Insurance and Expanding Systemic Risks (Policy Issues in Insurance, No. 5)

« Policy Issues in Insurance Insurance and Expanding Systemic Risks This volume is the fifth in a series devoted to ma...

Author: Michael Faure | T. Hartlief

15 downloads 666 Views 1MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

«

Policy Issues in Insurance

Insurance and Expanding Systemic Risks This volume is the fifth in a series devoted to major policy issues in insurance. It contains an in-depth analysis of the assessment, management and compensation of the so-called “expanding systemic risks”, to which market players and insurers are exposed.


This comprehensive study responds to the growing concerns of economic, financial, political and social actors regarding the ever increasing exposure to new expanding risks. These risks are particularly related to natural disaster/environment pollution, technology, health and terrorism. For insurers the difficulty is encountered in adequately appraising and covering the potential liability stemming from these risks. It also sketches out some policy recommendations for decision makers in governments and in the business community on how to limit, prevent and manage such risks. In this perspective it will constitute a unique reference work for the attention of both OECD countries and emerging economies.

This book is available to subscribers to the following SourceOECD theme: Finance and Investment/Insurance and Pensions Ask your librarian for more details of how to access OECD books on line, or write to us at

[email protected]

No. 5

Insurance and Expanding Systemic Risks

OECD's books, periodicals and statistical databases are now available via www.SourceOECD.org, our online library.

Insurance and Expanding Systemic Risks

www.oecd.org

-:HSTCQE=VUW]^\:

ISBN 92-64-10289-2 21 2003 04 1 P

No. 5

© OECD, 2003. © Software: 1987-1996, Acrobat is a trademark of ADOBE. All rights reserved. OECD grants you the right to use one copy of this Program for your personal use only. Unauthorised reproduction, lending, hiring, transmission or distribution of any data or software is prohibited. You must treat the Program and associated materials and any elements thereof like any other copyrighted material. All requests should be made to: Head of Publications Service, OECD Publications Service, 2, rue André-Pascal, 75775 Paris Cedex 16, France.


Insurance and Expanding Systemic Risks No. 5

Professor Michael G. FAURE and Professor Ton HARTLIEF

ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT

ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT

Pursuant to Article 1 of the Convention signed in Paris on 14th December 1960, and which came into force on 30th September 1961, the Organisation for Economic Co-operation and Development (OECD) shall promote policies designed: – to achieve the highest sustainable economic growth and employment and a rising standard of living in member countries, while maintaining financial stability, and thus to contribute to the development of the world economy; – to contribute to sound economic expansion in member as well as non-member countries in the process of economic development; and – to contribute to the expansion of world trade on a multilateral, non-discriminatory basis in accordance with international obligations. The original member countries of the OECD are Austria, Belgium, Canada, Denmark, France, Germany, Greece, Iceland, Ireland, Italy, Luxembourg, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, Turkey, the United Kingdom and the United States. The following countries became members subsequently through accession at the dates indicated hereafter: Japan (28th April 1964), Finland (28th January 1969), Australia (7th June 1971), New Zealand (29th May 1973), Mexico (18th May 1994), the Czech Republic (21st December 1995), Hungary (7th May 1996), Poland (22nd November 1996), Korea (12th December 1996) and the Slovak Republic (14th December 2000). The Commission of the European Communities takes part in the work of the OECD (Article 13 of the OECD Convention).

Publié en français sous le titre : Aspects fondamentaux des assurances

© OECD 2003 Permission to reproduce a portion of this work for non-commercial purposes or classroom use should be obtained through the Centre français d’exploitation du droit de copie (CFC), 20, rue des Grands-Augustins, 75006 Paris, France, tel. (33-1) 44 07 47 70, fax (33-1) 46 34 67 19, for every country except the United States. In the United States permission should be obtained through the Copyright Clearance Center, Customer Service, (508)750-8400, 222 Rosewood Drive, Danvers, MA 01923 USA, or CCC Online: www.copyright.com. All other applications for permission to reproduce or translate all or part of this book should be made to OECD Publications, 2, rue André-Pascal, 75775 Paris Cedex 16, France.

FOREWORD

Foreword

T

his fifth volume of the series “Policy Issues in Insurance” is devoted to an in-depth analysis of the impact of expanding systemic risks on enterprises, insurance companies and governments. In this report, systemic risks refer mainly to risks arising from natural disasters/environment pollution, technology related risks, health-related risks and terrorism risks.

Difficulties encountered in assessing, managing and covering emerging systemic risks and the liability they may entail for enterprises raises the question of their very insurability. This study emphasises the importance of current evolutions of legal systems – and typically liability law – as a key determinant of the ability of the various economic actors to manage systemic risks. It also reviews the various remedies for dealing with expanding liability for systemic risks and presents the main alternative compensation mechanisms and new financing techniques that may complement traditional insurance schemes. Finally, the report adresses the potential role of governments in the management of emerging systemic risks through safety regulation, social security, compensation funds or introduction of insurance compulsion for instance. Authors conclude with a series of recommendations addressed to policy makers in government and in the business community in order to improve the prevention of and protection against expanding systemic risks. This study, updated as of September 2002, is part of a broader OECD project on emerging systemic risks. This report was elaborated by Professors Michael Faure and Ton Hartlieff from Maastricht University. The views expressed here are the sole responsibility of the authors and do not necessarily reflect those of the Insurance Committee, the Secretariat or OECD member countries.

INSURANCE AND EXPANDING SYSTEMIC RISKS No. 5 – ISBN 92-64-10289-2 – © OECD 2003

3

TABLE OF CONTENTS

Table of Contents Chapter 1. Introduction ..................................................................................

9

A. Background of the study........................................................................... 1. Newly emerging systemic risks ......................................................... 2. Implications for liability and insurability......................................... 3. Consequences for insurers and government................................... 4. Scope and limits of the study ............................................................ 5. Methodology ......................................................................................... B. Set-up of the study .................................................................................... 1. Expanding liability for systemic risks? ............................................. 2. Insurability of systemic risks ............................................................. 3. Influence of liability law on insurability .......................................... 4. Alternative compensation mechanisms .......................................... 5. Role of government ............................................................................. 6. Recommendations............................................................................... C. Acknowledgments ..................................................................................... Notes..................................................................................................................

10 10 11 11 12 13 14 14 15 16 16 17 19 19 19

Chapter 2. Expanding Liability for Systemic Risks? .................................

21

A. Developments in liability law .................................................................. 1. Introduction.......................................................................................... 2. The international level........................................................................ 3. National law ......................................................................................... 4. Summary............................................................................................... B. Negligence or strict liability for systemic risks?.................................... 1. The model ............................................................................................. 2. Negligence............................................................................................. 3. Strict liability ........................................................................................ 4. Legal justifications for strict liability ................................................ 5. Strict liability for systemic risks? ...................................................... 6. A few refinements ............................................................................... C. Developments in social security.............................................................. 1. Introduction.......................................................................................... 2. Different principles.............................................................................. 3. Changes in the relationship between tort and social security......

22 22 23 27 34 35 35 36 38 40 41 42 44 44 44 49


5

TABLE OF CONTENTS

D. Towards a “claim culture”? ...................................................................... 1. Introduction.......................................................................................... 2. Victims .................................................................................................. 3. The Role of the Bar .............................................................................. 4. The Judiciary......................................................................................... 5. Concerns regarding the “willingness to claim” ............................... Notes .................................................................................................................

6

56 56 56 57 59 60 62

Chapter 3. Insurability of Systemic Risks ...................................................

81

A. Introduction................................................................................................ B. Basic principles of insurance ................................................................... 1. Risk aversion and insurance .............................................................. 2. Predictability of systemic risks as the corner stone ....................... 3. Remedies for “insurer ambiguity”? ................................................... 4. Co-operation between insurers versus competition policy........... C. Capacity....................................................................................................... 1. General .................................................................................................. 2. Pooling by insurers: the nuclear example ........................................ 3. Limited capacity and public policy towards insurance.................. 4. Financial limits on liability? ............................................................... D. Moral hazard .............................................................................................. 1. What is moral hazard?........................................................................ 2. Remedies............................................................................................... E. Adverse selection....................................................................................... 1. Avoiding the “market for lemons” .................................................... 2. Danger of “correlated risks” ............................................................... F. The remedy: risk differentiation ............................................................. 1. Risk differentiation: theory ................................................................ 2. Risk differentiation: the example of environmental risk............... 3. The need for specialisation ................................................................ G. Summary..................................................................................................... Notes..................................................................................................................

82 83 83 84 85 87 88 88 90 94 97 107 107 107 109 109 109 110 110 111 112 112 115

Chapter 4. Remedies for Dealing with Expanding Liability for Systemic Risks .......................................................................

121

A. Introduction................................................................................................ B. Foreseeable risks........................................................................................ 1. Principle: insurability .......................................................................... 2. Example: European Products Liability Directive ............................. C. Retrospective liability................................................................................ D. Shifting the risk of causal uncertainty ................................................... E. Joint and several liability, and channelling of liability......................... 1. Deterrence ............................................................................................ 2. Insurance ..............................................................................................

122 122 122 123 124 125 126 126 127


TABLE OF CONTENTS

F. Channelling of liability.............................................................................. 1. Deterrence ............................................................................................ 2. To increase insurability?..................................................................... G. Other remedies for insurers against expanding liability ..................... 1. Cover over time .................................................................................... 2. Policy conditions as a remedy for expanding liability ................... H. Summary..................................................................................................... Notes..................................................................................................................

127 127 128 129 129 132 134 136

Chapter 5. Alternative Compensation Mechanisms and New Financing Techniques ...............................................

141

A. Introduction................................................................................................ B. The use of capital markets to cover risks: a brief introduction .......... C. Self-insurance and captives ..................................................................... 1. Reserves or insurance? ....................................................................... 2. Captives................................................................................................. 3. Reserves as guarantees? ..................................................................... 4. Self-insurance as a means of controlling moral hazard ................ 5. Limits and possibilities of self-insurance ........................................ D. First-party and direct insurance schemes ............................................. 1. Introduction.......................................................................................... 2. First-party versus third-party insurance: theoretical differences..... 3. Environmental damage insurance in practice: the Dutch example.. 4. Summary............................................................................................... E. Risk-sharing agreements .......................................................................... 1. Risk-sharing by operators: principles ............................................... 2. Practicability of pools: the nuclear example.................................... 3. Risk-sharing agreements: examples ................................................. 4. Risk-sharing versus insurance for systemic risks........................... F. Ex ante guarantees and deposits.............................................................. 1. Obligation of financial security: a balanced approach ................... 2. Optional financial caps ....................................................................... 3. Deposits................................................................................................. 4. A flexible solution................................................................................ G. Summary..................................................................................................... Notes..................................................................................................................

142 143 144 144 145 145 146 146 147 147 148 154 161 162 162 164 165 167 168 168 169 171 171 172 175

Chapter 6. The Role of Government.............................................................

181

A. Introduction................................................................................................ B. Safety regulation........................................................................................ 1. Criteria for safety regulation .............................................................. 2. The need to regulate systemic risks ................................................. 3. Safety regulation in practice: the example of environmental risks ..

182 183 183 186 187


7

TABLE OF CONTENTS

8

4. The need to combine liability and regulation ................................. 5. Justificative effect of regulatory compliance ................................... C. Social security ............................................................................................ 1. Different starting points ..................................................................... 2. Moral hazard ........................................................................................ 3. Health care insurance via the private market or government provision?.............................................................................................. 4. Summary............................................................................................... D. Compensation funds ................................................................................. 1. Introduction.......................................................................................... 2. The various funds ................................................................................ 3. General principles of fair and efficient compensation................... 4. Funds versus insurance ...................................................................... 5. A compensation fund for damage caused by systemic risks? ...... 6. Financing of historical damage ......................................................... 7. Summary............................................................................................... E. Compulsory insurance? ............................................................................ 1. Introduction.......................................................................................... 2. Economic arguments........................................................................... 3. Potential dangers of compulsory insurance .................................... 4. Further warnings ................................................................................. 5. Policy recommendation ...................................................................... F. Facilitative strategies ................................................................................ 1. Compulsory direct insurance or first-party insurance?................. 2. Necessity of a competition policy ..................................................... 3. Information policy ............................................................................... Notes..................................................................................................................

188 189 191 191 193 196 199 201 201 201 204 205 206 208 209 211 211 211 215 216 218 220 220 222 225 226

Chapter 7. Recommendations .......................................................................

235

Bibliography......................................................................................................

239

List of Abbreviations........................................................................................

259


ISBN 92-64-10289-2 Policy Issues in Insurance Insurance and Expanding Systemic Risks: No. 5 © OECD 2003

Chapter 1

Introduction

INSURANCE AND EXPANDING SYSTEMS RISKS No. 5 – ISBN 92-64-10289-2 – © OECD 2003

9

1.

INTRODUCTION

A. Background of the study 1. Newly emerging systemic risks It is well known that many new risks are emerging as a result of various causes. Technological progress may have led to substantial benefits and gains for society but an unavoidable side-effect has been that substantial health risks have arisen and that serious environmental pollution is threatening the global commons. By the latter, we refer to global warming and the depletion of the ozone layer. Some argue that there is a direct causal link between human activities and climate change. As a result of these well-known trends, which have been extensively described in the literature, many have argued that humanity is today increasingly exposed to a whole new type of risks. Sociologists such as Ulrich Beck have studied this phenomenon closely and have even described our society as a “risk society”.1 Mankind, they argue, is more and more exposed to risks; citizens seemingly are no longer involved in decisions concerning the acceptance of those risks. Although this evolution has been studied at many levels and has prompted many policy and legislative responses (e.g. the increasing importance given to the so-called precautionary principle),2 it has apparently not come to an end yet. In the summer of 2001, after the conference of the parties in Bonn reached an agreement saving the Kyoto protocol3 and the corresponding climate change regime, there was initially great optimism concerning the possibility that a global agreement could be reached and implemented to reduce the threat of climate change. However, merely six weeks after the agreement was reached in Bonn, the world was shocked by yet another event in the US which may change the course of history. It is obvious that it is not within the scope of this study to even start thinking about a possible solution for dealing with all of these emerging risks. However, they have a number of consequences for industrial enterprises, which merit specific attention. In particular, the role that the legal system can play (positively or negatively) with respect to these seemingly ever-expanding risks will be examined. Undoubtedly, the emergence of these risks will have serious consequences for industrial enterprises. Just thinking about the events in the US one can imagine the catastrophic financial consequences for the business community. Again, it is obviously not possible within the scope of this study to examine how the law could help to prevent all of these risks.

10


1. INTRODUCTION

2. Implications for liability and insurability There is, however, one specific aspect of these emerging risks that merits closer attention: the implications of these increasing and expanding risks for the liability of enterprises.4 For example, new health risks (such as electromagnetic fields or repetitive strain injury as a result of computer use) may lead to increased liabilities for enterprises. The same is obviously true not only of health risks but for other risks as well, such as the consequences of natural disasters, technologyrelated risks and environmental pollution. Regarding increased corporate liability for such risks, the question of their insurability inevitably arises. 5 Indeed, one can imagine that many catastrophes are of such a magnitude that they can hardly be considered insurable, at least in the long run. For this reason, the insurance and business community have developed new mechanisms for generating the financial resources which may be needed if such high-risk events materialise.

3. Consequences for insurers and government These developments, e.g. towards an increasing use of capital markets and alternative risk transfer mechanisms, have to some extent taken place on a voluntary basis, this is to say without formal intervention or regulation by government. However, one can imagine that the increasing emergence of risks obviously has consequences not only for the business and insurance world but may also shed a different light on the role of government. On the one hand, the emergence of new risks will always lead to calls for regulatory intervention by the government with a view to preventing them;6 on the other hand, after a catastrophe has occurred, victims may well turn to the government for compensation. One has to realise that the ability of the liability system to provide coverage for specific risks depends on the extent to which a liable, solvent party can be identified in the event of an accident. Even then, one might ask whether there should not be a specific role for the government, either in facilitating the financing of new risks or in mandating the provision of financial coverage, e.g. by introducing compulsory insurance.7 But again, in addition to such intervention, one should realise that many risks and catastrophes (especially national disasters but also terrorist attacks) may well be caused by unknown sources, or at least by sources that cannot be held liable or are insolvent. 8 The question then arises as to whether, for such cases, the government should intervene, not only with precautionary measures to prevent the risks but also ex post with funding to compensate the victims. It may be noted that, in addition to the social security already provided by most governments in Western Europe, some legal systems tend to introduce compensation funds to ensure that victims are compensated.


11

1.

INTRODUCTION

4. Scope and limits of the study Within the scope of this project and within the limits sketched out above, some of these issues will be examined. As far as the scope of the project is concerned it should be clear that it is not possible in any way to discuss, describe or analyse the nature of newly emerging systemic risks. These are also too diversified and complex to be described in one project in an adequate manner.9 For the same reason it is not possible either to analyse which precautionary measures could be taken to prevent those risks. What can be analysed, however, is what the role could be of traditional legal institutions in the prevention and compensation of the consequences of those risks. Therefore attention will be given to liability law, and the question will be asked whether a trend towards expanding liability in law is noticeable as a result of the emerging risks. This trend towards expanding liability undoubtedly has consequences for the business world. Therefore the question has to be asked in what way the legal system can expect the business world to cover the consequences of increased liability. In this study we will mainly address the insurance issues related to the emergence of systemic risks, as defined in the OECD study on emerging systemic risks. In this study, these are defined as risks that occur as a result of current or future functioning of major systems. These major systems lead to complex interactions, which therefore lead to increasing risks. Examples that are given in the OECD studies relate to the provision of health services, transport, energy, food and water supplies. We will therefore focus on the consequences of natural disasters, technology-related risks and healthrelated risks. The recent events in the US highlight moreover the importance of terrorism-related risks, so these will be addressed as well. Obviously the reader should be aware of the fact that this is a preliminary study providing merely an overview of the possible consequences of emerging systemic risks for insurers and the possible reactions of insurers. The responses to these increasing systemic risks have a lot in common; however, we will give examples from different areas to show that there are differences as well. The traditional system providing compensation for the consequences of risks is obviously insurance. The question therefore arises as to what the consequences of newly emerging risks and expanding liability for the insurance system are. In this, respect one can also ask whether alternatives to traditional liability insurance exist that might better be able to provide compensation. Finally, within the overall framework of liability and insurance, one may ask what the appropriate role of government should be, and whether government should merely provide regulation that facilitates the provision of financial coverage by private markets or whether more regulatory-type intervention is necessary.10

12


1. INTRODUCTION

Some of these issues will be highlighted within the scope of this project. However, it will be clear that even when one focuses just on liability, insurance and the role of government in the emergence of new risks, the topic is that broad that it might warrant various sub-studies. Within the scope of this study, it is possible only to highlight several issues that will play a role in liability, insurance and the corresponding role of government as a result of emerging new risks.

5. Methodology This study will have several thrusts. First, as a general background economic analysis of law to law will be used.11 This seems to be a highly important and interesting approach which has been developed in the United States and is now well-known in Europe as well. The economic literature analysing the role of law has the advantage that it deals specifically with the way the law should respond to newly emerging risks and the role of liability and insurance in that respect. Moreover, the law and economic literature have always stressed, which is highly important in the context of this project, that any system of financing cover for risks should try to achieve two goals, on the one hand prevention and on the other hand compensation12. Indeed, some seem to take an excessively ex post approach to risk, merely stressing the need for compensation, whereas economists stress that the primary role of law should be accident prevention. Moreover, the influence of compensatory measures (either insurance or compensation funds) on the prevention of those risks has to be examined. In addition to the economic analysis of law, a legal analysis will also be provided. Within the framework of this study, it is obviously not possible to provide detailed information on developments in liability or insurance law in many countries. Some information in this respect may be provided on the basis of a questionnaire. However, examples will be given of the way the law deals with new risks in various countries, to illustrate particular points made in the study. A central issue in this study will be the question of how emerging systemic risks can be prevented by using legal rules. Therefore the role of preventive measures and more particularly the preventive effect of both the liability and the insurance system will be stressed. It will be argued, moreover, that no matter what compensation scheme is chosen, this may never lead to an increase in the accident risk. Hence, the idea that “prevention is better than cure”, will be a guiding principle when analysing all of the legal and insurance mechanisms in this study. Only when prevention has failed – and thus a risk has materialised – will the question be asked as to what kind of financial mechanisms should be used to cover the consequences i.e. the compensation, of the damage. With this as a background, the following issues will be examined in more detail.


13

1.

INTRODUCTION

B. Set-up of the study Within the limits sketched above, it is clearly not possible to address all the possible aspects of liability, insurance and financing of new systemic risks. Within the scope of this study, the following aspects will be addressed more closely:

1. Expanding liability for systemic risks? 1.1. Developments in liability law First of all, the factors influencing the expanding liability of enterprises for health risks will be examined. There have been many changes in tort law with respect to the scope of liability. We shall briefly address the shift towards expanding liability that has taken place at the international level (conventions, Council of Europe and European Union) as well as in some national laws. In various national laws, expanding enterprise liability both in legislation and case law may be noted. In case law, one can note an increasing shift from negligence towards strict liability. There are, moreover, several trends in case law or legislation, which may severely increase the liability of enterprises. They concern, on the one hand, the notion that enterprises can be held liable for behaviour that was not considered wrongful at the time when it was committed. Only through changes in case law or regulatory standards will specific behaviour in the past be considered wrongful today. This is a so-called retroactive application of standards. Another, potentially dangerous trend in liability law is the shift of the risk of causal uncertainty to enterprises. There are various situations, for instance in the areas of product liability and employer’s liability, where uncertainty often exists concerning the causal relationship between certain behaviour by an enterprise and specific health damage. If the risk of causal uncertainty is de facto shifted to the enterprise (e.g. through a reversal of the burden of proof or through a joint and several liability rule), this may seriously increase corporate exposure to liability. The general trend towards expanding liability will also be critically analysed by means of an economic analysis of tort law. Indeed, it is particularly in the economic analysis of tort law that attention has been paid to the specific functions of deterrence and compensation in tort law. These economic tools will therefore be used to review critically some of the aforementioned trends.

1.2. Developments in social security The expanding exposure to liability is not only determined by developments in liability law itself, but also by developments in social security law. Indeed, a basic assumption for a long time was that victims in Western Europe, unlike their American counterparts, did not have many incentives to bring a liability suit, since the social security system in many Western European countries provided

14


1. INTRODUCTION

for relatively wide coverage of many expenses that a victim incurs when an accident happens. Partly as a result of international conventions and European directives, many countries have put in place systems of compulsory first-party insurance covering medical expenses and lost income. Therefore, if one wishes to analyse the scope of liability in a particular legal system, one should also take into account the extent to which damage is already compensated for via other (private or public) compensation mechanisms which remove the need to use the tort system. In addition, if one wishes to analyse changes in the scope of liability, one should also address possible changes in social security law. The withdrawal of government from the social security arena may well increase enterprises’ liability exposure.13

1.3. Towards a “claim culture”? Finally, exposure to liability is influenced by a notion which is less tangible but which is sometimes referred to as the “willingness to claim” or the “claim culture”. Indeed, the question whether in a particular case a law suit will be brought will to a large extent depend on the role of the actors in the legal system, whether they are the victims or consumers themselves in the first place, the providers of legal services – mostly attorneys, or the judiciary. The attitude of these groups may to some extent influence the probability that the tort system will be used.14

2. Insurability of systemic risks In a later chapter, the insurability of liability risks will be analysed. Here attention will be given to the influence of increasing liability as a result of expanding health risks for the insurability of the liability of enterprises. Indeed, in many countries, enterprises are now held liable for so-called new systemic risks, which were largely unknown, say, ten years ago. The question obviously arises as to how this affects insurability and, more generally, how the insurability of systemic risks can be defined. Attention will thus be given to the importance of the predictability of risks but also to the importance of sufficient capacity for insurance providers. The latter point is obviously important given the increasing scale and concentration of emerging systemic risks. This may have an important effect for the insurance and reinsurance industry. Moreover, when addressing insurability, attention also has to be given to the potential remedies for uninsurability. In many countries insurers are now well aware of the fact that liability for systemic risks is expanding. As a result, they are taking all kind of measures to limit their exposure to liability. Some of these remedies, such as increased risk differentiation but also changes in insurance cover and the limitation of coverage, will be studied in further detail.15


15

1.

INTRODUCTION

3. Influence of liability law on insurability The next chapter addresses how the legal system can affect insurability. After the developments in liability have been sketched, they are then reviewed critically from an insurance angle. The question arises whether some of the developments in liability law towards increasing liability for systemic risks are such that it is to be feared that they will negatively affect insurability. For example, will the shift from negligence towards strict liability lead to insurance problems? The question can also be asked whether the issues mentioned above, namely retroactive liability and the shifting of the risk of causal uncertainty, will have an influence on insurability.

4. Alternative compensation mechanisms Next, the question that inevitably has to be addressed is whether useful alternatives could be worked out for liability law and liability insurance to cover increasing health risks. If it has been noted that to some extent liability risks are hardly insurable, the specific reasons for these insurance problems have to be identified and the question has to be asked whether alternative ways could be found, both within and outside the insurance business, of providing financial coverage for these risks on another basis. Insurance is basically a system of risk transfer and risk-spreading but there may be other mechanisms that can achieve the same goal. This is, by the way, obviously not a theoretical question either. Indeed, one can notice that both insurers and enterprises have not only tried to protect themselves against expanding liability by limiting their exposure but they have also looked for alternative mechanisms to cover the financial consequences of new health risks. This obviously is in the interests of enterprises. If, as a consequence of expanding liability, insurers limit the cover on liability insurance, this would mean that enterprises would increasingly face those risks themselves. Therefore, one can understand that they have been looking actively for alternatives.

4.1. Capital markets One alternative that has been developed especially in the United States, but is still less developed in Europe, is the use of capital markets. Economists have long predicted that capital markets may well be able to provide financial coverage for catastrophic risks. The question arises as to how it would precisely work, what the benefits could be compared to insurance, and what the practical consequences would be of such a system.

4.2. Self-insurance In addition to using the capital markets to cover risks, some enterprises have looked for other alternatives in the simple form of self-insurance, reserves or captives. The basic principle of all of these alternatives, obviously with different modifications and conditions, is that companies, either alone or jointly

16


1. INTRODUCTION

with other companies facing similar risks, set aside a specific sum to cover future risks. These systems of self-insurance and so-called captives thus also have to be analysed. A critical review has to be made of them, both from the perspective of the enterprises themselves (there may be advantages from the insurance standpoint, but also risks) as well as from a policy perspective (what is the guarantee that the money will actually be available when the risk materialises).

4.3. First-party or direct insurance There is a major trend in the insurance world today to move to a different system of insurance to cover all kind of new risks, namely, a shift from third-party to first-party or direct insurance. Insurers have long claimed that third-party liability insurance has a lot of disadvantages as far as predictability and hence insurability is concerned. Risk differentiation with third-party liability insurance is difficult and it is often hard to predict under what kind of circumstance a judge will hold a particular insured injurer liable. Insurers have therefore moved in many areas away from liability insurance and prefer to cover similar risks on a first-party basis. A clear trend in this respect can be found in many countries in the area of occupational diseases and accidents at work. Before, many of these risks were covered by liability cover, linked to the employers’ liability. Now many countries have abrogated employers’ liability (except for cases of gross negligence or intent) and have set up financial schemes (often in the form of social security) which provide cover for all the employees of a particular employer. Today, only in the Netherlands and the United Kingdom are occupational diseases and work accidents still financed via employers’ liability. But one can also find this shift from liability towards first-party or direct insurance in areas like medical malpractice (a move towards patient insurance) and environmental liability (a move towards environmental damage insurance). We will therefore analyse what the specific benefits may be of first-party/direct insurance schemes and whether they can be considered better equipped to cope with the newly emerging systemic risks. In general, it seems important to examine the possibilities of first-party coverage for all these risks where no liable injurer can be found (e.g. with natural disasters) or where the damage largely exceeds the possibilities of the liability insurance of the injurer (e.g. in case of a terrorist attack). In all these cases the question can be asked whether the potential victims can seek (firstparty) coverage themselves.

5. Role of government Finally, the role of government with respect to these newly emerging risks should be addressed. Is there any particular action that can or should be expected from a government? Here one can see a variety of roles that a government could play, both in the prevention of systemic risks as well as in the compensation sphere.


17

1.

INTRODUCTION

5.1. Safety regulation As far as prevention is concerned, the question obviously arises as to whether one can expect that liability law will give sufficient incentives to prevent health risks. There may be strong arguments in favour of regulation of certain activities or products that could cause health risks. Hence, the criteria for safety regulation should be examined as well as the question of how regulation can be combined with liability law.

5.2. Social security Secondly, as far as compensation is concerned, the question of the role of the government in social security needs to be addressed. In a number of European countries it may be seen that the government is increasingly withdrawing from social security, as a result of which victims have to rely more on tort law. Therefore, some basic differences between tort law and social security as far as the financing of health risks is concerned, have to be examined, and, more particularly, the difference between social security and private health care insurance.

5.3. Compensation funds Thirdly, the question can of course also be asked whether there should be a particular role for government in the compensation of those types of damage that occur from unknown sources and where no liable injurer can be found (as in the case of natural disasters).16 Many catastrophic accidents lead to serious damage and go far beyond the possibilities of liability law. In these cases, the question can be asked whether compensation funds can provide an adequate solution. Again one may see in many areas and many legal systems, the introduction of compensation funds (e.g. for asbestos victims), especially when traditional tort law fails to provide coverage. This could obviously also play a role when a very large number of victims is involved, such as in the case of a terrorist attack.

5.4. Compulsory insurance Fourthly, the question can of course also be asked whether there should be regulation obliging potential injurers and those who may cause systemic risks to secure financial coverage for those risks. This is the traditional question of compulsory insurance. For many areas, e.g. in the above-mentioned sphere of occupational health and work accidents, compulsory insurance exists. Therefore, the reasons for introducing compulsory insurance have to be examined carefully, and the question can be asked whether it may be warranted to extend compulsory insurance to other areas as well.

5.5. Facilitative strategies Finally, the question can be asked whether, even if one leaves the provision of financial coverage for systemic risks totally to the market, the government can take measures to facilitate the optimal provision of financial coverage for

18


1. INTRODUCTION

those risks. Information strategies could be envisaged to inform both potential victims and potential injurers of preventive measures that could be taken, of the availability and benefits of insurance, and of their exposure to risk. In addition to information strategies, other policies are possible. In this respect, one could mention the importance of an adequate competition policy for the optimal provision of insurance. The past has shown that if insurers conclude cartel agreements whereby they agree not to cover particular risks (this happened with respect to natural disasters such as flood risks), then the availability of insurance will inevitably be limited as a result of those agreements. Some attention should therefore be given as well to the importance of competition policy for the optimal provision of financial coverage for health risks.

6. Recommendations The study will conclude with some recommendations and a few final remarks.

C. Acknowledgments We have largely benefited from discussions with many of our colleagues in Belgium, the Netherlands and in other OECD countries concerning the liability for and insurance of systemic risks. There are too many to be enumerated here, but we highly appreciate all the useful comments we have received during our research. We more particularly benefited from discussions with the experts at the OECD, who provided us most useful comments on an earlier draft of this paper. We more particularly are grateful to Reza Lahidji, André Laboul and Cécile Vignial for useful comments and a pleasant cooperation. We are also grateful to Liselotte Vanhoppe of Maastricht University for highly useful research assistance. Moreover, we have benefited from highly useful practical information on the insurance of systemic risks in Europe. Without this practical information, we would have missed a lot as far as current developments are concerned. In this respect, we are particularly grateful to Peter Kamp (Nationale Nederlanden) and Jan Pieter Six (Interpolis) for keeping us informed on a regular basis on various recent developments relevant to our work. All remaining errors obviously remain ours. We have tried to update our study until 1 September 2002.

Notes 1. Beck, U., Risikogesellschaft. Auf dem Weg in eine andere Modeine, Frankfurt, 1986. 2. See O’Riordan, T., Cameron, J. and Jordan, A., (eds.), Reinterpreting the precautionary principle, London, Cameron, May, 2001.


19

1.

INTRODUCTION

3. See Grimeaud, D.J.E., “An overview of the policy and legal aspects of the international climate change regime”, Environmental Liability, 2001, 39-52. 4. See Faure, M. and Hartlief, T., “Towards an expanding enterprise liability in Europe”. How to analyse the scope of liability of industrial operators and their insurers, Maastricht Journal of European and Comparative Law, 1996, 235-270. 5. Faure, M., “The Limits to Insurability from a Law and Economics Perspective”, 1995, 454-462. 6. Compare Morris, J., (eds.), Rethinking Risk and the Precautionary Principle, Oxford, Butterworth, 2000. 7. See Jost, P.J., “Limited Liability and the requirement to purchase insurance”, International Preview of Law and Economics, 1996, 259-276. 8. Shavell, S., “The Judgment Proof Problem”, International Review of Law and Economics, 1986, 43-58. 9. This issue has, moreover, already been addressed in an OECD study on emerging systemic risks. 10. Compare Shavell, S., “Liability for Harm versus Regulation of Safety”, Journal of Legal Studies, 1984, 357-374. 11. For an overview see Cooter, R. and Ulen, Th., Law and Economics, 3rd ed., Addison Wesley, 2000. 12. Compare Shavell, S., Economic analysis of Accident Law, Cambridge, Harvard University Press, 1987. 13. Compare Faure, M., “The applicability of the principles of private insurance to social health care insurance, seen from a law and economics perspective”, The Geneva Papers on Risk and Insurance, 1998, 265-293. 14. See Faure, M. and Hartlief, T., “Het kabinet en de claimcultuur”, Nederlands Juristenblad, 1999, 2007-2015. 15. See Faure, M. and Hartlief, T., “Remedies for expanding liability”, Oxford Journal of Legal Studies, volume 8, 1998, 681-706. 16. See Faure, M. and Hartlief, T., “Compensation Funds versus Liability and Insurance for Remedying Environmental Damage”, Review of European Community and International Environmental Law, 1996, 321-326.

20



Chapter 2

Expanding Liability for Systemic Risks?


21

2.

EXPANDING LIABILITY FOR SYSTEMIC RISKS?

A. Developments in liability law 1. Introduction Although the main focus of this study is on the implications of newly emerging systemic risks for the insurability of those risks, the importance of developments in the legal system will also be underlined Indeed, one consequence of increasing systemic risks will be the fact that more accidents may obviously happen – accidents, moreover, with a higher damage potential. One (important) way that the consequences of these accidents can affect insurers is through the use of liability law. Indeed, the first question that will often arise when a group of persons is victimised as a result of a systemic risk, is whether a third party can be made liable for the consequences of an accident. The legal tool to force the third party to pay compensation to the victim is obviously liability law, also referred to (in some legal systems) as tort law. As a result of newly emerging risks, in many countries there has been a legal trend towards expanding liability. Indeed, in many legal systems legislators and judges have become increasingly sympathetic to the interests of victims and have therefore tried to accommodate them as much as possible so that they could get compensation for their losses via liability law. The tools to achieve this goal were given either by the legislator or by judges, in the latter case, by means of broad victim-friendly-interpretations in case law. Although this trend towards an expansion of liability law may to a large extent have accommodated the interests of victims, the other side of the medal is obviously that it has led to an increasing pressure on enterprises and therefore also on their insurers. It is particularly against the increasing use of liability law that insurers have been looking for remedies. In some legal systems, such as that in the US, scholars like Priest have argued that it has been the expansion of liability law, which has led to an insurance crisis.17 We will therefore focus on these developments in liability law to see whether they have had a negative impact on the insurability of risks. In the next section (Ch. 3) we will then address the insurability of systemic risks more broadly, and then come back to the developments in liability, considering in what way they negatively affect insurability (Ch. 4). The trend towards an expansion of legal liability, on which we will briefly focus in this part of the study, can be seen in many legal systems. Within the scope of this exploratory study it is obviously not possible to focus on developments in all the legal systems of OECD countries. We will therefore

22


2.


mainly select a few examples from Europe, the US and Canada and provide further references in the notes. In many countries one can find authors to support the argument that the scope of liability has expanded in recent years.18 This expansion of liability relates in general to enterprise liability, but also to the liability of professionals, such as physicians, accountants and lawyers.19 In this section we shall try to sketch some of the general developments that might have influenced this expansion. Since we have indicated that we are interested in addressing the consequences of the expansion in liability both for industrial operators and for their insurers, we shall also briefly address the consequences of changes in tort law for the insurability of certain risks. As far as insurability is concerned, one should keep in mind that the magnitude of the risk is determined by the probability (p) that an accident with damage (D) will occur.20 Changes in material tort law that affect the conditions for liability usually relate to (p). In many cases it will not be possible to indicate precisely what the financial consequences of a certain change in tort law will be. It will be hardly possible, for example, to ascertain whether a certain shift in case law will lead to a 10%, 25% or greater increase in the chances of an industrial operator being found liable. However, one can at least try to indicate whether an expansion in liability is due to changes in legislation or case law with respect to material tort law or to changes in the social security system. We shall now discuss some developments in material tort law that might affect the probability of an industrial operator being found liable, distinguishing between developments at the international level (B) and at the national level (C).

2. The international level 2.1. Conventions Several international conventions may be of importance for the scope of liability. For instance, with respect to civil liability for marine oil pollution, or the liability of nuclear power plant operators, conventions have adopted the principle of strict liability, but at the same time have limited its scope by introducing financial caps.21 Many of these conventions have a similar thrust: the operator is strictly liable, liability is channelled to one individual, short statutes of limitation apply and financial caps are introduced.22 These conventions usually only apply to a specific area such as oil pollution, nuclear accidents or civil aviation and, therefore, probably have little bearing on the general question addressed in this study of whether industrial operators face expanding liability.23 Many of the issues regulated by these international conventions could certainly be described as “systemic” risks. Moreover, it is not at all clear whether these conventions do in fact lead to an expansion of liability. Through the introduction of strict liability, it could be argued that the potential for being held liable once an accident has happened is obviously greater


23

2.


than under a negligence rule. However, the number of potential injurers against which a victim could bring a claim has been reduced through the channelling of liability.24 In addition, the amount at stake is usually seriously limited by the introduction of financial caps.25 Therefore, it can be argued that the scope of liability of a potential injurer under these conventions is usually more limited than under regular tort law.26 Moreover, the preparatory proceedings of some of these conventions also make it clear that their goal was not so much to increase protection for victims but to limit risks e.g. for a nuclear power plant operator.27 Hence, it can be argued that if the existing conventions do have any influence on the scope of liability, they tend to limit rather than expand it.28

2.2. Council of Europe As far as the scope of liability is concerned, we should, within the European context, certainly mention the Convention of Lugano on Civil Liability for Damage Resulting from Activities Dangerous to the Environment.29 In this convention the signatory parties agreed to strict liability for environmental damage. In fact, this reflects a trend in the legal systems of the various states towards the introduction of strict liability for environmental damage. Hence, the question arises as to whether it is fair to argue that this convention as such will lead to an expansion of the scope of liability. This would undoubtedly be true for those signatory states that had not adopted a strict liability rule for environmental damage in their legal system up to that point, either by statute or through developments in case law. In the case of the Netherlands, for example, the scope of liability was already relatively large as a result of Art. 6:175 of the Dutch Civil Code30 which came into effect on 1 February 1995.

2.3. European Union Originally very few initiatives were taken at the European level with respect to civil liability. However, given the initial goal of market integration of the European Economic Community, the Commission realised that it could use the fundamental harmonisation article, Article 100 of the original Treaty (before the Maastricht Treaty came into force), to harmonise legislation in the member States in order to equalise marketing conditions. Levelling the playing field was the message. The most noteworthy initiative with respect to civil liability was taken by the European Commission in the field of product liability as early as 1975.31 In the statement preceding the text of what ten years later would become the Product Liability Directive of 25 July 1985,32 one could read that market integration was not the only goal of the Directive. It also aimed at a higher level of consumer protection.33 Again, it could be asked whether the European Product Liability Directive constitutes an expansion of the scope of liability. Obviously this will again depend upon the state of liability law in the various European member States. For some member States which already had product liability regimes in place

24


2.


at the time of the promulgation of the Directive, the European initiative probably did not substantially enlarge the scope of liability for manufacturers. This was probably the case for Germany,34 but also for the Netherlands.35 For these member States, one can even argue that national case law and/or legislation provides for an even wider scope of liability, since national law does not have the restrictions (e.g. with respect to the statute of limitation, scope of damages etc.) that are included in the Directive.36 But for other member States, such as Spain and Greece,37 the Directive did probably substantially change the scope of liability of producers. In some legal systems, the effect might have been somewhat mixed. For instance, in Belgium, traditional case law had awarded a large degree of protection to victims forming part of a contractual chain with the producer (based on the seller’s responsibility to provide a guarantee against the consequences of defects of the products he sells),38 but for third parties not standing in a contractual relationship, classic tort rules applied. Even though some case law had argued that the fact of bringing a defective product on the market already constituted a tort, it seems fair to state that in Belgium the situation of third-party victims was improved (and the corresponding liability of the manufacturer expanded) with the coming into force of the Belgian Product Liability Act implementing the European Directive.39 Moreover, the European Product Liability Directive is also important, since it reflects a trend in European policy to opt for an increased consumer/victim protection and therefore also for an expansion of the scope of enterprise liability.40 Obviously it would be worthwhile investigating what the policy options of the European Commission are with respect to future initiatives concerning liability law. If these are, as some could argue with respect to the European Product Liability Directive, merely a codification of trends in national liability law, industrial operators should again not necessarily fear legislative initiatives coming out of Europe any more than legislative or jurisprudential developments in their own country. Moreover, the scope of civil liability has traditionally been a field that has been considered part and parcel of national legal culture and, therefore, not well suited to a total harmonisation.41 As a result, attempts to achieve at European level, a degree of consumer/victim/environmental protection which is higher than what has been customary within the member States, will inevitably have to cope with a lot of resistance.42 A perfect illustration constitutes the aforementioned European Product Liability Directive, which finally saw the light only after 9 years. Another illustration is the evolution of a draft of Directive concerning the liability of suppliers of services.43 This Directive on liability for services could indeed substantially expand the scope of liability. A first text opted for clear strict liability for damage caused by defectively supplied services. A later text weakened this a little, instituting a negligence rule with a reversal of the burden of proof. However, the Commission recently withdrew this draft Directive completely and is now considering promulgating specific


25

2.


rules for specific services (e.g. construction) instead of a generalised liability rule for all kinds of services.44 It generally seems fair to argue that whenever a European initiative is taken, it usually consists of a strict liability rule or a reversal of the burden of proof, and is accompanied by the argument that this is better suited to the needs of victim protection. Whether this always constitutes an expansion of the operator’s scope of liability will inevitably depend upon the existing liability rules in national law.45 In this respect, the evolution of the draft Directive concerning liability for services perfectly illustrates the fact that European initiatives that are too far-reaching will often not be acceptable to the member States. There are two more recent European documents concerning liability law that should be mentioned. The first is European Directive 90/270 of 29 May 1990 concerning the use of display screens (visual display units), which has to be implemented in national law by 1 January 1997.46 To some extent this Directive responded to the so-called `repetitive strain injury’ problem (RSI), a blanket term used to describe several health problems caused by repetition of a similar movement. In the UK, RSI has led to several successful liability claims.47 The Directive does not itself institute a liability regime, but forces employers to follow specific safety instructions that should prevent repetitive strain injury or other health problems connected with computer screens. Although one could question the efficiency and usefulness of this Directive at such an early stage, when little is known about the possible consequences,48 the regulatory devices provided for in the Directive will undoubtedly also affect liability law. If the safety precautions prescribed by the Directive have not been followed by an employer, a victim could claim that damage resulting from this omission has to be compensated for by the employer under the negligence rule.49 A second domain in which European action can be expected in the near future is environmental liability. In this respect, mention should be made of the green paper on remedying environmental damage that was launched in 199350 and at the White Paper presented in 2000.51 In these policy documents, the Commission discusses several options for regulation with respect to environmental liability. Once again, the Commission’s preference for a strict liability rule is clear. Given the policy options set out in the Green and White paper and the recently adopted Convention of Lugano, it seems likely that whenever another initiative is taken with respect to environmental liability, it will involve the adoption of a strict liability rule. This has meanwhile been done with the presentation of a proposal for a directive52. Again, we should repeat that such an initiative would not necessarily mean that the scope of liability will also be expanded. This will to a large extent depend upon the scope of national liability law.

26


2.


Finally we should mention important recent rulings by the European Court of Justice, more specifically Francovich53 and Brasserie du Pêcheur54 which will undoubtedly have an important impact on state liability in case of violations of Community law.55

3. National law 3.1. Legislation Much legislation has been passed in several European countries that has led to an expansion of the scope of liability of enterprises.56 Indeed, for several specific areas, legislators have opted for a shift from fault to strict liability. Some of these legislative trends were prompted by the international developments mentioned earlier. For example, the legislative move towards strict liability for damage caused by product defects was to a large extent initiated by the European Product Liability Directive, especially in countries that did not have a specific product liability regime at the time of the Directive’s introduction. But one can also point to other areas where either the burden of proof has been shifted or the fault concept has been broadened. This is the case in several countries with respect to environmental liability and related areas. Many legal systems have also introduced strict liability for environmental damage and the Merchant Shipping Act.57 One example is the German Water Pollution Act of 195758 and the subsequent Umwelthaftungsgesetz of 1990;59 another is the Danish Act on Strict Liability for Environmental Damage.60 One can also note a shift towards a form of strict liability in the related area of liability for hazardous substances and/or activities. Examples can be found in Art. 6:175 of the new Dutch Civil Code,61 in German law62 and in Belgium.63 What can be said about this trend as far as its influence on the insurability of certain risks is concerned?64 Again, it will largely depend upon the already existing legal regime in a specific country as to whether or not such a legislative move really constitutes an expansion of liability. If a clear negligence rule existed, a shift towards strict liability will undoubtedly constitute an expansion. If, on the other hand, national case law already interpreted the negligence rule relatively broadly, the shift towards a formal strict liability rule might not constitute such an important change in a firm’s exposure to liability. Secondly, as long as the new strict liability regime only applies to situations that occur after the new regime has come into force, the shift towards strict liability should not pose excessively serious problems from an insurance perspective. Indeed, as long as the new risk remains predictable the risk remains insurable.65 It may be noted that after the implementation of the new European Product Liability Directive, for instance, insurance policies adapted relatively smoothly to the new regime.66 Thirdly, one should not forget that enterprises providing services or producing goods usually stand in a contractual relationship with a customer.


27

2.


This means that the consequences of an expansion in liability can usually be passed on to consumers via the price mechanism.67 Moreover, it is very likely that in the near future the current trend towards introducing strict liability for specific areas and activities may be compounded by the introduction of generalised strict enterprise liability in either legislation or in case law.68 Even if this were the case it could still be argued that as long as the new legal regime only applies to activities that occur after the new regime came into force, this will allow an insurer to predict and calculate the new risk ex ante and to set premiums accordingly.69 Unfortunately, however, an insurer will often be unable to do so in the case of long-term insurance policies, which have not foreseen the possibility of a widening of the scope of liability.

3.2. Case Law 3.2.1. Shift from negligence to strict liability We have already discussed the shift from negligence towards strict liability that can be noticed in the legislation of some countries with respect to many activities involving industrial operators. This trend was often first visible in case law. In many areas the shift took place without the formal introduction of a strict liability rule, but instead simply by broadening the fault concept. In this respect, one can point to many developments. Firstly, in numerous countries, for example, the Netherlands, courts often seem to favour victims through broad interpretations of the negligence rule as soon as physical harm has been inflicted as a consequence of an industrial activity.70 Again, product liability law could serve as an example to illustrate this evolution. For instance, in Belgian tort law some courts ruled that a product has to be considered defective by virtue of the simple fact that it has been brought onto the market and apparently caused damage. Similar decisions could already be found before the European Product Liability Directive in the Netherlands and Germany.71 Also, in many cases where physical harm is inflicted, a "duty of care" will be interpreted extensively. In the recent Margereson and Hancock v. J.W. Roberts case, the High Court of Justice of Leeds District Registry held that a company was liable for negligence for personal injury caused by an asbestos factory to local residents as a result of its operations over sixty years.72 It was held that the company’s duty of care extended beyond its employees to the factory’s immediate surroundings. Secondly, in some cases the negligence rule continued to exist but a reversal of the burden of proof took place. This was the case in Belgian contract law with respect to product liability. According to a decision of the Cour de Cassation of 4 May 1939,73 the seller of a defective product is presumed to have known the defects of the products he sold.74 A later decision held that the seller could still prove otherwise, but only if he could show that it would have been totally impossible for him to have discovered the product defects;

28


2.


he has to show invincible ignorance.75 Whereas in Belgium the presumption can, therefore, still be rebutted, in France, on the contrary, case law has held that the presumption of knowledge of the defects for the professional seller was a presumption iure et de iure and therefore not reputable.76 Thirdly, one could also point to developments in case law concerning environmental damage. Especially as far as the liability for soil clean-up costs and other environmental damage is concerned, the foreseeability requirement has often been relaxed, which also substantially lowers the burden of proving negligence. This jurisprudential evolution has led to interesting debates, for example in the Netherlands.77 As early as 1988, Van Dunné claimed that liability for environmental damages had generally shifted from negligence to strict liability.78 He also claims that Dutch law now has a strict liability for soil pollution.79 Vranken on the contrary argues that it is merely a broadly defined negligence rule, but not strict liability.80 This debate illustrates that, even for specialists, it is no longer easy to distinguish between the negligence and the strict liability rule.81 Shifts towards strict liability for environmental damage can obviously also be found outside Dutch tort law – for example, the famous British Rylands v. Fletcher case where a strict liability rule for damage caused by burst reservoirs was instituted.82 This 19th century doctrine is now widely applied to many cases of environmental harm. Another example can be found in German case law where a reversal of the burden of proof was accepted in a 1985 environmental case.83 Fourthly, one can also note a broad interpretation of the negligence rule, especially as soon as physical harm is inflicted, through the fact that in many legal systems the violation of a specific regulatory safety standard automatically constitutes a fault.84 By raising the standard of care, courts can imperceptibly shift a negligence rule to strict liability.85 Finally, one could also illustrate the jurisprudential trend towards a widening of liability by reference to the domain of professional liability. Indeed, the liability of professionals, e.g. physicians, attorneys and architects as well as accountants, seems to be expanding in many legal systems as well.86 Especially liability for medical malpractice is of growing importance in the field of tort law.87 Remarkable developments in this context are the socalled wrongful birth and wrongful life cases. Already in 1997 the Dutch Hoge Raad held a doctor liable for the costs of the upbringing of a child, because he failed to replace a birth control devise.88 Although the Dutch lawyers are used to the wrongful birth claims by now, the wrongful life claims are still more debated. This issue has been extensively discussed in Dutch literature, which also takes into account evolutions in this respect in neighbouring countries.89 Also in France the discussion was fuelled by recent decisions of the Cour de


29

2.


Cassation in which the court reconfirmed his vision that a handicapped child can claim compensation if there is sufficient causal link between the handicap of the child and the fault of the defendant (the doctor/gynaecologist) which deprived the mother of the chance to choose for abortion.90 A key issue, which is discussed in many of these commentaries, is whether there should at all be any role for liability law in these particular cases. As far as the question of whether these jurisprudential developments constitute a broadening of the scope of liability is concerned, we can again refer to what was mentioned earlier. A shift from a negligence rule towards strict liability certainly expands the scope of liability. In many countries these jurisprudential developments have preceded legislative changes, which, in turn, have preceded international developments. The basis for changes could often be found in the jurisprudential trend to provide victim or consumer protection in individual cases. As far as insurability is concerned, it should be mentioned that a strict liability is also insurable as long as it only applies to new risks.91 There are, however, two threats to insurability that can also be noted in certain jurisprudential developments. 3.2.2. Retroactive liability The most serious threat posed by jurisprudential and legislative developments for industrial operators and their insurers is probably not the expansion of liability as such, whatever form it might take. Problems will however arise if retroactive liability is introduced. In that case, the specific enterprise will be held liable for behaviour that was not considered wrongful at the time when the act was committed. Only through changes in case law or regulatory standards will specific behaviour in the past be considered wrongful today. Such a retroactive application of new standards seems to go against one of the main principles of tort law, namely that the foresight of a liability ex post should give incentives ex ante for the prevention of accidents. Since the behaviour was not considered wrongful at the time, a retroactive application of new standards, either through case law or regulation, could therefore never affect incentives for future behaviour of the specific operator.92 Moreover, such retroactive liability may lead to uninsurability. The insurer was not aware of the risk at the time of the potentially wrongful behaviour of the insured. Since the risk was, therefore, not foreseeable at all, no premium could be charged for it, no specific preventive mechanisms could be required via the policy conditions and no reservations for losses could be made.93 The reason retroactive liability is nevertheless discussed often has to do with compensation. But even from a distributive point of view, retroactive liability could be criticised. Indeed, in many cases, the operators who caused, for example, soil pollution in the past, may be out of business by now. It is, therefore, questionable whether it is fair to shift the costs of soil clean-up

30


2.


operations to the unlucky operators who are still in business today but whose behaviour at the time was not even considered wrongful.94 Nevertheless, examples of retroactive liability can be found in the case law of several legal systems.95 Since such retroactive liability makes future risk calculation for operators and their insurers impossible, this jurisprudential trend seems highly problematic. Nevertheless, it cannot be excluded that the redistributive desire to provide victim protection may lead to an increased use of retroactive liability, especially in cases where victims are suffering physical harm. However, when environmental damage is concerned, e.g. liability for soil clean-up costs, increasingly judges or legislators seem to be becoming aware of the fact that it is not reasonable to shift all the costs of historic pollution to the operators still in business today. For example, the Dutch Supreme Court limited the polluter’s liability to the period starting 1 January 1975.96 A subsequent Soil Protection Act also limits the `retroactive’ effect of this liability in the Netherlands.97 In Flanders a Soil Clean-up Decree has been promulgated following the work of an inter-university commission on the reform of environmental law in Flanders.98 This commission, and the subsequent Decree, provide a ground for excuse for the owner of polluted soil who has acted in good faith, i.e. who was not aware of the fact that the specific soil was polluted.99 Also in the recent House of Lords decision, Cambridge Water co v. Eastern Counties Leather PLC,100 liability was denied because the foreseeability requirement was not met in this case of historic pollution.101 Note, however, that a different approach was taken in the United Kingdom’s Environment Act 1995, where strict retrospective liability was instituted with regard to the restoration of contaminated sites.102 3.2.3. Causal uncertainty There is a second trend in liability law, which may seriously endanger the predictability, and therefore the insurability, of certain risks. This is the trend to shift the risk of causal uncertainty to enterprises (and therefore to their insurers).103 Two well-known situations that have arisen in Dutch case law illustrate this trend. The first example concerns the drug DES. It is certain that this product caused birth defects and that specific daughters of mothers who took DES during pregnancy suffered physical harm. The causal link between the use of DES by the mother and the symptoms by the daughter were not disputed. It was also known that certain manufacturers had brought DES onto the market. Uncertainty existed though with respect to which manufacturer sold a specific product to a particular mother. Several “DES daughters” sued all producers that brought DES onto the market, although they could not prove from which manufacturer their mother had bought the drug. This gave rise to a debate concerning the question of whether a type of proportional liability rule should be applied to apportion the burden of liability between the manufacturers. A


31

2.


market share liability would be an example of such a proportionality rule.104 The Dutch Supreme Court, however, applied a so-called “alternative causation” rule, meaning that the DES daughters have the possibility of claiming full compensation from any of the manufacturers.105 The manufacturer can still rebut the presumption by proving that he did not sell DES to the particular mother, but this will often be impossible in practice.106 Hence, this result equals a joint and several liability rules.107 A second example of shifting the burden of causal uncertainty relates to employers’ liability for occupational diseases. In a well-known Dutch Supreme Court case, Cijsouw/De Schelde, a victim of asbestosis could not prove at what time he had been confronted with the fatal asbestos crystal that had caused the disease. The dating of the exposure was crucial for the case since Cijsouw had worked for several years for the defendant firm, but had earlier worked for another enterprise, where he had also worked with asbestos. The Supreme Court once more shifted the uncertainty concerning causation to the enterprise by holding that it was presumed that the employee had been confronted with the fatal asbestos crystal during the period of his employment for the defendant.108 The defendant could rebut this presumption by proving that it was not during the period that Cijsouw was employed by the specific defendant that he came into contact with the fatal crystal. 109 Obviously this would be practically impossible for the employer. In 2000 the Dutch Hoge Raad ruled a case that was even more spectacular with regard to causation. In this case the question was not when the employee was confronted with the asbestos crystal, but whether the disease the employee suffered from was actually caused by chemicals and solvents he was confronted with on his workplace. The causal link was attributed to the employer unless he could prove he had taken sufficient measures to prevent his employee from exposure to hazardous substances.110 The burden of proof resting on an employer can be highly disadvantageous for the employer, especially in those cases where a longer time elapsed since the accident or where no adequate report concerning the circumstances of the accident has been made. Recently the Hoge Raad held that equity could in specific circumstances require that this specific rule of evidence is not applied, so that the general rule of evidence of article 177 of the code of civil procedure would be applied.111 In both cases one is confronted with a causal uncertainty, whereby the burden of proof is shifted to the enterprise. Since causation issues are very often difficult to prove with scientific certainty, this shift of the risk of proof will often be decisive for the outcome of the case. Legal systems other than the Dutch system have also been confronted with the consequences of causal uncertainty. In a Canadian case the Supreme Court of Canada rejected applying an apportionment of damages approach. In this case a plaintiff suffered a herniated disc. Two previous automobile accidents were found by the trial

32


2.


judge to have been a relatively minor contributing factor (no more than 25 percent) to a pre-existing back condition. However, the plaintiff was able to recover full compensation, as the purpose of tort law was to restore the plaintiff to the position he or she would have enjoyed but for the negligence of the defendant. The 25 per cent contribution of the two car accidents to the disc herniation assessed by the trial judge fell outside the de minimis range and therefore constituted a material contribution sufficient to render the defendant fully liable for the damages flowing from the disc herniation.112 Thus, this case confirmed that causation is not to be applied too rigidly. The Alberta Court of Queen’s Bench recently followed this approach in Jones v. Mobil Oil Canada Ltd. Mister Jones’ cattle suffered from diseases, infertility and high mortality rates. The defendant was held liable despite the fact that the poor condition of the cattle could be the result of selenium deficiency of the soil as well and despite the disagreement of the experts on the actual cause.113 Causal uncertainty is especially important with regard to so-called toxic torts,114 whereby a part of the population has been exposed to hazardous substances or radiation and subsequently a certain disease e.g. cancer is discovered. The problem is that unfortunately people may get this particular disease, cancer, from various sources. In that case therefore the identity of the injurer is certain, but uncertainty exists with respect to the victim. Indeed, they may well have received their disease through the so-called background risk and not through the presence of a nuclear power plant for example.115 These kinds of questions have also arisen in Belgium and the United Kingdom. Causal uncertainty played a role in the famous British Sellafield case where an English court had to decide on the causal relationship between childhood leukaemia and the nearby presence of a nuclear power plant at Sellafield.116 Similarly, Belgian courts have been confronted with the question of deciding whether the physical complaints of the inhabitants of the community of Mellery in the Walloon Region were caused by emissions from a nearby waste site.117 An example may illustrate the difficulties that may arise in deciding these kinds of questions. Suppose that in an average population only 20 persons would be the victims of a cancer, whereas it is suddenly established that this number is significantly higher in a population living near a power plant for example. Assume that in that particular population, we find, for example 30 cases of cancer. If we further assume that the marginal increase in the number of victims by 10 has been caused by the presence of the power plant, the question then arises as to which of the 30 victims received cancer through the background risk and who were really victims of the presence of the power plant. The Dutch Supreme Court cases that shifted the risk of causal uncertainty to the enterprise would lead to dramatic consequences in this example.118 In the example under consideration, it would mean that we would presume that all 30 cases of cancer


33

2.


were caused by the power plant, unless the plant could prove that some of the cancers were not caused by its presence. This shift of the risk of proof will put an impossible burden on the plant and will effectively lead to the result that the plant will have to compensate all 30 victims. The result is that the plant pays in 20 out of 30 cases for victims whose cancer has not been caused by its presence.119 This perverse result could be avoided through a proportionate liability rule, such as a market share liability in product liability cases.120 The negative consequences of causal uncertainty could then be limited. A proportionate liability rule is less rigorous than the all or nothing approach of the reversal of the burden of proof.121 The proportionate liability rule would mean that all victims could claim a proportion of their damage equal to the amount, which the power plant contributed to the loss. Thus the power plant’s exposure to liability corresponds precisely to its contribution to the risk.122

4. Summary What does this brief overview of international and national trends in legislation and case law teach us about the scope of liability for systemic risks? Whether international trends are decisive for determining the scope of liability will to a large extent depend upon existing liability law in the particular country. In some countries where far-reaching liability law already existed, international conventions or European Directives will often not substantially broaden the scope of liability; this might, however, be the case in countries where this development has been absent. In any case, European Directives were often promulgated in the first place to harmonise existing legislation, since it is argued that differences could hamper competition. Even though, for example, the Product Liability Directive also argues that it wishes to achieve a higher level of consumer protection in many legal systems, it did not substantially broaden the scope of liability. The scope of national tort law can obviously be broadened through changes in legislation. These changes are, however, in many cases the codification of judge-made law. Indeed, the jurisprudential wish to provide victim or consumer protection has initiated a shift from fault to strict liability. As examples from some legal systems showed, this trend towards strict liability has usually been limited to specific areas of liability law, but a general trend towards no-fault enterprise liability can be noted in Europe as well. However, the soil pollution cases in the Netherlands and the UK equally show that some defences are still allowed under strict liability, e.g. the unforeseeability of the risk. Ultimately, therefore, the difference between a broadly interpreted negligence rule and a strict liability rule may not be that large in practice.123 Obviously, it will largely depend upon the legal tradition in the specific country whether a shift towards a broadening of enterprise liability takes place through case law or through legislation. In the common law, the role of courts is traditionally larger,124

34


2.


whereas one would expect legislation to play a greater role in continental systems.125 However, some examples showed that an important role has also been played on the continent by the judiciary in expanding enterprise liability.126 The normative explanation for this change is usually the victim protection argument. The wish to provide victims with compensation from a solvent enterprise leads to broad interpretations of the negligence standard. In some cases this hidden redistributive agenda may also lead to liability of enterprises for risks that they have not caused themselves (in case of causal uncertainty) or for risks that were unforeseeable at the time when the tort was committed (in case of retroactive liability). These trends may be more dangerous than the shift towards strict liability in itself. Whereas it can be argued that strict liability as such can still be insurable, this is no longer true if retroactive liability is introduced or the risk of causal uncertainty is shifted to the enterprise. Hence, one should be extremely cautious precisely with respect to these tendencies which may affect the insurability of risks. The effects of these changes in liability law on the insurability will be discussed in Ch. 4.

B. Negligence or strict liability for systemic risks? Above, we indicated that both negligence and a strict liability rule could give incentives for prevention, if we assume that only one party influences the accident risk. Let us now address the working of the regimes more in detail and see how they are fitted to be applied to systemic risks.

1. The model As was just mentioned, in the economic analysis of law, tort law is seen as an instrument to deter activities, worth to be avoided on efficiency grounds through liability rules.127 The expectation to be held liable ex post should induce parties ex ante to take care or change the activity level in view of reducing the accident risk.128 Liability rules should, according to this economic model, thus be used in order to reach efficiency. In order to establish the efficient solution, economists state that the goal of tort law is to minimise the sum of accident costs and the costs of accident avoidance.129 The sum, called the social costs of accidents, can be presented as follows:130 C = p(x, y) L + A(x) + B(y), where: C = the sum of expected accident costs and the costs of care, A = the victim, B = the injurer, x = level of care of the victim, y = level of care of the injurer, p = probability that an accident will occur, L = magnitude of the loss.


35

2.


It is assumed that both parties are risk neutral, that the magnitude of the loss (L) is independent of the level of care, that more care will reduce the probability (p) of an accident and that only the victim (A) suffers a loss. To minimise the social costs (C), the levels of care must be set at x = x* for the victim and y = y* for the injurer. At these efficient levels, the marginal benefits from an increase in care (reduction of p(x, y) L equal the marginal cost of greater care.131 From an economic point of view, optimal care is thus not equal to the highest care possible. The highest care might lead to spillage, since the marginal costs would be greater than the marginal benefit in reducing the expected loss. Calabresi formulated this as follows: “Our society is not committed to preserving life at any cost. In its broadest sense, the rather unpleasant notion that we are willing to destroy lives should be obvious. Wars are fought... Ventures are undertaken that, statistically at least, are certain to cost lives. We take planes and cars rather than safer, slower means of travel. And perhaps most telling, we use relatively safe equipment rather than the safest imaginable, because – and it is not a bad reason – the safest costs too much.”132 The difficulty is therefore to find the efficient levels of care x* and y*. In a tort situation, transaction costs are mostly prohibitive, meaning that the efficient levels x* and y* cannot be reached through voluntary negotiations.133 Therefore, the legal system should intervene to provide liability rules which will lead to the efficient level of care.

2. Negligence 2.1. Optimal incentives One rule which may give incentives to the injurer to follow the efficient level of care is the negligence rule. The negligence rule is defined as a rule according to which the injurer will only have to bear the loss if he uses less than a legally required level of care, referred to as the due care level. The negligence rule as defined here, means that the injurer will be held liable if he spends less than the due care level required by the legal system, in other words if he acted wrongful. Assuming that the due care level required by the legal system is equal to the optimal level of care (y*), the injurer will always follow the optimal care level. This is indeed the cheapest solution for him. If the injurer would spend less than y* on care, his total costs would be equal to: p(o, yo) L + B(yo) If, on the other hand, he spends the efficient level of care, he will not have to bear the expected loss. Hence, the expected costs of the injurer are in that case only his costs of taking efficient care: B(y*). The question whether B will take efficient care or not will, therefore, depend upon: p(o, yo) L + B(yo)>< B(y*)

36


2.


