comparison of the actual existence of harm (on the real facts) and the non-existence harm (on the hypothetical facts). The Court has confirmed this caselaw several times in judgments given on 21 April 2016 (RG no C.15.0286.N.), on 17 October 2016 (RG no C.11.0062.F.) and on 13 April 2018 (RG no 13.0302.F). 40 R von Ihering, L’esprit du droit romain dans les diverses phases de son développement, 3rd edn, t. IV (French translation by O de Meulenaere, Bologna, Forni Editore Bologna, 1969) 321; on the distinction between droit subjectif, intérêt and liberté civile, see T Léonard, Conflits entre droits subjectifs, libertés civiles et intérêts légitimes, Un modèle de résolution basé sur l’opposabilité et la responsabilité civile (Bruxelles, Larcier, 2005) 893; F Ost, Droit et intérêt, vol 2, Entre droit et non-droit: l’intérêt (Bruxelles, Publications des Facultés universitaires Saint-Louis, 1990) 199.

418  Bernard Dubuisson

B.  The Infringement and its Repercussions The distinction between the infringement and its repercussions is not insignificant from the point of view of causation, as it implies that causation must be established not merely between the action giving rise to liability and the infringement, but also between the infringement and the claimant’s harm. In this respect, we should note that a single action on which liability may be based can result in several infringements of different natures and that each one of these infringements can result in its turn in multiple repercussions, whether economic or non-economic. For example, a public employer who brings a recourse action against a third party held liable for injuring one of his employees under articles 1382 and 1383 of the Belgian Code civil and claims reimbursement of the salaries which it had to pay by law during the employee’s unavailability to work may rely on an infringement of its contractual right to the employee’s work since it could not enjoy this counterpart to the remuneration which it has paid. The Belgian case law allows such a personal action and therefore authorises a public employer in this situation to reclaim from the person liable the gross salaries which it has paid without return.41 Following this suggested analysis, we must consider that, unlike the position in German law, an employer’s contractual right (le droit de créance) is an interest which deserves legal protection in Belgian law and that the harm which employers suffer consists of the loss resulting from their employee’s lack of work (their prestations), a harm which is assessed at the level of their gross salary.

C.  The Nature and Extent of Harm Article 5.171(1)2° then makes clear that the nature and the extent of harm is to be deduced by comparing two factual situations: the one which existed before the infringement took place and the one which existed after it. On reflection, it would perhaps have been preferable to state more simply that harm results from such a comparison without mentioning at this stage the extent of the harm so as to avoid any confusion with the rule stated in article 5.180(1): the substance of the harm should not be confused with the method which is used for its reparation. Harm is not necessarily to be reduced to a difference in value in this way, for otherwise in cases where one cannot refer to a difference in market value for interest infringed, there would be no harm at all: harm can exist and be the object of reparation even where the ‘thing’ which is harmed is not given any market value. For example, the extinction of a rare bird does not necessarily have a market value but the harm still exists as the bird which was present before the infringement has disappeared.

41 Cass (2) 20 February 2001 (one decision), Pas 2001, 335; Cass (3), 19 February 2001 (four decisions), Pas 2001, 322, 327, 329 and 333, RGDC 2003, 182 note S Heremans, RW 2001–2002, 238, Dr circ 2001, 221, Bull ass 2001, 769, 771, 780 note P Graulus.

The Projet de Réforme du Code Civil Belge  419 This loss can attract reparation by means of a reimbursement of the costs needed to breed a bird of the same species artificially and to reintroduce it into the wild.

D.  Legitimate Interest and Losses Resulting from an Unlawful Activity Beyond article 5.171(1), article 5.171(2) opens the way to a second control on the admissibiliy of a claimant’s harm. Under existing Belgian law, all questions relating to such admissibility are grouped together under the banner of legitimate interest. This has given rise to an often chaotic case law. According to the Belgian Cour de cassation, an interest is considered illegitimate where an action for reparation tends to recognise, maintain or consolidate an unlawful situation. Thus, an action for reparation cannot enable the reimbursement of income obtained by the victim where it results from a business activity en noir, that is, in breach of employment, social security or fiscal legislation.42 Similarly, the loss of the benefit of abstracting water taken unlawfully after pollution caused by a third party cannot be compensated.43 Conversely, the Cour de cassation has quashed a decision of a lower court which refused any reparation to the owner of a caravan destroyed by a tree on the ground that the caravan had been installed without the requisite permission to build for a fixed installation. The Court criticised the lower court in that it had failed to hold that the compensation would have had as its object the maintenance of a situation contrary to public policy (ordre public).44 It is true, though, that in this case the compensation which would have been received by the owner would not necessarily authorise him to reconstruct the caravan in the same place and he could have used it for other purposes (such as rebuilding it elsewhere). On the other hand, the Court has also stated that the mere fact that a claimant is in an unlawful situation at the time of the occurrence of the harm does not necessarily mean that the victim cannot rely on an infringement of a legitimate interest.45 As a result, a claim for reparation of the consequences of personal injury suffered by a passenger following a railway accident is not necessarily refused on the ground that he or she did not possess a valid ticket. The result must also be the same for a claim by a foreigner injured in a road accident whose presence in the country is unlawful at the time. Article 5.171(2)2° (formerly article 5.171(2)) intends to provide a more solid basis for these different results. The requirement that the harm must result from the infringement of a legally protected interest is probably not enough to resolve all the problems tied to the lawful character of harm. As the avant-projet distinguishes



42 Cass

14 May 2003, RGAR 2003, no 13767, Bull ass 2003, 818. 3 October 1997, Pas 1997, no 387. 44 Cass 2 April 1998, Pas 1998, no 188; see also Cass 4 November 2011 RG no C.08.0407.F. 45 Cass 10 October 2017 RG n° P.17.0096.N. 43 Cass

420  Bernard Dubuisson between the infringement of an interest and harm, it is necessary to provide for a second control as to the actual existence of the harm suffered by the victim, that is to say, on the repercussions of this infringement.46 Article 5.171(2)2° therefore rules out any reparation of losses resulting from an unlawful activity imputable to the injured party, ie for which the injured party was responsible. For example, where a worker suffers a loss of income gained en noir, the infringement of physical integrity of which he or she complains clearly constitutes an infringement of a legally protected interest, but he or she would not be able to recover reparation of this harm, as the income in question resulted from an unlawful activity for which he was responsible. The provision expressly indicates that the harm must find its source in the unlawful activity itself, which assumes the existence of a strict relationship between the harm in question and the unlawful situation or activity. The injured passenger who did not have a ticket at the time of the accident would therefore be able to obtain reparation in respect of the consequences of his or her personal injury.

E.  Infringement of a Personal or a Collective Interest Article 5.172 of the avant-projet picks up another traditional condition relative to the admissibility of harm. The harm must result from the infringement of a personal or a collective interest. However, an infringement of a collective interest is reparable only in the situations and subject to the conditions determined by legislation. This form of words is inspired by article 8(2) of the Avant-projet Terré.47 At this stage, the Belgian reform avant-projet, unlike the French Projet de réforme (articles 1279-1 to 1279-6), does not deal with the reparation of purely environmental harm. Given the complexity of questions and issues to which this gives rise, the avant-projet leaves this question to consideration by the legislator without more.

F.  Certainty of Harm, Patrimonial and Extra-patrimonial Harm, and Indirect Harm The contrast between the infringement of a legally protected interest and its repercussions is particularly illuminating where we are concerned with the presentation of the characteristics of particular types of harm. Article 5.173 makes

46 This provision can also be compared with art 2:103 PETL which states that ‘[l]osses relating to activities or sources which are regarded as illegitimate cannot be recovered’. 47 The text of the Avant-projet Terré is entitled ‘Des délits’ (‘Delicts’) and concerns only extracontractual liability. For a commentary see F Terré (ed), Pour une réforme du droit de la responsabilité civile (Paris, Dalloz, 2011).

The Projet de Réforme du Code Civil Belge  421 clear that harm must be certain and that future harm is reparable where it is the certain consequence of an existing infringement of a legally protected interest.48 Article  5.175 defines ‘indirect harm’ (dommage par ricochet) as a person’s own harm resulting from a previous infringement to the interest of another person with whom the first person has a legal relationship or a sufficiently close relationship of affection. In doing so, article 5.175(2) formally recognises that a person injured indirectly would face the defences which the person liable could have put forward against the direct victim. The justification for this result is found in the relationship of dependence between the claims of direct and indirect victims instead of the ‘principle of family solidarity’ on which it is pinned at the moment by the Cour de cassation.49

VI.  Reparation of Harm The rules on reparation and the assessment of harm are provided by Section 5 relating to the consequences of liability (articles 5.179 to 5.188). In particular article 5.180(1) concerns the aims and modes of reparation of harm. Its first paragraph proclaims the rule known as ‘the negative difference’ according to which the reparation of patrimonial harm is intended to place the injured party in the situation in which he or she would have found themselves if the action giving rise to liability had not occurred.50 Unlike patrimonial harm, extra-patrimonial harm is the object of compensation and not reparation properly so-called, as is set out by article 5.180(1)(2).

A.  Reparation in Kind and Orders of Cessation of Unlawfulness Article 5.188 of the Belgian avant-projet provides that ‘without prejudice to the right to reparation of harm, the court may make an order or a prohibition against

48 According to a recent judgment given by the Cour de cassation in 2018 (Cass 3 January 2018 RG no P.17.0976.F.), the lower courts may perfectly well award damages for the loss which an injured party will establish in the future, on condition that the cause of that loss exists at the date of judgment in a way which allows the court to assess the harm which will necessarily result from it. 49 Cass 9 December 1962, RGAR 1963, no 7105; Cass 17 June 1963, RCJB 1964, 449 note Kirkpatrick; Cass 19 December 1967, Pas 1968, I, 537; Cass 6 January 1981, Pas 1981, I, 476; Cass 14 April 1981, Pas 1981, I, 915; Cass 1 February 1994, RGAR 1995, no 12.444; Cass 5 October 1995, Pas 1995, 873; Cass 5 September 2003, Pas 2003, 1360; Cass 28 June 2006, Pas 2006, 1534; Cass 16 February 2011, Pas 2011, 529. In any event, the Constitutional Court has not seen in this difference in treatment between categories of victim a discrimination prohibited by arts 10 and 11 of the Belgian Constitution (judgment of 17 July 2014 no 11/2014). 50 cf art 1258 of the Projet de réforme.

422  Bernard Dubuisson any person who, by their fault, risks causing harm to another person so as to prevent their wrongful action [acte fautif] or deter its repetition or continuation’. Under existing Belgian law, the juges du fond already have the power to impose a prohibition measure or order an injunction so as to stop harm being caused on the basis of the law of liability. Such an order or injunction can relate either to a harm which has already occured or to a possible increase of the harm. A measure so taken will prevent the pursuit or the repetition of the unlawful behaviour. On the other hand, there is no legal basis for such an order where the harm has not yet occurred and exists only in a fear that the harm will occur, without it being established with certainty that the harm will result in the future. Article 5.188 (formerly article 5.182) would provide a legal basis for such an injunction and this would confer a preventive effect on the law of liability. However, it should be noted that this provision does not apply except as regards liability for fault and cannot be relied on in the case of strict liability. Article 5.176 of the Belgian avant-projet further reinforces this preventive effect by providing that the costs of any measures taken to prevent an imminent harm constitute a reparable loss. Contrary to what is often written, prohibitions and injunctions are not conceived as a form of reparation in kind but as a preventive measure intended to ensure respect for the duty stated by article 5.146 not to cause harm by one’s fault. However, in practice, these measures can appear to be a form of reparation in kind where they involve, for example, putting an end to a long-lasting illegal situtation. Reparation in kind is itself the subject of specific provision, article 5.182(1) explaining that the aim of this mode of reparation is ‘to suppress the actual harmful consequences of the action giving rise to liability’.51 Article 5.182(1)2 confirms first that the court has very broad powers to give effect to a reparation in kind. In this context, ‘[t]he court may modify the legal situation of the parties or order that measures are taken by the person liable either personally or at his or her expense’. The court therefore has at its disposal a vast range of means of ensuring an effective reparation; for example, it can declare the annulment of a unlawful legal transaction, require the publication of its judgment in the press, or order the removal of an illegal plantation. Article 5.182(2) confirms the primacy of reparation in kind over reparation in the form of damages.52 An injured person therefore has the right, in principle, to reparation in kind but its availability does have certain limits.53 It cannot be

51 ‘La réparation en nature tend à supprimer concrètement les conséquences dommageables du fait générateur de responsabilité.’ 52 For this purpose, the expression ‘reparation by equivalent’ (réparation par équivalent) was deliberately avoided as it was considered ambiguous and ‘reparation in the form of damages’ (réparation sous forme de dommages et intérêts) that is, in money, was preferred. 53 Cass 26 June 1980, Arr cass 1979–1980, 1365; Bull ass 1980, 1341 concl J Velu, JT 1980, 707, Pas 1980, I, 1341 concl J Velu, RW 1980–1981, 1661, RCJB 1983, 173 note F Delpérée; Cass 20 January 1993, Arr cass 1993, 74, Bull ass 1993, 67, Pas 1993, I, 67; Cass 21 April 1994, Pas 1994, I, 388 (somm), Arr cass 1994, 392 (somm); Cass 5 May 2011, Pas 2011, 1272 concl Avocat général A Henkes, RCJB 2012, 363 note L Van Bunnen, RGAR 2012, no 14.846 note N Estienne, RGDC 2012, 247 note P Wéry.

The Projet de Réforme du Code Civil Belge  423 awarded if it is physically impossible, if it would constitute an abuse of rights, if it would involve recourse to the physical constraint of the person of the debtor, or if it would be contrary to human dignity. Apart from these particular cases, an injured person is not obliged to be content with damages. This primacy of reparation in kind is confirmed by Article 5.182(2)2 which in addition allows a person liable to offer to repair the harm in kind while leaving it to the injured person the possibility of refusing this offer for good reason (such as an irremediable loss of confidence or the serious breakdown of relations between them). This possibility stems from recent case law of the Cour de cassation.54

B.  Punitive Damages and the Assignment of Unlawful Gains The Belgian avant-projet de réforme says not a word about punitive damages. This reflects a choice which is more political than legal. When asked about this topic by the working group, the Minister of Justice clearly indicated that he did not wish this possibility to be laid down in the Code civil. Articles 5.179 et seq therefore formally recognise the mainly compensatory purpose of civil liability to the exclusion of any punitive purpose. The principle of full and actual reparation is repeated no less than three times in the two avants-projets (article 5.89 for contractual liability, article 5.179 for extra-contractual liability and article 5.311 for all obligations). On the other hand, article 5.183(2) (as amended) empowers the court to allocate to the injured person all or part of any unlawful gain made by the person who caused the harm, without derogating from the compensatory purpose of liability. Its wording is as follows: Where the person responsible, with the view to profit, has intentionally breached a right of personality of the person injured or has intentionally infringed his honour or his reputation, the court may assess the harm on the basis of the amount of profit realised or part of it, as long as this amount manifestly exceeds the amount which would be awarded for reparation of patrimonial or extra-patrimonial harm.

This provision, which did not exist in the first draft of the avant-projet, was added after the public consultation in order to discourage so-called ‘lucrative fault’, that is to say, fault committed by a person with the view to speculating on its advantages, where the amount of profits which he or she expects to gain exceeds the amount of any damages which could be awarded. In order to discourage this, article 5.184 gives the court the possibility of depriving the person who commits such a fault wholly or in part of the profits which he or she has gained unlawfully by transferring them to the injured party. As the provision is drafted, this does not involve a

54 Cass 3 April 2017, RW 2017–2018, 1414 note P Dion and Y Stevens, JLMB 2018, 1892 obs P Wéry: ‘La réparation en nature en matière extracontractuelle: quelques nouvelles avancées jurisprudentielles’.

424  Bernard Dubuisson civil penalty but is instead a particular way of making reparation of the harm. It remains the case, nonetheless, that this solution derogates from the rule stated by article 5.180(1) according to which ‘the aim of the reparation of patrimonial harm is to place the injured person in the position in which he would have been if the action giving rise to liability had not occurred’. The assignment of unlawful gains already exists in Belgian legislation in the situation of breach of intellectual property rights.55 Article 5.185(2) of the avantprojet (as amended) extends this possibility to breach of personality rights (for example, the right to respect of one’s private life, the right to one’s own image) and to the case of an infringement of one’s honour or reputation. This possibility is restricted as the transfer of gains is not to be considered where the amount of the profit realised manifestly exceeds the amount which would be awarded by way of reparation of patrimonial or extra-patrimonial harm. In determining the amount of the profit to be transfered, the court must take into account the net profit realised by the person liable thanks to the unlawful activity. This suggests that only the costs incurred specifically in order to obtain the profit can be deducted, to the exclusion of more general expenses.



55 Code

de droit économique, art XI, 335, §2.

20 The Reform of Delict in the Civil Code and Liability in Administrative Law JOHN BELL

French public law has tended to track French private law in many of its ideas. There is no Code, and the first creators of the subject were private lawyers. There is simply a lot more serious thinking about liability among private lawyers. That is clearly shown in the current books on French administrative law. When they come to talk about the liability of the state for fault, they immediately quote private lawyers talking about the Code civil. The influence of private law on administrative law liability was the theme of the thesis by the great administrative lawyer, René Chapus in 1954.1 But I want to argue that this is a misleading impression of harmony. I will argue that it is not merely the specific rules, but also the basic principles of liability that differ between public and private law, and the revisions proposed to the Code civil on delict bring this out clearly. In the end, there is no reason to expect one system, based on different legal texts operated by different judges and commented upon by different doctrinal writers for the last 200 years to be the same as the civil law system with its own code, judges and doctrinal writers. Private law writers are quick to point out the differences between administrative liability and civil law. As Viney neatly puts it: French law has given jurisdiction to decide litigation by individuals against the Administration to an autonomous judicial order which considers itself bound neither by the texts nor by the principles of civil law.2

Although there are occasional references to administrative law in civil law treatises and to civil law in administrative law treatises, the two systems largely carry on independently. The problem of consolidating civil law in a revised Code civil in isolation from the rules of administrative law is that this is very much a top-down reform, rather than taking advantage of a chance to review how the different rules

1 R Chapus, Responsabilité publique et responsabilité privée: Les influences réciproques des jurisprudences administrative et judiciaire (Paris, LGDJ, 1954). 2 G Viney, Traité de droit civil, Introduction à la responsabilité, 3rd edn (Paris, LGDJ, 2008) [1].

426  John Bell interact on the ground. Whereas English lawyers are forced by our system to ask the question how do the rules governing the liability of public authorities relate to the liability of private individuals, such debate is far less prominent in France.3 The difference between public law and private law on this topic matters because many public services are run as concessions by private companies. Whereas the relationship between the public authority and the concessionaire is governed by public law, the relationship between the concessionaire and the service user is governed by private law. So, if we have a service run directly by one commune any liability will be governed by administrative law. But in a neighbouring one of the 35,000 communes, the service might be put out to concession and the private law service provider would only be liable in private law. The prospect of a patchworkquilt pattern of liability law is not something which could be considered acceptable from the point of view of the constitutional principle of equality before the law found in article 6 of the Declaration of the Rights of Man and of the Citizen of 1789 (DDHC). It is not possible to cover all the areas of this large topic, so I will select four: (1) foundations of liability, (2) fault liability, (3) liability for others, and (4) no-fault liability. This chapter deliberately uses the first edition of Laferrière’s great treatise on administrative law because, in that way, it becomes clear that the differences between public law and private law in this area are of very long standing.

I. Foundations If we look hard enough, the roots of public law’s difference are found in the years just before the French Code civil of 1804. The Declaration of the Rights of Man has two relevant principles: article 4, that you can do anything you like unless you harm someone else, and article 13, that public burdens are to be shared by all. As we will see, these are constitutional principles in modern French law. The contemporary Code civil operates with two principles: fault (article 1240 Cc) and risk (articles 1242–44 Cc), and these are replicated in the Projet de réforme in articles 1241, 1242 and 1242-1, on the one hand, and articles 1243 and 1244, on the other. A better articulation of that second set of original ideas is found in the ­contemporaneous Introduction to the Prussian Allgemeines Landrecht of 1794: § 74 The furthering of the common good takes precedence over individual rights and privileges of members of the state, if a genuine conflict exists between these two positions. § 75 The State is, however, bound to compensate anyone who is forced to sacrifice his individual rights and privileges for the common good.

3 See N Albert-Moretti, F Leduc and O Sabard (eds), Droits privé et public de la responsabilité extra­ contractuelle: étude comparée (Paris, LexisNexis, 2017) as a rare example.

The Reform of Delict in the Civil Code and Liability in Administrative Law  427 These ideas were found in pre-Revolutionary French law but, at that stage, they typically led to ex gratia payments, rather than strict entitlement. The Revolution consolidated the principles.4 But the process of turning article 13 DDHC into a firm principle of liability was much slower. Deguergue5 argues that it was seen first as a principle governing taxation. The first signs of it turning into a more general principle governing the allocation of burdens within the state is traced to 1868. But she argues that it is really only with Hauriou in 1896 and then Jèze in 1910 that we see the articulation of equality before public burdens as a principle of state liability. She notes how the case law of the Conseil d’Etat moved gradually to recognising the principle between 1920 and 1923, leading to the decision in Couitéas,6 which accepted state liability for the consequences to a landlord of the refusal by public authorities to remove illegal occupants from its land who were disobeying a court order. The importance of Couitéas is that it moves beyond the idea of liability for risk which had recently been established in 1919 in Regnault-Desroziers.7 Here the state was held liable to neighbours for the explosion of an armaments store on the ground that the risks created exceeded normal risks between neighbours. Again, there was no need to prove fault. This followed on naturally from Cames, which was a factory accident caused to an employee.8 That decision did not articulate a principle, but the idea of liability for an abnormal risk took hold (and was then applied in private law in the following year9). This parallels the idea of liability based on risk under article 1384(1) Cc (now article 1242(1) Cc) developed by Salleilles and Josserand.10 This principle was then accepted by the Cour de cassation as giving rise to a presumption of fault under article 1384(1), also in 1919.11 Couitéas is much more clearly based on a principle of social solidarity, even if liability for risk might be seen as a sub-division. Rather than seeing harm as an accidental by-product of a created risk, the harm in equality before public burdens is a consequence of the normal operation of the administration which has to take decisions about where losses lie.12 The decision is part of a general movement of 4 A Jacquemet-Gauché, La responsabilité de la puissance publique en France et en Allemagne (Paris, LGDJ, 2013) 194–96. 5 M Deguergue, Jurisprudence et doctrine dans l’élaboration du droit de la responsabilité administrative (Paris, LGDJ, 1994) 138–39. 6 CE 30 November 1923, D 1923, 3, 59 concl Rivet, S 1923, 3, 57 note M Hauriou, RDP 1924, 75 and 208 note G Jèze; Deguergue (n 5) 140–42. 7 CE 28 March 1919, Regnault-Desroziers, RDP 1919, 239 concl M Corneille, note G Jèze; S 1918–19, 3, 25 note M Hauriou. 8 CE 21 June 1895, Cames, Leb 509 concl J Romieu, S 1897, 3, 33 note M Hauriou. 9 Cass civ 16 June 1896, Teffaine, D 1897, 1, 433 note R Saleilles. As a presumption of fault, it could be rebutted by showing no fault: Req 30 March 1897, Grange, D 1897, 1, 433, 440 note R Saleilles. 10 See J-L Halpérin, ‘French doctrinal writing’ in N Jansen (ed), The Development and Making of Legal Doctrine (Cambridge, Cambridge University Press, 2010) 73, 88–89, citing R Saleilles, Les accidents de travail et la responsabilité civile (Paris, Arthur Rousseau, 1897) and L Josserand, De la responsabilité des choses inanimées (Paris, Arthur Rousseau, 1897). 11 Cass civ 21 January 1919, Marcault, D 1922, 1, 25 note G Ripert. 12 Jacquemet-Gauché (n 4) 203–04.

428  John Bell the period to create solidarity for certain types of social harms. That idea was then constitutionalised in the Preamble to the 1946 Constitution: Para 11 [The nation] guarantees to all, especially to the child, the mother, and aged workers, the protection of health, material security, rest, and leisure. Any human being who by reason of his age, physical or mental health, or economic situation, is unable to work, has the right to obtain appropriate means of subsistence from the community. Para 12 The nation proclaims the solidarity and equality of all French men and women in the face of the burdens that result from national calamities.

The state is liable to pay not because it has done something wrong or created a danger for others, but simply because it is the conduit for redistributing losses that have fallen unfairly on specific individuals. So, in public law, there is the harm principle and the solidarity principle. The harm principle leads us to fault, but not necessarily to moral blame. As will be seen in the next section, fault is really a failure of the public service to meet its stated objectives. The solidarity principle of equal sharing of burdens is a distinct idea which goes beyond the socialisation of risk. Many authors have suggested over the years that fault liability was simply a subset of compensation for shared burdens.13 Unlawful acts by the administration are a burden that is more or less inevitable and so the victim should be compensated by society. This fits with the German development of the law on expropriation which now accepts the idea of ‘quasi-expropriation’ to cover compensation for unlawful acts of the administration.14 If this analysis is right, then public law and private law are not really based on the same principles. Private law is based on responsibility for wrong and for risk. Public law is based on either social solidarity on its own or social solidarity combined with liability for maladministration.

II.  Fault Liability Now my basic argument is not new. Laferrière wrote in 1887 that article 1240 (former article 1382) Cc did not apply in administrative law. On the one hand, it was too broad since the state is not liable to compensate for every act of fault, and on the other hand the state can be liable to compensate without fault.15

13 ibid 197–98, citing leading members of the Conseil d’Etat such as Latournerie (1945) and Rougevin-Baville (1992) and academics such as Amselek (1975). See also R Errera, ‘The Scope and Meaning of No-Fault Liability in French Administrative Law’ (1986) CLP 157, esp 171–72. 14 On the parallels between German and French laws in this area, see Jacquemet-Gauché (n 4) 196–200. 15 E Laferrière, Traité de la juridiction administrative et des recours contentieux (Paris, Berger-Levrault, 1887) vol 2, 176.

The Reform of Delict in the Civil Code and Liability in Administrative Law  429 Hauriou in 189616 drew the distinction between the foundation of an obligation to compensate for harm suffered and the conditions that had to be met before the obligation arose. For him, the foundation (or cause) of state liability was the mutual insurance run by the state against administrative accidents. Fault was not the cause of the obligation to compensation, but merely one of the conditions that had to be satisfied before an obligation arose, in other words it was a fait générateur. If we look at the conditions to be satisfied, then one might argue that there has been a convergence of civil and administrative liability for fault in the last half century. For Laferrière, the distinctiveness of public law was that there were important situations where simple fault was not enough to ground liability and there were important areas of immunity (which would be, in part, compensated by no-fault liability). In many ways that holds true. But now with a number of caveats. First, the principle in article 1240 Cc and the proposed article 1241 of the Projet that ‘A person is liable for the harm caused by his fault’ has been recognised since 1982 as a constitutional principle. Grounded in article 4 DDHC, it sets a baseline of liability from which it is only permissible to depart for good reason.17 In subsequent years, administrative law has been dismantling the exceptions to liability for simple fault of which Laferrière wrote. In particular, the areas in which the state is only liable for faute lourde (gross negligence) have diminished: hospitals, prisons and the public order activities have all moved to faute simple.18 As late as 1954, Chapus noted that there were two areas in which ordinary fault had been heightened to faute lourde: where a service was gratuitous (gratuitous passenger or a beneficiary of a public service) and difficult activities, especially medical cases, recovery of taxes and supervision of others (especially the mentally ill).19 Chapus already identified that the faute lourde requirement was out of line with the current climate of wishing to compensate victims as much as possible.20 Consequently in 1992, the Conseil d’Etat in M et Mme V21 abandoned the restriction of hospital liability to faute lourde and this was enacted in article 1142-1 of the Code de la santé publique from 2002.22 In prisons, immunity was replaced in 1958 by faute lourde and then by faute simple in 2007.23 Public regulation and public safety (police in the French meaning) was immune until 1905, then replaced by faute lourde. Now emergency services are liable merely for faute simple, despite the difficulty of the

16 M Hauriou, ‘Les actions en indemnité contre l’Etat pour les préjudices causés dans l’administration publique’ RDP 1896, 51, 53. 17 CC decision 82-144 DC of 22 October 1982, considérant no 3: ‘Considérant que, nul n’ayant le droit de nuire à autrui, en principe tout fait quelconque de l’homme, qui cause à autrui un dommage, oblige celui par la faute duquel il est arrivé, à le réparer’ with acceptance of exceptions in considérant 4 and the rejection of total immunity in considérant 5. 18 That now leaves as a curious exception liability for breach of EU law under art 340 TFEU. 19 Chapus (n 1) 360–66. 20 ibid 371. 21 CE Ass 10 April 1992, Leb 171 concl Legal. 22 Loi 2002-303 of 4 March 2002. 23 CE 9 July 2007, MD, AJDA 2007, 2094 note Arbousset.

430  John Bell task.24 Similarly the regulation of dangerous premises, phytosanitary measures and urgent removal of driving licences all now give rise to liability where there has been simple fault.25 Equally, tax authorities were protected by immunity until 1962, then subjected to liability only for faute lourde until this was all replaced by faute simple in 2011 in the Krupa case.26 The main area for faute lourde is one shared with private law – the operation of the justice system. The loi of 5 July 1972 (article L-141-1 Code de l’organisation judiciaire) imposes liability of the state for faute lourde or déni de justice. Although not directly applicable to administrative law, the Conseil d’Etat followed suit in Darmont in 1978.27 So it would be basically true that the public and private systems are aligned on this aspect of fault liability these days. As Frier and Petit put it, ‘without being totally abandoned [faute lourde] tends to retreat before the increasingly strong requirements of social solidarity’.28 Secondly, the areas in which there used to be no liability at all (acts jure imperii as opposed to jure gestionis)29 remain only to some limited extent. Fairgrieve points out that there is still no-fault liability for foreign affairs, legislation, and military or diplomatic activity.30 But this is misleading. Although acts of the legislature or treaties do not give rise to liability for fault, they may well give rise to no-fault liability as a matter of social solidarity. If some specific person’s rights are affected by legislation or a treaty, then the constitutional validity of that legal instrument will depend on the prior provision of adequate compensation.31 In addition, the La Fleurette case law allows for compensation for the unfair burdens created through a treaty, eg a landlord being unable to sue a tenant who is entitled to diplomatic immunity.32 It is true that this is not based on fault, but it does achieve the right of compensation. So, the general principle of liability for simple fault matters more than it used to and the enclaves of special rules are diminishing. Thirdly, we have to recognise that private law followed the lead of public law in creating no-fault liability for industrial accidents and then for cars from the 1890s. We will return to that later.

24 CE 29 April 1998, Commune de Hannapes, Leb 185 (fire services); CE Sect 13 March 1998, Améon, Leb 82 (sea rescue). 25 See respectively, CE 27 December 2006, Commune de Baalon, AJDA 2007, 385 note Lemaire; CE 7 August 2008, Ministre de l’agriculture et de la pêche, AJDA 2008, 1572; and CE 2 February 2011, M Gérard A, RFDA 2011, 451. 26 CE Sect 21 March 2011, M Christian Krupa, AJDA 2011, 1278 note Barque. 27 CE 29 December 1978, Darmont, Leb 542. 28 L Frier and J Petit, Droit administratif, 10th edn (Paris, LGDJ, 2018) [958]. 29 Laferrière (n 15) 174: ‘If one takes into account the differences in the liability of the state in relation to the different functions which it is called upon to fulfil, one sees that the liability is more and more limited the higher the level of function.’ 30 D Faigrieve, State Liability in Tort: A Comparative Law Study (Oxford, Oxford University Press, 2003) 14; Frier and Petit (n 28) [992], citing CE 25 March 1988, Société Sapvin, Leb 133 (refusal of France to negotiate with Spain or to refer a matter to the ICJ). 31 J Bell, French Constitutional Law (Oxford, Clarendon Press, 1992) 184–87. 32 CE Sect 29 October 1976, Dame Burgat, RDP 1977, 213 concl Massot.

The Reform of Delict in the Civil Code and Liability in Administrative Law  431 But there is still a more fundamental question which is raised by article 1242 of the Projet de réforme: ‘A violation of a legislative requirement or a failure in the general duty of care or diligence constitutes fault.’ The first of those two situations, illegality, poses no problem. Laferrière noted that damages might be awarded for illegal acts of the administration (eg closure of a match factory for fiscal reasons).33 The second idea of ‘the general duty of care or diligence’ is more problematic. What is the ‘general duty of care’ and is that what administrative law seeks? If we ask what constitutes ‘fault’ in administrative law, then it is really the defective functioning of the institution, not moral blameworthiness.34 As Fairgrieve rightly points out, the focus of faute de service is ‘a malfunctioning of the administrative machinery of the state’.35 It is more equivalent to the Ombudsman’s idea of ‘maladministration’, rather than fault in the civil law sense.36 As Delaunay neatly puts it, every breach of an obligation is a fault, but not every fault is a breach of an obligation.37 Specifically, he cites a decision of 1941 in which it is accepted that there was no duty on the mayor to order the repair or demolition of a dilapidated building, but there was fault. He comments ‘fault is not principally the breach of a pre-existing obligation, but the much more general failure of the administration in its duties, in a line of conduct in administrative action’.38 As a result, fault is a form of strict liability in the sense that you look at a result and if it should not have resulted from the normal functioning of the public service, then there is liability. There is never a need to identify a specific person who is responsible. Deguergue points out that the conception of fault in administrative law has its origins in the liability of the state for failure to maintain public works under the loi of 28 pluviôse an VIII (17 February 1800). There was either a failure to construct the public work correctly or a failure to maintain it.39 She notes that, after a period of hesitations during which private law ideas of attributing wrongdoing to individuals was adopted, by the 1870s the idea of faute du service was firmly accepted as the anonymous act of a public authority.40 Planiol’s idea that fault was a failure to conform to a pre-existing duty established by law, honesty or professional skill41 was a major reference point in public law writing, as well as in private law. Deguergue notes that, as private law ­objectified its notion of fault, it was easier to see similarities between public and private law

33 See CE 5 December 1879; Laferrière (n 15) 176, [1]. 34 Laferrière (n 15) 256–58. 35 Fairgrieve (n 30) 103. 36 ibid 104. This is developed by B Delaunay, La faute de l’administration (Paris, LGDJ, 2007) [169]–[177] and [593]. 37 Delaunay (n 36) 178. 38 ibid 180. 39 See Deguergue (n 5) 148–59. 40 ibid 162–63. 41 M Planiol, ‘Etudes sur la responsabilité civile’ RCLJ 1905, 277.

432  John Bell and this definition of Planiol is taken up by Chapus in his thesis.42 For her, as for a number of public law writers of the 1950s and 1960s, fault should be seen as the principal ground of administrative law liability.43 She cites Benoît in 1956 as defining administrative fault as ‘a dissonance between the way in which the service has functioned and the manner in which it ought to have functioned in accordance with laws and regulations’.44 But if the idea of non-conformity of actual conduct with some normative ideal seems similar between private law and public law, there are also important differences. In his commentary on the Avant-projet Catala, Borghetti suggests that the majority of private law scholars accept the idea that fault is the objective failure to conform to a pre-existing standard.45 But he also notes that the required standards are not set out in any detail and so much of the precision comes from the detailed analysis of the circumstances of each case. Fairgrieve also points to the use of a unitary conception of fault in civil law, which requires a detailed focus on the circumstances of the case.46 This view was paralleled in 1954 by Chapus. He suggested that in the process of determining whether there has been fault, the approach of civil and administrative judges is exactly comparable, for they ‘both pay attention very attentively to all the circumstances of the case in order to decide whether there has been a breach of an obligation’.47 Fairgrieve contrasts, however, administrative law in which there has been an attempt by some authors48 to provide a catalogue of duty situations. But he concedes that this is not consistently shown in the case law and in the works of many authors. In a sense, the focus on detailed analysis of situations, rather than general categories of duty (as in English law) or categories of protected interests (as in German law), prevents the development of categories of duty. But this analysis by Borghetti and Fairgrieve does suggest significant similarities in approach between civil law and administrative law approaches to fault liability, which doctrinal writers in both fields have drawn on. However, I want to argue that this similarity is misleading. The concept of a duty of care or pre-existing duty in private law implies a duty owed to the other person, the victim of a wrong, a correlative duty. After all, private law places delict within the ‘Law of Obligations’, the duties owed by one person to another. Winfield defined tort in ways that would have been familiar to Planiol: Tortious liability arises from the breach of a duty primarily fixed by the law: such duty is towards persons generally and its breach is redressible by an action for unliquidated damages.49

42 Deguergue (n 5) 622–26. 43 ibid 706–15. 44 ibid 623. 45 J-S Borghetti, ‘The Definition of la faute in the Avant-projet de réforme’ in J Cartwright, S Vogenauer and S Whittaker (eds), Reforming the French Law of Obligations, ch 12, 278–79. 46 Fairgrieve (n 30) 125–26. 47 Chapus (n 1) 357. 48 Fairgrieve (n 30) 125 cites in particular M Paillet, ‘Faute de service: Notion’ JCA 1994, no 818. 49 P Winfield, The Province of the Law of Tort (Cambridge, Cambridge University Press, 1931) 32.

The Reform of Delict in the Civil Code and Liability in Administrative Law  433 The duty is owed to persons in general and an obligation to pay is owed to any victim of a breach of that duty. The amorphous concept of ‘faute de service’ suffices and does not require a correlative duty.50 Jacquemet-Gauché argues that the distinctive feature of French administrative law is that it focuses on the functioning of the administration and so it does not need to ask the question whether individual rights are affected, as German law does.51 Rather, the duty owed by the administration is a non-correlative duty. If the duty is owed, it is owed to the public and is not directly owed to the victim. Rather the victim is compensated because the service has gone wrong and the victim is the person who bears a special burden, not because the rights of the victim have been affected. Braibant and Stirn describe it as the fault committed by an anonymous bureaucracy, where you cannot identify the person who committed the wrong act.52

III.  Liability for Others Articles 1245 to 1249 of the Projet de réforme set out a liability that is in essence vicarious. Giliker points out that this is really a codification of existing case law.53 As article 1245 suggests, ‘This liability rests on proof of an action of a nature to engage liability in the direct author of the harm.’ So parents, guardians and those controlling the way of life of a minor are all strictly liable for its acts (article 1246). The same is true of a body controlling the way of life of an adult (article 1247). Presumed fault governs the liability of contractors, businesses or professions for failing to supervise those whom they control (article 1248). Whereas proposed articles 1246 and 1247 are clear instances of vicarious liability, article 1248 looks more like what the English might describe as a non-delegable duty, a direct duty to supervise and so liability arises for failing to supervise, rather than for the acts of the supervisee. A significant difference between administrative law and civil law is that administrative law really has no role for vicarious liability. Article 75 of the Constitution of an VIII (1799) provided protection to civil servants against claims brought against them for fulfilling their functions.54 Faute de service is a defect of the institution and the equivalent to article 1242-1 of the Projet de réforme (a fault in one of its organs or a failure in its organisation or its functioning). As Laferrière wrote in 1887, ‘The liability of the state is not the liability for others envisaged by article 1242 (former article 1384) Cc, but direct liability: the public service is deemed to be

50 See Delaunay (n 36) [57]–[59] and at 178. 51 Jacquemet-Gauché (n 4) 481. 52 Cited in Fairgrieve (n 30) 39–40. 53 Giliker, ‘The Role of la faute in the Avant-projet de réforme’ in Cartwright, Vogenauer and ­Whittaker (n 45) ch 13, 296. 54 Now contained in art 11 of the law of 1983.

434  John Bell author of the fault; it is it, that is to say the state, which compensates’.55 His main examples of activities giving rise to compensation (ports and military exercises) have changed, but the principle remains fundamentally the same. We are dealing with direct liability, not vicarious liability. In Giliker’s terms, we are going down the track of non-delegable duties, rather than vicarious liability.56 Faute personnelle is relevant principally to permit disciplinary sanctions, though it may go so far as to take the act out of the sphere of state liability altogether. The absence of a vicarious liability is shown by the fact that you do not have to attribute the action of the public service to a particular individual. All you have to show is that the service failed to perform as expected. So ‘once the standard is fixed relating to what one can expect, if the performance is not satisfactory in the light of the obligation, fault is certain’.57 If the employee is not identifiable, that does not pose problems for the liability of the administration. Chapus argued that the foundation of the liability for the acts of state agents, as with that in private law, is the obligation of guarantee owed to the victim which follows from the fairness principle that he who profits from an activity should shoulder the risks. He drew on Planiol58 that the concept of the enterprise justified liability for those who worked within it. People create enterprises to realise what they could otherwise not do, and this creates risks – ubi emolumentum, ibi onus: ‘the similarity is the most complete on the fundamental principles which govern it and on the essential rules which constitute the regime of the liability for the acts of others’.59 Again, Chapus was trying to see the similarities between administrative law and civil law. I am not convinced that this is right. If anything, there is more in common between liability in administrative law and liability in the private law of contract than between administrative law and the private law of delict. As Viney pointed out, even if Blieck60 created a general principle of liability for the acts of others beyond the numerus clausus of article 1242 Cc (former article 1384 Cc), it is still based on failures of supervision and control, whether presumed in employment or proved in other cases. By contrast, in contract, you are liable for the collaborator in performing a service because you have a direct duty to perform it.61 The same is true of the public service. The obvious difference is that the duty in contract is owed to the victim, whereas the duty to perform a public service is owed to the public (often instantiated in a hierarchical superior) and not necessarily to the victim (especially where the victim is not the user of the public service in question). The duty to compensate arises as a result

55 Laferrière (n 15) 178. 56 P Giliker, ‘Analysing Institutional Liability for Child Sexual Abuse’ [2018] CLJ 506. 57 Frier and Petit (n 28) [956]. 58 M Planiol, ‘Etudes sur la responsabilité civile. Troisième étude: Responsabilité du fait d’autrui’ RCLJ 1909, 297. Chapus (n 1) also cites in particular C Eisenmann, JCP 1949 I 751, [10]. 59 Chapus (n 1) 263. 60 Ass plén 29 March 1991, D 1991, 324 note C Larroumet. 61 Viney (n 2) [170]–[170-3].

The Reform of Delict in the Civil Code and Liability in Administrative Law  435 of non-performance of a service from which you suffer, whereas Blieck creates a duty to compensate where the victim was owed a duty by the person committing the wrong. Whittaker helpfully pointed out in the course of the workshop that the liability of the administration has closer affinities to the liability of legal persons than to the liability of individuals, which is the paradigm within which most of the proposed revisions to the Code civil are couched.62 The editors of Winfield and Jolowicz helpfully explain why the personal liability of the corporate employer is a useful way of addressing many questions of liability: In many cases it is obviously much more convenient to say that a given state of affairs or a given event proves a breach by the employer of his personal duty than to say that some employee must somehow have been negligent for that state of affairs to exist or for that event to come about.63

Viney, Jourdain and Carval note the way in which the civil liability of corporations has long been accepted on the basis that they are liable for the acts of their employees. But they argue that it has now been supplanted by the direct liability of the corporation without having to identify any individual who is responsible.64 Viney noted in an earlier volume that administrative law had been much in advance of private law in establishing collective liability or fault and she saw this as an idea of ‘socialisation of risks’,65 though she did not think that civil law would go so far. Underpinning these developments, she saw a change in legal policy: the development of collective structures and the arrival of social philosophies giving a greater place to the notion of solidarity have directed minds to the idea that certain risks must not be left as a burden on individuals, but it is necessary to collectivise them in order to cushion their effects … In a general way, in order to compensate for the insufficiency of individual liability in its compensatory function, it has appeared necessary to have recourse to very different forms of compensation whose only common feature is that they give rise to a direct socialisation of compensation without passing through the stage of identifying an individual as “liable”.66

It seems to me that the formulation of the proposed reforms to the delict provisions, which were originally found in the Avant-projet Catala as article 1353

62 S Whittaker’s oral observation at workshop ‘French Civil Liability in Comparative Perspective’ (Oxford, 26–28 September 2018); see also Giliker (n 53) 295. 63 WE Peel and J Goudkamp, Winfield & Jolowicz on Tort, 19th edn (London, Sweet & Maxwell, 2014) [9-017]. 64 G Viney, P Jourdain, S Carval, Traité de droit civil. Les conditions de la responsabilité, 4th edn (Paris, LGDJ, 2013) [854], citing ‘une personne morale répond des fautes dont elle s’est rendue coupable par ses organes sans qu’il soit nécessaire de mettre en cause, sur le fondement de l’article 1384, alinéa 5, lesdits organes pris comme préposés’ (Cass civ (2) 17 July 1967, Bull civ II no 261, 182, RTD civ 1968, 149 obs Durry). 65 Viney (n 2) [28]. 66 ibid [27].

436  John Bell and are now found in article 1242-1 of the Projet de réforme, contain distinct strands: Fault in a legal person results from fault in one of its organs or a failure in its organisation or its functioning.

On the one hand, there is the strand of a corporate body being liable for the acts of its guiding officers (its ‘organs’). This is a strand which goes back to § 31 BGB and essentially fits what article 121-2 of the Penal Code provides, which is the liability of legal persons for the acts of their organs or representatives. As a result, Viney, Jourdain and Carval spend much of their time discussing the situations in which there will be overlap between the liability of a corporation and the liability of directors.67 It seems to me that the departure from established case law is much more in the phrasing ‘a failure in its organisation or its functioning’. As Giliker notes, this is ‘utilising an administrative law concept’.68 That is much more like faute de service, as Viney rightly detects. Rather than aligning civil liability to criminal law, it is aligned to contract where, to use the German terminology, the employees are seen as the Erfüllungsgehilfe, the people by whose instrumentality the obligor fulfils an obligation under § 278 BGB, rather than Verrichtungsgehilfe, people you direct to act on your behalf and for whose actions you are responsible under § 831 BGB.69 The non-performance of an obligation is evidence of a failed organisation or functioning of the public or private body. As in contract, non-performance is not necessarily a matter of blame, it is a matter of failure to achieve one’s goals. In contract, the promisor of an obligation of result is liable when that result does not materialise. In administrative law, the public service provider is also offering a (typically gratuitous) service and is liable when this does not occur. This is a fundamentally different way of thinking from liability for the acts of others. Now, to what extent can article 1242-1 of the Projet be seen as essentially the same as the liability of the public administration? Certainly, if the analogy with contract law is maintained, there is similarity. But it is not clear how far this applies in relation to delict. Under article 1241 of the Projet, the company will owe a duty of care (an obligation de moyens), rather than a duty of result as in contract. Under article 1243 of the Projet, the company as keeper of a thing is strictly liable for the harm it causes. This is no different from Jand’heur, where Maison Bumsel was liable as keeper for the harm caused by its vehicle and its driver.70 The change in wording has not obviously brought about any change. Chapus was right in thinking that in many situations, whether public or private, the employer is responsible for the acts of agents or employees. That does 67 Viney, Jourdain and Carval (n 64) [855]–[858]. 68 ibid (n 64) [53] at 295. See Exposé des motifs of the Avant-projet, [27] at 155: ‘La notion de “défaut d’organisation ou de fonctionnement”, couramment admise par les juridictions administratives, paraît utilement transposable en droit privé.’ 69 J Bell and A Janssen (eds), Markesinis’s German Law of Torts, 5th edn (Oxford, Oxford University Press, 2019) ch 5. 70 Ch Réun, 13 February 1930, D 1930, I, 57 note G Ripert.

The Reform of Delict in the Civil Code and Liability in Administrative Law  437 reflect an idea of an enterprise liability to some extent. But I do not think the basis is the same. Private law, as Planiol wrote over a century ago, is about the benefit and burden principle involved in the enterprise. The enterprise creates an activity by putting people in place and so should shoulder the burden, even if these activities are run by voluntary associations (as in Blieck). It is not about the socialisation of risk, but the internalisation of risk within the costs of the risk creator. Public law, by contrast, is about social solidarity in sharing the costs of public services by making the administration liable.

IV.  No-Fault Liability Duguit provided the essential rationale for no-fault liability in administrative law: The only way to affirm the liability of the State is to rely upon the idea of social insurance provided by the central budget and compensating those who have suffered a loss arising from a public service, the purpose of which is to benefit all.71

It is linked, as Errera pointed out, to article 13 of the DDHC, which is a constitutional principle.72 Errera’s article of 1986 is focused on the classical category of no-fault liability, which is what the Germans would classify as ‘sacrificial encroachment’ (Aufopferungsanspruch):73 (1) liability for the effect of legislation or treaties, (2) liability for failure of the administration to act, and (3) liability for the effects of new (and experimental) ways of managing public services like jails. The essential element here is that the acts of the administration are lawful. Even in failure to act, the police are exercising discretionary powers which are not unlawful. For example, they may refuse to clear strikers from a port even though this blocks British holidaymakers from returning on the August Bank Holiday weekend.74 The public order decision carried with it special consequences for a specific group of ferry travellers. In such cases, there are direct consequences which cause a disproportionate harm, that should be shared by the whole community. But there is no need to show that the act of the police was in any way ultra vires. Other areas really are governed by the same principle. In 1887, Laferrière wrote that liability for public works (under the loi of 28 pluviôse an VIII) is essentially based on the idea that the benefit of the road, canal or rail line was for the public benefit and those who suffer harm should be compensated in the same way as those who are expropriated.75 To include this under the heading of ‘presumed 71 L Duguit, Traité de droit constitutionnel, 3rd edn (Paris, E de Boccard, 1927) III, 466, cited by Errera (n 13) 171. 72 Errera (n 13) 171. See also CC decision no 73-51 DC of 27 December 1973, Ex Officio Taxation, Rec 25; Bell (n 31) 211–15 and 346–47. J-L Mestre, Introduction historique au droit administratif ­français (Paris, PUF, 1985) [161] shows that the principle is actually medieval. 73 See Bell and Janssen (n 69) Ch 7. 74 See CE 22 June 1984, Sealink UK Ltd, AJDA 1984, 708. 75 Laferrière (n 15) 151.

438  John Bell fault’, as Frier and Petit do,76 really misses the point of its basis. We can talk of défaut de l’entretien normal, but as Laferrière pointed out already in 1887, the person harmed does not have to demonstrate some defect in construction, a failure to maintain, or that the public work was worn out. It is sufficient to show that the public work had become harmful and that the victim suffered loss as a result.77 This is quite far from normal fault liability. Most of the French works begin their sections on no-fault liability with ‘liability for risk’ and distinguish this from both fault and equality before public burdens, but I do not think that is a separate basis of liability. Let us look at the cited situations: (a) dangerous things, such as the storage of munitions,78 the carrying of weapons by the police,79 and contaminated blood;80 (b) dangerous methods, such as freedom on licence for juvenile delinquents81 or experimental medical treatment;82 (c) dangerous risks, such as the exposure of a teacher to German measles whilst pregnant,83 or a consul victim of pillage while having to remain in post during the Korean war;84 (d) riot damage; (e) third-party victims of public works, such as the breach of a dam causing flooding;85 and (f) harm to occasional or volunteer helpers of the public service. This was used to create liability towards contractual workers in public industries.86 The commentary of the Grands Arrêts87 makes clear that, although cases like Cames in 1895 gave rise to an idea of liability for risk, the fundamental basis is equality before public burdens. It makes this remark specifically in relation to the harm caused to third parties by public works,88 but the argument could go wider. I think Hauriou’s distinction between ‘foundations’ and ‘conditions’ helps here. The reason why we are concerned to help non-users of the public service in (a), (b), (d) and (e) is because they should not have to shoulder the burden of a public

76 Frier and Petit (n 28) [966]. 77 Laferrière (n 15) 157. 78 CE 19 March 1919, Régnault-Desroziers, Leb 329. 79 CE 24 June 1949, Consorts Lecomte, Franquette et Deramy, Leb 307. 80 CE Ass 26 May 1995, Consorts N’guyen, RFDA 1995, 748 concl Dael. 81 CE Sect 3 February 1956, Thouzelier, Leb 49 and CE 17 December 2010, Garde des Sceaux c FGTVI, AJDA 2010, 1696 note Pollet-Panoussis. 82 CE Ass 9 April 1993, Bianchi, Leb 127 concl Dael. 83 CE Ass 6 November 1968, Dame Saulze, Leb 550. 84 CE Sect 19 October 1962, Perruche, Leb 555. 85 CE Ass 28 May 1971, Départment du Var, Leb 419. 86 CE 21 June 1895, Cames, Leb 509 concl Romieu; GA 6. 87 G Braibant et al, Les grands arrêts de la jurisprudence administrative, 20th edn (Paris, Dalloz, 2015) 44. 88 CE 2 October 1987, Spire, Leb 302.

The Reform of Delict in the Civil Code and Liability in Administrative Law  439 service available to all users. In (c) and (f) the victims are those who help to provide the service for others and so should also not have to shoulder excessive burdens. It may be that the trigger for identifying an exceptional burden is the existence of a risk. But it is not the only trigger of liability. In which case, risk is a condition of liability, not a foundation. Most of the instances of risk have been overtaken by statutory schemes which are very much based on social solidarity/social insurance (eg medical liability from 2001, riots and industrial employees). As was noted earlier, risk has been considered a justification for some situations of state liability – towards employees and neighbours in particular. Some authors saw it as a major ground of liability; others, such as Amsalek, as a ‘safety valve’ for fault liability – a situation where fault does not provide a fair solution to a problem.89 But Viney presents things more clearly. For her, the situations where the state takes on board certain ‘risks’ on the basis of social solidarity is an example of socialisation of risks, similar to the creation of compensation or guarantee funds.90 As we have seen earlier, Viney’s quotation above (at n 68) draws attention to the way the law has needed to create a notion of ‘social risks’ to balance the category of ‘individual risks’, which private law regulates.91 Private law risks, such as the running of a business that creates potential harm to employees or neighbours, are those which can be managed through the rules of private law, such as articles 1243 and 1244 of the Projet de réforme. But social risks typically involve the creation of distinct (public) compensation funds to cover certain industrial harms (eg asbestos) or the inevitable and unavoidable risks of contemporary life (eg vehicle accidents and medical injuries). This distinction between social risks and individual risks makes sense of the idea of Starck, her mentor, whose thesis focused on the need to provide security as the basis of civil liability.92 As Viney identifies, some of the instances of administrative law liability fall on the side of managing social risks, which can be justified really on the basis of social solidarity, rather than liability for risk. Now, of course, risk was abandoned as the basis for the current article 1242 (former article 1384) Cc in the Jand’heur decision of 1930,93 when it was said that dangerousness was not required. Likewise, the Exposé des motifs of the Avant-projet Catala, makes clear that some members felt liability for things was too wide and it should be reduced to liability for dangerous things.94 This was then reflected in a separate text on dangerous things in article 1362, which has not survived into the current proposals. So private law is based on one or more other bases. Whether it 89 See Deguergue (n 5) 709–12. 90 See Viney (n 2) [28-2] and [28-3]–[28-3.6]. 91 See Viney, quoted above, text to n 68. 92 B Starck, Essai d’une théorie générale de la responsabilité civile considérée en sa double fonction de garantie et de peine privée (Paris, L Rodstein, 1947). Also G Viney and A Guégan-Lécuyer, ‘The Development of Traffic Liability in France’ in W Ernst (ed), The Development of Traffic Liability (Cambridge, Cambridge University Press, 2010) 58–67. 93 Ch Réun, 13 February 1930, D 1930, I, 57 note G Ripert. 94 Exposé des motifs of the Avant-projet Catala, 159.

440  John Bell is Starck’s foundation in the guarantee of the rights of the citizen or some other basis, it is definitely not simply risk-based in the sense of looking at the dangers which the object poses.95 Indeed, the idea of guaranteeing fundamental rights seems to be gaining ground.96 Article 1243 of the Projet de réforme cites contact with person or property, defect or abnormal position as indicators of causation of harm, rather than as grounds of liability. If one asks why a person is liable for the ‘action of corporeal things within his keeping’, then the proposed Code has no answer. The concept of benefit and burden seems the most logical basis. That would fit with the logic of article 1244 of the Projet which sanctions abnormal nuisance between neighbours – a duty to compensate for excessive interference caused by your use of property. But these are all individual risks (created by a person or a company) which are on top of the risks of everyday life to which we expose ourselves. By contrast, administrative law liability arises where there are exceptional consequences not where there are exceptional risks. A good example is medical liability. In Bianchi,97 a hospital was held liable for the adverse consequences of a new medical treatment. There is no sense that this was a dangerous activity – merely a new technique. But the harm suffered was exceptionally severe and there needed to be compensation as a matter of social solidarity. In private law, the question would arise in terms of informed consent. If the patient had been informed of risks, but all the same consented, then there would be no liability on the part of the hospital or doctor.98 The individual risk would be allocated to the patient (and his insurer), rather than to the medics (and their insurers). If there is, as Viney suggests, some common thread in the notion of the socialisation of risks, they are socialised on different bases. No-fault liability in administrative law is not restricted to things and is based on social solidarity. It does however share the function of redistributing burdens of losses to the activities which create them. Whereas public law is fairly systematic and is based on constitutional principle, private law is less coherent and seems almost ramblingly opportunistic.

V. Conclusion This chapter began with the recognition that as regards liability, public law and private law in France are distinct. They started on different paths with different legal texts, they have been applied by distinct courts, and they have been commented on by distinct legal scholars. Basically, they started off differently and when new problems arise, they carry on within the framework which is familiar. This phenomenon



95 See

Viney (n 2) [54]. 122–27. 97 See above n 82. 98 See eg Cass civ (1) 12 July 2012, Y c X, MACSF, no 11-17510. 96 ibid

The Reform of Delict in the Civil Code and Liability in Administrative Law  441 is called ‘path dependence’.99 As I wrote elsewhere, ‘path dependence assumes that there is no single perfect system to which all legal systems should try to converge, but that there are a number of coherent sets of legal rules and institutions that can deal with the same social problem’.100 There I suggested that state liability was a good example of the deep embeddedness of legal rules in the principles and institutions of French and English laws.101 I think the analysis can apply equally well to the differences within French law between civil law and administrative law. In an earlier book, I wrote extensively about the differences between the culture of administrative law in France and the culture of private law.102 In particular, I drew attention not only to different rules and distinct judicial professions, but also to differences in legal education and in the recruitment of professors. That analysis has been endorsed by distinguished French scholars.103 Starting on the one side with liability for public works in 1800 and on the other from the Code civil of 1804, the distinctive rules have been developed over two centuries, each on their own track. Of course, lawyers on one path have had regard to what is happening on the other path and have sometimes mirrored what the other is doing. But this is not a systematic development. Not even academics, recruited through distinct agrégation routes since 1895 and teaching distinct courses in the universities, have had the incentive to view liability law holistically. Judges have dealt with cases as they arrived in court naturally on an incremental basis. Each system has its advantages and disadvantages, but it can cope with a significant number of problems that arise by and large in a fair and satisfactory way. So it is understandable why the two systems have simply co-existed on different paths. But this is an explanation, not a justification. For that, we need to look at how the law operates on the ground. Why does it matter whether public law and private law are aligned? The principal reason is that public services are not delivered exclusively by public employees. Although the French have not gone as far as the UK in the matter of privatisation of public services, a lot of public services are delivered by way of concession, rather than directly. If the service is in concession, then the relationship of the user to the service is governed by private law. Should it make a difference whether the service is run by a public authority or run by a private company? I remember a French judge explaining to me that, in the 1980s, blood transfusion was run either directly by public hospitals or was outsourced to private centres. So in different, but neighbouring, départements liability for contaminated blood could be regulated by different rules. That was leading to different results. It is not simply that public law and private law should be mutually influential, as Chapus described. They should lead to results that, taken together, are coherent. The user of public

99 See J Bell, ‘Path Dependence and Legal Development’ (2013) 87 Tulane L Rev 787. 100 ibid 797. 101 ibid 799–805. 102 See J Bell, French Legal Cultures (London, Butterworths, 2001) chs 3 and 5. 103 See F Audren and J-L Halpérin, La culture juridique française. Entre mythes et réalités. XIXe–XXe siècles (Paris, CNRS éditions, 2013) 9–10.

442  John Bell services is entitled to a sense of homogeneity – that, however a specific public body has organised the local service, there will be similarity in the way the user will be handled and treated in relation to the service and to complaints. There is still work to be done. Does the French experience have any lessons for the notion of tortious or delictual liability? If both civil and administrative law are developed around the need to indemnify the victim of a harm, then the question arises about when that occurs. As I noted earlier, Viney carefully charts the wide range of compensation systems which have grown up in France over the years to provide compensation to those whom society does not think should have to shoulder the burden alone. Social solidarity is a deeply embedded French idea and ‘fraternity’ has recently been recognised as a constitutional principle.104 But that is not the function directly of delictual liability. As my predecessor as editor of the Cambridge Law Journal and as professor of comparative law, Jack Hamson, pointed out, there is a distinction between duties of charity and duties under tort law to make reparation for wrongs done by the tortfeasors.105 Both deal with the victims of accidents, but on distinct moral bases. As Honoré pointed out, there is a difference between misfortune as a result of the acts for which another is responsible and as a result of bad luck.106 Where do our duties to pay compensation arise? For Hamson, responsibility to pay compensation arose from wrongdoing. In relation to strict liability, Honoré argued that outcome responsibility arises from inflicting a harm on another without justification.107 Both of these fit civil liability in France. Hamson’s justification fits fault liability and Honoré’s fits no fault – there is no justification for your thing or your supervisee to harm me, even if you are in no way to blame. In public law, there are clearly many situations of wrongdoing or unjustified creation of risk of harm. But faute de service goes beyond wrongdoing and includes simple non-performance, such as a failure to provide classes in a subject at school. Furthermore, equality before public burdens goes beyond both. If we ask why the state has to pay if returning British holidaymakers are unable to get home because protestors are blockading a French port, then the answer is really simply because a public authority’s decision is the proximate cause of not removing the blockade.108 In public law terms, the decision of the public authority is justified and its consequences are not wrongful. It is simply that it is unfair to leave the loss where it falls. In many ways, this is very like the liability of the State under article L-211-10 of the Code of Internal Security (from 2012) which makes the State civilly liable to the harm caused through violent offences committed by armed or unarmed mobs or gatherings. (This is the current equivalent of the English Riot Compensation Act 2016.) Indeed, the rationale is much the same, but the losses in question are

104 CC

decisions nos 717/718 QPC of 6 July 2018, Cédric H, [7]. Hamson, ‘The Moral Law and Professor Tunc’ (1973) CLJ 52, 53. 106 T Honoré, Responsibility and Fault (Oxford, Hart Publishing, 1999) ch 2. 107 ibid 78. 108 See Sealink (n 74). 105 CJ

The Reform of Delict in the Civil Code and Liability in Administrative Law  443 different and the activity of the protestors is not necessarily violent. So it is more like an extension by analogy of a principle recognised in medieval law and carried through into modern legislation about the kinds of loss an individual should be allowed to suffer. This is not to say that there are not significant areas of overlap between administrative law and civil law. But they are not necessarily doing the same job. When we compare French delict with English tort and German delict, we may come to similar conclusions. Tort law in each country does not do the same job. In particular, German law is much more rigorous in compartmentalising fault-based liability from other regimes of compensation, which exist outside the Code. We take from these national comparisons that tort law does not have to do the same job in every legal system. Similarly, when we compare administrative and civil liability regimes within France, we can accept that they need not perform exactly the same functions. Chapus concluded his major work on civil and administrative liability by expressing regret that the Conseil d’Etat did not articulate more clearly the principles on which administrative liability is based.109 In the subsequent 60 years, it is the work of Chapus which contributed greatly to systematise thinking in administrative law, even if a number of his conclusions might be contested. But, in looking at the reform of the Code civil, I am left sharing his regret, but directed this time at the civil law. There does seem to be a forest of single instances of liability without much general principle. Whittaker’s suggestion that we focus on comparing corporate liability in civil law with the liability of the administration is a good way forward, but more thought is needed about how different these should be in a modern society.



109 Chapus

(n 1) 558.