If the sum of expected loss and the actual costs of care are higher than the costs of efficient care (y*) the injurer will take efficient care. If, on the other hand, these costs would be lower than the costs of taking efficient care, it would be cheaper for the injurer not to take efficient care. However, since we defined y* as the point where the social costs are as low as possible, p(o, yo)L + B(yo) will always be higher than the costs of taking y*. Hence, the simple conclusion is that under a negligence rule, the injurer will always have an incentive to spend y* on care and an efficient outcome will be reached. This is true as long as the legal system defines y* as due care.134 According to this economic model, an injurer will be found liable under a negligence rule if he spends less than the due care the legal system required from him (y < y*). The crucial question therefore is, from an economic point of view, how the judge should establish this efficient level of care. Spending more on care will reduce the probability that an accident may occur additionally. Requiring the injurer to spend more on care is efficient as long as the marginal costs of care are lower than the additional benefit in reduction of the expected loss. The efficient level of care y* is found where marginal costs of care equal the marginal benefits in reduction of expected loss.135 This is also the point of view of the White Paper, where it is stated: If polluters have to pay for damage caused, they will cut back pollution up to the point where the marginal cost of abatement exceeds the compensation avoided.136 Economic analysis does, however, not assume that judges make explicit mathematical calculations of marginal costs and marginal benefits to establish the efficient care level. Often such a marginal cost/marginal benefit weighing takes place in a rather implicit manner. Hence, economic scholars often argue that judges act as if the goal of their actions were to achieve wealth maximisation, even though their decisions are mostly not formulated in economic terms.137 Sometimes, however, a judge will explicitly argue that the injurer is liable since he could have reduced the likelihood of an accident easily be spending relatively low costs on additional care.

2.2. Learned Hand rule This – explicit – weighing of costs and benefits took place in a well-known American tort case, United States v Carroll Towing Co.138 Judge Learned Hand took economic considerations explicitly into account in judging comparative negligence in the following case: a barge was moored in a harbour, while a storm was approaching. The barge broke loose from its moorings and considerable damage to the ship was caused. The question arose whether the barge-owner had taken appropriate precautions in order to protect its own vessel. Judge Learned Hand wondered whether the accident could have been avoided by having a person watching the barge who could have prevented any collision of vessels. The costs of this preventive measure might have been considerably


37

2.


lower than the expected loss. According to Learned Hand, this question had to be judged by weighing the costs of preventive action versus the damage caused by the accident. Judge Learned Hand translated this question by asking whether the burden of precaution (B(y)) is lower or higher than the probability times the magnitude of the loss (p.L) the formula therefore reads:139 B(y) L >< p.L Although this case is often quoted in American law and economics textbooks, Learned Hand’s formula does not quite fit into the economic model of tort law.140 The problem is that judge Learned Hand expresses the costs of care and the expected loss in absolute numbers, whereas in the economic model y* is fixed by weighing marginal costs versus marginal benefits. Brown pointed at this shortcoming of the Hand formula and showed that the efficient level of care should be determined in an incremental and not in an absolute manner. 141 However, Landes and Posner provided a detailed analysis of American tort case law and argued that in legal practice American courts actually define the required care in incremental terms: “The courts asks: what additional care inputs should the defendant have used to avoid this accident, given his existing level of care?”142 This obviously invites a marginal analysis. Other criticism has been formulated upon the Hand formula as well, which can, however, not be discussed within the scope of this study143.

3. Strict liability From the above it follows that the negligence rule is in principle able to give a polluter an incentive to spend on care to reach the optimal standard. However, the negligence rule only works optimally if the legal system defines the due care level as the optimal standard. What about strict liability?

3.1. Optimal incentives Also a strict liability rule will lead to optimal incentives for care taking for the polluter, since taking efficient care will minimise the expected accident costs which the potential polluter has to bear under a strict liability system.144 Therefore, the literature generally accepts that both a negligence rule and a strict liability rule will provide a potential polluter with incentives to take the efficient care level. However, this is only valid in a unilateral accident setting, in an accident whereby only the injurer can influence the accident risk. Above we already stressed that if victims were also given incentives for accident reduction a contributory negligence defence should be added to the strict liability rule. Under negligence victims will always have an incentive to take efficient care as well since they will in principle not be compensated by the injurer who, under a negligence rule, will take efficient care to avoid liability.

38


2.


3.2. Influence of the activity level In the joint care (bilateral) case strict liability with contributory negligence and a negligence rule (with or without contributory negligence) will incite parties to adopt efficient levels of care. However, the accident risk cannot totally be minimised by increasing the levels of care. Accident losses depend also on the extent to which parties participate in the activity, which might cause the damage (for instance the miles driven). Therefore, reducing the activity level will also reduce the accident risk.145 The activity level can be interpreted as any control variable not taken into account in setting the optimal level of care. Under a negligence rule an injurer has no incentive to adopt an optimal level of activity. This cannot be remedied, because judges cannot easily calculate the optimal activity level into the due care standard146. A strict liability rule has the advantage that the injurer will automatically adopt an optimal activity level. This is also a means to minimise his costs. If the victim’s activity has no influence on the accident risk, strict liability might have a slight advantage, because it might also lead to an optimal activity level of the injurer. However, in a joint care case this advantage is cancelled out by the fact that the victim will not adopt an optimal activity level. This is due to the impossibility of calculating the activity level into the due care standard, when considering contributory negligence.147

3.3. A test for strict liability What is the importance of the activity level for the choice between negligence and strict liability? In a unilateral accident model, whereby only the behaviour of the injurer influences the accident risk, strict liability seems to be the preferred rule since it is the only liability rule which will lead both to efficient care and to an optimal activity level. In a bilateral case the answer is more balanced. Since activity level changes are not calculated into the due care standard, strict liability (with a defence of contributory negligence) will encourage activity level changes of the injurer. On the other hand, a negligence rule will encourage activity level changes of the victim. Therefore, several authors suggest that in bilateral cases strict liability will be a superior device if it is more important to give injurers an incentive to change the activity level than that victims be given a similar incentive.148 This implies that if the injurer’s activity is very dangerous and creates a high accident risk, even if optimal care is taken, it will be more desirable to control the injurer’s activity than it is to control the victim’s. For instance, if an injurer creates the collapse of an old house with explosives in the middle of a densely populated area, it is more desirable to control his activity than it is to control the victim’s. On the other hand, the advantage of a strict liability rule disappears if the injurer’s activity is not important enough to be controlled. For instance, if the injurer runs to


39

2.


catch a train and collides with another pedestrian, a negligence rule would be superior, because it is at least as important to control the victim’s activity as it is to control the injurer’s. If the activity of the injurer is not very dangerous, if exercised with reasonable care, it is desirable to give the victim an incentive to take an optimal activity level as well. Even though a clear-cut test is therefore difficult to give, Landes and Posner describe several factors which may lead to a preference for a strict liability rule.149 These elements are: 1. High-expected accident costs; 2. The impossibility that more care by the injurer would reduce the accident risk; 3. The impracticability to constraining the victim’s activity in favour of the injurer’s and 4. The desirability to reduce the risk by an activity level change of the injurer. How does this economic argument in favour of strict liability in unilateral accident cases compare to legal justifications for strict liability?

4. Legal justifications for strict liability 4.1. Improving the situation of the victim? The reason that is often advanced in legal literature in favour of strict liability is that strict liability will help the victim in obtaining compensation since he is released from the heavy burden of proving fault under the negligence rule.150 However, from a deterrence point of view victim compensation is not as such a goal of accident law. The duty of the injurer to compensate his victim is only an instrument to reach deterrence efficiency. Moreover, the victim compensation argument to introduce strict liability for environmental pollution is not that convincing in all cases. Indeed, many legal systems qualify every violation of a statutory or regulatory norm as a civil fault.151 Most industries are subjected to extensive safety regulation. Hence, in these systems the victim only has to prove the violation of one of these regulations to establish a fault. If, in addition, the victim can prove a causal relationship with the loss suffered, he will be able to claim compensation. In many accident cases this burden of proof will therefore not be as heavy as has been argued. It is, therefore, at least questionable whether a strict liability rule substantially improves the situation of the victim in comparison with an already existing broadly interpreted civil fault regime. It should also not be overlooked that under the general fault regime of tort law no limitations apply and the victim is entitled to full compensation. In many of the cases where strict liability was first introduced, more particularly in the international conventions concerning nuclear accidents and oil pollution,152 financial caps and other limitations on the victim’s rights were introduced. The alleged compensating benefit of the strict liability in those cases is therefore doubtful.

40


2.


4.2. Redistribution? One obviously should make a careful distinction between on the one hand the efficiency arguments, advanced in the Shavellian economic analysis on the one hand and distributional arguments on the other hand. So far we have focused on these efficiency arguments, which show that in particular cases there might be an argument in favour of strict liability. It should, however, be reminded that from an economic perspective these arguments in favour of strict liability for unilateral hazardous activities are only based on the fact that the injurer would then be in a better position to prevent the harm. Obviously arguments in favour of strict liability could also be based on distributional differences. There are more particularly distributional arguments which can often be found in the legal literature. Indeed, an important difference between strict liability and negligence is obviously that if the negligence rule works perfectly, the injurer will take the care the legal system requires from him and will therefore not be held liable. A perfectly working negligence rule will therefore give appropriate incentives to the injurer, but not provide compensation to the victim. A strict liability rule is a rule which in principle guarantees compensation to the victim (if one disregards the insolvency issue, which we shall discuss below). It is precisely because of this distributional difference that many lawyers favour the strict liability rule. Remember, however, that in bilateral cases from an economic perspective strict liability is efficient only if a defence is added (contributory or comparative negligence) to give the victim incentives as well for effective prevention. If we discuss strict liability in the remainder of this report we assume that a defence is always added to take account of the victim’s behaviour.

5. Strict liability for systemic risks? How do, in sum, the economic arguments in favour of strict liability, apply to the environmental case? Although the classic victim compensation argument may, as we just argued, as such not justify the introduction of strict liability for systemic risks, there are on the other hand economic reasons based on deterrence efficiency for introducing a strict liability rule.153 Systemic risks can in most cases certainly be considered a unilateral accident, i.e. an accident whereby only the injurer can influence the accident risk. In this case we noted that the economic model predicts that the advantage of the strict liability rule is that it will give the injurer an incentive both to adopt an optimal activity level and to take efficient care. Since the victim cannot influence the accident risk, strict liability seems to be the first best solution to give the potential injurer optimal incentives for accident reduction in those cases.154 There may, however, obviously be cases where other parties than the injurer could influence the systemic risk of environmental degradation. These are not always the victims on the traditional sense. One can imagine cases


41

2.


where e.g. public or private actors would be responsible to manage a natural resource area. It might be desirable in those cases that liability also aims at giving them appropriate incentives to take those preventive measures. In that case environmental pollution would constitute a bilateral risk on the condition that one considers that third actor a victim.155 However, since, in the example given, the influence of the injurer is probably still far more important than the influence of the other actors, the outcome doesn’t change: a strict liability rule still is warranted to give the party who has most influence on the accident risk (the injurer) the incentive to take preventive measures. It is, however, important to remember that in bilateral cases a defence should always be added to victims as well. Moreover, if also other parties than the polluter can influence the accident risk as well, they might be held liable as well to the amount in which they contributed to the loss. That is, however, not an argument against the strict liability of the injurer. So, if we apply the criteria of Shavell determining the choice between negligence and strict liability to systemic risks, there seem to be strong arguments in favour of an introduction of strict liability. In many cases systemic risks will be truly unilateral in the sense that only the injurer’s activity can influence the accident risk, which constitutes a strong case for strict liability.156 In other cases the victim will certainly be able to exercise an influence on the risk as well. One can more specifically think about situations where the victim has the possibility to mitigate damages after the accident occurred. However, in those cases it is not the victim’s activity level, but his level of care which influences the accident risk. As was mentioned, this can be controlled by adding a contributory or comparative negligence defence to the strict liability rule.

6. A few refinements 6.1. Information differences Many law and economics scholars argued that there is indeed a strong case in favour of strict liability in case of systemic risks: this will give the potential injurer optimal incentives for accident reduction and hence, for optimal internalisation.157 There is, however, another important aspect of the difference between negligence and strict liability which should be mentioned. This concerns the fact that the application of negligence requires high information costs from the judge, who will have to set the due care standard. The information necessary to weigh costs and benefits and to fix the optimal care may not be readily available for the judge. Strict liability shifts all costs to the injurer, who will then have to define the optimal care level. If one therefore assumes that, as may be the case with environmental harm, the information on optimal precaution is better available with industry than with the judges, this constitutes an argument for strict liability. Note that obviously in some

42


2.


cases there may be an information advantage with the regulator. This is an argument on favour of regulation, but not necessarily against strict liability. This information advantage may therefore constitute an additional argument in favour of strict liability for systemic risks. One should, however, remember that this finding only holds in all the models, such as the one which is e.g. developed by Shavell, which start from an assumption of risk neutrality. If risk aversion is introduced and the potential injurer is risk averse, Endres/Schwarze correctly argue that strict liability is only efficient if in some way risk can be removed from the risk averse injurer, e.g. through insurance158. Moreover, we assume that the judge has accurate information on the amount of the damage. If courts err in assessing damages, strict liability will lead to underdeterrence.159

6.2. Insolvency: strict liability versus negligence There are other reasons why the seemingly advantage of strict liability should be somewhat balanced. Until now, it was assumed until now that the injurer has money at stake to pay compensation to the victim. If, however, the amount of the damage exceeds the injurer’s wealth, a problem of underdeterrence will arise. Under strict liability the injurer will consider the accident as one which is equal to his total wealth and will therefore only take the care necessary to avoid an accident with a magnitude equal to his total wealth. If that wealth is lower than the magnitude of an accident he will take less than the optimal care and therefore a problem of underdeterrence arises under strict liability. Insolvency is less of a problem under negligence since under that rule the injurer will still have an incentive to take the care required by the legal system as long as the costs of taking care are less than his individual wealth. Taking due care remains indeed a way for the injurer to avoid having to pay compensation to the victim. If there would thus be a potential accident setting whereby the magnitude of the loss may be higher than the injurer’s wealth (which can often be the case in environmental liability) this constitutes an argument in favour of negligence rather than strict liability.

6.3. Positive externalities There is, in addition, another issue, which has been mentioned in the literature which may balance the arguments in favour of strict liability. Gilead rightly pointed out that many activities do not only externalise harmful effects, but also have positive effects, which may be a reason to be somewhat cautious with strict liability.160 That is a consequence of the fact that strict liability incorporates activity level changes, whereas a negligence rule does not. It is not always obvious that increases in activity levels are socially undesirable.161 Think about the manufacture of drugs. Most of this activity creates huge social benefits. One should therefore be careful with the argument that strict liability is needed to


43

2.


internalise risk and control activity levels if the benefits of an increased activity were neglected. There is finally, one other nuance which has to be added. So far we assumed a strict liability rule whereby the injurer is “merely” held to compensate the damage he caused to his victim, not less but also not more. Problems may arise if strict liability is combined with other features which may expand the burden of liability on enterprises. One could think of a removal of the burden of proving causation,162 joint and several liability and high (punitive) damages for (nonpecuniary) losses. These are formally different issues than the simple choice between negligence and strict liability. However, Trebilcock163 indicated that it was especially because of these other features that the strict liability regimes in the United States were experienced as “crushing”. Hence, the final judgement on the efficiency of strict liability may also depend upon these other issues, such as causation and the magnitude of the damages awarded.164

C. Developments in social security 1. Introduction One of the purposes of this study is to analyse what elements may affect the insurability of emerging systemic risks. We indicated that the insurability may be endangered by an increasing scope of liability. The question whether an enterprise may be found liable after a systemic risk caused an accident (and he will therefore be bound to pay compensation) may not only depend upon the developments in substantive tort law, which we have just discussed, but also upon the existing social security systems. In some countries, the social security system may be so elaborate that it eliminates the need for a victim to file a tort claim. Therefore, in many cases where potential liability exists damage suits are nevertheless not brought. Thus the scope of the social security system may well have an important impact on the insurability as well. Let us briefly address how the relationship between tort law and social security may affect the scope of liability and therefore the insurability of risks. We will focus briefly on the general relationship between tort law and social security and show how changes in this relationship may lead to a changing scope of liability. We will show this by providing an example from the Netherlands, but the same may be the case for other systems as well.

2. Different principles 2.1. Prevention versus compensation The literature indicates clearly that there are significant differences in function between tort law and social security in the field of compensating for personal injuries.

44


2.


In that respect we can be rather clear if we address this from an economic perspective: economists assume that tort law is an instrument to deter activities, worth to be avoided on efficiency grounds through liability rules. The idea is that the expectation to be held liable ex post will induce potential parties in an accident setting ex ante to take care or to change their activity level. For the economist the most important goal of tort law therefore is the prevention of accidents.165 Some have seen here a difference between the legal approach and the economic approach. Whereas economists stress the importance of tort rules as having an ex ante effect on prevention, traditional tort lawyers stress the ex post compensatory function of tort law. If one would exaggerate the difference between the traditional economic and the classic legal approach one would state that the economist is interested in prevention of accidents, whereas the lawyer is interested in victim compensation. However, as is often the case with generalities like these, reality is often more balanced. Some economists (or lawyers writing in the economic tradition of law) also point at the loss spreading function of the tort rules (and therefore implicitly at victim compensation) ;166 some lawyers (although there are also differences in legal systems) equally stress the preventive function of tort rules. This is e.g. the case in Austrian tort law.167 Moreover, in a brilliant study Schwartz has demonstrated that tort rules can be considered to serve both the goals of prevention (via deterrence) and of compensation (also often referred to as the corrective justice idea).168 Notwithstanding these nuances, which somehow reconcile legal and economic thinking concerning the goals of tort law it should be kept in mind that from an economic point of view liability rules are primarily considered as an incentive system. The idea therefore is that the foresight of being held liable after an accident will encourage a change of behaviour (towards taking precautionary measures) before the accident happens. In this basic framework the fact that a victim may be compensated after an accident is not considered as a goal of tort law, but as an instrument to provide incentives to a potential injurer for preventive behaviour. Many different legal rules can be used to reach this optimal deterrence. The literature has paid a lot of attention to the way in which a negligence regime or a strict liability regime can provide injurers with incentives to take optimal preventive measures to reduce a risk. An important difference between negligence and strict liability is that if the negligence rule works correctly, it will give an injurerincentives to follow the appropriate care required by case law. The effect will, however, be that if nevertheless an accident happens (although preventive measures and due care was taken) the injurer will not be bound to compensate the victim. Under a correctly working negligence rule the victim is hence not compensated. The result is in principle the opposite under strict liability: if one defines strict liability simply as a rule whereby the injurer is


45

2.


always held to compensate the victim the victim will under strict liability (if one disregards the insolvency problem) always be compensated. In sum: the function of tort law in the economic approach is clearly the prevention of accidents. Compensation is only an instrument to force a potential injurer to take preventive measures. The opposite is obviously the case if one regards a social security system from an economic perspective. The major goal of social security is precisely to provide, depending upon the services provided, compensation (e.g. for lost income) or health care, in principle irrespective of the source of the illness.169 If one would therefore like to identify a dividing line between tort law and social security law as far as personal injuries are concerned one could very generally state that from an economic perspective the general function of tort law is prevention of accidents whereas the general goal of social security law is providing compensation for the damage caused as a result of personal injuries. A major difference following from this dividing line is that tort law is triggered as a result of an accident whereby a third party is held liable for having caused the personal injury, whereas this is not at all necessary under social security. Social security is therefore a system whereby compensation is, if the legal conditions are met, provided directly to the protected victim, irrespective of whether his personal injury was caused by a liable third party. It may well be that social security intervenes in cases where it was entirely the victim’s fault that personal injury was caused.170 But the prevention of harm is also increasingly stressed as being important as well.171 However, the payments made under social security schemes cannot be considered as having mainly a prevention goal. Traditional social security still focuses on compensation of losses.172

2.2. Combined use in practice Although the starting points are therefore different (torts: deterrence; social security: compensation) the economic literature has indicated that if one would have pure systems relying totally on either tort or social security the systems would have to take care of compensation issues respectively prevention of accidents as well. We mean the following: even in a system which would primarily rely on torts, one will notice that the tort system (which economically serves the goal of deterrence) will be complemented with systems of insurance to meet the second goal (compensation). Thus a potential tortfeasor could take out liability insurance coverage which might (if the injurer is found liable) serve the interests of the victims as well, since it provides a guarantee against the insolvency of the potential injurer. Moreover, under a pure tort system victims (who would fear that they would receive no compensation) could chose to take out first-party insurance coverage to cover for personal injury losses. This is, by the way, what obviously happens in practice a lot:

46


2.


many potential victims will take out accident insurance coverage to provide (additional) coverage in case personal injury happens to them. Hence, the deterrence oriented tort system might be combined with (liability or firstparty) insurance schemes to serve the goal of compensation.173 The same applies, however, as well for a system whereby automatic compensation (irrespective of the behaviour of the beneficiary) would take place under social security. If such a perfectly working social security compensation mechanism would exist whereby all potential victims would be guaranteed compensation for their personal injuries some other systems than tort (assuming that this would not exist) would be needed to guarantee that those who may have an influence on the accident risk behave properly. The systems which are then advocated to be combined with social security are systems of safety regulation. Thus, the model then assumes that victims are compensated via social security, whereby this is combined with ex ante government regulation which is enforced via administrative or criminal law. Regulation, not tort law, then serves the goal of deterrence.174 This is a model which has been advocated by many economists: they have often argued that if society wishes to reach the goals of both deterrence of accidents and compensation of victims for personal injury, it could chose between on the one hand a tort system (for deterrence) + insurance systems (for compensation) or on the other hand a system of regulation (for deterrence) + social security (for compensation). This combination of private and public regulation of safety has especially been advocated by Skogh.175 In practice of course a combined variety of all of these instruments of deterrence and compensation exists, although the focus on each of the particular instruments may change per society and over time. However, as we will demonstrate with the example of the Netherlands, all of the systems (torts, insurance, regulation and social security) are interrelated and mutually influence each other. Hence, one can expect that e.g. if the government would withdraw from the compensation of personal injury via social security, victims would be forced to make an increasing use of tort law and insurance to meet this same goal. That these relationships may change can be illustrated by the example of employers’ liability in the Netherlands.

2.3. Example: employers’ liability In the Netherlands, in the case of an occupational disease, social security traditionally provided limited compensation of lost income and took care of health care expenses. This corresponded with the traditional view that it is social security which provides “Existenzsicherung” (subsistence protection).176 In addition to the social security system, tort law traditionally played a role in the case of occupational diseases, but only a modest one. Victims (employees suffering from an occupational disease) only used tort law to receive


47

2.


compensation for the top of their income (the part which was not covered by social security) and to get coverage for non-pecuniary loss. These were precisely the types of damage not covered under social security. Prevention of occupational diseases was to a large extent guaranteed through health and safety regulation, imposing specific safety duties on the employer. This resulted in a system whereby victims of an occupational disease received primarily compensation via the social security system to provide some “Existenzsicherung”. Tort law could be used, if the specific conditions were met, for that part of the damage which was not covered under social security. Obviously some interrelationships between those systems could exist in the sense that (again, under specific conditions) the social security system might use tort law to attempt to get recovery of the amounts paid to the victim. This is precisely the issue of recourse which we will discuss below. It is striking that in this traditional (Dutch) (but to some extent European) system tort law was a luxury system and at the same time of rather limited importance in the compensation of victims. Indeed, only a limited amount of the damage which occurs in society is covered via tort law.177 The largest part of damage was covered either via social security or via private first-party insurance. Tort law can be considered a luxury system in the sense that it provides a guarantee of in principle full compensation for the damage suffered and even compensation for non-pecuniary loss. That is a luxury, so it has been held in the literature, which the social security system cannot afford. 178 Indeed, the essence of an “Existenzsicherung” is that it provides a minimum, but not the “luxury” of full compensation. The economic reasons why the social security system cannot guarantee full compensation (including compensation for non-pecuniary losses) are manifold, the costs of full recovery would be high and would lead to higher premiums or an increased pressure on public budgets. In addition non-pecuniary loss will be different for every individual, whereas social security usually works with more or less fixed, at least standardised compensations.

2.4. Tort Law: luxury or “Existenzsicherung?” Therefore it was held that this “luxury” of tort law can only be provided in exceptional circumstances and when specific conditions are met. Hence, tort law cannot guarantee full compensation to every victim of personal injury, not even to every victim of an accident. It will in most systems be dependent on the behaviour of the injurer whether the victim can obtain this full compensation. That again corresponds with the economic insights sketched out above that tort law is, at least in systems where it does not have to provide this social security function, more a system to influence the behaviour of the injurer than to provide compensation. As we will indicate below, this is according to

48


2.


us a major difference between tort law as it is viewed in the US and tort law as it was traditionally viewed in Europe. Obviously the particular view a system takes on the function of tort law (providing Existenzsicherung or merely luxury system) will have important social consequences as well. If victims are forced to use tort law also to provide this basic “Existenzsicherung”, e.g. because social security would not provide this basic coverage, one can understand that victims would in fact be forced to increasingly use tort law and that judges would be tempted towards an expanding scope of liability. If, on the other hand, “Existenzsicherung” is basically provided via social security, the least one can say is that tort law does not have to be used to provide this basic compensation. Our claim is that in many Western-European systems (and more particularly in the Netherlands) the traditional (although maybe implicit) view was that the basic compensation is provided via social security; tort law intervened only if specific conditions (depending upon the behaviour of the injurer) were met and usually only for those types of damage not covered under social security. However, although this might have been the traditional approach, at least for compensation of occupational diseases, at least in the Netherlands, some changes seem to occur which might have an important influence for the way in which tort law is viewed and used. These changes in social security have not only occurred in the area of healthcare, but more specifically also within that domain of social security which aimed at providing replacement income in case of disability. As a result of these changes a full income guarantee is no longer unconditionally provided under social security, which obviously increases the pressure on tort law.

3. Changes in the relationship between tort and social security 3.1. Pressures on social security We already indicated that the traditional way that tort law was used – at least in Europe -, compared to social security was in fact as an exceptional system. Only a limited part of the damage occurring in society as a result of accidents was compensated via tort law. A much larger part was compensated via social security. Tort law only intervened in specific circumstances (wrongfulness with the injurer) and for particular types of damage: top of the income and nonpecuniary loss.179 Effectively, tort law was in that respect more or less replaced by social security e.g. in the field of occupational health. In the Netherlands such a system existed under the accidents at work act of 1901: this provided employees social security and granted immunity for liability to employers. Such a system is still in place in many legal systems. A legislative change later allowed claims on the employer, but they were effectively rarely used. The simple reason was that social security still granted compensation for the basic needs. Victims therefore e.g. in the Netherlands only used tort law in exceptional


49

2.


cases.180 However, several changes have occurred recently, in Europe in general and more particularly in the Netherlands as far as the financing of damage caused through accidents at work is concerned. This follows from a general debate on the reform of social insurance in Western Europe. In many legal systems it is now felt that the health insurance schemes that were developed a few decades ago have come under increasing pressure.181 The increasing costs of health care systems (and of social insurance in general) have led to a variety of proposals.182 In some legal systems it is noticeable that the government has progressively withdrawn from providing all-inclusive social insurance schemes and prefers to rely more heavily on market solutions and private insurance. This behaviour obviously fits with the deregulation wave that hit Western European politics in the 1980s. As a consequence of this deregulation – for example in the Netherlands – an increasing use of the liability system is advocated, so that tort law could be used more often to compensate accident victims, thereby releasing the social security system from this heavy burden.183 In that debate attention has also been paid to the role of private insurance as an alternative to social insurance. The question therefore arises of how to find a proper balance between the two systems.

3.2. Importance for the scope of liability Recent changes in social security, not only in the Netherlands but also in many other West European countries, may have had an important influence on the scope of liability of enterprises.184 Our basic assumption is that for a long time victims in Western Europe, contrary to their American counterparts, did not have many incentives to bring a liability suit, since the social security system in many Western European countries provided for a relatively wide coverage of many expenses that a victim incurs when an accident happens.185 Partially as a result of international conventions186 and European Directives,187 many countries have put in place systems of compulsory first-party insurance covering medical expenses. In addition, lost income is often taken care of as well. Depending upon the legal system, this is usually mandatory if one is employed and on a voluntary basis for self-employed individuals. In some cases even property losses are insured, depending upon the type of accident. This relatively elaborate first-party insurance system had as a consequence that a large part of the damage to victims was taken care of already. Victims individually therefore only had incentives to sue an injurer in tort for the part of the damage that was not taken care of via the first-party insurance scheme. This could in some cases be property loss, or the higher part of one’s income which would not be compensated and – mostly – compensation for so-called pain and suffering for which no first-party insurance is available.188

50


2.


Therefore, if one wishes to analyse the scope of liability in a particular legal system, one should also take into account the extent to which damage is already compensated for via other (private or public) compensation mechanisms which remove the need to use the tort system.189 In addition, if one wishes to analyse possible evolutions in the scope of liability one should also address possible changes in social security law. We shall discuss the Dutch case as an example to illustrate how changes in the social security system can influence the amount to which injurers will have to intervene to provide compensation to potential victims. Obviously, the power of one state to change its social security system is limited by its duty to fulfil its obligations under European and international law.190 A closely related question is whether first-party insurers or social security organisations have a right of redress against the injurer after having compensated the victim. One can also understand that if e.g. a health care insurer (or other social security agency) were not subrogated to the rights of the victim for the amounts paid in health care (or for other social security provisions), this would seriously limit the scope of liability of the injurer. From an economic perspective this may obviously endanger the incentives of the injurer for prevention. By the same token, proposed changes to increase the scope of redress for social insurers would obviously expand the scope of liability.

3.3. Recent changes in the Netherlands 3.3.1. In social security... Since it is not possible to discuss workers compensation issues at great length in this study, we will discuss a few recent changes in Dutch law that illustrate the trend of deregulation in this area which seems to be typical for the policy in the nineties in many Western European legal systems. It is important to give these issues some attention, since they are obviously crucial for the question how damage caused by systemic risks will be compensated. Generally the recent Dutch changes aim at increasing the individual responsibility of an employer, inter alia for the costs associated with absence from work of his employees. The basic idea is that by making employers feel the financial consequences of absence from work directly, this will increase their incentives to implement an effective preventive policy. One can understand that many of these recent changes have been highly debated in the Netherlands. These changes are driven primarily by the idea of making employers individually liable for social security payments. This obviously contrasts sharply with the idea of “risque social” that has traditionally underpinned Dutch social security law. 3.3.1.1. Sickness pay Both with respect to workers compensation and safety at work, two important changes have taken place since 1994, and other recent developments merit further discussion as well. As far as workers compensation is concerned,


51

2.


since 1 January 1994 an employer is forced to pay his employee 70% of his salary for a period of 6 weeks.191 Recently this system was even further changed since the Dutch government decided to completely privatise the workers compensation scheme at the beginning of 1996. The duty to pay the salary has now been extended to 52 weeks and an employer can purchase insurance coverage against the risk of having to continue to pay the salary of his employee.192 The fact that the employer purchases insurance coverage is again criticised since it is held that this might precisely dilute the incentives for an effective control of the employees by the employer, for which the system was introduced in the first place. Notice that the insurers claim that it is hardly possible to effectuate an adequate control of employers, as was suggested would be the result of the regulation. In addition to the expected incentive effect, for this new legislation was obviously that the government expected savings of up to 600 million Dutch guilders from the new privatised system in 1998. It is not immediately clear what the consequences of these recent changes will be for the scope of liability.193 Obviously, this privatisation means that the costs of providing workers compensation are no longer borne by the social security system (and hence redistributed to the society at large), but are shifted to the individual employer. This duty to pay the salary exists irrespective of any liability. In principle the employee is still compensated so that one could argue that his incentives to use the liability system for compensation remain in principle unchanged. However, there is one change which may affect the need of victims to use the tort system for compensation, i.e. the salary is only paid up to 70% unless the employer agreed to pay more than 70%.194 If the inability to work is therefore caused by a tort, a victim still has incentives to sue for the remaining 30% and obviously always for pain and suffering.195 Moreover, we should add that, as was mentioned above, the duty imposed upon the employer to pay the salary of the employee is limited up to 52 weeks. After the 52 weeks, the employee falls within the “classic” social security system, although the replacement payments are now seriously reduced. After one year of illness (paid by the employer) the employee who is still ill will fall into the already mentioned WAO. Depending upon the applicable system, this will provide some (although limited) compensation for lost income, but it is also limited in time. Hence, as a result of all of these changes tort law has again become important, more particularly for the long lasting income damage. Indeed, after a certain period of this WAO some employees in the Netherlands might end up with a relatively low amount of social security benefits which is barely enough to live. That obviously explains the increasing importance of the tort system, more particularly for lost income in case of long lasting diseases.

52


2.


3.3.1.2. Safety in the workplace A second evolution concerns safety in the workplace. The Dutch Health and Safety at Work Act (Arbeidsomstandighedenwet) has also undergone several changes since 1 January 1994. Especially important are in this respect plans that have been launched by a commission that proposes a total deregulation of the Safety at Work Act. A commission, chaired by Kortmann, proposed replacing the existing regulations aiming at safety at the workplace through a civil liability in tort to prevent accidents at work. The Safety at Work Act should – in the newderegulated system – consist of only a few basic principles, supported by civil liability. Although these proposals have been seriously criticised,196 and it is thus not clear whether they will be politically viable,197 it does not seem entirely unlikely that some system of deregulation, introducing a generalised duty of care for the employer, will be introduced. Although the future evolutions with respect to these revolutionary ideas are still unclear, it is obvious that a far-reaching deregulation of the Safety at the Workplace Act, aiming at an increased use of the civil liability system, may increase the scope of liability.198 In that case the changes will not be caused by evolutions in the tort system itself, but by external factors that force victims to make an increased use of possibilities already available within the tort system. 199 Again, it should be stressed that all these recent legislative changes we just described lead to a reduction of the scope of social security. This is more particularly important as far as the changes in the Dutch WAO (providing replacement income in case of disability) are concerned. As a result of these changes social security payments become more limited in scope and in time which may increase the incentives for victims to use the tort system. 3.3.2. … and in tort law The amazing fact is hence that at the political level in the Netherlands some voices were heard that victims (of accidents at work) should increasingly use the tort system, since this was then considered as a means to reduce the pressure on the social security system. Apparently politicians were at the time (beginning of the 90s) not aware (or neglected) the fact that a reduced social security coverage might lead to an increased use of the tort system, but could even expand beyond the reasonable (insurable) abilities of this tort system. In addition to the mentioned changes in the social security system (and maybe partially as a result of it) victims of accidents at work increasingly used the tort system to obtain compensation. Hence, the decrease of social security protection was paralleled by a development in tort law in the Netherlands towards increased victim protection and an expanding employers’ liability for occupational diseases.


53

2.


This can be easily demonstrated by looking at case law. The legislation changed in the Netherlands in 1997 the employers’ liability, introducing a reversal of the burden of proof concerning the fault of the employer (see Art. 7:658 BW) .200 In addition, there is growing trend in case law towards increased victim protection in case of employers’ liability.201 More particularly the Dutch Supreme Court also reversed the burden of proof in case of causal uncertainty concerning the precise time when an employee had inhaled a fatal asbestos crystal resulting in asbestosis. Some of this case law was already discussed above in Ch. 2. In a well-known Supreme Court case, Cijsouw v. De Schelde, a victim of asbestosis could not prove at what time he had been in contact with the fatal asbestos fiber that caused his disease. The determination of this moment was crucial for the case since Cijsouw had worked for the defendant firm for several years, but in a first period the employer could not have known that he had to take measures to protect his employee against asbestos and thus could not be liable. The Supreme Court shifted the uncertainty concerning causation to the enterprise by holding that it was presumed that the employee had been in contact with the fatal asbestos fibre during the second period of his employment with the defendant.202 This presumption could have been rebutted if the defendant had been able to prove that it was not during the second period that Cijsouw was employed by the defendant that he was in contact with the fatal fiber.203 Obviously, this would have been practically impossible for the employer. The threshold for employer’s liability is relatively low since it suffices that the employee holds (and in cases of motivated discussions proves) that he suffered damage as a result of the execution of his work.204 However, even this low threshold for liability can constitute a serious barrier with specific occupational diseases. Sometimes the employer can as a defence argue that the damage has another cause than the execution of his work by the employee. The employer could for instance hold that (a specific type of) long cancer can be caused by working with asbestos, but could equally have other causes, such as smoking. The same is true in case of new health risks such as RSI (repetitive strain injury) and stress. An employer would of course suggest that these complaints have to do with a variety of other factors, which have nothing to do with the work (like lifestyle, working at home, family situation etc.). This issue has played a role in an asbestos procedure. There was doubt whether a victim of mesotheliom had incurred the mesotheliom when working for the employer De Schelde. The Hoge Raad held that there is no reason to expand the liability of the employer in such a way as that he would also be held liable for damage which the employee has suffered and for which it may be possible that this has been caused as a result of the execution of the work by the employee.205 The key problem is of course that there is uncertainty concerning the causal link. In the end, it will be possible to hold with certainty that the

54


2.


damage was incurred at the occasion of the execution of his work. Recent case law (although not from the Hoge Raad) shows that this lead to success for the employee. For instance the Kantonrechter in Middelburg accepted that complaints concerning repetitive strain injury could by their natureindicate a link with the execution of the work.206 Moreover, the contributory negligence defence in case of employers’ liability was severely reduced as a result of the case law of the Supreme Court. It has inter alia been decided that there can only be a reduction of the employer’s liability in case of gross negligence of the employee.207 Moreover, it was decided that gross negligence had to be interpreted as intent or wilful recklessness.208 This wilful recklessness can only be accepted if it is clear that the employee in his conduct immediately preceding the accident was actually aware of the reckless character of his behaviour.209 In other words, the contributory negligence defence almost became without contents in the context of employers’ liability. Moreover, although formally employer’s liability is still based on negligence, the scope of liability and the duty of care have been expanded in such a way that some claim that employer’s liability in the Netherlands is actually based on strict liability, even though this is according to case law not formally the case yet.210 The result is that both as far as the finding of liability is concerned, the burden of proof, causation as well as the interpretation of contributory negligence, all tendencies go toward an increased victim protection and in favour of an increased employer’s liability. The employee is not so much protected against the dangers at work, but even against himself. Some have claimed that such a trend is contrary to the fundamental principle of liability law.211

3.4. The best of both worlds Several trends are pointing in the same direction: social security coverage has been decreasing in the Netherlands (as far as damage resulting from occupational diseases is concerned) and the scope of employers’ liability has been increased through case law. The same is, moreover, true for other fields of liability as well. A result of this development is that probably the traditional difference between tort law and social security might become smaller. The traditional idea was that social security would provide for easy compensation (with a low threshold), but also for a limited amount, whereas full compensation could only be awarded when the more complex conditions of liability under tort law were met.212 Victims of accidents now seek “the best of both worlds” within tort law: they seek the low threshold for compensation of social security, to be combined with full compensation under tort law. Indeed, as a result of the mentioned developments within liability law employees now can claim compensation from the employer in tort on the basis of very low thresholds (which were customary in social security), but with high


55

2.


compensation (typical for torts). This happens, moreover, in a period wherein social security protection reduces and therefore the need for victims to need tort law even increases. This evolution may, however, lead to problems at the insurance level. Traditional tort law and liability insurance have not been developed as mechanisms which should guarantee full compensation to all victims of accidents.213 One problem is that tort law is now expanded or combined with pleas in favour of no fault compensation schemes whereby often the question of who should finance this expanded protection seems to be neglected. The future will probably show that victims and policymakers will have to make a choice between either an automatic compensation which can be warranted through no fault compensation schemes (social security, firstparty insurance or compensation fund), but then the damages awarded are necessarily limited (and the question still has to be answered how incentives for prevention can be given) or to rely still on tort law with its full compensation for a necessarily limited number of victims.

D. Towards a “claim culture”? 1. Introduction In Part I of this section, we sketched several developments at the international level and in national legislation and case law that may affect the likelihood of a law suit being brought against an industrial operator. In the previous part, we discussed the importance of factors outside tort law that may equally affect the probability that victims will be forced to use the tort system. All of this will affect the likelihood that claims resulting from emerging systemic risks end up with insurers. We shall now discuss a few factors that do not affect material tort law itself but may prompt victims to have increasing recourse to the existing possibilities in tort law. Indeed, whether or not a law suit will be brought in a particular case will to a large extent depend on the role of the actors in the legal system – firstly, the victims or consumers themselves, secondly the providers of legal services – mostly attorneys, and thirdly, the judiciary. The attitude of these groups may to some extent influence the likelihood that the tort system will be used.

2. Victims Whether or not a victim will bring a liability suit will to some extent depend upon the existence of procedural instruments that can improve the position of a victim, for example that could make access to justice easier, such as a collective right of action. Such a collective right of action has been introduced recently in many Western European legal systems, especially in environmental and consumer law, and could trigger a higher number of claims.214 In the Netherlands, even a generalised right of action for groups was introduced in Art. 3:305 a and b of the Dutch Civil Code.215

56


2.


Nevertheless a collective right of action will usually not lead to damages being awarded to the acting group, but mostly only to an injunction. US-style class actions are not known in European procedural law.216 In Dutch DES case, for example, every individual victim acts as a plaintiff in the court, although all separate cases can be joined.217 In the area of workers’ compensation or work accidents, labour unions will usually play a large role in the litigation. With respect to the victim’s willingness to file a claim, the role of insurance should be pointed out. If the victim has insurance coverage for his legal fees, this barrier to starting a legal action will be removed. Obviously the insurer will exercise a marginal control via his assessment of the chances that the claim will be successful, but the insurer will deny coverage only in cases where it is ex ante undoubtedly clear that no serious claim exists. In many countries the number of insurance policies for legal fees is increasing rapidly.218 The removal of this barrier through insurance coverage may, in combination with greater awareness of the possibility of filing a claim, lead to an increase in the number of law suits, although there is no empirical research to support this assumption as yet.219 The issue of whether the victim has insurance coverage for legal fees or not can obviously affect the decision of the insured to file a law suit. In addition, one should note that the willingness to claim may increase greatly when victims become organised. Some patients’ associations aim to increase the awareness of patients that they can file a lawsuit against a physician in cases of medical malpractice. In the Netherlands, almost every new disease spawns a patients association. The introduction of such an association, for instance in the case of whiplash damage, often has a substantial influence on the number of victims as well. In many cases the mass media will also bring the possibility of bringing a liability suit to the attention of victims.220 This has been the case for recent phenomena such as Repetitive Strain Injury (RSI) and the Organo-Psycho-Syndrome (OPS). For instance, the University of Amsterdam published a brochure indicating how an employee could bring a lawsuit against an employer who was liable for OPS.221 Whether such claims will eventually be successful will, however, depend to a large degree upon the organisation of the legal services.

3. The Role of the Bar 3.1. Monopoly Rights Obviously the Bar can play an important role in the decision to file a lawsuit. However, although many monopoly rights still exist in this respect,222 in many cases other professionals have a right to represent a victim in court as well. For example, in many legal systems a representative of a trade union may represent an employee in court in a claim against his employer. This is the case in the Netherlands before the “Kanton-juddge” and in Belgium before


57

2.


the labour court. However, given their monopoly rights, attorneys will still play an important role. Nevertheless, it should be mentioned that in many legal systems, attorneys’ monopoly rights are the subject of serious criticism. In the Netherlands, for example, Brunner noted that pleas may shortly be heard in favour of an abrogation of monopoly rights for attorneys as early as 1980.223 Recently, the wave of deregulation in the Netherlands also reached the area of legal services. The Dutch Ministers for Justice and Economic Affairs have argued in favour of deregulation of the Bar.224 The question, however, arises as to whether there is any relationship between the monopoly rights of attorneys and the possible number of claims. It is highly doubtful that the abrogation of a process monopoly will lead to more claims. In the Netherlands and in many other legal systems, attorneys already do not enjoy a monopoly concerning workers compensation claims. Nevertheless, many victims still use the services of attorneys and there is also no proof that attorneys bring fewer claims than other providers of legal services such as trade union representatives.225 Therefore, there seems to be little reason to assume that the mere abrogation of the monopoly to plead would by itself lead to an increase in the number of claims.226

3.2. Contingency fees Another development that might influence the willingness to claim relates to the way in which attorneys are remunerated for their services. In most continental legal systems, attorneys are not allowed to make fees dependent upon the outcome of the case.227 In some countries this prohibition of the socalled pacta de quota litis (known in common law countries as contingency fees) has been laid down in the Code of Civil Procedure e.g. in Belgium.228 Although contingency fees have therefore traditionally been prohibited for a long time, some voices have been heard recently in favour of allowing this system of payment in the Netherlands. One reason is that Dutch law firms operate on an international market in competition with – inter alia – American law firms, which could have a comparative advantage if they were allowed to charge contingency fees. Therefore it is not unlikely that contingency fees will be allowed in the Netherlands in the near future.229 More important than the question of whether or not contingency fees will be allowed is obviously the question of the influence a contingency fee system would have on the incentives of attorneys to file a law suit. The feeling exists that the tort crisis of the US was partially due to attorneys being paid on a “no cure, no pay” basis, and bringing frivolous lawsuits. The reluctance to introduce contingency fees in the Netherlands is also attributed to the fear that it would lead to an increase in the number of claims.230 From an economic point of view it is true that a contingency fee system may provide better incentives to attorneys to invest a lot of effort in winning a case. On the other hand, the

58


2.


system should not necessarily lead to more claims, since an attorney will only file a law suit if he estimates that he has a reasonable chance of success, given that his personal interest is at stake. Claims that have no chance of being won in court will also not be filed in a contingency fee system.231 However, there may be one reason to assume that contingency fees may increase the number of claims. This has to do with the fact that until now many victims did not use the possibilities of the tort system in Europe, either because they were unaware of them or because the damage they incurred was taken care of through the social security system. If one takes into account a deregulation of the social security system in combination with greater public awareness of the possibilities of the tort system, the contingency fee system may obviously provide incentives to attorneys to file additional claims.

3.3. Reimbursement of legal fees by the losing party A final element that may play an important role in the decision to file a law suit or not has to do with the question of whether a potential victim can recover his legal fees from the defendant if the latter loses the case, and equally the question of whether unsuccessful victims face the risk of having to reimburse defendants’ legal fees. The US system globally accepts that every party takes care of their own legal fees. It has been argued that the fact that an American plaintiff cannot be held liable for the legal fees of the defendant in cases of unsuccessful claims provides too little incentive to the victim to think twice before starting a lawsuit, and hence leads to an excessive high number of claims.232 In most European countries it is accepted that a successful plaintiff may get (part of) his legal fees back from the defendant. In the Belgian system these amounts are limited and much lower than the actual fees to be paid to an attorney; in Germany the successful plaintiff can recover part of his legal fees from the defendant, which are fixed according to the BRAO.233 In the Netherlands a similar principle applies; the obligation to pay the costs is limited to a certain amount and does not cover all costs;234 in the Netherlands, case law has increased the possibility for victims to claim reasonable costs from the injurer.235 Anyway, compared to the American system, the fact that a losing party can be bound to pay part of the winner’s legal fees will obviously be an important deterrent against frivolous law suits; one can expect a lower number of claims than in the American system where a plaintiff only risks the legal fees he paid to his own attorney in case of failure.236

4. The Judiciary Obviously the role of the judiciary will have an important bearing on whether a claim will be successful or not. We have already mentioned that, to a large extent, the jurisprudential shift from negligence towards strict liability has been caused by the wish of judges to award compensation as soon as


59

2.


physical harm is inflicted. Obviously, the probability that claims will be brought is only affected by the judiciary in an indirect way, e.g. by broadening the negligence standard. The judiciary itself can obviously not decide upon the number of claims that will be brought. However it can play an important supportive role once claims have been brought.

5. Concerns regarding the “willingness to claim” We showed above that in countries like the Netherlands, reforms have been considered which influence the relationship between tort and social security. The reduction of the scope of social security has certainly placed increasing demands on tort law. Interestingly enough, this has recently been a matter of concern to the Dutch government as well. The latter noticed that there was a risk that a “claim culture” would emerge in the Netherlands.237 According to the Dutch government, a claim culture is one whereby citizens would claim against each other regularly and for high amounts to obtain compensation for damage.238 The government indicates a few factors, which may have contributed to such a claim culture and examines some of the instruments which could be used to counter it. Attention is also being given to the relationship between tort law and social law239. The Dutch government indicates that the reduction of social security coverage may have led to an increasing need to use tort liability. In addition, it realises that the attitude of the government towards generalised compensation (to include victims of floods for example) may have given support to the impression that damage suffered must in principle always be compensated. It has been argued in the Dutch literature that automatic compensation funds providing compensation for flood damage, for example, may precisely give rise to such a claim culture.240 Also the literature has warned that deregulation and privatisation in and of social security may lead to an increasing use of tort law.241 Indeed, it was claimed that the reduction of social security had led to a situation whereby in the Netherlands too, tort law was becoming important for “Existenzsicherung”.242 The government’s withdrawal from social security in the domain of primary compensation of healthcare costs and income loss, for example, will inevitably lead to an increasing use of tort law.243 The Dutch government in fact realised that to some extent it itself had caused the increasing use of tort law. It is remarkable that some years ago a discussion took place concerning the deregulation of the Safety at Work Act. It was even suggested that the whole body of safety regulation should be repealed and replaced by tort liability.244 At that time, a lot of attention was paid to the potential blessings of tort law, whereas only now is the other side of the medal (a growing claim culture) being addressed. It is remarkable that a lot of attention is being paid to the danger of the emergence of an “American claim culture” but the Dutch government apparently

60


2.


neglects the fact that the high claim culture in the US may to a large extent be due precisely to the fact that victims of torts in the US will have to use tort law for “Existenzsicherung” whereas that was until recently not the case in the Netherlands.245 In an earlier paper, we claimed that the difference in the level of social security between Europe and the United States is probably one of the most important causes for a difference in “claim culture” between both systems. The number of claims has risen as a result of a deregulation and privatisation in the field of social security. It has thus been argued that a remedy should probably be looked for primarily in a reform of social security rather than in reforms of tort law.246 There is undoubtedly in the Netherlands today a problem in tort law as a result of an increase in claims. The main reason for this increase is probably the fact that Dutch victims today, in contrast with about ten years ago, are now also forced to use tort law to cover primary needs. This ties in with the claim made by George Priest that the American tort and insurance crisis was to a large extent due to the fact that third-party liability insurance and tort law were increasingly used to cover primary needs, for which the tort system would not be suited.247 One can indeed note certain developments within liability law which seem to conflict with the traditional starting points of tort law and stem from the idea that all damage should be compensated at all costs, combined with a reduced social security. The result is a paradox: a “socialisation” of tort law and a “privatisation” of social security. Some Dutch scholars are less pessimistic concerning the growing claim culture. For example, Bergkamp argues that less social security does not necessarily lead to more claims in tort law.248 However, legal doctrine and the trade unions clearly indicate that victims are, as result of state withdrawal from social security, forced to make an increasing use of tort law.249 Even the Dutch government realises that less social security may lead to an increase in claims.250 Bergkamp has equally argued that from a public policy point of view it would make no difference whether compensation is paid via tort law or via social security. This is, in our view, a serious misunderstanding. Tort law and social security are, as we have argued in this paper, different systems, which are not automatically interchangeable. A negative upshot of the cut-back in social security at a time when the thresholds for tort law are also being reduced is that the increasing claims on tort law and liability insurance may indeed endanger insurability.251 Insurers have argued that the increasing use of tort law for compensating primary needs may lead to an unpredictability of claims and hence to uninsurability. In the Netherlands, insurers have already defended themselves against this claim culture by changing the system of coverage, policy conditions etc. As result of these steps, some victims may not receive compensation any longer via tort law.252 If this is combined with reduced social security, some


61

2.


victims will not be compensated by either system any longer and it is very likely that, at least in the field of occupational health but also in other fields, some reforms of financing systems will take place in the Netherlands in the near future, which may lead to a re-allocation of tasks between tort law and social security. To be clear: the mere fact that more claims are brought under tort law is not necessarily negative per se. If all of these additional claims comply with the fundamental principles of tort law, there should be no cause for worry at the policy level. However, increased recourse to tort law (partially as a result of reduced social security) in cases where the use of tort law does not correspond with fundamental principles, may cause problems. This is particularly true if tort law is used, for example, in cases where the establishment of liability could never have positively affected the incentives for prevention, or in cases where a victim receives full compensation notwithstanding his contributory negligence. Examples of these trends have been found in the case law concerning employers’ liability we described above. Such use of tort law may put strong pressure also on liability insurance and therefore endanger the insurability of systemic risks.

Notes 17. See Priest, G., “The current insurance crisis and modern tort law”, Yale Law Journal, 1987, 1521-1590. 18. For the Netherlands see e.g. J. Spier, De maalstroom van het aansprakelijkheidsrecht, Houthoff-lecture, 24 March 1992 (Houthoff, 1992) and for Germany Schmidt-Salzer, J., “Unternehmens-und Mitarbeiterhaftung im deutschen und europäischen Produkt- und Umwelthaftungsrecht” in Ahrens, M. and Simon, J. (eds.), Umwelthaftung, Risikosteuerung und Versicherung, (Erich Schmidt Verlag, 1996), 59 and Kötz,H., Deliktsrecht, (Alfred Metzner, 1991), 5th ed., 15-17, as well as generally Rogers, W.V.H., Spier, J. and Viney, G., “Preliminary observations” in Spier, J. (ed.), The Limits of Liability, Keeping the Floodgates Shut, Kluwer, 1996, 1-2; Spier, J.,“How to keep liability within reasonable limits? A brief outline of Dutch law” in Spier, J. (ed.), The Limits of Liability, 98-99 and Bolt, A.T. and Spier, J., De uitdijende reikwijdte van de aansprakelijkheid uit onrechtmatige daad, advisory-report Nederlandse Juristenvereniging, (Tjeenk Willink, 1996). 19. See with respect to developments in professional liability, the special issues of The Geneva Papers on Risk and Insurance, October 1989 and January 1990 and specifically Whelan, C. and McBarnet, D., “The “Crisis” in Professional Liability Insurance”, The Geneva Papers on Risk and Insurance, 1989, 296. 20. On insurability generally, see Faure, M., “The Limits to Insurability from a Law and Economics Perspective”, The Geneva Papers on Risk and Insurance, 1995, 454. For a more specific example see for instance: Bergkamp, L., “Allocating unknown risk: liability for environmental damages caused by genetically modified organisms”, Tijdschrift voor Milieuaansprakelijkheid, 2000, 61-70 and 104-114.

62


2.


21. See with respect to nuclear accidents OECD, Liability and Compensation for Nuclear Damage, An Iinternational Overview, OECD, 1994; Faure, M. and Skogh, G., “Compensation for damages caused by nuclear accidents: A convention as insurance”, The Geneva Papers on Risk and Insurance, 1992, 499; Deprimoz, J., “Regime juridique des assurances contre les risques nucléaires”, 555 Juris. Classeur, 1995, 1 and with respect to civil liability for marine oil pollution Faure, M. and Heine, G., “The insurance of fines: the case of oil pollution”, The Geneva Papers on Risk and Insurance, 1991, 39 and for recent evolutions Brans, E.H.P., “Liability for ecological damage under the 1992 protocols to the civil liability convention and the fund convention and the oil pollution act of 1990”, Tijdschrift voor Milieuaansprakelijkheid, 1994, 61-67 and 85-91. 22. For a comparison of various conventions see Faure, M. and Van den Bergh, R., Objectieve aansprakelijkheid, verplichte verzekering en veiligheidsregulering, Maklu, 1989, 257-264. 23. The Basel Protocol on Liability and Compensation for Damage resulting from Transboundary Movements of Hazardous Wastes and their Disposal instituted a liability regime with regard to the Basel Convention. However, the Protocol never entered into force due to the fact that no Parties of the Convention ratified, acceded, approved, accepted or formally confirmed it. 24. In this particular case, the channelling of liability for nuclear accidents and oil pollution means that the victim can only base his suit on the statutory liability as provided by the convention and excludes the right to bring a suit under the common tort rule of negligence. 25. Zweigert, K. and Kötz, H., Introduction to Comparative Law, Clarendon Press, 1994, 3rd ed., 655. 26. With respect to liability for nuclear accidents, see Faure, M. and Van den Bergh, R.,“ Liability for nuclear accidents in Belgium from an interest group perspective”, International Review of Law and Economics, 1990, 241. 27. In this connection, see for example, Van Maanen, G.E., “Pleidooi voor verbetering van de rechtspotitie van slachtoffers van kernongevalle”, Nederlands Juristenblad, 1986, 1342 and Van Maanen, G.E., “De civielrechtelijke aansprakelijkheid voor kernongevallen naar Nederlands recht” in Faure, M., (ed.), Aansprakelijkheid voor het nucleaire risico, Maklu, 1993, 19. 28. Note however, that with respect to both oil pollution and nuclear accidents, there is a trend, as a result of political pressure, to amend existing conventions with a view to increasing the available amounts. 29. For comments on this convention see e.g. De Boer, J., “Risicoaansprakelijkheid voor gevaarlijke stoffen en milieuverontreiniging”, Nederlands Juristenblad, 1993, 225; Sterk, C.H.W.M., Verhoogd gevaar in het aansprakelijkheidsrecht, diss. KUB, Kluwer, 1994; Spier, J. and Sterk, C.H.W.M., “The draft convention on civil liability for damage resulting from activities dangerous to the environment”, Tijdschrift voor milieu en recht, 1992, 591 and Bauw, E., Buitencontractuele aansprakelijkheid voor bodemverontreiniging, Kluwer, 1993. and Bergkamp, L., “Allocating unknown risk: liability for environmental damages caused by genetically modified organisms”, Tijdschrift voor Milieuaansprakelijkheid, 2000, 68-69. 30. See in this connection, Messer, E.A., Risicoaansprakelijkheid voor milieuverontreiniging in het WW, diss., RUU, Gouda Quint, 1994 and Spier, J. and Sterk, C.H.W.M., Aansprakelijkheid voor gevaarlijke stoffen, Kluwer, 1995.


63

2.


31. See for a recent discussion concerning a European civil liability regime Banakas, S., “European Tort Law: is it Possible?”, European Review of Private Law, 2002, 363-375. 32. 1985 O.J. L210/29. 33. This notion can be found at several instances in the statements preceding the text; see Faure, M., “Harmonization of product liability law in Europe” in GIRGIS (ed.), L’impresa Europea nel 1992: problemi economici e giuridici nella prospettiva del mercato unico, Il fisco, 1992, 395. 34. Through the “Hühnerpest” case [Bundesgerichtshof 28 November 1968, Neue Juristische Wochenschrift, 1969, 269] the risk of proving the cause of the defect was shifted to the producer. See for a discussion of other German cases on product liability prior to the Directive, Deutsch, E., “Die neuere Entwicklung der Rechtsprechung zum Haftungsrecht”, Juristenzeitung, 1984, 208; Kullmann, H., “Die Rechtsprechung des BGH zur deliktischen Haftung des Herstellers für Schäden an der von ihm hergestellten Sache”, Betriebsberater, 1985, 409 and Kötz, H., Deliktsrecht, 124. See also Brüggemeier, G., “Liability for Water Pollution under German Law: Fault or Strict Liability” in Van Dunné, J. (ed.), Transboundary Pollution and Liability: the Case of the River Rhine, Vermande, 1991, 83-85 and Markesinis, B.S., The German Law of Torts, A Comparative Introduction, Clarendon, 1994, 3rd ed., 79-95. 35. The Dutch Supreme Court had already provided for substantial protection for victims of defective products. See in this connection the wellknown Ford case [Court of Amsterdam 27 June 1957, Nederlandse Jurisprudentie, 1958, 104] and (after implementation of the Directive but based on Dutch tort law) the famous Halcion case [Dutch Supreme Court 30 June 1989, Nederlandse Jurisprudentie, 1990, 652], and the Vrumona case [Dutch Supreme Court 24 December 1993, Nederlandse Jurisprudentie, 1994, 214]. 36. See Haazen, O.A. and Spier, J., in Bolt, A.T. and Spier, J. (eds.), De uitbreidende reikwijdte van de aansprakelijkheid uit onrechtmatige daad, Advisory Report, Nederlandse Juristen Vereniging, Tjeenk Willink, 1996, 33. 37. Bourgoignie argued that the Directive brought about drastic changes for Greece, Italy, Spain and Portugal. Other countries already had – in his view – a product liability system, which equalled the liability system of the EC directive (Bourgoignie, T., “Responsabilité du fait des produits: arguments connus pour un nouveau débat”, Revue Européenne de droit de la consommation, 1987, 17). 38. Fontaine, M. and Bourgoignie, T., Le droit de la consommation en Belgique et au Luxembourg, Van Nostrand Reinhold, 1981, 130. 39. On the consequences of a “dualistic” product liability regime (combining old national and new European law) see Hondius, E.H., “Produktenaansprakelijkheid: de voordelen van een dualistische rechtsorde”, Ars Aequi 1996, 38. 40. For a critical discussion see Stapleton, J., Product Liability, Butterworths, 1994, , A and others v. the National Blood Authority [2001] 3 ALL ER 289, Howells, G. and Mildred, M., “Infected Blood: Defect and Discoverability. A First Exposition of the EC Product Liability Directive”, The Modern Law Review, 2002, 95-106 and Kidner, R., Casebook on Torts, Oxford University Press, 2002 7th. ed., 338-348. 41. On problems of national legal culture and sovereignty, see De Witte, B., “Sovereignty and European Integration: the Weight of Legal Tradition”, Maastricht Journal of European and Comparative Law, 1995, 145, and for a model concerning the division of labour between Europe and the member States, Van den Bergh, R., “The Subsidiarity Principle in European Community Law: Some Insights from Law and Economics”, Maastricht Journal of European and Comparative Law, 1994, 337. The task

64


2.


of the European Union with respect to liability law has been addressed by Kötz, A., “A common private law for Europe: perspectives for the reform of European legal eduction” in De Witte, B. and Forder, C. (eds.), The common law of Europe and the future of legal education, Kluwer, 1992, 31. On this topic see also Barendrecht, J.M., “Produktenaansprakelijkheid: Europees burgerlijk recht?”, Preadvies voor de vereniging van burgerlijk recht, Vermande, 1987. 42. See Bell, S. and Mc Gillivrary, D., Environmental Law, London: Blackstone Press, 2000, 5th ed., 212. 43. On this proposal, see Spier, J., “Wederom: De EEG en gebrekkige diensten”, N ed erl an d s Juri s ten bl ad , 19 91 , 6 63 ; Van Dam, C.C., “De EG-richtlij n dienstenaansprakelijkheid”, Aansprakelijkheid en Verzekering, 1994, 113 and De Boeck, A., “Het voorstel van EG-richtlijn inzake de aansprakelijkheid voor gebrekkige diensten”, Rechtskundig Weekblad, 1993-94, 585, and for an economic analysis of the draft directive, see Curran, C., “The burden of proof and the liability for suppliers of services in the EEC”, The Geneva Papers on Risk and Insurance, 1994, 85, as well as Faure, M., “Enkele rechtseconomische kanttekeningen bij de dienstenaansprakelijkheid”, Aansprakelijkheid en Verzekering, 1994, 33. 44. Van Dam, C.C., “De EG-richtlijn dienstenaansprakelijkheid: afstell”, Aansprakelijkheid en Verzekering 1994, 149, as well as Van Dam, C.C., in Opdracht en Dienstverlening, Series praktijkhandelingen, Tjeenk Willink, 1994. 45. Also in British law it is considered unclear whether the Consumer Protection Act 1987 implementing the Product Liability Directive will actually improve on the common law of negligence (see Mullis, A. and Oliphant, K., Torts, MacMillan, 1993, 132-142). 46. 1990, O.J., L156/14. 47. See in this connection Dommering-van Rongen, L., “Aansprakelijkheid van de werkgever voor toetsenbordletsel (“repetitive strain injury”)”, Aansprakelijkheid en Verzekering, 1995, 27. For employer’s liability with regard to RSI in a Dutch context see Hartlief, T.,“Wat kan artikel 7: 658 BW betekenen voor werknemers met RSI?”, Nederlands Tijdschrift voor Burgerlijk Recht, 2002, 2-6. 48. For a critical analysis of this directive, see Ogus, A.I., “Quality Control for European Regulation”, Maastricht Journal of European and Comparative Law, 1995, 325, especially 330. 49. See Haazen, O.A. and Spier, J., in Bolt, A.T. and Spier, J. (eds.), De uitbreidende reikwijdte van de aansprakelijkheid uit onrechtmatige daad, 29. 50. On this green paper, see Deketelaere, M., “Civielrechtelijke aansprakelijkheid voor milieuschade als instrument van Europees milieubeleid: het “groen boek inzake herstel van milieuschade” van de Commissie van de Europese Gemeenschappen” in Deketelaere, M., (ed.), Recente ontwikkelingen inzake de aansprake lijkheid voor milieuschade, Die Keure, 1993, 99 and see Spier, J., “Green boek on remedying environmental damage”, Aansprakelijkheid en Verzekering, 1993, 13. 51. For comments on this White Paper, see among others Bergkamp, L., “The Commission’s White Paper on environmental liability: a weak case for an EC strict liability regime, European Environmental Law Review, 2000, 105-114 and 141-147; Rice, P.,” “From Lugano to Brussels via Arhus: environmental liability White Paper published,” Environmental Liability, 2000, 39-45 and Rehbinder, E., “towards a community environmental liability regime: the Commission’s White Paper on environmental liability,” Environmental Liability, 85-96 and Arcuri, A., “Controlling


65

2.