444

21 The Importance of Terminology in the Law of Civil Liability OLIVIER DESHAYES

I.  The Importance of Terminology in the Law The law is made of rules, and these rules are made of words. In this way terminology is of decisive importance for jurists, for on it depends the meaning of the rules and our understanding of their scope. As a result, the accessibility and predictability of the law, those two mainstays of the legal order, also depend on the use of an adequate terminology in the law. As Camus famously observed, ‘to name things badly is to add to the unhappiness of the world’; the jurist could paraphrase this as ‘to employ a defective terminology in the formulation of a legal rule is to add to the unhappiness of litigants and to weaken the law as a whole’. Perhaps, more optimistically, we could borrow from the great Cornu and also say that ‘la justice du droit [gagne] quelque chose à la justesse des mots’.1 This is, of course, all very well-known and need not detain us long.

II.  The Importance of Terminology in the Law of Civil Liability In comparison with other areas of the law, does the law of civil liability2 bear any special feature requiring a separate treatment of the wider question of the importance of legal terminology? In my view, it clearly does not. The choice of words is just as important here as elsewhere – no more and no less.

1 G Cornu, Linguistique juridique, 3rd edn (Paris, Montchrestien, 2005) foreword, v. In English, this could be rendered as ‘legal justice gains from the correctness of words’, but this loses the elegant play on the near homonyms of la justice and la justesse. 2 In this chapter, by ‘civil liability’ I refer to extra-contractual or delictual liability.

446  Olivier Deshayes The most that can be said is that the function of civil liability is to distinguish between what is permitted and what is forbidden in a different way from, for example, criminal law. Indeed, in the case of the latter the principle of the legality of criminal offences and penalties requires a particularly strict definition of what is forbidden with the consequence that behaviour which is not precisely foreseen by a criminal law rule cannot attract any sanction. As a result, once the legislator has identified those unlawful acts which it considers ought to attract a sanction it then tirelessly seeks to define them precisely. By contrast, in the case of civil liability, a broad flexibility is both possible and necessary: all harms unfairly caused to another person should lead to an obligation in the person who caused them to make reparation. In defining what is a harm unfairly caused, it is impossible to be perfectly precise, no matter how much care and attention is devoted to terminology. One can, of course, divide liability into a series of bite-sized requirements (harm, unlawful act or fait générateur, causation, etc). But one always comes up against the same difficulty in the end, which is that these requirements refer to each other. A widely accepted starting-point is that liability should be imposed where the normal course of events has been deviated by behaviour which is inappropriate, unlawful and which results in a negative and detrimental result for somebody else. But then, what is a deviation from the ‘normal course’ of events if not a modification of the course of events by an unlawful act? And what is an unlawful act for the purposes of civil liability if not one which is disapproved by the legal system on the ground that it exposes others abnormally to a risk of harm? And what is a ‘harm’ if not the difference between what is and what should have been, that is, a detrimental deviation from the course of events? And so the argument goes round and round. The unlawfulness of the action which gives rise to liability, its fait ­générateur, is linked to causation, which then refers us to harm, and so on. Everything is connected to everything else. We should not conclude from this that terminology does not matter, but simply that it cannot do everything. In my view, we cannot avoid the fact that the reasoning employed in the law of civil liability is to a large extent circular. On the other hand, what terminology can do is to help us structure our reasoning, to make it more reliable, less subjective and more transparent. It is a lack of transparency which can be seen as the weakness of the French law of civil liability.

III.  The Importance of Terminology in the French Law of Civil Liability The French law of civil liability possesses some characteristics which are very distinctive, imprinted on it by its history and by the generality of the legislative provisions which govern it. It may be compared to a DNA ribbon which has not been unravelled and whose main mechanisms are therefore hidden from view. The  French reform which is now in the course of discussion would lead to the

The Importance of Terminology in the Law of Civil Liability  447 partial stretching out of this DNA ribbon. This may be a cause for celebration, but we should not hide the difficulties of the undertaking.

A.  Characteristic Features of the French Law of Civil Liability If one leaves aside the special regimes of civil liability which have themselves appeared only quite recently, from 1804 to the present day the French law of civil liability rests on five articles of the Code civil: formerly articles 1382 to 1386 and, since the Ordonnance of 2016, articles 1240 to 1244. We know that these ­provisions are not only very few; they are also seriously incomplete. Indeed, all they do is list the actions which may give rise to liability (fault, and the action of things, of other persons, of animals, and of buildings). They contain no details on the extent of reparable harms, on the ways in which reparation is to be effected, on liability for resulting loss, on causation, on possible grounds of exoneration etc. It has therefore actually been left to the courts to fashion the law of liability as we know it today. As Denis Mazeaud has expressed it: today it is difficult for litigants, for whom the law is made, and even for jurists, who have a duty to know it, to gain access to the actual law of civil liability, of which our Code gives only a very misleading picture. The Code is, in fact, incomplete.3

The development of the law by the courts has been effected by means of only the handful of conceptual tools which are present in the Code civil: ‘fault’ (faute), ‘action’ (fait), ‘harm’ (dommage), ‘keeper’ (gardien), and ‘causation’ (cause). This has had two consequences. First, from the point of view of what may be called technique, the courts have had a free hand. Nothing has required them, for example, to explain the nature of the causal relationship between the action attributed to the defendant and the claimant’s loss, whether in terms of accepting the theory of adequate causation, the theory of the equivalence of conditions or any of their numerous variants. The courts have indeed taken great care not to take sides in this dispute, so that even today there remains a degree of mystery on this aspect of causation.4 Nor does anything require French courts to distinguish between types of harm according to a set pattern, and this to such an extent that one sometimes cannot exactly know which harms are in fact being compensated. Secondly, and above all, as to the politics of the law of civil liability, French courts have found no obstacle in the legislation to the compensation of losses, no barriers against the claims of victims. As has been said, the Code civil placed 3 D Mazeaud, ‘Réflexions sur le projet de réforme de la responsabilité civile du 13 mars 2017’ in Études à la mémoire de Philippe Neau-Leduc (Paris, LGDJ, 2018) 711, [3]. 4 See eg G Viney, ‘Conclusion prospective’ RLDC Suppl to no 40 on “Les distorsions du lien de causalité”’ (1 July 2017) 101ff.

448  Olivier Deshayes no condition as to a person’s ability to recover in respect of harm, no condition as to causation, nor did it set out any grounds of exoneration for those otherwise liable. At most, we can find traces of conditions in the provisions governing liability for the actions of others.5 But the most important provisions are in contrast completely general. First among these is the famous former article 1382 Cc (now article 1240), according to which: Any human action whatsoever which causes harm to another creates an obligation in the person by whose fault it occurred to make reparation for it.

This certainly does not mean that French courts have accepted the reparation of all harms in all circumstances without limit. For example, they laid down the famous, if now outdated formula, that harm can be the subject of reparation only if it was ‘direct, current and certain’.6 But restrictions such as this can only be reduced in number for the good reason that where legislation does not make any distinction, in principle no distinction should be made.7 For example, in 1863 the Cour de cassation was obliged to hold that ‘article 1382 … has no restrictions at all as to the nature of the proven harm’.8 All these circumstances have led, particularly from the end of the nineteenth century and on the development of insurance, to the development of a system of liability which is very favourable to victims, indeed so favourable that one scholar has gone as far as to talk of a true ‘ideology of reparation’.9 To summarise, the French law of civil liability possesses two characteristics: it uses a small number of concepts that are broad and largely undetermined and it has few – if any – brakes on claims by victims leading to a broad and undifferentiated compensation of harms of all types. The use of better terminology could help the French law of civil liability in the following respects. As regards judges, it would give a clearer view of how the rules are applied, unravelling hidden reasoning and helping them to draft solutions that are more convincing from a technical and a logical point of view. As regards the legislator, it would oblige the law to spell out its policies in relation to civil liability. Of course, we still have to agree on what is meant in the present context by a better terminology. The French language is not short of words and French law is not short of jurists keen to ensure good usage. It may be thought, though, that what is lacking are the concepts, the tools by which we can understand the inner workings of civil liability which would allow us to stretch out the chain of ­liability

5 Fathers and mothers are liable only for harm caused by their minor children who live with them; teachers and artisans are liable for harm caused by their pupils and apprentices only in respect of the period where they are under their supervision. 6 See P le Tourneau (ed), Droit de la responsabilité et des contrats, Régimes d’indemnisation (Paris, Dalloz, 2018) [2123.61], 517. 7 This evokes the Latin maxim ubi lex non distinguit, nec nos distinguere debemus. 8 Cass crim 20 February 1863, S 1863, 1, 321, rapp Nouguier. 9 L Cadiet, ‘Sur les faits et les méfaits de l’idéologie de la réparation’ in Mélanges offerts à Pierre Drai: Le juge entre deux millenaires (Paris, Dalloz, 2000) 495ff.

The Importance of Terminology in the Law of Civil Liability  449 and analyse it in a more detailed and refined way. As it is, and if one keeps to what is called the general law of civil liability, there are only three elementary driving forces at work: actions which give rise to liability (faits générateurs), harm (dommage), and causation, as is set out at Table 21.1. Table 21.1  Analytical Table of Civil Liability in France Action giving rise to liability

→ Causal connection

Harm

− fault − the action of things − the action of others − nuisance (trouble du voisinage)

B.  Towards a Better Terminology for the French Law of Civil Liability The DNA ribbon which forms civil liability may be unravelled by pulling it at either end. Let us start with the final element, that is to say, with harm. Following a classic presentation, harm is patrimonial, corporeal or moral.10 The tendency has long been to allow the victim to recover in one claim all kinds of harm so as to ensure his or her full compensation. But it is obvious that corporeal harm (or, to use the more usual English expression, personal injury) can have implications both of a patrimonial and of a moral nature and that it therefore does not constitute a category of harm at the same level as the others. It is logically better to distinguish between harms which originate in personal injury (which can themselves be ­patrimonial or moral) and those which do not originate in personal injury (which can equally be either patrimonial or moral). Over the last few years, a way of drawing a reliable dividing line has been put forward by scholars. It consists of distinguishing between harm (dommage) and loss (préjudice),11 following the example of many foreign laws. According to this 10 The significance of a patrimonial harm (dommage patrimonial) is that it affects the patrimoine, that is, the wealth or estate of a person: these harms include what an English lawyer would see as damage to property or financial loss. ‘Moral harm’ (dommage moral) refers to harms which are not patrimonial and include grief, upset, and distress. ‘Moral harm’ usually derives from a personal injury but sometimes derives from a physical damage to property. 11 In favour of this distinction see in particular L Cadiet, ‘Le préjudice d’agrément’ (Thesis, ­Poitiers, 1983) [323]ff; Y Lambert-Faivre and S Porchy-Simon, Droit du dommage corporel, 6th edn (Paris, Dalloz, 2008) [86]; P le Tourneau (ed), Droit de la responsabilité et des contrats, Régimes d’indemnisation, 11th edn (Paris, Dalloz, 2017) [1304]; J-S Borghetti, ‘Les intérêts protégés et l’étendue des préjudices réparables en droit de la responsabilité civile extra-contractuelle’ in Études offertes à Geneviève Viney (Paris, LGDJ, 2008) 145.

450  Olivier Deshayes approach, harm is the infringement of a protected interest (for example, property, the human body or the environment) while losses are the different negative repercussions of such a harm for the victim (whether financial or moral loss).12 The adoption of this distinction allows a better classification and a better ­differentiation of the kinds of loss. In terms of the future reform of the law, the distinction between harm and loss would allow us to retain a general rule according to which any harm gives rise to an obligation to make reparation while modulating this reparation according to the types of losses in question. Similarly, it would allow us to impose one requirement in order to establish the causal connection between the harmful action and the harm itself while leaving open the possibility of imposing a different and possibly more flexible requirement in order to establish the causal connection between the victim’s harm and his or her loss. In sum, it could put into effect a legal policy.13 The DNA ribbon of civil liability may also be pulled from the other end, that is, actions giving rise to liability (faits générateurs). Until now in French law only the concept of causation is entrusted with the task of describing the connection which must exist between a fait générateur and the victim’s harm. However, it has been pointed out that the scope of causation would be improved if it were given a more limited role consisting of attaching the harm to the immediate unlawful action which is its source. The relationship which allows the identification of the person having to answer for the fait générateur would instead be one of­ imputation.14 In cases where the fait générateur is fault, such a requirement of imputation seems to merge with causation: it indicates the person who committed the fault (its ‘author’) as the person liable. In cases of liability for the action of a thing, the relationship of imputation would appear more distinctly: it points to the keeper of the thing whose behaviour can be without any causal connection with the action of the thing and therefore with the harm. In cases of liability for the action of others, the point of distinguishing between causation and imputation is even more striking. Causation allows us only to tie the harm to the unlawful action committed by its direct perpetrator; only imputation can allow us to tie this

12 According to J Flour, J-L Aubert, and E Savaux, Droit civil, Les obligations, 2. Le fait juridique (Paris, Sirey, 2011) [133]: ‘the distinction [between harm and loss] has, among other things, the merit of recalling that every injury to an interest cannot automatically give rise to reparation … [The distinction] is for the moment insufficiently conceptualised and put into practice in the positive law to provide a general and secure guide for analysis of the conditions to which reparation of harm is subject.’ 13 This is the position adopted by the Avant-projet Terré. On this point see P Rémy and J-S Borghetti, ‘Présentation du projet de réforme de la responsabilité délictuelle’ in F Terré (ed), Pour une réforme de la responsabilité civile (Paris, Dalloz, 2011) 61ff, esp [16]: ‘[t]he distinction between harm and loss is not, however, posed in the [Avant-projet Terré] in a dogmatic way (see art 49); it is only a conceptual tool to be used as the means of exposition which appears clearest. To distinguish between harm and loss seemed to be the best way of untangling the two related but nevertheless distinct questions of the extent and of the manner of reparation.’ 14 On which see in particular F Leduc, ‘Causalité civile et imputation’ RLDC Supp to no 40 on ‘Les distorsions du lien de causalité’ (1 July 2017), 21ff. It should be noted that the positions adopted in the present chapter do not reflect those taken by this author.

The Importance of Terminology in the Law of Civil Liability  451 action to the person liable for the action of another who may again possibly have no connection with the harm, that is, took no part in its occurrence. Going even further, some scholars even consider that the action of another person is not a fait générateur of liability to be placed at the same level as fault or the action of a thing. To be precise, they consider that liability for the action of another person is grafted onto a fait générateur such as fault, the action of a thing or a nuisance by allowing the imputation (whether independently or cumulatively) of the harm to a person other than the one indicated by causation. It is this position which was defended by the Projet Terré. If these changes were accepted, the table of liability would find itself modified in the way shown at Table 21.2. Table 21.2  Possible Analytical Table of Civil Liability in France

Person liable

← Imputational connection

Action giving rise to liability

− ‘author’ of fault

– fault

− keeper of the thing

– the action of things

− ‘author’ of nuisance

– nuisance (trouble du voisinage)

→ Causal connection

→ Causal connection Harm ? Losses

+ Person liable for the action of another person

The Projet de réforme of 2017 adopts some of these terminological clarifications. First, it distinguishes between harm (dommage) and loss (préjudice). Thus, for example, it includes a section entitled ‘Special rules governing the reparation of losses resulting from certain types of harm’ within which is found in particular a sub-section entitled ‘Special rules governing the reparation of losses resulting from personal injury’. This choice goes hand-in-hand with the appearance of rules which treat different types of harms differently: this is highlighted in the Projet de réforme by the special favourable exemptions from particular rules which it reserves for the reparation of personal injuries. In addition, the Projet de réforme adopts the presentation of the law according to which liability for harm caused by another person is not itself a fait générateur of liability but is the result of the imputation on the defendant of the fait

452  Olivier Deshayes générateur caused by the person directly responsible for the harm.15 The Projet de réforme adds that liability for the action of another person is premised only on ‘proof of an action of a nature to attract liability in the direct author of the harm’ (article 1245).16 It is pointless, though, to insist further on the advantages that would be gained by clarifications of this kind. Instead, I would like to say a few words about the difficulties which they would involve.

C.  Difficulties with Terminological Change in the French Law of Civil Liability Any attempt of modifying current terminology must face two difficulties. The first relates to the persons for whom the new rules are intended and is due to the fact that it would require to give to words which are already known and used, definitions other than their current ones. It is inevitable that there would be opposition to such a change. What sociologists call ‘path dependency’ makes any change difficult.17 But where the change involves the modification of the meaning of words in common usage, we should expect the opposition to be even stronger. This could be particularly true in the case of the distinction between harm and loss. The vast majority of French lawyers and judges are not used to making this distinction and it is highly probable that it would take decades to change their habits. The second difficulty relates to the legislator itself, as the terminological clarifications involved would require particular care. To avoid confusing their meanings, each use of the terms chosen would have to correspond properly to their new definitions and their significance would have to be understandable without too many doubts. Unfortunately, two illustrations will show that here the Projet de réforme is a little disappointing. The first relates to limitation clauses governing a person’s liability for the action of others. According to the Projet de réforme, contract terms restricting or excluding liability are in principle valid. Nevertheless, article 1283 of the Projet provides that ‘in extra-contractual matters, a person cannot exclude or limit his liability for fault’, implying that a distinction should be made between those liabilities arising from a fault and those which do not. What about liabilities for the actions of others, which, as has already been explained, are not treated in the Projet de réforme as liabilities based on a fait générateur in the defendant but rather as a mechanism of imputing to a third party harm caused

15 See arts 1245ff of the Projet de réforme on ‘l’imputation du dommage causé par autrui’. 16 This would put an end to the case law according to which fathers and mothers are liable strictly for harm caused by their infant children as long as the latter have physically caused the harm, even in cases where the children would not themselves be liable. 17 See eg B Palier, ‘Dépendance au chemin emprunté’ in L Boussaguet, S Jacquot, and P Ravinet (eds), Dictionnaire des politiques publiques (Paris, Presses de Sciences-Po, 2010) 411ff.

The Importance of Terminology in the Law of Civil Liability  453 by the fait ­générateur of another person? Should we hold that terms restricting or excluding such a liability are always valid on the ground that liability for the action of others is never based on a fault in the defendant? Or should we distinguish between cases where the direct author of the harm has or has not himself committed a fault? The drafters of the Projet de réforme should have addressed this question which is raised by their own decision to set the liability for the action of another aside from faits générateurs. The second illustration relates to the specific regime that applies to personal injuries. On a number of occasions, the Projet uses the expression ‘in the case of personal injury’ (en cas de dommage corporel). For example, article 1254(2) provides that ‘[i]n the case of personal injury, only gross fault can lead to partial exoneration’. But the meaning of the expression ‘in the case of personal injury’ is not always clear. Used in article 1254, does it mean that, for  the reparation of losses resulting from personal injury, only gross fault in the victim can lead to the partial exoneration of the person liable? If so, where there is an ordinary fault (faute simple) in the victim, we should distinguish between losses resulting from personal injuries (where full reparation should be awarded) and losses arising from harm other than personal injury (where only partial reparation should be awarded). Or does the expression instead mean that ‘in the case of personal injuries’ the whole of any liability incurred is subjected to the rule ­according to which ordinary fault in the victim has no exonerating effect?

IV. Conclusion In conclusion, the present state of the French law of civil liability is characterised by the existence of a handful of very general rules and by the limited number of conceptual tools employed. This has a number of advantages but also of drawbacks. Clarification of the terminology used may be desirable. This clarification would consist of the sub-division of the key concepts of liability (harm, causation, faits générateurs). It would allow us to refine the reasoning used and also to increase the number of elements on which the imposition of liability could be based so as to enable a diversification of the solutions or, to put it another way, so as to adopt more tailor-made solutions. This is in part the path which the Projet de réforme has chosen to take. It distinguishes between harm and loss and subjects liability to a series of distinct conditions depending on the nature of the loss. The Projet also places the liability for the actions of others in a separate category rather than treating it as a fait générateur on the basis of a distinction between causation and imputation. These terminological clarifications will not be without difficulty, as with all changes. They will disturb, irritate and may attract challenge. Nevertheless, it is a risk worth taking. Let us just hope that the drafters of the new legislative provisions will show a little more care as to terminological issues than the drafters of the Projet de réforme. The meaning of the words or concepts must not be blurred by provisions which use them in a way which is not wholly consistent.

454

22 Principles of Liability or a Law of Torts? JEAN-SÉBASTIEN BORGHETTI AND SIMON WHITTAKER

English law and French law are often contrasted as regards their different approach to tort law – or should we say the law of torts? English law is famous for its discrete approach to the imposition of liability in the absence of breach of contract, as it imposes liability only on the basis of a range of distinct legal grounds (‘torts’), in the absence of which liability will not be imposed. By contrast, the French law of extra-contractual liability1 seems to love nothing better than to posit general and potentially all-encompassing principles, a distinctively French taste which is fully reflected in the Projet de réforme. In our view, however, a closer analysis of the Projet de réforme as well as of French case law reveals that the French law of liability (contractual and extra-contractual) may not be structured along legal grounds or principles that are as general as its usual presentation suggests. This concluding chapter of the book seeks to explain this view, drawing on the discussions developed in earlier chapters. How much principle really is there in the French law of civil liability?

I.  A Taste for General Rules According to article 1240 Cc (long famous as article 1382 Cc), Any human action whatsoever which causes harm to another creates an obligation in the person by whose fault it occurred to make reparation for it.2

1 The terminology of French civil liability remains unsettled, but for present purposes we shall refer to ‘civil liability’ as a broad category containing all examples of non-criminal (and non-administrative) liability (responsabilité), including cases of liability based on contractual non-performance and cases other than contractual non-performance: see further ch 1, ch 2 and ch 3. 2 ‘Tout fait quelconque de l’homme, qui cause à autrui un dommage, oblige celui par la faute duquel il est arrivé à le réparer.’

456  Jean-Sébastien Borghetti and Simon Whittaker It is not clear how the drafters of the original article 1382 of the Code civil (now article 1240) understood the significance of this provision. Was it intended only as an umbrella under which the various instances of fault-based liability that had been recognised over time could shelter, or was it understood as a genuinely general legal rule applicable to all the factual situations in which its conditions were satisfied? While Domat, whose work was clearly the direct inspiration of the formula used by the Code civil,3 may have intended it only in the first sense, the travaux préparatoires of the Code civil suggest that, at the beginning of the nineteenth century, it was understood as stating an independent rule, and not merely as a shorthand for the various discrete cases in which a person could be made liable for harm caused through his fault.4 In other words, article 1382 was seen as expressing a very broad legal rule: where A’ s fault causes harm to B, A is liable to B.

A.  The Significance of Principe At this point, it is probably worth clarifying what is meant by the term principe (principle) in this context. English lawyers tend to distinguish between principles and rules, and while there is no fixed usage here, typically there are three features of ‘principles’ in particular when contrasted with mere rules: first, principles are of relatively general application, not being specific to a particular set or sets of facts; secondly, often a principle is a legal proposition of some importance; and, thirdly, even where a principle is applicable (that is to say, the facts would fall within its terms), it may not actually be applied, depending on the circumstances. So, while a simple ‘rule’ may be expressed in a syllogistic form (if facts A, B and C exist, then legal result X follows), a principle appeals for its own application in the absence of some good reason why it should not apply. In the context of the law of torts, the prominent example of this can be seen in the evolution of the way in which the so-called ‘neighbour principle’ set out in Lord Atkin’s famous speech in Donoghue v Stevenson has been viewed.5 Even at the time, this formulation was not seen as

3 Jean de Domat, Les lois civiles dans leur ordre naturel, 2nd edn (Paris, Auboüin, Emery and C ­ louzier, 1697), vol II 131 (‘Toutes les pertes, & tous les dommages qui peuvent arriver par le fait de quelque personne, soit imprudence, legereté [sic], ignorance de ce qu’on doit sçavoir [sic], ou autres fautes semblables, si legeres [sic] qu’elles puissant être, doivent être reparées par celuy [sic] dont l’imprudence, ou autre faute y a donné lieu’). On the interpretation of Domat’s position, see the leading work by O Descamps, Les origines de la responsabilité pour faute personnelle dans le Code civil de 1804 (Paris, LGDJ, 2005) 423–29. 4 See Tribun Tarrible, ‘Discours, Discussion devant le Corps législatif ’ in PA Fenet (ed) Recueil complet des travaux préparatoires du Code civil (Paris, Marchand de Breuil, 1827) vol XIII, 478 at 488, who, in discussing arts 1382–1383 Cc, declared that ‘[c]ette disposition embrasse dans sa vaste latitude tous les genres de dommages et les assujetti à une réparation uniforme qui a pour mesure la valeur du préjudice souffert’ and referring to the law in these provisions as a principe. On the conceptions of the Code’s drafters, see Descamps (n 3) 462–68. 5 Donoghue v Stevenson [1932] AC 562, 581.

Principles of Liability or a Law of Torts?  457 requiring later courts to impose liability by recognising a ‘duty of care’ in the tort of negligence where harm was reasonably foreseeable, even though one could gain this impression from the terms of the dictum, but rather as allowing or perhaps inviting later courts to find a duty of care and so impose liability where this is the case even where earlier authority appears to deny it. Indeed, Lord ­Macmillan’s dictum in Donoghue that the ‘categories of negligence are never closed’6 may well better reflect the decision’s initial impact.7 Later courts seem to take a more extensive approach, with judges at least appearing to recognise that foreseeability of the claimant’s harm (possibly combined with ‘proximity’) would prima facie lead to the recognition of a duty of care, unless there are good reasons for there not being one.8 This approach was later seen as potentially too expansive, and from the early 1990s the courts generally preferred to adopt the ‘Caparo test’ (where foreseeability and proximity must be supplemented by asking whether it is ‘fair, just and reasonable’ to impose a duty of care9) or, even more conservatively, to adopt an ‘incremental approach’ and so seek analogies from the body of decided cases.10 Indeed, in Robinson v Chief Constable of West Yorkshire Police Lord Reed JSC recently considered that the point of the Caparo case was to repudiate the idea that there is a single test which can be applied in all cases in order to determine whether a duty of care exists, and instead to adopt an approach based, in the manner characteristic of the common law, on precedent, and on the development of the law incrementally and by analogy with established authorities.11

As a result, even in the case of the tort of negligence, which is the most general and broad of the common law torts, English judges reject the idea of a principle in the sense of a genuine test for the existence of a duty of care and, therefore, of the foundation of liability. In novel cases the proper approach is seen as the drawing of analogies from decided cases, which may include weighing up the reasons for or against liability.12 And while some English judges may feel able to identify circumstances in which ‘established’ or ‘ordinary principles’ of liability for negligence recognise the existence of a duty of care, notably, in the case of liability

6 ibid 619. 7 WE Peel and J Goudkamp, Winfield and Jolowicz on Tort, 19th edn (London, Sweet & Maxwell, 2014) [5-016]. 8 See, in particular, Home Office v Dorset Yacht Co Ltd [[1970] AC 1004, 1027 (Lord Reid) and, even more famously Anns v Merton LBC [1978] AC 728, 751–52, where Lord Wilberforce distinguished a two-stage test in which first it was asked whether there was a sufficient relationship of proximity (when prima facie duty of care would arise) and, secondly, whether there were ‘any considerations which ought to negative, or to reduce or limit’ that duty. 9 Caparo Industries Plc v Dickman [1990] 2 AC 605, 617–18 (Lord Bridge of Harwich). 10 This approach was famously advocated by Brennan J in the High Court of Australia in ­Sutherland Shire Council v Heyman (1985) 60 ALR 1, 43–44, but has been much approved, including by Lord Bridge in Caparo, [1990] 2 AC 605, 618. 11 Robinson v Chief Constable of West Yorkshire Police [2018] UKSC 4, [2018] AC 736 at [21] (with whom Baroness Hale of Richmond and Lord Hodge JJSC agreed). 12 ibid [29].

458  Jean-Sébastien Borghetti and Simon Whittaker for positive action causing personal injury, others consider that even here these broad propositions are accurate only in general.13 By contrast, French lawyers see article 1240 Cc (and, indeed, article 1242(1) Cc) as setting out a principe général de responsabilité and in doing so they use the word principe as all but synonymous with règle (rule): principe de droit and règle de droit can equally consist of legal propositions which, if their conditions are satisfied, require that a certain result should follow.14 Indeed, this possible way of understanding the significance of principe may explain a change in the drafting of the reform of the law of contract put in place by the Ordonnance of 2016. In an earlier version of the proposed reform, the new introductory provisions of the Code civil on contract were to be prefaced by a section entitled Principes directeurs (‘guiding principles’), these consisting of freedom of contract, the binding force of contract and good faith.15 However, the Ministry of Justice later abandoned the language of principe to describe these three broad propositions in the text of the reform itself.16 At first sight, this might have been thought surprising, given that the parliamentary legislation which empowered the government to enact the Ordonnance17 itself referred to these ‘general principles of the law of contract’, but the Ministry’s report to the President of the Republic on the Ordonnance explained that this change sought to avoid giving courts the impression that these propositions enjoyed a higher status and therefore could be used to justify a greater judicial activism.18 For the French lawyer, the terminology of principes may go too far in terms of normative effect than is desirable. This may have been thought to be particularly the case of the three legal propositions in question, as freedom of contract and the binding force of contract, and, indeed some particular rules subsequently laid out 13 cf Lord Reed JSC’s appeal to ‘established principles’ and ‘ordinary principles’ of liability in negligence in Robinson v Chief Constable of West Yorkshire Police [2018] UKSC 4 at [26], [27], [29], [30], [42], [64] and [67] in recognising a duty of care in the police for their action causing personal injuries, and the approach of Lord Mance JSC (at [85]–[91], who emphasised (at [85]) that this ‘principle [is] generally correct’ but who was ‘not persuaded that it is always a safe guide at the margins’) and Lord Hughes JSC (at [118]–[120], who emphasised the importance of policy considerations in the context). 14 Principe can have distinct meanings in French law (see P Deumier, Introduction générale au droit, 3rd edn (Paris, LGDJ, 2015) 23–25), but it is undisputed that a principe can have the same normative value as a règle: see eg P Morvan, Le principe en droit privé (Paris, Editions Panthéon-Assas, 1999) 349ff. 15 Avant-projet de réforme du droit des obligations (document de travail) circulated by the French Ministry on 23 October 2013. 16 The substantive provisions remained as arts 1102–1104 Cc. 17 Loi no 2015-177 of 16 February 2015 ‘relative à la modernisation et à la simplification du droit et des procédures dans les domaines de la justice et des affaires intérieures’, art 8 referring to ‘les ­principes généraux du droit des contrats tels que la bonne foi et la liberté contractuelle’. 18 ‘Unlike some of the European instruments, the Ordonnance has not opted for a preliminary c­ hapter dedicated to “guiding principles” [principes directeurs] of the law of contract. Instead, a choice was made to draw inspiration from the existing first chapter of Title III of the Code civil entitled “Preliminary provisions” so as to indicate that, even though intended to give guiding lines [lignes directrices] for the law of contract, the general rules thereby set out as required by article 8 of the authorising legislation [loi no 2015-177, above n 17], do not in fact constitute rules of a higher level than those which follow and which the courts could use as a basis to justify an increased interventionism: they instead consist of principles intended to facilitate the interpretation of all the rules applicable to contracts, and to fill in any gaps between them’: Report to the President the Republic on Ordonnance of 2016.

Principles of Liability or a Law of Torts?  459 in the text itself, could be seen as countered or capable of circumvention by the principle of good faith.19 In referring to the principes généraux of liability for fault in article 1382 (now 1240) and of liability for the actions of things in article 1384(1) (now article 1242(1)) of the Code civil, French scholars do not therefore mean that these articles provide a legislative umbrella under which particular liability rules can shelter, nor that they provide a guideline on the basis of which the courts can create or discover new particular liability rules, the offspring or applications of that principle. Rather, the principles in articles 1240 or 1242(1) constitute legal rules, albeit rules with very broad areas of application. This does not mean, of course, that the courts have no role to play in putting these principles into effect in practice, but in doing so they are not seen as creating new rules. Rather, the courts’ role is seen as interpreting and applying the various elements set out by these general provisions, for example, in the case of liability for fault, the notions of ‘fault’, ‘harm’ and the causal relationship between the two. In his famous introduction to the Code civil, Portalis explained its drafters’ approach to regulation, observing famously that: We have not thought it necessary to simplify legislation to the point of leaving citizens without rules and without a guarantee of their most important interests. We have also managed to avoid the dangerous ambition of wishing to regulate everything and foresee everything. Who would think that it is the very same people to whom a code always appears too voluminous who imperiously dare to impose on the legislator the terrible task of leaving nothing at all to the decision of the judge?20

From this perspective, it might perhaps be better, therefore, not to translate the French principe as ‘principle’ in English, but instead to adopt the German expression ‘general clause’ (used to describe the delict provisions in §§ 823 and 826 BGB), even though to most English lawyers a ‘clause’ suggests a term of a contract rather than a legislative provision.21 Finally, we should make clear that the approach of French lawyers to the principes contained in articles 1240 and 1242(1) Cc can also be seen in relation to principles which are not, at least as yet, explicitly codified such as the ‘principle of full reparation’ (le principe de la réparation intégrale) of

19 On the idea of the principle of good faith as a counter-principle in French contract law see S Whittaker, ‘Contracts, Contract Law and Contractual Principle’ in The Code Napoléon Rewritten, ch 3, 44–46. 20 ‘Nous n’avons … pas cru devoir simplifier les lois au point de laisser les citoyens sans règles et sans garantie sur leurs plus grands intérêts. Nous nous sommes également préservés de la dangereuse ambition de vouloir tout régler et tout prévoir. Qui pourrait penser que ce sont ceux mêmes auxquels un code paraît toujours trop volumineux, qui osent prescrire impérieusement au législateur la terrible tâche de ne rien abandonner à la décision du juge?’ J-E-M Portalis, Discours préliminaire sur le projet de Code civil (presented 1 pluviôse an IX = 21 January 1801) in Discours et Rapports sur le Code civil (Caen, Centre de philosophie politique et juridique, 1989) 6–7. 21 See S Whittaker, ‘Theory and Practice of the “General Clause” in English Law: General Norms and the Structuring of Judicial Discretion’ in S Grundmann and D Mazeaud (eds), General Clauses and Standards in European Contract Law, Comparative Law, EC Law and Contract Law Codification (The Hague, Kluwer Law, 2006) ch 4, 60.

460  Jean-Sébastien Borghetti and Simon Whittaker a claimant’s harm or loss.22 As the proposed provision in the Projet de réforme makes clear, the purpose of reparation (whether in kind or by way of damages) is to ensure the full reparation of the claimant’s loss, neither more nor less.23 The all-encompassing nature of the principle of full reparation has led some French courts to seek ways of avoiding its consequences by fixing requirements as to the nature of the harm to be subject to reparation,24 but if anything this emphasises the way in which the principle is generally viewed.

B.  The Tendency to Generalise As was explained in the Introduction to the present work,25 the evolution of French liability over the last two centuries is often seen as reflecting a move towards very general bases of liability, with at least two general ‘principles’ (for fault and for the action of things) and with at one time a possible third general ‘principle’ (for the actions of others).26 This tendency to generalise and to bring together different things under a single concept or idea, rather than emphasising their differences may be seen as a wider intellectual trait of French civil lawyers, who are keen to ‘systematise’ the raw legal materials which they find either in the codes or in the practice of the courts. Certainly, the majority of French legal scholars appear to be entirely comfortable with and, indeed, to celebrate the breadth of treatment and practical expansion of liability which has resulted from developments in the private law of liability, principally on the basis that this allows French law to vindicate the rights of ‘victims’, in particular of personal injuries and death.27 Rather than questioning the generality of such concepts as harm or fault, legal scholars have endorsed this way of thinking wholeheartedly. This is reflected in the structure of almost all treatises, with harm and causation presented as conditions common to all regimes of liability, together with a single set of defences, in particular force majeure and contributory fault in the claimant (faute de la victime). Similarly, the ‘effects of liability’, ie the rules on reparation and the assessment of damages, are presented as being common to all types of liability, at least where

22 On this principle and its proposed explicit codification by the Projet de réforme arts 1258 and 1259, see above ch 9 and ch 10. Most French lawyers would see this as recognised by art 1231-2 Cc (former art 1149 Cc), even though this relates directly only to damages for contractual non-performance. 23 Art 1258: ‘The aim of reparation is to replace the victim as much as is possible in the situation in which he would have been if the harmful action had not taken place. It must cause him neither a loss nor an advantage’; art 1259: ‘Reparation may take the form of reparation in kind or damages, these two types of measures being able to be combined so as to ensure full reparation of the loss.’ 24 Ch 9 above pp 185–189. 25 Above, pp 3–5. 26 See ch 8 above pp 152–153. 27 See esp G Viney ‘Pour ou contre un “principe general” de responsabilité civile pour faute?’ in Etudes offertes à P Catala (Paris, Litec, 2001) 555, 557 who observes that almost no one in France challenges the existence of the general principle of liability for fault, the exception being P Remy, ‘Critique du système français de responsabilité civile’ (1996) Droit et Cultures 31.

Principles of Liability or a Law of Torts?  461 this is extra-contractual.28 According to this now standard presentation, the only differences between the ‘main’ liability regimes, ie those based on the historical provisions in the Code civil, consist in their faits générateurs, that is, the actions (or perhaps events) which may give rise to liability. A striking exception to this approach is found in Remy’s work. He remarked on French law’s characteristic ‘constantly to reduce the casuistry of the faits générateurs of delictual liability to an abstract principle (fault or risk)’ and he saw this as problematic as it explained why strict liability is seen as based on a ‘principle’ competing with the principle of liability for fault, whereas it should be seen only as an exception.29 In his view, the ‘excessive generality’ of the French law of delict has led to too broad an imposition of strict liability, which in its turn has led to occasional attempts to return liability to fault by circuitous and hidden routes.30 As has been seen in earlier chapters, the two principles of French civil liability law advocated by scholars and recognised by the courts are indeed of an extraordinary breadth. So, liability for fault does not formally distinguish between different types of fault (intentional, grossly negligent (faute lourde) or merely negligent);31 it does not distinguish between different types of harm (dommage), allowing personal injury (dommage corporel, which also includes death), damage to ­physical property, ‘moral’ harm of different types (dommage moral) and economic loss.32 While the original context of ‘liability for the action of things’ was accidents at work and then road accidents, this strict liability has never been restricted to these contexts, nor even to the recovery by claimants of compensation for personal injuries or death.33 To an extent, the Projet de réforme can be seen as the apotheosis of this generalising tendency of French liability law. It brings together la responsabilité délictuelle (now firmly renamed la responsabilité extracontractuelle) and la responsabilité contractuelle (a category which, as has earlier been seen, was itself ­controversial) together under a single legislative category of la responsabilité civile or civil ­liability.34 Moreover, most of the rules on liability in the Projet de réforme are presented as being common to both these branches of liability: this applies to the rules governing harm/loss and causation, the effects of liability and the grounds of ­exoneration.35 The analytical starting-point is clearly intended to be general 28 In the contractual context, there have always been the special rules as to foreseen and foreseeable damages etc in arts 1150–1151 Cc (now arts 1231-3–1231-4 Cc). 29 Rémy (n 27) 32–33. 30 ibid 37 (such as the requirement that the ‘thing’ played an ‘active role’ in the creation of the claimant’s harm and the definition of the ‘keeping’ (la garde) of the thing as a power of its ‘use, control and direction’). 31 See ch 5 esp at p 80. 32 See ch 9, p 181. 33 See above, p 4. 34 See above ch 2 and ch 3, pp 39–47. 35 This is particularly clear from the proposed Sub-Title II, Ch II (‘Conditions of Liability’), which sets out ‘common provisions’ for both contractual and extra-contractual liability for ‘reparable loss’ and ‘causal relationship’, Ch III (‘Grounds of Exoneration or of Exclusion of Liability’) and Ch IV (‘The Effects of Liability’).

462  Jean-Sébastien Borghetti and Simon Whittaker for these fundamental elements and for the most important consequences of contractual and extra-contractual liability.

II.  Diversity Beneath the Generality However, in our view, such a picture of the French law of liability is fundamentally misleading. The patterns of liability are much more complicated and diverse that this ‘classical’ and, to an extent, reductionist picture allows. This may be seen in two main ways. First, there is a considerable diversity of treatment beneath the generality of the classic principles of liability themselves – as is, in some respects and to an extent, reflected in the Projet de réforme itself. Secondly, the French legislature has enacted a series of what have become known as ‘special regimes of liability’, with particular rules adapted for their contexts and with varying relationships with the ‘general law’ arising under the classic principles.

A.  The Nuanced Bases of Liability First, in the case of liability for fault under article 1240 of the Code civil which is seen as providing the fundamental starting-point for liability (le droit commun de la responsabilité) liability is not as general in its application as at first appears, quite apart from the specific legislative regimes which explicitly exclude its application (such as the loi Badinter on motor-vehicle accidents36). As regards the categories of people who can be liable (so-called restrictions rationae personae), while in principle all persons (apart from public bodies37) are covered, the courts have held that article 1240 does not apply to company directors (dirigeants sociaux)38 nor to employees (préposés) acting within the scope of their mission both of whom enjoy a certain immunity from liability.39 As regards the subject-matter of a claim (so-called restrictions rationae materiae), article 1240 does not apply to cases of abuse of freedom of expression against persons;40 it does not apply in cases of the infringement of a person’s privacy,41 nor does it apply in 36 Loi 85-677 of 5 July 1985 (loi Badinter). 37 Public bodies (including the State and local authorities) are often subject to the special rules set out by administrative law, on which see ch 20 above. 38 Cass com 20 May 2003 no 99-17092, Bull civ IV no 84, D 2003, 2623 note B Dondero. 39 In the case of employees (préposés), this was recognised first by Ass plén 25 February 2000 no 97-17378 (affaire Costedoat) D 2000, 673 note P Brun and see P Malaurie, L Aynès and P Stoffel-Munck, Droit des obligations, 10th edn (Paris, LDGJ, 2018) 96–97 noting that later case law has restricted the scope of the immunity in particular by excluding cases in which the employee intentionally commits a criminal offence (Ass plén 14 December 2001 no 00-82066 (affaire Cousin) D 2002, 1230 note J Julien) and has held that it does not extend to the employee’s own liability insurer: Cass civ (1) 12 July 2007, Bull civ I no 270, D 2007, 2908 note S Porchy-Simon. See further ch 8 above pp 154–155. 40 Ass plén 12 July 2000 no 98-11155, D 2000, 463 obs P Jourdain. 41 Cass civ (1) 5 November 1996 no 94-14798, JCP 1997, II, 22805 note J Ravanas.

Principles of Liability or a Law of Torts?  463 the very particular case of ‘faults’ committed on a sports ground which have an effect on bets which were placed on the outcome of the game.42 Interestingly, the Projet de réforme does not say anything about these judicial restrictions on the application of liability for fault. Despite this silence, these exceptions are likely to survive the enactment of the Projet de réforme. Moreover, even where the courts have recognised that, in principle, article 1240 does apply, they have then treated the conditions of liability specially, notably by requiring a particular type of fault in order for liability to be imposed. This was recognised by Starck, who identified a broad distinction between claims for personal injuries and damage to property, where the slightest fault is enough, and claims for economic and ‘moral’ harm, where the courts require particular types of harm (des fautes suffisament caractérisées).43 For example, in claims based on the abuse of rights, in some types of case an intention to harm is required, whereas in other types of case the court considers whether the defendant’s exercise of their right is ‘excessive’;44 in claims in respect of the breaking-off of pre-contractual negotiations by a party, the courts impose liability only in special circumstances, as ‘fault consists generally in a sudden volte-face putting an end to long negotiations which were capable of giving an impression that the contract would be concluded’;45 in claims against third parties to a contract as ‘accessory to a contracting party’s non-performance’, their liability arises only if they knew of the party’s contractual obligation in question;46 and in claims against competitors their behaviour must have been unfair (la concurrence déloyale).47 This final example is particularly illuminating. In the nineteenth century, the courts created the possibility of liability for ‘unfair competition’, so as to balance the otherwise very general principle of freedom of commerce and industry and, therefore, of competition recognised by the revolutionary Décret d’Allarde of 1791.48 They did so on the basis of the general ground of liability for fault in (then) article 1382 Cc, the courts assessing whether the unfairness or fairness of competition in terms of the presence or absence of ‘fault’.49 More recently, however, the legislature has supplemented this case law under the general law by imposing liability in particular circumstances and subject to particular conditions in the Code de commerce.

42 Cass civ (2) 14 June 2018 no 17-20046, D 2018, 1784 note JS Borghetti. 43 B Starck, H Roland and L Boyer, Obligations, 1. Responsabilité délictuelle, 5th edn (Paris, Litec, 1996) 149–50. 44 Malaurie, Aynès and Stoffel-Munck (n 39) 70. 45 ibid 268 noting that this liability is normally delictual and referring to Cass com 26 November 2003 no 00-10243 and no 00-10949, D 2004, 869 note AS Dupré-Dallemagne. 46 Such a person is termed a tiers complice à l’inexécution d’une obligation contractuelle. The condition of knowledge is clear from Ass plén 9 May 2008, Bull civ Ass plén no 3, RTD civ 2008, 485 note P Jourdain. 47 Cf the position of employees acting in the course of their employment whose liability can arise only where they deliberately committed a criminal offence: see above, p 462. 48 Loi of 2 and 17 March 1791. 49 Louis Vogel, Traité de droit commercial Georges Ripert/René Roublot, 18th edn (Paris, LDGJ, 2001), vol 1 590–91 and 601ff.

464  Jean-Sébastien Borghetti and Simon Whittaker For example, since 2008 the Code de commerce has provided that various types of commercial person are ‘liable’ for ‘subjecting or attempting to subject’ their commercial partner to ‘obligations creating a significant imbalance in the rights and obligations of the parties’.50 Secondly, in the case of liability for the actions of things under article 1242(1) (former article 1384(1) Cc), the formal position has become ever more general since the two particular examples of ‘liability for things’ in the Code (liability for animals and for ruinous buildings) have been interpreted so as to lose their distinctiveness51 to the extent that they are not specifically regulated by the Projet de réforme. However, underneath the generality of the principle of this strict liability for the action of things, the courts have accepted a number of rather strange distinctions in deciding whether there can be said to be an ‘action of the thing’ (le  fait de la chose), distinguishing between moving things in contact with the claimant or their property, non-moving things, and things which suffer from defects or ‘abnormality’.52 This complex case law may be open to criticism, but its content was nevertheless included explicitly in summary form in article 1243 of the Projet de réforme.53 More generally, as Knetsch observes, ‘it is not by chance that presumptions of causation find their preferred area of application in the field of strict liability’.54 Moreover, while liability for the actions of things under article 1242(1) Cc (formerly article 1384(1) Cc) is typically presented as very general, its practical scope of application has become increasingly restricted, and its applications in the modern case law are surprisingly few and far between. In particular, it does not apply to motor-vehicle accidents which fall within the scope of the special rules provided by the loi Badinter of 1985,55 nor to claims falling within the special rules governing product liability where the particular claim against a producer is necessarily based on the lack of safety of a product.56 To the extent to which the Projet de réforme would apply the special regime governing motor-vehicle 50 This provision was enacted as art 442-6-I.2o Ccom, but was amended and renumbered by Ordonnance 2019-359 of 24 April 2019, art 2 as art L 442-1-I Ccom. 51 Malaurie, Aynès and Stoffel-Munck (n 39) 107. See art 1385 Cc (now art 1243 Cc); art 1386 Cc (now art 1244 Cc) (ruinous buildings). 52 For a summary in English see S Whittaker ‘The Law of Obligations’ in J Bell, S Boyron and S Whittaker (eds), Principles of French Law, 2nd edn (Oxford, Oxford University Press, 2008) ch 10, 386–88. 53 Art 1243(2)–(3): ‘(2) An action of a thing is presumed wherever, while moving, it comes into contact with the person or property which is harmed. (3) In other cases, it is for the victim to prove the action of the thing by establishing either its defect or the abnormality of its position, its state or its behaviour.’ For an overview of the case law see F Terré, P Simler, Y Lequette and F Chénedé, Droit civil, Les obligations, 12th edn (Paris, Dalloz, 2018) 1057ff. 54 See ch 7 above p 140. 55 The regime of liability under the loi Badinter is termed ‘autonomous’ meaning that no victim of an accident can rely on the general law of liability (whether for fault or, more importantly, for the actions of things): Cass civ (2) 4 May 1987, Bull civ II no 87. 56 Com civ (1) 11 July 2018 no 17-20154 thereby interpreting strictly art 1386-18 (now 1245-17) Cc which allows a person falling within the product liability regime to claim on the basis of the law of contractual or extra-contractual liabilities or special liability regimes.

Principles of Liability or a Law of Torts?  465 accidents, it would restrict further the practical application of the ‘general law’ of liability for the actions of things.57 Thirdly, in the case of liability for the actions of other people, it is often said that in its famous decision in Blieck in 1991 the Cour de cassation recognised a new general principle of liability for another’s action based on article 1242(1) (formerly article 1384(1)) of the Code civil mirroring the principle of liability for the action of things.58 However, this is not in fact the case – indeed, it would be difficult to see how a rule could work if it stated simply that ‘one is liable for the action of other persons’. Certainly, if one looks at the case law after Blieck, it is clear that article 1242(1) has instead been treated as providing an umbrella under which two distinct and well-defined situations of liability for another’s action can shelter: liability of persons in charge of organising and controlling the ‘lifestyle’ (le mode de vie) of other persons for harm caused by those persons; and, secondly, liability of sports associations for harm caused by their members in the practice of the sport.59 Beyond these two recognised categories of liability for the actions of others, the Code civil itself still sets out a series of particular examples: liability of employers (commettants) for their employees (préposés), liability of parents for their minor children, and liability of teachers or artisans for their pupils or apprentices.60 As Häcker has explained, the Projet de réforme implicitly rejects the idea that liability for the action of others is based on a ‘general principle’ as article 1245(1) of the Projet providing that ‘[a] person is liable for harm caused by another person in the cases and subject to the conditions laid down by articles 1246 to 1249’ and the latter provisions then set out a series of p ­ articular situations drawn from the existing law in which such a liability is recognised.61 Of these situations, three are stated to be strict liability (the liability of parents and others for minor children, the liability of a person charged with the care of an adult, and the liability of employers for their employees’, whereas one rests on a rebuttable presumption of fault (‘other persons who take on by contract, and by way of their business or profession, a task of supervision of another person or the organisation and control of the activity of another person’).62

57 Arts 1285–1288 of the Projet de réforme. 58 See eg M Fabre-Magnan, Droit des obligations, 2 – Responsabilité civile et quasi-contrats, 4th edn (Paris, PUF, 2018) 458–62 who considers that the principle has remained marginal compared to the particular examples in the Code but that the scope and regime governing the new ‘general principle’ of liability for the actions of others have been made clearer, although uncertainties remain. 59 See ch 8 above pp 152–153. 60 Art 1243(3)–(8) Cc (formerly art 1384(4)–(8) Cc). 61 Ch 8 above, pp 155–156 where it is also explained that, unlike the existing law, the Projet de réforme does not treat these cases as examples of ‘actions which give rise to liability’ (faits générateurs de responsabilité) but rather as cases in which there is ‘proof of an action of a nature to engage liability in the direct author of the harm’: art 1245(1) of the Projet de réforme. In this way, as the title of the relevant section states, liability arises as a result of the ‘imputation of harm caused by another person’. 62 Arts 1246–1249 Cc.

466  Jean-Sébastien Borghetti and Simon Whittaker Overall, therefore, we can see that there is a complexity and a diversity of treatment in the established law beneath the generality of the ‘principles’ of ­liability for fault, for the action of things and for the actions of other people.

B.  A Unified Regime for ‘Civil Liability’? Many general works on the law of obligations set out the law in a way which treats the rules governing liability to a common framework, apart from the various sources of the liability, its faits générateurs. This can be seen in the treatment of causation, harm or loss (where a principle of ‘full reparation’ applies), and defences (such as the victim’s contributory fault and force majeure). This approach is reflected in the Projet de réforme as it would create a common framework of treatment of ‘civil liability’ so as to include both contractual and extra-contractual liability.63 However, on closer examination a number of rules which are presented as applying to all types of liability are in fact specific either to extra-contractual liability or to contractual liability – or even to the ‘special regimes of liability’ which the Projet includes within its scope. In some cases, the distinct treatment of the rules governing contractual and extra-contractual liability is explicit in the text of the provisions themselves, as in the case of the definitions of force majeure,64 the rules governing the effect of exclusion clauses65 and possibly also penalty clauses,66 and the restriction of actions for cessation of unlawful behaviour and civil penalties to extra-contractual liability.67 There are, however, a number of rules which appear to apply equally to contractual and to extra-contractual liability, but in fact are likely to apply in practice only to extra-contractual liability. This stems from the new and radical treatment of liability for personal injury (dommage corporel) found in article 1233-1(1) of the Projet, which provides that: Losses resulting from personal injury are subject to reparation on the basis of the rules of extra-contractual liability, even if they are caused in the course of performance of a contract.

Although article 1233-1(2) recognises an exception to this rule for the case where an express contractual stipulation is more favourable to the claimant than the 63 Above, ch 1. 64 Art 1253 of the Projet de réforme sets out the definition of force majeure for extra-contractual liability and cross-refers to art 1218 Cc (in force) for the definition of force majeure for contractual liability. 65 Arts 1282 and 1283, providing exceptions to the general recognition of the validity of contract terms both for contractual and extra-contractual matters of ‘contract terms whose object or effect is to exclude or to limit liability’. 66 Although art 1284’s rules on clauses pénales reflect established law (art 1231-5 Cc) and are generally seen as concerning penalties imposed for contractual non-performance, Z Jacquemin (ch 15 above p 326) raises the question whether art 1284 could equally apply to penalties agreed to govern extracontractual liability. 67 Art 1266 (actions for cessation on which see ch 18 above); art 1266-1 (civil penalties on which see ch 16 above).

Principles of Liability or a Law of Torts?  467 a­ pplication of the rules of extra-contractual liability, this is unlikely to be significant in practice. If this provision were enacted, the vast majority of claims for damages for personal injuries would necessarily be on the basis of extra-contractual liability.68 This means that where the Projet sets out rules governing personal injury which apparently apply both to contractual and to extra-contractual liability, in practice they will apply almost exclusively to extra-contractual liability. This is true of the special rule of causation governing personal injury caused ‘by an undetermined person among two or more identified persons acting in concert or exercising a similar activity’,69 the special rules governing the assessment of reparation for personal injuries,70 and the exclusion of liability for personal injury.71 This proposed treatment is particularly interesting for our present concern. For while it is the case that the extraordinary developments in the French law of civil liability, both extra-contractual (liability for the action of things) and contractual (liability arising from non-performance of ‘safety obligations’), were inspired by a concern to ensure compensation for victims of personal injuries, the liabilities themselves are not so restricted.72 The Projet de réforme would give formal recognition both to the importance and to the distinctiveness of this type of harm in contrast to the classically undifferentiated approach earlier described. As will be explained, this reflects a more general shift in approach towards the recognition of distinctions in the law of liability according to the nature of the harm suffered by the claimant. We can see a similar contrast between initial appearance and reality in the Projet de réforme’s treatment of defences. For while Chapter III of the Projet de réforme presents the ‘grounds of exoneration or of exclusion of liability’ in a way which looks as though it applies to liability in general, this presentation is ­misleading.73 A first and obvious point which we have already noted is that the Projet explicitly accepts that the definition of force majeure should differ as between contractual and extra-contractual liability,74 but we should add that in practice in the context

68 One of the apparent consequences of this approach is the intended abandonment of the complex case law which has recognised for more than a century that contracts may contain (contractual) obligations as to the safety of the other party or their property (obligations de sécurité). These obligations may either require proof of negligence (in the case of obligations de moyens) or they may instead impose strict liability with a defence of force majeure (in the case of obligations de résultat). On this existing law see Malaurie, Aynès and Stoffel-Munck (n 39) 542–44. 69 Art 1240 of the Projet de réforme, on which see ch 12 above pp 251–252, and ch 13, pp 277–280. 70 Arts 1267–1277 of the Projet de réforme. 71 Art 1281(2) of the Projet de réforme. 72 Thus, both these grounds of liability may apply to recover for damage to property (dommage materiel), such as in a motor-accident before the loi Badinter. See eg Cass civ (2), 26 September 2002, no 00-18627, Bull civ II no 198, JCP G 2003, I, 154, no 34, obs G Viney, ruling that the owner of a restaurant which had to be closed due to the risk that rocks would fall from a nearby cliff could claim damages for pure economic loss against the owner of that cliff on the basis of art 1384(1); Cass civ (1), 9 July 2015, no 14-13423, ruling that the the company in charge of carrying passengers with their luggage has an obligation (obligation de sécurité) to ensure that the luggage does not get damaged in the course of travel. 73 For further analysis of the Projet de réforme’s treatment of causes d’exonération see ch 14 above. 74 Art 1253 of the Projet de réforme and art 1218 Cc.

468  Jean-Sébastien Borghetti and Simon Whittaker of extra-contractual liability this defence applies only to liability for the action of things.75 Secondly, article 1257 of the Projet de réforme provides for ‘grounds of ­exclusion’ of civil liability applicable generally, stating that: An action causing harm does not give rise to liability where its author finds himself in one of the situations foreseen by articles 122-4 to 122-77 of the Criminal Code.

As Steel has explained, these faits justificatifs are largely drawn from the provisions of the Criminal Code governing, for example, self-defence or necessity.76 It is also clear that the way in which these exclusions of liability take effect is that their existence is incompatible with fault in the ‘author’, that is the person who acts and whose liability is in issue.77 This has a direct implication for the practical ambit of these ‘grounds of exclusion’ as they will be significant predominantly where liability is itself sought on the basis of the author’s ‘fault’. Thirdly, the victim’s contributory fault is not applied uniformly across the whole of the law of civil liability, in particular being subject to a series of special rules in the law governing road accident and product liability.78 Finally, the existing law governing a victim’s ‘acceptance of risks’ apparently differs according to the nature of the harm, a broad distinction being drawn between interferences with property and privacy and related rights on the one hand and most cases of personal injury on the other.79 In this respect, ­article 1257-1 of the Projet de réforme appears to concern only the first category of case, leaving the second unregulated.80

C.  Distinctions According to the Type of Interest of the Claimant One of the axioms of the French law of civil liability is that it does not distinguish according to the nature of the claimant’s harm or, to put it another way, according to the nature of the interest of the claimant which has been prejudiced

75 See J-S Borghetti, ‘L’avant-projet de réforme de la responsabilité civile. Commentaire des ­principales dispositions’ D 2016, 142, [34]. Force majeure cannot coexist with a fault on the part of the defendant, which means that, if fault has been established by the claimant, the defence of force majeure cannot be raised. Moreover, force majeure is not accepted as a defence in most special regimes of liability, including those covered by the Projet, ie liability for the action of motor-vehicles and liability for defective products. Finally, the case law is not clear as to whether force majeure can apply in cases of abnormal nuisance between neighbours. 76 Ch 14 above pp 294–297. 77 ibid p 295. 78 ibid pp 301–303. 79 ibid p 304 and ch 15 above, pp 318–319. 80 Ch 14 above, p 303.

Principles of Liability or a Law of Torts?  469 by the defendant’s action or omission.81 But is this position reflected in the Projet de réforme? While the ‘principles’ of liability for fault and for the action of things in articles 1241 to 1243 of the Projet make no distinctions of this sort, on looking more closely, there appear to be a number of ways in which the type of harm can indeed have an impact on whether liability will be imposed. First, as has earlier been noted, if implemented, the Projet de réforme would generally bar the application of contractual liability in cases of personal injury, and the effect of this would be to apply specific rules governing this type of harm, which are broadly intended to be more favourable to claimants. This would mark a radical change from existing law. Secondly, this time reflecting existing law, both the ‘principal special regimes of liability’ included in the Projet (and therefore to be included in the Code civil) do in fact distinguish according to the nature of the claimant’s harm. So while the ground of liability for the ‘driver or keeper’ of a vehicle for road accidents does not distinguish according to the victim’s harm, the treatment of the victim’s fault does indeed do so, it having no effect where the victim has suffered personal injury (with a series of exceptions and special rules), whereas it may limit or exclude liability where the victim has suffered physical damage to property.82 And while the special legislative regime governing harm caused by defective products in the Projet de réforme applies without qualification where the victim has suffered personal injury, it applies only to certain cases of damage to property, namely where the damage is other than to the defective product itself, causes a loss above an amount set by decree (currently 500 euros) and ‘the property is of a type normally intended for private use or consumption and was used by the victim mainly for his own private use or consumption’.83 Thirdly, the particular situations which were earlier outlined where liability for fault under article 1240 Cc (and apparently equally under articles 1241 and 1242 of the Projet de réforme) is excluded by reason of the subject-matter of the claim, typically involving claims for financial or purely ‘moral’ harm.84 Finally, and much more generally, even as regards the application of liability for fault, in reading the case law one is left with the impression that the courts, especially the Cour de cassation, are much more willing to accept findings of ‘fault’ by the lower courts, the juges du fond, where the claimant has suffered a serious harm.85 And very broadly, while ‘virtual fault’ (la faute virtuelle, where fault is 81 This is often contrasted with the approach of German law, whose § 823(1) BGB lists the interests of a person to be protected. 82 Arts 1287 and 1288 of the Projet de réforme reflecting loi no 85-677 of 5 July 1985, arts 2–5. 83 Art 1290 of the Projet de réforme. The first two restrictions reflect the current position in the Code civil (art 1245-1 Cc), but the restriction as to the character of the property damaged is not contained in the Code civil except for the purposes of the validity of exclusion clauses: art 1245-14 Cc. This restriction is, however, contained in art 9 of the Product Liability Directive 1985 which these special liability provisions seek to implement. 84 Above, pp 462–463. 85 cf Starck, Roland and Boyer (n 43) 149–50. A good example is Cass civ (2) 18 May 2000 no 98-12802, Bull civ II no 85; in this case, where a person fell while climbing a rock, bringing together in

470  Jean-Sébastien Borghetti and Simon Whittaker really fictitious) has been recognised in cases of death or personal injury, in claims for purely financial losses a claimant may often find establishing fault in the defendant much more difficult. Indeed, at times it could be thought that the general law of civil liability operates under a points system under which each element of liability (fait générateur, causation, harm) is worth a certain number of points and claimants win where they gain a certain total number of points. Let us say, for example, that liability is established as soon as the claimant scores a total of 10 points. If the defendant was at fault but was only slightly careless, his fault will count as two points; whereas, if he was guilty of gross negligence or even of an intentional fault, this fait générateur will be worth three or four points, or maybe even five points. Likewise, if the causal connection between the fait générateur and the harm was very close and obvious, it will be ‘worth’ more points than if it was more remote and/or uncertain. As for harm, it will ‘weigh’ a number of points proportional to its nature and its intensity: serious personal injury may be worth six, seven or eight points, whereas loss of expected profits or harm to reputation will only be worth, say, one or two points. This means that if a claimant can establish that he suffered serious personal injury (eight points), it will be enough for him to establish a very small fault on the defendant’s part (one or two points), or even a ‘virtual fault’, and a more tenuous causal link (one point) to get compensated. However, if the harm he suffered is not so serious (say four points), the court will be more demanding as regards the reality and seriousness of fault and causation. Of course, this ‘points system’ is only a metaphor, and we are by no means suggesting that French courts explicitly follow this sort of reasoning when dealing with liability cases. However, the study of the French case law suggests that serious harm or gross fault do attract liability more readily than minor harm or slight negligence – even though, in theory, neither the seriousness of fault nor the nature or intensity of harm should have an influence on the existence of liability. As Carbonnier expressed it so vividly: Fault is no more the sum of physical, psychological and social elements than life is the sum of oxygen, hydrogen, carbon etc. There must be something more: a spontaneous and intuitive judgment by the court … Nowhere more than in relation to fault do the judges proceed by way of a judgment based on fairness [équité], condemning or pardoning in the name of society … The Cour de cassation’s assessment is a global one, acting as a court of review of excesses of fairness. Rather than ensuring that the law [la loi] is respected, its role here appears to be to prevent (in the interest, most often, why shouldn’t one say it? Of insurance companies and parties with deep pockets) an over-charitable fairness, which would let the evaluation of needs and resources come before the morality of the case.86

his fall the climber behind him, the Cour de cassation ruled that the defendant was liable under article 1382 Cc, given that ‘to bring about another climber’s fall constitutes a fault’ (‘le fait de provoquer la chute d’un autre grimpeur constitue une faute’). 86 J Carbonnier, Droit civil, vol II, Les biens, Les obligations (Paris, PUF, 2004) 2306 (emphasis added).