Environmental Risks in Europe: the complementary role of an EC environmental liability regime”, Tijdschrift voor Milieuaansprakelijkheid, 2000, 37-45. 52. See Bergkamp, L., “The Commission July 2001 working paper environmental liability: civil or administrative law to prevent and restore environmental harm?”, Environmental Liability, 2001, 207-216 and Brans, E.H.P., “Nieuwe stap in de ontwikkeling van een EU milieuaansprakelijkheidsrichtlijn”, Aansprakelijkheid, Verzekering en Schade, 2002, 3-11 and Brans, E.H.P., “Voorstel van de Europese Commissie voor een richtlijn milieuaansprakelijkheid (COM (2002) 17)”, Milieu en Recht, 2002, 136-144. 53. Joint cases C-6/90 and 9/90 Francovich, 1991, ECR, I 5357. 54. C-46/93 Brasserie du Pêcheur [1996] ECR I- 1029, C-48/93 Factortame III [1996] ECR I – 1029, C-392/93 British Telecommunications [1996] ECR I-1631. 55. For a discussion of Community tort law after Francovich and Brasserie du Pêcheur see Van Gerven, W., “Bridging the unbridgeable”, International Comparative Law Quarterly, 1996, forthcoming, building on his earlier paper “Bridging the gap between Community and national laws, towards a principle of homogeneity in the field of legal remedies”, Common Market Law Review, 1995, 679. 56. Important differences exist between legal systems, amongst others regarding the question of what is considered to be a protected interest in tort law. We will discuss some of these issues in Ch. 5.C. See also Berkamp,L., Liability and Environment, The Hague: Kluwer Law International, 200152-56. 57. See Bell, S. and Mc Gillivrary, D., Environmental Law, London, Blackstone Press, 2000, 5th ed., 287. 58. Brüggemeier, G., in Van Dunné, J. (ed.), Transboundary Pollution and Liability, 85-91. 59. Brüggemeier, G., in Van Dunné, J. (ed.), Transboundary Pollution and Liability, 91-92; Simon, J., “Das Umwelthaftungsgesetz” in Ahrens, M. and Simon, J. (eds.), Umwelthaftung, Risikosteuerung und Versicherung, Erich Schmidt Verlag, 1996, 13 and Wagner, G., “Die Aufgaben des Haftungsrechts – eine Untersuchung am Beispiel des Umwelthaftungsrechts – Reform’, Juristenzeitung, 1991, 175. A translation of the German Environmental Liability Act can be found in Markesinis, B.S., The German Law of Torts, A Comparative Introduction, 871. 60. For a comment see Pagh, P., “The New Danish Act on Strict Liability for Environmental Damage”, Environmental Liability, 1995, 15. According to Betlem, liability under this Act is already excluded if the conditions of the licence are followed (Betlem, G., “It’s no use crying over spilt chemicals”, Maastricht Journal of European and Comparative Law, 1995, 302); this would indeed seriously limit the scope of liability. 61. See Spier, J., in Spier, J. (ed.), The Limits of Liability, 121-123, Von Bar, C., The Common European Law of Torts (vol.2), Oxford: Clarendon Press, 2000, 431, 432 and Betlem, G. “Strict Environmental Liability and NGO Damages and Enforcement of Claims: a Dutch International Law Perspective’, European Environmental Law Review, 2001, 315. 62. Kötz, H., Deliktsrecht, 1991, 130; Kötz, H., “Strict Liability in German Law” in In Memoriam Jean Limpens, Kluwer, 1987, 75; Markesinis, B.S., The German Law of Torts, 692-720; Zweigert, K. and Kötz, H., Introduction to Comparative Law, 693-699. 63. For an overview see Bocken, H., “La responsabilité sans faute en droit belge” in In Memoriam Jean Limpens, 85.

66


2.


64. See generally on the conditions for insurability Faure, M., “The Limits to Insurability from a Law and Economics Perspective”, The Geneva Papers on Risk and Insurance, 1995, 454. 65. If at least the other conditions of insurability (control of moral hazard, avoidance of adverse selection, availability of sufficient coverage and a competitive insurance market) are met. We will come back to this issue in further detail in Ch. 4. 66. But one should note that Art. 17 of the European Product Liability Directive explicitly states that the Directive only applies to products put into circulation after the the legislation implementing the Directive came into force (see Storm, P., “Een gebrekkig produkt”, Tijdschrift voor Vennootschappen, Verenigingen en Stichtingen, 1985, 245). 67. Note, however, that this might cause perverse distributional effects in as much as the buyers of the product are not the potential victims. This means that customers pay a higher price for increased protection provided to third parties (see Faure, M. in GIRGIS (ed.), L’impresa Europea nel 1992, 395). 68. A similar trend could be noted in the United States; see Priest, G., “The Invention of Enterprise Liability: A Critical History of the Intellectual Foundations of Modern Tort Law”, Journal of Legal Studies, 1985, 461. 69. One is left with Priest”s assertion that the shift towards enterprise liability will lead to an increased use of thirdparty liability insurance; it is less easy to control adverse selection with such insurance than with first-party insurance under the negligence rule (see Priest, G., “The Current Insurance Crisis and Modern Tort Law”, Yale Law Journal, 1987, 1521 and see the criticism by Viscusi, K., “The Dimensions of the Product Liability Crisis”, Journal of Legal Studies, 1991, 147). 70. For this trend in the Netherlands, see for example: Van Maanen, G.E., Spier, J. and Sterk, C.H.W.M., “Van schuld naar risico? Enkele opmerkingen over ontwikkelingen in het aansprakelijkheidsrecht” in Raaymakers, M.J.G.C. (eds.), Aansprakelijkheden, Kluwer, 1990, 65 and for Belgium see Bocken, H., “Van fout naar risico. Een overzicht van de objectieve aansprakelijkheidsregelingen naar Belgisch recht”, Tijdschrift voor Privaatrecht, 1984, 329. 71. See the references in notes 16 and 17. 72. Margereson and Hancock v. J.W. Roberts, Queens Bench Division, Leeds, 27 October 1995, Product Liability International, 1995, 179; ENDS Report, 251, December 1995, 40. 73. Pasicrisie belge, 1939, I, 223. 74. And is, therefore, bound to compensate all damage to the victim according to Art. 1645 of the Belgian civil code. 75. Cour de Cassation, 6 October 1961, Pasicrisie belge, 1962, I, 152. 76. Cour de Cassation, 1 July 1969, Dalloz, 1970, 40; Cour de Cassation, 20 January 1970, La semaine Juridique, 1972, 17280; Cour de Cassation, 15 November 1971, Dalloz, 1972, 211. 77. See Spier J. in Bolt, A.T. and Spier, J., De uitdijende reikwijdte van het aansprakelijkheidsrecht?, 250-266. 78. Van Dunné, J.M., “De rechtspraak inzake milieuaansprakelijkheid uit onrechtmatige daad: van schuldbeginsel naar risicobeginsel”, Tijdschrift voor Milieuaansprakelijkheid, 1988, 33. He claims that in case of a creation of a


67

2.


dangerous situation a “pseudo strict liability” applies (Van Dunné, J.M., “Environmental liability: continental style”, Review of European Community and International Environmental Law, 1992, 396). 79. On Dutch soil pollution cases, see also: Taams, R. and Uithoorn, M., “The Dutch Soil Pollution Cases: Shell, Sovay, Duphar and Fasson/Van den Brink”, Environmental Liability, 1995, 39. 80. See Vranken, J.B.M., “Zorgvuldigheidsnorm en aansprakelijkheid voor bodemverontreiniging uit het verleden”, Weekblad voor Privaatrecht Notariaat en Registratie (not dated), 5953 and see Van Dunné”s reaction: Van Dunné, J.M., “Een kamikazeactie op De Rotte. De visie van Vranken op de aansprakelijkheid uit art. 1401 BW, in het bijzonder bij bodemvervuiling uit het verleden”, Weekblad voor Privaatrecht Notariaat en Registratie (not dated), 5976. 81. See in this respect also Spier, J., De maalstroom van het aansprakelijkheidsrecht, 34. 82. See Zweigert, K. and Kötz, H., Introduction to Comparative Law, 666-667 and Mullis, A. and Oliphant, K., Torts, 195-200; Von Bar, C., The Common European Law of Torts (vol.2), Oxford: Clarendon Press, 2000, 387-389; Bell, S. and McGillivray, D., Environmental Law, London: Blackstone Press, 2000, 5th ed., 270-272; see also the Cambridge water case, discussed in the next section. 83. Discussed by Markesinis, B.S., The German Law of Torts, A Comparative Introduction, 880-881. 84. For Dutch case law, see ,for example, the so-called Jumbo II case of 1 October 1993 (Nederlandse Jurisprudentie, 1995, 182) and Hartlief, T., “Open normen in het schadevergoedingsrecht; de artikelen 6:98, 99 en 101 BW”, Weekblad voor Privaatrecht, Notariaat en Registratie, 2002, 8-10. See also Restatement of the Law. Torts: Liability for Physical Harm (Basic Principles), Tentative Draft No. 1, The American Law Institute, March 28, 2001, 76. 85. So Zweigert, K. and Kötz, H., Introduction to Comparative Law, 690. 86. See e.g. Bolt, A.T. in Bolt, A.T. and Spier, J., De uitdijende reikwijdte van het aansprakelijkheidsrecht?, 118-160; Dam, C.C. van en Wessels, B., (eds.), Opdracht en Dienstverlening, serie Praktijkhandleidingen, Tjeenk Willink, 1994; Arisz, F.H.A., “Inleiding: Beroepsaansprakelijkheid”, in Arisz, F.H.A. (ed.), Beroeps aansprakelijkheid, Recht op een scheve schaats, Bundel ter gelegenheid van het Jonge Balie Congres 1991, Tjeenk Willink, 1991, 1 and see specifically for the professional liability of the notary public Wessels, B., “Beroepsaansprakelijkheid van de notaris”, Weekblad voor Privaatrecht Notariaat en Registratie (not dated), 5992. See generally also Stolker, C.J.J.M., “Aansprakelijkheid voor beroepsfouten; van droomcarrière naar nachtmerrie”, Ars Aequi, 1995, 13; see also note 3 and from a comparative perspective Zweigert, K. and Kötz, H., Introduction to Comparative Law, 690-693 and for Sweden see Dufwa, B., “Strict liability in tort law”, Svensk Juristtidning, 1987, 269. 87. See for a comparative overview of this development in several European countries, Faure, M. and Koziol, H. (eds), Torts and Insurance Law, vol.1, Cases on Malpractice in a Comparative Perspective, Wien: Springer, 2001. 88. See Hoge Raad, 21 February 1997, [1999] NJ 145 (Wrongful Birth). 89. See for instance Van Maanen, G.E., “Onrechtmatig leven”, Nederlands Tijdschrift voor Burgerlijk Recht, 2001, 57, Nieuwenhuis, J.H., “De dag verga, waarop ik geboren werd”, RM Themis, 2001, 97-98, Schoordijk, H.C.F., “Wrongful life mede vanuit rechtsvergelijkend perspectief”, Nederlands Tijdschrift voor Burgerlijk Recht, 2001, 212 and Faure, M. and Hartlief, T., in Spier, J., Yearbook on Insurance and Liability, to be published.

68


2.


90. See the cases of the Cour de Cassation of 13 July 2001, respectively the numbers 97.-17.359, 97.-19.282, 98.-19.190 to be found on www.courdecassation.fr/agenda/ arrets/arrets and Vranken, J.B.M., “Wrongful life in Frankrijk: drie recente uitspraken”, Weekblad voor Privaatrecht, Notariaat en Registratie, 2001, 752-755. 91. We will come back to this immobility issue in Ch. 3. 92. See Van Boom, W.H., “Anticiperen op nieuwe Aansprakelijkheid, Verzekering en Schade, 2001, 7-11.

gezondheidsrisico’s”,

93. The (un)insurability of retroactive liability will be discussed in Ch. 4. 94. Some have therefore argued that actions to recover clean-up costs in fact have the character of a lottery. 95. For examples in Dutch law see Hartlief, T. and Spier, J., “Verzekering en aansprakelijkheid met terugwerkende kracht", Aansprakelijkheid and Verzekering, 1994, 27 and Haazen, O.A., and Spier, J., Bolt, A.T. and Spier, J. (eds.), De uitbreidende reikwijdte van de aansprakelijkheid uit onrechtmatige daad, 45-46. 96. Dutch Supreme Court, 24 April 1992, Tijdschrift voor Milieuaansprakelijkheid, 1992, 131; Milieu en Recht 1992, nr. 87; Nederlands Juristenblad, 1992, nr. 121. 97. Soil Protection Act of 10 May 1994; on its importance, see: Spier, J., “De Novelle inzake art. 47 lid 5 ontwerp inbouwwet bodemsanering”, Aansprakelijkheid en Verzekering, 1994, 43 and Betlem, G., Maastricht Journal of European and Comparative Law, 1995, 300. 98. See Interuniversitaire Commissie tot Herziening van het Milieurecht in het Vlaamse Gewest, Voorontwerp Decreet Milieubeleid, Die Keure, 1995; the Commission”s proposals concerning civil liability have been discussed by Deketelaere, K., “De voorstellen van de Interuniversitaire Commissie tot Herziening van het Milieurecht in het Vlaamse Gewest inzake aansprakelijkheid voor milieuschade” in Deketelaere, M. (ed.), Recente ontwikkelingen inzake de aansprakelijkheid voor milieuschade, Die Keure, 1993, 75. 99. According to Art. 31 Ch. 2 of the Soil Clean-up Decree of 22 February 1995, there is no duty to clean up in cases of historic pollution if the user of the soil was in good faith (see Deketelaere, M., “Aansprakelijkheid voor historische, nieuwe en gemengde bodemverontreiniging” in Deketelaere, K. (ed.), Het decreet betreffende de bodemsanering, Die Keure, 1995, 135). 100. Journal of Environmental Law, 1994, 137 with case note by Ogus. 101. For a discussion see Betlem, G., Maastricht Journal of European and Comparative Law, 1995, 289 and Rogers, W.V.H., “Liability for environmental pollution in the common law: the Cambridge water case”, Aansprakelijkheid en Verzekering, 1994, 64 ff. 102. See Jones, B., “The identification and remediation of contaminated sites: the United Kingdom’s Environment Act 1995”, Tijdschrift voor Milieuaansprakelijkheid, 1995, (159), 163. 103. See Frenk, N., “Toerekening naar kansbepaling”, Nederlands Juristenblad, 1995, 482. 104. See for example Banakas, S., “European Tort Law: is it Possible?”, European Review of Private Law, 2002 363-375, in which a common European Law of civil liability is discussed. The author considers the market-share-liability an important development in (American) tort law, which should be adopted in a new European Law of civil liability. An example of a market-share-approach can be found in the


69

2.


Tobacco Damages and Health Care Cost Recovery Act [SBC 2000] chapter 30 of British Columbia, Canada, section 7. 105. For a discussion of alternative causation under German law see Köndgen, J., “Multiple causation and joint tortfeasors in pollution cases according to German law”, in Van Dunné, J.M. (ed.), Transboundary Pollution and Liability, the Case of the River Rhine, Vermande, 1991, 99 and the interesting Article of Bodewig, Th., “Probleme alternativer Kausalität bei Massenschäden”, Archiv für die civilistische Praxis, 1985, 505. 106. Note, however, that the Dutch Supreme Court only considered the causation question. Formally, it still has to be decided whether bringing DES onto the market was wrongful per se. See Dutch Supreme Court, 9 October 1992, Nederlandse Jurisprudentie, 1994, 535 (C.J.H.B.). See on this case Spier, J. and Wansink, J.H., “Joint and Several Liability of DES Manufacturers: a Dutch Tort Crisis”, International Insurance Law Review, 1993, 176 and Hartlief, T., “Open normen in het schadevergoedingsrecht; de artikelen 6:98, 99 en 101 BW”, Weekblad voor Privaatrecht, Notariaat en Registratie, 2002, 8-10. 107. See Spier, J., in Spier, J. (ed.), The Limits of Liability, 123-124. 108. See Spier, J., in Spier, J. (ed.), The Limits of Liability, 124-125. 109. Dutch Supreme Court 25 June 1993, Nederlandse Jurisprudentie, 1993, 686. 110. Hoge Raad, 17 November 2000, [2001] NJ 596, (Unilever-Dikmans). See for a comment on this case Van Dunne, J.M., “Schending van veiligheidsnormen en causaliteit: conditio sine qua non als rudimentair vereiste”, Tijdschrift voor Milieuaansprakelijkheid, 2001, 6. 111. See Hoge Raad, 29 June 2001, [2001] NJ 476, (Industromontaza Odo/ Banfic). This exception was however not applied in this particular case. See also Faure, M. and Hartlief, T., in Spier, J., Yearbook on Insurance and Liability, to be published. 112. See Supreme Court of Canada, Athey v. Leonati [1996] 3 SCR 458, Supreme Court of Canada Farell v. Snell [1990] 72, D.L.R. (4th) 289 and Tillemann, W.A., Lucas, A.R. and Hughes, E., Environmental law and policy, Toronto: Edmond Montgomery Publications, 1998, 114. 113. Alberta Court of Queen”s Bench Jones v Mobil Oil Canada Ltd. [1999] A.J. No. 979. 114. Trauberman, J., “Statutory reform of “toxic torts”. Relieving legal, scientific and economic burdens on the chemical victim”, Harvard Environmental Law Review, 1983, 177 and Tillemann, W.A., Lucas, A.R. and Hughes, E., Environmental law and policy, Toronto: Edmond Montgomery Publications, 1998, 101-122. 115. Estep, E., “Radiation injuries and statistics: the need for a new approach to injury litigation”, Michigan Law Review, 1960, 259; Van, A., “Statistisch bewijs van causaal verband” Discussion of the High Court of Justice 8 October 1993 (Sellafield Claims), Tijdschrift voor Milieuaansprakelijkheid, 1994, 109 and Van, A., Onzekerheid over daderschap en causaliteit, Gouda Quint, 1995, 145-154. 116. Gardner, M., “Results of a case-control study of leukaemia and lymphoma among young people near Sellafield nuclear plant in West Cumbria”, British Medical Journal, 1990, 423. 117. For a discussion of that case, see Lavrysen, L., “Judicial responses in the nineties to Dutch (and German) shipments of waste to Belgium in the eighties”, Maastricht Journal of European and Comparative Law, 1995, 219.

70


2.


118. Nevertheless, this has been argued by some authors. See e.g. Van Maanen, G.E., “De civielrechtelijke aansprakelijkheid voor kernongevallen naar Nederlands Recht” in Faure, M. (ed.), Aansprakelijkheid voor het Nucleaire Risico, Maklu, 1993, 19. 119. For a further discussion of these problems, see: Faure, M., (G)een schijn van kans. Beschouwingen over het statistisch causaliteitsbewijs bij milieu- en gezondheidsschade, Maklu, 1993. 120. This proportionate liability rule has been defended by several American scholars (Rosenberg, D., “The causal connection in Mass Exposure cases: a ‘public law’ vision of the tort system”, Harvard Law Review, 1984, 851) and is also defended in the economic analysis of law (Shavell, S., “Uncertainty over causation and the determination of civil liability”, Journal of Law and Economics, 1985, 587). The Dutch Attorney General Hartkamp also defended a market share liability in the DES case (Tijdschrift voor Consumentenrecht, 1992, 241). In addition Spier pleaded in favour of proportionate liability for latent diseases in his inauguration address, Sluipende schade, Kluwer, 1990. 121. See Brüggemeier, G., in Van Dunné, J. (ed.), Transboundary Pollution and Liability, 88-91. 122. Robinson, G., “Probabilistic causation and compensation for tortious risk”, Journal of Legal Studies, 1985, 798. 123. This point has been rightly mentioned by Betlem, G., Maastricht Journal of European and Comparative Law, 1995, 297; see also Rogers, W.V.H., Spier, J. and Viney, G., in Spier, J. (ed.), The Limits of Liability, 13 and for French law: RémardGouillond, M., “Faute ou Risque?” in Van Dunné, J.M. (ed.), Transboundary Pollution and Liability, Vermande, 1991, 67. 124. Zweigert, K. and Kötz, H., Introduction to comparative law, 605, 606, discuss the role of the English judiciary in expanding liability to provide victim compensation. 125. Betlem, G., Maastricht Journal of European and Comparative Law, 1995, 303-304. 126. Brüggemeier, G., “Judizielle Schutzpolitik de lege lata – Zur Restrukturierung des BGB-Deliktsrechts”, Juristenzeitung, 1986, 972. Although Betlem favours a statutory no-fault liability, “given the inherent uncertainty in judicial lawmaking” (Betlem, G., Maastricht Journal of European and Comparative Law, 1995, 305). 127. Although we stress the preventive function of tort law in this chapter, Calabresi has pointed at the fact that liability rules may equally aim at loss spreading (Calabresi”s secondary costs) and at a reduction of administrative costs (Calabresi”s tertiary costs). See G. Calabresi, The Costs of Accidents. A Legal and Economic Analysis (1970). 128. See on the goals of tort law G. Schwartz, The Ethics and the Economics of Tort Liability Insurance, [1990] Cornell Law Review (CLR), 313-365 and G. Schwartz, Mixed Theories of Tort Law: Affirming both Deterrence and Corrective Justice, [1997] 75 TLR, 1801-1834. 129. See G. Calabresi, The Costs of Accidents. A Legal and Economic Analysis (1970). See also S. Shavell, Economic Analysis of Accident Law (1987), pp. 5-6. 130. This is the basic model presented by S. Shavell in Strict Liability versus Negligence, [1980] JLS, 1-25. 131. J.P. Brown, “Toward an Economic theory of liability”, [1973] Journal of Legal Studies (JLS), pp. 323-349.


71

2.


132. G. Calabresi (supra note 129) p. 17. 133. Otherwise the Coase theorem would teach us that the efficient outcome would automatically follow through voluntary negotiations (R.H. Coase, The Problem of Social Cost, [1960] 1 Journal of Law and Economics (JLE), 1-44). 134. See S. Shavell (supra note 129) p. 14. 135. See S. Shavell, “Strict Liability versus Negligence”, [1980] JLS, 1-25 and W. Landes/ R. Posner, “The Positive Economic Theory of Tort Law”, [1981] Georgia Law Review (GLR), 851-924. 136. White Paper on environmental liabiliy, p. 14. 137. This claim has especially been made by W. Landes/R. Posner (supra note 135) p. 851. 138. 159F 2d 169 (2d Cir. 1947). 139. See for a discussion of this Learned Hand case also R. Cooter/Th. Ulen, Law and Economics, 3rd edition (2000), pp. 313-316; M. Gimpel-Hinteregger, Grundfragen des Umwelthaftung, (1994), pp. 51-56 and G. Schwartz, “Mixed Theories of Tort Law: Affirming both Deterrence and Corrective Justice”, [1997] 75 TLR, p. 1819. 140. See W. Landes/R. Posner (supra note 135) pp. 884-885. 141. J.P. Brown, “Toward an Economic theory of liability”, [1973] Journal of Legal Studies (JLS), 323-349. 142. W. Landes/R. Posner (supra note 135) p. 885; see also R. Cooter/Th. Ulen Law and Economics, 3rd edition (2000), pp. 313-316p. 283. 143. See more particularly the paper by I. Gilead, Tort Law and Internalization. The Gap between Private Loss and Social Cost, [1997] 17 International Review of Law and Economics (IRLE), 589-608, who showed that the Hand formula disregards third parties and positive externalities. 144. A.M. Polinsky, Introduction to Law and Economics, (1983) p. 39; S. Shavell, “Strict Liability versus Negligence”, [1980] JLS, p. 11 and S. Shavell, Economic Analysis of Accident Law (1987), p. 8. 145. P. Diamond, “Single Activity Accidents”, [1974] JLS, p. 974; A.M. Polinsky (supra note 144) p. 44 and S. Shavell (supra note 144). 146. It is difficult to determine both the optimal and the actual activity level (for examples, see A.M. Polinsky (supra note 144) p. 47 and S. Shavell (supra note 144) p. 2). 147. P. Diamond (supra note 145); W. Landes/R. Posner (supra note 135); A.M. Polinsky (supra note 144) and S. Shavell (supra note 144). 148. W. Landes/R. Posner (supra note 135) p. 877 and S. Shavell (supra note 144) p. 7 and p. 19. 149. W. Landes/R. Posner (supra note 135) p. 907. 150. For an overview of all legal arguments in favor of strict environmental liability see B. Jones, Deterring, Compensating and Remedying Environmental Damage: the Contribution of Tort Liability, in: P. Wetterstein (ed.), Harm to the Environment: The right to compensation and the assessment of damages (1997), pp. 11-27. 151. See for Dutch case law e.g. the so called Jumbo II case of 1 October 1993 (Nederlandse Jurisprudentie (1995), p. 182).

72


2.


152. See with respect to nuclear accidents T. Vanden Borre, Efficiënte preventie en compensatie van catastroferisico’s. Het voorbeeld van schade door kernongevallen (2001); OECD, Liability and Compensation for Nuclear Damage, An International Overview, (1994); M. Faure/G. Skogh, Compensation for Damages Caused by Nuclear Accidents: A Convention as Insurance, [1992] The Geneva Papers on Risk and Insurance (GPRI), 499-513; J. Deprimoz, Régime Juridique des Assurances contre les Risques Nucléaires, [1995] Juris. Classeur, 1-24 and with respect to civil liability for marine oil pollution M. Faure/G. Heine, The Insurance of Fines: The Case of Oil Pollution, [1991] GPRI, 39-54 and for recent evolutions E.H.P. Brans, Liability for Ecological Damage under the 1992 Protocols to the Civil Liability Convention and the Fund Convention and the Oil Pollution Act of 1990, [1994] TMA, 61-67 and 85-91. 153. See also B. Pozzo Zanchetta, The Liability Problem in Modern Environmental Statutes, [1996] 4 ERPL, 112-129. 154. See, for an application to nuclear liability: M. Faure, Economic Models of Compensation for Damage Caused by Nuclear Accidents: Some Lessons for the Revision of the Paris and Vienna Conventions, [1995] European Journal of Law and Economics (EJLE), 21-43. 155. Although it is then probably more a case where more parties can influence the accident risk (and should therefore be given appropriate incentives) since the actors in the example given can not be considered traditional victims who suffer the loss personally (see also G.J. Niezen, “Aansprakelijkheid voor milieuschade in de Europese Unie” in: Ongebonden Recht Bedrijven (2000) p. 171). 156. In some cases it will be the victim”s activity that caused the harm, e.g. if the victim knowingly came to the nuisance. This may then lead to a denial of a claim on compensation. See in that respect the discussion on the coming to the nuisance doctrine, by D. Wittman, First come, first served: an economic analysis of “coming to nuisance”, [1980] JLS, 557-568. 157. See with respect to environmental liability A. Endres/B. Staiger “Ökonomische Aspekte des Umwelthaftungsrecht” in: M. Ahrens/J. Simon (eds.), Umwelthaftung, Risikosteuerung und Versicherung (1996), pp. 79-93 and A. Monti, “Environmental Risk: A comparative law and economics approach to liability and insurance”, [2001] ERPL, pp. 56-57. 158. See A. Endres/R. Schwarze, Allokationswirkungen einer Umwelthaftpflichtversicherung, [1991] Zeitschrift fürUmweltpolitik und Umweltrecht, 1-25. 159. If, in other words, courts can more easily observe the socially desirable level of precaution, than the exact amount of external harm, a negligence rule should be favored. This point has been made by R. Cooter, “Prices and Sanctions”, [1984] 84 Columbia Law Review, 1343-1523. 160. See I. Gilead, “Tort Law and Internalization. The Gap between Private Loss and Social Cost”, [1997] 17 International Review of Law and Economics (IRLE), p. 589. 161. S. Gilles, “Rule-Based Negligence and the Regulation of Activity Levels”, [1992] 22 JLS, 319. 162. See M. Faure/T. Hartlief, “Remedies for Expanding Liability”, [1998] 18 Oxford Journal of Legal Studies (OJLS), 681-706. 163. M.J. Trebilcock, “The Social Insurance-Deterrence Dilemma of Modern North American Tort Law: A Canadian Perspective on the liability Insurance Crisis”, [1987] 24 San Diego Law Review (SDLR), 929-1002.


73

2.


164. Causal uncertainty will be discussed below in relation to its importance for insurability. See Chapter 4 § 6. 165. For some of the basic insights in the economic analysis of tort law, see G. Calabresi, The costs of accidents, 1970 and S. Shavell, Economic analysis of law, 1987. Differences between social insurance and liability law are also addressed from an economic perspective by M.J. Trebilcock, The Social Insurance-Deterrence Dilemma of Modern North American Tort Law: A Canadian Perspective on the Liability Crisis, 1987, 24, San Diego Law Review, 929-1002. 166. This is e.g. the case in Calabresi”s notion of secondary costs, which aims at the loss spreading. 167. The various goals of tort law have, from a legal perspective, also been elaborated in the famous handbook on Austrian tort law of H. Koziol, Österreichisches Haftpflichtrecht (Band I, Allgemeiner Teil, 3rd. edn. 1997), pp. 8-13. 168. See G. Schwartz, Mixed theories of tort law: affirming both deterrence and corrective justice, 1997, 75 Texas Law Review, 1804-1834. 169. The way the particular system is organised obviously differs a lot and is dependent upon the particular legal system. Pieters, Introduction in the basic principles of social security, 1993, and D. Pieters, (ed.), Introduction into the social security law of the member States of the European community, 1993. 170. This principle that social security provides compensation irrespective of the behaviour of the beneficiary may well be different in some legal systems; one can imagine cases where compensation under social security is denied if e.g. the personal injury was caused intentionally. 171. See S. Klosse, Schadeleer: geschikt fundament voor sociale zekerheid in de 21 e eeuw?, in S. Klosse (ed.) Sociale zekerheid: een ander gezichtspunt. Toekomstperspectief vanuit vier disciplines, 2000, p. 12. Prevention in social security is equally stressed by J. van Kessel, Schade bij werknemers: aandachtspunten voor deze eeuw. Betere werking van het huidige stelsel door nauwere afstemming op schadeleer, meer flexibiliteit en verdergaande integratie, in S. Klosse (ed.) Sociale zekerheid: een ander gezichtspunt. Toekomstperspectief vanuit vier disciplines, 2000, pp. 257-260. We come back to the importance of regulation in ch. 6. 172. S. Klosse (supra note 171) pp. 8-9. 173. See T. Hartlief, Ieder draagt zijn eigen schade, 1997, p. 28. 174. See for criteria for safety regulation S. Shavell, Liability for harm versus regulation of safety, 1984, Journal of Legal Studies, 57-374 and S. Shavell, A model of the optimal use of liability and safety regulation, 1984, Rand Journal of Economics, 271-280. 175. See G. Skogh, Public insurance and accident prevention, 1982, 2 The International Review of Law and Economics, 67-80 and G. Skogh, The combination of private and public regulation of safety, in M. Faure/R. Van den Bergh (eds.), Essays in law and economics. Corporations, accident prevention and compensation for losses, 1989, pp. 87-101. 176. See H. Koziol, Die Arzthaftung im geltenden und künftigen Recht in: Haftungsrechtliche Perspektiven der ärztlichen Behandlung, 1997, pp. 21-35. 177. This has also been proven empirically (by calculating the specific contribution of tort law compared to social security) by A.R. Bloembergen, De invloed van verzekeringen in: Schade lijden en schade dragen, 1980, pp. 16-17. See for a brief overview of the relation between social security and tort law in European

74


2.


countries Koch, B.A. and Koziol, H. (eds.), Unification of Tort Law: Strict Liability, London, Kluwer Law International, 2002. 178. T. Hartlief (supra note 173) pp. 29-30. 179. T. Hartlief (supra note 173), p. 29. 180. See M. Faure/T. Hartlief, Verzekering en financiering van beroepsziekten: enkele tips voor de SER, 1998, Nederlands Juristenblad, 1135. 181. See with respect to the American health care reform debate some of the relevant insurance economics issues discussed by H. Aaron, Issues Every Plan to Reform Health Care Financing must Confront, 1994, 3 Journal of Economic Perspective, 31-43; R. Frank, Arch. Phys. Med. Rehabil, 120-124 and M. Pauly, Taxation, Health Insurance and Market Failure in the Medical Economy, 1986, Journal of Economic Literature, 629-675; see also the recent survey by J. Newhouse, Reimbursing Health Plans and Health Providers: Selection versus Efficiency in Production, 1996, 34 Journal of Economic Literature, 1236-1263; see equally the discussion of this debate from a public finance perspective by R. Zeckhauser, Public Finance Principles and National Health Care Reform, 1994, Journal of Economic Perspective, 55-60. 182. See P. van Wijngaarden/F. van den Heuvel, Privatisering sociale zekerheid ter discussie, 1995, 12 Sociaal Bestek, 2-6. 183. Although it has been argued that this evolution may put too heavy a burden on tort law as well; M. Faure/T. Hartlief, Towards an Expanding Enterprise Liability in Europe? How to Analyze the Scope of Liability of Industrial Operators and Their Insurers, 1996, Maastricht Journal of European and Comparative Law, 235-270. 184. See on the importance of the reduced social security for the scope of liability also S. Klosse/G.Vonk (supra note 171) p. 204-217. 185. See also G. Brüggemeier, Judizielle Schutzpolitik de lege lata – Zur Restrukturierung des BGB-Deliktsrechts, 1986, Juristenzeitung, 969. 186. See e.g. the European Social Charter of 18 October 1961, and ILO-conventions nr. 24 concerning sickness insurance for workers in industry and commerce and domestic servants, nr. 25 concerning sickness insurance for agricultural workers and nr. 102 concerning minimum standards of social security. 187. For an overview see P.F. van der Heijden, Privatisering van de sociale zekerheid: begrensd door (internationale) sociaalrechtelijke normen in: W.J.P.M. Fase (ed.), Sociale zekerheid: Privaat of Publiek, 1994, p. 58. 188. For an economic explanation see M. Adams, Warum kein Ersatz von Nichtvermögensschäden? in: C. Ott/H.B. Schäfer (eds.), Allokationseffizienz in der Rechtsordnung, 1989, p. 210. 189. See for example O.A. Haazen/J. Spier/A.T. Bolt (eds.), De uitdijende reikwijdte van de aansprakelijkheid uit onrechtmatige daad, 1996, p. 32 and p. 44. 190. See for the Netherlands D. de Wolff, Internationaal recht verbiedt privatisering ziektewet, 1993, Nederlands Juristenblad, 1517. 191. For further details, see M.G. Faure/A.J.C.M Geers/T. Hartlief (eds.), Verzekering en de groeiende aansprakelijkheidslast. Een juridisch, gezondheidskundig en econom isch onderzoek naar ontwikkeling en met betrekking tot de aansprakelijkheidslast en de consequenties voor verzekeraars naar aanleiding van de werkgeversaansprakelijkheid voor bedrijfsongevallen en beroepsziekten, 1995, pp. 38-40.


75

2.


192. For further details see S. Klosse/G. Vonk (supra note 171) p. 207-208. 193. Indeed, the employer still has a right of redress for the amounts he paid to his employee against a liable third party on the basis of art. 6:107 a BW. 194. In practice a 100% compensation is often still provided based on Collective Labour Agreements. 195. In fact for the 70% (or more) he paid the employer has an incentive to sue. Indeed due to recent legislative action he has a right of redress (art. 6:107a BW). 196. See e.g. M. Faure, Rechtseconomische kanttekeningen bij de deregulering van de arbeidsomstan-dighedenwet, 1995, Sociaal Recht, 140. See also R.A. Ball/C.J.J.M. S t o l k e r, D e reg u l e re n i n A r b o - l a n d : e e n h e ro r i ë n t a ti e o p de arbeidsomstandighedenwet, 1997, 22 Nederlands Juristenblad, 969-975. 197. This does not seem very likely, since a total deregulation of the Safety at the Workplace Act would collide with many European directives. 198. The precise effects of these changes are hard to predict. Indeed, one could also argue that the current system with very detailed, precise norms should make it even easier for victims to prove the employer”s negligence compared to a system where a breach of a general duty of care has to be proven. 199. For further details see M. Faure/T. Hartlief, 1996, Maastricht Journal of European and Comparative Law, 235-270. 200. See on this reduced burden of proof for the injured employee Dutch Supreme Court 10 December 1999, 2000, Nederlandse Jurisprudentie, 211 and Hoge Raad 15 December 2000, [2001] NJ, 198 (Van Uitert/ Jalas). 201. See in this respect L. Bier, Aansprakelijkheid voor bedrijfsongevallen, dissertation, 1988; S. Klosse, Bedrijfsongevallen en beroepsziekten: individuele of collectieve verantwoordelijkheid?, 1993, Nederlands Juristenblad, 1608 et seq.; S.D. Lindenbergh, Arbeidsongevallen en beroepsziekten, 2000, and T. Hartlief, Van art. 7A: 1638x naar art. 7:658 BW in: S.C.J.J. Kortmann (ed.), Onderneming in vijf jaar nieuw burgerlijk recht, 1997, pp. 495 et seq. 202. J. Spier (ed.), The Limits of Liability, Keeping the Floodgates Shut, 1996, pp. 124-125 and Faure, M. and Hartlief, T., Nieuwe risico’s en vragen van aansprakelijkheid en verzekering, Deventer: Kluwer, 2002, p. 45-51. 203. Dutch Supreme Court, 25 June 1993, 1993, Nederlandse Jurisprudentie, 686. 204. See Hoge Raad, 29 June 2001, [2001] NJ 476 and Hoge Raad 17 November 2000, [2001] 596 (Unilever-Dikmans). 205. See Hoge Raad 26 January 2001, [2001] NJ 597 (Weststrate/De Schelde). For a comment on this decision see Giesen, I., “Werkgeversaansprakelijkheid, blootstelling aan asbest en bewijslastverdeling op basis van de strekking van de materiele norm”, Nederlands Tijdschrift voor Burgerlijk Recht, 2001, 242. 206. See Kantonrechter Middelburg, 4 December 2000, [2001] JAR 5, Kantonrechter Haarlem, 14 March 2001, [2001] JAR 88 and Faure, M. and Hartlief, T., in Spier, J., Yearbook on Insurance and Liability, to be published. 207. Dutch Supreme Court, 9 January 1987, 1987, Nederlandse Jurisprudentie, 948. 208. Dutch Supreme Court, 27 March 1992, 1992, Nederlandse Jurisprudentie, 496. 209. Dutch Supreme Court, 20 September 1996, 1997, Nederlandse Jurisprudentie, 198.

76


2.


210. See T. Hartlief/R.P.J.L. Tjittes, De aansprakelijkheid voor bedrijfsongevallen en– ziekten. Recente ontwikkelingen met betrekking tot tewerkstelling en bewijslast, 1990, Sociaal Recht, 286. 211. So T. Hartlief (supra note 173) pp. 43-44. 212. See H. Koziol, Die Arzthaftung im geltenden und künftigen Recht in: Haftungsrechtliche Perspektiven der ärztlichen Behandlung, 1997, pp. 33-34. 213. See T. Hartlief (supra note 173) pp. 56-57. 214. For a comparative overview, see the various contributions in Bocken, H. (ed.), Vorderingsbevoegdheid voor milieuverenigingen, le droit pour les associations de défence de l’environnement, d’ester en justice, Kluwer, 1988. 215. For the collective right of action in the Netherlands, see Bierbooms, P.F.A. en De Vries, L.J.A., “Wetsvoorstel collectief actierecht: de rechtspraak aan banden gelegd”, Nederlands Juristenblad, 1993, 96-97; Klik, P., “Het wetsvoorstel vorderingsrecht van belangenorganisaties bij de Eerste Kamer. Terug naar af”, Tijdschrift voor Milieuaansprakelijkheid, 1994, 34 as well as Wijmen, P.C.E. van, “Bescherming van natuur- en milieuwaarden in rechte; de verdediging van collectieve goederen”, Tijdschrift voor Milieu en Recht, 1994, 234, which discusses the proposal and Frenk, N., Kollektieve akties in het privaatrecht, diss. RUU, Kluwer, 1994 as well as Teunissen, J.M.H.F., Het Burgerlijk kleed van de Staat, diss. RL, Tjeenk Willink, 1996, which critically discusses the new provisions in the Civil Code. 216. For a general discussion of this issue, see Frenk, N., Kollektieve akties in het privaatrecht, and on the collective action rights of consumers, the dissertation of Mölenberg, L.J.H., Het collectief actierecht voor consumentenorganisaties op het terrein van de algemene voorwaarden, Gouda Quint, 1995 and Nagareda, R.A., “Autonomy, Peace and Put-Options in the Mass Tort Class Action”, Harvard Law Review, 2002, 749-829. 217. Note, however, that pleas can be heard in favour of a limitation on the number of claimants in mass torts cases such as the DES case for example, in order to make a settlement with all potential victims possible. See the inauguration address of Dommering-van Rongen, L., Schade vergoeden door fondsvorming, Kluwer, 1996. 218. For example, the number of insurance policies has doubled in the Netherlands since 1988 (see Verzekerd van cijfers, Verbond van Verzekeraars, 1994, 42). 219. Note that we are not arguing that the number of claims will increase because of moral hazard created through the insurance coverage (the issue of moral hazard will be discussed in Ch. 3). There might simply be a number of justified claims that would not have been brought otherwise given the process risk and the amount of legal fees. 220. In this connection, see also Stolker, C.J.J.M., Ars Aequi, 1995, 23. 221. The brochure is called Je hersens op het nachtkastje (Your brains on the night table). 222. For an overview see Faure, M., Finsinger, J., Siegers, J. and Van den Bergh, R. (eds.), Regulation of professions, a law and economics approach to the regulation of attorneys and physicians in the US, Belgium, the Netherlands, Germany and the U.K., Maklu, 1993, and Schneider, H., Die Anerkennung von Diplomen in der Europäischen Gemeinschaft, Maklu, 1995, 265-350. 223. Brunner, C.H.J., “Verplichte procesvertegenwoordiging in civiele zaken”, Advocatenblad, 1980, 473. 224. Nederlands Juristenblad, 1995, 35-36.


77

2.


225. There is, on the other hand, Belgian empirical research that has shown that the success rate of trade union representatives is greater than that of attorneys: K ey e r s , L . a n d Pe e t e r s , B. , “ R e c h t s h u l p b i j b e t w i s t i n g e n v o o r d e arbeidsgerechten”, Belgisch Tijdschrift voor Sociale Zekerheid, 1984, 177. 226. In economic theory it has, however, been pointed out that when the number of market participants increases, so will competitive pressure, which may lead to an increase in the number of claims (Cooter, R. and Ulen, Th., Law and Economics, Scott Foresman, 1988, 484, although the authors admit that there is no empirical support for the proposition that “more lawyers causes more claims”). 227. Fr the Netherlands, see: Hellingman, K., “An economic analysis of the regulation of lawyers in the Netherlands” in Faure, M. et al, Regulation of Professions, 167-169. 228. Art. 459, al. 1 of the Code of Civil Procedure prohibits any ex ante agreement on the fee which is dependent on the result of the case. See Faure, M., “Regulation of attorneys in Belgium” in Faure, M. et al, Regulation of Professions, 97-98. 229. See Van der Schans, E., “Contingency fees”, Advocatenblad, 1994, 663; Brenninkmeijer, A.F.M., “Advocatentarieven: niet op prijs gesteld”, Nederlands Juristenblad, 1994, 1246. Recently the Dutch Competition Authority (Nederlandse Mededingings autoriteit) fuelled the discussion as it found the no-cure-no pay prohibition in violation of Dutch (and maybe European) competition law. See on this decision Snoep, M., “NMa mist kans om eind te maken aan discussie”, Advocatenblad, 2002, 194-198. 230. Van der Schans, E., Advocatenblad, 1994, 665. 231. See Posner, R., Economic Analysis of Law, Little Brown, 1992, 567-571. 232. This point has been made repeatedly by Gordon Tullock. See, for example, his Trials on trial, Columbia University Press, 1980. 233. Bundesrechtsanwaltsordnung. 234. So Hugenholtz, W. and Heemskerk, W.H., Hoofdlijnen van Nederlands Burgerlijk Procesrecht, Lemma, 1996, 113-118 and Spier, J., “De grenzen der buitengerechtelijke kosten”, Aansprakelijkheid en Verzekering, 1995, 55. This rule, however, only applies to attorneys” fees for litigation; for so-called “extra court” costs, full compensation is due, provided these costs are reasonable as required by Art. 6:96 of the Dutch Civil Code. 235. See Dutch Supreme Court 3 April 1987, Nederlandse Jurisprudentie, 1988, 276 and also Sieswerda, P., “(On)voorwaardelijke toevoeging en de vergoeding van preprocessuele kosten”, Advocatenblad, 1987, 513, De Hoog, P.A., “Naar een werkelijke rechtsbescherming van medisch gelaedeerden”, Nederlands Juristenblad, 1988, 8-10 and Kremer, F.Th., “Moeten alle buiten-gerechtelijke kosten vergoed worden; en zo ja, waarom niet?” in Wansink, J.H. (ed.), Een Salomons Oordeel, Tjeenk Willink, 1991, 25-31. 236. Cooter and Ulen point out that the European rule of “loser pays all” instead of the American rule of “each pays his own expenses” combined with the possibility of contingency fees, cause more trials in the US (Cooter, R. and Ulen, Th., Law and Economics, 493; see also Posner, R., “An Economic Approach to Legal Procedure and Judicial Administration”, Journal of Legal Studies, 1973, 399 as well as Shavell, S., “Suit, Settlement and Trial: A Theoretical Analysis under Alternative Methods for the Allocation of legal Costs”, Journal of Legal Studies, 1982, 55). 237. This document is published in Handelingen II, 1998/99, 26630 nr. 1. For a summary see Oprukkende claimcultuur moet halt toegeroepen worden, 1999

78


2.


Nederlands Juristenblad, 1175-1178 and for critical comments see M. Faure/ T. Hartlief, Enkele opmerkingen naar aanleiding van de brief van het kabinet over de claimcultuur, 1999, Aansprakelijkheid en Verzekering, 75-84 and M. Faure/ T. Hartlief, Het Kabinet en de claimcultuur. Over de (onbezonnen) Hollandse vrees voor Amerikaanse toestanden of het -Hollands medicijn voor Amerikaanse ziektes, 1999, Nederlands Juristenblad, 2007-2015. 238. II. 1998/99, 26630, nr. 1, p.2 239. See S. Klosse/G.Vonk (supra note 155) p. 214-217. 240. See R.J. Tjittes, Hollandse toestanden, 1999, RM Themis, 141-142. 241. See M. Faure/T. Hartlief, 1998, Nederlands Juristenblad, 1135-1142. 242. See W.J.P.M. Fase, Wederzijds begrip: over de aansluiting tussen het arbeids- en sociaal verzekeringsrecht, 1993 en W.J.P.M. Fase, Privatisering van de ziektewet, 1995, Sociaal Maandblad Arbeid, 348-362. 243. See also N. Frenk, De directe schadeverzekering als vervanging van aansprakelijkheid, 1999 Nederlands Juristenblad, 15-47 and see N.F. van Manen, De paradox van individualisering en collectivisering in: N.F. van Manen/ R.H. Stutterheim (eds.), Wie draagt de schade?, 1998, pp. 95-113. 244. Critical comments on the proposals have been formulated by M. Faure, R e ch t s e c o n o m i s ch e k a n t t e k e n i n g e n b i j d e d e reg u l e r i n g va n de arbeidsomstandighedenwetgeving, 1995, Sociaal Recht, 140-149 and by T. Wilthagen, Normativiteit en effectiviteit van de regulering van arbeidsomstandigheden in: P. de Jong and others (eds.), Arbeid, recht en risico, 1995, pp. 85-111. 245. For a critical analysis see M. Faure/T. Hartlief, 1999, Nederlands Juristenblad, 2014. 246. See also L. de Leede, Meer aandacht voor beroepsziekten, 1998, Nederlands Juristenblad, 1779-1780. 247. G. Priest, The current insurance crisis and modern tort law, 1987, Yale Law Journal, 1521-1590. 248. See L. Bergkamp, Aansprakelijkheid is geen schadeverzekering, 2000, Nederlands Juristenblad, 273-276. 249. In this connection, see inter alia B. de Vroom (ed.) Betwijfelde zekerheden. Reacties op nieuwe risico’s in Nederland, 1998, p. 14 and see the conclusion of attorney general J. Spier in the case Rouwhof/Eternit (number C98/220) of the Dutch Supreme Court under 4.4.3. 250. See the letter of the Dutch government to the Chamber of Representatives 1998-99, 26630, number 1, pp. 2-3. 251. See in his respect the highly interesting Tilburg lecture of J. Spier, Een nieuwe dageraad voor het aansprakelijkheidsrecht, 1999 and see M. Faure/T. Hartlief, Claimcultuur in Nederland. Over diagnoses en medicijnen, 2000, Nederlands Juristenblad, 276. 252. For this reason there is a lot of criticism of the trend trend towards claims-made coverage. See e.g. S. Klosse/G. Vonk (supra note 171) p. 216-217.


79


Chapter 3

Insurability of Systemic Risks


81

3.

INSURABILITY OF SYSTEMIC RISKS

A. Introduction In this part of the study, we will try to focus on the key question, namely, whether emerging systemic risks are insurable. We will therefore address the theoretical conditions that have to be fulfilled to guarantee the insurability of a risk. It is therefore necessary briefly to show how insurance works. First, we will mainly focus on the functioning of liability insurance, although some generalities apply to other forms of insurance as well. Next, we will address other insurance techniques such as first-party and direct insurance. Although we will basically address traditional insurance (as provided by insurance companies) in this section, other forms of financial security could be examined, such as self-insurance, bank guarantees and participation in funding pools. However, the general principles of insurability which we will discuss in this section are also relevant for alternative forms of financial security. There is one important point (and inherent limitation) to this analysis which has to be made clear from the outset. Since this study is explicitly theoretical, we will describe which conditions have to be fulfilled to promote the insurability of systemic risks. Although this might seem merely theoretical, some of this information will be highly practical since it can identify, for example, which conditions the legislator or insurers must fulfil to optimally guarantee the insurability of systemic risks. In that sense, some of the issues identified in this chapter which may affect the insurability of liability certainly have a relevance to policy. Moreover, in some cases we will also indicate changes that have taken place in several countries either in legislation or in insurance practice (such as the emergence of new techniques of coverage) in response to newly emerging systemic risks. However, this study is not empirical in that we cannot show why, even if the theoretical conditions are met, insurers may not wish to provide coverage. In other words, we can – at the academic level – identify theoretical conditions, which may help to improve the insurability of systemic risks. Ultimately, however, it is the providers of financial security (in practice mostly insurance companies) who decide whether they consider risks to be insurable or not. This is obviously a critical issue in judging the insurability of systemic risks. We may well establish that certain systemic risks are theoretically insurable. If, however, the market decides not to cover a particular risk, the theoretical insurability would remain a dead letter.

82


3.


There is one other aspect which may affect insurability, but which may not sufficiently be taken into account in this study – the fact that insurance business is, not only in Europe but also world-wide, a regulated financial services sector. This means that insurers are subject to prudential rules and ongoing supervision by competent authorities. The goals of insurance regulation are usually the protection of the insured persons. This may have an impact on the market behaviour of insurance companies. The reader should therefore be aware that the possibilities for an insurance company to set premiums and other policy conditions according to the economic principles set out in this chapter, may sometimes be limited as a result of (national or European) regulation. Also, the possibilities to extend capacity may also be restricted by regulation e.g. a minimum solvency margin. Finally, the reader should be aware that within the scope of this study and the time limits set, we cannot discuss the complex issue of insurability of systemic risks in all its aspects. We will therefore mainly focus on those issues, which may be crucial to assessing the insurability of systemic risks. Moreover, we will try to identify – at the practical level – measures that either the legislator or insurers could take to increase the insurability of systemic risks. Some examples can be given of changing insurance techniques in response to newly emerging systemic risks. However, it is obviously not possible to explain in any detail how systemic risks are insured in practice in many OECD countries. The reader who is interested in more practical details is referred to the literature mentioned in the notes and listed in the bibliography at the end of this study.

B. Basic principles of insurance 1. Risk aversion and insurance Economists have used the concept of risk aversion to explain that many people will be averse to risks with a relatively low probability of occurring but with a possible large magnitude when they do occur. The utilitarian approach with respect to insurance has demonstrated that risk creates a disutility for people with risk aversion. Their utility can be increased by loss spreading, or if the small probability of a large loss is shifted from the injurer in exchange for the certainty of a small loss.253 The latter is of course what constitutes insurance. The risk-averse injurer has a demand for insurance; he prefers the certainty of a small loss (the payment of the insurance premium) whereby the probability of a larger loss is shifted to the insurance company, thereby increasing the utility of the injurer.254 It is remarkable that in this utilitarian approach to insurance, liability insurance is in the first place regarded as a means to increase the utility of a risk-averse injurer, and not so much as a means of protecting protect victims as is sometimes argued by lawyers.


83

3.


The reason an insurance company can take over the risk of the injurer is well known: because of the large number of participants, the risk can be spread over a larger group of people. The insurer only has to ensure that he builds relatively small risk groups in which the premium is as much as possible aligned on the risk of the members of that group. In addition to this utility-based theory of insurance, which sees insurance as an instrument to increase the expected utility of risk-averse persons through a system of risk-spreading, Skogh has powerfully argued that insurance may also be used as a device to reduce transactions costs.255 The basic principle of insurance is therefore relatively simple: because individuals have an aversion to risk, they will seek insurance coverage. The insurer is able to aggregate many risks of risk-adverse individuals facing the same risk profile. Using the law of large numbers, the insurer is able to spread the risk over a larger risk group. However, this supposes that the insurer is able to calculate an adequate premium based, on the one hand, on the probability (p) that a certain accident will occur and, on the other, the potential magnitude of the damage (D). The latter is obviously important because the insurer has to make sure that he has enough capacity at the moment a certain systemic risk with a potential large magnitude emerges. This capacity problem will be addressed in the next section (C). We will now address briefly the element of predictability, given that even insurers may be uncertain concerning the probability of the emergence of a specific risk.

2. Predictability of systemic risks as the corner stone Hence, one can argue that these general principles will apply also with respect to systemic risks: insurance can provide protection for risk-averse injurers. By accepting a certain loss in the form of the payment of a premium, the future risk that an accident will happen can be removed from the injurer. This arrangement will be utility maximising for the particular injurer and may reduce transaction costs since ex ante it has been decided who will intervene in the event an accident occurs. The insurance company will be able to provide coverage if he can aggregate similar risks in risk groups and thus spread the risk among his insured via the law of the large numbers. Obviously for every insurance scheme and for the insurance of systemic risks too, it is crucial that the insurer possesses accurate information on the likelihood that the event will occur (the probability) and on the possible magnitude of the damage once the accident has occurred. These expectations concerning the probability and magnitude of the loss are essential for the insurer to be able to calculate his so-called actuarially fair premium. Increased by so-called loading costs (inter alia administrative expenses) and, depending on the market structure, a profit margin, this will constitute the premium to be paid by the insured.

84


3.


In this respect, the insurance of systemic risks is obviously not different from any other type of insurance; for these general principles, one can refer to the insurance economics literature mentioned in the previous notes. However, it is important to stress that one crucial element in insurance, in order to make an accurate premium calculation possible, is precise information on the probability that a certain loss will occur and the possibility of making a more or less accurate estimate of the potential magnitude of the damage. This is not only necessary in order to make an accurate estimation of the premium to be charged, but also to set aside a reserve in case the accident for which insurance coverage was taken out, occurs. It may be concluded from this brief introduction to the working of liability insurance that an insurer needs information concerning the probability that a certain event (the fact that the insured party will be held liable) will occur and that he will need information on the possible magnitude of the damage. The multiplier of these (pxD) constitutes the fair premium. The reason why the insurer can take over this risk is the law of large numbers: a larger group of insurers with a similar risk can be brought together in a risk group and thus risk-spreading becomes possible. If the insurer ideally has ex ante perfect information on the predictability of the probability and the magnitude of the damage, we call the particular risk insurable. It is precisely on the basis of statistics that the insurer will require information on the likelihood that the risk will occur with a particular insured party; statistics may also provide information on the possible magnitude of the damage. Both these requirements may, however, be a problem in the case of the insurance of systemic risks. Several elements may negatively influence the ex ante predictability of the risk.256 The ex ante information on the predictability of the risk is often low, given the relatively new character of many systemic risks. Reliable statistics may sometimes be missing both with respect to the probability of the event occurring and with respect to the damage.257 Hence, there may not be a “law of large numbers” to be applied. This obviously is not only a problem for systemic risks but also occurs in every case where insurers are confronted with new risks for which reliable data may be lacking. As far as natural resource damage is concerned, an additional problem lies in the fact that generallyaccepted measurement techniques for quantifying environmental damage may be lacking.258

3. Remedies for “insurer ambiguity”? The predictability of the liability risk is obviously a crucial element to guarantee the insurability of systemic risks. The question therefore arises whether the predictability of the liability risk can be increased even in the absence of reliable statistics, or whether the particular risk should be judged as uninsurable when such statistics are not available. The literature has


85

3.


indicated that uncertainty concerning the probability or the damage is of course an element, which the insurer can – in principal – take account of ex ante. If there is uncertainty, because of a lack of reliable statistics, this should not necessarily lead to the conclusion that a particular risk is uninsurable. We are then dealing with the concept referred to as “insurer ambiguity” addressed by Kunreuther, Hogarth and Meszaros.259 They argue that the insurer can respond to this uncertainty concerning either the probability of the event or the magnitude of the damage by charging a so-called risk premium to account for this unpredictability. Hence, an insurer can in principle also deal with a “hard-topredict” event, by charging an additional premium. Although theoretically the additional risk premium is the answer to insurer ambiguity, in practice the insurer will at least need some information to make more than an educated guess concerning the risk premium he has to charge. Moreover, given the fact that an insurer finds himself in a competitive environment, market forces may well drive him to engage in liability insurance even when an appropriate risk premium cannot be charged. Whether this will in practice actually be done is hard to predict. On the one hand, we have to bear in mind that insurance business in many countries is to a large extent regulated. This might limit insurers’ ability anyhow to engage in extremely risky activities by changing high premiums. The direct link between tarification and provisioning is the subject of insurance supervision. Moreover, since many insurers are traditionally conservative and prudent, they might be inclined to be quite cautious about providing cover if uncertainty should be considered too large. The stance of this argument is hence that the predictability is crucial in requirement to keep a risk insurable, but that a lack of predictability should not immediately lead to the conclusion that risks are uninsurable. The crucial question is whether insurers are able to cope with uncertainty by charging an additional risk premium. However, as we just indicated, it may in practice not always be possible to charge this additional risk premium. In some cases, think more particularly of risks such as GMO’s or terrorism, there may be so little information (and hence the range of uncertainty can be quite high) that it may be hardly possible to calculate the additional risk premium which may be adequate to deal with insurer ambiguity. Moreover, the reaction of insurers to calculate an additional risk premium obviously works only if there is a willingness to pay of insured. This willingness to pay will, however, to a large extend depend upon the extend to which the insured recognizes that there are, all be it uncertain, additional risks for which additional cover needs to be extended. If, as a result of information deficiencies, the potential insured do not recognize these risks, they will not be willing to pay the additional risk premium and insurance cover will not take place. It may more particularly be this lack of information (and the resulting unwillingness to pay) which may explain why these additional risk premiums are seldom charged. This may,

86


3.


however, change if more information on risks becomes available or an increasing awareness of risks becomes apparent. Think in that respect again of the case of terrorism. The willingness to pay an additional risk premium will obviously have increased after 11th September 2001.

4. Co-operation between insurers versus competition policy Hence, other ways of increasing the predictability of systemic risks need to be examined. It is worth bearing in mind that there is obviously not just one insurer confronted with systemic risks. Hence, co-operation between insurance companies, maybe even with the help of an association of insurers, may lead individual insurers to disclose information on the risk to the advantage of the whole market. However, competition authorities and economists are traditionally cautious when it comes to any form of co-operation between enterprises such as insurance companies. The risk is always there that such a co-operation may lead to welfare losses as a result of cartel agreements. Is it possible to enjoy the benefits of co-operation concerning information on risk without the potentially adverse effects of cartelisation? The question of whether such co-operation is possible has been dealt with in Europe by European Commission regulation No. 3932/92 of 21 December 1992 on the application of the ex-Article 85 (3) of the EC treaty to certain categories of agreements, decisions and concerted practices in the insurance sector.260 This exemption clearly allows co-operation between insurers for the purpose of exchanging information. Consideration 6 preceding the regulation specifies: “Collaboration between insurance undertakings or within associations of undertakings in the compilation of statistics on the number of claims, the number of individual risks insured, total amounts paid in respect of claims and the amount of capital insured makes it possible to improve the knowledge of risks and facilitates the rating of risks for individual companies.” “Joint studies on the probable impact of extraneous circumstances that may influence the frequency or scale of claims, or the yield of different types of investments, should also be included. It is, however, necessary to ensure that the restrictions are only exempted to the extent to which they are necessary to attain the objectives”. From an insurance perspective it can be important to carry on joint research with respect to the insurability of certain risks such as environmental liability. This joint research may bring about important economies of scale. It does not seem very useful for every individual insurer separately to examine the probability of damage to bio-diversity in specific cases. Co-ordination and collaboration to this end seems important. In that respect, there can be an important role for an association of insurers, for example. This could gather statistical information and provide it to individual insurers. If the provision of this information then allows individual insurers to differentiate risks in an


87

3.


optimal way, the role of the association has thus served a social interest by the same token. This type of collaboration with a view to ensuring an optimal exchange of information is therefore allowed, at least in Europe, under the exemption No. 3932/92. It is, moreover, worth mentioning a recent report by the Commission to the European Parliament and the Council on the operation of the aforementioned regulation No. 3932/92.261 This Commission report, issued on 12 May 1999, examines how the exemption regulation has worked in practice and examines whether there is a need to amend it. This report also has a bearing on the area of liability insurance since several practices which are relevant for the domain of liability insurance are discussed in it. In this report the Commission also examines whether the keeping registers or exchanging information on aggravated risks are allowed from a competition policy perspective. The Commission holds262 that these agreements on keeping registers or exchanging information have the aim of making it possible for insurers to know better the nature of the risks to be insured. These agreements do not fall formally within the scope of Article 81-1 (ex-Article 85-1) of the EC Treaty if they restrict themselves to giving information on aggravated risks. In any case, the Commission holds that the mere exchange of information on the nature of a risk does not appear to have the aim of restricting competition between insurers. However, it is careful and argues that the situation is different if the exchange of information is accompanied by an agreement aiming to adopt a common attitude with regard to the risk in question. For example, recommendations to refuse to cover the aggravated risks in question or to raise the risk premiums for these risks fall clearly within the scope of Article 81-1 (ex-Article 85-1) and do not appear exempt under the terms of Article 81-3 (ex-Article 85-3). This report therefore clearly states that the exchange of information on risk between insurers is also allowed from a European competition policy perspective – at least in Europe – in order to increase the availability of reliable statistics and data. At the same time, the limits of such co-operation are made clear as well: exchanging information is allowed, formulating recommendations to refuse coverage or to raise risk premiums is obviously not.263

C. Capacity 1. General We already mentioned that a requirement for insurability is that the insurer should have ex ante information on the predictability of the risk and on the magnitude of the damage. So far we have dealt with the importance of the predictability of liability. However, gauging the magnitude of the harm may also constitute a problem.264 The magnitude of the harm may cause a variety of problems for an insurer. The basic principle of insurance, stated very

88


3.


simply, is that the total amount of premiums received (that is to say, the actuarially fair premiums) should be sufficient to guarantee that the expected damage can be compensated in the event of an accident. We refer here to the expected damage, that being the absolute amount of the damage multiplied by the probability that it will occur (p. x D). However, even if the probability is very low, there is also an upper limit on the magnitude. Indeed, the basic idea is obviously that if this – even very low probability – event occurs, the insurer needs sufficient capacity to cover that particular risk. Therefore, insurers will build reserves to be able to provide coverage for the damage once it occurs. However, in many cases the expected loss may exceed the possibilities of the individual insurer. In that case, the insurer can use various traditional insurance techniques to cope with this capacity problem. One possibility is to insure a similar risk jointly with a few insurers (so-called co-insurance); another possibility is re-insurance. One other solution which is often used in environmental liability insurance is pooling of capacity by insurers. In many countries insurers have shared risks in mutual pools on a non-competitive basis to be able to provide coverage, and also for risks with a relatively high potential magnitude. This is typically the case for the nuclear risk.265 In sum, the mere fact that the magnitude of the damage may be large does not make the risk uninsurable per se. The aforementioned insurance techniques, such as co-insurance and re-insurance, will allow insurers to cover even potentially very large risks. Generally speaking, legal doctrine has pointed to the fact that the possibilities of insuring large risks have increased as a result of changes in the insurance industry. One effect of creating a European internal insurance market had been that the size of insurance markets (and hence capacity) has considerably increased.266 But notwithstanding all these techniques to increase capacity, risks may still be of such a large magnitude that even with the use of all the aforementioned techniques, capacity may not be sufficient to cover the risk once it emerges. Especially with systemic risks (we already mentioned the nuclear example) whose magnitude can be very high, this may well play a role. At the policy level, there is relatively little that one can advocate: even insurance markets have their limitations. All one can say in this respect is that insurers themselves obviously have the ability to limit the risk of a high-damage event, more specifically by putting an upper limit on coverage. For the legislator, the fact that high- damage events may not be insurable on the traditional market may lead to the question of whether public compensation mechanisms such as funds should be used as an alternative. We will come back to this point later, in Chapter 6.


89

3.


Moreover, the potentially large magnitude of damage in the case of emerging systemic risks may cause another problem, having to do with scale effects. One only has to think of events like Hurricane Andrew or the damage caused to the twin towers on 11th September 2001. In these cases, it is obviously not only the high amount of the damage which may put off insurers but also the fact that many claims will arise at the same time, probably with a similar insurer. Again, insurance theory offers some traditional “hints” for avoiding such a risk. For example, the golden rule for a fire insurer is obviously never to insure all the flats in one apartment building. To some extent, this is equal to the risk of adverse selection, which we will address below. But even if we just focus on the problem that capacity may largely outweigh the individual possibilities of insurers, the question still arises as to what the effect of the large capacity should be at the policy level. One potential response is for insurers to pool their resources in order to be able to provide higher amounts of coverage. But then of course the question arises as to whether such a pooling is acceptable under competition law. We will address this question below (B). Furthermore, at a general level, the policy-maker may well ask whether the limited capacity should lead to all kinds of legislative intervention e.g. to limit liability. We will argue that this is not the case since it is also very difficult for a policy-maker to receive adequate information on the available capacity (C). We will also argue below that the limited capacity of insurance markets should not as such be an argument for governments to limit the liability of industrial operators (D).

2. Pooling by insurers: the nuclear example 2.1. The nuclear example Let us now address one example of a systemic risk which is of particular relevance from an insurance perspective – the nuclear risk. Indeed, almost all the questions that arise concerning the insurability of emerging systemic risks have also arisen concerning the nuclear risk. This is typically a low-probability high-damage event, where there is moreover large uncertainty concerning both the probability of occurrence and the probability of the damage. Particularly for the nuclear risk, all kind of solutions have therefore been examined to increase insurability and more particularly capacity. The solutions found in nuclear insurance are therefore highly interesting when addressing the insurability of emerging systemic risks. Insurance of nuclear risks is provided by the nuclear pools. Since the risks faced are considered to be very serious, the major national insurance companies in every nuclear country decided to pool their resources in the fifties on a noncompetitive basis in order to be able to provide coverage for them267. The result of this pooling arrangement is that at the beginning of each year the insurers determine the amount that they are prepared to commit for every nuclear

90


3.


installation. The total of contributions made by the participating insurers at the beginning of the year constitutes the capacity of the national pools involved. Pools are organised on a national basis. Reinsurance is obtained from similar “pools” or “syndicates” in other countries. Nowadays 28 pools exist worldwide.268 Through this organisation of reinsurance with other pools, the whole nuclear insurance is connected worldwide.269 For example, Belgian insurance companies that joined the Belgian nuclear pool SYBAN paid part of the damages caused by the nuclear accident at Three Miles Island in the United States.270 These pools do not only provide coverage for third-party liability; they also provide first-party insurance for the damage caused to the nuclear power plant itself. This is an aspect of considerable importance since the amount available for third-party liability insurance is limited by the amounts made available for the first-party insurance for the nuclear power plant. In some literature it is even argued that the amounts that should be available for coverage of the nuclear power plant should be much higher than the amounts to be made available for the liability of the licensee of the power plant toward third parties.271 According to some authors, first-party insurance even has priority over third-party liability insurance since a nuclear accident will probably first affect the installation itself and only then the surrounding area.272 This combination of first-party and thirdparty liability insurance through the nuclear pools therefore leads to a limitation of the amount available for third-party liability insurance. It is also remarkable that brokers argued during an OECD conference in Munich in 1984 that first-party insurance should be removed from the pools; the premiums would be relatively high due to the monopoly. They also argued that an exclusion of competition was possibly justified 30 years ago at the beginning of the nuclear age but could not be defended nowadays.273 Given this high concentration on the nuclear insurance market in some countries, initiatives have been taken by the nuclear industry, in co-operation with some brokers, to withdraw first-party insurance from the nuclear pools and to cover this through a new mutual insurance fund of nuclear power plant operators.274 Pooling is not only used in nuclear liability insurance, but also with other systemic risks. Take for instance the case of environmental liability insurance. In the Netherlands, for example, when environmental insurance was still relatively new, a pool provided for environmental liability insurance. Environmental pools exist in many other countries as well.275 Now many more companies also provide coverage individually for environmental risks. It is interesting to note that the nuclear example showed the possible negative consequences of pooling in the long run: if pools act like monopolistic insurers, premiums might be relatively high and unattractive for specific operators. This led some nuclear power plants to withdraw from the pools as far as first-party coverage was concerned and to examine the possibilities of a mutual between operators. Hence, one should clearly distinguish this pooling of insurers from the pooling of risks by operators


91

3.


through risk-sharing agreements.276 The pooling of risks by operators as a result of risk-sharing agreements is considered something different than liability insurance and will therefore be discussed in further detail in the chapter dealing with alternatives to liability insurance. For the moment, it is important to realise that co-insurance, re-insurance and pooling are instruments which allow individual insurers to provide coverage for amounts which can largely exceed their individual capacity. Financial capacity should therefore not be judged solely by looking at the individual financial possibilities of one insurance company. However, with respect to these collaborative arrangements between insurance companies, the question could be asked as to how they can be reconciled with a competition policy which generally looks with suspicion at any form of co-operation between enterprises.

2.2. Pooling versus competition policy The issue of pooling has also been the subject of discussion in Europe. It has been explicitly addressed in the aforementioned European Commission Regulation 3932/92 concerning the application of competition policy on insurance markets. Co-operation between insurers is undoubtedly necessary to increase the capacity, which may be required to cover large risks, not only for environmental liability but also for the insurance of other systemic risks. As we just mentioned, pooling of risks is well-known in the liability insurance business, more particularly with respect to nuclear liability but also with respect to environmental liability. Dutch liability insurance was for a long time provided through a so-called environmental pool, which consisted of co-operating insurers who provided liability coverage. The European Commission is, from a competition policy perspective, rather cautious with respect to allowing collaboration between insurance undertakings for the purpose of pooling. Article 11 of the exemption regulation provides that the exemption shall apply on the condition that the insurance products underwritten by the participating undertakings represent: ●

in the case of co-insurance groups, not more than 10% of all the insurance products that are identical from the point of view of the risks covered and

●

in the case of co-re-insurance groups, not more than 15% of all the insurance products.

It is doubtful whether nuclear liability pools can be justified within the strict limits set by the European Commission. Indeed, most of the insurance companies which co-operate within the Dutch atomic pool, for example, have a market share which is much larger than the strict limits set by the Commission.277 The exemption regulation does not seem to provide a specific exception for nuclear or environmental risks, so that some of the pools mentioned which have

92


3.


a market share of more than 15%, seem incompatible with the Commission’s competition policy. From an economic point of view, one can argue that pooling of risks can in itself be efficient, since pooling also allows smaller companies to operate on the market for catastrophic risks. Without pooling only larger companies could cover systemic risks. Hence, pooling could lead to a larger number of market participants. 278 If, on the other hand, pooling does not lead to increased competition, which seems to be the case at least in the nuclear field, pooling remains inefficient. The issue of pooling has also been discussed by the European Commission in the aforementioned report to the European Parliament and Council of 12 May 1999 concerning the operation of the exemption regulation 3932/92. The Commission clearly discusses the common coverage of certain risks, which we have referred to here as “pooling”. The Commission states that the starting point is still that any institutionalised grouping is in itself restrictive of competition. However, a pool can benefit from the block exemption if the market share of its members does not exceed the thresholds as specified in the exemption regulation279. The Commission however remains flexible and recognises that in some areas of insurance an insurer must, in order to be present on a market without incurring excessive risk, insure a sufficient number of risks so that the risk profile of its portfolio corresponds to the average for the totality of risks in the category. The Commission continues: “There therefore needs to be a strong probability that the real level of claims incurred by the insurer will be the same as the average level of claims of all insurers. This strong probability can only be obtained above a certain number of risks covered by the insurer. Certain catastrophic risks may be such that no individual insurer is capable of insuring it alone. In such a case, the pooling of capacity does not restrict competition. If anything, the pool strengthened competition since it allows several insurers who are unable alone to provide cover for the risk at hand to put their resources in common and create a new competitor for the benefit of customers in need of such cover”. The Commission even holds: “In any event, the Commission will consider that pools no matter how high their market share is, are not covered by Article 81-1 (ex-Article 85-1) when they are necessary to allow their members to provide a type of insurance they could not provide alone.”280 This probably may save nuclear and other pools who could argue that (no matter how high their market share is) they are necessary since without a pool this type of insurance for systemic risks could otherwise not be provided. The European Commission also discusses the prospects for the future and announces that its services have just launched their investigation into


93

3.


“co-insurance or co-reinsurance pools dealing with environmental risks and nuclear risks. Several of those pools have been notified (the French environmental pool Assurpol was actually granted an exemption in 1991. This exemption expired last year). All these pools will be assessed in the light of the three-tier legality test spelled out above”.281 In other words, one can expect more news from the competition authorities of the European Commission with respect to the legality of the co-operation within environmental and nuclear pools.

3. Limited capacity and public policy towards insurance Let us now once more address the question of the impact of the given fact that insurance companies will not be able to cover every case of damage resulting from systemic risks, simply because their capacity is limited. We will argue here that capacity may obviously be limited but this should not necessarily lead to legislative intervention, for example to limit the liability of enterprises. First, we will argue that it is very difficult to provide accurate information on the actually available amounts of capacity; second, one should not only look at traditional insurance markets when judging capacity. If capacity on traditional insurance markets is limited and enterprises remain exposed to liability, this may give them an incentive to develop other financial mechanisms to cover risks. Third, limiting the liability of enterprises would not solve but only shift the problem. The financial consequences of systemic risks would not be borne any longer by those who caused them, but the damage would still be there. In other words, as a result of a limitation, the damage costs would then be shifted to the public at large (the general tax payers) which can hardly be seen as an adequate solution.

3.1. Limited capacity and insurability We can summarise by saying that the high possible magnitude of the expected damages does not make a risk uninsurable as such. One should remember that competitive insurance markets have worked out all kinds of devices to cope with relatively large risks, including systemic risks. Tyran and Zweifel report, for instance, that if today there were an earthquake of the magnitude of the 1906 San Francisco earthquake, there would be insurance coverage available up to US$ 39.5 billion.282 As far as environmental insurance is concerned, Ranson recently reported that the current insurance capacity on the European market would be € 100.000.000 per insurance policy,283 which is quite substantial as well. More fundamentally, in the discussion on caps below it will be argued that the capacity problem is not an argument for introducing a financial limitation on liability.284 The individual insurer can always limit insurance coverage up to the amount for which he is willing to provide coverage, either himself or in combination with co-insurance, re-insurance or pooling. The capacity problem is therefore not a reason to argue that certain risks would be

94


3.


uninsurable; the capacity will only define the amount of coverage available which will be defined in contractual limitations. It is obviously possible that the magnitude of the harm could still be larger than the insured amounts (even with pooling and re-insurance). This may then constitute grounds for examining whether alternative compensation mechanisms could provide larger amounts than insurance but there is no reason to assume that this will automatically be the case. Moreover, the capacity to cover systemic risks depends on a variety of complex elements and not only upon the individual reserves of one insurer or his capacity to obtain reinsurance. For instance, the method of coverage in time may have a far more important influence on capacity, as we will show below.

3.2. Reliability of information on capacity At the policy level, it is dangerous to infer anything from the possibly limited capacity for covering systemic risks, since it appears to be extremely difficult to obtain reliable information on insurability in general, and capacity in particular, from the insurance market. This difficulty is connected with the fact that insurance markets in some OECD countries are relatively highly concentrated. Lobbying theories have predicted that when an industrial sector has been well organised (e.g. in a cartel) their transaction costs will be low and their potential success in the field of lobbying may be large. It is therefore reasonable to question the extent to which liability should be judged on the basis of information provided by monopolistic insurers concerning insurance capacity. If the argument that liability should be “insurable” is taken seriously, it is obviously very important that the policy-maker dispose of reliable information on the actual insurability of certain risks. A well-functioning competitive market is important precisely for this reason, so that the policy-maker can ask several companies what the precise possibilities of insurance coverage are. Experience with the nuclear field has shown that the information provided may be unreliable if the policy-maker becomes completely dependent upon information provided by a monopolistic insurer. Take the example of nuclear insurance, which, as we just mentioned, is dominated in every country by socalled nuclear pools. In the Netherlands the Dutch government relied almost blindly on information provided by the Dutch nuclear pool on the availability of coverage for liability insurance when fixing the liability limit in the new Dutch Act on Nuclear Liability of 26 June 1991. Minister Kok declared during the parliamentary debate that “during the whole preparation of the draft, negotiations have taken place with the nuclear pool. In all cases the nuclear pool could agree with the proposals. There has hence been an optimal involvement of the sector”.285 Critical voices have asked whether the liability limit for the licensee of a nuclear power plant had to be set at the amount of


95

3.


500 million Dutch guilders and should not be tested periodically according to the increasing possibilities of coverage on the private insurance market,286 but the availability of insurance remained entirely based on information provided by the nuclear pool.287 The fact that the policy-maker often relies on information provided by monopolistic insurers to judge the capacity of the insurance market is obviously not merely a Dutch phenomenon. Precisely the same took place when the Belgian Act of 22 July 1985 concerning the liability of the licensee of a nuclear plant was discussed in parliament. Also in Belgium, the government contacted the Belgian nuclear pool referred to earlier, Syban, to find out whether more than BF 4 billion would be available in third-party liability coverage. Syban fiercely denied this. Later it turned out that the nuclear power plant itself was insured under firstparty insurance (property insurance) for more than BF 40 billion. It is obviously relatively unclear why the nuclear pool only had BF 4 billion available for thirdparty liability insurance, whereas the damage to the nuclear installation itself could apparently be insured for BF 40 billion.288 The Belgian Act has, by the way, recently been changed, since parliament accepted recently a legislative proposal (launched by two Ecolo ministers) to increase the amount of the licensee of a nuclear power plant to BF 12 billion.289 But that obviously does not change the points made here: again the amount was based on the insurability according to information provided by insurers. These “nuclear” examples obviously also have importance for our study on the insurability of systemic risks. They show that one should be careful of judging the “insurability” of a particular risk, more particularly concerning the capacity aspect, on the basis of information provided by insurers, at least when there is a high degree of concentration on as on this insurance market. It is striking that national nuclear pools do not compete (in order to increase capacity) but again co-operate. They only insure the nuclear installations on their own territory, so that there is no competition between them. This example shows, once more, that in the nuclear case, pooling takes the restrictions of competition further than would be necessary to increase the insurability of the nuclear liability risk. This discussion shows that one has to be very careful with the argument that capacity may be limited. The policy recommendation one can make is that co-operation between insurers should be allowed on the condition that it increases competition, which is precisely the spirit of the report of the European Commission of 12 May 1999 on the operation of the exemption regulation No. 3932/92. Discussions on capacity obviously often play a role in justifying a financial cap on liability. We will, however, argue below, that a limited capacity (however difficult it may be to judge) should not necessarily lead to a limitation of liability via financial caps but may possibly lead to a limitation on the amount of coverage provided by an individual insurer.

96


3.


Finally, it should be recalled that in this section we merely dealt with some insurance economic techniques for increasing capacity. These possibilities may, however, in some cases be limited as a result of regulation of the insurance industry.

4. Financial limits on liability? 4.1. Introduction The risk that the amount of the damage may largely exceed the individual wealth of an injurer or potentially even of his liability insurer, has caused some legal authors (followed by some legislators) to argue that in the event that the expected amount of damage would be extremely high, a financial limit on the compensation due to the victim should be introduced. These ceilings on the compensation due to a victim are known in the literature as limitations of compensation or “financial caps”. Often they can be found jointly with strict liability. Apparently the (implicit) underlying argument is that the strict liability forcing a potential injurer to pay compensation to the victim even if he is not at fault can only be justified if this is compensated by a financial limit on the amount of compensation. This is also one of the arguments advanced in legal literature in favour of financial caps: in combination with strict liability the victim would still be better off with caps. He would be relieved of the heavy burden of proving the fault of his injurer and would thus be certain of receiving some form of compensation. This certainty would then outweigh the fact that the amount the victim receives might be less than the total amount of the loss. Caps can often be found in cases where the expected losses may be large, as is precisely the case for systemic risks. It is apparently judged that, without a cap, an injurer would anyway be unable to compensate the victim. A combination of strict liability and financial caps can be found in many international conventions dealing with systemic risks, such as the conventions on nuclear liability and on liability for marine oil pollution. But financial caps are certainly not merely “modern” responses to “crushing” strict liabilities. In former times, caps could be already found in maritime law in cases of loss of the vessel or cargo. The effects of a limitation on liability have been addressed in relation to corporate liability by Hansmann/Kraakman, 1991 290 and with respect to nuclear liability by Faure/Vandenbergh (1990) and Trebilcock/Winter (1997).291 Boyd/Ingb erm an (1994) analys ed w hether caps could reme dy the underdeterrence under strict liability, caused by insolvency.292

4.2. Financial caps in the legal debate Limitations on the amount to be paid by the injurer can be found in various legal systems and are defended on various grounds. Some legal systems explicitly give the power to the judge to limit the amount due to the victim on the basis of


97

3.


a hardship clause. In circumstances where it is considered to be unjust to require the injurer to pay full compensation to the victim, the judge can in an individual case decide to lower the amount due by the tortfeasor. This hardship clause293 only allows the judge in individual cases to reduce the compensation due to the victim for various reasons. It can be applied in cases where a relatively poor injurer causes harm to a relatively wealthy victim; in such cases, the power of the judge to lower the amount of compensation is thus an instrument of redistribution. Usually, this power of the judge is, however, limited to certain specific cases prescribed by statute and controlled via case law.294 Far more important are the cases where the amount of compensation due to the victim is limited in a more general way by statute and thus not left to the discretion of the judge in an individual case. These statutory limitations can also be found in cases which relate to systemic risks, such as nuclear accidents or marine oil pollution. The reasons given by the legislators (or drafters of the conventions) and some legal doctrine to justify these financial caps have already been mentioned. It is argued that the magnitude of an accident may be so high that it would be unjust to hold the injurer to be fully liable for these high amounts. Sometimes financial caps are also defended as compensation for the introduction of strict liability. It is often mentioned in legal doctrine that whenever strict liability is introduced, this cannot be unlimited. Finally, it is sometimes also argued that when the amount of a major accident will be so high that it will in any event exceed the injurer’s individual wealth, the injurer will anyway be judgement proof. Hence, unlimited liability exceeding the injurer’s wealth would make no sense.295 This is then combined with the argument that the victim would be much better off with a regime that guarantees the solvency of the injurer, such as compulsory insurance. But in order to make the risk insurable the amount of compensation should then be limited. Nevertheless, in legal doctrine, arguments are also advanced against financial caps. Legal authors of course argue that a limitation would be contrary to the principle of full compensation of victims;296 also the economic argument is advanced that a limitation – specifically in case of professional liability – would lead to a reduction of the quality of the services.297 Finally, it is also argued that granting a limitation to a specific group of professionals would violate the equality principle since there are no specific reasons to protect one group by financial caps whereas others would still be exposed to full liability.298

4.3. Financial caps in tort liability: the risk-neutral case In the literature it has been argued that there may be good reasons in favour of a strict liability rule for systemic risks, the main reason being that only a strict liability rule would lead to a full internalisation of such highly

98


3.


risky activities.299 This strict liability rule is especially put forward in so-called unilateral accident situations where only one party influences the accident risk. Only with strict liability would the potential injurer also have an incentive to adopt an optimal activity level. This full internalisation is obviously only possible if the injurer is effectively exposed to the full costs of the activity he engages in, and is therefore in principle held to provide full compensation to a victim. An obvious disadvantage of a system of financial caps is that this will seriously impair the victim’s rights to full compensation. But if the cap is set at a much lower amount than the expected damage, this would not only violate the victim’s right to compensation, but the above-mentioned full internalisation of the externality would not take place either. From an economic point of view a limitation of compensation therefore poses a serious problem since there will be no internalisation of the risky activity. Indeed, if one believes that exposure to liability has a deterrent effect, a limitation of the amount of compensation due to victims poses another problem. There is a direct linear relationship between the magnitude of the accident risk and the amount spent on care by the potential wrongdoer. If the liability therefore is limited to a certain amount, the potential injurer will consider the accident as one with a magnitude of the limited amount. Hence, the amount he will spend on care to avoid an accident will be equal to the limited amount and he will not spend on care the amount necessary to reduce the total accident costs. Obviously, the amount of care spent by the potential injurer will be lower and a problem of underdeterrence arises. The amount of optimal care, reflected in the optimal standard, being the care necessary to reduce the total accident costs efficiently, will be higher than the amount the potential injurer will spend to avoid an accident equal to the statutory limited amount.300 Thus, as a result of the cap, too little care is taken.301 The conclusion is, however, different in the case of bilateral accidents, where the victim’s behaviour may also affect the accident risk. The standard argument against providing full compensation to victims (also for non-pecuniary losses) in such accidents is that victims can take precautionary measures which are not always observable for judges and which can therefore not be fully accounted for in contributory or comparative negligence defences.302 A limit on the compensation for bilateral accidents may therefore be useful in cases where victims should be given additional incentives to reduce the accident risk. Whether caps are efficient in specific bilateral accident cases will depend on the circumstances. The question arises – inter alia – whether exposing the victim to risk is necessary to provide these additional incentives or whether the victim’s incentives can be optimally controlled via the contributory negligence defence. Also the amount of the cap remains important. If the cap were set too law, this would give incentives to the victim but it could equally lead to serious underdeterrence of the injurer.


99

3.


Several scholars have applied these insights to specific systemic risks such as the domain of nuclear liability, where tight limits on liability are in place both in international conventions and in national legislation. It has been argued that these caps inefficiently damper the operator’s incentives to take precautions.303 4.3.1. Subsidisation effect Another effect of protecting a particular industry through a statutory limitation is that this constitutes an indirect subsidisation of that industry.304 This point was also raised in the Netherlands during the parliamentary debate preceding a statutory change to the nuclear liability statute.305 The amount of guarantee provided by the Dutch State was increased to the exceptional amount of 5 billion Dutch guilders. It was mentioned in parliament that the Ministry of Finance would have to charge the licensee of a nuclear power plant for this guarantee provided by the state.306 Otherwise, nuclear energy would remain too cheap since the energy price would not reflect the true costs of the nuclear risk. This is in line with the traditional argument that not exposing the injurer to the full costs of his activity will lead to a too high activity level. Of course this problem of ’over-consumption of nuclear power’ would be reduced if other energy producers enjoyed a limitation of liability as well. In that case a second-best solution could be achieved. However, it seems that nuclear energy producers are alone in enjoying this benefit of the limitation of liability. The magnitude of this subsidisation effect has been estimated by Heyes/ Liston-Heyes and is considered to be relative low,307 meaning that the price of e.g. nuclear power is not substantially lower as a result of the subsidy. However, the flip-side of this is that if the financial cap were abolished (which has actually happened in some countries), the exposure to liability would apparently not endanger the entire industry.

4.4. Financial caps under risk aversion without liability insurance We now drop the assumption of the traditional models of accident law that both victim and injurer are risk-neutral and we assume the injurers’ risk aversion in a world in which no liability insurance was available. We showed above that when an injurer is risk-neutral, the injurer has to pay damages, which are equal to the amount of the victim’s loss. This will (in the absence of a judgement proof problem) lead to optimal incentives for optimal care. This is the case under both a negligence and a strict liability rule. When an injurer is risk-averse, full compensation under strict liability is no longer optimal, given risk aversion. This might lead an injurer to take inefficiently high care or to avoid certain risky activities, even though they are socially beneficial: too much care and a too low activity level will follow.308

100


3.


Here a problem arises because the legal system cannot both achieve the goal of optimal deterrence and cure the problem of risk aversion. From the perspective of optimal deterrence, the amount of damages to be paid by the injurer should be equal to the victim’s loss; from the perspective of avoiding risk aversion, it might not be optimal to expose a risk-averse injurer to full liability.309 If one addresses this problem merely from the perspective of optimal risk-spreading, one could even argue that the optimal amount of compensation to be paid by the injurer would be zero in order to avoid his exposure to risk. The obvious disadvantage of this solution would be that the injurer will follow a level of care which is equal to zero and therefore suboptimal. A first-best solution can therefore not be achieved in this situation of risk aversion by the injurer and no availability of insurance. In that case, a second-best solution might consist in a financial cap in the sense that the amount to be paid by the injurer to the victim would be fixed at a lower amount than the real loss of the victim.310 Thus both injurer and victim would carry a part of the risk and the injurer would still have incentives to take efficient care, although this solution is second best. In sum, there may only be an economic argument in favour of financial caps in case of third-party liability when injurers are risk-averse and no liability insurance is available.311 Nowadays, with the availability of insurance, this hypothesis seems at first sight to have little practical relevance. However, this obviously depends upon the availability of full insurance coverage. Even if insurance coverage is available, usually the insurer does not provide full coverage so that the injurer is still exposed to risk. However, the risk aversion of the injurer is not necessarily an argument for a cap. Indeed, one should stress that in many cases the injurer might not be more risk-averse than the victim. In addition it should be recalled that the solution of a financial cap is at best “second-best” since the negative influence on the incentives to take care still remain. 312 We shall now turn to the situation where liability insurance is available and ask whether financial caps can play a role in increasing insurability.

4.5. Financial caps to increase insurability? 4.5.1. Capacity as insurability problem If we now turn to the situation where injurers are risk-averse but liability insurance is available, the question should be asked whether statutory caps on liability are a necessary tool to guarantee the insurability of risks. This argument is often advanced in the context of compulsory insurance. Often the legislator introduced compulsory insurance (as a result of international conventions) and consequently argued that the amount of compensation in tort liability should be limited to make the particular systemic risk insurable.


101

3.


Generally one can argue that within liability insurance it is usually not (only) the amount of the expected damage that causes uninsurability of risks, but – as we argued above – more often the unpredictability of certain risks.313 The insurability question is analysed by looking both at the probability and the magnitude of the risk. The amount is not necessarily the main problem since we indicated that competitive insurance markets have worked out all kinds of devices to cope with large risks as well. Reinsurance, co-insurance, or pooling of risks are well-known phenomena that allow insurers to provide large amounts of insurance coverage. The high magnitude of the risk itself therefore does not make certain industrial accidents uninsurable per se. Moreover, as we shall discuss below, by adjusting the policy conditions the insurer can limit the amount for which he is willing to provide coverage. Usually the problem of insurability of major systemic risks refers to the “hard-to-predict” character of those risks which may make insurers both ambiguous and averse towards these risks. They might respond to insurer ambiguity314 by requiring an additional risk premium. The insured, however, may not be willing to pay the additional risk premium if they do not recognise the ambiguity an insurer is confronted with.315 4.5.2. Limitation of the duty to insure More principally, one can also argue that even in cases where there is a limited availability of insurance coverage (which is already hard to judge for the legislator, if possible at all), this should not necessarily lead to a limitation of the liability of the injurer who causes the particular systemic risk. If it does appear that the possibilities of obtaining liability insurance coverage are limited to a certain amount, there is no reason to limit the liability itself to that same amount. A clear alternative would be to introduce a duty to insure up to the available amount of insurance coverage, but to keep the liability of the injurer unlimited. This would, on the one hand, have the advantage that the duty to insure is limited to realistic amounts, while on the other, the incentives for the injurer to take care would remain at least partially in existence because the injurer would still be exposed to risk in the event that the magnitude of the harm were be higher than the insured amount. In sum, from an economic point of view there are, in unilateral accident situations, very few convincing reasons to limit the amount of compensation due to the victim of a systemic risk. If insurability problems exist they can be solved by limiting the duty to insure. Recent examples have also shown that, with respect to nuclear liability conventions, some countries have introduced a duty to insure up to a limited amount but have left the liability of the licensee of the nuclear power plant itself unlimited. This has been done for instance in Austria,316 Germany, Japan, Switzerland and Sweden.317 The advantage of this

102


3.


approach is that in those cases where injurers have assets at stake that outweigh the limited amount for which they had to purchase insurance coverage, they will still have incentives further to reduce the accident risk. A generalised limit on liability does not take into account the differing financial possibilities of injurers and their insurers. 4.5.3. Contractual limitations Although there are, in sum, very few arguments in favour of a generalised statutory limitation of liability, this does not mean that there may be no reasons for a contractual limitation in insurance policies, covering systemic risks. In many insurance policies these limitations already exist since an insurer will hardly ever provide unlimited coverage for the particular risk concerned. Thus an insurer could consider lowering the amount of liability coverage generally. This is, by the way, what is done in most insurance policies in practice: almost no policy will provide coverage to a potential polluter without a financial limit. 4.5.4. Unlimited liability to control moral hazard There is, finally, one general argument related to insurance which can be put forward against financial caps introduced in legislation. Statutory limitations could be contrary to the insurer’s interests, since it eliminates one way of reducing the moral hazard problem, which we will discuss in further detail below. There we will argue that the remedy is to expose the insured party to risk for the uninsured top slice of liability.318 It should apparently be in the insurer’s interest to have a system of unlimited liability where a partial exposure to risk may be used by the insurer as a device to control moral hazard and where on the other hand the insurer may put contractual limitations on the amount of coverage (in the absence of a duty to insure up to a certain amount) depending upon the demand for insurance of the particular injurer and the willingness to provide coverage of the insurer. Contractual limitations thus seem a better way of allowing for optimal differentiation of risk, thereby providing optimal control of moral hazard.

4.6. Policy considerations So far we have argued that financial caps introduced in legislation might cause efficiency problems since they dilute the deterrent effect of tort rules, especially in those unilateral accident cases where the expected amount of the loss exceeds the limited amount of the cap. Caps would only make sense in a bilateral setting if one would argue that the victim’s care cannot be sufficiently controlled through a contributory negligence defence that he should still be exposed to risk. That seems, however, doubtful, given the risk aversion of victims. If the legislation introducing financial caps were to pursue


103

3.


a public interest goal, one would therefore in principle expect it to refrain from the introduction of such caps. Reality is, however, often very different. For instance, in the area of the nuclear risk (which can most certainly be considered unilateral) some countries have – as we indicated – seriously limited the liability of the licensee of a nuclear power plant. For instance, in Belgium the limitations are so low that the victim’s right to compensation is reduced to less than 1% of the average costs of an accident.319 The reason why these caps are often introduced can be found in interest group theories. Up to now we have adopted the relatively unrealistic assumption that politicians act in the public interest and will therefore promulgate legislation with respect to liability rules and safety regulations for systemic risks only if this is welfare-maximising. Reality, however, shows that financial caps are not in the public interest. Especially the victim-protection argument which is sometimes used to defend financial caps is a very weak defence for existing liability schemes, e.g. with respect to a systemic risk such as nuclear liability320. Given the low limitations in some of the national implementing legislation, one can conclude that in some cases the victims were better “protected” before the implementation of the nuclear liability legislation than after.321 These inefficiencies can be explained by public choice theory which regards regulation as the product of demand for regulation by interest groups and supply by wealth-maximising politicians.322 How did this lobbying take place for instance with respect to the nuclear liability conventions? In the 1950’s the nuclear industry feared that the future of nuclear power could be endangered by unlimited liability. In the preparatory documents preceding the Paris Convention on Nuclear Liability, it can clearly be read that the goal of the regulation of nuclear liability was not so much the protection of victims but the protection of the nuclear industry itself.323 There is qualitative evidence of the influence of the nuclear industry both in drafting the international conventions and especially in the national implementing legislation.324 Both insurers and the nuclear industry lobbied in a joint profitmaximising strategy for the limitation of liability. Indeed, some have argued that the object of these nuclear liability and oil pollution conventions was not so much to increase protection for the victim but to limit the risks of, for example, a nuclear power plant operator.325 One can obviously wonder why both insurers and nuclear power plant operators lobbied in favour of a limitation of compensation. The interests of industry are clear: financial caps lower the exposure to liability and (in case of insurance), the insurance premiums. Obviously industry will lobby in favour of limiting its liability to the insured amount available. In that case the licensee of a nuclear power plant would bear no liability apart from the insured amount. At first glance it might seem strange that the insurance industry, for instance in

104


3.


Belgium, also favoured a reduction of the liability of the operator of a power plant. We already indicated that this may reduce the possibilities for the insurer to control moral hazard; in addition it also reduces the demand for insurance. However, one can clearly note in many national parliaments that the insurance industry lobbied for a limitation of the third-party liability of the operator to the “insurable” amount. One possible explanation is that, because of the unpredictability of the loss, premium calculation in a profit-maximising manner is impossible, whereas premiums in other classes of insurance are profitable. Indeed, we have mentioned that the insurer of the nuclear risk prefers to cover property damage instead of third-party liability. One possible reason for this preference might be the fact that administrative costs can be much higher in a third-party nuclear insurance. This seems to be a plausible additional explanation for the lobbying in favour of a reduced third-party liability by the insurance industry. Hence, the influence of industrial pressure groups in the process of drafting legislation explains to some extent why financial caps have nevertheless been introduced. In the European context, the European Product Liability directive had the optional possibility of putting a financial ceiling on the liability of manufacturers in case of serial damage.326 However, only Germany, Spain and Portugal made use of this option.327 Now the Green Paper has opened the debate on whether the ceiling should be increased to € 140 million,328 but the question is equally asked whether the existence of financial limits is strictly justified.329 The limit in the product liability directive is thus now apparently on the agenda for reform.

4.7. Policy recommendation Hence, the conclusion is relatively simple: although we realise quite well that the capacity of insurance and financial markets may be lower than the damage which may be caused by emerging systemic risks, there seem to be relatively few reasons to introduce financial caps for specific systemic risks. This outcome is certainly true when the parties who could be involved in an accident are risk-neutral, but it also applies in cases of risk aversion, at least when insurance is available. Even if liability insurance were not available, the appropriate policy answer hardly seems to be the introduction of a cap, given the risk aversion of victims. The better option seems to be to agree on the optimal amount of compensation via contract, which is obviously only possible when transaction costs are low, e.g. when victim and injurer stand in a relationship to each other via the price mechanism. The basic argument against financial caps assumes that the injurer has assets at stake which exceed the amount of the financial cap and that the expected amount of the damage will equally be higher than the cap. However, a judgement proof problem may arise. In that case the appropriate answer is not to


105

3.


limit liability to the amount the injurer has available but to seek insurance coverage. Through diversified contractual arrangements between the insurer and the injurer, an optimal amount of coverage can be determined in an individual case. The incentives can then be controlled via the risk differentiation in insurance, and unlimited liability can apply for the excess in case the expected loss were (ex post) higher than the insured amount and the injurer still had assets at stake. In some cases compulsory insurance might be an appropriate mechanism to control the judgement proof problem330 but even compulsory insurance is no reason to introduce financial caps in legislation. The duty to purchase insurance coverage can be limited to a certain amount but liability could remain unlimited. We showed some of the disadvantages of financial caps by pointing to the example of nuclear liability where the national implementing legislation of international conventions has very often relied upon legislative caps. The analysis obviously applies to environmental damage as well. Of course, the disincentive effects of caps in these cases might to a large extent be undone by the safety regulation which governs the nuclear risk for example.331 But that would only be a valid point if the beneficial effects of caps could be identified, which is doubtful. At the normative level, economic analysis therefore calls for caution with respect to the introduction of financial caps to cope with systemic risks, given their possible negative incentives on deterrence and the possible negative distributional consequences. Other mechanisms may be available (such as compulsory insurance) which might cure an insolvency problem without the aforementioned disadvantages. How serious these disadvantages are will obviously depend upon the specific amount of the cap. Moreover, one should note that caps do not fully expose the potentially responsible parties to the damage they may cause. Finally, a policy conclusion may also be that lawyers analysing these problems should take into account the fact that, within the process of drafting legislation with respect to the consequences of major systemic risks, numerous industrial pressure groups might intervene. They will, as the nuclear liability example showed, obviously try to limit the extent of liability of the industry involved. Industry might therefore often argue that strict liability is an insurmountable extension of their responsibility which can therefore only be accepted if the amount is statutorily limited. The examples of the convention on civil liability for oil pollution332 and the Paris and Vienna convention on the liability for nuclear accidents, which all introduced strict liability with serious financial caps, show that pressure group intervention has so far been relatively effective.

106


3.


D. Moral hazard 1. What is moral hazard? A further condition for insurability is that moral hazard can be cured. All insurance schemes are vulnerable to moral hazard. Moral hazard is the well-known phenomenon that the behaviour of the insured injurer (and every insured for that matter) will change as soon as the risk is removed from him. This is precisely the essential contradiction in liability insurance. The disutility the injurer suffers because of his exposure to risk was precisely needed to give him correct incentives for care taking. If risk is fully removed from the injurer and shifted to the insurer, the injurer will lack the incentive to take care that was given to him by the deterrent effect of having to pay compensation in the event of an accident. Marc Pauly has, by the way, pointed out that in fact this behaviour is not immoral but completely rational since the injurer simply reacts to varying costs for his behaviour.333 For the insurer of course, the problem arises as to how incentives can nevertheless be given to the insured to behave in exactly the same way as if no insurance were available. This is of course the goal of an optimal control of moral hazard.334

2. Remedies 2.1. Monitoring In the literature two ways of controlling the moral hazard problem are indicated.335 The first is control of the insured and adapting the premium appropriately; the second is exposing the insured partially to risk. A first-best solution is detailed control of the insured.336 In such a scenario, the premium conditions would be exactly adapted to the behaviour of the insured and the premium would reflect the care taken by the insured. In an optimal world this should give the insured incentives to behave exactly as if no insurance were available and the premium would reflect the true accident risk. Of course this first-best solution is only possible in the ideal world where control by the insurance company would be costless and information on the behaviour of the insured readily available. In practice this is of course not true. There are, however, various ways of controlling the insured and a differentiation of premium conditions is possible according to certain groups of risk. This can either be an ex ante screening with a higher premium for certain high risk groups or an ex post premium increase or change of policy conditions, based on previous loss experience. This is the so-called experience rating. Much insurance legislation is also aimed at reducing moral hazard. For example, many insurance laws prohibit insurers from covering accidents caused with intent.337


107

3.


2.2. Exposing the insured to risk A second-best solution is exposing the insured partially to risk. This is considered second-best because insurance should ideally exactly aim at removing risk from the injurer. Exposing the insured to risk will mean that some degree of risk aversion will remain. This has on the other hand the advantage that the insured injurer will still have some incentives to take care even though he is insured. This exposure to risk can be either at the lower level of damage or at the higher level. One could imagine a system with a deductible whereby a lower threshold applies or one could introduce an upper limit on coverage whereby the insured bears his own loss if the damage exceeds the insured amount.

2.3. Combination In practice one will of course see a combination of both systems of controlling moral hazard. Usually there is some degree of differentiation within the policy conditions, a deductible and an upper limit on coverage. Of course the methods used depend upon the information costs, but also on the value of the insurance policy.338 Obviously an insurer will be more willing to invest resources in making a nicely-tailored insurance policy for a large company that pays a substantial premium than in the case of consumer risks. If moral hazard is controlled optimally through the use of the abovementioned devices, the insured will again behave as if no insurance coverage were available, with the benefit that the disutility of risk is removed from him. The incentives to take care are in that case no longer given by liability law since the threat to have to pay compensation to a victim is shifted to the insurance company. In case of insurance, the injurer is given the incentive to take care through an appropriate adaptation of the policy conditions to the behaviour of the individual insured. This also explains why liability insurance has a very important social function. Under liability insurance, the insurer has to guarantee that the insured will take efficient care and thus have an incentive to avoid accidents. Appropriate control of moral hazard is thus not only in the interest of the individual insurer but also of society. If there were no efficient control of moral hazard, insurance would on the whole do more harm than good.

108


3.


E. Adverse selection 1. Avoiding the “market for lemons” A further condition for insurability is that the adverse selection problem can be overcome. This problem can be described briefly as follows. It has already been explained that insurance is based on a system of lossspreading. Therefore the insurer needs a minimum number of similar risks to insure. At the same time risk pools have to be constructed as narrowly as possible, meaning that the average premium in the risk pool should correspond to the risk of most of the members in the particular pool. If this were not the case, then the average premium would be relatively high for low-risk members, who would then leave the group. In that case, the well-known phenomenon of adverse selection could emerge. This has been described in Akerlof’s seminal paper on the market for lemons.339 Adverse selection will, in other words, arise if potentially responsible parties fail to disclose their true risk profile appropriately, which may impair the narrowing of risk pools.340 Rogge holds that, in Belgium, the financial capacity to insure environmental risks would be limited341 precisely because only bad risks would have a demand for insurance. If this cannot be "compensated" by good risks, an incurable problem of adverse selection would remain.342 Thus lack of demand would be matched by lack of supply"343 Let us now address the solutions prescribed in the literature to the problem of adverse selection.

2. Danger of “correlated risks” A problem which is linked to the adverse selection problem is that insurance supposes that the insurer makes adequate, narrow risk pools of similar, but unrelated risks. The requirement of bringing similar risks into the risk pool is necessary in order to avoid the market for lemons, which could result in adverse selection. However, the requirement that the risks in the pool should be uncorrelated, is equally important. This simply has to do with a basic insurance requirement that the realization of one risk may not at the same time cause the realization of other insured risks as well. The traditional example is the fact that a fire insurer would always be advised not to insure all the buildings within one street. These buildings would indeed be correlated risks, at least as the risk of fire is concerned. Hence, the avoidance of uncorrelated risks is an essential requirement for the ability to pool individual risks. The correlation between individual risks has been one of the reasons why a real insurance market for flood risks has not developed in many countries. Indeed, there is a serious risk that if a flood occurs, this does not only affect one insured individual, but most likely a large, correlated group.344


109

3.


A number of other systemic risks nowadays seem to share this feature, in particularly due to the development of economic and physical networks. Think in this respect of the risk of cyber terrorism. If it would indeed appear that some risks are correlated, this may endanger the availability of insurance, since the potential to pool the individual risks is then limited. However, the answer to this problem is obviously that one insurer should not engage in the insurance of many correlated risks. Risk spreading via other insurance companies, operating on a competitive market, could provide a (although partial) remedy for this problem.

F. The remedy: risk differentiation The literature has indicated that the appropriate remedy for both the problems of moral hazard and adverse selection is to be found in adequate risk differentiation by insurers. We will first identify the theoretical benefits of risk differentiation; then we will apply these insights to one specific systemic risk – environmental risk, showing how an insurer could optimise risk differentiation. This will lead us to show that an insurer could, through appropriate specialisation and monitoring, check the ecological reliability of an enterprise.

1. Risk differentiation: theory It follows from the economic principles of liability insurance that tailoring of policy conditions to the individual risk is essential in order to control both moral hazard and adverse selection. George Priest has claimed that the adverse selection problem has caused an insurance crisis in the United States and that it can only be cured by an appropriate differentiation of risk. 345 If the insurance policy requires preventive action from the insured party and provides a corresponding reward for doing so in the premium, this should give optimal incentives to the insured to reduce accidents. Risk pools should therefore be constructed as narrowly as possible so that the premium reflects the risk of the average member of that particular pool.346 Further differentiation of the risk is obviously efficient only as long as the marginal benefits outweigh the marginal costs of doing so.347 Risk differentiation certainly does not mean that insurers would have to use an individual tariff in each case.348 The possibilities for individual differentiation will inevitably also depend upon the value of the particular insurance policy. For mass insurance products with a low premium, risk differentiation can only take place in general categories. In professional liability insurance, however, the benefit of detailed differentiation, rewarding an enterprise for preventive action, may well outweigh the costs. It is thus not difficult to make an economic argument in favour of effective risk differentiation as a remedy for emerging systemic risks. If good risks are not rewarded for preventive action, either they will not have an incentive for

110


3.


prevention or they will leave the risk pool, and consequently the risk pool will be unravelled, as described by Priest.

2. Risk differentiation: the example of environmental risk These general principles concerning the control of moral hazard obviously apply to systemic risks as well. We will illustrate this using the example of environmental risk. As we indicated earlier, the ideal and first-best method of control is to monitor the behaviour of the insured as much as possible and to react to his behaviour by requiring specific preventive measures through the policy conditions and by tailoring the premium accordingly. Such individual risk monitoring may obviously be costly. On the other hand, in environmental insurance the amounts at stake are obviously larger than in consumer insurance policies, for example. Hence, this may merit investment in obtaining information on the individual risk. Moreover, some classification based on prior experience is possible as well. It is thus essential for the insurer of an environmental risk to obtain accurate information on the risk ex ante and during the performance of the insurance policy, and preferably at the lowest cost possible.349 The insurer can of course rely on existing mechanisms to control the "ecological reliability" of his insured. In this respect, one should not forget that most insured parties for environmental risks are licensed operators. An absolute minimum would thus be that the insurer checks whether the insured operator in fact possesses a valid licence. In addition, he could require expressly in the policy conditions that the operator follow the conditions of the licence and could make the insurance coverage even dependent on it. In other words, the insurer could exploit the fact that a decision concerning the way to reduce environmental harm has already been made by an administrative agency and possibly laid down in the administrative licence. This is the current practice in Belgian environmental liability insurance. Many insurance policies stipulate that there would be no coverage if the damage resulted from a violation of regulatory norms.350 There is thus a close interdependence between regulation and insurance. By relying partially on regulation (by requiring that the insured should as a minimum comply with regulatory standards), the insurer can take advantage of the work of the regulator to control moral hazard. On the other hand, the control of regulatory compliance by the insurer will also have a public policy benefit, and the insurer thereby becomes a regulatory enforcer. Hence, regulation and the control via the insurer become mutually dependent. Optimal control of systemic risks will inevitably require an optimal mix of private and public regulatory tools. As we will show below, there may be economies of scale to be derived from regulatory standard-setting, but on the other hand regulation cannot foresee every possible situation ex ante. Therefore private controls (e.g. by insurance companies) will have to supplement regulatory requirements.


111

3.


However, it is important to stress that the literature indicates that regulatory compliance is, from the standpoint of tort law, usually merely considered a minimum. If socially efficient care is higher than the regulatory standard, the potential polluter may be held liable, even in the case of regulatory compliance.351 However, from the insurer’s perspective it seems useful to rely at least on regulatory standards, even though they may not always constitute the optimal norm and even though the insured may be held liable even if tough regulatory standards have been met. Compliance with regulatory standards is, in other words, a minimum, which an insurer could require. But other techniques should be used as well to ascertain whether the insured could have taken more cost-effective prevention measures than the ones required by regulation.

3. The need for specialisation All these examples show that there are a variety of techniques that insurers can use to control moral hazard in the case of systemic risks. However, they also make clear that such control is only possible if the insurer has adequate information on the risks he is covering. For an average insurer who covers e.g. traffic liability, fires and professional liability, and who occasionally writes an insurance policy for environmental liability as well, this may be a problem. Information may be purchased at lower cost if the insurer is willing to specialise in particular systemic risks. This specialisation may give the insurer adequate information on the appropriate techniques to reduce specific systemic risks. Thus he may attract good risks by offering them a lower premium if certain preventive measures are taken. In the end, therefore, specialisation may give the insurer a comparative advantage. At the same time, specialisation may be the only way in which some complex systemic risks can be insured anyhow, controlling moral hazard in an optimal way.

G. Summary In this chapter, we briefly addressed the insurability of emerging systemic risks by outlining the basic conditions of insurability. On the one hand we addressed the issue from a theoretical perspective; on the other hand, examples were given of several systemic risks (such as nuclear risk, environmental and health-related risks). These examples made clear the extent to which insurance markets are able to use a variety of insurance techniques to guarantee better availability of insurance. We indicated at various places in this section that insurability is not a black-and-white issue. It is not a matter of saying that a certain risk is insurable or not. It is more akin to a sliding scale on which one can plot the factors that may positively or negatively affect insurability. Thus it also becomes possible to indicate how insurers or the legislator can positively influence the availability of insurance. In this chapter, we addressed these issues from a insurance policy

112


3.


perspective; in the next chapter we will see how expanding liability, as discussed in the previous chapter, may lead to further problems for insurers and how either the legislator or insurers can intervene to provide remedies to them. But even this chapter, which merely focused on the general conditions of insurability, already gave some indications of the insurability of systemic risks. Again, it should be stressed that whether a specific systemic risk will be considered insurable or not will to a large extent depend upon the type of risks and the specific insurance market situation in a particular country. Within the scope of this study, only some general concepts could be outlined. We started from the idea that the predictability of risks is the cornerstone of insurability. In that sense, systemic risks are no different from other insurable risks. Moreover, we also showed that there is apparently de facto insurability of many systemic risks, albeit not for the full amount. The general principles of insurance may be undermined inter alia by the fact that the information necessary to make an accurate assessment of the risk will typically be lacking in the case of systemic risks. In particular, because these are newly emerging risks, the insurer will often have too little information to make an adequate assessment of the probability of an accident and the possible magnitude of the damage if the risk materialises. This basic problem, i.e. the lack of information, is obviously not different for emerging systemic risks than for other risks, albeit with high-technology-related risks there may be even less information than with traditional risks. The key issue is obviously that, with newly emerging risks, there is not enough statistical data available to make an adequate assessment of the risk. However, we equally indicated that lack of information and doubts concerning the predictability do not as such make a risk uninsurable. The literature indicated that insurers can also cope with “hard-to-predict” events by asking for an additional risk premium to cope with the ambiguity of those events. However, this supposes that the insured also recognises the uncertainty and the potential of a new risk. Otherwise, the insured may not be willing to pay the additional risk premium. We equally indicated that the uncertainty does not only relate to the accuracy of the risk assessment (necessary to calculate the actuarially-fair premium) but also to the preventive measures the insurer should require to prevent systemic risks. We indicated that the insurer can remedy the problems of moral hazard and adverse selection by providing an optimal risk differentiation, and thereby increase insurability. However, this supposes that the insurer can identify good and bad risks and should therefore know some of the possible causes of systemic risks and the potential preventive strategies. Government regulation in particular can play an important role in this respect. If the government issues regulatory standards concerning the prevention of systemic


113

3.


risks, the insurer can in turn use these standards e.g. by requiring that the insured at least comply with mandatory government regulation. Finally, it should be remembered that adequate risk differentiation is only possible if the insurer has adequate information on the systemic risks he is covering. This may well require the insurer to specialise in specific risks. Moreover, we indicated that there may be a potential conflict between, on the one hand, the need for adequate risk differentiation and, on the other, the need for insurers to co-operate in order to obtain adequate information on risk prevention. This co-operation may come into conflict with the need for an adequate competition policy. However, it should be possible to co-operate merely on exchanging statistical information on accident risks, for example, while still competing with respect to premiums. However, this is a conflict, which cannot always be resolved precisely because information is also an important competitive tool. Charging an appropriate risk premium, risk differentiation, information provision and specialisation, are thus instruments that the insurer can use to increase the insurability of systemic risks. We used the example of environmental risks to show how this may work in practice. However, insurability is not only influenced by the ability to predict the probability of harm (p), but also by the magnitude of the damage (D). The capacity of insurers to cope with very large accidents (think about the destruction of the Twin Towers on 11th September 2001 or other terrorist attacks) may well be limited. We showed that although the individual capacity of one insurer may be limited, there are various insurance techniques, which may be used to increase capacity. Co-insurance, re-insurance and pooling have been used in the past, e.g. to cover earthquake risks, and provided coverage for relatively spectacular amounts. But still, there is the inevitable fact that the magnitude of some risks will be so high, that even with re-insurance and pooling techniques, coverage will not be available. This may prompt recourse to alternative financing strategies, some of which we will discuss below. On the other hand, some may argue that the limited availability of coverage on financial and insurance markets should lead legislators to limit the liability of enterprises. Indeed, one can seriously question whether enterprises should be held liable if it is judged that financial coverage to compensate the losses if the risks materialise is not available to the full extent. Nevertheless, we argued that it would be a wrong policy to cap the amount of damages due by a party who is responsible for a systemic risk, since this may have various negative effects. Keeping full liability in principle in place may provide exactly the necessary incentives to search for financial and insurance products to cover those risks. Therefore we argued that although capacity will unavoidably lead to serious limitations on the possibility of insurance markets to cover the consequences of catastrophic accidents, this should not as such be an

114


3.


argument for limiting the liability of those enterprises (or systems) that are responsible for systemic risks. In sum, in this chapter we addressed some of the general conditions of insurability, applying them to various systemic risks and indicated what insurers can do to increase the insurability of systemic risks. In the next chapter, we will come back to the phenomenon of expanding liability for systemic risks and ask ourselves in what way legislators or insurers can or should react to it. Indeed, the question arises as whether some of the trends in legislation or case law that we identified in Chapter 2 may negatively affect insurability as it has been identified in this chapter.

Notes 253. See K. Arrow, Aspects of the Theory of Risk-Bearing, Helsinki, Yrjö Jahnssonin Säätiö, 1965 and K. Borch, The Utility Concept Applied to the Theory of Insurance, 1961, The Astin Bulletin, 245-255. 254. See S. Shavell, Economic Analysis of Accident Law, 1987, p. 190. 255. This argument has been made by G. Skogh, The Transactions Cost Theory of Insurance: Contracting Impediments and Costs, Journal of Risk and Insurance, 1989, 726-732. 256. Monti rightly points out that the fact that there may be both factual and legal uncertainty: A. Monti, Environmental Risk: A comparative law and economics approach to liability and insurance, ERPL, 2001, 51-79. 257. J. Rogge, Les assurances en matière d’environnement, 1997, p. 4. 258. White Paper on environmental liability, p. 23. 259. H. Kunreuther/R. Hogarth/J. Meszaros, Insurer Ambiguity and Market Failure, Journal of Risk and Uncertainty, 1993, 71-87. We will come back to the importance of “insurer ambiguity” when dealing with the insurability of retrospective liability. 260. Official Journal L398/7 of 31 December 1992. For a comment on this exemption regulation see G. Levie/H. Cousy (eds.), La politique Européenne de concurrence en matière d’assurance, 1994. 261. COM, 1999, 192 final. 262. See report, No. 47. 263. On the applicability of the cartel exemption to insurance more generally, see M. Faure/R. Van den Bergh, Restrictions of Competition on Insurance Markets and the Applicability of EC Anti-Trust Law, 1995 Kyklos, 65-85 and M. Faure/R. Van den Bergh, Aansprakelijkheidsverzekering, concurrentie en ongevallenpreventie, in: T. Hartlief/M.M. Mendel (eds.), Juridische beschouwingen over de maatschappelijke rol van verzekeringen en verzekeringsmaatschappijen, 2000, pp. 315-342. 264. See J. Rogge (supra note 257) pp. 3-4. 265. In addition an increasing number of alternative possibilities for financing potential liability for systemic risks are emerging. These will be discussed in Chapter 5 of this report.


115

3.


266. H. Cousy, Recent developments in environmental insurance, in: F. Abraham/ K. Deketelaere/J. Stuyck (eds.), Recent Economic and Legal Developments in European Environmental Policy, 1995, pp. 227-241. 267. See M. Faure/R. Van den Bergh, Liability for Nuclear Accidents in Belgium from an Interest Group Perspective, 1990, IRLE, 250-251. 268. S.M.S. Reitsma, Nuclear Insurance Pools: History and Development, paper presented at the OECD Symposium on Nuclear Accidents, Liabilities and Guarantees, Helsinki, Finland, September 1992. 269. See J. Dow, The Organization and Development of International Liability Capacity and National Market Pools, with special reference to new “nuclear countries”, in: Nuclear Third Party Liability and Insurance, 1985, pp. 172-182. 270. For details on the functioning of SYBAN see T. Vanden Borre, Dekking van het nucleaire risico op nationaalrechtelijke basis of via internationale verdragen: de Verenigde Staten versus Europa, in: K. Deketelaere/M. Faure/G, Verhoosel (eds.), Grensoverschrijdende milieuproblemen: uitdagingen voor de nationale en internationale rechtsorde, 1998, pp. 472-477. 271. See J. Dow (supra note 269) p. 180. 272. Muller notes in this respect: “In view of the rising cost of erecting nuclear energy plants, nuclear property insurance, which is likewise borne by the nuclear pools, is under considerable pressure and, in turn, represents an involvement by the insurance industry to the machinery insurance which, in the case of a nuclear power plant, also goes into the billions. Both forms of cover have priority over liability insurance, since a theoretical large scale nuclear occurrence would probably first affect the material assets within the plant, then the surrounding area. It is naive to consider only the third party suffering loss or damage – as occasionally happens – and to regard property insurance as an unnecessary appendage which only absorbs capacity. Every reasonable person knows that a nuclear power plant requires a heavy investment and that not only the operators, but also their creditors, should be protected. It is quite simply foolish to regard the loss of this investment as a sort of “punishment” for having brought about a nuclear occurrence and to ignore the interests of the power supply company and the investors in safeguarding their material assets” (W. Müller, The role of insurance industry in covering nuclear third party liability risks, in: Nuclear Third Party Liability and Insurance, 1985, p. 171). 273. See the discussion in Nuclear Third Party Liability and Insurance, 1985, pp. 192-196. 274. In Belgium a European Mutual Association for Nuclear Insurance (EMANI) was founded in 1978. By 1984 it already provided cover of BF 1.4 billion for property damage per nuclear installation in Belgium. 275. See H. Cousy (supra note 266) p. 234 and J. Rogge (supra note 257) pp. 36-37. 276. This will be discussed below in relation to the alternative compensation mechanisms in Chapter 5. 277. On this point, see for example M. Hamburger, De aansprakelijkheidsverzekering in Nederland voor schade door kernongevallen in West-Europa, 1996, NJB, 1340. 278. See C.W. III Havens/R.M. Theisen, The Application of United States and ECC Antitrust Laws to Reinsurance and Insurance Pooling Arrangements, 1986, The Antitrust Bulletin, 1301. 279. Report, No. 26. 280. Report, No. 28 (B).

116


3.


281. See report, No. 32. 282. See J.R. Tyran/P. Zweifel, Environmental risk internalization through capital markets (Ericam): the case of nuclear power, 1993, IRLE, 431-444. 283. This remarkable statement is a literal quote from D. Ranson, Verzekering van milieuaansprakelijkheid, Milieu- en Energierecht, 2000, 66-73. p. 72. 284. See Chapter 7. 285. Parliamentary Documents, Second Chamber of Parliament, 23 April 1991, 72-4061. 286. See inter alia the critical questions by Van Rijn-Vellekoop, Second Chamber of Parliament, 23 April 1991, 72-4046 and of the De Korte, Second Chamber of Parliament, 23 April 1991, 72-4052. 287. For further details see M. Faure, De verzekering van het nucleaire risico, in: In volle verzekerdheid, Essays offered to Prof.Mr. A.J.O. Van Wassenaer van Catwijck, 1993, pp. 241-254. 288. For a critical analysis see also M. Faure/R. Van den Bergh (supra note 26) pp. 241-254. 289. See Parliamentary Documents, Belgian House of Representatives, Doc 1999/2000, 50 0560/001, of 3 April 2000. The Act was subsequently passed, becoming the Act of 11 July 2000, Moniteur Belge, 4 October 2000. For a comment see T. Vanden Borre, Recente ontwikkelingen in het nucleaire aansprakelijkheidsrecht: innovatie of renovatie?, 2000, Milieu- en Energierecht (MER), 25-49 and T. Vanden Borre, Efficiënte preventie en compensatie van catastroferisico’s. Het voorbeeld van schade door kernongevallen, 2001, pp. 323-326. 290. H. Hausmann/R.H. Kraakman, Toward Unlimited Shareholder Liability for Corporate Torts, 1991, 100 YLJ p. 1879 291. M. Trebilcock/R. Winter, The economics of nuclear accident law, 1997, IRLE, pp.215-243. 292. J. Boyd/D. Ingberman, Noncompensatory damages and potential insolvency, 1994, JLS, 895-910. 293. In Dutch referred to as a “rechterlijk matigingsrecht”. For details see M. Faure/ T. Hartlief, Gevolgen van de uitbreidende werkgeversaansprakelijkheid: beleidsconsequenties voor verzekeraars?, in: M. Faure/T. Hartlief (eds.) Verzekering en de groeiende aansprakelijkheidslast, een juridisch, gezondheidskundig en economisch onderoek naar ontwikkelingen met betrekking tot de aansprakelijkheidslast en de consequenties voro verzekeraars naar aanleiding van de werkgeversaansprakelijkheid voor bedrijfsongevallen en beroepsziekten, 1995, pp. 309-315. 294. On the difference between global (statutory) limits and ceilings versus ad hoc control by the judiciary see: W.V.H. Rogers/J. Spier/G. Viney, Preliminary Observations, in: J. Spier (ed.), The limits of liability. Keeping the Floodgates Shut, 1996, pp. 1-15. 295. For an overview of the arguments advanced in legal doctrine in favour of caps see M. Faure/R. Van den Bergh, Objectieve aansprakelijkheid, verplichte verzekering en veiligheidsregulering, 1989, pp. 339-342 and F.J. De Vries, Wettelijke limitering van aansprakelijkheid, 1990. 296. See T. Hartlief/R.P.J.L. Tjittes, Verzekering en Aansprakelijkheid, 1994, and T. Hartlief/R.P.J.L. Tjittes, Invloed van verzekering op de civiele aansprakelijkheid, Preadvies voor de Vereniging voor Burgerlijk Recht, 1990. 297. See B. Wessels, Zorgen om morgen: beheersing van beroepsaansprakelijkheid in discussie, 1995, NJB, 235 and B. Wessels, Risicobeheer bij preadvisering, in Preadvies Ne Voa, 1995, p. 54.


117

3.


298. See M. Faure/T. Hartlief (supra note 41) p. 313. The fact that the legislator nevertheless often introduces financial caps for one specific activity can be explained as a result of effective lobbying by the specific interest group representing that activity. 299. S. Shavell, Strict Liability versus Negligence, JLS, 1980, p. 8 and p. 11. 300. See M. Faure, Economic Models of Compensation for Damage Caused by Nuclear Accidents: Some Lessons for the Revision of the Paris and Vienna Conventions, 1995, European Journal of Law and Economics (EJLE), 21-43. 301. The reason for the underdeterrence is obviously the same as for the un derdeter ren c e which res ults from the inso lvency of the injurer. Underdeterrence arises because the injurer is not exposed to full liability, either as a result of his insolvency or as a result of a cap. 302. This point has been made by S. Rea, 50-52, but also by M. Adams, Warum kein Ersatz von Nichtvermögensschäden, in: C. Ott/H.B. Schäfer (eds.), Allokationseffizienz in der Rechtsordnung, p. 214 and by C. Ott/H.B. Schäfer, Schmerzensgeld bei Körperverletzungen. Eine Ökonomische Analyse, 1990, JZ, 564-565. 303. See M. Faure/R. Van den Bergh (supra note 267), M. Faure, 1995 and M. Trebilcock/ R. Winter (supra note 291). 304. M. Radetzki, Private arrangements to cover large-scale liabilities caused by nuclear and other industrial catastrophes, 2000, GPRI, 180-195. 305. M. Faure (supra note 287) pp. 247-260. 306. A similar argument has been made in Germany. See G. Wagner, Versicherungsfragen der Umwelthaftung, in: M. Ahrens/J. Simon (eds.) Umwelthaftung, Risikosteuerung und Versicherung, 1996, pp. 101-102. 307. A. Heyes/C. Liston-Heyes, Capping Environmental Liability: the Case of North American Nuclear Power, 2000, GPRI, 196. 308. This has been demonstrated with a numerical example by S. Shavell (supra note 291) p. 209. 309. See also M. Faure/R. Van den Bergh (supra note 22) pp. 130-131. 310. This has been suggested by S. Shavell (supra note 12) p. 210. 311. This argument in favour of caps in specific circumstances is also made by Boyd and Ingberman (supra note 292) pp. 895-910. They however, apply this specifically to the case where an injurer would be confronted with a set of possible losses rather than with one single loss. Moreover, their analysis does not include the effects on insurance. 312. Although risk aversion may also lead to higher care, as Shavell indicated. 313. An event is defined as insurable when insurers can set a premium which reflects the risk and enables them to make a profit and subsequently a market emerges (M. Faure/P. Fenn, Retroactive liability and the insurability of long-tail risks, 1999, IRLE, 487-500. See generally on insurability problems the papers published in The Geneva Papers on Risk and Insurance, 1995, 407-462 and M. Faure, The limits to Insurability from a Law and Economics Perspective, 1995, GPRI, 454-462. 314. See for the basic literature on insurer ambiguity H. Kunreuther/R. Hogarth/ J. Meszaros (supra note 259) pp. 71-87. 315. These problems have been discussed extensively in M. Faure/P. Fenn (supra note 313).

118


3.


316. See M. Hinteregger, La nouvelle loi autrichienne sur la responsabilité civile pour les dommages nucléaires, 1998,62 Bulletin de Droit Nucléaire, 27-34. 317. See M. Trebilcock/R.A. Winter (supra note 291), p. 221. 318. On devices to control moral hazard in general, see S. Shavell, On Moral Hazard and Insurance, 1979, Quarterly Journal of Economics (QJE), 541-562. See also M. Faure/T. Hartlief (supra note 15) pp. 681-706. 319. For estimates see M. Faure (supra note 300) pp. 29-31. As a result of recent changes the amount has, however, been increased. See T. Vanden Borre, Efficiënte preventie en compensatie van catastroferisico’s. Het voorbeeld van schade door kernongevallen, 2001, pp. 47-48. 320. See for instance M. Faure (supra note 300) p. 33. 321. This point has been made by M. Faure/R. Van den Bergh (supra note 26) p. 241. 322. See the basic insights by J.M. Buchanan/G. Tullock, The calculus of consent, Ann Arbor, 1962; M. Olson, The logic of collective action, 1971; S. Peltzman, Toward a more general theory of regulation, 1976, JLE, 211-240 and R. Posner, Theories of Economic Regulation, 1974, Bell Journal of Economics, 335-358. 323. This has been convincingly proven in the recent dissertation by T. Vanden Borre, Recente ontwikkelingen in het nucleaire aansprakelijkheidsrecht: innovatie of renovatie?, 2000, Milieu- en Energierecht (MER), pp. 101-111 and pp. 225-246. 324. M. Faure (supra note 300) pp. 33-35. 325. See, for instance, G.E. Van Maanen, Pleidooi voor verbetering van de rechtspositie van slachtoffers van kernongevallen, 1986, NJB, 1342; and G.E. Van Maanen, De civielrechtelijke aansprakelijkheid voor kernongevallen naar Nederlands Recht, in: M. Faure (ed.), Aansprakelijkheid voor het Nucleaire Risico, 1993, p. 19. Note, however, that both with respect to oil pollution and nuclear accidents, as a result of political pressure there is a tendency to amend the existing conventions with a view to increasing the available amounts. See, for the nuclear conventions, T. Vanden Borre (supra note 323) pp. 40-42. 326. Article 16 of the Product Liability directive. 327. See the overview of the transposition into domestic law of the Product Liability Directive, provided in the Green Paper on liability for defective products, pp. 35-36. 328. It was set at 70 million ECU in Article 16 of the Directive for damage resulting from death or personal injury if it was caused by identical items with the same defect (serial damage). 329. Green paper on liability for defective products, p. 26. 330. We will show this below in Chapter 6. 331. The combination of regulation with capped liability is studied by M. Trebilcock/ R. Winter (supra note 291) pp. 215-243. 332. The 1969 Brussels convention on civil liability for oil pollution damage. 333. M. Pauly, The Economics of Moral Hazard: Comment, 1968, AER, 531-545. 334. See G. Wagner (supra note 306) pp. 104-105. 335. See S. Shavell, On Moral Hazard and Insurance, 1979, Quarterly Journal of Economics (QJE), 541-562. 336. M. Spence/R. Zeckhauser, Insurance, Information, and Individual Action, 1971, AER, 380-391.


119

3.


337. See van J.C. Eijk-Graveland, schadeverzekeringsrecht, 1998.

Verzekerbaarheid

van

opzet

in

het

338. J. Marshall, Moral Hazard, 1976, AER, 880-890. 339. See G. Akerlof, The market for “lemons”: quality, uncertainty and the market mechanism, 1970, QJE, 488-500. 340. See also G. Wagner (supra note 306) pp. 105-106. 341. Although he mentions that some larger companies are (in 1996) able to get coverage up to 1 billion Belgian Francs for environmental risks. 342. J. Rogge (supra note 257) p. 5. 343. So J. Cowell, Compulsory environmental liability insurance, in: H. Bocken/ D. Ryckbost (eds.) Insurance of environmental damage, 327. 344. However, the reason why a flood insurance did not develop had also to do with anticompetitive behaviour of insurance companies. See in this respect below Chapter 6, F, 2.2. 345. G. Priest, The current insurance crisis and modern tort law, 1987, YLJ, 1521-1590. Priest has been criticised by Viscusi, who claims that there were other reasons for the product liability crisis in the US than adverse selection on its own (W.K. Viscusi, The Dimensions of the Product Liability Crisis, 1991, JLS, 147-177). 346. K. Abraham, Environmental Liability and the Limits of Insurance, 1988, 88 CLR, 949-951. 347. Generally, the question is whether the benefits of particularization outweigh the costs, an issue which has to be addressed in tort law when a standard of care is defined (R. Posner, Economic Analysis of Law, 1998, pp. 187-189) but also when legal rules are made (I. Ehrlich/R. Posner, An Economic Analysis of Legal Rule-Making, 1974, JLS, 257) or standards are set (A.I. Ogus, Quantitative Rules and Judicial Decision-Making, in: P.Burrows/C. Veljanovski (eds.), The Economic Approach to Law, 1981, pp. 210-225; A.I. Ogus, Standard Setting for Environmental Protection: Principles and Processes, in: M. Faure/J. Vervaele/A. Weale (eds.), Environmental Standards in the European Union, 1994, pp. 25-37). 348. On the costs of risk differentiation, see S. Bohrenstein, The Economics of Costly Risk Sorting in Competitive Insurance Markets, 1989, IRLE, 25-39; and W.P.J. Wils, Insurance Risk Classifications in the EC: Regulatory Outlook, 1994, OJLS, 449-467. 349. On the importance of information for the insurer of the environmental risk, see A. Endres/R. Schwarze Allokationswirkungen einer Umwelthaftpflicht-versicherung, 1991, Zeitschrift fürUmweltpolitik und Umweltrecht, 1-25. 350. J. Rogge (supra note 257) pp. 28-29. 351. This will be discussed below.