Principles of Liability or a Law of Torts?  471 Ironically, therefore, the very generality of approach of the law and of the concepts of which the law is constructed allows French courts to proceed in a distinctly casuistic, not to say, policy-driven way.

III.  The Increased Importance of the ‘Special Regimes’ of Liability and Compensation The heartland of the French law of civil liability is typically seen in the classic law of liability for fault, for the action of things and the action of others and these bodies of law often remain the focus of discussions of the law of responsabilité délictuelle or responsabilité civile in the treatises on the law of obligations. However, it is widely realised that this body of law has been much supplemented and at times excluded by special regimes of liability or of compensation. These special legislative regimes of liability have been created for a number of reasons and in a number of particular circumstances, but many have been concerned to take further the primary concern of French lawyers of making easier the recovery of damages for accidental personal injury and death. The seeds of these important developments can be seen early in the case of accidents at work. Towards the end of the nineteenth century some French lawyers became sympathetic to the difficulty which workers injured at work encountered in establishing the fault of their employers. Various suggestions were put forward, including treating the employer’s liability as based on a contractual obligation as to their employees’ safety and, even more famously, holding the employer as ‘keeper’ of the thing (typically machinery) which caused the employee’s injury.87 While both these suggestions flourished in the sense that the idea of ‘contractual obligations of safety’ and liability for the action of things became settled features of the law, in the specific case of accidents at work in 1898 the legislature intervened and imposed liability on employers to provide fixed rates or tariffs of compensation to their workers for death or personal injuries without proof of fault.88 Where an employee had such a claim, neither the employer’s fault nor the employee’s fault would have any effect on the amount of employee’s compensation except where the fault was ‘inexcusable’.89 However, where the employee had such a claim, he could not rely on any other legal grounds of claiming in respect of the accident.90 In this way, compensation for accidents at work is often seen as having ‘escaped’ the law

87 On the idea of using contract as a means of imposing strict liability, see ch 2 above p 16. On the use of ‘liability for the actions of things’ see Cass civ 16 June 1896 (arrêt Teffaine) S 1897, 1, 17 note P Esmein and ch 7 above p 129. 88 Loi of 9 April 1898 ‘sur les responsabilités des accidents dont les ouvriers sont victimes dans leur travail’, art 1. 89 ibid art 20. 90 ibid art 2(1).

472  Jean-Sébastien Borghetti and Simon Whittaker of civil liability (almost) altogether.91 Moreover, of the ‘classic triple requirement’ of harm, fault and a causal link between them, this regime waives the requirement of fault; it substitutes for the normal rule of causation a requirement of a ‘causal link, even if rather stretched’ between the employee’s work and the harm;92 and it disapplies the general principle of ‘full reparation’ in favour of a scale of fixed sums according to the seriousness of the effects of the accident, providing a system of compensation which is said to be forfaitaire. None of this law governing compensation for accidents at work is contained in the Code civil as it is seen as belonging instead to the law of social security and therefore appropriate for the Code de sécurité sociale. More recently, other special regimes of compensation (régimes d’indemnisation) as opposed to regimes of liability (régimes de responsabilité) have been created, notably, for the cases of compensation of victims of crime93 and of victims of contamination by blood products.94 Secondly, as recognised by the Projet de réforme, there are two ‘principal’ special regimes of liability: liability for road accidents and liability for defective ­products.95 The first was born out of a series of decisions of the Cour de cassation in relation to the application of liability for the actions of things in article 1384(1) Cc (now article 1242(1) Cc)96 and was put into effect in 1985 by special legislation outside the Code civil.97 As earlier noted, where this special regime applies, the victim of a road accident cannot rely on the general law.98 The origins of the product liability regime are rather different, being international rather than particular to France. In 1985 the EEC enacted a directive requiring Member States to recognise a regime of liability on ‘producers’ where a defect of safety in their product caused harm (personal injuries or ‘consumer property’).99 The implementation of this directive in France proved particularly difficult, in part owing to the (correct) sense that the European regime would be less protective of the victims of products than existing French approaches under the general law, and in part owing to the political and legal fall-out of the ‘affair of contaminated blood’.100 Eventually, it was

91 Malaurie, Aynès and Stoffel-Munck (n 39) 26. The relevant legislative scheme is now contained in art L 411-1 ff Code de sécurité sociale. 92 Terré, Simler and Lequette (n 53) 1222. 93 This compensation is provided by the Fonds d’indemnisation des victimes des actes de terrorisme et d’autres infractions as provided for by arts 706-3–706-15 Code de procédure pénale. 94 Arts L 3122-1–3122-6 Code de la santé publique. 95 There are other special regimes of liability, for example, as regards nuclear accidents: see art L 597-1ff Code de l’environnement. 96 Notably, the arrêt Desmares Cass civ (2) 21 July 1982, D 1982, 449 concl Charbonnier, note Larroumet. 97 Loi 85-677 of 5 July 1985, loi Badinter. 98 Above, p 464. 99 Council Directive 1985/374/EEC of 25 July 1985 concerning liability for defective products, art 9. The directive was preceded by a European Convention on products liability in regard to personal injury and death of 27 January 1977, European Treaty Series No 91 which was signed by only four states, Austria, Belgium, France and Luxembourg. 100 On this see (in English) S Whittaker, Liability for Products: English Law, French Law, and European Harmonization (Oxford, Oxford University Press, 2005) 450–61.

Principles of Liability or a Law of Torts?  473 implemented by legislation inserted (controversially) into the Code civil.101 Where the special product liability regime applies, in principle the victim of a product can claim only on the basis of a ‘ground of liability’ other than a defect of safety.102 In the case of both liability for road accidents and liability for defective p ­ roducts, the bases of liability and the defences available differ significantly from the approach under the general laws, whether for fault or for the action of things. The Projet de réforme follows the common academic approach by presenting liability for ‘the action of motor-vehicles’ (ie for traffic accidents) and liability for defective products as ‘special liability regimes’ after the main body of the law of civil liability in Chapter VI, even though the Projet recognises that the structure of these regimes is similar to the other liability regimes, and is therefore based on the triple requirement of ‘fait générateur-causation-harm’103 and although the practical relevance of these regimes is much greater than that of liability for the action of things or liability for ‘abnormal nuisance between neighbours’, both of which are included simply and prominently as faits générateurs of liability in Section 2 of Chapter II. Apart from the power of established practice, the reason why liability for traffic accidents and liability for defective products remain regulated separately probably lies in the fact that these regimes stand apart not only by reason of their faits générateurs, but also by reason of their grounds of exoneration and of the way in which they differentiate various kinds of harm – and the product liability regime also has specific prescription and limitation periods.104 However, these types of differences are not as unusual as the Projet de réforme’s drafters seem to believe and in particular the grounds of exoneration mentioned in Chapter III of the Projet de réforme are not in fact common to all types of liability. Finally, beyond these two principal special regimes of liability, the Projet de réforme includes some particular treatments of liability, but excludes others. The Projet includes a new provision on ‘abnormal nuisance between neighbours’ which has no equivalent in the existing Code civil, though it clearly seeks to give legislative recognition to a long-established body of case law.105 In doing so, it treats ‘abnormal nuisance’ as a distinct fait générateur de responsabilité. Secondly, in 2016 new provisions were inserted into the Code civil immediately after the rules governing product liability which concern ‘reparation of environmental loss’.106 While at very first glance one could imagine that these provisions impose a special liability in respect of environmental loss, in fact 101 Art 1386-1–1386-18 (now 1245-1 – 1245-17) Cc (first inserted in 1998). 102 Above, p 464, n 56. 103 In the case of motor-vehicle accidents this had, indeed, been disputed on the basis that the motor-vehicle merely had to be ‘involved’ or ‘implicated’ (impliqué) in the accident which led to the claimant’s harm: loi no 85-677 art 1 and see Terré, Simler and Lequette (n 53) 1249–51. 104 See art 1298 of the Projet de réforme reflecting art 1386-6 Cc (special grounds of exoneration); art 1290 of the Projet de réforme reflecting art 1386-2 Cc (special treatment of harm); arts 1299-1 and 1299-2 of the Projet de réforme reflecting arts 1386-16 and 1386-17 Cc (special prescription and delay periods). 105 See ch 11 above. 106 Arts 1246–1252 Cc referring to ‘un préjudice écologique’.

474  Jean-Sébastien Borghetti and Simon Whittaker they do not; rather, they regulate the nature of reparation for environmental loss where liability is established, that is, on independent grounds.107 This character of these new rules is reflected properly in their position in the Projet de réforme as they appear there in a sub-section of the new Section 2 of Chapter IV entitled ‘Special rules governing the reparation of losses resulting from certain types of harm’, here ‘­environmental harm’. On the other hand, there is no mention in the Projet de réforme of the special rules governing medical liability. Until 2002 the liability in private law108 of hospitals, clinics and doctors was based on non-performance of a contractual ‘safety obligation’ which was at first seen as requiring proof of lack of care (obligation de moyens),109 later though the courts recognised that liability could sometimes be based on a stricter contractual obligation, notably in the case of hospital-acquired infection.110 However, in 2002 legislation was enacted which provides that, whether working in the public or private sector, medical professionals and ‘any establishment, service or bodies in which individual acts of prevention, diagnosis or care take place are liable for the harmful consequences of acts of prevention, diagnosis or care only in the case of fault’.111 The legislation recognises two exceptions to this position: cases involving the application of the special product liability provisions and cases involving hospital-acquired infection in the absence of cause étrangère, that is, force majeure.112 Despite its practical significance, this law is not contained in the Code civil nor even referenced in the Projet de réforme.

IV.  Concluding Observations Overall, therefore, the generality and sheer breadth of treatment of the grounds of liability and also of their accompanying regime of rules in existing French law is not merely reflected, it is magnified in the new provisions proposed by the Projet de réforme. However, it soon becomes clear that this first impression of generality must be qualified in a number of ways, both in the case of existing law as developed by the courts on the basis of the classic provisions in the Code civil and as supplemented by ‘special’ legislation, and in the case of the Projet de réforme itself. 107 This is clear from the terms of art 1246 Cc, which states that ‘Any person liable for ecological loss [un préjudice écologique] is bound to make reparation for it’; art 1247 Cc adds that ‘Ecological loss is subject to reparation subject to the conditions in the present Title, where it consists of a non-negligible injury to the elements or functions of ecosystems or to the collective advantages derived by man from the environment’. 108 Different rules applied to the liability of public hospitals in the administrative courts, though in 1992 the Conseil d’Etat changed the general basis of liability from faute lourde (gross fault) to faute simple (ordinary fault): CE 10 April 1992, JCP 1992, II, 79027 note Moreau, Leb 1992 171, concl Legal. 109 This was first accepted by the Cour de cassation in Cass civ 20 May 1936, DP 1936, 1, 88 rapp Josserand, concl Matter. 110 Cass civ (1) 29 June 1999, rapp Sargos, JCP 1999, II, 10138. 111 Art L 1142-1-1(1) Code de la santé publique. 112 Art L 1142-1-1(2) Code de la santé publique.

Principles of Liability or a Law of Torts?  475 What we find in practice is a law teeming with distinctions: as to the subject-matter of the claim, as to the formal legal ground of the claim (contract or non-contract, fault, the ‘action of things’, product liability etc), as to the context in which it arises, and as to the kind of harm for which the claimant seeks compensation. Some of these distinctions are crisp and depend on quite finely drawn legal categories (for example, the separate treatment of injuries in road accidents, injuries by defective products and injuries in the course of medical treatment); but many of these distinctions are much more fluid, notably where they are drawn by the courts in the course of their application of the ‘chameleon concept’ of extra-contractual fault.113 The title of this chapter poses a question about the French law of civil liability, asking whether it consists of ‘principles of liability’ or a ‘law of torts’. Of course, a first response to this question could challenge its terms of reference by pointing out the inappropriateness of seeking to apply a way of thinking (‘a law of torts’) taken from the common law to the very different materials of French civil law. However, we can, we think, rescue our title’s legitimacy if we explain that the key feature of the common law of torts is that it consists of a series of particular liabilities, each with their own concepts, context and history and that we are therefore asking if any of this finds echoes in the French law. The existence of the ‘law of torts’ in English law does not deny the possibility that these ‘nominate torts’ are related (as, for example, liability in private nuisance and liability in the rule in Rylands v Fletcher), nor that they can overlap in particular circumstances. Nor does it deny the possibility of their sharing common concepts or being based on an overall theoretical position, for example, that they are all based on infringement of a claimant’s rights.114 However, what it reflects is the importance of a series of distinctions, in particular, according to the nature of the claimant’s interest (or right) which has been infringed, according to the nature of the defendant’s conduct and according to the practical context more generally. It also reflects an assumption that outside the scope of each of the particular torts (and also breach of contract or breach of trust), the law does not impose liability: each tort constitutes an island of liability set in a sea of freedom from liability.115

113 B Starck, ‘Des contrats conclus en violation des droits contractuels d’autrui’ JCP 1954, I, 1180 [38] referring to fault as a ‘veritable legal chameleon, changing according to its context’. 114 See, notably, R Stevens, Torts and Rights (Oxford, Oxford University Press, 2007). See also T Weir, Tort Law (Oxford, Oxford University Press, 2002) 12, arguing that, whatever the tort, there are three ‘focal points’: what the tortfeasor did, what the claimant suffered in consequence, and how the suffering resulted from the conduct. 115 cf Whittaker (n 52) 365. We avoid referring to a ‘sea of immunity’ as ‘[an] “immunity” is generally understood to be an exemption based on a defendant’s status from a liability imposed by the law on others, as in the case of sovereign immunity’: Michael v Chief Constable of South Wales Police [2015] UKSC 2, [2015] AC 1732 per Lord Toulson JSC. This approach was adopted by the Supreme Court in Robinson v Chief Constable of West Yorkshire Police n 11, [2018] UKSC 4 at [55] (Lord Reed JSC) and [99] (Lord Hughes JSC) who identified the question there as being whether the police owed a duty of care in the tort of negligence (the threshold condition for liability) in the circumstances rather than whether the police enjoyed an immunity from liability in those circumstances.

476  Jean-Sébastien Borghetti and Simon Whittaker At one level, this must also be true of French law: if none of the broad classic grounds of liability in the Code civil, nor any of the particular ‘special’ grounds of liability beyond the Code apply on the facts, then the defendant should not be held liable. However, this way of thinking is not, in our view, recognisably French, and for two main reasons. First, the point of the general principle of liability for fault in article 1240 Cc (formerly article 1382 Cc) is exactly that all ‘victims’ of any harm in all situations are covered by its protection, leaving to the application of the conditions of liability which that provision poses (‘fault’ ‘causing’ ‘harm’) to determine whether a particular defendant is liable to a particular claimant. There are, therefore, no situations in which a defendant is in principle free from liability (and even more free from responsabilité), even though there may be situations in which a defendant is not liable.116 Secondly, an assumption that the law allows gaps in the law of liability ignores the centrality of ‘victims’ of death and personal injuries in the thinking and in the development of the French law of civil liability and the general commitment of French lawyers to finding any and every way to ensure that these victims are properly compensated.117 This explains what to a foreign lawyer can look like the distortions of the notion of delictual fault (especially ‘virtual fault’) and the extraordinarily inventive approach to interpretation of the Code civil required for the invention of liability for the actions of things. As a result, the idea of areas of freedom from liability (where a person cannot in law be liable) would be seen as incompatible with the ‘principles of liability’ which are set out in the Code and also with the social and legal policy of well over a century. In our view, there is, therefore, a commitment by French lawyers both to the idea that their law of extra-contractual liability takes as its starting-point a series of ‘general principles’ of liability and also to a social policy of ensuring compensation for the victims of personal injuries. Indeed, it would not be going too far to say that it is a point of faith among many French legal scholars to uphold the existence of ‘principles of liability’ rather than a law of individual civil wrongs (torts or delicts) with intervening gaps. As we have observed, this does not mean that the treatment of cases by French scholars or courts is undifferentiated. Quite the reverse. Both legislation and even more case law contain a myriad of distinctions, some underneath the generality of the ‘principles’ of liability themselves and some by reason of the particular legislative treatments of the various ‘special regimes of liability’. But these are islands of particular treatment set in a sea of liability.

116 The defendant may not be liable broadly in two situations. First, where one of the legal conditions of liability are not satisfied (such as an absence of fault for the purposes of liability under art 1240 Cc) and, secondly, where exceptionally a defendant is immune from liability in the sense that, although the conditions of that defendant’s liability are actually met, a special immunity rule protects the defendant. In French law, such an immunity normally exists where the claimant can turn to another defendant for compensation: see O Deshayes (ed), Les immunités de responsabilité civile (Paris, PUF, 2009). 117 J-S Borghetti, ‘The Culture of Tort Law in France’ (2012) 3 JETL 158, 173.

Appendix Projet de réforme de la responsabilité civile Présenté le 13 mars 2017, par Jean-Jacques Urvoas, Garde des Sceaux, Ministre de la Justice suite à la Consultation Publique Menée d’avril à juillet 2016 Reform Bill on Civil Liability Presented on the 13 March 2017 by Jean-Jacques Urvoas, Garde des Sceaux, Minister of Justice after a Publication Consultation Conducted between April and July 2016 English Translation by Simon Whittaker in Consultation with Jean-Sébastien Borghetti

478  Appendix

Translators’ Preface This translation was commissioned by the Direction des affaires civiles et du sceau, Ministère de la Justice, République française. We have endeavoured to translate the French legal terminology in a way which is both accurate and faithful to French legal thought and yet comprehensible to readers of English, whether or not they are familiar with the concepts of the civil law tradition. We have added some notes of our own to the text, some relating to the translation itself, others seeking to explain briefly those concepts or rules which are most unfamiliar to a person not trained in the civil law. Simon Whittaker Jean-Sébastien Borghetti English translation © Simon Whittaker and Jean-Sébastien Borghetti 2017, revised 2019. All rights reserved.

Appendix  479 Titre Ier

Title I

DISPOSITIONS RELATIVES AU LIVRE III DU CODE CIVIL

PROVISIONS RELATING TO BOOK III OF THE CIVIL CODE

Art 1. – I. Le titre III du livre III du code civil est ainsi modifié:

Art 1. – I. Title III of Book III of the Civil Code is amended as follows:

1. Sont abrogés les articles 1231 à 1252.

1. Articles 1231 to 1252 are repealed.

2. L’article 1231 est ainsi rédigé: «Le créancier d’une obligation issue d’un contrat valablement formé peut, en cas d’inexécution, demander au débiteur réparation de son préjudice dans les conditions prévues au sous-titre II.»

2. Article 1231 is composed as follows: ‘The creditor of an obligation arising from a validly formed contract may, in case of its non-performance, claim from the debtor reparation for his1 loss2 under the conditions provided by sub-title II.’

3. Le sous-titre II du titre III du livre III est ainsi rédigé:

3. Sub-title II of Title III of Book III is composed as follows:

Sous-titre II

Sub-title II

LA RESPONSABILITE CIVILE

CIVIL LIABILITY

Art 1232. – Les dispositions des chapitres I à IV s’appliquent sous réserve des dispositions propres aux régimes spéciaux. Chapitre Ier DISPOSITIONS LIMINAIRES

Art 1232. – The provisions of Chapters 1 to IV are applicable subject to provisions particular to the special regimes of liability. Chapter I INTRODUCTORY PROVISIONS

Art 1233. – En cas d’inexécution d’une obligation contractuelle, ni le débiteur ni le créancier ne peuvent se soustraire à l’application des dispositions propres à la responsabilité contractuelle pour opter en faveur des règles spécifiques à la responsabilité extracontractuelle.

Art 1233. – In the case of non-performance of a contractual obligation, neither the debtor nor the creditor may escape the application of provisions special to contractual liability in order to opt in favour of rules specific to extra-contractual liability.

Art 1233-1. – Les préjudices résultant d’un dommage corporel sont réparés sur le fondement des règles de la responsabilité extracontractuelle, alors même qu’ils seraient causés à l’occasion de l’exécution du contrat.

Art 1233-1. – Losses resulting from personal injury are subject to reparation on the basis of the rules of extra-contractual liability, even if they are caused in the course of performance of a contract.

1 General note. Throughout the translation we follow the convention of English statutory drafting and use the masculine singular personal and possessive pronoun (which are to be read as referring equally to the feminine or neuter) rather than using ‘he/she’, ‘his/her’ etc., or some form of circumlocution. 2 There is an important distinction in the text between le dommage (translated as ‘harm’ or ‘the harm’) and le préjudice (translated as ‘loss’ or ‘the loss’) which the claimant (often in the text, the ‘victim’) suffers as a result of the infliction of this harm. Le dommage is translated as ‘harm’ rather than ‘damage’ (which is obviously closer linguistically) as le dommage can include a number of non-physical harms (such as the grief or upset included under le dommage moral), whereas ‘damage’ in English is typically (though not exclusively) understood as being physical. On this distinction especially see above, ch [9] and ch [10].

480  Appendix Toutefois, la victime peut invoquer les stipulations expresses du contrat qui lui sont plus favorables que l’application des règles de la responsabilité extracontractuelle.

However, the victim may invoke express stipulations of a contract which are more favourable to him than the application of the rules of extra-contractual liability.

Art 1234. – Lorsque l’inexécution du contrat cause un dommage à un tiers, celui-ci ne peut demander réparation de ses conséquences au débiteur que sur le fondement de la responsabilité extracontractuelle, à charge pour lui de rapporter la preuve de l’un des faits générateurs visés à la section II du chapitre II.

Art 1234. – Where non-performance of a contract causes harm to a third party, the latter can claim reparation of its consequences from the debtor only on the basis of extra-contractual liability, and subject to that third party’s establishing one of the actions giving rise to liability targeted by Section II of Chapter II.

Toutefois, le tiers ayant un intérêt légitime à la bonne exécution d’un contrat peut également invoquer, sur le fondement de la responsabilité contractuelle, un manquement contractuel dès lors que celui-ci lui a causé un dommage. Les conditions et limites de la responsabilité qui s’appliquent dans les relations entre les contractants lui sont opposables. Toute clause qui limite la responsabilité contractuelle d’un contractant à l’égard des tiers est réputée non écrite.

Nevertheless, a third party who has a legitimate interest in the proper performance of a contract can equally invoke, on the basis of contractual liability, a contractual failing3 where the latter has caused him harm. The conditions and limitations on this liability which are applicable in the relations between the contracting parties may be set up against him.4 Any contract term which limits the contractual liability of a party to the contract in relation to a third party is deemed not written.

CHAPITRE II

CHAPTER II

LES CONDITIONS DE LA RESPONSABILITÉ

CONDITIONS OF LIABILITY

SECTION 1

SECTION 1

Dispositions communes aux responsabilités contractuelle et extracontractuelle

Provisions common to contractual and extra-contractual liability

Sous-section 1

Sub-section 1

Le préjudice réparable

Reparable loss5

Art 1235. – Est réparable tout préjudice certain résultant d’un dommage et consistant en la lésion d’un intérêt licite, patrimonial ou extrapatrimonial.

Art 1235. – Any certain loss is reparable where it results from harm and consists of an injury to a lawful interest, whether patrimonial or extra-patrimonial.6

3 ‘Contractual failing’ translates the phrase un manquement contractuel (which appears otherwise only in art 1254). The more technical expression for contractual non-performance is l’inexécution d’une obligation contractuelle or, less commonly but as earlier in this same provision, l’inexécution du contrat. 4 ‘May be set up against’ translates opposable. 5 ‘Loss’ here translates [le] préjudice, on which see above, n2. 6 ‘Patrimonial’ refers to a person’s wealth or estate (le patrimoine).

Appendix  481 Art 1236. – Le préjudice futur est réparable lorsqu’il est la prolongation certaine et directe d’un état de choses actuel.

Art 1236. – Future loss is reparable where it is the certain and direct prolongation of an existing state of things.

Art 1237. – Les dépenses exposées par le demandeur pour prévenir la réalisation imminente d’un dommage ou pour éviter son aggravation, ainsi que pour en réduire les conséquences, constituent un préjudice réparable dès lors qu’elles ont été raisonnablement engagées.

Art 1237. – Expenses incurred by a claimant in order to prevent the imminent occurrence of harm or to avoid its getting worse, as well as in order to reduce its consequences, constitute a reparable loss as long as they were reasonably undertaken.

Art 1238. – Seule constitue une perte de chance réparable, la disparition actuelle et certaine d’une éventualité favorable.

Art 1238. – A loss of a chance is reparable only where it is the present and certain disappearance of a favourable eventuality.

Ce préjudice doit être mesuré à la chance perdue et ne peut être égal à l’avantage qu’aurait procuré cette chance si elle s’était réalisée.

This loss must be calculated by reference to the chance lost and cannot be equal to the advantage which this chance would have procured if it had been realised.

Sous-section 2

Sub-section 2

Le lien de causalité

Causal relationship

Art 1239. – La responsabilité suppose l’existence d’un lien de causalité entre le fait imputé au défendeur et le dommage.

Art 1239. – Liability supposes the existence of a causal relationship between the action attributed to the defendant and the harm.

Le lien de causalité s’établit par tout moyen.

A causal relationship may be established by any means of proof.

Art 1240. – Lorsqu’un dommage corporel est causé par une personne indéterminée parmi des personnes identifiées agissant de concert ou exerçant une activité similaire, chacune en répond pour le tout, sauf à démontrer qu’elle ne peut l’avoir causé.

Art 1240. – Where personal injury is caused by an undetermined person among two or more identified persons acting in concert or exercising a similar activity, each person is liable for the whole, unless he shows that he could not have caused it.

Les responsables contribuent alors entre eux The persons so liable make contribution to à proportion de la probabilité que chacun ait each other in proportion to the probability causé le dommage. that each had caused the harm. SECTION 2

SECTION 2

Dispositions propres à la responsabilité extracontractuelle

Provisions special to extra-contractual liability

Sous-section 1

Sub-section 1

Le fait générateur de responsabilité extracontractuelle

Action giving rise to extra-contractual liability

§1 La faute

§1 Fault

Art 1241. – On est responsable du dommage Art 1241. – A person is liable for the harm causé par sa faute. caused by his fault.

482  Appendix Art 1242. – Constitue une faute la violation d’une prescription légale ou le manquement au devoir général de prudence ou de diligence.

Art 1242. – A violation of a legislative requirement or a failure in the general duty of care or diligence constitutes a fault.

Art 1242-1. – [La faute de la personne morale résulte de celle de ses organes ou d’un défaut d’organisation ou de fonctionnement.]

Art 1242-1. – [Fault in a legal person results from fault in one of its organs or a failure in its organisation or its functioning.]

§2 Le fait des choses

§2 The action of things

Art 1243. – On est responsable de plein droit Art 1243. – A person is liable strictly7 des dommages causés par le fait des choses for harm caused by the action of corporeal corporelles que l’on a sous sa garde. things within his keeping. Le fait de la chose est présumé dès lors que celle-ci, en mouvement, est entrée en contact avec le siège du dommage.

An action of a thing is presumed wherever, while moving, it comes into contact with the person or property which is harmed.8

Dans les autres cas, il appartient à la victime de prouver le fait de la chose, en établissant soit le vice de celle-ci, soit l’anormalité de sa position, de son état ou de son comportement.

In other cases, it is for the victim to prove the action of the thing by establishing either its defect or the abnormality of its position, its state or its behaviour.

Le gardien est celui qui a l’usage, le contrôle et la direction de la chose au moment du fait dommageable. Le propriétaire est présumé gardien.

The keeper of a thing is the person who has the use, control and direction of the thing at the time of the action causing the harm. The owner of a thing is presumed to be its keeper.

§3 Les troubles anormaux de voisinage

§3 Abnormal nuisance between neighbours

Art 1244. – Le propriétaire, le locataire, le bénéficiaire d’un titre ayant pour objet principal de l’autoriser à occuper ou à exploiter un fonds, le maître d’ouvrage ou celui qui en exerce les pouvoirs, qui provoque un trouble excédant les inconvénients normaux de voisinage, répond de plein droit du dommage résultant de ce trouble.

Art 1244. – The owner, lessee, holder of a title whose principal object is a permission to occupy or exploit land or a building,9 or a person who commissions work on land or enjoys the latter’s authority, who causes a nuisance10 exceeding the normal inconveniences of being neighbours, is liable strictly for the harm resulting from the nuisance.

7 ‘Strictly’ translates de plein droit, more literally, ‘by operation of law’. The significance, however, is that the liability is imposed without reference to the fault of the person held liable. Cf the translation of de plein droit in art 1280(2) as ‘by operation of law’. 8 ‘Person or property which is harmed’ translates le siège du dommage, more literally, the ‘seat (or locus) of the harm’. 9 ‘Land or a building’ translates un fonds, which refers here to ‘a thing which is immovable by its nature’: cf art 518 Cc. 10 ‘Nuisance’ translates here troubles and refers here and in the context of art 1244 (which concerns les troubles anormaux de voisinage) to something similar to the common law tort of private nuisance.

Appendix  483 Lorsqu’une activité dommageable a été autorisée par voie administrative, le juge peut cependant accorder des dommages et intérêts ou ordonner les mesures raisonnables permettant de faire cesser le trouble.

Where a harmful activity has been authorised by an administrative means,11 the court may, however, award damages or order reasonable measures permitting the nuisance to be stopped.

Sous-section 2

Sub-section 2

L’imputation du dommage causé par autrui

The imputation of harm caused by another person

Art 1245. – On est responsable du dommage causé par autrui dans les cas et aux conditions posés par les articles 1246 à 1249.

Art 1245. – A person is liable for harm caused by another person in the cases and subject to the conditions laid down by articles 1246 to 1249.

Cette responsabilité suppose la preuve d’un fait de nature à engager la responsabilité de l’auteur direct du dommage.

This liability rests on proof of an action of a nature to engage liability in the direct author of the harm.

Art 1246. – Sont responsables de plein droit du fait du mineur:

Art 1246. – The following are liable strictly for the action of a minor:

–– ses parents, en tant qu’ils exercent –– his parents, to the extent to which they l’autorité parentale; exercise parental authority; –– son ou ses tuteurs, en tant qu’ils sont –– his guardian or guardians,12 to the chargés de la personne du mineur; extent to which they are charged with care of the minor’s person; –– la personne physique ou morale chargée –– par décision judiciaire ou administrative, d’organiser et contrôler à titre permanent le mode de vie du mineur. Dans cette hypothèse, la responsabilité des parents de ce mineur ne peut être engagée. Art 1247. – Est responsable de plein droit du fait du majeur placé sous sa surveillance la personne physique ou morale chargée, par décision judiciaire ou administrative, d’organiser et contrôler à titre permanent son mode de vie.

a physical or legal person charged by judicial or administrative decision with organising and controlling the minor’s way of life on a permanent basis. In these circumstances, the parents’ liability of such a minor cannot be engaged.

Art 1247. – A physical or legal person charged by judicial or administrative decision with organising and controlling an adult’s way of life on a permanent basis is liable strictly for the action of such an adult placed under their supervision.

However, the French term trouble itself can refer to harmful activities in a much broader sense and outside the context of relations between neighbours. This broader sense may be seen in the Projet in arts 1266 and 1279-6 in the context of the cessation of unlawful activity, which refer to le trouble illicite (‘unlawful nuisance’) on which see above, ch 18, esp p 381 ff. 11 This translates [la] voie administrative. This very broad expression is used so as to include the different types of authorisation by the administration which could be envisaged. 12 ‘Guardian’ translates here [le] tuteur. Guardianship is provided for by art 390 ff Cc.

484  Appendix Art 1248. – Les autres personnes qui, par contrat assument, à titre professionnel, une mission de surveillance d’autrui ou d’organisation et de contrôle de l’activité d’autrui, répondent du fait de la personne physique surveillée à moins qu’elles ne démontrent qu’elles n’ont pas commis de faute.