120



Chapter 4

Remedies for Dealing with Expanding Liability for Systemic Risks


121

4.

REMEDIES FOR DEALING WITH EXPANDING LIABILITY FOR SYSTEMIC RISKS

A. Introduction In the previous section, we discussed a variety of instruments that are available for insurers for dealing with increasing systemic risks. An upper limit on coverage may be an appropriate remedy for uncertainty concerning the magnitude of the damage; charging an appropriate risk premium can be used to deal with insurer ambiguity. Lastly, risk differentiation is the general means of coping with expanding systemic risks. In this section, we will turn back to Chapter 2 where we indicated that it was expanding liability for risks that was causing major worries to insurers and prompting them to adapt their behaviour accordingly. Expanding liability for risk leads, as we already mentioned in the introduction, to claims by the insurers that unlimited liability may lead to uninsurability of certain risks. In this section, we will focus more closely on the relationship between expanding liability for systemic risks and insurability. We will first of all address various aspects of expanding liability law by asking whether they endanger the insurability of risks. We will show that some of them do whereas others do not. Moreover, we will look at various other remedies that may be available for insurers to cope with expanding liability. We have already asked whether financial caps on liability should be introduced as a remedy for uninsurability, but we answered in the negative. In addition, we made the case for risk differentiation as a general tool for coping with increasing liability risks. However, there may be other insurance techniques as well which could be used by insurers (and have been used in practice as well) to deal with increasing liabilities for systemic risks. The relationship between liability and insurance will therefore be the central focus of this chapter. Note, however, that most of the remedies discussed in this chapter are not merely remedies against expanding liability, but rather, more generally, remedies against increased exposure to risk.

B. Foreseeable risks 1. Principle: insurability As long as insurers can foresee new risks, they can calculate the risk beforehand, demand a corresponding premium and build up a reserve. Indeed, as long as the insurer is able to demand a premium that is equal to the probability (p) of an event with a damage (D) occurring, together with a surplus

122


4.


for administrative costs, there should not be a problem of insurability provided that the problems of adverse selection and moral hazard mentioned above can be controlled as well. Irrespective of whether the change in the scope of liability involves a higher probability of being found liable (an increase in p in other words) or an increase in the damage (D), a competitive insurance market will, in principle, be able to cope with the accompanying rising demand for insurance, provided that there is sufficient capacity. Many of the changes taking place in liability law which tend to increase liability for systemic risks do not as such cause uninsurability because they are relatively foreseeable. Changes in liability law have been taking place since the beginning of the 1960s and most of these changes have only resulted in an increase in the liability of businesses. Since new case law or legislation applies to future situations, an insurer can assess whether this increase in liability amounts to a change in (p) and/or (D), and therefore necessitates a change in premium or a change in policy conditions. Legal practice has also demonstrated that insurers are well able to deal with some of these changes.

2. Example: European Products Liability Directive A relevant example is the European Product Liability Directive of 25 July 1985.352 For some European member States, which already had product liability regimes at the time of the promulgation of the directive, this directive probably did not even substantially enlarge the scope of liability for manufacturers. For other member States, such as Spain and Greece,353 it probably did. However, the strict liability regime of the new directive only applies to situations that occur after the new regime came into force. Art. 17 of the Product Liability Directive explicitly states that the directive applies only to products put into circulation after the legislation implementing the directive has come into force. The implementation of the European Product Liability Directive does not therefore seem to have posed too many problems from an insurance perspective, since it respected the golden rule that the new regime was only applicable to new risks. This seems to comply with the recent findings of the so-called “Green paper on liability for defective products”. The Product Liability Directive of 1985 had called for a five- year review to analyse its working.354 This five-year review resulted in a Commission Green Paper issued on 28 July 1999, which was addressed to the larger European business and consumer community with two aims: 1. To seek information for the purpose of assessing its application in the field in view of the experience of those concerned (in particular industry and consumers), and to establish definitely whether it is achieving its objectives; 2. To “gauge” reaction to a possible revision as regards the most sensitive points of this legislation.355


123

4.


The European Commission wishes to promote reflection and debate and therefore invites replies based on facts. Obviously the European Commission has laid down guidelines for discussion concerning all of the important topics in the product liability area, such as the existence of financial limits, the-10 year deadline, the burden of proof, the assessment of the insurability of risks, suppliers’ liability and the type of goods and damage covered. Although the green paper clearly states that it does not prejudge the European Commission’s position on these topics, the fact that a lot of topics which were previously controversial are now on the agenda for possible reform is interesting in the light of the analysis provided in this study. Indeed, as far as product liability insurance is concerned, the European Commission wishes to analyse how the cases brought under the directive have affected the insurance sector,356 but apparently it is assumed that the product liability risk is covered through insurance. This also follows from a brochure put out by the Swiss Reinsurance Company on product liability claims in Europe, which describes a wide range of cover for the product liability risk in all member States.357 This obviously ties in with the general finding in the legal and economics literature that strict liability will lead to a demand for third-party insurance from risk-averse injurers.358 A shift from negligence to strict liability will, therefore, lead to a shift from first-party victim insurance to third-party liability insurance. As long as strict liability also allows a narrowing of risk groups, this shift towards strict liability should not as such lead to uninsurability.359

C. Retrospective liability This picture, however, changes when liability expands in such a way that it is no longer foreseeable and policy conditions can hence not be adapted ex ante. The question whether a change in the liability regime leads to uninsurability, when it is applied retrospectively, thus seems to be answerable relatively easily. At first sight one could argue that this certainly is the case. If the insurer were not aware that the behaviour of the insured party might potentially have been considered wrongful, no premium would have been charged for this risk, no preventive measures would have been required in the policy conditions and no reserves against losses would have been set aside. Indeed, insurance assumes that the insurer covers future risks which are at least to some extent foreseeable. Insurance requires some degree of predictability. However, the mere fact that insurers of, say, industrial waste disposal sites in the 1960s did not – as a matter of fact – foresee that the activities of their insured parties could lead to a liability in the future, does not make the event totally unforeseeable. The possibility of a change in the scope of liability is an uncertain element which the insurer can – in principle – take into account ex ante. We are dealing

124


4.


here with the concept of "insurer ambiguity" addressed by Kunreuther, Hogarth and Meszaros,360 which was discussed above. If insurers could foresee the likelihood of a possible change in the liability system, they could react to uncertainty by estimating the probability that this event would occur and charge an additional risk premium to account for the legal uncertainty. In sum: from an ex ante perspective one can argue that nothing is totally unforeseeable or unpredictable; insurers can in principle cope with "hard-to-predict" events such as the introduction of retrospective liability by charging an additional premium. However, from an ex post perspective this message is not very helpful for insurers who, at the time, did not take the risk into account and now have to provide cover to enterprises for risks which the insurers considered apparently unforeseeable. Hence, no additional premium was charged and no reserves were constituted, which explains why the retrospective liability, which is now, laid down, say, for soil clean-up costs, leads to major problems for insurers.361 There is thus no problem of uninsurability of retrospective liability as such, but only the simple fact that insurers did not take these risks into account when the policy was drawn up. It is apparently primarily the retrospective character of the so-called Superfund regime put in place by CERCLA in the United States that has caused the most worry to enterprises and their insurers. From the above it follows that real retroactive liability, where any change in the law was not foreseeable, would pose problems for insurers. Also insurers argue in their publications that liability for past pollution is uninsurable.362 The retrospectivity problem may especially arise with environmental risks such as soil pollution. It could, however, play a role with other systemic risks as well. Think of the cases where employees have been exposed to dangerous substances and where the negative health effects only arise years later. If the employer were held liable years later whereas the exposure was not wrongful when it occurred, this may lead to uninsurable retrospectivity. Our theoretical analysis supports this claim. However, the problem of a long time-lapse between the wrongful event (e.g. the emission of a toxic substance) and the damage will obviously often arise in environmental and other systemic risks. A question which is closely related to the issue of retrospectivity is whether the system of coverage should be changed to be able to cope with these long-tail risks. More particularly, the question of how insurance cover in time should be taken care of arises. This issue of cover over time will be discussed below as one of the possible remedies for expanding liability. There are, however, two other features in liability law that we should discuss as well, because they may also endanger the insurability of a risk.

D. Shifting the risk of causal uncertainty Several examples of this shift of the risk of causal uncertainty have been presented in Chapter 2.363 Above we already argued that such a shift may


125

4.


counter to the principles of liability law. But It may also endanger the insurability of systemic risks. The danger of shifting the burden of causal uncertainty to the enterprise is that the insurer of the specific employer or producer will be required to compensate for damage which, on the whole, had probably not been caused by the insured party.364 Unless a proportionate liability rule is followed, it is not possible to cover a risk if that would mean that the insurer would not only cover the damage of his own insured parties but also the damage that might possibly have been caused by another party. The upshot of these trends is that enterprises would be liable for risks that they have not caused themselves (in the case of causal uncertainty) or for risks that were not foreseen at the time when the tort was committed (in the case of retrospective liability). They stem largely from a hidden redistributive agenda: the wish to provide victim protection no matter what it may cost. These trends may be far more problematic from an insurability point of view than the shift towards strict liability itself. Indeed, whereas we argued that strict liability as such is insurable, this is no longer true if retrospective liability is introduced or the risk of causal uncertainty is shifted to the enterprise.

E. Joint and several liability, and channelling of liability All of the aforementioned features, which depart from the principle that an insurer can only be held liable for damage caused by the party insured by him, may lead to difficulties when it comes to insuring environmental liability. Joint and several liability exhibits another departure from the principle that an insurer should only be held liable for the damage that the insured party has caused. This may cause difficulties when it comes to insuring systemic risks.

1. Deterrence This is an area, which is closely related to the issue of causal uncertainty that we have just discussed. There is a trend to hold joint tortfeasors jointly and severally liable for all the damage to which their behaviour might have contributed. The reasons for doing so are well known. For the victim, it is often difficult to prove a clear causal link with the actions of one particular enterprise. This may sometimes lead to alleviation of the burden of proof, or to holding several insurers jointly and severally liable. Again, the often-debated Superfund (CERCLA) in the United States is an example of such a joint and several liability regimes. The effect of joint and several liability is obviously that the risk of insolvency is shifted to the injurer who will be sued by the victim. Indeed, joint and several liability means that the victim can claim full compensation from one injurer who then can seek redress from the other parties who contributed to the loss in proportion to their contribution. If, however, the other parties are all insolvent, the injurer will have to cover the total loss, including the losses he has not caused.

126


4.


In addition the risks of uncertainty concerning the causal link are, under joint and several liability, also shifted to the one injurer who is sued in the particular case. The victim just has to sue one of the many potentially liable insurers and claim full compensation. If the injurer does not succeed in proving that others contributed to the loss, liability for the damage will ultimately fall on him. Although it could be argued in favour of joint and several liability that, being ex ante, it should give excellent incentives for mutual monitoring of potential injurers365, there seem to be disadvantages as well. Joint and several liability may violate the principles of fair and efficient compensation which hold that an injurer should in principle only be required to provide compensation in proportion to his contribution to the loss.366 The effects on deterrence largely depend upon the legal regime chosen and upon the insolvency of the injurers.367

2. Insurance From an insurance perspective, joint and several liability may be dangerous for the simple reason that the insurer is then no longer merely insuring the risk posed by his insured party (which he can still control), but also the risk caused by all the others.368 The welfare losses resulting from such a system of joint and several liability may be large. Mutual monitoring may ex ante not always be possible and the transactions costs involved (also in the systems of redress) can be huge. Hence, on balance it is doubtful that joint and several liability will have positive incentive effects. The insurance effect is, obviously, as the case of causal uncertainty illustrated, that an insurer will be held liable for the risks that his insured party never caused. The example of the Des-case, where the Dutch Supreme Court applied a joint and several liability rule de facto, may be recalled. This means that any victim can claim full compensation from any of the manufacturers. The insurer of the particular manufacturer then becomes de facto the insurer of the whole market. This trend towards joint and several liability seems therefore to endanger insurability.369

F. Channelling of liability 1. Deterrence There is another feature of some statutes and/or conventions dealing with systemic risks like environmental or nuclear risks which again departs from the principle that only the injurer who caused the damage should be held fully liable for the loss. This is the so-called channelling of liability. Whereas with joint and several liability a victim can in principle claim full compensation from any of the multiple insured, under channelling the liability is channelled to one party who than becomes fully liable for the damage. The channelling, which indicates which party will be held liable for the loss, is often exclusive; it means that the victim can only sue the “channelled injurer” and not another party who might have contributed to the loss as well. Channelling


127

4.


can be found again in international conventions concerning nuclear liability and oil pollution. In nuclear liability conventions, the liability is channelled to the licensee of a nuclear power plant; in the conventions concerning damage caused by marine oil pollution, liability is channelled to the tanker owner. It has been argued that this channelling is inefficient because it has perverse effects on the incentives to take care when the liability applies exclusively to one operator.370 This is the case if it means that victims no longer have the right to sue another party who could influence the accident risk as well. Excluding that third party from liability is inefficient since his incentives to take preventive measures would be diluted. This effect is obviously reduced if the licensee or operator who would be held liable still has a right of recourse against the third party or if liability could be passed on, e.g. on the basis of a contract. In that case, one could argue that the liability is simply transferred and that such a reallocation complies with the principles of the Coase theorem.371 However, this private reallocation of liability may not always be possible and some of the conventions, moreover, even restrict the possibilities of a right of recourse. Channelling can hence hardly be considered an efficient mechanism for preventing accidents.

2. To increase insurability? The reason that we have discussed channelling in this chapter on the relationship between liability and insurability is that some have argued that channelling might improve the insurability of risks. Some have held that as far as nuclear liability is concerned, for example, channelling has the advantage that only the licensee of a nuclear power plant would have to take insurance coverage.372 Also, channelling in oil pollution cases is defended by arguing that it would reduce insurance costs since only the tanker owner would have to take insurance coverage373. This insurance argument is wrong.374 Let us assume that other parties than the licensee or tanker owner would be held liable; they could obviously purchase liability insurance as well. Otherwise, with an exclusive channelling the insurer of the channelled operator would have to cover accidents also in case these have not been caused by his insured, but are allocated to him because of the channelling. Channelling thus creates a greater risk exposure for the operator and therefore creates higher uncertainty for the insurer. If the channelling has any effect on insurability, it is more likely to decrease insurability than to make liability more insurable, as is sometimes argued.375 In sum, there seem to be few arguments in favour of channelling of liability. This may dilute the incentives to act preventively and could even endanger insurability.

128


4.


G. Other remedies for insurers against expanding liability We just discussed a few features, which may seriously endanger the insurability of systemic risks. One concerned the shifting of the risk of causal uncertainty; another one was related to the retrospective liability. As far as the trend in legislation or case law to shift the risk of causal uncertainty to enterprises is concerned, there is relatively little an insurer can do to protect himself from that risk. One can only suggest to the legislators and the judiciary that be relatively careful with this instrument. However, there are various ways insurers can protect themselves against retrospective liability (at least to some extent) and against expanding liability in general.

1. Cover over time 1.1. Options We have already indicated that insurance problems may arise with socalled "long-tail exposure", i.e. when damage occurs much later than the wrongful act. Many systemic risks may have a long-tail character. In such cases it is of utmost importance to examine what the precise period of cover under the particular insurance policy is.376 In this respect, a distinction is usually made between three systems:377 a) The act-committed system: the wrongful act must have taken place during the period of insurance cover; b) The loss-occurrence system: the damage must have taken place within the period of insurance cover; c) The claims-made system: the claim for damages must have been received by the insured or his insurer during the period of insurance cover. Nowadays, a marked increase in claims-made policies throughout Europe is clearly apparent.378 This is due to the obvious disadvantages, at least for insurers, of other two systems in cases of long-tail exposure.

1.2. Changes In the act-committed system, which was especially popular in the 1950s and 1960s, there is only cover if damage can be attributed to a wrongful act of the insured that took place during the period of insurance cover. Obviously, this system is advantageous for the insured party since the potential claim remains in existence until a statute of limitation has passed. In addition, cover is in accord with the policy conditions valid at the moment that the damage was caused, meaning the moment that the wrongful act occurred. As a consequence, limitations of cover that might be introduced at a later stage cannot be invoked against the insured. For insurers, this act-committed system can be highly disadvantageous.379 If an insurer provided cover, for example, for product liability and the policy expired shortly after a certain product was brought onto the


129

4.


market, under an act-committed system of cover the insurer could still be held liable if, several years later, it emerged that this particular product caused damage to a victim. This may be problematic, especially when the insurer has not ex ante foreseen the possibility of "long-tail" exposure and has therefore not charged an adequate premium. These problems caused the international reinsurance market to introduce a loss-occurrence system of cover. Under loss-occurrence, the damage must have taken place within the period of cover. Under loss-occurrence, therefore, cover is also provided if the wrongful act that caused the damage took place before the insurance contract entered into force. However, a loss-occurrence system does not completely eliminate the problem of the long-tail; indeed, if an injury manifests itself during the period of cover, the insurance policy will apply for future consequences over, say, twenty years in the case of long-lasting diseases.380 Therefore, insurers changed to claims-made cover. Although claims-made policies may exist in different forms,381 it is essential to a claimsmade system that the claim for damages should have been received by the insured party or his insurer during the period of insurance cover. Although claims-made policies are probably the only way insurers can protect themselves adequately against long-tail risks Abraham argued that these policies could dilute the deterrent function.382 The argument is that policy-holders under the claims-made system would only internalise the costs that arise during the year of coverage and not the future costs.383 This argument is not totally convincing since, irrespective of insurance, the policyholder will be liable for those future costs as well. Nevertheless, an insolvency problem may arise in the "future" if as a result of the claims-made coverage the covers were limited in time.

1.3. Legal responses Although this system may have some benefits for the insured party as well,384 the most important effect is that it will allow the insurer effectively to control the long-tail risk. This "long tail" may be a particular problem in cases where insurers cannot acquire information on the (perhaps highly technical) risks and thus cannot demand a risk premium to cover insurer ambiguity. Indeed, this hardly-manageable long-tail risk is fully excluded in a claimsmade policy. As soon as the insurance policy expires, the cover ends. This also applies for claims filed after the period of insurance cover for damage that occurred during the period of insurance cover. For this reason, many legal systems have been rather critical of so-called claims-made cover.385 Spanish case law has held that claims-made cover is void.386 In Belgium, the new Insurance Act of 25 June 1992 originally did not allow claims-made cover. However, in both countries this rather rigorous point of view with respect to claims-made cover has recently been relaxed. A recent legislative change in

130


4.


Spain allows claims-made cover on condition that it extends at least one year after the insurance contract expires. Moreover, the Belgian Insurance Act has been amended with respect to the problem of insurance cover over time. The starting point is still the loss-occurrence system, but the Act now provides for the possibility of allowing claims-made cover for certain types of liability insurance via royal decree.387 Again, the Belgian legislation prescribes that some long-tail risk must be covered: cover should extend at least to claims made within 36 months after the insurance contract expires which relate to damage that occurred during the period of cover or to circumstances that could give rise to a claim and which have been announced to the insurer during the period of the insurance cover.388 This brief overview of legislative and jurisprudential responses to claimsmade cover shows that although claims-made cover may be the appropriate response by insurers to long-tail risks,389 legislators and judges apparently pay a lot of attention to the interests of the insured, which might be seriously restricted under claims-made cover. The most important argument against a pure claims-made system is that insured parties may not be awarded cover due to coincidental circumstances, although they paid a premium for the risk that occurred. In addition, claims-made cover may cause problems if the insured party wishes to move to a new insurer who will immediately have to cover the “old” risk. Therefore, legislators or judges, assuming that insured parties may lack appropriate information on the precise consequences of claims-made cover, may react by limiting the effects of claims-made cover in order to protect the insured parties. Nevertheless the increasing trend towards claims-made cover is apparently unstoppable. In the Netherlands recently, the Insurers’ Association advised its members to move to a claims-made system of cover,390 the idea being that the interests of the insured parties can be protected by mitigating the harshest consequences of claims-made cover. One possibility is to provide for an extension of the period of cover after the insurance contract has expired; another possibility is to stipulate that the insured party has the possibility of reporting circumstances or facts that may lead to a claim to the insurer during the period of insurance cover although no damage has yet occurred. Provided that these relevant circumstances or facts have been reported during the period of the insurance contract, cover will be awarded even if the damage only occurs after the insurance contract has expired.391 This is apparently also the system currently provided by the Belgian Insurance Act.

1.4. Acceptability of “claims-made” cover? In fact, there does not seem to be anything inherently unfair about claims-made cover: the cover ends in principle when the insurance policy expires, but the insured party obviously does not pay for cover after expiry


131

4.


either. Another system whereby the "long-tail" risk is shifted to the insurer is worthy of consideration. In that case, one would revert, for example, to traditional act-committed cover systems. The insurer, knowing that he will have to cope with the uncertain long-tail risk will then charge a risk premium so that in the end the premium charged would be higher than under a claimsmade cover system. Again one should realise that the sudden enthusiasm of insurers for claims-made cover can largely be explained as an ex post – and therefore late – reaction to risks which they had not foreseen at the time when the insurance policies were drawn up. Anyway, claims-made policies seem to be a major success in environmental liability insurance and there is definitely a trend to introduce them for other types of systemic risks as well. Whereas authors report that in Belgium, for example, traditional liability policies provided coverage for amounts up to BF 50 million for personal injury damage, the amount of coverage for environmental claims would now be high as BF 1 billion as a result of the introduction of claimsmade coverage.392 This shows that the system of cover in time can apparently have an important effect on the insurability of systemic risks. Although many varieties of claims-made policies exist, the bottom line is that the claim must have been received by the insured or his insurer during the period of insurance cover. Thus the insurer is able to exclude “long-tail” coverage, which seems an adequate remedy for risks which are hard to predict ex ante. However, claims-made policies are of only limited help for dealing with the problem of retrospective liability. Suppose that an insurer providing coverage on a claims-made policy is suddenly confronted with a change in the liability regime as a result of which he has to provide coverage for old risks, which might have been unforeseeable at the time when the insurance policy was concluded. It might then obviously be tempting for the insurer to quickly terminate all the insurance policies on which claims can be expected. But that is obviously an aspect of claims-made policies that many legislators and judges in many European legal systems do not like. There will, therefore, always be the risk that either the claims-made policy will be declared void or that some coverage must still be extended even after the expiration of the insurance policy. Hence, claims-made policies cannot provide insurers with a full guarantee against politicians or judges introducing retroactive liability.

2. Policy conditions as a remedy for expanding liability 2.1. Reactions to changes with respect to the scope of liability We indicated above that problems may arise if the scope of liability is changed to make it retroactive. If the policy conditions do not provide for a solution for this problem, coverage will be provided for events for which no premium was paid, no reserve was set aside and for which no preventive measures have been required through policy conditions simply because the

132


4.


insurer did not actually foresee the particular risk. Some of the problems of retroactive application of new liability regimes may be dealt with by insurance policies with a relatively short period of coverage based on a claims-made system.393 However, this cannot completely remedy the retroactivity problem. Ex ante the ideal solution may be to assess the risk of a change in the scope of liability and to charge a corresponding premium. However, this may not always be possible because of competitive pressures or simply because the insurer lacks information to set an accurate risk premium.394 Therefore, one has to look for alternative solutions as well. One possible solution is to formulate the coverage ex ante in such a way that it only refers to the scope of liability applicable at the time that the insurance contract was concluded.395 Another possibility is to stipulate that the insurer has the right to terminate the insurance contract as soon as the scope of liability is increased through changes in legislation or case law. Obviously, the advantages of such policy conditions lie with the insurer. The insurer will know under which legal regime he offers coverage and can ask for a corresponding premium in accordance with the principles set out above. The disadvantage for the insured is that they are only insured in a very relative sense. Indeed, they may be held liable according to a new trend in case law, for facts that occurred in the past and hence lack insurance coverage. A practical disadvantage of policy conditions limiting insurance coverage to the scope of liability applicable at the time of conclusion of the insurance contract is that it is often not very clear whether there is a change in the scope of liability. Especially as far as case law is concerned, courts will seldom indicate that they are formulating a rule that constitutes a change in comparison to earlier rulings. Sometimes courts also suggest that they are only applying already existing legal rules (e.g. the negligence rule of Art. 1382 of the French Code Civil of 1804) to new situations, so that in fact no new legal rule is created. In these kinds of situations, the insurer may often not be able to avoid providing coverage.396

2.2. Exclusion of certain risks Insurers of enterprises are not only concerned about changes in liability law, but also, for example, about the fact that new diseases are discovered and that higher amounts may be awarded for specific claims. As far as the risk is concerned that new diseases such as repetitive strain injury (RSI) or the Organo Psycho Syndrome (OPS) may give rise to claims, the insurer could opt for coverage limited to the European list of occupational diseases. The advantage is that this makes clear which diseases are covered. Current insurance policies in the Netherlands, for example, provide open coverage for all work-related diseases. Another option would be to exclude certain risks from coverage. One risk that may give rise to large claims is stress at work. Insurers may wish to


133

4.


exclude this type of disease. Obviously the disadvantage of such an approach is that insurers may then choose to make certain risks uninsurable whereas this may not be necessary as long as it is predictable that claims will be made for these “new” diseases. A corresponding premium can then be charged and a reserve set aside.

2.3. Limiting or excluding certain components of the damage If the insurer fears the unpredictability of the amounts awarded for specific types of damage, limitations or exclusions could be written into the contract. One could think of the exclusion of punitive damages or, although this may lead to problems in case law, exclusion of damages awarded for pain and suffering. The most important reason for the increase in the scope of liability seems to be the trend to award higher amounts specifically for pain and suffering. It might be more attractive to opt for maximum amounts for specific components of the damage. The damage as such, including pain and suffering, is obviously not unpredictable. Only the amount may be. Therefore, one could introduce limitations for specific components of the damage instead of general limitations for the whole insurance contract, as is often the case nowadays.

2.4. Acceptability The options we discussed in this section obviously cannot provide a solution to the general problems of uninsurability but they may resolve specific problems insurers encounter in regard to the predictability of certain risks. A key question will, however, often be whether the solutions suggested are also socially acceptable. From a strictly legal point of view, there should not be a problem as long as the insured and the insurer freely choose a specific limited amount of coverage for which the insured pays a corresponding low premium. This is also in keeping with economic principles of insurance. Problems may, however, arise especially in fields where compulsory insurance exists and where it is widely held that the risk of unpredictability should rest with insurers instead of placing it on the shoulders of the insured. This may put pressure on insurers, not only from politicians but also from powerful brokers, to be extremely cautious about introducing policy conditions as suggested above. The legal ex post approach i.e. nullifying certain policy conditions deemed unreasonable in order to protect an individual insured party, may in the end have the adverse effect that certain risks will become uninsurable.

H. Summary In this chapter, we dealt with the relationship between liability and insurability. In Chapter 2, we focused on the reasons for the expanding scope of liability for systemic risks. In that chapter, we argued that there were a

134


4.


number of reasons why it was realistic to assume that there would be greater recourse to suing enterprises for liability in the near future, and thus that there would be a greater claim on their insurers. Alongside changes in tort law, we also pointed to an increased willingness to claim but also to the trend to privatise social security. In the past, many claims were not made because the real costs of an accident were spread over the community by means of social security. It is argued that now– at least in Europe – the social security system is in danger of becoming unaffordable and that governments are increasingly withdrawing from that arena. These developments have led insurers to complain bitterly that the growing liability for systemic risks is no longer insurable. This chapter dealt with the question of whether an increase in the scope of liability would necessarily lead to uninsurability. We have argued that this need not necessarily be the case when the changes are predictable, so that the insurer can demand a commensurate premium beforehand and set aside a reserve. In the previous chapter, we already argued that statutory caps on liability are not an appropriate answer to growing liability since they may lead both to underdeterrence and to undercompensation of victims. They are therefore neither efficient nor just. It is important to focus clearly on the causes of so-called uninsurability before advocating remedies. Indeed, one cannot escape the impression that among certain professions, a generalised sense of panic is spreading about expanding liability for systemic risks, and the argument is advanced that a further expansion of liability may make such risks uninsurable. Apart from the question of whether this is necessarily the case for all systemic risks, it seems questionable whether the liability issue itself depends upon the insurability of a certain risk. Our general finding was that as long as the general principles of liability and insurance are respected, and new rules only apply to new risks, there should be no insurability problems, as the example of the introduction of the European product liability directive has shown. There was no major problem with the implementation of this directive because the directive itself stipulated that the new regime was solely applicable to products which were brought onto the market after the national amending legislation had come into force. There was, in other words, no retrospective force. This chapter also focused on what could be done in respect to the expansion of liability or, more generally of systemic risks. Obviously, one could first have analysed whether there are mechanisms within tort law to “keep the floodgates shut”.397 But we have addressed the problem from the angle of the insurer. In the previous chapter, we already indicated that adequate risk differentiation may provide a good remedy to the problem of insurability. Obviously, problems may


135

4.


still arise if the amount of the damage largely exceeds the capacity of an individual insurer. But we argued that reinsurance, pooling or co-insurance of risks, may provide at least part of the answer. Clearly, the insurer will, especially in cases of catastrophic systemic risks, always put an upper limit on coverage. Economic analysis provides specific hints to legislators or judges for dealing with these issues: first, be careful with the introduction of retrospective liability; second, do not shift the risk of causal uncertainty to one enterprise and third, be careful with the introduction of joint and several liability or channelling of liability. As far as advice to individual insurers on how to cope with expanding systemic (liability) risks is concerned, the need for adequate risk differentiation, stressed in the previous chapter, must be re-emphasised. In addition, we suggest that various policy conditions could be used to limit the risk of unexpected changes in the law. Moreover, a claims-made policy may provide an adequate answer for long-tail risks with unknown consequences, as is typically the case with systemic risks. As far as advice to the legislator is concerned, it seems important that legislators realise that these policy conditions may be necessary to ensure that a risk remains insurable. Some politicians tend to believe insurers should provide compensation for any loss, including those resulting from systemic risks, no matter what the reason is. The desire to find a solvent debtor may lead a judge or legislator, for instance, to prohibit policy conditions excluding specific opponents of the damage or declaring, for example, a claims-made policy void. These reactions would inevitably have the perverse result that insurers would have to withdraw from specific markets all together. Hence, it seems to be more appropriate to focus on specific conditions that have to be fulfilled ex ante in order to benefit from insurance cover, and to respect them, instead of aiming for short-term ex post success which would protect some consumers or victims today, but may result in the total lack of availability of insurance for some systemic risks in the long term.

Notes 352. Official Journal, L 210/29 of 7 August 1985. 353. Bourgoigny argued that the Directive brought about drastic changes for Greece, Italy, Spain and Portugal. Other countries already had – in his view – a product liability system which equalled the liability system of the EC Directive (T. Bourgoigny, Responsabilité du fait des produits: arguments connus pour un nouveau débat, 1987, Revue Européenne de Droit de la Consommation, 17). 354. See Article 21 of the product liability Directive which states “Every five years the Commission shall present a report to the Council on the application of this Directive, and, if necessary, shall submit appropriate proposals to it”.

136


4.


355. See Green paper on liability for defective products, p. 2. 356. Green paper on liability for defective products, p. 17. 357. S. Narita, Product Liability Claims www.swissre.co.uk.e/publications

in

Europe,

1996,

available

through

358. S. Shavell (supra note 12) p. 211. 359. Priest claimed that the liability insurance crisis in the US has to a large extent been caused by a shift to third-party liability insurance, combined with too little risk differentiation by insurers (G. Priest (supra note 17) pp. 1521-1590). The need for effective risk differentiation has been discussed above in this study. 360. H. Kunreuther/R. Hogarth/J. Meszaros (supra note 259) pp. 71-87. 361. See also R. Zeckhauser, 19th Annual Lecture of the Geneva Association and Catastrophes, 1996, GPRI, 5, who equally argues that retrospective liability may affect the predictability of risks; as well as K. Abraham (supra note 346) pp. 957-959. 362. For Belgium see e.g. J. Rogge (supra note 257) p. 6. 363. See Chapter 2, A, 3.2.3. 364. Also K. Abraham (supra note 346) pp. 959-960 and M. Katzman, Pollution liability insurance and catastrophic environmental risk, 1988, Journal of Risk and Insurance (JRI), 89-90. See also L. Bergkamp, The Proper Scope of Joint and Several Liability, 2000, TMA, 154-155. 365. An argument along these lines is put forward by T. Tietenberg. 366. See also L. Bergkamp (supra note 364) pp. 153-154 who argues that joint and several liability may be unfair and may lead to overdeterrence. 367. For a detailed analysis of joint and several liability when all defendants are fully solvent see L. Kornhauser/R. Revesz, Sharing Damages among multiple Tortfeasors, 1989,YLJ, 831-884, and for an analysis of the case of limited solvency see L. Kornhauser/R. Revesz, Apportioning damages among potentially insolvent actors, 1990, JLS, 617-651. 368. See H. Cousy (supra note 266) p. 235. 369. A. Monti (supra note 256). And see L. Bergkamp (supra note 364) p. 154 who argues that under joint and several liability, civil liability becomes unpredictable and hence uninsurable. 370. For a critical economic analysis of the channeling of nuclear liability, see T. Vanden Borre, Transplantatie van “kanalisatie van aansprakelijkheid” van het kernenergierecht naar het milieu (aansprakelijkheids-)recht: een goede of een gebrekkige zaak?, in: M. Faure/K. Deketelaere (eds.), Ius Commune en Milieurecht, Actualia in het Milieurecht in België en Nederland, 1997, pp. 329-382. 371. See M. Trebilcock/R. De Winter (supra note 291), 232-235. 372. See H. Cousy, Een nieuwe vorm van schuldloze aansprakelijkheid voor schade veroorzaakt door het vreedzaam gebruik van kernenergie, 1974-1975, Jura Falconis, 46. 373. See A. Popp, Liability and compensation for pollution damage caused by ships revisited – report on an international conference, 1985, Loyds Maritime and Commercial Law Quarterly, 120 (bibliography 118-131). 374. So T. Vanden Borre (supra note 370) pp. 366-367.


137

4.


375. For further arguments see M. Faure/R. Van den Bergh, Objectieve aansprakelijkheid, verplichte verzekering en veiligheidsregulering, 1989, pp. 357358. 376. On the importance of this issue for environmental liability insurance, see K. Abraham (supra note 346) pp. 970-972. 377. See also H. Cousy (supra note 266) pp. 237-239. 378. So J.H. Wansink, De algemene aansprakelijkheidsverzekering, 1994, p. 110; J.H. Wansink, De aansprakelijkheids-verzekering en de dekking van “long-tail risico’s”, 1995, A&V, 1-5 and J. Spier, De uitdijende reikwijdte van het aansprakelijkheidsrecht, Preadvies Nederlandse Juristenvereniging, Chapter 2. 379. In this connection, see inter alia the discussion of the Dutch DES-case above. 380. See F. Van Huizen, Enkele begrenzingen van de (beroeps)aansprakelijkheidsverzekering, in: Drie treden (De Ruiter-bundel), 1995, pp. 325-339. 381. See a discussion by J.H. Wansink (supra note 378) pp. 110-112, and also O.A. Haazen/J. Spier, in: De uitdijende reikwijdte van de aansprakelijkheid uit onrechtmatige daad, pp. 56-79. 382. K.S. Abraham, Cost Internalization, Insurance, and Toxic Tort Compensation Funds, 1982, 2 Virgina Journal of Natural Resources, 123-131. 383. B.J. Richardson, Financial Institutions for Sustainability, 2000, Environmental Liabilty, 52-64. 384. Insurance cover is awarded immediately, even if the damage had its origin long before the insurance policy entered into force. 385. For a detailed discussion, see especially J. Spier/O.A. Haazen, Aansprakelijkheidsverzekeringen op claims’ made grondslag, 1996, p. 125 and J.H. Wansink (supra note 378) p. 111; J.H. Wansink (supra note 378) pp. 1-5; I.W. Verloren van Themaat, Claims made dekking in Europa, Beursbengel, 11 April 1994. See L. Dommering-Van Rongen, Risico’s met een lange staart, in: Miscellanea Iurisconsulto Vero Dedicata, Essays offered to Prof. Mr. J.M. van Dunné, 1997, pp. 35-37. 386. S. Hankey, Claims-made Policies and Choice of Law in the European Union, 1994, IILR, 267 and A.M. Alvarez-Baron, Spain’s Claims-Made Crisis, 1994, IILR, 316-319. 387. See J.H. Wansink, De Belgische wet op de verzekeringsovereenkomst, 1995, Verzekeringsrechtelijke Berichten, 38 and H.A. Cousy, ’Panta Rei’: een kort bericht over de wijzigingen van de gewijzigde wetgeving inzake verzekeringen, 1995, Tijdschrift voor Belgisch Handelsrecht (TBH), 470-471 and C. Van Schoubroek/G. Schoorens, De aansprakelijkheidsverzekering: a never ending story, 1995, TBH, 644-660. 388. On the scope of coverage of pollution insurance in Belgium see J. Rogge (supra note 257) pp. 18-24. 389. Katzman argues that claims-made policies are less risky for insurers than occurence-based ones, since there is no commitment inthe indefinite future (M. Katzman (supra note 364) p. 88). 390. See, in this connection, O.A. Haazen/J. Spier, Amerikaanse toestanden en de nieuwe aansprakelijkheidsverzekering voor bedrijven en beroepen, 1996, NJB, 45-50. 391. See J.H. Wansink (supra note 378) pp. 116-118.

138


4.


392. M. Deketelaere (supra note 50) pp. 706-709. 393. See Wansink, J.H., “Het DES-arrest in het perspectief van verzekerbare slachtofferbescherming”, Aansprakelijkheid en Verzekering, 1993, 7-12. 394. In addition Hogarth/Kunreuther showed that the insured’s willingness to pay may be less than the required risk premium, since individuals may underestimate these low probability-events (Hogarth, R. and Kunreuther, H., “Ambiguity and Insurance Decisions”, American Economic Review, vol. 75, 1985, 386-390) and Katzman, M., (supra note 364), 86. 395. This and other suggestions have been formulated by Hartlief, T. and Spier, J., “Verzekering en aansprakelijkheid met “terugwerkende kracht”, Aansprakelijkheid en Verzekering, 1994, 28-29. 396. See Dommering-Van Rongen, L., “Risico’s met een lange staart”, in Miscellanea, Jurisconsulto vero Dedicata, opstellen aangeboden aan J.M. van Dunné, Deventer, Kluwer, 1997, 28-29. 397. See Spier, J. (ed.), The limits of liability, Keeping the floodgates shut, Deventer, Kluwer, 1996.


139


Chapter 5

Alternative Compensation Mechanisms and New Financing Techniques


141

5.

ALTERNATIVE COMPENSATION MECHANISMS AND NEW FINANCING TECHNIQUES

A. Introduction In this study, we have so far mainly discussed the applicability of traditional insurance schemes as an instrument to compensate damage caused by systemic risks. Moreover, we focused principally on liability insurance. In this section, we will show that the general principles of insurance, which were discussed above, are also relevant even when traditional insurance undertakings are unable to provide cover in the form of liability insurance, and other compensation mechanisms will be outlined. For instance, the general idea that, no matter what type of compensation system is put in place, it should always be organised in such a manner that the party creating the risk should also bear the financial consequences of it insofar as possible, is a principle which is not only crucial for liability insurance. This idea of risk differentiation and the fact that a liability system should provide adequate incentives for prevention remains valid no matter what type of compensation mechanism one seeks. Indeed, the potential damage of systemic risks may be so large, as we have indicated above, and the scale effect so spectacular (viz. 11th September 2001), that the capacity of traditional insurance undertakings may well be too limited to provide adequate coverage. We will now analyse other financial arrangements than traditional liability insurance. We will have a brief look at some of these alternative mechanisms, not only from a theoretical point of view, but also by looking at some compensation mechanisms which are used in practice. Indeed, for many risks, such as hightechnology risks, nuclear risks and other typical systemic risks, industrial operators have moved away from traditional liability insurance and have searched alternative financial mechanisms to cope with the risks to which they are exposed. We shall first of all briefly discuss the tendency in the literature to advocate the use of capital markets for covering systemic risks (B); then we shall review the notion of self-insurance, examine it critically and look at the benefits and disadvantages (C). There is a major trend in the insurance world away from liability insurance toward first-party or direct insurance. This is true for many systemic risks, more particularly in the areas of medical malpractice, occupational health and even environmental risks (D). Moreover, we shall examine whether risk-sharing agreements (pooling) between operators could provide compensation at lower costs than traditional insurance (E). Finally, we shall look at the possibility of introducing ex ante guarantees and deposits as instruments of financial coverage (F).

142


5.


It is obviously not possible to discuss all of these possible alternatives in detail within the scope of this study. We shall therefore not develop all of them alternatives at great length, but shall focus on some of the mechanisms which can actually be found in practice, the advantage being that there is some empirical information on the way they work. Therefore we will only briefly address the theoretical advantages of using capital markets to cope with systemic risks, but far more attention will be given to first-party or direct insurance schemes, of which there seems to be more (positive) experience. Obviously some of these alternative compensation mechanisms and new financing techniques are akin to traditional insurance (such as first-party of direct insurance schemes). They can therefore be considered as new ways in which insurers and industrial operators have reacted to newly emerging systemic risks. Other alternative financing techniques such as the use of capital markets or risk-sharing agreements between operators should be seen more as mechanisms which have been worked out by the industrial world itself to deal with the unavailability of insurance coverage.

B. The use of capital markets to cover risks: a brief introduction It is obviously not possible within the scope of this study to discuss every potential alternative way of providing compensation for damage caused by systemic risks. In this respect, we can refer to the literature on other alternatives.398 Some economists, more particularly Tyran and Zweifel, have advocated the use of capital markets to provide coverage for environmental damage.399 They begin with the statement that third-party liability as it exists fails to internalise the externality caused by the nuclear risk. They also criticise the argument of the uninsurability of nuclear risks, noting that earthquakes are also not predictable even for large amounts.400 They also point to the disadvantages of the pooling system: cartels charge excessive premiums. In addition they criticise the limitation of liability to a value that falls far short of a possible loss, since this constitutes a subsidisation of nuclear power. As an alternative, they advocate the use of capital markets to provide additional coverage for liability. Rather than turning to nuclear insurance pools, Tyron and Zweifel suggest that plant operators should be permitted to take risk-participation shares on capital markets. Investors would have the opportunity of buying such a share against deposit of financial assets, creating a warrant in favour of the plant of their choice. This warrant can be exercised by the agency when liability claims are presented. Since operators must have sufficient coverage, they will compete for coverage capital and bad risks will not be able to get coverage. Hence Tyran and Zweifel argue that their ERICAM system can correct market failure and provide for higher coverage of nuclear risks and improved safety. The idea behind ERICAM is that the interest rate on the bonds issued reflects the accident rate. In so far as the care taken by operators is observable,


143

5.


it will be reflected in interest rates. But when accidents are a rare experience, rating will be impossible and government regulation will still be necessary. The ERICAM system401 certainly has advantages compared to the current system of limited liability and non-competing insurance pools. This idea of using capital markets to finance potential liabilities from environmental harm is, moreover, not merely theoretical. Several experiments are taking place with forms of so-called Alternative Risk Transfer (ATR) or securitization.402 Although it is apparently possible today to buy “catastrophe bonds”, inter alia at the New York Stock Exchange,403 we will not address these alternative mechanisms in much detail since according to insurers they do not play a major role in Europe yet.404 Moreover, it is argued that these “new” financing instruments, based on capital markets, will never totally replace traditional products.405 Since it is not possible to discuss all of these possible alternatives in detail within the scope of this paper, we have preferred to focus on some alternatives which can actually be found in practice, so that some empirical information on their working can be provided as well.

C. Self-insurance and captives 1. Reserves or insurance? One can question whether self-insurance should be discussed at all within the scope of a chapter dealing with alternatives to liability insurance, since it may be asked whether it can honestly be considered as a realistic alternative. The reason we nevertheless want to mention self-insurance is that, although it cannot be considered as “insurance”, it is certainly a form of financial security. The reason we are rather sceptical concerning this concept of “self-insurance” is that it is a nice word denoting a situation where potentially responsible parties make reserves themselves for potential losses. However, reserves made by the potential injurer himself cannot be considered “insurance” in the traditional sense, for the simple reason that there is no risk-spreading, risk distribution and consequently no loss spreading after an accident happens. Indeed, with selfinsurance the risk will not be transferred to an insurer, a mutual, a bank or a pool, which is typically the feature of most insurance schemes. To be blunt: selfinsurance is not an insurance scheme but a system whereby potentially responsible parties make reserves for future losses. Reserves can, however, be considered as tools for guaranteeing financial assurance; this is, however, different from insurance since reserves lack the element of risk-spreading.406 Although we argue that self-insurance does not belong to the category of financial securities which can be considered as “insurance” allowing a potential responsible party to make reserves for future losses, it is obviously highly desirable within the central concern of this study, namely to analyse

144


5.


how adequate compensation can be provided for damage caused by systemic risks. Within this focus on compensation alternatives other than insurance have to be considered as well.

2. Captives One could therefore question why there is often discussion concerning self-insurance if it constitutes nothing else than private reserves by the potentially responsible party for future losses. The reason is usually a fiscal one. If these reserves were made without any specific goal, they could be considered as hidden profit by tax authorities and could thus be taxed. If on the other hand the tax system allows them (and could even encourage them by making them deductible), self-insurance becomes a way in which the potentially responsible parties could make reserves for future losses in a taxfriendly way. Often these reserves are referred to as “captives”.407 These captives, to which there is often reference in the literature, are in some cases again nothing else than reserves made by larger industries to cover future losses. A captive is, according to Ranson, a reinsurer which would be owned by the insured.408 In some case, but not necessarily, various industries posing the same risk could join forces via mutuals. The self-insurance would then amount to pooling. In other cases the “captives” are used for reinsurance. The company would, for example, take out first-liability insurance with an insurance company, which we refer to as a “fronting company”, which may reassure with a captive. These captives can be managed by banks or (re)insurance companies.409

3. Reserves as guarantees? A positive aspect of these tax devices allowing reserves to be set aside for future losses is obviously that there may be at least a minimum guarantee that these amounts will be used to cover future environmental damage. However, from a policy perspective, the authorities responsible for verifying that minimum financial security exists for example, will usually be rather cautious about accepting self-insurance as proof of such a guarantee. The fact that these reserves are made today does not necessarily mean that they will still be available when the loss occurs. This is obviously the case if it is only the potentially responsible party that has the right to decide the ultimate goal of reserves made in the form of self-insurance. Moreover, even if the reserve is still available when environmental damage, for example, happens (and the potentially responsible party has thus not decided to take the money with him to the Bahamas), the question will arise as to whether the reserve can effectively be used to cover the environmental damage. In the absence of specific statutory provisions to protect the reserve as a specific security, it will simply be considered as one of the assets of the company and the creditors may try to get hold it. Therefore, if selfinsurance in the form of reserves and captives is to be allowed as financial


145

5.


security for environmental damage, additional statutory measures are necessary to regulate the goal of the money reserved by the potentially responsible party. Self-insurance may therefore certainly play a role as a guarantee of financial security, provided that the conditions set out above are met. This means that there needs to be a guarantee (via regulation) that the amounts set aside for this goal cannot be reached by other creditors.

4. Self-insurance as a means of controlling moral hazard However, in many cases self-insurance obviously already exists today, even without this regulatory protection. In practice, many potentially responsible parties decide to self-insure for a large amount and only purchase “excess” insurance when liability would exceed a specific ceiling. Therefore, in practice a combination of self-insurance and liability insurance, whereby self-insurance can take the form of a deductible, can be found. In this respect, we refer once more to the discussion of the moral hazard problem above, where it was argued that exposing the insured partially to risk can be considered a means of controlling control moral hazard. Self-insurance, as described above, obviously has the advantage that it is probably a lot less costly than systems of risk distribution, especially if risks are shifted to an insurance undertaking. As we have explained above, the amount paid (via a premium or a contribution) by the potentially responsible party might often be much higher than the actuarially-fair value of the risk. Thus one can understand why potentially responsible parties might want to self-insure and only purchase excess insurance for high amounts.

5. Limits and possibilities of self-insurance However, there are certain risks and disadvantages with self-insurance as well. One obvious point is that victims still should have the guarantee that a potentially responsible party who is found liable will be able to pay the compensation. Self-insurance is not necessarily a waterproof guarantee against insolvency. This would only be the case if regulation could guarantee that the money set aside for covering environmental damage would only be used for that specific goal. In addition, risk-spreading via systems such as (liability) insurance or pooling has the major advantage that it makes possible riskspreading on the basis of economies of scale. A pool or insurer may bring together similar but unrelated risks and can thus increase the expected utility of all the insured by reducing risk aversion. The major benefit of insurance or pooling – risk-spreading, is obviously lost with self-insurance. Moreover, insurance or pooling systems may have the advantage that specialised insurance undertakings (or brokers) may acquire accurate information on the risk and could thus, via the insurance policy provisions, require specific preventive measures from the potentially responsible parties. In Chapter 3 we explained that efficient insurance policies may lead to a reduction of the dangers caused by systemic

146


5.


risks if an adequate risk differentiation is performed. All these advantages are lost with self-insurance. Finally, self-insurance of a responsible party can potentially lead to redistribution problems. Assume that a potentially responsible party does not purchase liability insurance (or any other form of financial security). In that case he would simply run the risk of having to pay major amounts as a result of his liabilities, and if they materialised he would pass on the costs to the taxpayer because he was insolvent. Externalisation of the risk is precisely the problem, which will occur as a result of insolvency. Therefore we will argue below that this constitutes one of the major arguments in favour of compulsory insurance. In sum: self-insurance may be a low-cost useful instrument which makes it possible to set aside assets (make reserves) to cover future losses. However, in order to avoid the risk of externalisation of harm (as a result of insolvency), self-insurance can only be considered effective financial security if guarantees can be provided through regulation that the reserves set aside will actually be used for the potential losses for which they were earmarked. Moreover, the basic problem remains that potentially responsible parties for systemic risks may cause losses that largely exceed even the assets, which they might have set aside in a reserve (under so-called self-insurance). Hence, additional mechanisms will still have to be examined to provide adequate coverage for damage caused through systemic risks.

D. First-party and direct insurance schemes 1. Introduction In Chapter 2 we explained that the scope of liability for systemic risks seems to be expanding. We equally explained that this has led to great concern for liability insurance. More particularly, elements such as the shifting of the risk of causal uncertainty and the retrospective effects of new liability rules were considered to be dangerous for liability insurers. An interesting effect of all these changes in liability law is that both insurers and industrial operators have been looking for alternatives to liability insurance to cover systemic risks. Several alternatives have been developed, from the use of capital markets (discussed above) to the creation of compensation funds (to be discussed below).410 There is one particular alternative, which merits a closer look. It concerns more particularly the trend in some countries to move away from third-party liability insurance towards first-party insurance. It is precisely this trend that we would like to examine more closely in this paper. The reason is that this shift from third-party towards first-party insurance seems at first blush to correspond nicely with George Priest’s suggestion that this would be an appropriate remedy for the American insurance crisis.411 Some insurers seem to have taken Priest’s warnings of a liability crisis seriously and the same is


147

5.


apparently true for the remedies he proposed. One can, especially in the environmental sphere, notice an increasing trend towards first-party insurance. This is, by the way, not only the case in the environmental area but also for many other systemic risks, for instance the areas of occupational health and medical malpractice. An increasing use of first-party insurance is even also being discussed in the traffic accident area. Therefore, this phenomenon definitely merits a closer look. In this section we will first of all look at the theoretical differences between first-party and third-party insurance. The question arises how, at least in theory, the two differ as far as their ability to prevent systemic risks is concerned as well as the possibility to provide adequate compensation to accident victims. Then we will take a practical example of one systemic risk – environmental harm. The reason is that in the Netherlands, Dutch insurers recently decided to more or less abolish environmental liability insurance and to make a radical shift to environmental damage insurance. We will have a critical look at this change and examine whether this is an example which should also be followed for other systemic risks. The question obviously arises why certain risks would be insurable in a first-party insurance scheme whereas they would not be insurable under liability insurance.

2. First-party versus third-party insurance: theoretical differences 2.1. Introduction In the economics of accident law and insurance, the reason why persons seek insurance coverage have been explained. The utilitarian approach with respect to insurance has demonstrated that risk creates a disutility for people with risk aversion. Their utility can be increased by loss-spreading or if the small probability of a large loss is taken away from the injurer in exchange for the certainty of a small loss.412 The latter is of course exactly what constitutes insurance. The risk-averse injurer has a demand for insurance; he prefers the certainty of a small loss (the payment of the insurance premium) whereby the probability of a larger loss is shifted to the insurance company, thereby increasing the utility of the injurer.413 It is remarkable that in this utilitarian approach to insurance, liability insurance is in the first place regarded as a means of increasing the utility of a risk-averse injurer, not so much as a means to protect victims, as is sometimes argued by lawyers. The reason an insurance company can take over the risk of the injurer is well known: because of the large number of participants, the risk can be spread over a larger group of people. The insurer only has to take care to constitute relatively small risk groups in which the premium is as much as possible aligned on the risk of the members of that group. In addition to this utility-based theory of insurance which sees insurance as an instrument to increase the expected utility of risk-averse persons through

148


5.


a system of risk-spreading, Skogh has powerfully argued that insurance may also be used as a device to reduce transactions costs.414 Thus one can easily apply these general principles to explain why there will be a demand for environmental liability insurance: it can provide protection for risk-averse injurers. However, it has been argued that several conditions have to be met to keep liability for systemic risk insurable. The question we are interested in is obviously whether the conditions of insurability might be easier to meet with first-party insurance than with third-party insurance.

2.2. Advantages of a first-party insurance Liability insurance is third-party insurance, whereby the insurer covers the risk that the insured (the potentially responsible party) would have to provide compensation to a third party. First-party insurance is a system whereby the insurance coverage is provided and compensation is awarded directly by the insurer to the victim. Whether such first-party insurance can be considered as an efficient alternative for third-party liability cannot be answered in general terms. It depends to a large extent on the details of such a system and more particularly on the question whether the first-party insurance is combined or not with the liability of the potentially responsible party. The underlying principle in first-party insurance is that the insurance undertaking – in principle – pays as soon as damage occurs, provided that it can be proven that the particular damage has been caused by the insured risk. Payment by the insurance undertaking occurs irrespective of the fact whether there is liability. The arguments advanced in the literature in favour of firstparty insurance are that the transaction costs would be lower and that risk differentiation might be a lot easier.415 The reason is simply that with firstparty insurance the insurer covers directly the risk of damage to a particular victim or a particular site. The idea is that it is therefore much easier for the insured to signal particular circumstances which may influence the risk to the insurer. The problem with liability insurance is that the insurer is always insuring the risk that the insured (the potential injurer) will harm a victim (a third party) of which the properties are unknown ex ante to the insurer. Moreover, under liability insurance there are many uncertainties, e.g. how the judge will interpret this specific liability of the insured. In the ideal world of first-party insurance, the insurer directly covers the victim, and thus the risk. He can therefore monitor directly the risk and in principle provide much better risk differentiation.416 If one took the importance of risk differentiation and the risks of adverse selection as seriously as Priest did, the shift towards first-party insurance would seem to be most promising. One can – at least at a theoretical level – understand why first-party insurance would be beneficial to insurers: it is


149

5.


obviously much easier to control and assess ex ante the risk that a particular victim would suffer damage instead of assessing the risk that the insured potential injurer would cause harm to a third party and thus be found liable. There are many more uncertainties in third party liability which could make adequate risk differentiation more complex. Obviously, in liability insurance risk differentiation is possible as well, in the sense that, as we have explained above, the insurer can adequately monitor whether his insured is a good or a bad risk (e.g. in the sense of taking preventive measures), which could then be rewarded with a lower or higher premium. But even if ideal risk differentiation were applied, there are still many uncertainties in liability insurance. For instance, it will depend on the state of case law and the interpretation by judges whether the insured potential injurer will be found liable for a specific behaviour and a specific damage; it is apparently this legal uncertainty which insurers dislike in particular. Moreover, with third-party liability insurance it is ex ante not possible to know whether the insured injurer will cause damage to a very high-income or a very low-income victim. That type of uncertainty can of course be avoided in first-party insurance where – at least in the ideal situation – the victim chooses ex ante the insurance coverage he wishes according to his own demand and expected losses. Notwithstanding these theoretical advantages, several questions still remain. More particularly the question arises of how this scheme could be used to enhance the insurability of an environmental risk, which is the central topic of this paper.

2.3. Direct insurance for environmental risks: a few questions Let us now see how such first-party insurance could be used as an alternative to liability insurance to cover environmental risks. Several practical questions will arise. 2.3.1. First-party versus direct insurance As we mentioned above, first-party insurance is based on the principle that the victim seeks coverage directly from the insurer. Such a system can of course hardly be applied to the full extent in the environmental sphere. Theoretically, in pure first-party insurance it is the victim who takes out insurance coverage and who therefore also pays the premium. Such a pure first-party insurance scheme would probably not be very practical for environmental risks unless one imagines industrial operators seeking coverage e.g. for the pollution that may occur in their own plant. In that case, the operator would, in first-party insurance, seek coverage for the damage he may suffer himself. Otherwise first-party insurance would mean that private citizens who fear to suffer environmental harm, e.g. in their garden, seek insurance coverage themselves.

150


5.


However, there is an alternative which looks like first-party insurance, and which is called direct insurance. In a direct insurance scheme, one would imagine the potential polluter, e.g. the party which possesses a particular site, seeking insurance coverage that provides protection also to third parties who could suffer damage resulting from that particular site. Such a direct insurance scheme is applied in some countries with respect to occupational health. The employer takes insurance coverage on behalf of his employees who could claim directly on the policy. Under direct insurance, the policyholder is not the victim, but the potential injurer. It is not, however, liability insurance since it is no longer liability but the mere existence of damage that triggers the compensation. For environmental risks, one could thus examine both pure first-party insurance, whereby potential victims cover losses themselves, and direct insurance, whereby a potential polluter seeks insurance coverage which is also to the benefit of a third party (the victim). As we will see below, in practice there may be a combination of first-party and direct insurance; the owner of a particular site from which soil pollution may occur, for example, would seek coverage both for his own damage and for the damage which may be caused to third parties. 2.3.2. Scope of coverage and causation The main difference between traditional liability insurance and either pure first-party or direct insurance is that the coverage under the latter policies is no longer for liability. Hence, a consequence (and insurers claim an advantage) is that liability is no longer required to be able to claim on the policy. Traditional insurance would say that one would need an “accident”. However, it is well known that, given the gradual nature of many pollution cases that they can hardly be considered as sudden events like accidents. Therefore, in environmental insurance, damage will probably suffice, but it will be important to describe clearly what constitutes the insured damage which may gave a rise to a claim on the policy. Depending on how this is formulated in the policy, the insured will have to claim that the damage is caused by the insured activity. In the hypothesis of direct insurance, which benefits third-party victims, problems could arise in case of multiple causes. In that case, the victim will obviously have to claim and prove that the damage was caused by the particular insured risk. However, with environmental damage insurance causation does not seem to be a major problem (other than that it may always be a problem in any environmental liability case as well). This is different in cases where firstparty insurance solutions are presented for health damage, such as in the case of medical malpractice or for occupational diseases. In all of these cases, the


151

5.


difficulty will be that the question will have to be settled whether the personal injury suffered by the victim is actually caused by medical malpractice or by an occupational disease. These types of causation issues, relating to the fact that it should be clear that the harm is caused by the insured risk, may arise also in environmental insurance, but seem less serious than in cases where personal injury is insured. 2.3.3. Financing Although less interesting from a theoretical perspective, a practical question that will always arise at the policy level is who should pay for a firstparty or direct insurance scheme. A consequence of a pure first-party system is that it theoretically is the victim who seeks insurance coverage himself and who therefore finances the first-party insurance. To the extent that this victim is also the industrial operator who caused the risk, polluter and victim will be the same and it will effectively be the polluter who finances the insurance. If one were, however, to adopt a first-party insurance scheme for “innocent” victims, politicians would probably argue that victims would be forced to finance e.g. the clean-up of the polluted soil in their own garden (by paying the premium for the first-party insurance) whereas the polluter should pay for it. The answer would obviously be that if the first-party insurer of the “innocent” victim were allowed a right of redress against the polluter, the polluter-pays principle would still be satisfied and the polluter would still be given appropriate incentives for prevention. However, it is most likely that at the policy level, discussions would lean towards direct insurance since under such a system the polluters finance the insurance coverage, and also the coverage for damage caused to third-party victims. 2.3.4. Prevention A question which will obviously be asked is: how can polluters be given appropriate incentives to prevent environmental harm if the risk is fully covered under insurance. The answer here is obviously that this poses no difference with traditional liability insurance. All insurance systems potentially lead to the wellknown moral hazard problem.417 But we equally indicated that adequate risk differentiation can be the appropriate remedy to moral hazard. First-party and direct insurance schemes were advocated precisely because they would enable a better risk differentiation. If that were the case, they could even lead to better results as far as prevention is concerned. In practice, the risk differentiation under first-party and direct insurance would mean that if a particular site were insured, the insurer would use all the ex ante monitoring and ex post devices to check the “ecological reliability” of the

152


5.


particular operator, which should provide optimal incentives for prevention. Hence, if the theoretical possibilities of risk differentiation are used in an optimal manner, first-party or direct insurance schemes should cause no problem as far as prevention is concerned. 2.3.5. Relationship with liability law An interesting issue, since liability is no longer the trigger for coverage, is of course the relationship to liability law. In many instances where first party of direct insurance schemes are advocated (such as in the areas of medical malpractice and occupational health), they are often supposed to totally replace the liability system. In the area of occupational health in Europe, for example, direct insurance schemes (providing coverage to all insured employees of one employer) with an immunity for liability of the employer except in cases of intent or gross negligence. In many countries these insurance systems are embedded in social security law. A traditional argument, e.g. in the area of medical malpractice, against first-party (patient) insurance schemes and in favour of tort law is that the incentives of the health care provider would be diluted without a liability system.418 At the theoretical level, there would be no reason to grant immunity to the polluter even if a mandatory first-party or direct environmental insurance scheme were introduced. Even if a third-party victim were compensated by the insurer, it would still be possible to use liability law (via the right of recourse) to provide the polluter with appropriate incentives for prevention. Therefore, there seems to be no reason to let first-party insurance replace liability law. Notwithstanding this theoretical starting point, it may be often noticed that in practice first-party or direct insurance schemes are only accepted (that is to say, their mandatory introduction) if, e.g. in the area of occupational health, the potential injurer (the employer) receives some kind of “compensation” in the form of a partial immunity of liability. 2.3.6. Guarantee of coverage Finally, one could ask to what extent a first-party or direct insurance scheme can provide a guarantee that funds will indeed be available at the crucial moment when they are needed, e.g. when polluted soil needs to be cleaned up. This question has important policy implications and has several aspects. First of all, it relates to the question whether such a first-party or direct insurance scheme should be introduced compulsorily. Indeed, the market could devise wonderful insurance devices but if no operator makes use of them there would still be no guarantee that, given the insolvency risk, funds would actually be available when they were needed. However, this is obviously not a particular problem of first-party insurance but a more general problem


153

5.


relating to the question of whether at the policy level a duty to seek financial coverage for environmental risk should be introduced. One issue, which is related to direct insurance, is whether one can provide a direct right of action for a third-party victim. This concerns those situations where the insurance policy of the industrial operator provides that his coverage also extends to damage caused to third parties. Normally these third parties do not, given the provisions of the contract, have a claim on the insurance policy unless they are formally made third-party beneficiary, as provided by direct insurance. The question hence arises as to whether one sees such third-party beneficiary clauses in practice. After having identified the potential benefits and several points relating to first-party and direct environmental insurance, let us now look at how these new policies are introduced in practice.

3. Environmental damage insurance in practice: the Dutch example 3.1. A trend towards first-party coverage At the end of the last section, it was argued that although first-party insurance might seem very attractive at first sight at the theoretical level, since it might facilitate a narrowing of risk pools, many questions still remain, especially as regards its application to environmental risks. It is probably best to examine these questions by looking at a practical example where first-party insurance of polluted sites has been implemented. Before doing so, it is probably important to stress that this presentation of first-party insurance is not merely theoretical but has a certain practical relevance. Indeed, the general liability committee of the Comité Européen des Assurances (CEA) carried out a study on first-party legal obligations for cleanups and corresponding insurance covers in European countries.419 This study shows that although the insurance situation between the European countries still differs to a large extent, first-party insurance coverage seems to be available in several member States.420 In one country, more particularly in the Netherlands, insurers have deliberately chosen to provide coverage of polluted sites on a basis. The idea is that first-party coverage should replace traditional environmental liability insurance. Hence, it seems interesting to take a closer look at the insurance situation in the Netherlands.

3.2. Environmental damage insurance in the Netherlands 3.2.1. Dissatisfaction with existing coverage The concern of the Dutch insurers was prompted by the fact that all the theoretical problems discussed above in Para. 2 concerning the insurability of environmental liability had also played a major role in Dutch environmental practice. This had to do with the fact that environmental risk is an example of a “long-tail risk”, whereby the insurer today could be confronted with events

154


5.


which occurred in a distant past and would lead to liability of the insured now. Insurers held that this generally endangered the predictability of the risk. Most of these problems were hence related to the fact that environmental harm does not constitute a sudden event, as is the case with most “traditional” accidents which are insured in liability insurance. The Dutch insurance market therefore used cover environmental risks through a variety of insurance policies, the most important of which were: ●

liability insurance (AVB)421 for sudden risks and occupational health risks which were related to the environment;

●

environmental liability insurance (MAS) for risks of a more gradual nature, excluding personal injury;422

●

fire insurance for clean-up costs after fires (although the precise coverage of that policy was debated).