Art 1248. – Other persons who take on by contract, and by way of their business or profession,13 a task of supervision of another person or the organisation and control of the activity of another person, are liable for the action of the physical person supervised unless they show that they did not commit any fault.

Art 1249. – Le commettant est responsable de plein droit des dommages causés par son préposé. Est commettant celui qui a le pouvoir de donner au préposé des ordres ou des instructions en relation avec l’accomplissement de ses fonctions.

Art 1249. – An employer is liable strictly for harm caused by his employee. An employer is a person who has the power to give orders or instructions to his employee in relation to the performance of his functions.14

En cas de transfert du lien de préposition, cette responsabilité pèse sur le bénéficiaire du transfert.

In the case of transfer of the relationship of employment, this liability is borne by the beneficiary of the transfer.

Le commettant ou le bénéficiaire du transfert n’est pas responsable s’il prouve que le préposé a agi hors des fonctions auxquelles il était employé, sans autorisation et à des fins étrangères à ses attributions. Il ne l’est pas davantage s’il établit une collusion du préposé et de la victime.

An employer or a beneficiary of such a transfer is not liable if he proves that the employee acted outside the functions for which he was employed, without authorisation and for purposes alien to his attributions. Nor is he liable if he establishes collusion between his employee and the victim.

Le préposé n’engage sa responsabilité personnelle qu’en cas de faute intentionnelle, ou lorsque, sans autorisation, il a agi à des fins étrangères à ses attributions.

An employee is not subject to any personal liability except in the case of intentional fault, or where without authorisation he acted for purposes alien to his attributions.

SECTION 3

SECTION 3

Dispositions propres à la responsabilité contractuelle

Provisions special to contractual liability

Art 1250. – Toute inexécution du contrat Art 1250. – Every non-performance of a ayant causé un dommage au créancier oblige contract which has caused harm to the le débiteur à en répondre. creditor gives rise to an obligation in the debtor to be liable for it.

13 This translates à titre professionnel, professionnel in French being broad enough to include both business and profession. 14 As will be seen from the definition in this provision, le commettant and le préposé are understood more widely than the English terms ‘employer’ and ‘employee’ which are normally restricted to the persons party to a contract of employment.

Appendix  485 Art 1251. – Sauf faute lourde ou dolosive, le débiteur n’est tenu de réparer que les conséquences de l’inexécution raisonnablement prévisibles lors de la formation du contrat.

Art 1251. – Except for gross or dishonest fault, a debtor is bound to make reparation only for those consequences of non-performance which were reasonably foreseeable at the time of the formation of the contract.

Art 1252. – La réparation du préjudice résultant du retard dans l’exécution suppose la mise en demeure préalable du débiteur. La mise en demeure n’est requise pour la réparation de tout autre préjudice que lorsqu’elle est nécessaire pour caractériser l’inexécution.

Art 1252. – The reparation of loss resulting from delay in performance is premised on the prior giving of a notice to perform to the debtor. Notice to perform is not required for reparation of any other loss except where it is necessary in order to characterize the non-performance.

CHAPITRE III

CHAPTER III

LES CAUSES D’EXONÉRATION OU D’EXCLUSION DE LA RESPONSABILITÉ

GROUNDS OF EXONERATION OR OF EXCLUSION OF LIABILITY

SECTION 1

SECTION 1

Les causes d’exonération de responsabilité

Grounds of exoneration of liability

Art 1253. – Le cas fortuit, le fait du tiers ou de la victime sont totalement exonératoires s’ils revêtent les caractères de la force majeure.

Art 1253. – A fortuitous event, or an act of a third party or of the victim, provide a total exoneration where they bear the characteristics of force majeure.

En matière extracontractuelle, la force majeure est l’évènement échappant au contrôle du défendeur ou de la personne dont il doit répondre, et dont ceux-ci ne pouvaient éviter ni la réalisation ni les conséquences par des mesures appropriées.

In extra-contractual matters, force majeure is an event escaping the control of the defendant or of a person for whom he is responsible, whose occurrence and whose consequences the latter could not avoid by appropriate measures.

En matière contractuelle, la force majeure est définie à l’article 1218.

In contractual matters, force majeure is defined by article 1218.

Art 1254. – Le manquement de la victime à ses obligations contractuelles, sa faute ou celle d’une personne dont elle doit répondre sont partiellement exonératoires lorsqu’ils ont contribué à la réalisation du dommage.

Art 1254. – A failure by the victim in his contractual obligations, his own fault or that of a person for whom he is responsible, provide a partial exoneration where they contributed to the occurrence of the harm.

En cas de dommage corporel, seule une faute lourde peut entrainer l’exonération partielle.

In the case of personal injury, only gross fault can lead to partial exoneration.

Art 1255. – Sauf si elle revêt les caractères de Art 1255. – Unless it bears the la force majeure, la faute de la victime privée characteristics of force majeure, fault in de discernement n’a pas d’effet exonératoire. a victim who lacks discernment15 has no exonerating effect.



15 Such a ‘lack of discernment’ would be established in cases of relevant mental ill health or disability.

486  Appendix Art 1256. – La faute ou l’inexécution contractuelle opposable à la victime directe l’est également aux victimes d’un préjudice par ricochet.

Art 1256. – Fault or contractual non-performance which may be set up against16 the direct victim may also be set up again any indirect victims of a loss.

SECTION 2

SECTION 2

Les causes d’exclusion de responsabilité

Grounds of exclusion of liability

Art 1257. – Le fait dommageable ne donne pas lieu à responsabilité lorsque l’auteur se trouve dans l’une des situations prévues aux articles 122-4 à 122-7 du code pénal.

Art 1257. – An action causing harm does not give rise to liability where its author finds himself in one of the situations foreseen by articles 122-4 to 122-7 of the Criminal Code.

Art 1257-1. – Ne donne pas non plus lieu à responsabilité le fait dommageable portant atteinte à un droit ou à un intérêt dont la victime pouvait disposer, si celle-ci y a consenti.

Art 1257-1. – Equally, there is no room for liability where an action causing harm prejudices a right or an interest over which the victim has a power of disposal if the latter has consented to it.

CHAPITRE IV

CHAPTER IV

LES EFFETS DE LA RESPONSABILITÉ

THE EFFECTS OF LIABILITY

SECTION 1

SECTION 1

Principes

Principles

Art 1258. – La réparation a pour objet de replacer la victime autant qu’il est possible dans la situation où elle se serait trouvée si le fait dommageable n’avait pas eu lieu. Il ne doit en résulter pour elle ni perte ni profit.

Art 1258. – The aim of reparation is to replace the victim as much as is possible in the situation in which he would have been if the harmful action had not taken place. It must cause him neither a loss nor an advantage.

Art 1259. – La réparation peut prendre la forme d’une réparation en nature ou de dommages et intérêts, ces deux types de mesures pouvant se cumuler afin d’assurer la réparation intégrale du préjudice.

Art 1259. – Reparation may take the form of reparation in kind17 or damages, these two types of measures being able to be combined so as to ensure full reparation of the loss.

Sous-section 1

Sub-section 1

La réparation en nature

Reparation in kind

Art 1260. – La réparation en nature doit être spécifiquement propre à supprimer, réduire ou compenser le dommage.

Art 1260. – Reparation in kind must be specifically appropriate to suppress, reduce or make up for the harm.

16 This translates opposable. 17 ‘Reparation in kind’ translates une réparation en nature. There is, therefore, a close link with l’exécution forcée en nature in the context of contractual non-performance, on which see arts 1221 and 1222 Cc. This relationship comes out very strongly in the provisions governing la réparation en nature in arts 1261(2) and (3) of the Projet.

Appendix  487 Art 1261. – La réparation en nature ne peut être imposée à la victime.

Art 1261. – Reparation in kind cannot be imposed on the victim.

Elle ne peut non plus être ordonnée en cas d’impossibilité ou de disproportion manifeste entre son coût pour le responsable et son intérêt pour la victime.

Nor can it be ordered in the case of impossibility or of manifest disproportionality between its cost for the person liable and its interest for the victim.

Sous les mêmes réserves, le juge peut également autoriser la victime à prendre elle-même les mesures de réparation en nature aux frais du responsable. Celui-ci peut être condamné à faire l’avance des sommes nécessaires.

Subsequent to the same qualifications, equally a court may authorise the victim himself to take measures of reparation in kind at the expense of the person liable. The latter may be ordered to provide an advance of the money necessary for this purpose.

Sous-section 2

Sub-section 2

Les dommages et intérêts

Damages

Art 1262. – Les dommages et intérêts sont évalués au jour du jugement, en tenant compte de toutes les circonstances qui ont pu affecter la consistance et la valeur du préjudice depuis le jour de la manifestation du dommage, ainsi que de son évolution raisonnablement prévisible.

Art 1262. – Damages are assessed as of the day of judgment, taking into account all the circumstances which could have affected the make-up and the value of the loss since the day of the manifestation of the harm, as well as its reasonably foreseeable development.

En cas d’aggravation du dommage postérieurement au jugement, la victime peut demander un complément d’indemnité pour le préjudice qui en résulte.

In the case of the worsening of the harm subsequent to judgment, the victim may claim supplementary compensation for the loss which results from it.

En cas de dommage corporel, elle peut également réclamer une indemnisation complémentaire pour tout chef de préjudice préexistant non inclus dans la demande initiale.

In the case of personal injury, the victim can also claim a supplementary compensation for any pre-existing head of loss18 not included in the initial claim.

Chacun des chefs de préjudice est évalué distinctement.

Each head of loss is evaluated separately.

18 ‘Head of loss’ translates [le] chef de préjudice. While these are mentioned again by the Projet (later in art 1262 and again in art 1276), they are not explicitly explained or enumerated; instead, the projet sets out a series of special rules governing damages for personal injury in arts 1267 to 1277. For this purpose, art 1269 refers to particular ‘items of loss’ (postes de préjudices) to be fixed by decree made after consulting the Conseil d’Etat.

488  Appendix Art 1263. – Sauf en cas de dommage corporel, les dommages et intérêts sont réduits lorsque la victime n’a pas pris les mesures sûres et raisonnables, notamment au regard de ses facultés contributives, propres à éviter l’aggravation de son préjudice.

Art 1263. – Except in the case of personal injuries, damages are reduced where the victim did not take safe and reasonable measures, notably having regard to his ability to pay, appropriate to avoid the worsening of his own loss.

Art 1264. – La victime est libre de disposer des sommes allouées.

Art 1264. – The victim is free in his use of any sums awarded.

Sous-section 3 L’incidence de la pluralité de responsables

Sub-section 3 The effect of plurality of persons liable

Art 1265. – Lorsque plusieurs personnes sont responsables d’un même dommage, elles sont solidairement tenues à réparation envers la victime.

Art 1265. – Where one or more persons are liable for the same harm, they are jointly and severally bound to make reparation for it to the victim.

Si toutes ou certaines d’entre elles ont commis une faute, elles contribuent entre elles à proportion de la gravité et du rôle causal du fait générateur qui leur est imputable. Si aucune d’elles n’a commis de faute, elles contribuent à proportion du rôle causal du fait générateur qui leur est imputable, ou à défaut par parts égales.

If all or certain of them have committed a fault, they make contribution to each other in proportion to the seriousness and the causal role of the action giving rise to liability which is attributable to them. If none of them has committed a fault, they make contribution in proportion to the causal role of the action giving rise to liability which is attributable to them, or, if this is not possible, in equal parts.

Sous-section 4

Sub-section 4

La cessation de l’illicite

Cessation of unlawful action

Art 1266. – En matière extracontractuelle, indépendamment de la réparation du préjudice éventuellement subi, le juge peut prescrire les mesures raisonnables propres à prévenir le dommage ou faire cesser le trouble illicite auquel est exposé le demandeur. Sous-section 5

Art 1266. – In extra-contractual matters, independently of any reparation of loss which may have been suffered, a court may prescribe reasonable measures appropriate to prevent harm or to see that an unlawful nuisance to which a claimant is exposed is stopped. Sub-section 5

L’amende civile

Civil penalty

Art 1266-1. – En matière extracontractuelle, lorsque l’auteur du dommage a délibérément commis une faute en vue d’obtenir un gain ou une économie, le juge peut le condamner, à la demande de la victime ou du ministère public et par une décision spécialement motivée, au paiement d’une amende civile.

Art 1266-1. – In extra-contractual matters, where the author of the harm has deliberately committed a fault with the view to making a gain or to saving money, a court may, at the request of the victim or the ministère public19 and by specially justified decision, condemn him to the payment of a civil penalty.

19 The ministère public (sometimes called the parquet) is a particular category of magistrat (broadly, a member of the judiciary) whose role in civil matters is to join proceedings (and sometimes

Appendix  489 Cette amende est proportionnée à la gravité de la faute commise, aux facultés contributives de l’auteur et aux profits qu’il en aura retirés.

Such a penalty is proportionate to the seriousness of the fault committed, to the ability to pay of the author of the harm, and to any profits which he may have made from it.

L’amende ne peut être supérieure au décuple du montant du profit réalité́.

The penalty cannot be higher than 10 times the amount of any profit made.

Si le responsable est une personne morale, l’amende peut être portée à 5 % du montant du chiffre d’affaires hors taxes le plus élevé réalisé en France au cours d’un des exercices clos depuis l’exercice précédant celui au cours duquel la faute a été commise.

If the person liable is a legal person, the penalty can be as high as 5% of the highest amount of its revenue excluding value-added tax realised in France in the course of one of the fiscal years ending after the fiscal year before the one in the course of which the fault was committed.

Cette amende est affectée au financement d’un fonds d’indemnisation en lien avec la nature du dommage subi ou, à défaut, au Trésor public.

Such a penalty is allocated to the financing of a compensation fund related to the nature of the harm suffered or, if this is not possible, to the public Treasury.

Elle n’est pas assurable.

It is not insurable.

SECTION 2

SECTION 2

Règles particulières à la réparation des préjudices résultant de certaines catégories de dommages

Special rules governing the reparation of losses resulting from certain types of harm

Sous-section 1

Sub-section 1

Règles particulières à la réparation des préjudices résultant d’un dommage corporel

Special rules governing the reparation of losses resulting from personal injury

Art 1267. – Les règles de la présente soussection sont applicables aux décisions des juridictions judiciaires et administratives, ainsi qu’aux transactions conclues entre la victime et le débiteur de l’indemnisation.

Art 1267. – The rules in the present sub-section are applicable to decisions of private law courts and of administrative courts,20 as well as to settlements concluded between the victim and the person who owes the compensation.

initiate them) and submit oral or written arguments to the ‘sitting’ judges (the magistrats du siège) as a matter of the public interest. The ministère public is answerable to the Minister of Justice. 20 ‘Private law courts and administrative courts’ translates [les] juridictions judiciaires et administratives. In the French legal system, the ‘judicial jurisdiction’ covers all courts and tribunals which do not form part of the administrative law system of adjudication. As a result, ‘private law courts’ includes courts dealing with a very wide range of cases, including commercial and employment matters.

490  Appendix Art 1267-1. – Toute stipulation contraire aux dispositions de la présente sous-section est réputée non écrite à moins qu’elle ne soit plus favorable à la victime.

Art 1267-1. – Any stipulation contrary to the provisions of the present sub-section are deemed not written unless they are more favourable to the victim.

Art 1268. – Les préjudices doivent être appréciées sans qu’il soit tenu compte d’éventuelles prédispositions de la victime lorsque l’affection qui en est issue n’a été provoquée ou révélée que par le fait dommageable.

Art 1268. – Losses must be assessed without taking into account any possible predispositions of the victim where the ailment which arose from them was provoked or revealed only by the action which caused the harm.

Art 1269. – Les préjudices patrimoniaux et extrapatrimoniaux résultant d’un dommage corporel sont déterminés, poste par poste, suivant une nomenclature non limitative des postes de préjudices fixée par décret en Conseil d’État.

Art 1269. – Patrimonial and extrapatrimonial21 losses resulting from personal injury shall be determined item by item following a non-exclusive terminology of items of loss fixed by decree made after consulting the Conseil d’Etat.

Art 1270. – Sauf disposition particulière, le déficit fonctionnel après consolidation est mesuré selon un barème médical unique, indicatif, dont les modalités d’élaboration, de révision et de publication sont déterminées par voie règlementaire.

Art 1270. – Subject to any special provision, a functional deficiency which has stabilised shall be measured according to a single medical scale whose manner of elaboration, revision and publication shall be determined by administrative regulation.

Art 1271. – Un décret en Conseil d’État fixe les postes de préjudices extrapatrimoniaux qui peuvent être évalués selon un référentiel indicatif d’indemnisation, dont il détermine les modalités d’élaboration et de publication. Ce référentiel est réévalué tous les trois ans en fonction de l’évolution de la moyenne des indemnités accordées par les juridictions.

Art 1271. – A decree made after consulting the Conseil d’Etat shall set the items of extrapatrimonial losses which can be assessed according to an indicative compensation index, whose manner of elaboration and publication it shall decide. This index shall be revalued every three years as a function of developments in the average of compensation awards made by the courts.

A cette fin, une base de données rassemble, sous le contrôle de l’État et dans des conditions définies par décret en Conseil d’État, les décisions définitives rendues par les cours d’appel en matière d’indemnisation du dommage corporel des victimes d’un accident de la circulation.

For this purpose, a database shall bring together, under the control of the State and under conditions defined by a decree made after consulting the Conseil d’Etat, final decisions rendered by courts of appeal in the matter of compensation of personal injury suffered by victims of traffic accidents.



21 See

above n 6.

Appendix  491 Art 1272. – L’indemnisation due au titre de la perte de gains professionnels, de la perte de revenus des proches ou de l’assistance d’une tierce personne a lieu en principe sous forme d’une rente. Celle-ci est indexée sur un indice fixé par voie règlementaire et lié à l’évolution du salaire minimum.

Art 1272. – In principle compensation which is due under the heading of loss of business gains, loss of revenue to persons near to the primary victim22 or of assistance by a third person shall take place in the form of a periodic payment. The latter shall be indexed on the basis of an index fixed by way of administrative regulation and shall be tied to changes in the minimum wage.

Avec l’accord des parties, ou sur décision spécialement motivée, la rente peut être convertie en capital selon une table déterminée par voie règlementaire fondée sur un taux d’intérêt prenant en compte l’inflation prévisible et actualisée tous les trois ans suivant les dernières évaluations statistiques de l’espérance de vie publiées par l’Institut national des statistiques et des études économiques.

With the agreement of the parties or on the basis of a specially justified judicial decision, the periodic payment may be converted into a capital sum according to a table determined by way of administrative regulation based on a rate of interest taking account of foreseeable inflation and updated every three years following the latest statistical evaluations of life expectancy published by the Institut national des statistiques et des études économiques.

Lorsqu’une rente a été allouée conventionnellement ou judiciairement en réparation de préjudices futurs, le crédirentier peut, si sa situation personnelle le justifie, demander que les arrérages à échoir soient remplacés en tout ou partie par un capital, suivant la table de conversion visée à l’alinéa précédent.

Where a periodic payment has been awarded by agreement or by judicial decision as reparation for future losses, the beneficiary of the periodic payment may, if his personal situation so justifies, claim that the payments falling due in the future are replaced in full or in part by a capital sum, following the conversion table specified by the preceding paragraph.

Art 1273. – Les sommes versées à la victime à des fins indemnitaires par les tiers payeurs ne donnent lieu à recours subrogatoire contre le responsable ou son assureur que dans les cas prévus par la loi.

Art 1273. – Sums of money paid to a victim for the purpose of compensation by third party payers23 give rise to a right to subrogation against the person liable or his insurer only in the situations provided for by legislation.

22 ‘Loss of revenue to persons near to the primary victim’ translates in an explanatory way the more allusive French la perte de revenus des proches. In this respect, while proches typically includes any relatives of the primary victim, it may extend further. 23 A ‘third party payer’ may include the victim’s own first party insurer or the State by way of social security benefits.

492  Appendix Art 1274. – Seules les prestations énumérées ci-après versées à la victime d’un dommage corporel ouvrent droit à un recours contre la personne tenue à réparation ou son assureur:

Art 1274. – Only those benefits which are hereafter listed and which are paid to the victim of personal injury give rise to a right of recourse against the person bound to make reparation or his insurer.

1. Les prestations versées par les organismes, 1. Benefits supplied by organisations, établissements et services gérant un régime bodies or services operating a compulsory obligatoire de sécurité sociale; regime of social security; 2. Les prestations énumérées au II de l’article 1er de l’ordonnance n° 59-76 du 7 janvier 1959 relative aux actions en réparation civile de l’État et de certaines autres personnes publiques;

2. Benefits listed in paragraph II of article 1 of the Ordonnance no 59-76 of 7 January 1959 concerning actions for civil reparation brought by the State and by certain other public bodies;

3. Les sommes versées en remboursement des frais de traitement médical et de rééducation;

3. Sums paid by way of the reimbursement of the cost of medical treatment and rehabilitation;

4. Les salaires et les accessoires du salaire maintenus par l’employeur pendant la période d’inactivité consécutive à l’évènement qui a occasionné le dommage;

4. Salaries and their incidental costs which an employer continues to pay during a period of inactivity following the event which caused the harm;

5. Les indemnités journalières de maladie et les prestations d’invalidité versées par les groupements mutualistes régis par le code de la mutualité, les institutions de prévoyance régies par le code de la sécurité sociale ou le code rural et de la pêche maritime et les sociétés d’assurance régies par le code des assurances;

5. Daily sickness payments and invalidity benefits paid by friendly societies governed by the Code of Friendly Societies, provident institutions governed by the Code of Social Security or the Rural Code and insurance companies governed by the Insurance Code;

6. Les prestations prévues à l’article L.245-1 du code de l’action sociale et des familles.

6. Benefits as provided for by article L.245-1 of the Code of Social Action and Families.

Art 1275. – Les employeurs sont admis à poursuivre directement contre le responsable des dommages ou son assureur le remboursement des charges patronales afférentes aux rémunérations maintenues ou versées à la victime pendant la période d’indisponibilité de celle-ci. Ces dispositions sont applicables à l’État par dérogation aux dispositions de l’article 2 de l’ordonnance n° 59-76 du 7 janvier 1959 précitée.

Art 1275. – Employers are permitted to pursue directly the person liable for harm or his insurer for the reimbursement of employer contributions relating to remuneration maintained or paid to the victim during the latter’s period of unavailability. These provisions are applicable to the State by derogation from the provisions of article 2 of the Ordonnance no 59-76 of 7 January 1959 cited above.

Appendix  493 Art 1276. – Les prestations donnant lieu à recours s’imputent poste par poste sur les seules indemnités dues par le responsable pour les chefs de préjudice pris en charge par le tiers payeur, à l’exclusion des préjudices extrapatrimoniaux.

Art 1276. – Benefits giving rise to recourse are to be imputed item by item solely to the compensation due by the person liable for the heads of loss24 for which a third-party payer25 has taken responsibility, to the exclusion of extra-patrimonial losses.

Dans le cas où l’insolvabilité du responsable empêcherait l’indemnisation intégrale de la victime, celle-ci sera préférée au tiers payeur pour ce qui lui reste dû par le responsable.

In the case where the insolvency of the person liable would prevent the full compensation of the victim, the latter will be given priority over the third-party payer as regards what remains due to him by the person liable.

La faute de la victime ne peut réduire son droit à indemnisation que sur la part de son préjudice qui n’a pas été réparée par les prestations du tiers payeur. Celui-ci a droit au reliquat de la dette mise à la charge du responsable.

A victim’s fault can reduce his right to compensation only as regards that part of his loss which has not been subject to reparation by benefits provided by a thirdparty payer. The latter has a right to the remainder of the debt owed by the person liable.

Art 1277. – Hormis les prestations mentionnées aux articles 1274 et 1275, aucun versement effectué au profit d’une victime en vertu d’une obligation légale, conventionnelle ou statutaire n’ouvre droit à une action contre la personne tenue à réparation du dommage ou son assureur.

Art 1277. – Apart from the benefits mentioned in articles 1274 and 1275, no payment made to the benefit of a victim by virtue of a legislative, contractual or statutory obligation26 gives rise to an action against the person bound to make reparation for the harm or his insurer.

Toutefois lorsqu’il est prévu par contrat, le recours subrogatoire de l’assureur qui a versé à la victime une avance sur indemnité du fait de l’accident peut être exercé contre l’assureur de la personne tenue à réparation dans la limite du solde subsistant après paiements aux tiers visés à l’article 1274. Il doit être exercé, s’il y a lieu, dans les délais impartis par la loi aux tiers payeurs pour produire leurs créances.

Nevertheless, where it is provided by contract, recourse by way of subrogation by an insurer who has paid to the victim an advance of compensation as a result of an accident can be brought against the insurer of the person bound to make reparation up to the limit of the remaining balance after payments to third parties as set out by article 1274. Where this applies, it must be brought within the periods allowed by legislation to third-party payers to put forward their claims.

24 See above, n 18. 25 See above, n 23. 26 ‘Statutory obligation’ translates obligation statutaire. As will be apparent from the text, such an obligation is to be distinguished from one arising directly from legislation or a contract; it could include, for example, an obligation arising from a collective agreement.

494  Appendix Sous-section 2

Sub-section 2

Règles particulières à la réparation des préjudices résultant d’un dommage matériel

Special rules governing the reparation of losses resulting from physical damage to property

Art 1278. – En cas d’atteinte à un bien corporel, l’indemnité est de la plus faible des deux sommes représentant le coût de la remise en état et celui du remplacement du bien, sans qu’il soit tenu compte de sa vétusté ni de la plus-value éventuellement inhérente à la réparation.

Art 1278. – In the case of damage to physical property, compensation is the lower of the two sums representing the cost of repair and the cost of replacement of the property, without any account being taken of its age nor of any possible increase in value resulting from the repair.

Lorsque le bien ne peut être ni remis en état, ni remplacé, l’indemnité est de la valeur qu’aurait eue le bien au jour de la décision, dans son état antérieur au dommage.

Where the property cannot be repaired nor replaced, compensation is the value which the property would have had on the day of judgment in its state before the harm took place.

Si, à la demande de la victime, le bien endommagé n’est pas remis au responsable dans son état actuel, sa valeur résiduelle est déduite de l’indemnité.

If, at the request of the victim, the property which was damaged is not handed over in its present state to the person liable for the damage, its remaining value is deducted from the compensation.

Art 1279. – Le cas échéant, l’indemnité compense également la privation de jouissance du bien endommagé, les pertes d’exploitation ou tout autre préjudice.

Art 1279. – As the case may be, compensation shall also make up for any deprivation of enjoyment of the property which was damaged, operating losses or any other loss.

Sous-section 3

Sub-section 3

Règles particulières à la réparation des préjudices résultant d’un dommage environnemental

Special rules governing the reparation of losses resulting from environmental harm

Art 1279-1. – Toute personne responsable d’un préjudice écologique est tenue de le réparer.

Art 1279-1. – Every person liable for an ecological loss is bound to make reparation for it.

Art 1279-2. – Est réparable, dans les conditions prévues par la présente soussection, le préjudice écologique consistant en une atteinte non négligeable aux éléments ou aux fonctions des écosystèmes ou aux bénéfices collectifs tirés par l’homme de l’environnement.

Art 1279-2. – Subject to the conditions provided by the present sub-section, an ecological loss is subject to reparation where it consists in a non-negligible damage to the constituent parts or the functions of ecosystems or to the collective benefits drawn by man from his environment.

Appendix  495 Art 1279-3. – L’action en réparation du préjudice écologique est ouverte à toute personne ayant qualité et intérêt à agir, telle que l’État, l’Agence française pour la biodiversité, les collectivités territoriales et leurs groupements dont le territoire est concerné, ainsi que les établissements publics et les associations agréées ou créées depuis au moins cinq ans à la date d’introduction de l’instance qui ont pour objet la protection de la nature et la défense de l’environnement.

Art 1279-3. – An action for reparation for ecological loss is open to any person having the standing and interest to sue, such as the State, the Agence française pour la biodiversité, local authorities and their groups whose area is affected, as well as public bodies and associations accredited or created at least five years before the date of introduction of the proceedings which have as their purpose the protection of nature and the defence of the environment.

Art 1279-4. – La réparation du préjudice écologique s’effectue par priorité en nature.

Art 1279-4. – As a matter of priority, reparation for ecological loss is to be effected in kind.

En cas d’impossibilité ou d’insuffisance des mesures de réparation, le juge condamne le responsable à verser des dommages et intérêts, affectés à la réparation de l’environnement, au demandeur ou, si celui-ci ne peut prendre les mesures utiles à cette fin, à l’État.

Where such reparation measures are impossible or insufficient, a court shall order the person liable to pay damages, which are set aside for the reparation of the environment, to the claimant or, if the latter cannot take measures effective for this purpose, to the State.

L’évaluation du préjudice tient compte, le cas échéant, des mesures de réparation déjà intervenues, en particulier dans le cadre de la mise en œuvre du titre VI du livre Ier du code de l’environnement.

Where appropriate, assessment of the loss shall take into account any reparation measures which have already taken place, in particular within the framework of the operation of Title VI of Book I of the Environmental Code.

Art 1279-5. – En cas d’astreinte, celle-ci est liquidée par le juge au profit du demandeur, qui l’affecte à la réparation de l’environnement ou, si le demandeur ne peut prendre les mesures utiles à cette fin, au profit de l’État, qui l’affecte à cette même fin.

Art 1279-5. – As regards any money sum ordered by the court to be paid while its order is not obeyed,27 this is fixed by the court for the benefit of the claimant, who must set it aside for the reparation of the environment or, if the claimant cannot take any effective measures for this purpose, for the benefit of the State, who must set it aside for the same purpose.

Le juge se réserve le pouvoir de la liquider.

The court retains a power to fix the sum later.

27 ‘[A]ny money sum ordered by the court to be paid while its order is not obeyed’ translates in an explanatory way l’astreinte. The relevant provisions are contained in arts L131-1 to L131-4 of the Code of Civil Enforcement Procedures. Astreintes are ordered to be paid by the day, week or month and are either ‘provisional’ or ‘final’. An astreinte is ‘provisional’ where the sum to be paid is finally fixed by the court after the person ordered by the court has failed to obey and its amount is then fixed taking into account that person’s behaviour and any difficulties he may have encountered: this is the default position. An astreinte is ‘final’ where it is first ordered and may not later be amended.

496  Appendix Art 1279-6. – Les dispositions de l’article 1266 sont applicables au trouble illicite auquel est exposé l’environnement.

Art 1279-6. – The provisions of article 1266 are applicable to any unlawful nuisance28 to which the environment is exposed.

Sous-section 4

Sub-section 4

Règles particulières à la réparation des préjudices résultant du retard dans le paiement d’une somme d’argent

Special rules governing the reparation of losses resulting from delay in payment of a sum of money

Art 1280. – Le préjudice résultant du retard dans le paiement d’une somme d’argent est réparé par les intérêts au taux légal.

Art 1280. – Loss resulting from a delay in payment of a sum of money is subject to reparation by payment of interest at the legal rate.

Ces intérêts sont dus sans que le créancier soit tenu de justifier d’aucune perte. Ils ne sont dus que du jour de la mise en demeure, excepté dans le cas où la loi les fait courir de plein droit.

This interest is due without the creditor being bound to justify any loss. It is due only from the day of notice to perform, except in the case where legislation makes them run by operation of law.

Le créancier auquel son débiteur en retard a causé un préjudice supplémentaire, peut obtenir des dommages et intérêts distincts des intérêts moratoires de la créance.

A creditor who has been caused a supplementary loss by his debtor’s delay may obtain damages distinct from the interest generated by the delay in performance.

CHAPITRE V

CHAPTER V

LES CLAUSES PORTANT SUR LA RESPONSABILITE

CONTRACT TERMS CONCERNING LIABILITY

SECTION 1

SECTION 1

Les clauses excluant ou limitant la responsabilité

Contract terms excluding or limiting liability

Art 1281. – Les clauses ayant pour objet ou pour effet d’exclure ou de limiter la responsabilité sont en principe valables, aussi bien en matière contractuelle qu’extracontractuelle.

Art 1281. – Contract terms whose object or effect is to exclude or to limit liability are in principle valid, in contractual as well as in extra-contractual matters.

Toutefois, la responsabilité ne peut être limitée ou exclue par contrat en cas de dommage corporel.