Environmental liability insurance was provided by an environmental pool, referred to as MAS.423 In this pool, 50 (re)insurers work together to cover the environmental liability risk. The MAS was hence constructed as an environmental pool, although the individual insurers in it contracted the MAS policy individually under their own label. There was, however, a lot of criticism of the system, which can be summarised as follows424. ●

First of all, the whole division of coverage between the AVB and the MAS was based on the idea that the AVB would cover sudden risks and the MAS would cover risks of a more gradual nature. As is well known, in practice it was not always possible to make a clear distinction between sudden and gradual risks, which led to uncertainties concerning the scope of the coverage of both policies. This was obviously the result of the fact that Dutch insurers did not decide to exclude the environmental risk all together from the traditional liability insurance of companies (AVB).425

●

Environmental liability policy (MAS) was considered to be rather complicated and had a very complicated procedure for accepting policyholders. The policy was in addition rather expensive and hence difficult to sell.

●

A further problem was that neither the general environmental liability policy (AVB) nor the environmental liability insurance policy (MAS) provided any coverage for damage caused to the insured’s own site. This caused a problem for the insured party, since according to the case law of the Dutch Supreme Court, companies could also be held liable for pollution of their own site. In addition the fact that pollution caused to the site of the insured was not covered inevitably caused further uncertainty concerning the scope of coverage. One could imagine cases where polluted groundwater went from the site of the insured to a neighbouring site. This caused problems since the


155

5.


pollution to the neighbour’s site was insured whereas the pollution to the insured party’s own site was not. ●

Furthermore there were more uncertainties concerning the question of whether clean-up costs were covered under the fire insurance policy. Under the clean-up costs after a fire, fire insurance also covered environmental damage that might have occurred, but only on the condition that there was a prior fire. No account was taken of the fact that, after a fire, serious soil and water pollution could also occur. It was not always clear whether the soil clean-up costs resulting from the fire were also covered under the fire insurance.

●

Finally there were obviously all the traditional problems related to environmental liability such as the question whether specific damage was caused as a result of an insured risk. Also Dutch case law concerning environmental liability was extended in a way that had not been foreseen by insurers. Hence, liability risk was increasingly considered unpredictable and hence uninsurable as far as environmental harm was concerned.

3.2.2. Environmental damage insurance: main features This led the Dutch Insurers’ Association to present a new product – environmental damage insurance (MSV).426 This policy has been available since 1 January 1998 and it takes a radically different approach from traditional environmental liability insurance.427 This new environmental damage policy provides for various new elements. First of all, it provides integrated coverage of all the environmental damage which occurs on or from the insured site. The prerequisite is that it must concern pollution of the soil or water. The integrated coverage means that the new environmental damage insurance will replace the traditional pollution insurance (for sudden pollution) provided by the AVB and the liability insurance of the MAS (for gradual pollution). The whole idea is that this coverage constitutes a direct insurance. In other words, the insured site is insured even when clean-up costs are incurred on the site of a third party. Coverage takes effect as soon as the insured site is polluted as the result of the insured risk, irrespective of whether the insured could be held liable for the damage or not. In some cases, the third party (the victim), benefits from a direct action for compensation on the basis of the environmental damage insurance policy. The trigger for compensation under this policy is therefore no longer tort law but the insurance policy concluded between the insured and the insurance company. This policy thus constitutes first-party insurance or, as it is called in the Netherlands, direct insurance. It is direct insurance to the extent that it also benefits third parties. It is not the third party victim who purchases insurance (although the insured may be the victim) but someone who has

156


5.


responsibility for a site on or from which water or soil pollution may occur. The policy benefits third parties as well, at least when this is provided in the policy. Obviously, environmental damage insurance cannot set aside tort law, but the main advantage according to Dutch insurers is that the coverage is not triggered on the basis of liability. The advantage from the victim’s point of view is obviously that coverage can be provided more rapidly and probably at lower transaction costs than through the courts, as has to be done under liability law. The environmental damage insurance provided by the Dutch Insurers’ Association consists of several categories with different policies. The insured can opt for various insurance policies. This obviously shows that first-party insurance more effectively provides optimal risk differentiation since every party will be able to purchase insurance coverage according to his own preferences. The damage on the insured location itself is insured. This at least provides coverage for clean-up costs, which are rather broadly defined. Also costs for the repair of damage are included. Soil pollution, which is also envisaged, obviously falls within environmental damage insurance. Note, however, that under this policy it is only the costs of “clean-up” which are covered.428 Also, the damage caused from the insured site and suffered by third parties is covered. This obviously provides a broader scope of protection. The insured remains in principle fully liable, although the third party (beneficiary) that would be protected under the new MSV policy could claim directly on the policy and would hence in principle not have an interest in using liability law. However, it might be that the insured has taken too limited coverage, in which case the third party would still (have to) use liability law. In that case, the liability itself is not covered (and an insolvency risk remains), but the MSV policy provides for legal aid in a couple of specific cases,429 for example, if the sum which is insured under the MSV coverage would not be sufficient to pay for the clean-up costs incurred by the government. Or if a third party chose liability law instead of direct action under the MSV policy.430 The main feature of the new system of environmental damage insurance in the Netherlands is obviously that it is no longer liability insurance but only first-party (or direct) insurance. The advantage for the insurer (and for the insured) is that the difficulties of an action under liability law are avoided. Whether liability law will still be used is uncertain. Third parties could still use liability law, although it is obviously easier for victims to take direct action under the MSV policy. There is, however, one important weakness, which unavoidably remains, namely that the environmental damage insurance is not compulsory. Hence, there may remain situations where companies in the Netherlands have purchased no insurance coverage at all or situations where only basic coverage was taken for damage on the particular site, but only to a


157

5.


limited extent for damage to third parties. 431 In such cases, third parties claiming against the responsible party might still be confronted with an insolvent polluter. Moreover, the new MSV regime is exclusive, meaning that the coverage for (sudden) soil and/or ground water pollution has now been removed from the liability insurance policy. This means that if an insured party has, for example, only taken out MSV coverage for the insured site, and a loss occurs to a third party on another site, this third party would probably use liability law against the polluter. In that particular case the polluter cannot call on his general liability insurance (AVB) since the environmental risks have now been completely removed from that policy as a result of the coming into force of the MSV.432 But obviously, this can hardly be considered a weakness of the system of first-party insurance: the insured obviously does not get more than what he pays for. Since the MSV is a general policy with numerous options for the insured, premiums and amount of coverage can vary. The type of costs which are insured are, however, identified in the general policy and according to the CEA study on first-party insurance, the total amount of coverage available under the new environmental damage insurance in the Netherlands is 25 million Dutch guilders. 3.2.3. Evaluation 3.2.3.1. Mixed first-party and direct insurance Turning back to the specific issue of first-party/direct insurance, it may be noted that the Dutch scheme is apparently a mix of first-party and direct insurance. It is principally first-party insurance in the sense that the insured site of the potential polluter is covered. It is also direct insurance since the policy not only covers damage on the insured site itself but also the damage suffered by a third party. The advantage for the insured is clearly that this first-party insurance now provides coverage for the damage on the site of the insured as well, which is a major change compared to the past. Very often the authorities forced polluters to clean up their own polluted soil. Formally this was not constructed as liability and therefore no insurance cover was available. As far as the insurance policy also covers the damage suffered by third parties, it effectively functions as a liability cover. The trigger for coverage is no longer liability but damage. There is a clear causation requirement in the policy, since the pollution must be the direct and exclusive result of an emission caused by one of the insured risks. Moreover, only the costs of rehabilitation of the insured site are covered, which excludes the uncertainties involved with assessing environmental damage.

158


5.


3.2.3.2. Advantages As far as prevention via optimal risk differentiation is concerned, insurers obviously claim that for them, this policy has the major advantage that they can adequately control ex ante the quality of a particular soil and the production methods applied by the insured. On that basis, an adequate assessment of the risk can be made. However, some company lawyers have been critical of this new insurance product. They argue that in practice insurers look so closely at risks before providing coverage that in practice only very good risks would be able to receive coverage.433 Hence, they claim that one cannot blindly argue that the environmental damage insurance would be a remedy for all insurability problems concerning environmental risks. Indeed, the flip side of effective ex ante monitoring is obviously that insurers could chose either to cover only good risks or to cover bad risks only for a high premium. But one can hardly reproach the insurers with applying insurance economic principles correctly. Moreover, the environmental damage insurance introduced by the Dutch insurers is a completely voluntary system, introduced without any government interference. A consequence of this is that liability law is not affected by it at all. Victims suffering harm as a result of soil pollution can still sue the polluter. If the insured has accepted the possibility of claims by third parties, the victim can even claim directly on the policy as third-party beneficiary. If, on the other hand, the MSV policy did not contain a third-party beneficiary clause, the victim will, just as is the case today, in the Netherlands, have to sue the liable polluter. A problem in that case is, however, as we indicated above, that the MSV is supposed to replace liability insurance, meaning that a liability insurance will no longer be available, at least for the environmental risks covered under the MSV. According to information provided by the Dutch Insurance Association, this new product is working remarkably well. They claim that the interest shown by enterprises in the new environmental damage insurance is much larger than their interest in traditional environmental liability insurance. Whether the new product is actually a success is more difficult to judge. There is, however, undoubtedly increasing interest on the part of industry in the new environmental damage insurance. The fact that more extensive financial security for environmental damage is provided in the Netherlands as a result of this product should definitely be considered as positive. Moreover from the victim’s (mostly the government) perspective the fact that the environmental damage insurance provides for a direct action for victims should be considered as positive as well. However, one should note that the third-party right of action only applies if the insured has accepted it. Moreover, the advantage for the insured is that damage caused to his own site is also covered, which was obviously not the case under liability insurance.434


159

5.


If, at a policy level, it is concluded that an environmental liability system should be combined with some form of guarantee that financial security will be available, one should at least leave the option open to industry to provide that financial security via environmental damage insurance. The Dutch example shows that first-party type coverage seems able to achieve that end.435 Moreover, it seems that it is being followed in other countries as well.436 3.2.3.3. Disadvantages However, there are inherent limitations in the new environmental damage policy as well. First of all, one should note that it only applies to soil pollution, ground water pollution and the subsequent remediation costs. It is therefore not applicable in cases of other types of environmental harm like noise, surface water pollution or air pollution. But obviously it is also more difficult to apply liability law to those types of environmental damage as well. Another point of criticism is that the amount of coverage that would be available seems to be relatively low. The CEA study reports that the amount of coverage available would be 25 million Dutch guilders. This would be sufficient to cover the average clean-up costs in case of soil pollution but not for larger cases. However, in practice policies are not always awarded for this maximum amount of 25 million. The fundamental question one can also ask at the policy level is whether the introduction of this new environmental damage insurance was really necessary to deal with the uncertainties of environmental liability, as the Dutch insurers claimed. One could argue that the tricky aspects of environmental liability such as causal uncertainty and long-tail risks could be dealt with either by the legislator or through adequate policy provisions. Note that as a reaction to the long-tail risk arising under liability insurance, claims-made coverage was introduced. But this also applies in the MSV. Moreover, risk differentiation may theoretically be easier in first-party insurance, but if liability insurers monitored potential polluters adequately (it is doubtful whether that happened in the past), the risk differentiation in environmental liability could be as adequate as in firstparty insurance. One has the impression that the real reason for Dutch insurers moving to environmental damage insurance was that they wanted to exclude the uncertainties having to do with the involvement of the judges (will there be wrongfulness, liability in this particular case? Will he accept causation or not?). Insurers may now have excluded many of these uncertainties but they will remain for polluters who are still fully exposed to environmental liability. This leads to a particular tricky aspect of the new regime, which really is a source of worry – the fact that the new MSV is supposed to replace general liability insurance, at least for the environmental risks covered under the MSV. As a result of this, industrial operators in the Netherlands today no longer

160


5.


have the possibility to obtain liability coverage for their environmental risks. Third parties who suffer harm may be protected under the new MSV, but only under the assumption that the MSV coverage is sufficient to cover their losses. If this is not the case, the third party will simply have to use tort law and bring a lawsuit against the polluter. But then the problem will arise that the polluter no longer is covered under liability insurance. The victim then may find his defendant to be judgement proof.

4. Summary In this section special attention was given to a new environmental insurance product introduced recently in the Netherlands and gaining popularity in other countries as well. The reason for giving so much attention to this new product, referred to an environmental damage insurance, is that the insurers claim that many of the problems that arise when insuring environmental risks are in fact problems related to environmental liability. In other words: they claim that environmental risks would be more insurable if they could be insured on a firstparty basis. Many of the problems that appear under liability insurance would then disappear, or at least be reduced, so it is held. From a theoretical perspective, the Dutch insurers certainly have a point: already in 1987 George Priest claimed that the American liability and insurance crisis was caused by a shift from first party to third-party insurance. The obvious remedy for him was to move back from third party (liability) insurance to firstparty insurance. However, while the first-party model does seem an appropriate remedy for problems in liability cover, in fact it is valid only for one aspect of the insurability of the environmental risk, namely the issue of risk differentiation. Indeed, first-party insurance policies are held to provide better risk differentiation than traditional liability insurance. Of course this is an important issue since adequate risk differentiation has always been advanced as the appropriate remedy for the dangers of moral hazard and adverse selection. But on the other hand, other problems concerning the insurance of environmental risk will remain – for instance, the issue of limited capacity in the case of catastrophic environmental accidents, the risk of judges shifting causal uncertainty to an industrial operator and the danger of a retrospective application of liability laws. Also under first-party insurance, capacity will not be unlimited; first-party insurance requires a condition of causation (it must be proven that the damage was caused by the insured risk) and the problem of retroactive liability resulting from the “long-tail” character of the environmental risk may be resolved both under third-party and under first-party insurance (at least to some extent) by the introduction of claims-made cover. Dutch insurers introduced their new environmental damage insurance with great enthusiasm, claiming that they would now be able to increase the manageability and predictability of risks. Their basic idea is that it is much


161

5.


easier to monitor and predict the potential damage ex ante for one particular site than to predict the chances that one operator may be held liable to pay damages to a third party. One important factor of uncertainty (the judge) has thus been excluded. Although one can understand the enthusiasm of the insurers and – to some extent – the corporate world (since under first-party insurance they also receive coverage for their own losses) there are clearly disadvantages as well. The major disadvantage is not at all related to the fact that first-party environmental damage insurance would not be an adequate insurance product but to the fact that its introduction was accompanied by the abolition of liability insurance for environmental risks. As a result of this, some victims may not receive compensation any longer, at least in those cases where the damage insurance does not provide sufficient coverage and the polluter is insolvent. Therefore, at the policy level we should examine whether it would not be necessary to introduce a compulsory insurance scheme, be it liability or first-party based. Given the risk of insolvency, we held that a serious problem of underdeterrence could arise which could be solved by introducing some obligation to provide financial security to cover the environmental risk. In addition, the fact that Dutch insurers are collectively moving from liability to environmental damage insurance raises questions concerning competition on the insurance market. Apparently the co-operation between insurers in the Netherlands is so close that, although formally the standard policy only is an “advise”, in practice operators cannot obtain coverage for liability risks (for instance, for soil pollution) any longer. This example shows that restrictions of competition on the insurance market may seriously limit product differentiation and thus endanger the insurability of the environmental risk. It shows that an effective competition policy is of utmost importance to ensure that a wide variety of efficient insurance policies is available on the market for environmental risks. Otherwise, changes such as the one in the Netherlands could limit rather than enhance the insurability of environmental risks.

E. Risk-sharing agreements 1. Risk-sharing by operators: principles Above we discussed the possibility of pooling as a means of increasing insurance capacity and providing higher amounts of coverage. We mentioned the possible negative consequences of pooling in the long run, citing the example of the nuclear pools. If pools act like monopolistic insurers, premiums might be relatively high and unattractive for specific operators. Pooling of insurance resources was also used for environmental insurance, as the example of the Dutch environmental pool showed.

162


5.


An alternative is risk-pooling by plant operators. Faure and Skogh have proposed risk-pooling by operators, a risk-sharing agreement as an alternative compensation scheme which could provide higher amounts of coverage to deal with nuclear risks.437 Their reasoning, which could also be used to increase capacity for environmental risks, is as follows: Although large risks may be uninsurable on the traditional insurance market, they might be shared by ex ante agreements. Such agreements are often used in cases where the risk is uncertain and possibly difficult to estimate, at least initially before information on the probability and possible magnitude of the accident is available. The risk sharing could be realised through an international convention in order to get as many plant owners involved as possible. It could, however, also function within the existing conventions on a voluntary basis. It would function as follows: ●

the owner of a nuclear power plant is strictly liable for losses caused by an accident, including third-party liability.

●

in principle, as was mentioned above, strict liability should be unlimited. It may, however, be necessary to limit initially the liability to e.g. US$ 100 billion, which would already be substantially larger than the size of the liability today. This would also allow compensation for medium-sized accidents.

●

every plant owner should contribute to a mutual guarantee fund. This mutual pool covers liability in the case of an accident. If there are 100 nuclear power plants involved in the pool, and thus 100 members, each plant owner then has to contribute US$ 1 billion in the event of an accident with costs rising to US$ 100 billion.

●

the maximum potential claim on each plant is therefore US$1 billion. This makes reinsurance necessary. A fraction of the desired coverage can be provided by the international insurance market; another part can be obtained through reinsurance from the signatory states.

A major difference with the existing system is that, in the proposed risksharing agreement, all plants share the costs of accidents wherever they occur in the signatory states. This makes nuclear safety a collective economic responsibility for the industry. The liability is to a large part transferred from the single plant owner to the collective guarantee fund mutually owned by the nuclear industry in the contracting states. The mutual fund will face moral hazard problems caused by individual members. It will therefore be in the interest of the fund to control its members and to reduce risks, for instance by imposing membership requirements. Hence, the industry will be forced by the mutual fund to take preventive measures as a condition of membership. Thus an important effect of the risk-sharing agreement would be that accident prevention would gain a new economic dimension. Unlimited liability of plant owners in combination with a mutual guarantee fund would


163

5.


create a collective interest on the part of the industry in reducing risks and controlling pool members. The working of this ex ante agreement can be illustrated by one example. Assume, for instance, that there is an accident in France that causes US$ 60 billion damages in France, Belgium and Germany. Each of the 100 plant owners will then contribute US$600 million. The 12 plants in Sweden would have to cover US$ 7.2 billion of losses. A large part of these costs would be insured by the insurance market (and maybe by the existing nuclear insurance pools) and there would be reinsurance by the states and to some extent by the international insurance market. The possibility of reinsurance by the state is already covered under the Paris Convention. For instance in the Netherlands the Dutch State provides a guarantee of 5 billion Dutch guilders (almost US$3 billion). This shows that although it might be impossible to receive insurance coverage of up to US$ 1 billion for each plant, an ex ante commitment by each power plant to intervene in the costs of a nuclear accident up to a maximum amount of US$ 1 billion might well be given if it is supported by insurance, reinsurance and a state guarantee.

2. Practicability of pools: the nuclear example The statements made above that the existing limitations on liability neither provide for adequate compensation of losses nor for full internalisation of nuclear risks are of course not new. The IAEA and the Nuclear Energy Agency of the OECD also see an increase in the amounts available as a necessary change to the conventions, especially in the light of the Chernobyl accident. We also have reason to believe that it is in the interest of the nuclear industry today not to oppose such increases. With the current limitations property damages will to a large extent be uncovered by insurance and the international conventions. A nuclear accident in Central Europe, for example, may destroy land or property in surrounding nations with serious diplomatic controversies as a consequence. An accident that makes it necessary to abandon the area around a plant may also cause immense losses of land and property. The effects of these losses on the financial system are often overlooked. Since most credits are guaranteed by real estate, many creditors will not recover their loans. The consequence can be the bankruptcy of many financial institutions, which may disturb the international financial system. The impact on public budgets may therefore be large. Since not all the losses will be compensated, the distribution of the available compensation will be a political matter. In general, undercompensation will lead to a general distrust of nuclear power, which will put considerable pressure on politicians to reduce the use of nuclear power to an even greater extent. This can only be avoided if the nuclear industry can come up with an arrangement that provides at least for adequate compensation if an accident happens. The classic system of individual liability of a nuclear power plant operator with individual insurance by the national pools has not proven able to generate sufficient amounts. As we have

164


5.


demonstrated, an ex ante risk-sharing agreement on an international level might provide more adequate victim compensation. In a way, an ex ante agreement is simply a decision on how to distribute the costs of a potential accident ex ante. An accident today could lead to random victimisation, political conflicts and large economic disturbances. An ex ante regulation is therefore in the public interest but also in the interest of the industry. Such a risk-sharing agreement will reduce the public’s disutility of risk aversion and thus the fear of nuclear power in general. The ex ante agreement will also reduce the risk for the nuclear industry of political failures due to hasty opinions, poorly informed politicians etc. An opinion in favour of an ex ante risk-sharing agreement may be put forward. First, countries without nuclear power will only benefit from the system. Second, countries with nuclear power may have very large losses. It is therefore in the interest of these countries to clarify the liability and the sharing of the potential costs ex ante. Third, pressure groups which are very strong economically have much to gain from an ex ante risk-sharing agreement, especially creditors and insurers with a lot of real estate in their portfolio or as surety.

3. Risk-sharing agreements: examples 3.1. Price-Anderson Act The implementation of such a risk-sharing agreement should not be considered impossible, for several reasons. A comparable, although not completely identical, system was introduced by the most recent amendment to the US Price-Anderson Act. The principle in US law is still individual liability of the operator of the power plant up to an amount of US$ 200 million,438 for which the operator will seek insurance coverage with the classic nuclear insurance pools. If the damage is higher than U$ 200 million, all nuclear power plants in the United States will share on a pro rata basis up to a total amount of US$ 7.2 billion for the 114 American power plants.439 The amounts have been raised since, so that the total now comes close to US$ 9 billion.440 This example shows that, through risk-spreading between all nuclear power plants, considerably higher amounts can be made available than with a liability limit which is set on the basis of the insurability of the risk by a national insurance pool.441

3.2. P&I Clubs Risk-sharing agreements are well-known in other fields of liability. For instance marine oil pollution is insured by the so-called Protection and Indemnity Clubs (P&I Clubs). The members of these clubs are tanker owners. They provide insurance on a non-profit basis for the members. At the beginning of each year, a “call” is made which should cover claims and administrative costs.442 These P&I Clubs function as a mutual insurance company. Profits and losses are shared amongst the members. If the receipts for one year are insufficient to cover losses, an additional call can be asked from the members.


165

5.


Obviously co-operation between commercial undertakings could again be viewed as endangering competition, also when it takes place to cover risks. One should note that as far as insurance undertakings are concerned, the exemption regulation 393/92 applies. In this respect we can refer to the discussion above concerning the conditions under which co-operation between insurers in a pooling arrangement falls under the exemption. The pooling by P&I Clubs had already been exempted previously from the application of the old article 85, 1 EEC Treaty, by a Commission decision of 16 December 1985. This individual exemption was criticised in the literature since it was held that the conditions for exempting P&I Clubs from competition law were not met 443 . These voices have apparently not convinced the Commission since in the recent report of 12 May 1999 on the exemption regulation, the Commission holds once more that co-operation between P&I Clubs is necessary to offer an adequate coverage (although the companies involved together have a market share of 89%!).444 Anyway this shows that, on the basis this recent report, at least P&I Clubs should not worry that their anti-competitive pooling strategies would be considered a violation of European competition law.

3.3. Nuclear liability: recent developments445 Also within the revision of the Paris and Vienna Convention, the drafting committees have discussed several alternatives in order to produce higher amounts of coverage. In this respect a proposal by the delegations of the United Kingdom and France is worth mentioning. They proposed a 4-tier system whereby the compensation would be provided as follows: a) “up to the amount provided under the legislation of a contracting party, by insurance or other financial security of the operator liable in accordance with the Vienna or the Paris convention; b) between that amount and (A) million Special Drawing Rights, art of funds provided by the operator liable which, subject to §4 of this Article, the operator shall have secured through his membership of a risk pool established by free association of operators; c) between (A) and (B) million Special Drawing Rights, out of public funds to be made available by the contracting party in whose territory the nuclear installation of the operator liable is situated; d) between (B) and (C) million Special Drawing Rights, out of public funds to be made available by the contracting parties according to the formula for contributions specified in Article 5.” The proposal moreover stipulated that the risk pool shall secure the availability of the funds through contracts binding upon all participants and

166


5.


shall allow for the accession of new participants on agreed terms which may take account, inter alia, of risk. Unfortunately, however, these proposals were not taken up either in the protocol to amend the Vienna Convention or in the Convention on Supplementary Funding, both of 12 September 1997. The idea remains that if the damage exceeds the liability of the operator, the state will have to intervene with public funds.

4. Risk-sharing versus insurance for systemic risks Obviously, pooling of resources by operators might be a good way of providing coverage for systemic risks. In that respect the examples of the nuclear area are totally applicable to other systemic risks as well. The crucial question is obviously whether the operators themselves are better equipped to achieve the fundamental goals of an efficient system of prevention and compensation for damage caused by systemic risks. Since we explained that such a system can only work efficiently insofar as risk differentiation can be applied, the crucial question is whether the operators of the pool can estimate the costs of liability claims and can judge, on the basis of actuarial expertise, which potentially responsible parties pose higher liability risks than others. Also, the technical annex made clear that pooling can only be considered an efficient funding system if the managers of the pool are able to identify the potentially responsible parties who are higher liability risks than others and if they can adjust their contributions to the pooled funding scheme accordingly.446 If, therefore, there are situations where one can assume that the risk is so technical and complicated that only the operators themselves can be judged to be capable of monitoring the risk and instituting preventive measures accordingly, a pooling system run by operators might create more efficient results than insurance. There is, moreover, one other important advantage of mutual risksharing via pooling. Skogh has shown that insurance requires that pricing be undertaken before the occurrence of the insured event. An agreement to mutually share each other’s losses does not necessarily require the ex ante payment of a premium. Mutual risk-sharing can be based on an agreement to share losses ex post, after the accident occurs and does not require actuarial information ex ante.447 Mutual risk-sharing seems therefore an ideal solution if the probability of an accident is statistically unpredictable. It is however, also necessary that moral hazard be controlled via mutual monitoring, e.g. by individuals in the same business. This explains why originally risk distribution started via mutual risk-sharing. Insurance can only develop when actuarial information is available. Hence, mutual risk-sharing e.g. via pools of operators, seems warranted when precise actuarial information on the predictability of the risk is lacking, as in the case of damage to biodiversity.


167

5.


However, one should realise that this is not always necessarily the case. Indeed, insurers are specialised in dealing with risk and hence, have also a good knowledge of risk differentiation. Moreover, the costs of running a pooling system might be high as well. In sum, the question of whether a pooling system is more efficient than an insurance scheme will not only depend upon the ability of the insurers or the managers of a pooling system to monitor the risks involved, but also upon the transaction costs of both schemes. In any event, a pooling system should not be excluded as a way of providing financial security to cover damage caused by systemic risks. However, if pooling systems are used, the relevant administrative agency should obviously receive sufficient guarantees that at the end of the day the pool can meet the obligations of the operators involved. As was shown in the nuclear example, some reinsurance by the pool may be necessary.

F. Ex ante guarantees and deposits Many states provide the possibility for operators of classified activities to provide financial security via other means than through liability insurance. The same is true for international conventions, which impose on licensees an obligation to provide financial security. To illustrate the wide range of possible financial guarantees, we propose to have once more a look at the Flemish example and more particularly at the proposals of the Inter-University Commission for the revision of environmental law in the Flemish region, which include elaborate provisions concerning financial guarantees.

1. Obligation of financial security: a balanced approach The obligation to provide financial guarantees is provided in part 9, dealing with the compensation of damage caused by pollution. The main provisions concerning environmental liability, as proposed by this commission, have been set out above. Chapter 3 of part 9 specifies in section 1 that the Flemish government shall designate those categories of classified installations and activities for which the permit-holder or the holder of a notification certificate is bound to provide financial guarantees for the purpose of guaranteeing liability for damage or for the impairment of the environment which may be caused by the installations or activities.448 The advantage of this system is obviously that there is no general duty for all operators of classified installations to constitute financial security. It will be the government or the administrative authorities that decide in which cases financial security is required. However, Article 9.1.14 Para. 3 of the draft decree on environmental policy states that there shall be an obligation to provide financial guarantees for a soil clean-up operation. It is interesting that, as far as the nature of the financial guarantees is concerned, Article 9.1.16 of the draft decree on environmental policy does not

168


5.


only rely on liability insurance, but stipulates that the financial guarantee in question may assume any of the following forms: ●

Insurance policy;

●

Guarantee provided by a financial institution;

●

Any other form of personal or collateral security;

●

A deposit paid by way of security on a separate account with the environmental guarantee account.

The amount of the financial guarantee shall be fixed by the Flemish government by means of general, sectoral and integral conditions, in the light of the dangers presented to man and to the environment by various categories of installations and activities. In addition, Article 9.1.17 Para. 1 stipulates that, for each category, separate guarantees are provided for a. damage caused by death or by personal injury and b. any other type of damage or impairment to the environment. The Inter-University Commission clearly assumed that the amount couldn’t be mentioned in general terms but will have to depend upon the type of installation. Apparently the Inter-University Commission also judged it important to provide for specific amounts for personal injury on the one hand, and for environmental damage on the other (as the type which is defined in the White Paper). Obviously Article 9.1.22 of the draft decree on environmental policy also stipulates that the money available under financial guarantee may only be claimed for the purpose of repairing damage or impairment of the environment. This is necessary, as was mentioned above, to prevent the guarantee which might have been set aside to cover environmental damage from being used by other creditors.

2. Optional financial caps The liability in the draft decree on environmental policy is in principle an unlimited liability. Hence, the provisions concerning the amounts only concern the amount of financial security. However, the draft decree on environmental policy introduces the possibility of a system of financial caps, which should not be ruled out immediately as being inefficient. Part 9 of the draft decree of the Inter-University Commission lays down rules concerning compensation of damage caused by pollution. Title 1 of this Part 9 provides for rules concerning liability and financial guarantees. Within this Title 1, Chapter 4 deals with “Limitation of liability by establishing a voluntary guarantee-fund”. Article 9.1.24 of the draft decree provides in para. 1 that the operator may limit any liability which he may incur pursuant to the provisions of this title


169

5.


for any damage to the environment, to the amount which he is bound to provide by way of financial guarantee where he is able to prove that a guarantee-fund guarantees the compensation of any damage or impairment of the environment exceeding this amount within the limits and subject to specific conditions. The limit only applies when the guarantee-fund provided by the operator guarantees the compensation over and above the amount for which the operator is liable anyway to provide a financial guarantee, and should attain specific amounts to be determined by implementing regulations. In addition, the possibility of limiting liability only applies when the guarantee fund proposed by the operator has been recognised by the Flemish government. When the guarantee-fund proves to be inadequate for the purpose of paying compensation within the limits set, the operator shall have unlimited liability in respect of these amounts. This does not mean that there is unlimited liability, but liability up to the amount established by implementing regulations.449 Where an intentional fault on the part of the operator occurs, the operator shall continue to have unlimited liability.450 The explanatory memorandum451 makes clear that the general rule in the environmental liability regime proposed by the Commission still is the unlimited liability. The draft Article 9.1.24 only provides for an optional mechanism which in fact trades a limitation of the operator’s liability against the guarantee of the availability of additional capital for the compensation of victims. Operators already have to provide a minimum financial guarantee for their liability. The proposed article only offers them the possibility of limiting their personal liability by providing a second, substantially higher amount of guarantees to a compensation fund. (The word compensation fund, which is used in the Flemish text is somewhat misleading since it concerns a fund which only serves to compensate the damage caused by the particular injurer). Although the text of the Flemish proposal is not clear in every respect, this proposal may well correspond to the principles of the theoretical model. A basic financial guarantee will always have to be provided on a mandatory basis by the operator. The draft only concerns the payment of a substantial higher amount as a second layer, which constitutes an “optional limitation fund”. This additional guarantee may be covered by insurance, which gives the insurer incentives to exercise additional control. The efficiency of the regime will obviously depend upon the expected amount of the damage. But the administrative authority (Flemish government) can apparently control the amount in every specific case. Thus the government makes a trade-off and accepts the optional limitation fund (for an amount which should in principle be sufficient to cover every loss) which provides a higher certainty of compensation for victims on the one hand, and certainty for risk-averse injurers on the other. This is the main difference from generalised statutory caps, discussed above.452 These generalised caps were judged inefficient because they do not take into account the individual financial situation of the polluter nor the damage he

170


5.


may cause. This is different from the individualised optional caps proposed by the Flemish Inter-University Commission. The advantage for operators in participating in this regime is obviously that they can limit ex ante their exposure to liability by providing a guarantee for an amount, which in principle will cover all losses. Thus the system also allows for a differentiation of risks and injurers. Moreover, the injurer’s incentives will remain unaffected since unlimited liability still applies in the case of an international fault (Article 9.1.24 para. 6).

3. Deposits In addition to these interesting provisions concerning the voluntary guarantee fund, the draft decree on environmental policy also contains provisions regarding the obligation to provide deposits. In some cases, the permit-holder may be required to pay a deposit into an environmental guarantee account. Interestingly enough, these deposits are held in a separate account for the benefit of the depositor and can only be used to cover expenses related to the treatment they are intended to cover.453 When all the obligations have been fulfilled by the permit-holder, the deposit and the interest on it have to be repaid to the permit-holder. This means that effectively the deposit is a guarantee that remains at the disposal of the authorities who run the environmental guarantee account and generates proceeds in the form of interest which is returned, for example, when an operation is terminated and no damage has occurred. Title 4 of part 9 deals in more detail with the environmental guarantee account.

4. A flexible solution In sum, this Flemish example shows that there are many more financial techniques available to provide financial security for repairing harm to biodiversity or to the soil than merely liability insurance. The draft decree not only refers to guarantees provided by a financial institution, personal or collateral security, but even to deposits paid to an environmental guarantee account. In all of these cases, the goal is to make sure that when environmental damage occurs, sufficient coverage is available to compensate the damage the potential responsible party may cause. All these instruments are, in other words, meant to resolve the insolvency problem with the related risks of undercompensation and underdeterrence. At the policy level, the Flemish example is important since the proposals of the Inter-University Commission for the revision of environmental law in the Flemish region were put forward after long consultations with both the insurance industry and financial institutions and the industrial world. In particular, the idea of deposits paid into an environmental guarantee account, which could generate interest payments to the benefit of the industrial operator, seemed very attractive to industry, especially if the payments were made tax-deductible. The major advantage is that the payments of deposits into a guarantee account are not lost in the event that an activity is exercised


171

5.


without any damage (hence, it provides excellent incentives for prevention), whereas the premiums paid to an insurance undertaking are obviously lost for industry, even if no damage occurred during the operation of a plant. The Flemish example shows that mandatory financial security can be required on a flexible basis, without necessarily having to recourse liability insurance. It shows that it is better to work with flexible solutions where the form and amount of the financial security required is tailored according to the needs of the specific case. This calls for regulation either on a sectoral base or via permit conditions set by administrative agencies, instead of general rules which would lack the flexibility required.

G. Summary In this chapter we analysed whether there were other financial arrangements than traditional liability insurance that could be used to cope with damage caused by systemic risk. The question of alternative financial arrangements and new financing techniques is an essential one in the light of the fact that liability rules will have a deterrent effect only if the insolvency risk can be deal with through the availability of financial security. Otherwise the insolvency problem may lead to underdeterrence. We also noted that the development of the liability insurance market in some countries is, with respect to particular systemic risks, relatively weak. This was for instance the case with environmental risks, but may be the case as well with other newly emerging systemic risks such as those related to new technologies. The problem in these cases is that the information required to make an adequate risk assessment may be lacking, as a result of which traditional insurance markets may not immediately wish to engage in the coverage of these risks. Moreover, we also indicated that the problems of scale and capacity with large systemic risks may be so huge that traditional insurance markets may not be able to deal with them adequately. Alternatives thus need to be sought both within and outside traditional insurance markets. We have looked at various alternatives, both in theory and in practice. In various areas of risk which could be described as “systemic”, the development of new financing techniques for dealing with risk may be noted. One of the well-known alternatives outside the traditional insurance branch is the use of capital markets. This idea, which has been advanced mainly by economists, is theoretically sound. Investors could purchase “risk bonds” and stock prices would reflect investors’ perceptions concerning the riskiness of the activity. In theory this instrument may well work, although it is clear that only larger investors able to recognise good and bad risks will be engaged in such a market. Until now the use of capital markets to deal with risks has mainly been limited to the US, although it has its advocates in Europe as well.

172


5.


Obviously, self-insurance may be a valuable alternative as well. Larger enterprises in particular, which may be exposed (as injurer or victim) to larger systemic risks, may be able to cover a large part of the losses themselves, and hence not need to purchase additional financial security. One should note in this respect that the purchase of additional financial security will always lead to an additional cost. Thus one may well understand why enterprises exposed to risk may primarily self insured. However, although self-insurance is certainly useful, we also voiced several doubts about it. If, at the policy level, one wants to make sure that sufficient coverage will be available if an accident occurs, selfinsurance may not be a very reliable alternative. The key issue is that guarantees should be provided to ensure that victims will be compensated in case of an accident. Moreover, it should also be made clear that these funds are available for victims only and not for other creditors (or for the trustee in bankruptcy, for example). Self-insurance may not be able to provide these guarantees. Moreover, we indicated that in many cases self-insurance is not insurance in the traditional sense since it does not contain an element of risk-spreading. It can, however, be considered an instrument of financial security or, as it is sometimes referred to, financial “assurance”. The same is true for another alternative often mentioned – the use of socalled captives. Often captives are presented as separate forms of financial security, whereas it is in fact rather difficult to classify them precisely given that their nature may very much differ. In some cases, captives have a legal personality and constitute a separate capital which actually is a reserve set aside by a particular company for future risk. Such captives are not a form of insurance but have been created mainly for tax purposes. Otherwise a reserve for future risk may not always be tax deductible, this depending upon the tax rules applicable in the particular legal system. In some cases, captives are set up by various companies and imply some form of risk-spreading. However, although captives can be highly useful for the enterprises involved, they may pose the same problems as self-insurance, at least from a public policy perspective. Indeed, unless specific legal rules apply, there is not necessarily a guarantee that the capital available via the captive will also be available for accident victims ex post. Therefore, other alternatives have been examined as well. An interesting alternative within the insurance world itself is the trend towards first-party or direct insurance. Theoretically one can argue that as the key remedy to uninsurability is an adequate risk differentiation, it may be achieved better with first-party or direct insurance than with liability insurance. Hence, one can understand the trend in many countries to move toward first-party/direct insurance solutions for many systemic risks. This may be noticed for instance in the area of occupational health where employers will take out direct insurance to the benefit of their employees guaranteeing compensation in the event of an


173

5.


accident at work, for example, irrespective of the liability of the employer. A similar trend can be seen in environmental insurance as well. We presented the Dutch experience with environmental damage insurance, showing that this new instrument may be more effective in providing compensation than traditional liability insurance. Obviously the scope of coverage (and therefore the compensation involved) will depend upon the contents of the insurance policy, but initial experience with this new instrument seems quite promising. Clearly, this new insurance technique was introduced directly in response to on the growing problems of insuring systemic risks. We also showed that in some cases and for some systemic risks, risksharing agreements by operators are an alternative. They may provide good results if either capacity on insurance markets is limited or if it can be assumed that the information required to optimally control risk (and therefore moral hazard) is more available to the operators themselves than to insurers. Experiences with some larger systemic risks (oil pollution and the nuclear risk) have shown that such risk-sharing agreements may in some cases provide better results than traditional liability insurance. An advantage of these alternative arrangements is, moreover, that one does not necessarily need adequate information on the predictability of the risk ex ante, provided that adequate monitoring is possible. Obviously within the scope of this chapter we could only discuss a few new insurance and financing techniques for covering damage caused by systemic risks. However, a variety of other arrangements provided by financial markets could be used as well, such as of ex ante guarantees or deposits into an environmental savings account. We discussed briefly proposals by the Flemish Inter-University Commission for a variety of financial instruments which could be used by operators to meet their duty to provide financial security. On balance, at the policy level it does not seem useful or appropriate to make a decision in favour of one particular financial or insurance instrument. It seems better to take into account the existence and working of all of these financial and insurance institutions. If the policy-maker wished to make sure that financial guarantees were available so that ex post damage caused by particular systemic risks would be covered (and thus a duty to provide financial security – see below – would be introduced), it would be better to give a particular agency or administrative authority the task of assessing whether the financial security offered would be effective. This means that the agency would have to check that it can be guaranteed that the money will ultimately be available when required, meaning when the systemic risk materialises. This point obviously remains a crucial one and is a major weakness of so-called selfinsurance or captive schemes. However, the necessary flexibility remains important. Thus, depending upon the structure of the particular financial or

174


5.


insurance market, enterprises may wind up with a variety of instruments for covering future losses as a result of systemic risk. Experience shows that this flexibility will have the advantage of pushing industry towards investments in research to discover optimal ways to cover risks. Take the example of the nuclear risk: the problems on the traditional nuclear liability insurance market have led operators (and in some cases insurers) to develop new instruments such as the risk-sharing agreements mentioned above. Thus the policy-maker should always ensure that legislation is structured in such a way that it provides adequate incentives to operators to develop new financial and insurance techniques which may lead to additional coverage for damage caused by systemic risks. There remains, however, one important weakness with all of these alternative financial arrangements. All of them assume that a known source can be found to which the particular systemic risk can be attributed. It is then assumed that this known source (a polluter, tortfeasor or injurer) can be made liable and hence obliged to compensate. In many cases, however, it may not be possible to attribute the consequences of systemic risks to a particular source. The reasons may be manifold. There can either be problems in proving the causal link between the damage and a particular source or there may be problems due to the legal system e.g. in proving wrongfulness in tort. For this reason, many accidents may remain uncompensated. This again prompts calls for first-party insurance by victims, as discussed above. On the other hand, it may also be asked whether there is a need for government intervention. Indeed, first-party insurance may not be able to cope with all problems either. In some cases, the reasons for uninsurability in liability insurance will also remain if one turns to first-party insurance. One only has to think of terrorist attacks. The scale problems caused by terrorist attacks may be so huge that the capacity problems inherent in insurance will still remain if one moves to first-party solutions. Many problems facing the insurance and reinsurance world today as a result of the attack on the Twin Towers on 11 September 2001 appear more acutely in the first-party insurance sphere. If even the limits of first-party insurance coverage are reached, inevitably the public will call on the government for intervention. We thus have to address the role of government in dealing with systemic risks, which we shall do in the next chapter.

Notes 398. For a good overview see M. Radetzki (supra note 304) pp. 180-195. 399. See J.R. Tyran/P. Zweifel (supra note 282) pp. 431-444. 400. Tyran and Zweifel ((supra note 282), p. 433) report that the 1906 San Francisco earthquake caused insured damage of $ 39,5 billion (US). This sum is 100 times greater than the maximum nuclear coverage granted by private insurers. Other


175

5.


examples can be given where the probabilities of an accident are unknown, the expected accident costs are large, but large amounts of insurance coverage are nevertheless provided. 401. The possibilities of financial markets to manage the uncovered liabilities of industrial catastrophes are further explored in M. Radetzki/M. Radetzki (supra note 304) pp. 188-193. 402. See F. Wagner, Risk Securitization. An Alternative Risk Transfer of Insurance Companies, 1998, GPRI, 575. 403. See R.E. Smith/E.A. Canelo/A. Di Dio, Reinventing Reinsurance using the Capital Markets, 1997, GPRI, 26-37. 404. See R.E. Smith/E.A. Canelo (supra note 403) p. 31. 405. J. Zech, Will the international insurance market replace traditional insurance products, 1998, GPRI, 494. 406. Reserves, of course, allow for risk-spreading in time, but not between various parties exposed to risk. 407. On captives and the different forms they may take, see: P.A. Bawcut, Captive Insurance Companies, Establishment, Operation and Management, 1991; T. Dowding, Global Developments in Captive Insurance, 1997 and Captive Insurance: a threat to the Global Insurance Industry?, 1993. 408. Again, the word “reinsurance” is slightly misleading since it assumes that there is a basic insurance obligation which is reassured. 409. J. Rogge (supra note 257) p. 38. 410. For a good overview of alternatives see Tyran, J.R. and Zweifel, P., “Environmental risk internalisation through capital markets (Ericam): the case of nuclear power”, International Review of Law and Economics, 1993, 431-444; Radetzki, M. (supra note 304), 180-195. 411. Priest, G., “The current insurance crisis and modern tort law”, Yale Law Journal, 1987, 1521-1590. 412. See Arrow, K., Aspects of the Theory of Risk-Bearing, Helsinki, Yrjö Jahnssonin Säätiö, 1965 and Borch, K., “The Utility Concept Applied to the Theory of Insurance”, The Astin Bulletin, 1961, 245-255. 413. See Shavell, S., Economic analysis of accident law, Cambridge, Harvard University Press, 1987, 190. 414. Skogh, G., “The transactions cost theory of insurance: contracting impediments and costs”, Journal of Risk and Insurance, 1989, 726-732. 415. This argument is – again – especially advanced by Priest, G., “The current insurance crisis and modern tort law”, Yale Law Journal, 1987, 1521-1590. 416. Many other law and economics scholars have pointed out the advantages of first-party insurance. See e.g. Bishop, W., “The Contract-Tort Boundary and the Economics of Insurance”, Journal of Legal Studies, vol. 12, 1983, 241-266 and Epstein, R.A., Simple Rules for a Complex World, Cambridge, Harvard University Press, 1995, 221 and, in the product liability area, Epstein, R.A., “Products Liability as an insurance market”, Journal of Legal Studies, vol. 14, 1985, 645-669. 417. See Wagner, G., “Versicherungsfragen der Umwelthaftung” in Ahrens, M. and Simon, J. (eds.), Umwelthaftung, Risikosteuerung und Versicherung, Berlin, Erich

176


5.


Schmidt Verlag, 1996, 104-105 and Shavell, S., “On Moral Hazard and Insurance”, Quarterly Journal of Economics, 1979, 541-562. 418. See e.g Danzon, P., “Alternative liability regimes for medical injuries”, Geneva Papers on Risk and Insurance, 1999, 3-22 and see Koziol, “Die Arzthaftung im geltenden und künftigen Recht”, in Haftungsrechtliche Perspektiven der Ärztlichen Behandlung, Linz, Universitätsverlag Rudolf Trauner, 1997, 21-35. 419. CEA, Study on first-party legal obligations for clean-ups and corresponding insurance covers in European Countries, Paris, CEA, 21 October 1998. 420. See the summary tables in the CEA study, 32. 421. Aansprakelijkheidsverzekering bedrijven. 422. See Wansink, J.H., “De nieuwe milieuaansprakelijkheidsverzekering”, Milieu en Recht, 1985, 98. 423. Milieu-aansprakelijkheidsverzekering Samenwerkingsverband. 424. I am grateful to Mr. P.A.J. Kamp of Nationale Nederlanden who provided further insights into the reasons behind the changes regarding first-party insurance in the Netherlands. 425. For further details on these difficulties, see Wansink, J.H., “Hoe plotseling en onzeker is de verzekeringsdekking voor millieuaansprakelijkheidsrisico’s?” in Miscellanea. JurisConsulto vero dedicata, Essays offered to Prof.Mr. J.M. van Dunné, Deventer, Kluwer, 1997, 451-460. 426. Milieuschadeverzekering. 427. For a good description of the new policy see Wansink, J.H., “Verzekering en milieuschade als gevolg van vervoer/opslag van gevaarlijke stoffen”, Tijdschrift voor Milieuaansprakelijkheid, 1999, 77-82 as well as Janssen, C.A., “Aansprakelijkheid voor milieuschade en financiële zekerheid naar toekomstig recht: nieuwe oplossingen. Nederlands Recht” in Wiggers-Rust, L.F., en Deketelaere, K. (eds.), Aansprakelijkheid voor milieuschade en financiële zekerheid, Die Keure – Vermande, 1998, 111-112 and Drion, P.J.M., “Milieu onder één dak: milieuschadeverzekering (MSV)”, Verzekeringsrechtelijke Berichten, 1998/2, 19-21. 428. Article II.1.1. defines as the scope of coverage: “Insured are the costs of remediation of the insured site. This remediation must apply to a pollution which is the direct and exclusive result of an emission, caused by one of the insured risks...”. 429. This legal aid coverage, however, is not applicable if the insured took out a small amount of cover. 430. See Wansink, J.H., “Verzekering en Milieuschade als gevolg van vervoer/opslag van gevaarlijke stoffen”, Tijdschrift voor Milieuaansprakelijkheid, 1999, 81. 431. In principle, damage caused from the insured site to third parties is always covered but the amounts of coverage may be limited. 432. Wansink, J.H., “Verzekering en Milieuschade als gevolg van vervoer/opslag van gevaarlijke stoffen”, Tijdschrift voor Milieuaansprakelijkheid, 1999, 81-82. 433. Niezen, J., “Nieuwe milieuschadeverzekering – geen panacee”, Milieu en Recht, 1998, 114.


177

5.


434. Cousy, H., “Recent developments in environmental insurance”, in Abraham, F. Deketelaere, K. and Stuyck, J. (eds.), Recent economic and legal developments in European environmental policy, 240. 435. Also Bergkamp argues strongly in favour of first-party and against environmental liability insurance (Bergkamp, L., “The Commissions White Paper on Environmental Liability: A weak case for an EC Strict Liability Regime”, European Environmental Law Review, 2000, 112-114. 436. Ranson reports that in Belgium too, ’direct’ environmental insurance is offered which also cover gradual pollution (but excludes ecological damage). See Ranson, D., “Verzeskering van milieuaansprakelijkheid”, Milieu- en Energierecht, 2000, 68. The new policy also covers remediation costs like the Dutch system. It is offered by AIG and provides coverage up to BF 1 billion (see H. Kerremans, Aansprakelijkheid voor Milieuschade en verzekeringsmogelijkheden in Milieuzorg in de Onderneming, I., Juridische, fiscale en organisatorische aspecten, Antwerpen, Standaard, pp. 537-583). 437. See M. Faure/G. Skogh (supra note 16) pp. 499-513. 438. See M. Trebilcock/R. Winter (supra note 291) p. 221. 439. For a description of the American system, see J. Marrone, Closing the Circle of Protection for the Public-the Evolution of the System in the United States, paper presented at the symposium “Nuclear accidents-liability and guarantees”, Helsinki, Finland, September 1992; F. Heimann, The US Liability Protection System for Nuclear Power Plants, paper presented at the symposium “Nuclear accidents, liability and guarantees”, Helsinki, Finland, September 1992; M. Robesin, Aansprakelijkheid voor kernongevallen, ruim een jaar na Tsjernobyl, 1987, MenR, 228-229 and M. Rowden/J. Kraemer/L. Cuoco, La révision de 1988 de la loi Price-Anderson ou mieux vaut tard que jamais, Bulletin de droit nucléaire, Nr. 42, 79-101. 440. According to M. Radetzki (supra note 304) p. 188. 441. For an in-depth analysis of the liability regime under the Price-Anderson Act, and a comparison with European compensation schemes, see T. Vanden Borre (supra note 270) pp. 443-490 and see T. Vanden Borre (supra note 289) pp. 280-290 and 639-674. 442. See J. Bongaerts/A. Debièvre, Insurance for Civil Liability for Marine Oil Pollution Damages, 1987, GPRI, 145-187; M. Faure/G. Heine (supra note 21) pp. 39-54 and T.G. Coghlin, Protection and Indemnity Clubs, 1984, Lloyd’s maritime and commercial law quarterly, 403-416. 443. See M. Faure/R. Van den Bergh (supra note 375) pp. 331-336. 444. Commission report, No. 29. 445. For these recent developments, see also H. Conruyt-Angenent, L’évolution en droit belge de la réponation des dommages nucléaires à la lumière des conventions internationales, in: J. Rogge (ed.), Liber Amicorum René van Gompel, Études en assurances, 1998, pp. 75-108 and see T. Vanden Borre (supra note 289) pp. 25-49 and T. Vanden Borre (supra note 289) pp. 117-141. 446. Technical Annex, pp. 3-4. 447. See G. Skogh, Development Risks, Strict Liability and the Insurability of Industrial Hazards, 1998, GPRI, 247-265 and G. Skogh, Mandatory Insurance: Transaction Costs Analysis of Insurance, in: B. Bouckaert/G. De Geest (eds.), Encyclopedia of Law and Economics, 2000, pp. 521-537. 448. See the proposed Article 9.1.14 of the draft decree on environmental policy.

178


5.


449. See Article 9.1.24, para. 5. 450. Article 9.1.24, para. 6. 451. H. Bocken/D. Ryckbost/S. Deloddere, Liability and Financial Guarantees, in: H. Bocken/D. Ryckbost (eds.), Codification of Environmental Law. Draft Decree on Environmental Policy, 1996, p. 214-223. 452. See Chapter 7. 453. See H. Bocken/D. Ryckbost, Deposits, in: H. Bocken/D. Ryckbost (eds.), Codification of Environmental Law. Draft Decree on Environmental Policy, 1996, p. 224.


179


Chapter 6

The Role of Government


181

6.

THE ROLE OF GOVERNMENT

A. Introduction In this study we outlined various ways in which insurers can cope with newly emerging systemic risks, but indicated at the same time that there may also be limits. Emerging risks may be expanding to such an extent that the limits of insurability may be exceeded. Therefore, new financing techniques and alternative compensation mechanisms other than insurance (or within insurance schemes) were examined. However, we equally made clear that even these alternatives, which were discussed in Chapter 5, may have their limits as well. A question which therefore has to be addressed is what the much debated role of government should be in the regulation of systemic risks, and, more particularly, whether the government should intervene to provide for compensation if insurance and financial markets are not capable of covering risks. This may very well happen in the case of catastrophic risks. One only has to think of terrorist attacks. A first point we would like to address in this respect is whether society can rely solely on the deterrent effect of liability rules to prevent accidents. We stressed throughout this study that a compensation system cannot solely focus on the indemnification of victims, but should also always address the question of how a particular compensation system affects the incentives to prevent accidents. The key idea should still be that the best form of victim protection is prevention: it is better to have no victims at all than victims who are protected. The idea that all newly emerging systemic risks can be prevented through the use of liability law may be somewhat naive. Therefore we will first of all address the role of regulation in preventing risks (B). But there may also be a role for government in providing compensation for damage caused by systemic risks. We have already indicated that the increasing use of liability law, for example in Europe, may be largely due to the fact that many governments have withdrawn from social security provision. Hence the question arises whether the government should intervene e.g. via social security, as the new risks can cause huge damage which may not be covered via the liability system (C). Catastrophic risks (floods, earthquakes or terrorist attacks) often cause a high number of victims. This usually brings about a call for public intervention via compensation funds. This is especially the case when indemnification of victims is not possible, or not complete, via liability law. Therefore one should examine those cases in which compensation funds may provide an adequate response to damage caused by systemic risks (D).

182


6.


Finally, we will look at the question of how the government can intervene to promote the provision of financial security via private insurance and financial markets. Many will argue that, given the insolvency risk, there should be a guarantee that private operators have funds available to cover the damage caused by systemic risks. This raises the question of whether compulsory insurance should be introduced (E). On the other hand, we will also look at the various facilitative strategies that the government could adopt to promote the insurability of risks. One clear element in this respect is the role of competition law. Experience with nuclear risks, for example, has shown that if there is a high degree of concentration on insurance markets, this may limit the availability of insurance. Hence, competition law can be seen as an important tool for promoting insurability. Moreover, we have indicated that in some cases, it may be better that victims choose protection themselves, e.g. via first-party insurance, for newly emerging systemic risks. But in that case it would be important that victims be made aware of certain risks and of the possibilities of purchasing insurance coverage. These types of information provision can also constitute one of the facilitative strategies to be followed in public policy (F).

B. Safety regulation Above we assumed that liability rules are necessary to prevent harm caused by systemic risks. In practice, however, activities which may cause systemic risks are often subject to a large degree of safety regulation. These safety regulations also aim to prevent environmental harm. This therefore raises the question of whether liability rules still should have a deterrent effect, given the existence of regulation. In addition, one could ask why safety standards are set at all and why governments do not simply rely on the deterrent effect of liability rules. It is therefore important to point to the precise role of liability rules – taking into account the existence of extensive safety regulation – in controlling systemic risks.

1. Criteria for safety regulation Let us now examine under what kind of circumstances liability rules may not suffice to deter systemic risks and regulatory intervention may be necessary. The choice between regulation and liability rules was thoroughly examined by Steven Shavell in 1984 in a paper in which he advanced several criteria that influence the choice between safety regulation and liability rules.454

1.1. Information asymmetry as a criterion for regulatory intervention Information deficiencies have often been advanced as a cause of market failure and as the justification for government intervention through regulation.455 Also, for the proper operation of a liability system, information on existing legal rules, the accident risk, and efficient measures to prevent accidents,


183

6.


is a precondition for efficient deterrence. According to Shavell, the parties in an accident setting generally have much better information on the accident risk than that possessed by the regulatory body.456 The parties themselves have, in principle, the best information on the costs and benefits of the activity that they undertake and of the optimal way to prevent accidents. This “assumption of information” will, however, be reversed if it becomes clear that some risks are not readily appreciated by the parties in an accident setting. This may be particularly the case if costs are external. These cannot always be easily assessed by the parties involved. Therefore, for every activity the question that will have to be asked is whether either the government or the parties involved can acquire the information at the least cost.

1.2. Insolvency risk If the potential damages can be so high that they will exceed the wealth of the individual injurer, liability rules will not provide optimal incentives. The reason is that the costs of care are directly related to the magnitude of the expected damages. If the expected damages are much greater than the individual wealth of the injurer, the injurer will only consider the accident as having a magnitude equal to his wealth. He will take, therefore, only the care necessary to avoid an accident equal to his wealth, which can be lower than the care required to avoid the total accident risk.457 This is a simple application of the principle that the deterrent effect of tort liability works only if the injurer has assets to pay for the damages he causes. If an injurer is protected against such liability, a problem of underdeterrence arises.458 Safety regulation can overcome this problem of underdeterrence caused by insolvency.459 Efficient care will be determined ex ante by regulation and will be affected by enforcement instruments which induce the potential injurer to comply with the regulatory standard, irrespective of his wealth. A problem might still arise if regulation were also enforced by means of monetary sanctions. Again, if these were to exceed the injurer’s wealth, the insolvency problem would remain. Hence, if a safety regulation is introduced because of a potential insolvency problem, the regulation itself should be enforced by non-monetary sanctions.460

1.3. The threat of a liability suit Some activities can cause considerable damage but even so a law suit to obtain damages may be never brought. If this were the case, there would of course be no deterrent effect of liability rules. Therefore, the absence of a liability suit would again be an argument for enforcing the duty of efficient care by means of safety regulations rather than through liability rules.461 There can be a number of reasons why a law suit is not brought even though considerable damages have been caused.

184


6.


Sometimes an injurer can escape liability because the harm is thinly spread among a number of victims. As a consequence, the damage incurred by every individual victim is so small that he has no incentive to bring a suit. In particular, this problem will arise if the damage is not caused to an individual but to a common property such as the surface waters in which each member of the population has a minor interest. In addition, a long time might have elapsed before the damage becomes apparent, so that much of the necessary evidence may either have been lost or be impossible to obtain. Another problem is that if the damage only manifests itself years after the activity, the injurer might have gone out of business. A related problem is that it is often hard to prove that a causal link exists between an activity and a type of damage.462 The burden of proof of a causal relationship becomes more difficult with the increasing passage of time since the incident that caused the damage took place. Often a victim will not recognise that the harm had been caused by a tort but might think that his particular ailment, e.g. cancer, had a “natural cause” associated with general ill health. For all these reasons, a liability suit might not be brought and hence safety regulation is necessary to ensure that the potential injurer takes efficient care.463 An additional problem is that the liability system cannot always incorporate all social costs of harm as damage which has to be compensated by an injurer. This is typically the case for the so-called non pecuniary losses, such as pain and suffering. It is well-known that these non pecuniary losses may be very difficult to estimate. Think in this respect of the deterioration of image and reputation, but also the losses one suffers because of the death of a relative. The administrative costs for courts to valuate these losses may be huge. Economics generally makes the argument that tort law should, as far as possible, incorporate these types of non pecuniary losses as well, especially in unilateral accident types.464 However, the amount to which tort law succeeds in incorporating these non pecuniary losses in damage awards may be sub optimal. In some of these cases, society might simply want to prevent an action it considered inappropriate. Think again of the secondary economic consequences of e.g. BSE or the foot and mouth outbreaks in the UK. In these cases, the role of the liability system is modest and inadequate compensation can lead to underdeterrence. It is clear that in those cases other types of intervention are necessary, more particularly the ex ante prohibition of certain activities. This hence requires a regulatory intervention with the threat of sanctions on those who would violate such a prohibition.465

1.4. Administrative costs When examining the pro and cons of liability versus regulation, the administrative costs of both systems should also be compared. Liability rules are clearly costly in terms of time for both parties and in court fees. Part of these costs is borne by the whole community, such as the cost of the legal


185

6.


system, fees for the judges etc. Regulation produces costs for the community, including the costs of making the regulation, setting the standards, passing the statutes etc. and of subsequent enforcement.466 In this respect the liability system seems to have an advantage: the administrative costs of the court system are only incurred if an accident has actually happened. The main advantage of the tort system is that a lot of accidents will be prevented by the deterrent effect of being held liable and having to pay damages to the victim, whereas with safety regulation, the costs of passing the regulation and of enforcing it are always there whether there are accidents or not.

2. The need to regulate systemic risks After having discussed these criteria for regulation467 we can now turn to the question of how these criteria relate to systemic risks. If one takes the criteria for safety regulation discussed above and applies them to the potential damage caused by systemic risks, there is no doubt that liability rules alone are not sufficient. If one looks at the first criterion, that of information costs, it must be stressed that an assessment of the risks of a certain activity often requires expert knowledge and judgement. Small organisations might lack the incentive or resources to invest in research to find out what the optimal care level would be. Also, there would be little incentive to carry out intensive research if the results were automatically available to competitors in the market: this is the wellknown “free rider” problem. This problem can partially be countered by legal instruments granting an intellectual property right to the results of the research. However, the problem remains that it may not be possible for small companies to undertake studies on the optimal technology for preventing environmental damage. Therefore, it is often more efficient to allow the government itself to do the research on the optimal technology (e.g. in a government research institute). The results of this research can then be passed on to the parties in the market through the regulation. Hence, the setting of standards can be seen as a means of passing on information on the minimal technology required. Obviously, it is more efficient for the government to acquire information on the optimal standard than it would be for an individual firm to find out what additional reduction in risk would produce an optimal reduction of the expected damages from the particular systemic risk. There are undeniable “economies of scale” advantages in regulation. Also, the risk of insolvency buttresses the argument for regulation. Systemic risks can be caused by individuals or firms with assets which are generally lower than the damages they can cause as a result of the systemic risk. In this respect it should not be forgotten that even a small firm could cause harm to a large number of individuals. Take the example of terrorist

186


6.


attacks. The insolvency problem is very clear there (and liability rules could hardly deter such behaviour). But the same is even true for firms. Also firms can cause damage resulting from systemic risks that may largely exceed their individual assets. Moreover, most firms have been incorporated as a legal entity and therefore benefit from limited liability. Hence, the individual shareholders are not liable to the extent of their personal assets, but a creditor of the firm can only lay claim to part of the total assets purchased in the firm by the shareholders. Also the chance of a liability suit being brought for damage caused as a result of systemic risks is naturally very low. The damage is often spread over a large number of people who will have difficulties organising themselves to bring a law suit. In addition, the damage could only become apparent some years after the wrongful act took place. This will bring proof of causation and latency problems, which will only make it more difficult for a lawsuit to be brought against the party who caused the systemic risk. For these reasons, it is clear that some form of government regulation of systemic risks is necessary. To reformulate: we have shown that liability rules alone cannot suffice to prevent systemic risks but that there might be other, publicly imposed, instruments than command-and-control-type regulation which can be used to reach this goal. Taxes are obviously such an alternative. But these are also publicly imposed and can hence be considered as “regulation”. Another question, which will be discussed below, is whether this necessarily implies that systemic risks should solely be governed by regulation or whether liability can still fulfil a supplementary role.

3. Safety regulation in practice: the example of environmental risks Let us now focus again on one particular example of a systemic risk – environmental pollution. In the introduction we made clear that systemic risks can be of a very differentiated nature. Therefore it is difficult to provide general conclusions. It seems useful to take the example of environmental risk to show that, in that particular area, a strong argument can be made for laying down ex ante by regulation, the level of efficient care to be taken to avoid that particular systemic risk. Indeed, in many cases the regulation consists of licences or permits whereby an administrative authority fixes an emission standard which must be followed by the potential polluter. These licences play a crucial role in environmental policy in most countries. An improvement of environmental quality will mostly be effected by imposing more stringent emission standards in administrative licences. Hence, the general requirement that emissions be controlled through licences and that the quality and quantity of emissions be regulated by the conditions in this licence, is a cornerstone of environmental law. Since these licences are administrative acts, in most legal systems


187

6.


environmental law is considered to be a part of administrative law. Criminal law usually only comes into the picture in order to sanction a violation of administrative regulations or emission standards in the licences. Although environmental pollution is in the first place controlled through these administrative licences, in individual cases there can still be damage to the environment. Then again, liability under tort law comes into the picture and the question is raised of the influence of regulation on the liability system and vice versa.468 These complementarities between tort law and regulation will be discussed below. Although it is difficult to ascertain whether environmental regulation is generally also effective in reducing environmental harm, some studies have attempted to examine the effectiveness of safety regulation in controlling such harm. These studies do not address the specific quality of every environmental law, but examine whether regulation has generally been more important in reducing environmental harm than liability rules. Dewees demonstrated that in North America the quality of the environment has improved substantially as a result of regulatory efforts rather than in response to legal action in tort.469 The empirical evidence of the success of regulation compared with tort law has been stressed in the recent book by Dewees/Duff/Trebilcock.470 They hold that the major regulatory effort to improve the environment has met with considerable success when measured by the reduction of emissions, but that it is more difficult to argue that the environmental regulations of the 1970s in US equally had a considerable influence on the ambient environmental quality. Moreover, they also stress that while environmental regulation is a determining factor in pollutant emissions and ambient concentrations, other non-regulatory factors such as economic growth and even the weather also influence environmental quality.471

4. The need to combine liability and regulation We have just stressed that, according to Shavell’s criteria, there is a strong argument for controlling systemic risks through ex ante regulation (or taxes). However, in individual cases there can still be damage. Then again, liability under tort comes into the picture and the question has been addressed in the literature of how regulation influences the liability system and vice versa. These complementarities between tort law and regulation have more particularly been addressed by Rose-Ackerman.472, Faure/Ruegg,473 Kolstad/Ulen/Johnson474 and recently by Arcuri475 and Burrows476. Rose-Ackerman also compared US and European experiences in using regulation versus tort law in environmental policy.477 The first point which is often stressed is that the fact that there are many arguments in favour of ex ante regulation of systemic risks does not mean that the tort system should not be used any longer for its deterrent and compensating functions. One reason to still rely on the tort system is that the

188


6.


effectiveness of regulation is dependent upon enforcement, which may be weak. In addition, the influence of lobby groups on regulation, to which public choice theory has rightly pointed, can to some extent be overcome by combining safety regulation and liability rules. Moreover, safety regulation, for example via licences, can become rapidly outdated and often lacks flexibility, which constitutes a further argument for combining it with tort rules. From the above it follows that although there is a strong case for the use of safety regulation to control systemic risks, tort rules will still play an important role as well.478 This obviously raises the question of whether compliance with regulation will affect the liability issue, to which we shall now turn.