However, in the case of personal injury, liability cannot be limited or excluded by contract.

Art 1282. – En matière contractuelle, les clauses limitatives ou exclusives de responsabilité n’ont point d’effet en cas de faute lourde ou dolosive. Elles sont réputées non écrites lorsqu’elles privent de sa substance l’obligation essentielle du débiteur.

Art 1282. – In contractual matters, contract terms limiting or excluding liability have no effect in the case of gross or dishonest fault. They are deemed not written where they deprive the debtor’s essential obligation of its substance.



28 See

above, n 10.

Appendix  497 Art 1283. – En matière extracontractuelle, on ne peut exclure ou limiter la responsabilité pour faute.

Art 1283. – In extra-contractual matters, a person cannot exclude or limit his liability for fault.

SECTION 2

SECTION 2

Les clauses pénales

Penalty clauses

Art 1284. – Lorsque le contrat stipule que celui qui manquera de l’exécuter paiera une certaine pénalité à titre de réparation, il ne peut être alloué à l’autre partie une pénalité plus forte ni moindre.

Art 1284. – Where a contract stipulates that a party who fails to perform shall pay a certain penalty by way of reparation, the other party cannot be awarded a penalty which is either higher or lower.

Néanmoins, le juge peut, même d’office, modérer ou augmenter la pénalité ainsi convenue si elle est manifestement excessive ou dérisoire.

However, a court may, even of its own initiative, lower or increase the penalty so agreed if it is manifestly excessive or derisory.

Lorsque l’engagement a été exécuté en partie, la pénalité convenue peut être diminuée par le juge, même d’office, à proportion de l’intérêt que l’exécution partielle a procuré au créancier, sans préjudice de l’application de l’alinéa précédent.

Where a party’s undertaking has been performed in part, an agreed penalty may be reduced by the court, even of its own initiative, in proportion to the advantage which partial performance has procured for the creditor, without prejudice to the application of the preceding paragraph.

Toute stipulation contraire aux deux alinéas précédents est réputée non écrite.

Any stipulation contrary to the two preceding paragraphs is deemed not written.

Sauf inexécution définitive, la pénalité n’est encourue que lorsque le débiteur est mis en demeure.

Except where non-performance is permanent, a penalty is incurred only where the debtor was given notice to perform.

CHAPITRE VI

CHAPTER VI

LES PRINCIPAUX REGIMES SPECIAUX DE RESPONSABILITE

THE PRINCIPAL SPECIAL REGIMES OF LIABILITY

SECTION 1

SECTION 1

Le fait des véhicules terrestres à moteur

The action of motor-vehicles29

Art 1285. – Le conducteur ou le gardien d’un véhicule terrestre à moteur répond de plein droit du dommage causé par un accident de la circulation dans lequel son véhicule, ou une remorque ou semiremorque, est impliqué.

Art 1285. – The driver or the keeper30 of a motor-vehicle is liable strictly for harm caused by a traffic accident in which his vehicle, trailer or articulated trailer, is involved.

29 The French specifies that the véhicule à moteur is ‘terrestre’ (ie by land) but this is not necessary in the English as ‘motor-vehicle’ is used only of road vehicles. 30 ‘The keeper’ translates le gardien which is defined in art 1243(4) of the Projet for the purposes of the general regime of liability for the actions of things.

498  Appendix Les dispositions de la présente section sont d’ordre public. Elles s’appliquent même lorsque la victime est transportée en vertu d’un contrat.

The provisions of the present section are a matter of public policy. They apply even where the victim is transported under a contract.

Art 1286. – La victime ne peut se voir opposer le cas fortuit ou le fait d’un tiers même lorsqu’ils présentent les caractères de la force majeure.

Art 1286. – A fortuitous event or the action of a third party may not be set up as a defence31 against the victim even where they bear the characteristics of force majeure.

Elle n’a pas droit à réparation sur le fondement de la présente section lorsqu’elle a volontairement recherché le dommage qu’elle a subi.

The victim has no right to reparation on the basis of the present section where he voluntarily sought the harm which he suffered.

Art 1287. – En cas de dommage corporel, la Art 1287. – In the case of personal injury, faute de la victime est sans incidence sur son the victim’s fault has no effect on his right droit à réparation. to reparation. Toutefois, la faute inexcusable prive la victime de tout droit à réparation si elle a été la cause exclusive de l’accident.

However, an inexcusable fault deprives the victim of any right to reparation if it was the exclusive cause of the accident.

Lorsqu’elle n’est pas la cause exclusive de l’accident, la faute inexcusable commise par le conducteur du véhicule terrestre à moteur a pour effet de limiter son droit à réparation.

Where it was not the exclusive cause of the accident, an inexcusable fault committed by the driver of the motor-vehicle has the effect of limiting his right to reparation.

Les victimes, hormis les conducteurs de véhicules terrestres à moteur, âgées de moins de seize ans ou de plus de soixantedix ans ou, quel que soit leur âge, titulaires, au moment de l’accident, d’un titre leur reconnaissant un taux d’incapacité permanente ou d’invalidité au moins égal à 80 p. 100, sont, dans tous les cas, indemnisées des dommages corporels.

Apart from drivers of motor-vehicles, victims aged less than 16 years or more than 70 years or, whatever their age, those who at the time of the accident bear an entitlement recognising them as permanently incapacitated or disabled by at least 80 per cent, are compensated for their personal injuries in all circumstances.

Art 1288. – En cas de dommage matériel, la faute de la victime a pour effet de limiter ou d’exclure l’indemnisation de ses préjudices lorsqu’elle a contribué à la réalisation du dommage.

Art 1288. – In the case of physical damage to property, the victim’s fault has the effect of limiting or excluding compensation for his losses where it contributed to the occurrence of the harm.

L’exclusion de l’indemnisation doit être spécialement motivée par référence à la gravité de la faute.

An exclusion of compensation must be specially justified by a court by reference to the seriousness of the fault.



31 ‘Set

up as a defence’ translates opposer.

Appendix  499 Toutefois, les dommages causés à des fournitures ou appareils délivrés sur prescription médicale sont indemnisés selon les règles applicables au dommage corporel.

However, harm caused to supplies or apparatus provided on medical prescription are compensated according to the rules governing personal injury.

Lorsque le conducteur d’un véhicule terrestre à moteur n’en est pas le propriétaire, la faute de ce conducteur peut être opposée au propriétaire pour l’indemnisation des dommages autres que corporels. Le propriétaire dispose d’un recours contre le conducteur.

Where the driver of a motor-vehicle is not its owner, any fault in that driver can be set up against the owner as regards the compensation of harm other than personal injury. The owner has a recourse against the driver.

SECTION 2

SECTION 2

Le fait des produits défectueux

The action of defective products

Art 1289. – Le producteur est responsable de Art 1289. – A producer is liable strictly for plein droit du dommage causé par un défaut harm caused by a defect in his product. de son produit. Les dispositions de la présente section sont d’ordre public et s’appliquent même lorsque la victime est liée au producteur par un contrat.

The provisions of the present section are a matter of public policy and are applicable even where the victim is linked to the producer by a contract.

Art 1290. – Les dispositions de la présente section s’appliquent à la réparation des préjudices qui résultent d’un dommage corporel.

Art 1290. – The provisions of the present section are applicable to the reparation of losses which result from personal injury.

Elles s’appliquent également à la réparation du préjudice supérieur à un montant déterminé par décret, qui résulte d’une atteinte à un bien autre que le produit défectueux lui-même, à condition que ce bien soit d’un type normalement destiné à l’usage ou à la consommation privés et ait été́ utilisé par la victime principalement pour son usage ou sa consommation privée.

They also apply to the reparation of a loss above an amount determined by decree, which results from damage to property other than the defective product itself, on condition that the property is of a type normally intended for private use or consumption and was used by the victim mainly for his own private use or consumption.

Art 1291. – Est un produit tout bien meuble, même s’il est incorporé dans un immeuble, y compris les produits du sol, de l’élevage, de la chasse et de la pêche. L’électricité est considérée comme un produit.

Art 1291. – A product is any movable property, even if it is incorporated into immovable property, including products of the soil, of stock-farming, hunting or fisheries. Electricity is considered to be a product.

500  Appendix Art 1292. – Un produit est défectueux au sens de la présente section lorsqu’il n’offre pas la sécurité à laquelle on peut légitimement s’attendre.

Art 1292. – A product is defective within the meaning of the present section where it does not provide the safety which one can legitimately expect.

Dans l’appréciation de la sécurité à laquelle on peut légitimement s’attendre, il doit être tenu compte de toutes les circonstances et notamment de la présentation du produit, de l’usage qui peut en être raisonnablement attendu et du moment de sa mise en circulation.

In the assessment of the safety which one can legitimately expect, account must be taken of all the circumstances and in particular the presentation of the product, the use to which it may reasonably be expected to be put and the time of its being put into circulation.

Un produit ne peut être considéré comme défectueux par le seul fait qu’un autre, plus perfectionné, a été mis postérieurement en circulation.

A product cannot be considered as defective by the mere fact that another, improved product has subsequently been put into circulation.

Art 1293. – Est producteur, lorsqu’il agit à titre professionnel, le fabricant d’un produit fini, le producteur d’une matière première, le fabricant d’une partie composante.

Art 1293. – A manufacturer of a finished product, a producer of raw materials, and a manufacturer of a component part are producers if they act in the course of a business or profession.32

Est assimilée à un producteur pour l’application de la présente section toute personne agissant à titre professionnel:

Where acting in the course of a business or profession, the following persons are assimilated to a producer for the purposes of the present section:

1° Qui se présente comme producteur en apposant sur le produit son nom, sa marque ou un autre signe distinctif;

1o A person who presents himself as producer by attaching to the product his name, trademark or other distinguishing feature;

2° Qui importe un produit dans l’Union européenne en vue d’une vente, d’une location, avec ou sans promesse de vente, ou de toute autre forme de distribution.

2 o A person who imports a product into the European Union with the view to sale, hire (with or without a promise to sell), or any other form of distribution.

Ne sont pas considérées comme producteurs, au sens de la présente section, les personnes dont la responsabilité peut être recherchée sur le fondement des articles 1646-1 et 1792 à 1792-6.

Persons whose liability may be sought on the basis of articles 1646-1 and 1792 to 1792-6 are not considered to be producers within the meaning of the present section.



32 This

translates agissant à titre professionnel: cf above, n 13.

Appendix  501 Art 1294. – Si le producteur ne peut être identifié, le vendeur, le loueur, à l’exception du crédit-bailleur ou du loueur assimilable au crédit-bailleur, ou tout autre fournisseur professionnel, est responsable du défaut de sécurité du produit, dans les mêmes conditions que le producteur, à moins qu’il ne désigne son propre fournisseur ou le producteur, dans un délai de trois mois à compter de la date à laquelle la demande de la victime lui a été notifiée.

Art 1294. – If the producer cannot be identified, the seller, the hirer (with the exception of a finance lessor or a hirer comparable to a finance lessor) or any other supplier in the course of business or a profession is liable for a defect of safety in the product on the same conditions as the producer, unless he indicates his own supplier or the producer within a period of three months starting from the date on which the claim of the victim was notified to him.

Le recours du fournisseur contre le producteur obéit aux mêmes règles que la demande émanant de la victime directe du défaut. Toutefois, il doit agir dans l’année suivant la date de sa citation en justice.

Recourse by the supplier against the producer is governed by the same rules as the claim coming from the direct victim of the defect. However, he must sue within a year of the date of proceedings being brought against him.

Art 1295. – En cas de dommage causé par le défaut d’un produit incorporé dans un autre, le producteur de la partie composante et celui qui a réalisé l’incorporation sont solidairement responsables.

Art 1295. – In the case of harm caused by a defect in a product incorporated into another product, the producer of the component part and the person who effected its incorporation are jointly and severally liable.

Art 1296. – Le demandeur doit prouver le dommage, le défaut et le lien de causalité entre le défaut et le dommage.

Art 1296. – The claimant must prove the harm, the defect and the causal relationship between the defect and the harm.

Art 1297. – Le producteur peut être responsable du défaut alors même que le produit a été fabriqué dans le respect des règles de l’art ou de normes existantes ou qu’il a fait l’objet d’une autorisation administrative.

Art 1297. – The producer may be liable for a defect even if the product was manufactured in accordance with the rules of the trade or existing standards or if it was the object of administrative authorisation.

Art 1298. – Le producteur est responsable de Art 1298. – The producer is liable strictly plein droit à moins qu’il ne prouve: unless he proves: 1° Qu’il n’avait pas mis le produit en circulation;

1o that he had not put the product into circulation;

2° Que, compte tenu des circonstances, il y a lieu d’estimer que le défaut ayant causé le dommage n’existait pas au moment où le produit a été mis en circulation par lui ou que ce défaut est né postérieurement;

2o that, having regard to the circumstances, there is reason to think that the defect causing the harm did not exist at the time when the product was put into circulation by him or that the defect arose afterwards;

502  Appendix 3° Que le produit n’a pas été destiné à la vente ou à toute autre forme de distribution;

3o that the product was not intended for sale or any other form of distribution;

4° Que l’état des connaissances scientifiques et techniques, au moment où il a mis le produit en circulation, n’a pas permis de déceler l’existence du défaut;

4o that the state of scientific and technical knowledge at the time when he put the product into circulation did not allow discovery of the existence of the defect;

5° Ou que le défaut est dû à la conformité du produit avec des règles impératives d’ordre législatif ou règlementaire.

5o or that the defect is due to compliance of the product with mandatory legislative or administrative rules.

Le producteur de la partie composante n’est pas non plus responsable s’il établit que le défaut est imputable à la conception du produit dans lequel cette partie a été incorporée ou aux instructions données par le producteur de ce produit.

Moreover, a producer of a component part is not liable if he establishes that the defect is attributable to the design of the product in which this part was incorporated or to instructions given by the producer of that product.

Art 1298-1. – Le producteur ne peut invoquer la cause d’exonération prévue au 4° de l’article 1298 lorsque le dommage a été causé par un élément du corps humain ou par les produits issus de celui-ci, ou par tout produit de santé à usage humain mentionné dans le premier chapitre du titre II du livre Ier de la cinquième partie du code de la santé publique.

Art 1298-1. – A producer cannot rely on the defence provided by paragraph 4 of article 1298 where the harm was caused by an element of the human body or by products derived from it, or by any health product for human use listed in the first chapter of Title II of Book I of the Fifth Part of the Code of Public Health.

Art 1299. – Les dispositions de l’alinéa 2 de l’article 1254 ne sont pas applicables.

Art 1299. – The provisions of paragraph 2 of article 1254 are not applicable.

Art 1299-1. – La responsabilité fondée sur les dispositions de la présente section est éteinte dix ans après la mise en circulation du produit même qui a causé le dommage à moins que, durant cette période, la victime n’ait engagé une action en justice.

Art 1299-1. – Liability based on the provisions of the present section is extinguished 10 years after the actual product which caused the harm was put into circulation unless the victim brought legal proceedings during this period.

Art 1299-2. – L’action en réparation fondée sur les dispositions de la présente section se prescrit par trois ans à compter du jour où le demandeur a eu ou aurait dû avoir connaissance du dommage, du défaut et de l’identité du producteur.

Art 1299-2. – An action for reparation based on the provisions of the present section is subject to prescription after a period of three years starting from the day on which the claimant knew or ought to have known of the harm, the defect and the identity of the producer.

Art 1299-3. – Les dispositions de la présente section n’interdisent pas à la victime d’invoquer les dispositions d’autres régimes de responsabilité contractuelle ou extracontractuelle, dès lors que ceux-ci reposent sur un fondement autre que le défaut de sécurité du produit.

Art 1299-3. – The provisions of the present section do not forbid a victim from relying on provisions of other regimes of contractual or extra-contractual liability, as long as the latter rest on a basis other than a defect of safety in the product.

INDEX acceptation des risques doctrine, 304 accidents at work, see workplace accidents action of assumpsit, 28–29, 37 actions of others, liability for, 142–77 administrative law (French), 433 collective liability, 435 enterprise liability, 436–37 liability for the acts of state agents, 434 liability of corporations, 425 liability of directors, 426 liability of legal persons, 435–36 non-performance of a service, 434–35 vicarious liability, 433–34 connection with defendant’s actions, 145–46 English law, see vicarious liability French law, see fait d’autrui German law, 165–66 expansion of the law, 170–72 liability of representatives, 170 organisational failure, 169–70 parental liability, 166–67 wrongs committed by employees, 167–69 insurance, 172–73 liability dependent on insurance, 173–75 third-party liability insurance, 175–77 legal history: Middle Ages, 149–50 19th century, 150 Roman law, 148–49 objective model of extra-contractual liability, 147–48 Projet de réforme articles, 483–84 responsible person: German law, 146–47 subjective model of extra-contractual liability, 147–48 see also fait d’autrui; vicarious liability action of things, liability for, 123–25, 144–45 Belgian law, 399 causation, 139–40 Code civil, 10 defective products, 320 force majeure, 405, 467–68 hostility towards concept, 125

imputability, 405 origins, 129 Projet de réforme, 144, 469 articles, 482 road traffic accidents, 4–5, 461 sports injuries, 318 strict liability, 4, 464 workplace accidents, 461, 471–72 see also road traffic accidents; workplace accidents acts or omissions, 31, 110 attribution, 105 Dutch law, 116 Italian law, 270 pure omission/non-feasance/ misfeasance distinguished, 95 unlawful acts, 113 administrative law (French), 100, 106 action of others, liability for, 433 collective liability, 435 enterprise liability, 436–37 liability for the acts of state agents, 434 liability of corporations, 425 liability of directors, 426 liability of legal persons, 435–36 non-performance of a service, 434–35 vicarious liability, 433–34 civil liability, relationship with, 425–26 coming to the nuisance, 237, 244 duty of care or diligence, 431 defective functioning of institutions, 431–33 fault liability: conditions, 429 emergency services, 429–30 gross negligence, 429 hospital liability, 429 operation of the justice system, 430 prison liability, 429 regulation of dangerous premises, 430 state compensation for fault, 428–29 fundamental public law principles: Allgemeines Landrecht (Prussia), 426 Declaration des droits de l’homme (F), 426

504  Index pre-Revolutionary France, 427 principle of state liability, 427–28 gross negligence, liability for, 429 emergency services, 429–30 hospital liability, 429 operation of the justice system, 430 prison liability, 429 regulation of dangerous premises, 430 illegality, 431 path dependency, 440–43 state liability principle harm principle, 427–28 solidarity principle, 427–28 strict liability, 437 diplomatic activity, 430 foreign affairs, 430 industrial accidents, 430 liability for public works, 437–38 liability for risks, 438–40 military activity, 430 road traffic accidents, 430 social solidarity basis, 440 alternative causation: Austrian law, 248–49, 258–59 comparative law, 250–54 concurrent events both triggering liability, 247 actual causation, 250–51, 252, 259–60 intentional coordination, 252–53 personal injury limitation, 254–55 potential causation, 251 several tortfeasors, 251–54 concurrent events one of which triggers liability: certainty about causation, 255–56 limitations of liability, 257 loss of chance, 256 reparable loss, 256–57 uncertainty about causation, 255–56 German law, 248, 257 likelihood of causation, 249, 250, 254, 258, 260–61 loss of chance compared, 247–48, 249–50 Projet de réforme, 247–48, 254–55, 259–60 see also joint and several liability; loss of chance amende civile, 269, 329–30, 344–47 apportionment of damage, 249, 254, 258 apportionment of liability, 275–77, 278–79, 298–99 see also contributory fault in claimant/victim; contributory negligence

Aquilian relativity theory, 91–92, 141, 414 assessment of damages, see damages, assessment of assignment of rights, 55 assignment of unlawful gains: Belgian law, 423–24 assumption of responsibility: duty of care, 65–66, 90–91 third-parties, towards, 67–69 negligence claims, 90–91 Hedley-Byrne rule, 31 attribution, 105, 145, 146, 159–60 Belgian law: imputability, 405–6 employers’ liability, 160–61, 163, 170 German law, 145–46, 373 employers’ liability, 170 Austrian law: concurrent events: both triggering liability, 253–54 one of which triggers liability, 258–59 joint and several liability, 253–54, 271 partial causation, 247, 248–49 concurrent events one of which triggers liability, 258–59 concurrent events triggering liability, 253–54 Avant-projet Catala, 6, 8, 10, 27, 251–52, 316, 324, 343, 349, 380, 432, 435–36, 439 Avant-projet Terré, 6, 8, 10, 96, 251–52, 316, 380–81, 382–83, 393, 420, 451 Belgian law: admissibility of claimant’s harm: legitimate interest, 419–20 capacity: loss of consciousness, 414 minors, 413 Code civil, 397–98 reform, 398–99 coexistence of liabilities, 407 concurrence of liabilities: actions between parties, 407–10 actions against agent for performance, 410–11 contractual and extra-contractual liability, relationship between: abuse of rights, 402–3 contract terms excluding liability, 404 distinct regimes, 401–2 force majeure, 405–6 foreseeability of harm, 404–5

Index  505 formation of contracts, 403 imputability, 405–6 obligation to reduce one’s own harm, 406–7 pre-contractual liability, 403 extra-contractual civil liability: causation, 399 contractual liability, relationship between, 401–411 defective products, 400 effects of liability, 400 grounds for, 399 harm/dommage, 400 introductory rules, 399 recourse actions, 400 reparation, 400 strict liability, 400 fault, 411–12 assessment of fault, 414–15 capacity, 413 grounds of fault, 412–13 moral element of fault, 413 harm/dommage: admissibility of claimant’s harm, 419–20 certainty of harm, 420–21 extra-patrimonial harm, 420–21 indirect harm, 420–21 infringement of legally protected interest compared, 416–17 infringement of personal or collective interest, 420 legitimate interest, 419–20 nature and extent, 418–19 patrimonial harm, 420–21 reparation of harm, 421–24 infringement of legally protected interest, 416–17 infringement of individual right distinguished, 417 repercussions, 418 reparation of harm, 421 assignment of unlawful gains, 423–24 injunctions, 421–23 punitive damages, 423–24 reparation in kind, 422–23 BGB, see German Civil Code breach of contract: action of assumpsit, 28 contributory fault defence, 300 damages, 32–35 force majeure, 288–89 French and English law compared, 56–60

negligence, 59 restitution, 373 tort compared, 24–25, 29–36 see also contractual non-performance breach of legislative/statutory duty English law, 88, 91, 115–16 French law, 80–81, 83–94, 114–15, 431 burden of proof, 159–60, 192–93, 211–12 Belgian law, 409–10, 413 defences, 287, 288–89, 290, 293, 294 reversal, 140, 147 employers’ liability, 168 medical liability, 257 parental liability, 150, 164, 166–67, 176 several tortfeasors, 278 teachers, 152 capacity: Belgian law, 406, 413 criminal law, 106 English law, 31–32, 93, 105–7, 108 French law, 325 minors, 31–32, 93 tort law, 31–32, 93, 105–6, 108 causal relationship, see causation; lien de causalité causation, 107, 107 alternative causation, see alternative causation assessment of causation, 139–42 actual causation, 250–51, 252 likelihood of, 254, 258, 259–60 Belgian law, 399, 418 concurrent events, see concurrent events consequential harm, 61 contributory negligence, 299–300 French law, 92, 112, 184, 446, 447–48, 450–51, 453 causalité adequate, 73, 139–40 equivalence of conditions, 139–40 lien de causalité, 108, 139, 279, 292–93, 472 personal injury, 467 presumption of causation, 140–41 proof of causation, 140 strict liability, 139–41 German law, 141–42 Italian law, 208–9 liability for alternative causation, see alternative causation omission, 95 partial causation, see partial causation

506  Index personal injury, 467 potential causation, 251 several tortfeasors, 265, 274 fault and causation distinguished, 276 strict liability, 139–42 uncertainty, 255–56 see also fait générateur; loss of chance causes d’irresponsabilité: causes d’exonération, 292–93 causes d’irresponsabilité subjectives, 291–92 fait justificatifs, 291 cessation of unlawful action Projet de réforme articles, 488 see also injunctive relief chains of contracts, 55–56 Civil Codes see Code civil (French), Code civil (Belgian) and German Civil Code (BGB) civil fault: Code civil, 79 comparative law, 81–83 English law, 80–81 fault by carelessness, 87–89 fault by contravention, 87–89 Projet de réforme, 79 assessment of fault, 90–94 fault by commission, 94–96 fault by omission, 94–96 flexible approach, 85–87 liability for fault, 79–80 seriousness of fault, 96–97 traditional approach, 83–85 see also fault civil liability defined, 1 duality thesis, 16–17, 39, 40–43 unity thesis, 17–18 challenges, 19–20 see also contractual liability; extra-contractual liability, Code civil (Belgian), 397–99 see also Belgian law Code civil (French), 1–2 outline, 9–11 reform, 6–7, 20–21 see also Projet de réforme responsabilité contractuelle, 1–2 responsabilité délictuelle, 1–2 responsabilité extracontractuelle,1–2 collective liability rule, 250–51 administrative law, 435 see joint and several liability

coming to the nuisance: English law: decline, 229–31 easements principle, 230 increased favour, 231–32 interference with senses, 231–32 proprietary judicial reasoning, 229–30 public nuisance, 231 servitude, 230, 239–43 French law: administrative law, 237, 244 increased favour, 234–36 rejection of concept, 232–34 servitude, as a, 238–39 use of land restrictions, 236–37 compensation, 3–6, 329–30 consequential loss, absence of, 217–19 contractual liability, 5 criminal and tort law compared, 103–4 duty of care, 90–91 extra-contractual liability, 4 Projet de réforme, 10, 198–203 réparation intégrale, 200–3 road traffic accidents, 199 social security regime, 198–99 Italian law: compensation requirements, 208 danno inguisto, 207 danno-conseguenza, 207–8 danno-evento, 207, 208 liability towards third parties, 67–69 third-party actions, 65, 67 see also damages; reparation concurrent events: actual causation, 250–51 Austrian law: both triggering liability, 253–54 one of which triggers liability, 258–59 French law: both triggering liability, 251–52 one of which triggers liability, 256–57 German law: both triggering liability, 252–53 one of which triggers liability, 257 potential causation, 251 triggering liability (both events): claimant’s rights and interests, 260 defendant’s fault, 260 degree of coordination, 260 likelihood of actual causation, 250–51, 260

Index  507 triggering liability (one event only): defendant’s fault, 260 likelihood of causation, 255–56, 260 uncertain causation, 255–56 conflict of laws, 31–32 coming to the nuisance, 231–32 consent/consentement, 292 acceptation des risques, 304 autonomy of will, 324 freedom to consent, 321–24 interferences with property, 303 interferences with the body, 303–4 volenti non fit injuria, 304–5 consequential losses, 220–21 breach of duty of information, 217–18 compensation of harms, 217–21 emotional distress, 219 pain and suffering, 219 personal injury, 219 psychological injury, 219 recoverable damages, as, 220 unfair competition, 218–19 see also danno-conseguenza; préjudice contracts concerning civil liability, 309–311 clauses excluding and limiting liability, 311 content: restrictions as to, 324–26 contracts concerning liability defined, 312–13 compromise agreements distinguished, 313 settlement agreements distinguished, 313 contractual freedom: public policy, 314–16 validity of contracts, 316–17 effects, 326 enforceability, 321 freedom to consent, 321–23 autonomy of will, 324 method of offer and acceptance, 323–24 lawfulness: contractual freedom, 314–17 limitations, 317–21 limitations on excluding liability: fault, 319–21 personal injuries, 317–19 penalty clauses, 310–12, 326, 466 personal injuries, exception, 317–19 restrictions as to, 324–26 unequivocal acceptance, 321–23 validity of contracts, 316–17

contractual non-performance damages, measures for, 24–25, 41–42, 43–47 delictual/extra-contractual liability and, 16–20, 21–24, 40–43 Projet de réforme, 21–23, 25–27, 32–33. third party claims English law, 64 assumption of responsibility, 67–69 duty of care, 66 negligence, 66 requirements for success, 65–69 French law, 60–61 authorised actions, 61 contractual actions, 74–75 contractual failing, 62–63 Boot Shop case, 69–71 establishing non-performance, 61–62 fault in the defendant, 62 legitimate interest in bringing actions, 61, 63–64 ordinary actions, 61 see also breach of contract contribution, 275–77 Projet de réforme articles, 488 contributory fault in claimant/victim: apportioning fault and liability, 298–99, 301–2 breach of contract, 300 causation, 292, 299–300 criticisms of defence, 299 defence, as, 286, 292, 293, 460, 466 determination of fault, 301 English law, 298–301 exceptions: conversion claims, 300–1 deceit, 300 intentional wrongful interference with goods, 300 trespass to the person, 300 French law, 288, 298–99 intention, 301 product liability, 468 road traffic accidents, 468 parental liability, 154 Projet de réforme, 302–3 articles, 488 third-party payers, 199 contributory negligence, see contributory fault in claimant/victim

508  Index corporate liability, 425, 443 criminal law: remedies: parallel proceedings, 101, 111 prior convictions, 101–2 priority of civil/criminal remedy, 104 procedural advantages, 102 range of remedies, 104 victim compensation, 102–4 tort law, interaction with: institutional interaction, 100 legal reasoning, 100–1 parallel proceedings, 101, 111 prior convictions, 101–2 priority of civil/criminal remedy, 104 procedural advantages, 102 procedural connections, 101 range of remedies, 104 remedies, 101–5 substantive law, 101 victim compensation, 102–4 tort law, integration with: Dutch law, 116–17 English law, 115 French law, 114 general integrative techniques, 114–17 German law, 114–15 limitation of liability, 114 Spanish law, 115 unlawful acts, 113–14, 116–17 unity of criminal and civil faults, 99, 102, 104, 109–13, 118–19, 125, 295 damage, see dommage, harm damage to property: Projet de réforme articles, 494 damages: assessment, 43–44, 47, 192, 204, 326, 372, 375, 460–61 breach of contract, 32–33 contract and tort distinguished, 32–36 measures of damages, 32–36 performance interest measure, 25, 33 Projet de réforme articles, 487–88 reliance interest measure, 25, 33 reparation or performance, 43–44 see also compensation; exemplary damages; measure of damages; punitive damages; reparation danno biologico, 209 compensation, 217, 219

interpretation as danno-conseguenza, 210–11 parallel with préjudice/déficit fonctionnel, 211, 215 danno-conseguenza, 207–8, 210 danno-evento, distinction between, 208–9 interpretation of danno biologico, 210–11 préjudice distinguished, 219–20 see also danno inguisto danno esistenziale, 209, 211 compensation, 219 parallel with préjudice d’agrément, 211 danno-evento, 207, 208, 210 danno-conseguenza, distinction between, 208–9 dommage distinguished, 219–20 see also danno inguisto danno inguisto, 207 danno-evento and danno-conseguenza distinguished, 207 easement, 212 French law compared, 214–15 harm to someone else, 212 infringement of ownership, 212 recoverable losses, 213–14 right to possess someone else’s property, 212 servitude, 212 usufruct, 212 widening scope, 212–13 harm caused by public official, 213 violation of right in personam, 213 see also danno-conseguenza; danno-evento danno-morale (soggettivo): consequential loss, 210, 219 defective products, see product liability defences: burden of proof, 287, 288–89, 290, 293, 294 causes d’irresponsabilité: causes d’exonération, 292–93 causes d’irresponsabilité subjectives, 291–92 faits justificatifs, 291 common law taxonomy of defences, 294 consent/consentement, 303–5 acceptation des risques, 304 interferences with property, 303 interferences with the body, 303–4 volenti non fit injuria, 304–5 contributory fault in claimant/victim: apportioning fault, 301–2 causation, 299–300