5. Justificative effect of regulatory compliance An important aspect of the combination of liability and regulation is that the question arises of whether following regulation excludes a finding of negligence in tort law. Industry will often argue that as long as they have followed regulatory requirements, they cannot be held liable in tort any more. This is often referred to as the “justificative effect of a licence” or the “regulatory compliance defence”. The regulatory compliance defence is rejected in legal systems like those in Belgium and the Netherlands.479 One can find a clear economic rationale for this rule. If compliance with a regulatory standard or licence automatically resulted in a release from liability, the potential injurer would have no incentive to invest more in care than the regulation asks from him, even if additional care could still reduce the expected accident costs beneficially.480 The first reason to hold an injurer liable (if the other conditions for liability are met) even though he has followed the regulatory standard is that the standard is often merely a minimum. The complete "compliance defence" prevents any incentive to take precaution in excess of the regulated standard.481 Exposure to liability will give the potential injurer incentives to take all efficient precautions, even if this requires more than just following the licence. This, by the way, holds both under negligence and strict liability. Since the regulatory standard cannot always take into account all the efficient precautionary measures an injurer can take, testing the measures taken by the injurer even though the regulatory standard was followed, will provide additional incentives. Allowing a regulatory compliance defence would also largely remove the beneficial incentive effects of strict liability. As we argued above, strict liability has the advantage that it provides the injurer with incentives to take all efficient measures to reduce the risk (prevention and activity level), even if regulatory conditions were followed. This outcome has been shown formally by Kolstad/Ulen/Johnson482 and more recently by Burrows.483 They argue that the complete compliance defence prevents any precaution in excess of


189

6.


the regulated standard. If there is serious under-enforcement of standards, the role of liability as an incentive to take precautions remains important. A second reason is that exposure to liability might be a good remedy for the unavoidable capturing and public choice effects that play a role when permits are granted. If a permit always released from liability, all a plant operator would have to do would be to get a permit with easy conditions from a friendly civil servant. That would then exclude any law suit for damages from a potential victim. Obviously the capturing and public choice effects should be addressed also via direct tools. In this respect, the liability, even under criminal law, of the licensor is also an issue.484 Liability of the licensor (and appropriate sanctions within administrative law) can provide incentives to civil servants to act efficiently when granting licences.485 This, however, still requires tort law take into account the fact that regulatory standards are not always set efficiently. If the optimal level of care is higher than the regulatory standard, liability will efficiently provide additional incentives. Finally, tort law can also be seen as a “stop-gap” for situations not dealt with by statute.486 This makes clear that, exposure to liability notwithstanding, a permit is an important guarantee that the plant operator will take efficient care. Therefore, compliance with a licence or – more generally – regulatory standards, should not have a justificative effect in tort. The opposite would only be true if it were clear that the administrative agency took into account all the potential harm to all interested third parties when setting permit conditions. Indeed, theoretically regulators and licensors are supposed to set standards in regulations and permits in a way that reflects political choices about the level of risks that maximises welfare. Hence, ideally, when setting objective standards to control systemic risks, regulators are supposed to weigh costs and benefits of different norms and choose the standard that delivers the highest social net benefit. A judge in a civil liability suit should not be “second-guessing” efficient agency decisions. It is, however, rare that agencies will be able to take ex ante all these interests and possible damages into account when setting permit conditions. Hence, as a general rule, complying with licences or regulatory standards should not discharge from liability; the opposite would be the exception. This is the case both under a negligence as well as under a strict liability rule. Indeed, holding an injurer liable notwithstanding whether he followed regulatory standards will play an important role under a strict liability rule, since it will prompt the injurer to take efficient care and adopt an efficient activity level, i.e. to take all efficient measures to reduce the potential accident costs, although this might require the potential injurer to do more (as far as precautions are concerned) than the regulation requires. Under a negligence rule, this case law is also significant if the efficient care standard (which is assumed to be equal to the due care standard required by the legal system) is higher than the regulatory standard. The basic reason

190


6.


remains that efficient preventive measures can be taken above what is often prescribed in the norm. Requiring a potential injurer to take these efficient preventive measures thus increases social welfare.

C. Social security 1. Different starting points One could also imagine that a solution for potential uninsurability would be a further development of social security. With increasing damage from emerging systemic risks, it becomes clear that, although many of these risks may be insurable to some extent, there can be situations where the limits of insurability are reached. This may be the case because of a lack of predictability of risks or because of the capacity problem, especially with hazardous activities which may cause catastrophic risks. This means that there may be losses which cannot be recovered via liability law and liability insurance or via firstparty or other alternative financing techniques. It also raises the question of whether some of the losses that victims may suffer as a result of systemic risks should not be compensated via social security law. We have already discussed developments in social security and showed that social security and tort law are based on different principles, and that the expansion of exposure to liability (in Europe) was to a large extent due to the changes in the relationship between tort and social security. The question we would like to address now is whether social security could in principle replace liability law. We will use the example of health care to show that precisely because these systems have different underlying principles, it may not be that easy to argue that tort law could be replaced by social security.

1.1. Private insurance versus social security The cornerstone of social security is the idea of solidarity between citizens. All citizens contribute (in accordance with their income) to the social security system. Thus, the “working class” pays for those who, for a variety of reasons, cannot work (and hence cannot contribute to social security) but receive social security benefits. This again contrasts rather sharply with tort law, which does not invoke any principle of solidarity between injurers or victims. Only in liability insurance is it sometimes said that there is solidarity between all the insured. That, however, is a different type of solidarity than the one in social security. The solidarity referred to in insurance is only a system of risk-spreading whereby all the insured (via premium payments) contribute to the loss that occurs to one particular insured party. However, since the insurance premium is in principle risk-based, bad risks should pay more than good risks. Risk differentiation is therefore the leading principle in insurance, not the type of income- dependence solidarity that underpins social security.487 However, the differences are not always that sharp. For


191

6.


instance in the area of health care insurance premium is not always based on income, but may, at least particularly (and for some risks) be linked to preferences and the attitude and risk of the insured. Let us now look at how private (liability) insurance differs from social security as far as the financing of those systems is concerned. We will address this question by looking at basic differences between private insurance and social security as far as the financing of health care is concerned. In this respect we should add once more that although most of this analysis refers to one aspect of social security, the provision of healthcare, most of what follows also applies to other aspects of social security which are at least as important, such as the provision of replacement income in case of disability. Before addressing some of the features of private and social insurance in more detail, we shall briefly review the basic characteristics of both systems. Even this brief comparison will indicate how difficult it may be to apply principles of private insurance to social insurance schemes.

1.2. Risk premium versus income-dependent premiums Private insurance starts from the simple assumption that the demand for insurance arises from individuals who wish to protect themselves against certain risks. These risk-averse individuals demand insurance cover which is provided by insurers on competitive insurance markets. The latter aggregate similar but non-related risks into risk pools and are able to accept these risks because they can spread them over a large number of cases. To control moral hazard488 and adverse selection, the insurer distinguishes risks according to the individual risk posed by the particular insured party. Hence, a system of risk differentiation is applied whereby narrow risk pools are constructed in such a way that the premium charged corresponds with the risk posed by the average insured in that particular pool. An important feature of private insurance is thus that the price charged, i.e. the premium, corresponds to the risk. Again, there may be exceptions. In several OECD countries community rating in health insurance and other premium rules do not strictly follow this principle. The insurer often charges a higher premium than the actuarially price. The actuarially fair premium is the probability that the risk materializes multiplied with the magnitude of the damage once it occurs, hence that is the “value” of the risk. The premium the insured pays may, however, well be higher than his actuarially fair premium. This depends upon his market position and administrative costs.489 Depending upon the degree of risk aversion of the insured party and the premium charged, the individual demands insurance cover for specific risks to which he is averse. Social security in general differs from the situation sketched above in many respects.490 In many countries, health care is provided, at least as far as

192


6.


basic needs are concerned, either totally free of charge or with a modest deductible to almost all citizens. Legal systems may differ with respect to who can benefit from social security and with respect to the payment system. In some countries, health care is simply provided for free; in other countries, the patient must pay but can recover the expenses from his insurer or, in other cases, the (social) insurer pays the costs of the health care system directly.491 Within a social security system, equal access to health care to a smaller or larger extent is usually essential. In addition, the price of social security is usually charged by withholding the premium from the wages of all employed workers. Some of the premium is therefore usually paid by the employed individual himself; in addition, the employer usually contributes substantially as well.492 In addition to withholding the premium, it should also be mentioned that the amount of premium to be paid for social security is generally incomedependent. Thus, although there may be large differences between countries, some essential features of health care systems as provided through social insurance are common to them: contrary to the situation under private insurance, there is no risk differentiation but a principle of solidarity between good and bad risks.493 In addition, the premium is not dependent upon risk but upon income.

2. Moral hazard A central question, whether health care is provided through private insurance or via the government, is how an artificially high demand for health care can be avoided with insurance cover. Moral hazard is, as we have shown above, precisely the problem that the demand for a certain service (in this case health care) increases because the price for the service is reduced or eliminated. In a classic paper on moral hazard, Arrow pointed out that the demand for health care increases as soon as full insurance cover is available.494 Moral hazard does not occur only on the demand side. Given asymmetric information, the physician often determines what services and how much of them are needed. This may cause a moral hazard problem on the supply side as well. Here we focus on the victim demanding health care. The extent to which the moral hazard problem plays a role depends upon the elasticity of demand for the particular service. If demand is totally inelastic and hence not dependent upon the price level, there will be no moral hazard at all. If demand is, however, elastic, insurance cover will lead to a decrease in the marginal cost of medical care and hence a moral hazard emerges.495

2.1. Risk differentiation The traditional answer is that there are basically two possible remedies for moral hazard: either the behaviour of the insured party is controlled by a corresponding adaptation of the premium or by partially exposing the insured


193

6.


party to risk, for example, through a deductible or through an upper limit on cover.496 Considering the first potential remedy, which is advanced in traditional insurance economics, one might argue that there should be some link between the risk an insured party poses and the premium paid. In theory, this would mean that the more the health care system is used, the higher the premium should be. In insurance economics theory, such an individual would then be classified as belonging to a high-risk group. Through such an individualisation of the risk, the insured party would then behave as if he were not insured at all and moral hazard would be avoided. Observing the premium-setting practice of social security systems, it is immediately clear that the way premiums are charged in social security systems does usually not correspond to the answer of traditional insurance economics to the moral hazard problem. First, the premium charged will usually be income-dependent and not dependent upon the risk posed. Individual risk differentiation seems, therefore, impossible. Second, a high degree of risk differentiation seems to conflict with the idea of solidarity between good risks and bad risks which underlies many systems of social security.497 In practice, risk differentiation would indeed mean that high-risk individuals would have to pay more premium than low-risk individuals. From a policy perspective, the social security minister might not want to charge some individuals higher premium just because they had become ill. Moreover, the insurance economics’ idea of a risk-related premium seems to collide with the income-redistributive goals of social security.498 The question of course arises whether this conflict between the need for risk differentiation to control moral hazard, as argued by insurance economics, and solidarity between good risks and bad risks in social security, leads to insurmountable problems. The problem is obviously not that persons who get ill rely on the (private or social) health care system; the problem is more that some people might be inclined to visit their general practitioner too often or to purchase too much medicine if it can be obtained at no cost. To answer the question of whether there is a serious moral hazard problem, it is useful to distinguish between small and large risks. For large risks, such as surgery, the price elasticity of demand might be small. It is unlikely that anyone wants surgery just for fun; one can, therefore, that assume the number of operations would not increase drastically from the moment that the price fell.499 This might be different in the case of visits to a general practitioner or the purchase of medicine. For these kinds of health care services, the price might well have an influence on the demand for the particular service. A similar distinction between small and large risks can also be made as far as the moral hazard problem is concerned. No one will be inclined, for example, to have his leg amputated just because the operation is free. The fact that no distinction is made between low-risk and high-risk individuals does not, therefore, seem to

194


6.


be a problem as far as these large risks are concerned. The contrary might be true for small risks. For them it seems important to adapt the premium to the individual risk. Obviously we do realize that this is a very complicated issue, which is debated among health economists and policy experts. The basic idea is that exposing the consumer partially to risk (via e.g. a deductible) may have an influence on demand for some services, but not for others. This will, moreover, also depend upon – among others – the size of the copayment. Whether there is indeed a serious issue of moral hazard in health insurance is to a large extent also an empirical question, which we can not develop any further here.

2.2. Exposing the insured party to risk This brings us to a second potential method of controlling moral hazard – partially exposing the insured party to risk. Introducing an upper limit on cover does not seem to be an efficient remedy given the insolvency problem, which might lead to free-riding. Moral hazard can be controlled, however, more particularly as far as small risks are concerned, for example, through a deductible. The elasticity of demand for, say, a visit to a general practitioner, might well be rather large. If that were the case, moral hazard could be controlled through the introduction of a deductible.500 If one has to choose between the two ways of controlling moral hazard – adapting the premium to the individual behaviour of insured party or partially exposing the insured party to risk – one should be aware of the fact that partially exposing the insured party to risk, for example with a deductible, obviously involves less information costs than detailed control of the behaviour of the insured party. Detailed control of the behaviour of the insured party with corresponding risk differentiation could also conflict with the principle of solidarity between bad and good risks that underpins social security. Even for smaller risks it might be relatively harsh to charge, say, the elderly, who might need more medication and more visits to a general practitioner, a higher premium than is charged to healthier, younger individuals. Imposing a deductible on good risks and bad risks can deter some inefficient over-demand and at the same time avoid the harshest effects of strict differentiation between bad and good risks. Moreover, even with a social security system with premiums dependent on incomes, demand can be modified by imposing a deductible as far as the low risks are concerned. There is, therefore, an increasing trend in many Western European systems to introduce user charges into the social security system. If they are introduced for services for which the demand depends on the price, this trend can be supported from an economic point of view.501 Obviously, a deductible could be an appropriate way of to provide incentives against over-demand for health care services. 502 One should, however, be careful that it is only used in the case of risks where the high level


195

6.


of demand is an issue. Too high a deductible might have an adverse effect of over-deterrence, leading individuals who should see a general practitioner to the wrong decision in order to avoid a visit, and this might in the end lead to much higher costs (for example, hospital treatment) for the total health care system. Hence, one must be careful even about imposing a deductible to avoid over-deterrence.503

3. Health care insurance via the private market or government provision? 3.1. Private insurers versus bureaucracies A difference can obviously be made between social insurance, which in most Western European systems is provided by government intervention (the social law referred to in the project), and private insurance provided by insurance companies operating in a competitive market. Although the social security systems in Western Europe were very generous (largely for historical and political reasons), some systems seem once more to be making the traditional distinction, which we discussed before, between smaller risks and larger risks. For example, in Belgium, for self-employed workers, larger risks are covered via government-provided social insurance whereas smaller risks can be covered on the private insurance market. Moreover, this social insurance is generally compulsory whereas private insurance for smaller risks is optional. However, this need not necessarily be the case. In some countries, for instance in the Netherlands, low-income groups can purchase additional insurance for the smaller risks via (government financed) health care funds. The choice between social insurance and private (first-party) insurance is obviously closely linked to the choice between tort law and social law. On the choice between government-provided or private-market insurance, an economist would usually refer to the theory of bureaucracies to argue that government-owned operations that are not exposed to competitive pressures have little incentive for efficient production and might therefore be more expensive than private insurance companies providing the same product.504 One could therefore argue that insurance cover could be provided by the private market even if it were to be decided that the risks should be subject to compulsory insurance.505 There are indeed many examples of compulsory insurance where the cover is not provided through social insurance or government-owned institutions but through private insurance companies. This, however, assumes that private insurance markets are sufficiently competitive. In such markets, insurance theory predicts that a diversified supply of different insurance policies will be provided at competitive premiums. Only if the private insurance market is not capable of covering certain risks or provides inefficient results, for example because of a high degree of market concentration, is there an argument for government intervention.506 Another

196


6.


argument in favour of government intervention is that certain risks could be uninsurable on the private market, say, because of an irremediable problem of adverse selection. If the government could gather information on the risk at lower costs than insurance companies, this would be an argument for governmentprovided insurance.507 The foregoing shows that, in arguing that private insurers are in principle better than bureaucratic institutions at providing cover at competitive premiums, there is a strong case for private health care insurance even if the provision of insurance should, for reasons mentioned above, in some cases be made compulsory.508 However, the flip side of introducing compulsory insurance is that the policy-maker becomes dependent upon the private insurance market to insure risks. As long as the market is competitive and sufficient cover is provided, this should not in itself be a problem. However, some risks might be uninsurable on the private market or may only be insurable privately if certain regulatory safeguards are provided. Looking only at the moral hazard issue, one could, therefore, argue that they should not be insured at all, but in that case there could be undesirable results because these high risk individuals would still need and seek such care through any reliable public system or “safety net” coverage. Hence, even for compulsory insurance provided by the private market, there might be a case for the government providing insurance for those categories that could not obtain insurance cover on the regular market.509 But this obviously requires a certain regulatory structure. Actually, providing such cover for people who could not otherwise obtain insurance is not the same as providing a general social insurance scheme. In addition, even if the government intervenes to provide cover for this remaining category, risks should be differentiated as much as possible within this category for the reasons mentioned above. So far we have considered the choice between insurance via the private market or government provision, assuming that health care insurance can be provided via the private market, with risk-based premiums, taking into account general insurance principles, more specifically risk differentiation. From this point of view we have argued that government intervention is not necessary as long as insurance is provided in a competitive market and that moral hazard as well as adverse selection can be controlled. The problem is, once more, that social security policy apparently often aims at income redistribution as well. Therefore social security premiums are often not risk-based but income-dependent. This inevitably conflicts with the insurance principle that the premium should correspond to the real risk; the income redistributive goal can probably only be achieved through government intervention. If one disregards this redistributive goal, there are less arguments for health care insurance via the government than via the private market. From this it follows that the question of whether government intervention is necessary can only be answered when the aim of social security policy has been clarified. This choice inevitably incorporates a


197

6.


normative element. If income redistribution is considered to be the aim of social security, it is obviously not possible to accomplish risk differentiation at the same time.

3.2. Financing health care schemes A politically hot issue, which is closely connected to the choice between the private market and regulation,510 is obviously the way a health care scheme should be financed. Within a classic model of private insurance the answer is relatively simple. A risk-averse consumer will purchase (by choice or compulsorily) health care insurance on a competitive insurance market and pay a corresponding premium related to his demand for insurance and to the individual risk he poses. Traditional insurance theory holds that the premiums charged must correspond to the risk, not only to control moral hazard but also to guarantee that funds will be available when the insured risk manifests itself. The premiums paid will in principle be set aside to cover future risks. An insurer is under a contractual obligation to fulfil his contractual duties vis-à-vis the insured party. Hence, he will have to bear the consequences if bad risk calculations create trouble for him afterwards. A private insurer in principle does not have the possibility, say, of lowering the compensation due to the insured party or of increasing premiums for all existing insured parties.511 The latter, however, often happens for health insurance and other types of insurance. That may be the consequence of regulation on restrictions of competition. The “traditional” final consequence of bad calculations may be the traditional market sanction of bankruptcy. Cover under social insurance basically comes from a different angle. Since first being introduced by Bismarck in the late 19th century,512 social security systems assume that a minimum of social services, including health care, is delivered to those who could not afford to purchase insurance cover themselves. Again, this cover for low-income classes, which may be deemed necessary from a policy perspective, could be provided if the government, for example, provides the cover itself for this remaining category. However, in many cases the principle of solidarity between good risks and bad risks will lead to income-dependent premium-setting. Inevitably, this leads to the question of whether such income-dependent premium-setting can be reconciled with the insurance economics goal of avoiding moral hazard. As we indicated above, this may be problematic given the different starting points of the two systems. Unlike private insurers, the government providing the social insurance is not under a contractual duty to compensate the insured fully. However, there can still be statutory entitlements. These are legal obligations and in some countries (i.e. US) have been construed to be constitutional rights. Since the income of the social health care system is not risk-related (it may be age-

198


6.


related however), there is a chance that at some moment the amounts reserved may not be sufficient to pay for the health care scheme. In that case, no bankruptcy follows but the government may either lower the payments due (through regulatory intervention) or look for additional sources of funding (say, from general income tax revenues). This shows the basic difference between the financing of private and social insurance schemes. Laws can be changed as a result of political changes, but contracts should in principle be executed by the insurers. An important consequence is that the shift from social to private insurance would not necessarily be negative for the insured; the government can reduce the social insurance benefits because of budget constraints, whereas an insurer is bound to his contractual duties.513 Some scholars indeed argue that a shift to a privatised system has the advantage for the insured parties of reducing political risk. 514 However, this difference should not be exaggerated higher. The governments can not unexpectedly reduce insurance benefits. This, depending upon the statutory system, may still require a change of the statute, with the guarantees of democratic control. On paper, the execution of a contractual obligation by an insurer may be more certain than the fulfilment of a legal obligation by public authorities. In practice that may, however, not necessarily be the case.

4. Summary Comparison of the way social security and private insurance are financed highlighted a number of other different principles underlying private and social insurance. For example, social security systems seem to be relying on income-dependent premium setting, whereas private insurance in principle is based on a system where premiums are risk-related to provide a reserve for future claims. The simple reason is that, unlike in the case of social insurance, private insurers cannot rely on governments to cope with their deficits. If a private insurer did not relate premiums to risk, he could incur the market sanction of bankruptcy. This shows that not only do tort law and social security have different starting points, but that the same applies – not surprisingly – to the related insurance schemes: social insurance versus private insurance. Notwithstanding the differences between the two systems, there seem to be some similarities as well, or at least the systems seem to approach each other. One example is the idea of equal access to health care services that underpins many social security systems. This idea is obviously not carried through all the way. There certainly is equal access for basic medical needs in most legal systems, but in many countries individuals can purchase additional coverage according to their own demand and pay a supplementary (risk-related) premium. A debate is also going on in private insurance today about whether the principles of risk differentiation can be carried through in all respects. Cousy has often advocated


199

6.


that insurance companies should take ethical principles into account as well, which may lead them to accept some solidarity between good and bad risks. These ethical principles in other words impose certain limits on very detailed risk differentiation.515 Wils has also pointed out that detailed risk differentiation may conflict with basic human rights if, for example, risks are differentiated according to gender or sexual preferences.516 A basic difference between social and private insurance is obviously that the policy-maker wants to achieve income redistributive goals through the financing of the health care system. Obviously, however, this may have a price as far as the efficiency of the system is concerned. Today many also advocate a combination of limited social insurance schemes providing basic medical treatment and financed by income dependent premiums, with the possibility of purchasing additional insurance on the market according to the preferences and demands of the particular individual.517 Hence, it is not always necessary to make a strict choice between private or social (health care) insurance, but a combination of systems is possible. One could obviously conceive of a compulsory social insurance scheme for the larger risks where model hazard does not play an important role, combined with private insurance for the smaller risks. It has been suggested that the current crisis in tort liability should be remedied by tackling the source of all the problems, that being the deregulation and privatisation of social security. The remedy seems straightforward if one believes that it was mainly the result of changes in social security that the scope of liability was expanded. However, expanding liability obviously has more sources than merely changes in social security. Moreover, it seems politically unfeasible at a time when Western European governments believe more in liberalisation and privatisation than in regulation, to argue in favour of expanding social security schemes. Moreover, many criticisms could also be levelled at an expansion of social security. In fact, the introduction of privatisation and market forces in social security was obviously not a bad idea per se. It is only surprising that governments did apparently not realise that the privatisation of social security would inevitably put more pressure on the compensatory function of tort law. It is, however, important to remember that the ideal solution to the increasing damage resulting from systemic risks probably does not lie in a return to social security schemes which to some extent also malfunctioned in the past. It is probably better first to make optimal use of financial and insurance markets and to develop the possibilities of alternative compensation mechanisms and new financing techniques outlined in Chapter 5.

200


6.


D. Compensation funds 1. Introduction A question that inevitably arises when the issue of damage caused as a result of systemic risks is discussed is whether traditional liability law combined with insurance is at all able to provide compensation for such damage. Issues of fault or negligence have been avoided in many legal systems by a trend towards strict liability for many systemic risks. However, we have already argued that strict liability alone will of course not guarantee compensation, given the insolvency problem. We have also indicated that many problems may arise with the insurance of systemic risks. Therefore, in many legal systems the question has been asked whether compensation for damage caused by systemic risk should be provided through compensation funds. A well-known example in this respect is of course the United States Superfund, introduced under CERCLA, which has prompted a lot of criticism.518 Calls are also being heard in several European legal systems for the creation compensation funds to cover particular systemic risks. In the Netherlands this has been proposed for environmental damage519. In addition, the idea is now increasingly being floated in policy documents in the various member States formulating proposals for the reform of environmental law. In this connection, we can once more point to the InterUniversity Commission for the Reform of Environmental Law that recently proposed the introduction of a compensation fund in the Flemish Region,520 but also to a Dutch study performed on behalf of the Ministry of the Environment, which proposed the introduction of an environmental compensation fund in the Netherlands.521 But the importance of compensation funds is discussed not only in the context of environmental damage but also with respect to many other systemic risks, such as occupational health hazards caused by exposure to asbestos. In many countries, this has led to calls for compensation funds to compensate victims of asbestosis, especially in those cases where the liable employer had gone out of business or was found to be insolvent. Therefore, there are ample reasons to take a closer look at compensation funds and to examine the role they could play in compensating damage caused by systemic risks. It is especially important to compare the functions of compensation funds with those of traditional insurance.

2. The various funds One cannot escape the impression that often – especially at the political level – funds are advocated as a miracle solution to all systemic risks, but without any clear definition being given of the specific funds. This can be misleading since the term “fund” is often used for a variety of private or public financial arrangements, some of which have been discussed in this chapter – that may be quite different. We will briefly sketch the kind of funds that might


201

6.


play a role in compensating environmental damage. This short overview will make clear that these various funds usually have very different goals.

2.1. Limitation fund In the US literature, the idea of a fund is sometimes used to refer to the situation whereby mass damage is caused by similar products or services. The problem that usually arises in these cases of serial damage (e.g. with toxic torts) is that the liable enterprise may be willing to agree to a settlement with the victims on the condition that it offers a certain sum to all the victims in return for a final settlement for the damage caused by the specific tort. In a certain sense the manufacturer then raises a fund that will have to be used to compensate all the victims. It could be called a limitation fund since the enterprise agreeing to such a settlement usually wishes to limit its liability to the amount in the fund. No risk-spreading with other (potential) manufacturers takes place since only the liable manufacturer finances the fund. Examples of such a limitation fund can be found in the area of civil liability for oil pollution damage. Article V of the Convention of Brussels on Civil Liability for Oil Pollution Damage provides for the possibility for the liable tanker owner to provide a limitation fund to compensate victims.522 In return for the payment of a particular sum, the duty to compensate is usually limited to the amount made available, i.e. a limitation fund. Recently it has also been argued in the Netherlands that such a fund would make compensation for health damage caused by exposure to toxic substances a lot easier.523 A problem that will, however, inevitably arise with environmental damage is latency. It is essential in a limitation fund solution that latecomers be excluded. Therefore this might not be the ideal solution in the case of environmental damage. The goal of a limitation fund in these cases is not so much to provide a remedy against perpetrators who vanish, for example, but to have an adequate instrument for dividing the available proceeds among the victims in the case of serial damage.524 There could however be one role for such a limitation fund. If one fears that a certain activity may cause harm in the future, at the start of the operation the enterprise involved in the activity could be asked to pay a substantial amount ex ante which would become available if damage resulting from systemic risks occurs. If the damage occurs, the liability would then be limited to the amount the injurer had to pay. As we have just explained, this regime has been proposed in the Flemish Decree on Environmental Policy.525 The advantage of such a regime for the enterprise is obvious. It will have certainty with respect to the amount to be paid, which is never the case when it is subjected to full unlimited liability. In addition, the amount reserved for compensating environmental damage (through either insurance, a payment in cash, a bank guarantee or any other financial guarantee to be approved by the agency) can be paid back after the operations if no environmental damage

202


6.


occurs. Hence, the money is not lost as is the case with insurance premiums. From society’s point of view, the advantage is that there is at least the certainty that an amount of money deemed adequate to compensate damage will be available if an accident happens. This system would thus seem useful for other systemic risks as well.

2.2. Advance fund A second arrangement referred to is an advance fund. Prepayment or advance of a certain sum is a remedy for drawn-out civil procedures concerning liability and insurance coverage issues, that can last much longer than the life of the victim. Especially in the case of asbestos victims, it has been argued that it is highly unfair that (relatives of) victims only receive compensation post mortem because of the relatively short time between the discovery of the illness and their death. Therefore an advance fund has been proposed in the Netherlands as a remedy for asbestos victims.526 Examples of these advance funds can for instance be found in environmental legislation. In Belgium the Act of 10 January 1977 set up an advance fund for damage caused by groundwater extraction.527 Again it is uncertain whether this fund can play a large role in cases of systemic risks. The reason why compensation funds are sometimes defended with respect to systemic risks is not so much related to protracted legal procedures but more to the uninsurability of certain systemic risks or to the fact that no individual injurer can be identified. If these are the problems to be solved, an advance fund is not an adequate remedy.

2.3. Guarantee Fund Guarantee funds are well-known instruments for protecting victims against the possible insolvency of a liable injurer or his insurer. The advantage of a guarantee fund is that it only intervenes for the so-called excess risk, that being the risk for which, in the specific case, no insurance coverage is available. However, it is essential that a guarantee fund intervenes only if other compensation mechanisms such as insurance have failed. The possible role of a guarantee fund in compensating damage caused by systemic risk will be discussed below.

2.4. A general compensation fund A fourth alternative could be a compensation fund that would generally operate as a substitute for liability and insurance. Although it is often not very specified, one has the impression that it is usually in this sense that the term compensation fund is used.528 However, if such funds were to be used as alternative to the liability system, inevitably the question will arise of how they can be financed in an adequate way. Therefore, we shall now focus on basic differences between compensation funds on the one hand, and liability


203

6.


combined with insurance on the other, as devices for providing compensation for damage.

3. General principles of fair and efficient compensation No matter how a compensation system is organised, the incentives for preventing damage caused by systemic risks should always remain intact. We have already referred to these principles when discussing the question of whether liability should be retroactive. They are, however, important as well when it comes to judging the effectiveness of alternative compensation mechanisms. In Chapter 2 it was often stressed that liability rules have a preventive as well as a compensating role. Liability rules can only have a preventive effect if the onus to compensate is put on the party which actually contributed to the risk. This means that the duty to compensate should in principle only rest upon the party which actually contributed to the risk. A second, related principle is that this duty to contribute should also be related to the amount in which the specific activity or entrepreneur contributed to the risk. This principle is usually automatically respected in liability law. The duty to compensate under tort law is indeed usually limited to the damage that the specific tortfeasor caused himself.529 However, if a collectivisation of the compensation also takes place, it is important that the tortfeasor only contributes financially in proportion to the amount in which he contributed to the risk. This is reflected in insurance policies in the well-known idea of risk differentiation which we have discussed extensively. It simply means that bad risks pay a higher premium than good risks. This principle should also be applied if a compensation fund is set up, meaning that bad risks should contribute more to the compensation system than good risks. This remains important since it will give incentives for prevention to the contributors to the fund. Bad risks will be punished and good risks will be rewarded. These principles are not only important from an efficiency point of view (providing optimal incentives for prevention), but also include a fairness element. Indeed, if these principles were not followed, it would mean that good risks would have to pay for the bad risks as well and would therefore in fact subsidise bad risks. This negative redistribution should be avoided and therefore the compensation mechanism, fund or insurance, should be financed principally by the parties that actually contributed to the damage. In sum, the compensation mechanism should aim at a differentiation of the contributions due. This differentiation is only possible if the insurance company or agency administering the fund also possesses information on the amount in which the specific activity contributed to the risk. One key element that should determine the choice between insurance or funds is therefore who possesses the best information to control the risk.

204


6.


4. Funds versus insurance 4.1. Risk differentiation Applying the principles discussed above, there are not many reasons why, if both are – in theory – available, a compensation fund would provide better protection against insolvency than private insurance markets. One can assume that an insurer is better able to differentiate risks since it is specialised in risk differentiation and risk-spreading. Insurers therefore possess techniques for determining in what way insured parties contribute to the risk. Obviously this assumes that insurance markets are competitive. In the absence of competition on insurance markets, either the supply of insurance coverage could be too limited or premiums could be excessively high, which could justify a preference for a compensation fund.530 But if insurance markets are competitive, insurers can be assumed to be better able to deal with classic insurance problems such as moral hazard and adverse selection than the administrators of a compensation fund. One cannot see as a matter of principle why a government agency running a compensation fund would have better information on risks than an insurer. This might, however, be different if highly technical systemic risks are involved where operators of certain facilities are in a much better position than the insurance company to monitor each other. Some examples have been given above. This point has been made for instance concerning the compensation nuclear damage. One could argue that a risk-sharing agreement between nuclear plant operators could lead to optimal monitoring between the operators since they would possess much better information on prevention, good and bad risks than an insurance company would531. In maritime insurance too, the Protection and Indemnity Clubs already discussed, which are based on mutual risk-sharing between tanker owners, play a crucial role.532 With respect to these highly specialised matters, one could therefore argue that the operators themselves might in some cases be better equipped than an insurance company to control moral hazard since they are better able to process information on the particular risk. However, the examples given show that these risk-sharing agreements do not involve the use of a government-run compensation fund. In sum, if both insurance and compensation funds are available, there are no clear reasons why a fund would be the preferred solution. There may, however, be reasons why insurance may not provide coverage for certain risks. In that case, funds cannot be compared with insurance since insurance is not an alternative. However, as we have shown in Chapter 3, one should not conclude too quickly that systemic risks are uninsurable.533

4.2. Costs Comparing insurance with compensation funds one should also address the comparative costs of both instruments. Insurance will generally be cheaper because liability insurance policies are not concluded for one activity, but for a


205

6.


whole set of risks. There is hence one insurance policy with transaction costs that are incurred once and an administrative structure within an insurance company that will be forced to an adequate cost reduction by competitive pressures. The costs of risk spreading might also be lower with an insurance company than with a compensation fund. Insurers are specialised in acquiring information on differentiation of risks. In addition, it has been argued in the literature that insurance make possible a reduction in transaction costs between contracting parties because they can ex ante agree on a distribution of risks and losses in the event of an incident.534 The comparison will obviously also depend upon the type of compensation fund under discussion. In most cases one immediately thinks of a compensation fund run by a regulatory authority. But then one can of course refer to the literature on the negative effects of bureaucracies to argue that such a publicly-operated compensation fund would not necessarily provide compensation at lower costs than the private insurance market. These can be reduced if the fund is administered privately, but in that case competition with other funds has to be organised in order to provide incentives for cost reduction. Such a fund would only have cost advantages if it did not replace the liability and insurance system but instead were a limitation fund such as that proposed by the Inter-University Commission, whereby potential injurers make an ex ante reserve for potential future losses.

5. A compensation fund for damage caused by systemic risks? So far we have argued that, generally speaking, there are very few reasons to expect that a compensation fund would provide better compensation than a private insurance market. This is especially the case if one considers situations where the injurer can be identified and can be held liable. Fund solutions for damage caused by systemic risks may play a role in the situation where the injurer cannot be identified or is insolvent.535

5.1. Guarantee Fund A compensation fund may well be used to guarantee compensation in the event of insolvency of the injurer or his insurer. The fund does not replace the liability and insurance system but only intervenes in a particular case when the injurer or his insurer turns out to be insolvent. The combined use of the liability system, insurance and a guarantee fund for the insolvency risk, has the advantage that the incentives of the liability system will remain intact and that the fund will only have to intervene in the event of insolvency.536 However, the question arises whether such a fund could be used separately from compulsory insurance (in the broadest sense). Indeed, if insolvency of the injurer is the problem one fears, it seems more logical to discuss the introduction of a duty to insure instead of immediately advancing a fund-solution. If one looks at the use of guarantee funds now, one may note that they are usually used in

206


6.


cases where an insurance company goes bankrupt. This raises the question of whether it is at all useful to impose a duty to insure for particular systemic risks. This will be discussed in more detail in the next section. If we take the example of one particular systemic risk, environmental damage, we may note that so far only a few countries have introduced compulsory environmental liability insurance.537 However, it is important to stress that if one wishes to provide protection against the insolvency of the injurer, it is more appropriate to focus on the variety of financial mechanisms which we discussed in the previous chapter, that would force a potential injurer ex ante to provide compensation for future losses instead of focusing on an ex post fund solution. Generally speaking, one can doubt whether it is useful to focus on a guarantee fund that would intervene, for example, in the event of the insolvency of an insurer instead of focusing on other ex ante compensation mechanisms.

5.2. Restoration fund There is one other situation where insurance cannot intervene and a compensation fund could prove useful. This is the situation where no individual enterprise can be identified as liable for the harm caused.538 Since there is no individual injurer who is liable, there is logically also no insurer who will be bound to compensate. One can think of a deterioration of a specific habitat through acid rain. These are typically cases where liability is not a suitable instrument. One could consider a compensation fund for these specific cases where no individual injurer can be found. However, the compensation fund should then be limited to those situations, so that liability rules and insurance can still exercise their preventive effects in all the other cases where an injurer can be found. But the question of who should contribute to such a fund will still have to be answered. Ideally, it this would be funded by all contributors to the risk. However, unlike in the case of traffic accidents, for example, it is not easy to identify ex ante what kind of activities contributed to the loss. If it is generally possible to argue that, for example, sulphur dioxide emissions caused the loss, economists would obviously advocate a Pigovian tax on sulphur dioxide emissions to finance the fund. One reason, however, why environmental taxes are still relatively scarce in practice is that it will often be hard to determine in what proportion the specific activity contributed to the risk and hence to determine the optimal marginal tax rate. If this is not possible, the only alternative is that the government will pay for the restoration of the specific site, either directly or by financing a compensation fund that has to finance the clean-up activity. However, since in both cases it is the tax payer who will have to provide the financial means, one can question whether a compensation fund has any added value, apart from the fact that one is certain that specific public funds will be reserved for compensation of the particular damage.


207

6.


6. Financing of historical damage We have already addressed the issue of retroactivity and all the problems caused by so-called long-tail risks (where a long time elapses between the wrongful act and the damage). One of the consequences of this long-tail character of systemic risks is that society can be confronted with situations where an insurer denies coverage and hence the (probably insolvent) injurer finds himself without insurance or cannot be identified at all any more. It seems important, in the context of a discussion on compensation funds, to discuss, from an equity perspective, who could be made liable for financing such “sins” of the past.

6.1. Insurers? The problem is that these issues are often discussed in an ex post manner, i.e. when the problem arises today. Usually sins of the past are discovered, e.g. a toxic waste dump that was filled 20 years ago, and the bottom-line is that everyone worries, especially the public authorities, because no one can be found to pick up the bill for the clean-up costs. This will also put immense public pressure on the judiciary to find a liable party and an insurer to pay for these harmful consequences of past behaviour. We have already argued that such an approach is neither efficient nor fair. The question, however, arises as to what an insurer should do when he is nevertheless confronted with such a bill in the form of retroactive liability and whether there are any alternative financing mechanisms that can avoid the negative consequences of retroactive liability we discussed. When an insurer today is confronted with a duty to compensate for losses which he has not foreseen, he is obviously confronted with the question of how funds can be generated to meet this obligation. The only practical solution is to generate them through an increase of premiums for the insured of today. Such a premium increase leads to negative distributional consequences since the insured today will have to pay for the sins of industry in the past, while industry today is maybe following the new stringent standard of care and taking preventive measures. Hence, such redistribution runs counter to our principles of efficient and fair compensation. In addition, it should be mentioned that such a premium increase will only be possible on the assumption that all insurers in the market are hit by the retroactivity problem in the same manner. If this were not the case, then the unlucky insurer who, by hypothesis, is the only one confronted with many retroactive claims, would not even be able to raise premiums ex post because of competitive pressures. The ultimate consequence might then even be the bankruptcy of the particular insurer.

208


6.


6.2. Compensation funds? Assuming therefore that it is not efficient and fair to confront the insurers of today with the problem of financing the sins of the past, the question arises whether there are any realistic alternatives. An alternative often heard, especially at the policy level, is precisely the use of compensation funds discussed in this section. The argument goes that compensation funds should intervene if certain risks seem uninsurable, which is the case for example, with historic pollution. We have just argued however, that it seems that compensation funds would not be more able than insurers to cope with these problems. The crucial question is of course who has to contribute to the fund. At that moment, the same question will arise as with liability and insurance. The first and obvious answer is that the industry that caused the risk should contribute to the fund. But if these injurers are identifiable there is usually no need for a fund solution. It is precisely because tortfeasors of the past can no longer be found that arguments in favour of funds are often put forward. Second, often it is argued that all manufacturers still in business and their insurers should contribute to a fund. But then this means that the blameless people of today would be paying for the mistakes of the culpable parties of the past. This can never have a positive influence on incentives for future behaviour and amounts once more to negative redistribution.

6.3. Government? The third solution inevitably put forward is government intervention. It does seems hardly fair for instance to place the cost of old soil pollution caused in the past on the shoulders of the unlucky ones who can still be found today, and to make them pay a collective debt. Since the problem is a collective one, it seems in the end unavoidable that the collectivity of today’s tax payers should pay for the collective failure in the past.539 Thus a kind of intergenerational equity is restored whereby the present generation pays for the mistakes of the past. Obviously, one can then still discuss the details of government intervention. In fact, it makes little difference whether the government pays, for instance, for a fund to compensate victims of historic pollution or whether it pays the costs of soil clean-up directly.

7. Summary Many arrangements that include a private or public compensation scheme for damage caused by systemic risks are referred to as “fund” solutions. We took a brief look at some of them, comparing them with traditional liability and insurance. The arrangement most discussed is a compensation fund that would totally replace traditional insurance. When an injurer can be identified, can be held liable and is solvent, we doubt whether a compensation fund can play the important role it is expected to do. Generally speaking, insurance seems better able to control risks and can


209

6.


be provided at lower costs. In addition, some problems such as long-tails risk can be resolved through adequate insurance policies. It seems therefore more appropriate to use traditional liability and insurance as far as possible and to use funds only in cases where insurance markets fail and there is reason to believe that funds would be able to provide adequate compensation. In that respect, a guarantee fund seems appropriate to provide coverage if an insurance company were to go out of business. However, it does not seem useful to introduce a guarantee fund without a corresponding duty to insure. It might be more interesting to focus on various other legal instruments that aim to provide protection against the insolvency of the injurer, which were also discussed in this report. Also, the Convention of Lugano on civil liability for damage resulting from activities dangerous to the environment provides for such a flexibility by stating in Article 12 that operators are required to participate in a financial security scheme or to maintain a financial guarantee up to a certain limit. This does not necessarily have to be insurance. More flexible than a duty to insure is the proposal by the Flemish Commission, discussed previously, to force a potential injurer ex ante to provide any kind of compensation, bank guarantee, or other kind of payment ex ante as guarantee of compensation in the event damage occurs as a result of systemic risks. The advantage of this solution is that it does not replace liability and insurance, nor does it force the potential injurer to take out insurance coverage. It provides for a flexible system whereby the potential polluter can choose any market solution provided that it will adequately compensate damage ex post. Still this may not solve the problem of damage caused by systemic risks in cases with a variety of causes where no individual injurer can be identified. For these kinds of situations a fund might be warranted, although it is still unclear how it should be financed. If the injurers can be adequately identified as well as the amount in which they contributed to the risk, the fund could be financed through a tax on risky activities. Otherwise, the general tax system should intervene, but then the need for a fund disappears as well. Generally speaking, it seems important to analyse carefully the joint use of liability rules, insurance and funds.540 This may lead to a combined use of these systems whereby the incentives provided through liability rules can remain unaffected and a fund might provide a useful contribution for the specific case in which the damage caused by systemic risks cannot be tracked down to one particular injurer.

210


6.


E. Compulsory insurance? 1. Introduction We have to consider carefully the risk that an injurer would simply chose a “hit-and-run” strategy whereby he would cause risk, irrespective of the potential liability, since the liability rule could not affect him, given his insolvency. In legal doctrine, compulsory liability insurance is often advocated as a means of protecting the innocent victim.340 The duty of an injurer to purchase liability insurance would be a good way to protect victims against the insolvency of the injurer.341 Economists on the other hand, see, as we mentioned above, different benefits of liability insurance. For them, it is an instrument to increase the utility of a risk-averse injurer. Whereas economists would therefore see liability insurance mainly as an instrument serving the interests of the injurer, lawyers tend to rely more on the victim protection argument. Of course these views do not necessarily contradict each other. Sometimes they may both imply an obligation to introduce liability insurance. Let us now address the question of the circumstances under which the purchase of insurance should be made compulsory.

2. Economic arguments 2.1. Increasing expected utility A first way to look at this question is to go back to the basic utilitarian literature on the benefits of insurance.543 If insurance is beneficial since it removes risk from risk-averse persons and thus increases their utility, are not these benefits so large that they warrant the introduction of compulsory liability insurance? A problem with this argument is that the degree of risk aversion varies. A Rockefeller will probably not be averse towards the risk of losing US$ 1000, but a low-income family father probably will be. The lowincome father will therefore probably have a demand for insurance against this risk of a loss of US$ 1000, whereas a Rockefeller probably would not. This straightforward example makes clear that the introduction of a duty to insure might be inefficient insofar as it forces some people to purchase liability insurance that would normally not have a demand for it. Insurance does not increase their expected utility. A generalised duty to insure might therefore create a social loss. Whether this is the case will of course depend upon the number of people that are actually harmed by the introduction of a duty to insure. This might indeed be outweighed by the fact that others will certainly derive a benefit from insurance. This, however, does in itself not justify the need to introduce a duty to insure. Risk-averse individuals might have purchased insurance coverage anyway. This means that the simple fact that insurance increases utility cannot justify per se the introduction of a duty to insure as


211

6.


long as we assume that all individuals are perfectly informed about the risk to which they are exposed and the availability of insurance. This argument also rather paternalistically assumes that insurance is under all circumstances beneficial to potentially responsible parties. The argument neglects the fact that, as we have explained above when discussing the basic theory of liability insurance, the insured has to pay a price to have the risk removed from him. This price will unavoidably be a lot higher than the actuarially-fair premium, consisting in the multiplier of the probability x damage (pxD). Depending up on the administrative efficiency of the insurance undertaking, a high or low amount of loading costs will be added as well as a profit margin depending upon the degree of concentration on the market. In addition taxes may increase the premium as well. Moreover, we have indicated that if there is uncertainty concerning the risk, the insurer might compensate this insurer’s ambiguity with a risk premium. In sum, the premium charged may well be a lot higher then the actuarially-fair premium which is the objective value of the risk. For some responsible parties, this premium will still be attractive but for others it may not be so. Moreover, some may choose a balanced approach, e.g. self-insuring a basic expected damage and only taking insurance for the excess loss. Compulsory insurance generally neglects the fact that the demand for insurance may vary according to the individual risk situation (and financial possibility) of every responsible party. Other things being equal, there is therefore no reason for regulatory intervention based solely on the fact that insurance may increase expected utility. Regulatory intervention would have the disadvantage that it is counter to the more differentiated demand of the responsible parties. However, this assumes that the responsible parties have knowledge of their exposure to risk and the availability of insurance and make a well-informed decision accordingly. If this assumption is not met, the question could again be raised of whether insurance should be made compulsory.

2.2. Information problems Information problems might arise if the potential injurer cannot make an accurate assessment of the risk he is exposed to and the benefits of the purchase of insurance. An underestimation of the risk would in that case lead to the wrongful decision of the injurer not to purchase liability insurance. The legislator could remedy this information problem by introducing a general duty to insure. This information problem is probably a valid argument for introducing a generalised duty to insure for motor vehicle owners. Maybe the average driver of a car underestimates the benefits of insurance. If there were no information problem and the legislator nevertheless introduced a duty to insure because this would be “in the best interest” of the insured, this would of course be mere paternalism.

212


6.


If empirical evidence existed that most injurers greatly underestimate the costs of damage that could arise from systemic risks they may cause, and the probability that they will be held liable for such damage, injurers would not constitute adequate reserves to cover their potential liability. If these conditions are met and one can assume that injurers underestimate the cost of environmental damage, this information deficiency may be considered an argument in favour of compulsory insurance. But again, the policy argument based on information asymmetry relates merely to the fact that the injurer would underestimate the potential benefits of insurance. There may, however, be another argument why the (uninformed) decision of a polluter not to insure may lead to underdeterrence. This policy argument is precisely related to the insolvency risk mentioned in the introduction.

2.3. Insolvency Another reason for introducing compulsory insurance is the argument often used by lawyers – insolvency. The argument goes that the magnitude of the harm will often exceed the individual wealth of an injurer, as a result of which a problem of undercompensation of victims will arise. Lawyers would therefore push compulsory insurance as an argument for guaranteeing effective compensation to the victim. This – more distributional – argument obviously may play a role in the context of insurance of particular systemic risks as well. We have already pointed to the insolvency problem. To take the example of environmental pollution: if an injurer were found to be judgement proof and a polluted site, for example, were thus “orphaned”, the costs would be borne by society. It is, however, also possible to make an economic argument that insolvency will lead to underdeterrence problems which might be remedied through insurance. This so-called “judgement proof” problem has been extensively dealt with in the economic literature.544 Insolvency may, however, pose a problem of underdeterrence. If the expected damage largely exceeds the injurer’s assets, the injurer will only have incentives to purchase insurance up to the amount of his own assets. He is only exposed to the risk of losing his own assets in a liability suit. The judgement proof problem may therefore lead to underinsurance and thus to underdeterrence. Jost has rightly pointed out that, in these circumstances of insolvency, compulsory insurance might provide an optimal outcome.545 By introducing a duty to purchase insurance coverage for the amount of the expected loss, better results will be obtained than with insolvency whereby the magnitude of the loss exceeds the injurer’s assets.546 In the latter case, the injurer will only consider the risk as one where he could at most lose his own assets, and will set his standard of care accordingly. When he is under a duty to insure and exposed to full liability, the insurer will obviously have incentives to


213

6.


control his behaviour. Via the traditional instruments for the control of moral hazard discussed above, the insurer can make sure that the injurer will take the necessary care to avoid an accident with the real magnitude of the loss. Thus Jost and Skogh argue that compulsory insurance can, provided that the moral hazard problem can be remedied adequately, yield better results than under the judgement proof scenario. This is probably one of the explanations why traffic liability compulsory insurance for instance was introduced. Uninsured and insolvent drivers who have little money at stake which they may lose compared with the possible magnitude of accidents they may cause, may have little incentives to avoid an accident. Insurers might be better able to control this risk and could force the injurer to take care under the threat of refusing him insurance. Thus the insurer becomes obliged to insure the licensor of the activity. Indeed, this economic argument shows that insolvency may cause injurers to externalise harm: they may be engaged in activities which may cause harm and which largely exceed their assets. Without financial provisions, these costs would be transferred to society and would hence be externalised instead of internalised. Such internalisation can be achieved if the insurer is able to control the behaviour of the insured. As we have shown above when discussing how risk differentiation can be applied to systemic risks, the insurer could set appropriate policy conditions and an adequate premium. This shows that if the moral hazard problem can be dealt with adequately, insurance even leads to a higher deterrence than a situation without liability insurance and with insolvency. Of course, this argument in favour of compulsory insurance relies on a few assumptions and conditions which will be discussed in further detail below. One is obviously that the argument is only valid if moral hazard can be controlled adequately and insurers also have appropriate incentives to do so. Another condition is that insurance markets should be competitive. But from a legal and economic point of view, the potential insolvency of the injurer is a problem since it can lead to both underdeterrence and undercompensation. Compulsory insurance may remedy both problems since it may provide adequate victim compensation and – if certain conditions are met – remedy the risk of underdeterrence. An issue which merits attention is whether compulsory liability insurance is the best instrument to remedy the insolvency problem. Indeed, although one usually refers to liability insurance in this context, several alternatives may exist that might be able to remedy the insolvency problem at lower costs. A first obvious alternative would be to turn to (compulsory?) first-party insurance instead of third-party liability insurance, as we have argued above.

214


6.


In sum, one could argue that compulsory insurance might theoretically be an effective way of remedying the insolvency problem. However, it does not necessarily have to be liability insurance but may well be the first-party or direct insurance discussed in this paper.

3. Potential dangers of compulsory insurance 3.1. Moral hazard After having discussed these three basic criteria for compulsory insurance, two other points must be discussed. Firstly, one should remember that with insurance there will always be the moral hazard problem, as discussed above. This means that even if a legislator decides to introduce compulsory insurance, he should not restrain the ability of an insurer to control moral hazard. Otherwise compulsory insurance will create more problems than it solves. Nevertheless, there seem to be problems since the legislator often limits the possibilities to expose the insured to risk. Indeed, with compulsory insurance the duty to insure is often equal to the total amount of liability and deductibles are not allowed. Hence, the total risk is shifted to the insurer, which means that the only way the insurer can deal with the moral hazard problem is to monitor the insured. If this seemed difficult or very costly, the introduction of compulsory liability insurance might indeed create problems. Shavell even goes as far as to state that, if moral hazard cannot be controlled, the only regulatory intervention with respect to insurance should be a prohibition of liability insurance.547 In any case, an introduction of compulsory insurance does seem problematic if moral hazard cannot be controlled adequately.

3.2. Concentration on insurance Markets A second, related, issue is that until now we assumed that insurance markets are perfectly competitive and that thus premiums and policy conditions will be nicely tailored to the individual needs and the behaviour of the insured in order to control moral hazard optimally. In practice, however, many restrictions on insurance markets exist. We will give examples below. If monopolistic premiums can be set, an insurer will have fewer incentives to align his premiums on the individual behaviour of the insured and thus there is less control of moral hazard. From a policy viewpoint, it also seems highly problematic to make liability insurance compulsory on concentrated insurance markets. Indeed, in that case the inefficiencies on the insurance market would be reinforced by making the purchase of insurance compulsory. Here the interest-group theory of government can of course explain why insurers might want to lobby in favour of compulsory liability insurance. If they already can determine the supply-side of the market through monopolistic premium-setting, all they should strive for is to ensure that


215

6.


every possible injurer is forced to purchase insurance coverage. Through this regulatory intervention, a certain demand is then guaranteed as well. Nevertheless, in practice insurers are never enthusiastic about compulsory insurance, at least for systemic risks. Cousy claims that this is related to the fact that, as a matter of law, under compulsory insurance the insurer can often not invoke defences against the third-party beneficiary of insurance. Moreover there would be problems related to the implementation and actual carrying-out of the obligation to insure.548

4. Further warnings 4.1. Dependence upon the insurance market The qualifications and potential dangers that have just been outlined show that although theoretically a compulsory insurance mechanism may be desirable, and for systemic risks more particularly in order to remedy information problems and the underdeterrence risk as a result of insolvency, in practice one should be cautious about advising the introduction of compulsory insurance for specific systemic risks. There are some other reasons for such caution. One is that the legislator should be aware of the fact that, as soon as it introduces compulsory insurance, it depends on insurers to fulfil the duty to insure. The practical possibilities of an effective enforcement of that duty will obviously to a large extent depend upon the willingness to insure on that particular market. It will ultimately be the insurance market which decides whether they are willing to cover a certain risk. This could result in the undesirable situation whereby the legislator introduces a duty to take out compulsory insurance but that the market refuses to provide such coverage. Introducing a duty to insure makes the policy-maker highly dependent upon the insurance market. This seems to have led to problems with the German Environmental Liability Act of 1990 (Umwelthaftungsgesetz), which requires the owner of an installation that can cause significant damage to take out liability insurance or to have sufficient financial guarantees549. It should be realised that if one makes the availability of insurance coverage a prerequisite for the operation of an enterprise, insurance undertakings become the de facto licensors of the industry, which may be questionable from a policy point of view.550 For instance in the environmental case, the insurer becomes the "environmental policeman".551 If that means, however, that the insurer, as a “policeman” controls the environmental performance of the insured company, there is of course nothing wrong with that. It may be problematic only if insurance companies are actually able to decide which companies can exercise their activities. Obviously, this problem arises especially in a monopolistic market. This may, moreover, cause practical problems. Imagine that an insurer has stipulated in the policy conditions that coverage will be excluded in the case of

216


6.


non-compliance by the insured with mandatory government regulation. This may well be, as we have suggested above, an effective means of controlling moral hazard. If, however, an accident happens under compulsory insurance, the insured will not be able to call on this exclusion vis-à-vis the third-party beneficiary of the liability insurance policy. The fact that defences in the insurance contract are not opposable to third parties is a well-known problem under compulsory insurance. The insurer will thus have to compensate the victim and may have a (statutory or contractual) legal right of recourse against the insured, provided that the latter is solvent. This is, as we explained, one of the reasons why insurers are reluctant about compulsory insurance.

4.2. Necessity of co-operation with insurers One could obviously argue that these problems can be remedied if the policy-maker and the insurance industry co-operate closely, so that insurers inform the policy-maker on the insurability of systemic risks. However, practice has shown that information provided by insurers concerning the insurability of a certain risk or with respect to the available amounts of coverage may not always be reliable. Below we will discuss the example of the Dutch insurers who decided collectively that the flood risk would be uninsurable. In fact it was only this monopolistic decision that rendered the risk uninsurable. There seems to be a trade-off in this respect: introducing a duty to insure without any co-operation with the insurance world (which may have been the case in Germany) could have catastrophic results, with government forcing industry to take out insurance coverage which the market is unwilling to provide. However, close co-operation between the insurers (usually represented via one insurance association) and the government only increases the risk of high concentration on insurance markets.

4.3. Availability of insurance for systemic risks? A further problem is that the policy-maker should also realise that some particular systemic risks may be so new that insurance markets may not yet have developed to cover them. If a differentiated offer of insurance policies is limited, one could again question whether it makes sense to introduce mandatory insurance if such coverage could only be found to a limited extent (or without sufficient competition) on private insurance markets. Of course, the limited availability of insurance cover for systemic risks today is to a large extent caused by the adverse selection problem: since too little companies have a demand for insurance, optimal risk-spreading (via the law of large numbers) is not possible552. Moreover, only bad risks will have a demand for insurance, which precisely causes the adverse selection problem. Hence, one could naively argue that this problem could be remedied by forcing all injurers to take out insurance coverage for good and bad risks. However, it seems strange to seek to remedy the limited availability of insurance for environmental risks,


217

6.


which may largely be due to the fact that they are difficult to insure, by forcing all injurers to purchase coverage. In this connection, we can refer to the example of environmental insurance. Indeed, risk differentiation in environmental insurance in Europe is still in its infancy, and there is far more scope to relate policy and premium conditions in an appropriate way to the ecological reliability of firms. Hence, one can really question whether today insurance firms are yet able to differentiate environmental risks in such a way that one can argue that moral hazard can be controlled optimally on competitive insurance markets. The remedy to these problems is obviously not to make a poorly-functioning insurance system compulsory.

4.4. Duty to accept? Obviously, one could naively retort that if insurance markets refuse to provide appropriate coverage, then the policy-maker should not only introduce a duty for industry to take out mandatory coverage but also a duty for insurance companies to accept. Such a duty to accept would, it could be argued, at the same time remedy the risk that the insurance undertakings would de facto become the licensors of industry. However, introducing a duty to insure certain industries seems an extremely dangerous path to take given the importance of effective control of moral hazard. One important instrument insurers have for controlling moral hazard is precisely the possibility to monitor ex ante the risk which a particular insured party may pose. This could ultimately prompt an insurance undertaking to consider that the environmental risk a particular industry poses is too high. A logical consequence of the wish to have optimal control of moral hazard should be the right of insurance undertakings to freely decide which potentially responsible parties to insure and which not to. A duty to accept certain risks seems therefore to conflict with the basic principles which have to be respected to guarantee an effective functioning of insurance markets.

5. Policy recommendation From the above it follows that there are arguments in favour of introducing a compulsory insurance scheme on account of information deficiencies and the risk of underdeterrence as a result of insolvency. However, on theoretical grounds one can point to dangers as well, more particularly the fact that insurers may become the licensors of the activities which may cause systemic risks. This should not be remedied through a duty to accept, since it may cause irremediable problems of moral hazard. For all these reasons, the policy-maker should be extremely cautious about introducing a regulatory obligation to purchase insurance coverage. From this discussion on compulsory insurance, an important conclusion can probably be drawn. Although it may be important from a theoretical

218


6.


perspective to impose a duty on the permit-holder to secure appropriate means, it seems more appropriate to adopt a flexible system whereby the licensing authorities could judge on a case-by-case basis whether the obligation to provide financial security has been met. Such a system, whereby it is left to the administrative authorities to decide the form and amount of the financial obligation, seems more flexible and entails fewer risks and dangers than a generalised system of compulsory liability insurance. To reiterate, the principle that liability should be covered through some form of financial security (not necessarily insurance) should be laid down in legislation. The authorities would then only have to fix the amount, taking into account the expected damage (this will allow for an individualisation and differentiation of risks), and they would have to check whether the type of financial assurance offered by the potentially responsible party will be adequate to meet his financial obligations. Here maybe the example of the Flemish InterUniversity Commission can be useful553. The draft decree on environmental policy would not introduce compulsory insurance but would require the licensee of a classified activity to provide a deposit in order to guarantee that specific obligations would be complied with. As proposed by the Flemish Inter-University Commission, the authority which would fix the amount and check the provision of financial security could be the agency that issues the licence at the start of the operation of the activity.554 In terms of the IPPC Directive, this would be the competent authority granting the permit. 555 Sweden has a compulsory insurance scheme for environmental harm which seems to be working quite well. Some argue that this is because it is not based upon liability insurance but upon direct insurance.556 In sum, the liability regime should be combined with some kind of obligation to provide financial security for systemic risks if one can assume that an insolvency risk may emerge. But: ●

this financial security should not necessarily be liability insurance;

●

the policy-maker could indicate that a wide variety of mechanisms may be used to provide this financial security;

●

the type of financial security provided should not be regulated in a general way but its adequacy should be assessed, for example by the administrative authorities who can require financial security as a condition for issuing the licence;

●

at the same time, the administrative authorities can determine the required amount of financial security on a case-by-case basis;

●

a duty to accept risks should certainly not be imposed on liability insurers, since this may have negative effects on the control of moral hazard;


219

6.


●

the administrative authorities imposing such a duty to provide financial security should make sure that a sufficient variety of financial securities exist on financial and insurance markets in order to ensure that governments or administrative authorities do not become dependent upon the financial or insurance industry, which would then effectively become the licensor of industrial activities;

●

the proposed regime is in line with the proposals made by the InterUniversity Commission in the Flemish Region. These proposals were based on information provided by insurers, and the regime has been incorporated in the Flemish soil pollution decree. Hence, there is some empirical evidence which shows that such a balanced and mitigated obligation to provide financial security in limited cases may work effectively.

F. Facilitative strategies In this chapter we have addressed the possible role of government in coping with newly emerging systemic risks. On the one hand, we pointed to the importance of regulation in preventing damage caused by systemic risks; on the other hand, we examined the role government could play in providing compensation for damage caused by systemic risks. As far as the latter aspect was concerned, we were relatively cautious. Our starting point was that as far as possible private initiatives by the insurance industry and financial markets should be fostered, and only when they have failed or have proven impossible because of technical reasons should consideration be given to public compensation mechanisms such as compensation funds. However, we also argued that government can play an important facilitative role in making these private financial and insurance instruments possible, for example by not introducing retroactive liability or by not prohibiting a claims-made coverage system. In addition, one could imagine the introduction of a compulsory insurance system. So far, we have merely discussed a system of compulsory liability insurance, but one could also conceive of a system of compulsory firstparty insurance (1). Secondly, we also argued that the public authorities can play an important role in promoting insurance markets via an adequate competition policy (2). An adequate information strategy is also very important (3).

1. Compulsory direct insurance or first-party insurance? In the previous chapter, we were relatively enthusiastic about systems of direct insurance or first-party insurance. The question one could ask now is whether a duty should be introduced, for instance for each enterprise or employer, to take out direct insurance to the benefit of all his employees or, even more broadly, for all potential victims (potentially all citizens) who may be hurt by the particular systemic risk the enterprise causes. Such general accident insurance is available (on a non-mandatory basis) in France. The

220


6.


Fédération Française des Sociétés d’Assurances offers first-party insurance for all the damage caused by accidents (wherever they occur). This policy, with the very general description of “les garanties des accidents de la vie”, provides compensation irrespective of whether someone is liable or not. What is important is that the victim gets compensation as if liability law were applicable. In other words, pain and suffering are also compensated for. Obviously, the firstparty insurer who has provided coverage for “les accidents de la vie” has a right of recourse against the third party who may be liable for the loss. This insurance policy is obviously attractive for those who wish to receive compensation relatively fast and do not want to wait for the drawn-out procedures in tort law. Moreover, this policy is attractive for those who fear that they might be victim of an accident where no liable injurer can be identified. Such a policy providing protection against “les accidents de la vie” may theoretically be a good solution for damage caused by systemic risks and corresponds to the first-party insurance models described above. The advantage is, moreover, that the insurer can tailor coverage exactly according to the wishes and individual needs (including the risk profile) of the insured party. Optimal risk differentiation is therefore possible. The disadvantage is of course that the potential victims will have to finance the coverage themselves. The question of course arises of whether this French model is so attractive that it should be introduced on a mandatory basis. We do not believe this to be the case. The reason is that it does not seem wise to provide mandatory coverage also for non-pecuniary losses. In the literature, it has been argued that the most important feature of non-pecuniary losses is that they do not in principle generate a demand for extra money. Moreover, the fact that non-pecuniary losses do not cause ex post a demand for extra money also has as a consequence that non-pecuniary losses do not bring about ex ante a demand for insurance. Therefore, the victim would in principle not take insurance coverage ex ante to get compensation for non-pecuniary losses. The payment of a premium ex ante will indeed always cause a loss of income and it is doubtful whether the average citizen has a willingness to pay this premium to get compensation for his nonpecuniary losses. An additional problem is that non-pecuniary losses may be subjective and can be different for every individual. This makes risk differentiation ex ante very difficult precisely because of the subjective nature of non-pecuniary losses. Different people react differently to accidents. This also causes a moral hazard problem when the amount of compensation has to be fixed. The latter problem can to some extent be dealt with via a standardisation of the amounts of compensation, but that would again mean that the compensation is not fixed according to the individual means of every victim.