Index  509 criticisms of defence, 299 determination of fault, 301 English law, 298–300 French law, 298–300 Projet de réforme, 302–3 contributory negligence: see contributory fault in claimant/ victim English law, 287–88, 290 exclusion/limitation of liability, 293–94, 305 exculpation defence, 151–52 excusatory defences, 297–98 French law, 288, 289–90 illegality, 293–94 justifications, 294–95 necessity/état de necessité, 296–97 self-defence/légitime défence, 295–96 limitation of actions, 293–94 public policy defences, 305 English law, 306 French law, 305–6 responsibility-based defences, 298 contributory fault in claimant/victim, 298–03 tort law defences, 289 delictual liability, see extra-contractual liability dommage, 108 actionability of damage: Code civil, 185–89 directness of harm, 186 dommage certain, 186–88 legality of damage, 188–89 personal nature of the harm, 186 Projet de réforme, 189–90 Code civil: actionability of damage, 185–89 joined concepts of dommage and préjudice, 183–84 conséquences non-patrimoniales, 184 conséquences patrimoniales, 184 danno-evento distinguished, 219–20 dommage corporel, 183 préjudices extra-patrimoniaux, 184, 189–90, 193–200 préjudices patrimoniaux, 184, 193–200 dommage matériel, 183, 189–90 dommage moral, 183, 189 dommages et intérêts, 184 see also damages préjudice distinguished, 183–85, 205–07

Projet de réforme: actionability of damage, 189–90 dommage certain, 190–91 legality of damage, 190–91 losses arising from different categories of dommage, 185 separation of dommage and préjudice, 184–85 see also préjudice reparation in kind, 184 see also damage, harm dommage certain, 186–88, 191–92 dommage corporel, 183 alternative causation limitation, 255 consequential loss, 219 déficit fonctionnel, 195, 197 Loi Badinter, 194–96 measure of damages, 203 personal harm (pain and suffering), 194, 219 préjudice d’agrément, 196, 197, 198 préjudice de caractère personnel, 197 préjudice esthétique, 196 préjudices extra-patrimoniaux, 184, 189–90, 193–200 préjudices patrimoniaux, 184, 193–200 social security system, 194 third-party payers, 195–96 law reform, 196–98 dommage moral, 183, 189, 195, 202, 219, 255, 461 Dutch law: criminal and tort law, integration of, 116–17 joint and several liability, 271 unlawful acts, 113 duty, breach of legislative, see breach of legislative/statutory duty duty of care: administrative law (French), 431–33 English law assumption of responsibility, 90–91 carelessness, 86–88 compensation, 90–91 controlling extent of civil liability, 90–91 duty to act, 90, 105 fair, just and reasonable, 90–91 foreseeability, 90–91 intangible harms, 90 pure economic loss and, 90–91 scope of duty, 90

510  Index French law administrative law, 431–33 contractual non-performance and third party claims, 59, 66 fault (civil) compared, 86–87, 90–91 emotional distress/harm, 3, 183, 190, 194–96, 198, 210, 219 see also dommage moral employers’ liability: English law, 162–63 French law, 147, 151, 154–55 German law contractual and extra-contractual liability, 170–72 employees’ acts, 167–69 organisational failure, 169–70 representatives’ acts, 170 Projet de Réforme, 159–61 English law: action of assumpsit, 28 breach of contract: French and English law compared, 56–60 see also contractual non-performance, third party claims coming to the nuisance: decline, 229–31 easements principle, 230 increased favour, 231–32 interference with senses, 231–32 proprietary judicial reasoning, 229–30 public nuisance, 231 servitude, 230, 239–43 contract and tort distinguished, 27–28 bases of liability, 31 differences in substantive rule, 31–32 French law compared, 33–36 origins, 28–31 rules governing damages, 32–36 contractual non-performance, third party claims assumption of responsibility, 67–69 duty of care, 66 negligence, 66 requirements, 65–69 contributory fault in claimant/victim, 298–99 contributory negligence, see contributory fault in claimant/victim employers’ liability for employees’ actions, 162–63

exemplary damages, see exemplary damages fault, 80–81 breach of statutory duty compared, 88 tort of negligence, 88 groups of contracts, 58–59 injunctive relief, 384–85 damages in lieu of injunctions, 387–88 equitable relief, 385 interim injunctions, 386–87 judicial discretion, 387–88 mandatory injunctions, 385–86 prohibitory injunctions, 385 quia timet injunctions, 385, 386 insurance law and the liability regime: home/household insurance, 176–77 measures of damages, 32–36 contract law, 32–34 French law compared, 33–36 tort and contract compared, 34–36 tort law, 32 negligence, tort of, 31, 58–60, 63, 65, 66–69, 81–82, 88, 161, 164, 286, 291, 292, 456–8. nuisance, tort of: strict liability nature of nuisance, 223 see also coming to the nuisance parental liability, see parental liability public policy defences, 305 punitive damages: United States compared, 342 see also exemplary damages; punitive damages retribution, 339–42 servitudes, 230, 239–43 third-party insurance: home/household insurance, 176–77 vicarious liability, see vicarious liability victim compensation, 103–4 environmental liability, 123, 204, 474 concurrent causation, 276 Projet de réforme articles, 494–96 EU Law: competition law, 219 joint and several liability, 272 Principles of European Tort Law, 89, 94, 214, 272, 398 product liability, 5, 10–11, 56, 400, 472 public supply and public works, 213 exclusion of liability: Projet de réforme articles, 485–86 see also limitation of liability

Index  511 exculpation defence, 151–52, 154, 158, 170 decentralised exculpation, 169 employers’ liability, 169, 170 parental liability, 156, 166 exemplary damages, 329, 337, 350–51 criticisms, 337 instrumentalism, 338–39 criticisms of exemplary damages as a remedy, 337 interpersonal justice, 339–40 functional critique, 341–42 structural critique, 340–41 justification, 338 private retribution, 339 when awarded, 337, 338 see also amende civile; exemplary damages extra-contractual liability (generally), 1–6 action of things, see action of things, liability for action of others, 147 objective approach, 147–48 subjective approach, 147–48 contractual liability, relationship with, 9, 38 unification of regimes, 466–68 “fair, just and reasonable”: duty of care, 90–91, 457 fait d’autrui, liability for: actions of things, 144–45 Code civil, 152–53 employers’ liability, 151, 154–55 parental liability, 151–52, 153–54 professional liability, 151 harms caused by another, 144–45 Projet de réforme, 155–56 employers’ liability, 156, 159–61 persons in charge of adults, liability of, 156, 157 persons in charge of minors, liability of, 156, 156–57 persons voluntarily in charge of adults, liability of, 156, 157–59 see also action of others, liability for; vicarious liability fait générateur (de responsabilité): abnormal nuisance, 473 causal link with harm, 139, 141, 446, 450–51, 470 concept, 8, 108–9, 119, 141 contractual and extra-contractual parallels, 15–16, 21–23

employers’ liability, 159 fait dommageable, relationship to, 26 fait générateur de responsabilité extracontractuelle, 156 multiple tortfeasors, 269 Projet de réforme, 8, 22–23, 89, 452–53, 473 unjust enrichment, 366 fault, 3, 79–97 administrative law (French) organisational failure, 169–70 Aquilian relativity theory, 91–92 assessment of fault, 91–92 classification of fault: fault by carelessness, 86–88 fault by commission v fault by omission, 94–96 fault by contravention, 86, 87–88 seriousness of fault, 96–97 criminal and tort law, integration: Dutch law, 116–17 English law, 115 French law, 114 general integrative techniques, 114–17 German law, 114–15 limitation of liability, 114 Spanish law, 115 unlawful acts, 113–14, 116–17 fait générateur, 108 fault by contravention, 86, 87–88 flexibility of concept, 85–86, 88–89 failure of duty of care or diligence, 86–88 fault by contravention, 86, 87–88 intention to injure, 3, 94, 97, 112, 463 legislative definitions (French law), 83–85 legislative duty, breach of, see breach of legislative/statutory duty negligence, see negligence Projet de réforme articles, 481–82 proof of fault, 88–89 standards, 92–94, 107 unity in fault, 109–10 abstract v practical development, 117–18 conceptual understanding, 118 civil and criminal fault, relationship between, 110–12 Projet de réforme, 112–13 societal challenges to, 110 unjust enrichment: claimant’s fault, 367–69 defendant’s fault, 369–71 unlawful acts, 114–15 see also civil fault; duty of care; negligence

512  Index faute de la victime, see contributory fault in claimant/victim floodgates of liability, 3–4, 357, 361 force majeure, 84, 142, 152, 297, 460, 466, 467–68 Belgian law, 405–6, 413–14 burden of proof, 288–89, 294 causes d’exonération, 292–93 medical liability, 474 foreseeability, 5, 22, 44–46, 73, 93, 187, 457 Belgian law, 404–5, 413, 415 duty of care, 90–91 limitation of damages, 75 pure economic loss, 90–91 full reparation principle, 24–26, 35, 46, 80, 123, 125, 129, 136–37, 145, 152–153, 181–82, 347–49, 459–60 calculation of damages, 200–3 see also reparation generality of approach (French law), 455–462 action of things, liability for, 123, 125, 129, 460, 461 action of others, liability for, 165–172 adoption of special regimes of liability, 462 diversity within liability for fault: conditions of liability, 463–64 rationae materiae, 462–63 rationae personae, 462 special regimes, 462 diversity within liability for the action of things: animals, 464 moving things, 464 non-moving things, 464 road traffic accidents, 464–65 diversity within liability for the actions of others: employers’ liability, 465 liability of persons in charge of mode de vie, 465 liability of sports associations, 465 liability of those in charge of adults, 465 parental liability, 465 strict liability, 465 fault, liability for, 80, 460, 461 general tendency towards, 460–61 Projet de réforme, 461–62 German Civil Code, 107, 114, 116, 150, 209, 248, 280, 289, 381, 417, 436 see also German law

German law: attribution, 145–46 causation, 107 Aquilian relativity, 141 strict liability, 141–42 concurrent events: both triggering liability, 252–53 one of which triggers liability, 257 criminal and tort law, integration of, 114–15 employers’ liability for employees etc: contractual and extra-contractual liability, 170–72 employees’ acts, 167–69 organisational failure, 169–70 representatives’ acts, 170 fault, 3, 114–15 insurance law and the liability regime, 172–73 private/personal third-party liability insurance, 175–76 joint and several liability, 251, 252–53, 269–70 medical liability: no liability rule, 257 reversal of burden of proof, 257 parental liability, 166–67 partial causation: concurrent events one of which triggers liability, 257 concurrent events triggering liability, 252–53 German Civil Code, 247, 248 pure economic loss, 215 reparation, 26 strict liability, 137–39 strict liability, 4, 123, 136, 137–39, 141–42 third-party insurance, 172–73 private/personal third-party liability insurance, 175–76 workplace accidents, 134–35 gross negligence, 429 administrative law (French): emergency services, 429–30 hospital liability, 429 operation of the justice system, 430 prison liability, 429 regulation of dangerous premises, 430 contract terms which limit/exclude liability, 311–12 parental liability, 166–67

Index  513 groups of contracts: English law, 58–59 French law, 55–58, 64, 73, 324 harm, 108 Code civil, 181–82 joined concepts of dommage and préjudice, 183–84 distinguishing between nature of claimant’s harm, 468–71 dommage, 108 French law: Code civil, 183–84 joined concepts of dommage and préjudice, 183–84 Projet de réforme, 184–85 separation of dommage and préjudice, 184–85 Italian law: causation between danno-conseguenza and danno-evento, 208–9 compensation requirements, 208 danno biologico, 209, 210–11 danno esistenziale, 209 danno inguisto, 207 danno-conseguenza, 207–8, 210 danno-evento, 207, 208, 210 non-patrimonial harms, 209 unlawfulness of damage, 212–14 loss distinguished, 8 Projet de réforme: separation of dommage and préjudice, 184–85 see also dommage human rights: damages for violation of human rights, 221 injunctions, 387 road traffic accidents: personal and psychological injury, 209–12 illegality, 105, 116–17 administrative law, 431 defence, 293–94 see also breach of legislative/statutory duty; unlawfulness injunctive relief: anticipation of commission of a tort, in, 392–93 cessation de l’illicite: Projet de réforme, 381–82

damages in lieu of injunctions, 387–88 English law, 384–85 damages in lieu of injunction, 387–88 equitable relief, 385 interim injunctions, 386–87 judicial discretion, 387–88 mandatory injunctions, 385–86 prohibitory injunctions, 385 quia timet injunctions, 385, 386 equitable relief, 385 judicial discretion, 387–88 French law compared, 389–91 French law: cessation de l’illicite, 381–82 interlocutory référé procedure, 378–79 historic background, 378–81 troubles illicites, 382–84 interim injunctions, 386–87 interlocutory référé procedure, 378–79 historic background, 378–81 mandatory injunctions, 385–86 prohibitory injunctions, 385 quia timet injunctions, 385, 386 troubles illicites, 382–84 insurance law and liability, 172–73 English law: home/household insurance, 176–77 German law: private/personal third-party liability insurance, 175–76 parental liability, 174, 175–76, 177 separation principle, 173–75 third-party insurance as a determinant of liability, 173–75 see also third-party payers intangible harms: duty of care, 90 proximity, 90–91 pure economic loss and, 90–91 scope of duty, 90 intangible loss, 69, 90 intention to injure, 3, 94, 97, 112 abuse of rights claims, 463 negligence distinguished, 83, 133, 278, 470 interferences with property, 303, 468 interferences with the body, 303–4 Italian law: causation, 107 danno-conseguenza and danno-evento, between, 208–9

514  Index compensation, 5–6 compensation requirements, 208 danno inguisto, 207 danno-conseguenza, 207–8 danno-evento, 207, 208 harm, 108 causation between danno-conseguenza and danno-evento, 208–9 compensation requirements, 208 danno biologico, 209, 210–11 danno-conseguenza, 207–8, 210 danno esistenziale, 209 danno-evento, 207, 208, 210 danno inguisto, 207 non-patrimonial harms, 209 unlawfulness of damage, 212–14 joint and several liability, 270 limitations of liability, 213–15 pure economic loss, 215 road traffic accidents, 209–10 unlawfulness, 207, 212–14 wrongful birth, 216 wrongful death: danno biologico terminale, 217 danno catastrofale, 217 danno tanatologico, 216 pure non-economic loss, 216–17 see also danno biologico; danno-conseguenza; danno esistenziale; danno-evento; danno inguisto joint and several liability, 250–51, 263–64 Austrian law, 253–54, 271 civil law systems, 269–71 common law systems, 271 Dutch law, 271 EU law, 272 French law, 251–52 contribution and apportionment of liability, 275–77 defective products, 266 parental liability, 265–66 presumption of solidarity, 264–65 Projet de réforme, 268–69 same harm, 273–74 see also obligatio in solidum German law, 251, 252–53, 269–70 Italian law, 270 non-joint liability compared, 263 Portugese law, 270 Projet de réforme, 268–69 general rule on solidarity, 277, 280–82

specific rule on personal injury cases, 277–80 unidentified tortfeasor, 277–80 soft law, 272 judicial discretion, 6, 181–82 law of obligations, 6–7, 27, 37, 125, 265, 432, 466, 471 German law, 170–71 law reform, 150–51, 343 non-cumul rule, 47–48 tort law, 100 legality of damage/loss, 188–89, 190–91, 192–93 legislative duty, see breach of legislative/ statutory duty légitime défence, 295–96 lien de causalité, 108, 139, 279, 292–93, 472 see also causation limitations of liability, 57 consent, 305 contracts concerning civil liability, 309–311 clauses excluding and limiting liability, 311 lawfulness, 314–21 liability for fault, 319–21 personal injuries, 317–19 defences, as, 293–94 French and Italian law compared, 213–15 see also contracts concerning civil liability limitations on reparation, 137–39 non-material loss, 137–38 patrimonial loss, 137 loss, 108 harm distinguished, 8 see also dommage, préjudice loss of chance, 192 Austrian law, 258–59 French law, 247 likelihood of actual causation, 250 reparable loss, reconceptualisation as, 256–57, 259–60 see also alternative causation loss of reputation, 35, 416, 423–24, 470 measures of damages: Code civil, 181–82 contractual and extra-contractual liability compared, 24–27, 41–43, 47 dommage corporel, 203 full reparation principle, 200–3

Index  515 English law, 32–36 contract law, 32–34 French law compared, 33–36 tort and contract compared, 34–36 tort law, 32 préjudice d’agrément, 198 medical liability: breach of a duty of information, 217–18 German law: no liability rule, 257 reversal of burden of proof, 257 Projet de réforme, 474 risk, 439, 440 mental distress, 35, 90–91, 255 see also dommage moral misfeasance: pure omission/non-feasance distinguished, 95 misrepresentation, 70, 287 moral harm, 90–91, 255, 372 see also danno-morale (soggettivo); dommage moral Napoleonic Code, see Code civil necessity/état de necessité, 296–97 negligence, 3, 29–31, 456–57 contributory negligence, see contributory fault in claimant/victim criminal law, 111–12 fault compared, 82, 88, 90 French law employers’ liability to employees, 133, 148 fault by carelessness, 86–88 obligation de moyens, 5 gross negligence, see gross negligence intention distinguished, 83, 133, 278, 470 obligation de moyens, 5 parental accountability, 164–65, 167 tort of (English law) assumption of responsibility, 67–68, 91 breach of contract, third party claims, 31, 59, 65–68, 91 breach of duty of care, 81–82 breach of statutory duty distinguished, 88–89 failure to supervise children/adults sufficiently, 161, 164–65 integrity of, 114–15 limitations, 66–67 omissions, 31 pure economic loss, 31, 59

relative nature, 90–91 see also duty of care no-fault liability, see strict liability non-compensatory functions of civil liability, 8–9 non-cumul rule, 5, 9, 16, 38–39, 47–53, 56, 170 action civile and, 49–50 exceptions, 49 potential development of the rule, 51–52 significance, 48–49 subsidiary claims, 50–51 non-feasance: pure omission/misfeasance distinguished, 95 non-material loss: strict liability, 137–38 non-patrimonial loss, 182, 197–98, 199–200, 203 Italian law, 216–17 see also préjudices extra-patrimoniaux nuisance, 128 coming to the nuisance: English law, 229–32 French law, 232–37 see also coming to the nuisance English law: strict liability nature of nuisance, 223 French law: Code civil, 223–24 judicial development, 224 troubles de voisinage, 224 see also troubles de voisinage privacy and views, 228–29 private nuisance, 123 property law v tort law, 237–38 troubles de voisinage compared: locality of wrong, 226 secondary importance of public law, 226 strict liability torts, 225 types of damage covered, 226 types of interference covered, 225–26 troubles de voisinage contrasted: exposure to liability, 226 personal injury claims, 227 protection of property rights v limitation of property rights, 226 scope, differences in, 227 types of interference, 227–29 use of land, 228 who can be sued, 226–27 who can sue, 226–27

516  Index obligatio in solidum, 250–51, 263–64, 266–67 see also joint and several liability obligation de moyens, 5, 414, 436, 474 obligation de résultat, 5, 127 obligation de sécurité, 5, 127 obligations solidaires, 263–64 see also joint and several liability omissions, see acts or omissions pain and suffering, 194, 210, 219 see also dommage corporel parallel proceedings, 101, 111 parental liability: actions of others, liability for, 166–67, 465 action of others, liability for, 151–52, 153–54, 144–45, 155–56 contributory fault in claimant/victim, 154 English law, 164–65 exculpation defence, 156, 166 fait d’autrui, 151–52, 153–54 French law: Code civil, 147, 150, 151–52, 153–54 joint and several liability, 265–66 Projet de réforme, 156–59 German law, 166–67 gross negligence, 166–67 insurance law and the liability regime, 174, 175–76, 177 joint and several liability, 265–66 reversed burden of proof, 150, 164, 166–67, 176 third-party insurance, 174, 175–76 partial causation: Austrian Civil Code, 247, 248–49 concurrent events one of which triggers liability: Austrian law, 258–59 French law, 256–57 German law, 257 uncertain causation, 255–56 concurrent events triggering liability: actual causation, 250–51 Austrian law, 253–54 French law, 251–52 German law, 252–53 potential causation, 251 German Civil Code, 247, 248 Projet de réforme, 247, 249–50 alternative causation, 247 loss of chance, 247 see also alternative causation; concurrent events

patrimonial loss, 137, 182, 193–98 see also préjudices patrimoniaux penalty clauses, 310–12, 326, 466 Projet de réforme articles, 497 performance, 5 damages, 43–44 English contract law, 32–33 French law compared, 33–36 performance by equivalent, 41, 44–46 performance interest damages, 47 reliance interest damages compared, 25, 33 see also contractual non-performance personal injury compensation, 5, 42, 182 French social security system, 194 exceptions, 194 limitation/exclusion of liability, 317–19 non-cumul rule, 52–53 Projet de réforme articles, 489–94 troubles de voisinage, 227 Portugese law: joint and several liability, 270 préjudice, 108 Code civil: joined concepts of dommage and préjudice, 183–84 danno-conseguenza distinguished, 219–20 dommage distinguished, 183–85, 205–07 préjudices extra-patrimoniaux, 184, 189–90, 193–200 préjudices patrimoniaux, 184, 193–200 Projet de réforme: separation of dommage and préjudice, 184–85 unjust enrichment, 358, 361 action for enrichissement injustifié, 363–66 diverse approaches, 360–61 English notion of loss, 362 French notion of loss, 362–63 préjudices extra-patrimoniaux: Code civil, 184, 189–90, 193–98 Projet de réforme, 198–200 préjudices patrimoniaux: Code civil, 184, 193–98 Projet de réforme, 198–200 principles: abuse of rights, 403 applicability, 456 causation, 266–69, 273

Index  517 contractual, 309–10, 404 contractual non-performance, third party claims, 58–60, 65, 69–71 extra-contractual liability, 79–80, 123, 125, 129–30, 150, 153, 155, 154, 189, 192, 426, 455–62 full reparation principle, see full reparation principle general application, 456 legal proposition of importance, 456 negligence, tort of: assumption of responsibility, 31 duty of care test, 457–58 public law, 425–430, 434, 437, 440–443 rules distinguished, 456–57 separation, principle of (German law), 173 unjust enrichment, 353–57 see also generality of approach, principe explained principe explained: codification, lack of, 459–60 general clause, translated as, 459 liability for fault principle généraux, 459 rule, synonymous with, 458 priority of contractual/extra-contractual action, 73–74 see also non-cumul rule priority of civil/criminal remedy, 104 priority of civil liability/unjust enrichment, 371–72 private nuisance, 123, 475 product liability, 18 action of things, liability for, 464–65 contributory fault in claimant/victim, 468 EU law, 5, 10–11, 400, 472 exclusion and limitation clauses, 318 liability of manufacturers: joint and several liability, 266 Projet de réforme articles, 499–501 special liability regime, 472–74 professional liability, 18, 151 Projet de réforme de la responsabilité civile (Projet de réforme), 1, 7–8 contractual and extra-contractual liability distinguished, 9, 22 duality approach, 16–17, 39, 40–43 full text, 477–501 single civil liability framework, 15–16 public policy: contracts concerning civil liability, 314–16

public policy defences, 294, 305–6 English law, 306 punitive damages, 329 France: amende civile, see amende civile Avant-projet Catala, 343–44 lack of, 343 split-recovery scheme, 343 popularity in United States, 330–31 private retribution: civil recourse, 335 corrective justice, 334–35 criminal punishment, 335 interpersonal justice, 334–36 non-instrumentalism, 334–36 private punishment, 335 revenge, 334–36 social policy, 331 criminal-like nature of punitive damages, 331–32 deterrence policy, 332–33 instrumentalism, 331, 331–34 societal compensation, 333 pure economic loss, 31, 59 compensation, 215–17 duty of care and, 90–91 French law, 215 German law, 215 Italian law, 215 third-party actions, 65–67 assumption of responsibility, 67–69 pure non-economic loss, 215 compensation, 215–17 see wrongful birth; wrongful death quasi-contracts, 353–55, 367, 375 see also unjust enrichment quasi-delicts, 96, 110, 149, 181, 315, 321 Reform Bill on Civil Liability, see Projet de réforme de la responsabilité civile reliance interest damages, 45 performance interest damages compared, 25, 33 remedies: damages, see damages injunctions, see injunctive relief tort and criminal law, interaction between: parallel proceedings, 101, 111 prior convictions, 101–2 priority of civil/criminal remedy, 104 procedural advantages, 102

518  Index range of remedies, 104 victim compensation, 102–4 remoteness of damage, 32, 35 reparable loss, 10, 21, 23 Projet de réforme articles, 480–81 see also full reparation principle reparation, 26 damages, 43–44, 46–67 see also damages full reparation principle, see full reparation principle strict liability: French law, 136–37 full reparation principle, 136–37 German law, 137–39 limitations of reparation, 137–39 non-material loss, 137–38 patrimonial loss, 137 reparation in kind, 22, 23–24, 184, 202, 268–69, 383–84 Belgian law, 421–23 primacy, 400 Projet de réforme articles, 486–87 reputation, see loss of reputation responsabilité civile: nature of category, 15–25, 37–44 responsabilité administrative distinguished, 1 responsabilité pénale distinguished, 1 unity or duality, 9, 16–19, 26, 36 see also contractual liability; extra-contractual liability responsabilité contractuelle, see contractual liability responsabilité délictuelle, see extra-contractual liability responsabilité extracontractuelle, see extra-contractual liability responsabilité sans faute, see strict liability restitutio in integrum, 201 restitution, 26, 201 unjust enrichment, 355–56, 360–63, 365–66, 371–74, 375 retribution: English law, 339–42 French law, 343–49 United States, 334–36 reversed burden of proof, 140, 147 employers’ liability, 168 medical liability, 257 parental liability, 150, 164, 166–67, 176

road traffic accidents, 4–5, 10–11, 142, 430 excluding liability for fault, 319–20 Italy, 209–10 Loi Badinter, 4–5, 11, 109, 123, 194–96, 462, 464–65 Projet de réforme articles, 497–99 secondary liability, 163 vicarious liability, see vicarious liability self-defence/légitime défence, 295–96 servitudes: coming to the nuisance: English law, 230, 239–43 French law, 238–39, 241–43 danno inguisto, 212 restrictive covenants, 240–41 special regimes, 471 abnormal nuisance between neighbours, 473 defective products, 472–73 see also product liability environmental harm, 473–74 medical liability, 474 see also medical liability road traffic accidents, 473 see also road traffic accidents workplace accidents, 471–72 see also workplace accidents standard of care, 93, 127, 298 parental liability, 164–65, 174 statutory duty, see breach of legislative/ statutory duty strict liability, 4, 123 accidents at work, see workplace accidents administrative law, 437 diplomatic activity, 430 foreign affairs, 430 industrial accidents, 430 liability for public works, 437–38 liability for risks, 438–40 military activity, 430 road traffic accidents, 430 social solidarity basis, 440 causation, assessment of, 139–42 comparative law, 123–24 scope of application, 124–28 employers’ liability, 161 French law, 109 Germanic laws, 131 origins, 132 pre-Napoleonic Code, 129–32 Roman law, 130–31

Index  519 integration within liability systems: French law, 135–36 German law, 136 full reparation principle, 136–37 patrimonial loss, 137 non-material loss, 137–38 limitations of reparation, 137–39 obligation de résultat, 5 Saxon Mirror, 131 workplace accidents, see workplace accidents terminology, importance of: characteristics of French civil liability: small number of concepts, 447 no barriers against claims, 447–48 civil liability, 445–46, 453 French law, 446–53 concepts: action/fait, 447 actions giving rise to liability/faits générateurs, 449, 450–51 causation/cause, 447, 449, 450–51 fault/faute, 447 harm/dommage, 447, 449 keeper/gardien, 447 harm/dommage concept, 447, 449 dommage and préjudice, 449–50, 451 patrimonial and non-patrimonial, 449 law generally, 445 modifiying terminology, difficulties: intended targets of new rules, 452 legislators and users of new rules, 452–53 path-dependency, 452 Projet de réforme clarifications, 451–52 third parties, 38 assignment of rights, 55 ‘complete’ third parties, 56 contractual non-performance and third party claims, see contractual non-performance and third party claims groups of contracts, 55–56 third-party payers: allocation of tort damages, 196–97 insurance companies, 195–96 Projet de réforme, 198–200 social security, 194–95 transfer of accessory rights, 55 third-party insurance, 172–73 determinant of liability, as a, 173–75

English law: home/household insurance, 176–77 French law: parental liability, 177 German law: private/personal third-party liability insurance, 175–76 parental liability, 174, 175–76 separation principle, 173–75 determinant of liability, as a, 173–75 tort law: attribution, 105 capacity, 105–6 causation, 107 criminal law, interaction with, 100–5 institutional interaction, 100 legal reasoning, 100–1 parallel proceedings, 101, 111 prior convictions, 101–2 priority of civil/criminal remedy, 104 procedural advantages, 102 procedural connections, 101 range of remedies, 104 remedies, 101–5 substantive law, 101 victim compensation, 102–4 criminal law, integration: Dutch law, 116–17 English law, 115 French law, 114 general integrative techniques, 114–17 German law, 114–15 limitation of liability, 114 Spanish law, 115 unlawful acts, 113–14, 116–17 fault, 107 French tort law: distinctive characteristics, 99, 119 integration of duties and fault, 119 harm/damage/loss, 108 nuisance: property law v tort law, 237–38 unlawfulness/wrongfulness, 106–7 see also extra-contractual liability tortious liability, see extra-contractual liability trespass to land, 32, 373 troubles de voisinage, 224–25 nuisance compared: locality of wrong, 226

520  Index secondary importance of public law, 226 strict liability torts, 225 types of damage covered, 226 types of interference covered, 225–26 nuisance contrasted: exposure to liability, 226 personal injury claims, 227 protection of property rights v limitation of property rights, 226 scope, differences in, 227 types of interference, 227–29 use of land, 228 who can be sued, 226–27 who can sue, 226–27 privacy and views, 228–29 Projet de réforme articles, 482–83 unfair competition, 49–50, 313, 321, 345, 463 certainty of loss, 191 consequential loss, 218–19 injunctive relief, 379 unification of contractual and extra-contractual civil liability regimes, 466–68 United States: punitive damages, 329, 330–31, 340, 350 English law compared, 342 private retribution, 334–36 revenge, 334–36 social policy, 331–34 see also punitive damages unity, faits justiticatifs, 291, 293 unity (or duality) of civil liability, 9, 16–19, 26, 36 unity of criminal and civil faults, 99, 102, 104, 109–13, 118–19, 125, 295 unjust enrichment: civil liability distinguished, 358–59 loss/impoverishment, 360–66 préjudice and dommage, 359 different interpretations of, 353–54 enrichissement injustifié, 355, 356–57 cohabitational property disputes, 357 indirect enrichment, 357 service cases, 357 enrichissement sans cause, 354–55

fault, role of: claimant’s fault, 367–69 defendant’s fault, 369–71 full reparation principle, 201–2 French legal framework: Code civil, 354–55 quasi-contracts, 354–55 liability for deeds of things, 358 liability for fault, 358 loss/impoverishment, role of, 358, 361 action for enrichissement injustifié, 363–66 diverse approaches, 360–61 English notion of loss, 362 French notion of loss, 362–63 Projet de réforme, 353 quasi-contracts, 353 negotiorum gestio, 354–55, 357 recovery of undue payments, 354, 357 restitution, grounds for, 355–56 unlawful benefits, recovery of, 371–74 unlawful benefits: recovery of, 371–74 unlawfulness: administrative law, 431 cessation of unlawful action, see cessation of unlawful action civil wrongs, 52, 84, 87, 94, 96, 106–8, 113–17, 119, 131, 166, 288–89, 291, 412, 419–20 criminal law, 107 German law, 166–68, 289 harm, unlawful, 207–8, 210, 212–18 Italian law: danno inguisto, 207, 212–14 vicarious liability, 143 attribution, 105 employers’ liability, 162–63, 165 exemplary damages, 339 insurance regime, 174 joint and several liability, 271 origins, 148–50 Projet de réforme, 433–34 secondary liability, as, 163 strict vicarious liability, 154, 175 see also fait d’autrui; liability for the action of others victim compensation: English law, 103–4 French law, 102–3

Index  521 victim’s fault, see contributory fault in claimant/victim volenti non fit injuria, 52, 304–5 see also consent workplace accidents, 4–5 French law, 132–34 German law compared, 134–35 origins of liability law, 132–35 Prussian law, 134

wrongful birth: French law, 215–16 non-patrimonial losses, 216 Italian law, 216 unlawfulness requirement, 216 wrongful death: French law, 217 Italian law, 216–17 see also fatal accidents wrongfulness, 106–7, 109 see also unlawfulness

522