221

6.


If there are certain individuals who fear ex ante that they will suffer more non-pecuniary losses after an accident than others, and if they believe that these non-pecuniary losses will generate a demand for extra money, then they individuals may take out accident insurance on the French model which will provide insurance coverage and hence extra money if an accident occurs. Hence, we believe that it is better to let individuals purchase first-party insurance coverage according to their own demands and preferences and their own risk profile. It seems wrong to force all potential victims into a system whereby they would be obliged to pay a premium for additional protection (above that provided by the social security system) even if they have no demand for it. A general duty to purchase first-party insurance coverage may even have negative distributional effects. It would indeed force all victims to pay for additional insurance coverage whereas only some individuals may actually have a demand for it. The negative redistribution is caused by the fact that all citizens would have to pay whereas only some of them would derive a relatively high benefit (because of their subjective propensity to suffering). In sum, first-party insurance (e.g. in the form of French accident insurance) would certainly be a useful and desirable instrument for coping with damage caused by systemic risks. Given the fact that the demand for insurance of every potential victim may be different, a regulatory intervention whereby the government would force all individuals to purchase such insurance, would only have negative consequences. It may lead to inefficiencies and a negative distribution. Thus, although first-party insurance seems like an adequate answer to some systemic risks, the credo should still be that it should be purchased on a voluntary basis.

2. Necessity of a competition policy Finally, it is also important to remember that an effective competition policy may be quite an important tool to guarantee optimal availability of insurance coverage for systemic risks. Some concerted practices in the Netherlands with respect to catastrophic losses have shown that in some cases insurers agree not to cover certain risks although technically they were not uninsurable. Competition authorities therefore always have to be aware that a wide and varied offer of insurance products should be made available on the market for systemic risks. In this respect, we should also warn against the trend in some countries to use standard form policies, which may seriously limit the freedom of choice of the insured. Some practices on the European market show that a too limited degree of competition on insurance markets may seriously limit the availability of insurance.

2.1. Importance of competition on insurance markets We noted that the Dutch Insurers’ Association advised all their members to move towards environmental damage insurance and to let the policy replace

222


6.


traditional environmental liability insurance. The effect could be that liability insurance for environmental risk will no longer be available on the Dutch market. Moreover, in the past the Insurers’ Association took an opposite stance with respect to flood risk: it argued for a long time that these risks were uninsurable and even prohibited their members to provide coverage for flood risks. Earlier we showed that these anti-competitive pressures may be particularly dangerous since they will obviously restrict the provision of a wide and differentiated variety of financial and insurance products for covering environmental risk. It is interesting to have a closer look at some of these practices, given the fact that the boundaries of the application of European competition law to the insurance industry have recently been outlined in a report of the European Commission with respect to the functioning of a regulation concerning the insurance industry.557 Certain practices with respect to environmental liability insurance are explicitly addressed in this report and therefore merit some attention here. It should first of all be pointed out that Commission Regulation No. 3932/ 92 of 21 December 1992 exempts many cartel agreements in the insurance world from the prohibition under the old Article 85 (3) of the EEC treaty, provided that certain strict conditions were met.558 This exemption has already been heavily criticised by law and economics scholars, who have argued that competition policy should be fully applied to insurance markets as well. They have argued that an exemption from competition policy may lead to a premium increase, lower product differentiation and potentially even to an increase in accident costs (because the incentives to control moral hazard would be reduced)559. But these criticisms will probably not affect the insurance industry, which can benefit from the exemption. However, the recent report to the European Parliament of 12 May 1999 made clear that some current practices may not fall under the conditions for the exemption from the cartel prohibition:

2.2. Binding decisions In the 1950s the Dutch Insurers’ Association issued a so-called “binding decision” on all of its members, prohibiting them from insuring flood and earthquake risks (the latter being relatively small in the Netherlands with the exception of the area around southern Limburg). Their argument was that these risks were technically not insurable and that therefore its members should all refrain from covering them. The argument concerning their uninsurability seems highly doubtful560 and also clearly violated the conditions of Regulation 3932/92. Consideration 8 preceding the exemption clearly states that standard policy conditions may in particular not contain any systematic exclusion of specific types of risk without providing for the express possibility of including that cover by agreement. This is repeated in Article 7 (1) (a) of the exemption. In its report to the European Parliament of 12 May, the Commission explicitly discusses these


223

6.


binding decisions.561 The report states that as a result of the questions asked by the Commission, the Dutch association of insurers decided to bring its binding decision into line with Article 7.1 sub-paragraph a by simply converting it into a non-binding recommendation, leaving each insurer free to extend cover to flood risks. However, this example shows that non-competitive practices are apparently not an exception in Dutch insurance practice.

2.3. Standard form policies A point which inevitably needs to be addressed in the context of the changes in the environmental insurance structure in the Netherlands discussed above, is the extent to which these standard policy models are compatible with European competition policy. We already noted concerning liability insurance for companies in the Netherlands (AVB) that a new model policy came into being in 1996 and was put on the market.562 This AVB model policy suggested the change from loss-occurrence coverage to claims-made coverage.563 Also, as has just been mentioned, the Dutch Insurers’ Association decided to change the coverage of environmental risks from liability insurance to first-party insurance. The question which inevitably arises is whether it is compatible with competition policy to formulate advice via a standard policy which is de facto followed by the whole market. Again, this issue is covered by the exemption regulation 3932/92 of 21 December 1992. Article 6 of the exemption regulation provides that the exemption shall apply to standard policy conditions on the condition that it is made clear that these policy conditions are purely illustrative and that they expressly mention the possibility that different conditions may be agreed upon. In addition, the standard policy conditions must be accessible to any interested person and provided on request. The standard policy conditions used by the Dutch insurers in environmental insurance certainly can meet this test. Both the new insurance policy for company liability (AVB) and the environmental damage insurance (MSV) state explicitly that they are merely illustrative and that any individual insurer is free to deviate from the wording or the contents of the policy model.564 Formally, therefore, the standard policy conditions provided by the Dutch Insurers’ Association are compatible with the exemption regulation. Nevertheless, one cannot deny that the use of standard form policies will inevitably have the effect of restricting competition. Indeed, although these standard policy models all state that they are merely illustrative and that any insurer can deviate from them, in the Netherlands today it is almost impossible in practice to get liability insurance for the environmental risks which are now covered under the MSV (more particularly for soil pollution). However, these concerns will probably not be a great worry to Dutch insurers since the standard form polices used can easily pass the test of the exemption regulation.

224


6.


Indeed, the Dutch Insurers’ Association did not formally force all insurers not to provide liability cover for environmental risks, and the standard policies stipulate clearly that every insurer is free not to follow the recommendations made in the standard policy. Formally, Dutch practice is therefore compatible with the requirements of European competition policy. However, in practice almost all the insurers in the Dutch Insurers’ Association follow the recommendations in the standard policy. The effect therefore is that a Dutch company will no longer be able to purchase liability cover for soil pollution on the Dutch market. The same is also true for foreign insurance companies operating on the Dutch market. The only alternative for Dutch operators today is to seek liability cover abroad. Even though standard form policies formally comply with European competition policy, some scholars have pointed out that they lead to serious restrictions of competition in practice and to reduced product differentiation.565 In sum: although we have argued that it is understandable that the Dutch Insurers’ Association wished to offer environmental damage insurance on a first-party/direct insurance basis, it is, from a competition perspective, regrettable that the introduction of this new product was accompanied by the abolition (in practice) of environmental liability insurance in the Netherlands, at least for the risks covered under the MSV. The upshot therefore is to reduce the variety of insurance products for environmental risk.

2.4. Summary These examples, taken largely from the Dutch market, show that some of these practices are a cause for concern. Apparently, the co-operation between insurers in some markets is so close that although formally these standard policies only “advise”, in practice all insurers seem to follow the advice given by the association of insurers. The examples given show that restrictions of competition on insurance markets may seriously limit product differentiation and therefore also endanger the insurability of some systemic risks. This shows that an effective competition policy is of utmost importance to ensure that a wide variety of efficient insurance policies is available on the market for the particular systemic risks. Otherwise, the risk might emerge that, as a result of concerted practices, the availability of insurance would be limited instead of increased.

3. Information policy In our view, the government has a very important role to play in making citizens aware of newly emerging systemic risks. It is important that government take a pro-active role, for example by doing research on newly emerging systemic risks and on the preventive strategies that could be followed to prevent them. The results of the research undertaken at the government level could be translated into safety regulations, but other strategies could be imagined as well. In some


225

6.


cases, private-public co-operation, whereby joint investments in research and development are undertaken, could be beneficial as well. It is clear that adequate information on the nature and character of newly emerging systemic risks and on the preventive strategies to be followed is in the interest of both private actors and governments. The same is true for information concerning the potential consequences of newly emerging systemic risks. It seems important to examine the potential damage that may result from catastrophic events such as earthquakes, hurricanes etc. Simulations, including damage assessments, can be very important. We explained above that many citizens may underestimate the potential consequences of newly emerging systemic risks. Lack of information is a very real problem. Information deficiencies, it was argued above, can therefore constitute a reason for introducing a duty to insure if risk-averse individuals underestimate the advantages of insurance. However, we also argued that compulsory insurance may not be an adequate tool that could justify compulsory first-party or direct insurance. Other instruments could be used to remedy information deficiencies, for example an active government policy to inform citizens about the potential consequences of systemic risks. In that respect, the citizens are obviously better-off with honest and open information (even if this may cause major worries) than with strategies to conceal or hide relevant information. It would be important, however, also to inform citizens about the scope of coverage in the event that systemic risks cause major damage. The government in a particular country could start with active information strategies to inform citizens about the part of the damage that would be covered by social security, for example, and the part that would remain uncovered. On the basis of that information, well-informed utility-maximising citizens may then choose to purchase additional accident insurance on a first-party basis according to their own preferences and needs. Thus we argued that an active information policy by the government can help citizens to make correct choices.

Notes 454. S. Shavell, Liability for Harm versus Regulation of Safety, 1984, JLS, 357-374; S. Shavell, A Model of the Optimal Use of Liability and Safety Regulation, 1984, Rand Journal of Economics (RJE), 271-280 and S. Shavell (supra note 12) pp. 277-290. 455. See the basic article by G. Stigler, The economics of information, 1961, Journal of Political Economics (JPE), 213 and see A. Schwartz/L. Wilde, Intervening in markets on the basis of imperfect information: a legal and economic analysis, 1979, University of Pennsylvania Law Review (UPLR), 630-682 as well as E. Mackaay, Economics of information and the law, 1982. 456. S. Shavell (supra note 454) p. 359. 457. S. Shavell (supra note 454) p. 360.

226


6.


458. S. Shavell, The judgement proof problem, 1986, IRLE, 43-58. Above we mentioned that insolvency causes a problem especially under a strict liability rule but less so under negligence. 459. If insurance came into the picture it could overcome the problems of underdeterrence provided that the moral hazard problem caused by insurance could be resolved. 460. S. Shavell, Criminal law and the optimal use of non-monetary sanctions as a deterrent, 1985, Columbia Law Review, 1232-1262. 461. S. Shavell (supra note 454) p. 363. 462. See W. Landes/R. Posner, Tort law as a regulatory regime for catastrophic personal injuries, 1984, JLS, 417. 463. For alternatives to liability suits see: H. Bocken, Alternatives to Liability and Liability Insurance for the Compensation of Pollution Damages, 1987, TMA, 83-87 and 1988, TMA, 3-10. 464. Faure, M., “Compensation for non pecuniary losses from an economic perspective”, Magnus, U and Spier, J. (eds.), European Tort Law, Liber Amicorum for Helmut Koziol, Frankfurt am Main, Peter Lang, 2000, 143-159. 465. According to Cooter, this is the most important difference between tort law and criminal law: tort law simply puts a price (in the form of compensation due for harmful acts), whereas criminal law simply prohibits some activities from being undertaken, by imposing a sanction on the one who nevertheless engages in this activity (see Cooter, R., “Prices and sanctions”, Columbia Law Review, vol. 84, 1984, 1523. 466. S. Shavell (supra note 454) pp. 363-364. 467. These are often referred to as “public interest” criteria for regulation in order to contrast them with “private interest” explanations for regulation, as advanced by public-choice scholars. 468. Complementarities between tort law and regulation have been addressed by S. Rose-Ackerman, Environmental Liability Law, in: T.H. Tietenberg (ed.), Innovation in Environmental Policy, Economic and Legal Aspects of Recent Developments in Environmental Enforcement and Liability, 1992, pp. 223-243; S. Rose-Ackerman, Rethinking the Progressive Agenda. The Reform of the American Regulatory State, 1992, pp. 118-131 and S. Rose-Ackerman, Regulation and the Law of Torts, 1991, American Economic Review, Papers and Proceedings (AERPP), 54-58. 469. D. Dewees, The Comparative Efficacy of Tort Law and Regulation for Environmental Protection, 1992, GPRI, 446-467 and D. Dewees, Tort Law and the Deterrence of Environmental Pollution, in: T.H. Tietenberg (ed.), Innovation in Environmental Policy, Economic and Legal Aspects of Recent Developments in Environmental Enforcement of Liability, 1992, pp. 139-164. 470. D. Dewees/D. Duff/M. Trebilcock, Exploring the Domain of Accident Law; Taking the Facts Seriously, 1996. 471. D. Dewees/D. Duff/M. Trebilcock (supra note 470) pp. 307-323. 472. S. Rose-Ackerman (supra note 468); S. Rose-Ackerman (supra note 468) pp. 223-243 and S. Rose-Ackerman, Public Law versus Private Law in Environmental Regulation: European Union Proposals in the Light of United States and German Experiences, in: E. Eide/R. Van den Bergh (eds.), Law and Economics of the Environment, 1996, pp. 13-39.


227

6.


473. M. Faure/M. Ruegg, Standard Setting through General Principles of Environmental Law, in: M. Faure/J.Vervaele/A. Weale (eds.), Environmental Standards in the European Union in an Interdisciplinary Framework, 1994, pp. 39-60. 474. Ch.D. Kolstad/Th.S. Ulen/G.V. Johnson, Ex Post Liability for Harm vs. Ex Ante Safety Regulation: Substitutes or Compliments?, 1990, 80 American Economic Review (AER), 888-901. 475. A. Arcuri, Controlling environmental risk in Europe: the complementary role of an EC environmental liability regime, 2001, Tijdschrift voor Milieuaansprakelijkheid (TMA), pp. 39-40. 476. P. Burrows, Combining regulation and liability for the control of external costs, 1999, 19 IRLE, 227-242. 477. S. Rose-Ackerman, Public Law versus Private Law in Environmental Regulation: European Union Proposals in the Light of United States Experience, 4 Review of European Community and International Environmental Law (RECIEL), 312-332 and S. Rose-Ackerman, Controlling Environmental Policy: the Limits of Public Law in Germany and the United States, 1995. 478. For a recent different analysis, which also concludes that liability and regulation should be combined, see P.W. Schmitz, On the joint use of liability and safety regulation, 2000, IRLE, 371-382. 479. For a comparative analysis of the question of whether complying with a permit excludes criminal liability see, M. Faure/J.C. Oudijk, Die strafgerichtliche Überprüfung von Verwaltungsakten im Umweltrecht. Ein rechtsvergleichender Überblick der Systeme in Deutschland, den Niederlanden und Belgien, 1994, JZ, 86-91. 480. S. Shavell (supra note 10) p. 365; M. Faure/R. Van den Bergh, Negligence, Strict Liability and Regulation of Safety under Belgian Law: An Introductory Economic Analysis, 1987, GPRI, 110. 481. According to P. Burrows (supra note 476) p. 242. Recently Schwartz added to the debate by discussing whether compliance with federal safety statutes should have a justificative effect in state tort cases. See A. Schwartz, Statutory Interpretation, Capture, and Tort Law: The regulatory compliance defense, 2000, American Law and Economics Review (ALER), 1-57. 482. C.D. Kolstad/T.S. Ulen/G.V. Johnson (supra note 474) pp. 888-901. 483. P. Burrows (supra note 476) pp. 227-244. 484. M. Faure/I. Koopmans/J. Oudijk, Imposing criminal liability on Government Officials under environmental law: a legal and economic analysis, 1996, Loyola of Los Angeles International Comparative Law Journal, 529-569. 485. Note, however, that industry argues against such a liability of the licensor, claiming that this may entail the risk that licensors would be too reluctant to allow an emission if it could give rise to their liability (G.J. Niezen, Aansprakelijkheid voor milieuschade in de Europese Unie in: Ongebonden Recht Bedrijven, 2000, p. 171. 486. S. Rose-Ackerman (supra note 468) p. 123 and A. Arcuri (supra note 475) pp. 43-44. 487. This is further developed in M. Faure, The applicability of the principles of private insurance to social healthcare insurance seen from a law and economics perspective, 1998, The Geneva Papers on Risk and insurance, 265-293.

228


6.


488. See the discussion on moral hazard above. 489. Moreover, in case of insurer ambiguity, an additional risk premium may be charged to cope with their risk. See above Chapter 3, B, 3. 490. The principles of social security are discussed by D. Pieters, Introduction in the basic principles of social security, 1993. 491. The way the system is organised may depend upon the particular legal system. For an introduction to the social security system in Europe see D. Pieters (ed.), Introduction into the social security law of the member States of the European community, 1993. 492. From an economic perspective, one can of course argue that in fact the employer’s contribution to the premium is also charged to the employee in the form of reduced salary. 493. Also O. Mitchell/S. Zeldes, Social SecurityPrivatisation: A Structure for Analysis, 1996, p. 11 consider redistribution and national risk-sharing as essential features of the social security system. 494. K. Arrow, Uncertainty and the Welfare Economics of Medical Care, 1963, American Economic Review, 941-973. 495. See M. Pauly, The Economics of Moral Hazard: Comment, 1968, American Economic Review, 531-537. 496. S. Shavell, On Moral Hazard and Insurance, 1979, Quarterly Journal of Economics, 541-562. 497. One example is the solidarity between younger employees and older patients, having to do with the fact that most health care expenditure is usually incurred in the final stage of a person’s life. 498. Van Mierlo pointed out that the government in fact confusingly mixes two policy goals: guaranteeing equal access to health care with income redistribution (J.G.A. van Mierlo, Een alternatief voor het Plan-Simons? Overwegingen vanuit de welvaartstheorie, 1991, ESB, 1164). 499. This might, however, be different for “cosmetic” surgery such as breastenlargements, etc. These “beauty corrections” can be regarded as a marketable item, so A.H.M. Kerkhoff, The Physicians’ Fraternity: Safeguard for Quality or Cartel?, in: L.J. Gunning-Schepers/G.J. Kronjee/R.A. Spasoff (eds.), Fundamental Questions About the Future of Health Care, 1996, p. 117. 500. M. Pauly, The Economics of Moral Hazard, 1968, American Economic Review, 534-535 and J. Ehrlich/G. Becker, Market Insurance, Self-insurance, and Self Protection, 1972, Journal of Political Economy, 623-648. 501. On this growing debate on the promotion of preventative health care, see the contributions in P.J. van Wijngaarden/F.G. van den Heuvel/J.C. Vrooman (eds.), Preventie in de sociale zekerheid, 1992 and in J. van Steenberge/S. Klosse/L.J.M. de Leede (eds.), Preventie: een solide basis voor sociale zekerheid?, 1994. 502. M. Whinston, Moral Hazard, Adverse Selection and the Optimal Provision of Social Insurance, 1983, Journal of Public Economics, 49-71. 503. On possible conflicts between optimal incentives for prevention and optimal risk-spreading, see R. Zeckhauser, Medical Insurance: a Case Study of the Tradeoff Between Risk-spreading and Appropriate Incentives, 1970, 2 Journal of Economic Theory, 10-26.


229

6.


504. See J. Ferejohn/Ch. Shipan, Congressional Influence on Bureaucracy, 1990, 6 Journal of Law, Economics and Organization, 1-20. 505. Also the Netherlands Scientific Council for Government Policy argued in a 1994 report “Belang en beleid” (Interest and Policy) that the public sector lacks the incentive which the private sector has to reduce the volume of health care costs (see M.W. Dijkshoorn, Ontwikkelingen in de WAO, 1996, Verzekeringsarchief, 82-86). 506. On the negative consequences of restrictions on competition on the insurance market, see M. Faure/R. Van den Bergh, Het toelaten van kartels op de Europese verzekeringsmarkt: hogere premies, lagere kwaliteit en meer ongevallen?, 1993, Nederlands Juristenblad, 261-267 and M. Faure/R. Van den Bergh, Restrictions of Competition on Insurance Markets and the Applicability of EEC Antitrust Law, 1995, Kyklos, 65-85; and, with respect to health care insurance, M. Pauly, Competition in Health Insurance Markets, 1988, Law and Contemporary Problems, 237-271. 507. According to W. Johnson, Choice of Compulsory Isurance under Adverse Selection, 1977, Public Choice, 23-25; M. Pauly, Overinsurance and the Public Provision of Insurance: The Roles of Moral Hazard and Adverse Selection, 1974, Quarterly Journal of Economics, 44-62 and M. Rothschild and J. Stiglitz, Equilibrium in Competitive Insurance Markets: an Essay on the Economics of Imperfect Information, 1976, Quarterly Journal of Economics, 629-649. 508. We do realize obviously that this device between private and public insurance insolvency involves many more complex issues, which we cannot develop within the scope of this paper. 509. See, for example, on the Medicaid system in the US, W. van de Griendt, De Amerikaanse paradox, 1995, ⅞ Sociaal bestek, 9-12 and R. Frank, Lessons from the Great Battle: Health Care Reform, 1992-1994, 1997, Arch. Phys. Med. Rehabil, 122-123. 510. We present it here, for simplicity reasons, as two extremes: either a free private market or public health care provided via the government. In reality there are many “shades of grey” such as private market with regulation. It is therefore not such a clear-cut choice. 511. If an insurer did so, he would be in a disadvantageous competitive position. 512. See L.J.C.M. Le Blanc, Bismarck als benchmark. Enkele rechtseconomische beschouwingen over publieke sector en sociale zekerheid in Nederland, 1990, Ars Aequi, 654-659 and V. Fuchs, From Bismarck to Woodcock: the “Irrational” Pursuit of National Health Insurance, 1976, Journal of Law and Economics, 347-359. 513. So M.W. Dijkshoorn, Verzekeringsarchief, 86. 514. O. Mitchell/S. Zeldes, Social Security Privatisation: A Structure for Analysis, 1996, p. 11. 515. H.A. Cousy, Lectuurnotities en reflecties over ethiek in het zakenleven, in het bijzonder in de verzekeringen, in: F. Fleerackers (ed.), Mens en recht. Essays tussen rechtstheorie en rechtspraktijk, Liber Amicorum Jan M. Broekman, 1996, pp. 47-63. 516. W.P.J. Wils, Insurance Risk Classifications in the EC: Regulatory Outlook, 1994, Oxford Journal of Legal Studies, 449-467. See also G. Schoorens, Segmentering en discriminatie, in H. Cousy/H. Claassens/C. van Schoubroek (eds.), Competitiviteit, ethiek en verzekering, 1998, p. 217-277. 517. See J.G.A. van Mierlo, Hervorming van het Nederlandse stelsel van sociale zekerheid: financiële consequenties en politieke haalbaarheid, in: F.A.J. van den

230


6.


Bosch/C. Peterson (eds.), Economie en arbeidsongeschiktheid: analyse en beleid, 1983, pp. 215-241. This is also defended by C.A. de Kam, Privatisering van sociale zekerheid: een werkbaar alternatief?, in: W.J.P.M. Fase e.a., Sociale Zekerheid: privaat of publiek?, 1994, pp. 52-57. The possibilities of complementary health care insurance in addition to a basic social benefit system are also addressed, for example, by H.A. Cousy, De rol van de private verzekeringen in de aanvullende sociale verzekeringen: de regels van het spel, in: J. van Langendonck (ed.), Liber Amicorum Roger Dillemans, 1997, pp. 55-74. 518. Interesting papers analysing the Superfund can be found in R.L. Revesz/ R.W. Stewart, Analysing Superfund, Economics, Science and Law, Washington, 1995. 519. See P.A. Kottenhagen-Edzes, Onrechtmatige daad en milieu. Het gebruik van het privaatrecht bij het voorkomen van milieu-aantasting en het verhaal van milieuschade, 1992, pp. 297-298; E. Hulst, Grondslagen van Milieuaansprakelijkheid, 1993; J. Knottenbelt, Produktaansprakelijkheid, 1990 and A.J. Van (supra note 115). 520. See Interuniversitaire Commissie voor de Herziening van het Milieurecht in het Vlaamse Gewest, Voorontwerp Decreet Milieubeleid, 1995, pp. 943-985 and H. Bocken/W. Lambrecht/M. Boes/A. De Nauw/M. Faure/L. Lavrysen, “The Flemish Draft Decree on Environmental Policy: An Outline”, in: H. Bocken/D. Ryckbost (eds.), Codification of Environmental Law. Proceedings of the International Conference, 1996, pp. 31-32. 521. P. Gilhuis/J. Verschuuren, Een Milieuschadefonds in Nederland; een Onderzoek naar de Mogelijkheden, Publicatiereeks Milieubeheer, 1994/3. For comments see E. Hulst, De werkelijkheid rondom een algemeen milieuschadefonds, een commentaar, 1995, TMA, 167-173 as well as De Putter/J. Verschuuren, Een milieuschadefonds in Nederland?, 1995, MenR, 96-99. 522. But in that particular case, the liability itself has also been limited. 523. See the inauguration address of L. Dommering-Van Rongen, Schade Vergoeden door Fondsvorming, 1996. 524. This was why the European Directive on Product Liability also incorporated the idea of a limitation fund in Art. 16, which gave member States the option to limit liability in case of serial damage to an amount which may not be less than € 70 million. However, only a few member States have used this option. 525. See H. Bocken/W. Lambrecht/M. Boes/A. Nauw/L. Lavrysen (supra note 67) pp. 31-32. It is based on earlier work of – inter alia – H. Smets, Pour une indemnisation garantie des victimes de pollution accidentelle, in: H. Bocken/D. Ryckbost (eds.), Insurance of Environmental Damage, 1991, pp. 397-423. 526. More particularly by the Dutch Socialist Party (Socialistische Partij), see P. Swuste, Van Individuele “Asbestprocessen” naar een Asbestfonds?, 1996, Arbeidsomstandigheden, 119-122; for a critical analysis see M. Faure/T. Hartlief, Een Asbestfonds als Alternatief voor de Aansprakelijkheid van de Werkgever?, 1996, Tijdschrift voor Sociaal Recht, 37-43. 527. For a detailed discussion, see H. Bocken (supra note 70) p. 368 et seq. 528. For an overview concerning the current use of various compensation funds, see H. Bocken, Systèmes Alternatives pour l’Indemnisation des Dommages du à la pollution, 1990, Revue Génerale des Assurances et des Responsabilités, 11698-11714 and H. Bocken (supra note 463) and see H. Bocken, Complementary Compensation Mechanisms. A General Environmental Damage Fund?, in: H. Bocken/D. Ryckbost (eds.), Insurance of Environmental Damage, 1991, pp. 425-437.


231

6.


529. Unless there were joint and several or channeling of liability. 530. M. Faure/R. Van den Bergh (supra note 263) pp. 65-85. 531. See M. Faure/G. Skogh (supra note 21) pp. 499-513 and M. Faure (supra note 300) pp. 21-43. 532. See T.G. Coghlin (supra note 442) pp. 403-416. 533. With respect to insurance problems concerning environmental risks in Belgium, see H. Bocken, “L’Assurance Responsabilité Civile pour Dommages causées par la Pollution”, in: Les Assurances de l’Entreprise, Actes du Collogues tenu à l’Université Libres des Bruxelles les 2 et 3 décembre 1993, pp. 239-280 and H. Bocken, La réparation des Dommages causées par la Pollution au Droit Belge. La Situation en 1992, 1992, Tijdschrift voor Belgisch Burgerlijk Recht, pp. 284-327. See also H. Bocken/D. Ryckbost, Verzekering van Milieuschade, Insurance of Environmental Damage, 1991. 534. This argument has been made by G. Skogh, The Transactions Cost Theory of Insurance: Contracting Impediments and Costs, 1989, Journal of Risk and Insurance, 726-732. 535. See, for example,. H. Smets, COSCA: A complementary system for compensation of accidental pollution damage, in: Wettenstein, P, (ed.), Harm to the environment: The right to compensation and the assessment of damage, 1997, pp. 223-248. 536. This combined approach is also proposed by A. Monti (supra note 256). 537. For instance Germany, see G. Wagner (supra note 306) pp. 249-260; G. Wagner, Die Zukunft der Umwelthaftpflichtversicherung, 1992, VR, pp. 261-272 and G. Wagner (supra note 306) pp. 97-146. 538. Other causes of the failure of traditional liability and insurance are discussed by H. Bocken, Deficiencies of the system of liability and liability insurance as a mechanism for the indemnification of environmental damage suffered by individual victims, in: H. Bocken/D. Ryckbost (eds.), Insurance of Environmental Damage, 133-145. 539. Also Boyd and Kunreuther prefer this “public purse” solution to retroactive liability (J. Boyd/H. Kunreuther, Retroactive Liability or the Public Purse?, 1997, Journal of Regulatory Economics, JRE, pp. 70-90). 540. As A. Monti says correctly (supra note 256). 541. Compulsory environmental liability insurance is proposed by several scholars, including Cowell, J., “Compulsory environmental liability insurance”, in Bocken, H. and Ryckbost, D. (eds.), Insurance of environmental damage, 317-330 and Monti, A., “Environmental risk: A comparative law and economics approach to liability and insurance”, European Review of Private Law, 2001, 65. 542. Some even argue that mandatory financial responsibility could help implement the precautionary principle by ensuring availability of resources to meet the costs of any future environmental damage, so Richardson, B.J., “Financial Institutions for Sustainability”, Environmental Liability, 2000, 61. 543. Arrow, K., “Aspects of the Theory of Risk-Bearing”, Helsinki, Yrjö Jahnssonin Säätiö, 1965; Borch, K., “The Utility Concept Applied to the Theory of Insurance”, The Astin Bulletin, 1961, 245-255; Borch, K., “Recent Developments in Economic Theory and their Application to Insurance”, The Astin Bulletin, 1963, 322-341 and Pratt, J., “Risk Aversion in the Small and in the Large”, Econometrica, 1964, 122-136.

232


6.


544. More particularly by Shavell, S., “The judgement proof problem”, International Review of Law and Economics, 1986, 43-58. 545. Jost, P.J., “Limited liability and the requirement to purchase insurance”, International Review of Law and Economics, 1996, 259-276. A similar argument has recently been formulated by Polborn, M., “Mandatory Insurance and the Judgement-Proof Problem”, International Review of Law and Economics, 1998, 141-146 and by Skogh, G., “Mandatory Insurance: Transaction Costs Analysis of Insurance”, in Bouckaert, B. and De Geest, G. (eds.), Encyclopedia of Law and Economics, Cheltenham, Edward Elgar, 2000, 521-537. Skogh has also pointed out that compulsory insurance may save on transaction costs. 546. See also Kunreuther, H. and Freeman, P. in A. Heyer, (ed.), The Law and Economics of the Environment, 316. 547. Shavell, S., “The judgement proof problem”, International Review of Law and Economics, 1986, 43-58. 548. Cousy, H., “Recent developments in environmental insurance”, in Abraham, F., Deketelaere, K. and Stuyck, J. (eds.) Recent economic and legal developments in European environmental policy, 241 and Rogge, J., Les assurances en matière d’environnement, Loose-Leaf, Kluwer, 1997, 39. 549. See Wagner, G., Umwelthaftung und Versicherung, Versicherungsrecht, 1991, 249-260 and Wagner, G.,"Die Zukunft der Umwelthaftpflichtversicherung, Versicherungsrecht, 1992, 261-272. 550. This point is also made in the Green Paper 13. See also Rogge, J., Les assurances en matière d’environnement, Loose-Leaf, Kluwer, 1997, 40. 551. See Monti, A., “Environmental risk: A comparative law and economics approach to liability and insurance”, European Review of Private Law, 2001, 65. 552. See Rogge, J., Les assurances en matière d’environnement, Loose-Leaf, Kluwer, 1997, 38. 553. See Bocken, H. and Ryckebost, D. (eds.), Codification of Environmental Law, Draft decree on Environmental Policy, London, Kluwer Law International, 1996, 106. 554. Bocken, H., Ryckbost, D. and Deloddere, S., “Liability and financial guarantees”, in Bocken, H. and Ryckbost, D. (eds.), Codification of environmental law, Draft Decree on Environmental Policy, London, Kluwer Law International, 1996, 214-223. 555. See Council Directive 96/61/EC of 24 September 1996 concerning Integrated Pollution Prevention and Control, Official Journal, 10 October 1996, L 257/26. 556. So Bocken, H., “Rechtstreekse verzekeringen ten behoeve van derden en andere wisseloplossingen voor aansprakelijkheid en aansprakelijkheidsverzekering. Een typologie.”, in Liber Amicorum René van Gompel, 35. 557. See the report of the Commission to the European Parliament and to the Council of 12 May 1999 concerning the operation of the exemption regulation 3932/92 (COM (1999) 92 final). 558. Official journal, L 398/7 of 31 December 1992. 559. For this criticism, see Faure, M. and Van den Bergh, R., “Restrictions of competition on insurance markets and the applicability of EC antitrust law”, Kyklos, vol. 48, 1995, 65-85. 560. And was criticised among others by Faure, M. and Hartlief, T., “Een schadefonds als alternatief voor aansprakelijkheid en verzekering”, RM Themis, 1998, 220-222.


233

6.


561. See Report, No. 18, p. 9. 562. See Faure, M. and Hartlief, T., “Ontwikkelingen in de werkgeversaansprakelijkheid voor beroepsziekten: aanleiding voor een nieuwe AVB-polis?”, A&V, 1996, 140-151. For a critical comment see Haazen, O.A. and Spier, J., “Amerikaanse toestanden en de nieuwe aansprakelijkheidsverzekering voor bedrijven en beroepen”, NJB, 1996, 45-50. 563. See Spier, J. and Haazen, O.A., Aansprakelijkheidsverzekeringen op claims-made grondslag, Deventer, Kluwer, 1996 and Wansink, J.H., “Het polismodel AVB ’96 en de dekking voor long-tail risico’s”, A&V, 1996, 120-122. 564. See also Haazen, O.A. and Spier, J., “Amerikaanse toestanden en de nieuwe aansprakelijkheidsverzekering voor bedrijven en beroepen”, Nederlands JuristenBlad, 1996, 45. 565. See Faure, M. and Van den Bergh, R., “Aansprakelijkheidsverzekering, concurrentie en ongevallenpreventie”, in Hartlief, T. and Mendel, M., (red.), Verzekering en maatschappij, Deventer, Kluwer, 2000, 315-342.

234



Chapter 7

Recommendations


235

7.

RECOMMENDATIONS

O

ur recommendations will be brief. We have ended every chapter with a summary of the main results. It thus does not make any sense to repeat these summaries, which can be found at the end of each chapter. We will now simply recapitulate the most important points and formulate them as recommendations to the social actors addressed in this study – enterprises, insurers and governments. The central question for this study was how these social actors could protect themselves against expanding systemic risks. Obviously, the reader should remain aware of the limits of this study (and hence of the recommendations), since we mainly focused on the role of law and even then chiefly on the function of liability law and insurance. Also, the reader should be aware that these recommendations are set out very briefly without all the nuances and explanations. He or she should also be aware of the fact that they are formulated in a rather unbalanced manner, and should be complemented by reading the full text in the various chapters. With all these caveats in mind, the following recommendations follow from this study: 1. Liability law should still play an important role in the prevention of damage caused by systemic risks and should, moreover, be optimised. This means: ●

Liability law should be used only as a response to a shortcoming of the injurer and should be used only in those cases where the injurer made liable could have an influence on the accident risk;

●

Liability law should take into account the behaviour of the victim as well; this means that the claim on compensation should in principle be reduced to the extent that the victim has contributed to the loss;

●

Liability law should in principle not be retroactive;

●

The risk of causal uncertainty should not be shifted to one injurer;

●

Channelling of liability should be avoided;

●

Joint and several liability regimes should be avoided. The foregoing recommendations can in some cases lead to a reduction of the scope of liability in some legal systems; in other cases they may lead to an intensified use of liability law. Indeed, one feature of newly emerging systemic risks may be the fact that to some extent they are still largely unknown in the legal system and hence one can expect that for some of them, claims may arise in areas where they have not been brought up to now. Therefore, in some legal systems one may expect to see more use of tort law for these newly emerging systemic risks. Some may describe this

236


7.

RECOMMENDATIONS

trend as an evolution towards a “claim culture”, but this should not be judged negatively as such, provided that this use of tort law corresponds to its fundamental goals and functions. 2. In addition to liability law, safety regulation is also needed for the prevention of systemic risks. In that respect, regulatory intervention needs to be proactive and dynamic. The pro-active character of regulation means that regulation should not only come into being after known risks have effectively caused damage. A pro-active approach, based on the precautionary principle, means that regulatory intervention follows a preventive strategy to systemic risks. In addition, safety regulation should be dynamic instead of static. This means that it should be constructed in such a manner that as far as possible it can be changed and amended rapidly, and adapted to new insights with respect to systemic risks and new technological developments e.g. with respect to preventive strategies to remedy systemic risks. 3. As far as possible, liable injurers should be able to seek insurance coverage for their liability. Therefore, government strategy should aim at optimising liability insurance and hence promoting the insurability of liability. 4. In this respect, the government should: ●

organise the liability system in such a manner that the objective conditions for insurability can be met (e.g. by avoiding retroactive liability or shifting the risk of causal uncertainty);

●

the government may also wish to allow claims-made coverage for systemic risks;

●

moreover, the legislator should not introduce financial caps on liability for systemic risks on the ground that they are needed to increase insurability. Caps are not necessary to increase insurability since insurers could themselves put a financial limit on liability; caps may lead to underdeterrence and undercompensation.

5. Also, insurers can be given several pieces of advice to increase the insurability of systemic risks. They are, briefly: ●

control moral hazard effectively;

●

reduce adverse selection;

●

make use of all the available techniques to increase capacity;

●

apply effective risk differentiation and

●

provide coverage on a “claims-made” basis for all those systemic risks which have a “long-tail” character.

6. In addition, new instruments like direct and first-party insurance seem to open up very interesting possibilities for providing additional compensation to potential victims of newly emerging systemic risks. At the policy level, the


237

7.

RECOMMENDATIONS

government should not introduce mandatory first-party coverage, in order to avoid inefficiencies and a negative redistribution. 7. However, the government could implement a facilitative strategy to ensure that well-informed citizens can purchase on competitive insurance markets those insurance policies that fit their individual risk profile. To guarantee this, the government should: ●

pursue an active competition policy so that insurers are encouraged to provide a well-diversified range of insurance products on the market;

●

pursue an active information strategy to inform citizens about the potential consequences of new systemic risks and the proportion of the damage caused by such risks that may not be covered via social security or other available sources.

8. To some extent, at least in Europe, the problems with the financing of damage caused by systemic risks stem not from the fact that systemic risks are newly emerging but from the fact that the consequences of some of these risks are no longer financed via the social security system. To some degree, therefore, the straightforward remedy seems to halt the unravelling of the social security system that has taken place mostly in Europe. However, at the current time when deregulation and private markets are getting a lot of attention in the political arena, a call for a return to social security will not be very successful. It seems more feasible to move in the direction of direct or first-party insurance. To some extent, one can even argue that these insurance schemes are a privately (market) financed form of additional social security. 9. One should always keep in mind that there is not one system of compensation for damage caused by systemic risks which can be considered optimal for all kinds of situations. In most cases, the optimum can be found via a combination of (private and public) insurance and financial compensation systems. In seeking the appropriate combination of instruments, it is very important to look not only at their availability but also at the ability of the various instruments to achieve an optimal prevention of systemic risks as well as an optimal reduction of costs.

238



Bibliography Aaron, H., Issues Every Plan to Reform Health Care Financing must Confront, Journal of Economic Perspective, 1994, 3, 31-43. Abraham, K.S., “Cost Internalisation, Insurance, and Toxic Tort Compensation Funds”, Virginia Journal of Natural Resources, vol. 2, 1982, 123-131. Abraham, K., “Environmental Liability and the Limits of Insurance”, Columbia Law Review, vol. 88, 1988, 942-988. Adams, M., “Warum kein Ersatz von Nichtvermögensschäden” in Ott, C. and Schäfer, H.B. (eds.), Allokationseffizienz in der Rechtsordnung, Berlin, Springer, 210-217. Akerlof, G., “The market for “lemons”: quality, uncertainty and the market mechanism,” Quarterly Journal of Economics, 1970, 488-500. Akkermans, A.,Proportionele aansprakelijkheid bij onzeker causaal verband, Deventer, Tjeenk Willink, 1997. Alvarez-Baron, A.M., “Spain’s Claims-Made Crisis”, International Insurance Law Review, 1994, 316-319. Arcuri, A., Controlling environmental risk in Europe: the complementary role of an EC environmental liability regime, Tijdschrift voor Milieuaansprakelijkheid, 2001, pp. 39-40. Arisz, F.H.A., “Inleiding: Beroepsaansprakelijkheid”, in Arisz, F.H.A. (ed.), Beroepsaansprakelijkheid, Recht op een scheve schaats, Bundel ter gelegenheid van het Jonge Balie Congres 1991, Tjeenk Willink, 1991, 1. Arrow, Uncertainty and the Welfare Economics of Medical Care, American Economic Review, 1963, 941-973. Arrow, K., Aspects of the Theoryof Risk-Bearing, Helsinki, Yrjö Jahnssonin Säätiö, 1965. Ball, C.J.J.M. Stolker, Dereguleren in Arbo-land: een heroriëntatie op de arbeidsomstandighedenwet, Nederlands Juristenblad, 1997, 22, 969-975. Banakas, S., "European Tort Law: is it Possible?", European Review of Private Law, 2002, 363-375. Barendrecht, J.M., “Produktenaansprakelijkheid: Europees burgerlijk recht?”, Preadvies voor de vereniging van burgerlijk recht, Vermande, 1987. Bauw, E., Buitencontractuele aansprakelijkheid voor bodemverontreiniging, Kluwer, 1993. Bawcut, P.A., Captive Insurance Companies, Establishment, Operation and Management, 3rd edition, Woodhead-Faulkner, New York, 1991. Beck, U., Risikogesellschaft. Auf dem Weg in eine andere Modeine, Frankfurt, 1986. Bell, S. and Mc Gillivrary, D.,Environmental Law, London, Blackstone Press, 2000.


239

BIBLIOGRAPHY

Bergkamp, L., “The Proper Scope of Joint and Several Liability”, TMA, 2000, 154-155. Bergkamp, L., “Aansprakelijkheid is geen schadeverzekering”, Nederlands Juristenblad, 2000, 273-276. Bergkamp, L., “The Commissions White Paper on Environmental Liability: A weak case for an EC Strict Liability Regime”, European Environmental Law Review, 2000, 112-114. Bergkamp, L., “The Commission July 2001 working paper environmental liability: civil or administrative law to prevent and restore environmental harm?”, Environmental Liability, 2001, 207-216. Bergkamp, L., Liability and Environment, The Hague: Kluwer Law International, 2001. Betlem, G., “It’s Chemicals”, Maastricht Journal of European and Comparative Law, 1995, 300. Betlem, G., “Strict Environmental Liability and NGO Damages and Enforcement of Claims: a Dutch International Law Perspective”, European Environmental Law Review, 2001, 314-321. Bier, L., “Aansprakelijkheid voor bedrijfsongevallen en beroepsziekten”, diss. RUU, Deventer, Kluwer, 1988, 62-76. Bierbooms, P.F.A. en De Vries, L.J.A., “Wetsvoorstel collectief actierecht: de rechtspraak aan banden gelegd”, Nederlands Juristenblad, 1993, 96-97. Bishop, W., “The Contract-Tort Boundary and the Economics of Insurance”, Journal of Legal Studies, vol. 12, 1983, 241-266. Bloembergen, A.R., “De invloed van verzekeringen” in: Schade lijden en schade dragen, 1980, pp. 16-17. Bocken, H., “La responsabilité sans faute en droit belge”, in In Memoriam Jean Limpens, 85. B o c k e n , H . , “ Va n f o u t n a a r r i s i c o . E e n o v e r z i c h t v a n d e o b j e c t i ev e aansprakelijkheidsregelingen naar Belgisch recht”, TPR, 1984, 329-415. Bocken, H., “Alternatives to Liability and liability insurance for the compensation of pollution damages”, Tijdschrift voor Milieuaansprakelijkheid, 1987, 83-87. Bocken, H., “Systèmes Alternatives pour l’Indemnisation des Dommages du à la pollution”, Revue Géneral des Assurances et des Responsabilités, 1990, 11698-11714. Bocken, H., “Deficiencies of the system of liability and liability insurance as a mechanism for the indemnification of environmental damage suffered by individual victims”, in Bocken, H. and Ryckbost, D. (eds.), Insurance of Environmental Damage, 133-145. Bocken, H., “Complementary Compensation Mechanisms. A General Environmental Damage Fund?”, in Bocken, H. and Ryckbost, D. (eds.), Insurance of Environmental Damage, 1991, Brussels, Story-Scienta, 425-437. Bocken, H. and Ryckbost, D., (eds.), Verzekering van Milieuschade, Insurance of Environmental Damage, Brussels, Story-Scientia, 1991. Bocken, H., “La réparation des dommages causées par la pollution au Droit belge. La situation en 1992”, Tijdschrift voor Belgisch Burgerlijk Recht, 1992, 284-327. Bocken, H., “L’assurance Responsabilité civile pour dommages causés par la pollution”, in Les Assurances de l’Entreprise, Actes du Collogues tenu à l’Université Libre de Bruxelles les 2 et 3 décembre 1993, Brussels, Bruylant, 1993, 239-280.

240


BIBLIOGRAPHY

Bocken, H. and Ryckbost, D. (eds.), Codification of Environmental Law. Draft Decree on Environmental Policy, London, Kluwer Law International, 1996, 2-158. Bocken, H., Ryckbost, D. and Deloddere, S., “Liability and Financial Guarantees”, in: H. Bocken/D. Ryckbost (eds.), Codification of Environmental Law. Draft Decree on Environmental Policy, 1996, p. 214-223. Bocken, H. and Ryckbost, D., “Deposits”, in Bocken, H. and Ryckbost, D., (eds.), Codification of Environmental Law. Draft Decree on Environmental Policy, London, Kluwer Law International, 1996, 224. Bocken, H., Lambrechts, W., Boes, M., De Nauw, A., Faure, M. and Lavrysen, L., “The Flemish Draft Decree on Environmental Policy: An Outline”, in Bocken, H. and Ryckbost, D. (eds.), Codification of Environmental Law. Proceedings of the International Conference, London, Kluwer, 1996, 11-40. Bocken, H., Ryckbost, D. and Deloddere, S., “Liability and Financial Guarantees”, in Bocken, H. and Ryckbost, D. (eds.), Codification of Environmental Law. Draft Decree on Environmental Policy, London, Kluwer Law International, 1996, 214-223. Bocken, H., Ryckbost, D. and Deloddere, S., “Soil Clean-Up”, in Bocken, H. and Ryckbost, D. (eds.), Codification of Environmental Law: Draft Decree on Environmental Policy, London, Kluwer Law International, 1996, 232-241. Bocken, H., “Rechtstreekse verzekeringen ten behoeve van derden en andere wisseloplossingen voor aansprakelijkheid en aansprakelijkheidsverzekering. Een typologie”, in Liber Amicorum René van Gompel, 1998, 35. Bodewig, Th., “Probleme alternativer Kausalität bei Massenschäden”, Archiv für die civilistische Praxis, 1985, 505-558. Bohrenstein, S., “The Economics of Costly Risk Sorting in Competitive Insurance Markets”, International Review of Law and Economics, 1989, 25-39. Bolt, A.T. and Spier, J. (red.), De uitdijende reikwijdte van de aansprakelijkheid uit onrechtmatige daad, NJV Preadvies, Zwolle, Tjeenk Willink, 1996. Bongaerts, J. and Debièvre, A., “Insurance for Civil Liability for Marine Oil Pollution Damages”, The Geneva Papers on Risk and Insurance, 1987, 145-187. Borch, K., “The Utility Concept Applied to the Theory of Insurance”, The Astin Bulletin, 1961, 245-255. Borch, K., “Recent Developments in Economic Theory and their Application to Insurance”, The Astin Bulletin, 1963, 322-341. Bourgoigny, T., “Responsabilité du fait des produits: arguments connus pour un nouveau débat”, Revue Européenne de Droit de la Consommation, 1987, 17. Boyd, J. and Ingberman, D., “Should “relative safety” be test of product liability”, Journal of Legal Studies, 1997, 433-473. Boyd, J. and Kunreuther, H., “Retroactive Liability or the Public Purse?”, Journal of Regulatory Economics, 1997, 79-90. Brans, E.H.P., Liability for Ecological Damage under the 1992 Protocols to the Civil Liability Convention and the Fund Convention and the Oil Pollution Act of 1990, [1994] TMA, 61-67 and 85-91. Brans, A. and Uilhoorn, M., Liability for ecological damage and assessment of ecological damage, study executed for the Commission to prepare the White Paper (a summary is included in the White Paper, p. 46-48).


241

BIBLIOGRAPHY

B ra n s , E . H . P. , “ N i e u we s t a p i n d e o n t wi k k e l i n g va n e e n E U milieuaansprakelijkheidsrichtlijn”, Aansprakelijkheid, Verzekering en Schade, 2002, 3-11. Brown , J.P., “Toward an Economic theory of liability”, [1973] Journal of Legal Studies (JLS), pp. 323-349. Brüggemeier, G., “Judizielle Schutzpolitik de lege lata – Zur Restrukturierung des BGBDeliktsrechts”, Juristenzeitung, 1986, 972. Brüggemeier, G., “Liability for Water Pollution under German Law: Fault or Strict Liability”, in Van Dunné, J. (ed.), Transboundary Pollution and Liability: the Case of the River Rhine, Lelystad, Vermande, 1991, 88-91. Brunner, C.H.J., “Verplichte procesvertegenwoordiging in civiele zaken”, Advocatenblad, 1980, 473. Buchanan, J.M. and Tullock, G., The calculus of consent, Ann Arbor, University of Michigan Press, 1962. Burrows, P., “Combining regulation and liability for the control of external costs”, International Review of law and economics, vol. 19, 1999, 227-224. Calabresi, G., The Costs of Accidents. A Legal and Economic Analysis, New Haven, Yale University Press, 1970. Coase, R.H., The Problem of Social Cost, [1960] 1 Journal of Law and Economics (JLE), 1-44. Coghlin, T.G., “Protection and Indemnity Clubs”, Lloyd’s maritime and commercial law quarterly, 1984, 403-416. Conruyt-Angenent, H., “L’évolution en droit belge de la réponation des dommages nucléaires à la lumière des conventions internationales”, in Rogge, J. (ed.), Liber Amicorum René van Gompel, Études en assurances, Deurne, Kluwer, 1998, 75-108. Cooter, R., “Prices and Sanctions”, [1984] 84 Columbia Law Review, 1343-1523. Cooter, R. and Ulen, Th., Law and Economics, 3rd edition, Addison-Welseley, 2000. Cousy, H., “Een nieuwe vorm van schuldloze aansprakelijkheid voor schade veroorzaakt door het vreedzaam gebruik van kernenergie”, Jura Falconis, 1974-1975, 46. Cousy, H.A., “Panta Rei": een kort bericht over de wijzigingen van de gewijzigde wetgeving inzake verzekeringen”, Tijdschrift voor Belgisch Handelsrecht, 1995, 470-471. Cousy, H., “Lectuurnotities en reflecties over ethiek in het zakenleven, in het bijzonder in de verzekeringen”, in: F. Fleerackers (ed.), Mens en recht. Essays tussen rechtstheorie en rechtspraktijk, Liber Amicorum Jan M. Broekman, 1996, pp. 47-63. Cousy, H., “De rol van de private verzekeringen in de aanvullende sociale verzekeringen: de regels van het spel”, in: J. van Langendonck (ed.), Liber Amicorum Roger Dillemans, 1997, pp. 55-74. Cousy, H., “Recent developments in environmental insurance”, in Abraham, F., Deketelaere, K. and Stuyck. T. (eds.), Recent economic and legal developments in European environmental policy, 227-241. Cowell, J., “Compulsory environmental liability insurance”, in Bocken H. and Ryckbost, D. (eds.) Insurance of environmental damage, 317-330. Curran, C., “The burden of proof and the liability for suppliers of services in the EEC”, The Geneva Papers on Risk and Insurance, 1994, 85. D a m , C .C . va n en Wes s e l s , B., ( e d s . ) , O pd ra ch t en D ie n st verl e n in g, s e ri e Praktijkhandleidingen, Tjeenk Willink, 1994.

242


BIBLIOGRAPHY

Danzon, P., “Alternative liability regimes for medical injuries”, Geneva Papers on Risk and Insurance, 1999, 3-22. De Boeck, A., “Het voorstel van EG-richtlijn inzake de aansprakelijkheid voor gebrekkige diensten”, Rechtskundig Weekblad, 1993-94, 585. D e B o e r, J. , “ R i s i c o a a n s p r a k e l i j k h e i d v o o r g ev a a r l i j k e s t o f f e n e n milieuverontreiniging”, Nederlands Juristenblad, 1993, 225. De Hoog, P.A., “Naar een werkelijke rechtsbescherming van medisch gelaedeerden”, Nederlands Juristenblad, 1988, 8-10. De Kam, “Privatisering van sociale zekerheid: een werkbaar alternatief?”, in: W.J.P.M. Fase e.a., Sociale Zekerheid: privaat of publiek?, 1994, pp. 52-57. De Leede, “Meer aandacht voor beroepsziekten”, Nederlands Juristenblad, 1998, 1779-1780. De Putter, P. and Verschuuren, J., “Een milieuschadefonds in Nederland?”, Milieu en Recht, 1995, 96-99. De Vroom (ed.), Betwijfelde zekerheden. Reacties op nieuwe risico’s in Nederland, 1998, p. 14. De Witte, B., “Sovereignty and European Integration: the Weight of Legal Tradition”, Maastricht Journal of European and Comparative Law, 1995, 145. De Wolff, D., “Internationaal recht verbiedt privatisering ziektewet”, Nederlands Juristenblad, 1993, 1517. Dedicata, Essays offered to Prof. Mr. J.M. van Dunné, Deventer, Kluwer, 1997, 27-39. Deketelaere, M., “Civielrechtelijke aansprakelijkheid voor milieuschade als instrument van Europees milieubeleid: het “groen boek inzake herstel van milieuschade” van de Commissie van de Europese Gemeenschappen” in Deketelaere, M., (ed.), Recente ontwikkelingen inzake de aansprakelijkheid voor milieuschade, Die Keure, 1993, 99. Deketelaere, K., “De voorstellen van de Interuniversitaire Commissie tot Herziening van het Milieurecht in het Vlaams Gewest inzake aansprakelijkheid voor milieuschade”, in Deketelaere, M. (ed.), Recente ontwikkelingen inzake de aansprakelijkheid voor milieuschade, Bruges, Die Keure, 1993, 55-98. Deketelaere, M., “Aansprakelijkheid bij historische bodemverontreiniging”, in Deketelaere, K. (ed.), Het decreet betreffende de bodemsanering, Bruges, Die Keure, 1995, 135-142. Deketelaere, M., “Het Vlaamse decreet en reglement betreffende de bodemsanering”, in Deketelaere, K. (ed.), Milieurecht in België, Status Questionis Anno 1997, Bruges, Die Keure, 1997, 265-311. Deprimoz, J., “Régime Juridique des Assurances contre les Risques Nucléaires”, Juris. Classeur, 1995, 555, 1-24. Deutsch, E., “Die neuere Entwicklung der Rechtsprechung zum Haftungsrecht”, Juristenzeitung, 1984, 208. Dewees, D., “The Comparative Efficacy of Tort Law and Regulation for Environmental Protection”, The Geneva Papers on Risk and Insurance, 1992, 446-467. Dewees, D., “Tort Law and the Deterrence of Environmental Pollution” in Tietenberg, T.H., (ed.), Innovation in Environmental Policy, Economic and Legal Aspects of Recent Developments in Environmental Enforcement of Liability, Brookfield, Elgar, 1992, 139-164.


243

BIBLIOGRAPHY

Dewees, D, Duff, D. and Trebilcock, M., Exploring the Domain of Accident Law; Taking the Facts Seriously, New York, Oxford, Oxford University Press, 1996. Diamond, P., “Single Activity Accidents”, [1974] JLS, p. 974. Dommering-van Rongen, L., “Risico’s met een lange staart”, in Miscellanea Iurisconsulto Vero. Dommering-van Rongen, L., “Aansprakelijkheid van de werkgever voor toetsenbordletsel ('repetitive strain injury')”, Aansprakelijkheid en Verzekering, 1995, 27. Dommering-Van Rongen, L., Schade Vergoeden door Fondsvorming, Deventer, Kluwer, 1996. Dow, J., “The Organization and Development of International Liability Capacity and National Market Pools, with special reference to new nuclear countries”, in: Nuclear Third Party Liability and Insurance, Munich Symposium, Status and Prospects, Parijs, OECD, 1985, 172-182. Dowding, T., Global Developments in Captive Insurance, FT Financial Publishing, London, 1997. Drion, P. J.M., “Milieu onder één da k: milieuschadeverzekering (MSV)”, Verzekeringsrechtelijke Berichten, 1998/2, 19-21. Ehrlich, I. and Posner, R., “An Economic Analysis of Legal Rule-Making”, Journal of Legal Studies, 1974, 257. Endres, A. and Schwarze, R., “Allokationswirkungen einer Umwelthaftpflichtversicherung”, Zeitschrift für Umweltpolitik und Umweltrecht, 1991, 1-25. Endres, A. and Staiger, B., “Ökonomische Aspekte des Umwelthaftungsrecht” in: M. Ahrens/J. Simon (eds.), Umwelthaftung, Risikosteuerung und Versicherung, 1996, pp. 79-93. ENDS Report, 251, December 1995, 40. Epstein, R.A., “Products Liability as an insurance market”, Journal of Legal Studies, vol. 14, 1985, 645-669. Epstein, R.A., Simple Rules for a Complex World, Cambridge, Harvard University Press, 1995, 221. Estep, E., “Radiation injuries and statistics: the need for a new approach to injury litigation”, Michigan Law Review, 1960, 259-304. Eijk-Graveland, J.C., Verzekerbaarheid van opzet in het schadeverzekeringsrecht, Tjeenk Willink, Zwolle, 1998. Faure, M./R. Van den Bergh, Negligence, Strict Liability and Regulation of Safety under Belgian Law: An Introductory Economic Analysis, 1987, GPRI, 110. Faure, M. and Van den Bergh, R., Objectieve aansprakelijkheid, verplichte verzekering en veiligheidsregulering, Maklu, 1989, 257-264. Faure, M. and Van den Bergh, R., Objectieve aansprakelijkheid, verplichte verzekering en veiligheidsregulering, 1989. Faure, M. and Van den Bergh, R., “Liability for Nuclear Accidents in Belgium from an Interest Group Perspective”, International Review of Law and Economics, 1990, 241-254. Faure, M. and Heine, G., “The Insurance of Fines: the Case of Oil Pollution”, The Geneva Papers on Risk and Insurance, 1991, 39-54.

244


BIBLIOGRAPHY

Faure, M. and Skogh, G., “Compensation for Damages Caused by Nuclear Accidents: A Convention as Insurance”, The Geneva Papers on Risk and Insurance, 1992, 499-513. Faure, M., “Harmonization of product liability law in Europe” in GIRGIS (ed.), L’impresa Europea nel 1992: problemi economici e giuridici nella prospettiva del mercato unico, Il fisco, 1992, 395. Faure, M., Finsinger, J., Siegers, J. and Van den Bergh, R. (eds.), Regulation of professions, a law and economics approach to the regulation of attorneys and physicians in the US, Belgium, the Netherlands, Germany and the U.K., Maklu, 1993. Faure, M., (G)een schijn van kans. Beschouwingen over het statistisch causaliteitbewijs bij milieugezondheidschade, Antwerpen, Maklu, 1993. Faure, M., “De verzekering van het nucleaire risico” in In volle verzekerdheid, Essays offered to Prof.Mr. A.J.O.Baron Van Wassenaer van Catwijck, Zwolle, Tjeenk Willink, 1993, 241-254. Faure, M. and Van den Bergh, R., “Het toelaten van kartels op de Europese verzekeringsmarkt”, Nederlands Juristenblad, 1993, 261-267. Faure, M. and Ruegg, M., “Standard Setting through General Principles of Environmental Law”, in Faure, M., Vervaele, J. and Weale, A. (eds.), Environmental Standards in the European Union in an Interdisciplinary Framework, Antwerp, Maklu, 1994, 39-60. Faure, M. and Oudijk, J.C., “Die strafgerichtliche Überprüfung von Verwaltungsakten im Umweltrecht. Ein rechtsvergleichender Überblick der Systeme in Deutschland, den Niederlanden und Belgien”, Juristenzeitung, 1994, 86-91. Faure, M., “Enkele rechtseconomische kanttekeningen bij de dienstenaansprakelijkheid”, Aansprakelijkheid en Verzekering, 1994, 33. Fa ure, M. /A. J. C.M Ge er s/ T. Ha rtlie f ( eds. ), Ver zeke ri ng en de groe ien de aansprakelijkheidslast. Een juridisch, gezondheidskundig en economisch onderzoek naar ontwikkelingen met betrekking tot de aansprakelijkheidslast en de consequenties voor verzekeraars naar aanleiding van de werkgeversaansprakelijkheid voor bedrijfsongevallen en beroepsziekten, 1995, pp. 38-40. Faure, M./R. Van den Bergh, “Restrictions of Competition on Insurance Markets and the Applicability of EC Anti-Trust Law”, Kyklos, 1995, 65-85. Faure, M., “Economic Models of Compensation for Damage Caused by Nuclear Accidents: Some Lessons for the Revision of the Paris and Vienna Conventions”, European Journal of Law and Economics, 1995, 21-43. Faure, M., “The limits to Insurability from a Law and Economics Perspective”, The Geneva Papers on Risk and Insurance, 1995, 454-462. Faure, M. and Van den Bergh, R., “Restrictions of Competition on Insurance Markets and the Applicability of EC Anti-Trust Law”, Kyklos, 1995, 65-85. Faure, M. and Hartlief, T., “Gevolgen van de uitbreidende werkgeversaansprakelijkheid: beleidsconsequenties voor verzekeraars?”, in Faure, M. and Hartlief, T. (eds.), Verzekering en de Groeiende Aansprakelijkheidslast, een juridisch, gezondheidskundig en economisch onderzoek naar ontwikkelingen met betrekking tot de Aansprakelijkheidslast en de consequenties voor verzekeraars naar aanleiding van de werkgeversaansprakelijkheid voor bedrijfsongevallen en beroepsziekten, Deventer, Kluwer, 1995, 313.


245

BIBLIOGRAPHY

Faure, M./I. Koopmans/J. Oudijk, Imposing criminal liability on Government Officials under environmental law: a legal and economic analysis, Loyola of Los Angeles International Comparative Law Journal, 1996, 529-569. Faure, M. and Hartlief, T., “Compensation Funds versus Liability and Insurance for Remedying Environmental Damage”, Review of European Community and International Environmental Law, 1996, 321-326. Faure, M. and Hartlief, T. , “Towards an Expanding Enterprise Liability in Europe? How to Analyze the Scope of Liability of Industrial Operators and Their Insurers”, Maastricht Journal of European and Comparative Law, 1996, 235-270. Faure, M. and Hartlief, T., “Ontwikkelingen in de werkgeversaansprakelijkheid voor beroepsziekten: aanleiding voor een nieuwe AVB-polis?”, A&V, 1996, 140-151. Faure, M. and Hartlief, T., “Een Asbestfonds als Alternatief voor de Aansprakelijkheid van de Werkgever?”, Tijdschrift voor Sociaal Recht, 1996, 37-43. Faure, M. and Hartlief, T., “Remedies for Expanding Liability”, Oxford Journal of Legal Studies, vol. 18, 1998, 681-706. Faure, M. and Hartlief, T., “Een schadefonds als alternatief voor aansprakelijkheid en verzekering”, RM Themis, 1998, 220-222. Faure, M. and Hartlief, T., Verzekering en financiering van beroepsziekten: enkele tips voor de SER, Nederlands Juristenblad, 1998, 1135. Faure, M., “The applicability of the principles of private insurance to social healthcare insurance seen from a law and economics perspective”, The Geneva Papers on Risk and insurance, 1998, 265-293. Faure, M. and Fenn, P., “Retro active liability and the insurability of long-tail risks”, International Review of Law and Economics, 1999, 487-500. Faure, M. and Hartlief, T., “Het kabinet en de claimcultuur”, Nederlands Juristenblad, 1999, 2007-2015. Faure, M., “Compensation for non pecuniary losses from an economic perspective”, Magnus, U and Spier, J. (eds.), European Tort Law, Liber Amicorum for Helmut Koziol, Frankfurt am Main, Peter Lang, 2000, 143-159. Faure, M., “Regulation of attorneys in Belgium” in Faure, M. et al, Regulation of Professions, 97-98. Faure, M. and Koziol, H. (eds), Torts and Insurance Law, vol.1, Cases on Malpractice in a Comparative Perspective, Wien: Springer, 2001. Faure, M. and Hartlief, T., Nieuwe risico’s en vragen van aansprakelijkheid en verzekering, Deventer: Kluwer, 2002. Faure, M. and Van den Bergh, R., “Aansprakelijkheidsverzekering, concurrentie en ongevallenpreventie”, in Hartlief, T. and Mendel, M.M. (eds.), Juridische beschouwingen over de maatschappelijke rol van verzekeringen en verzekeringsmaatschappijen, E.M. Meijers Instituut, 2000, 315-342, to be published. Faure, M. and Hartlief, T., in Spier, J., Yearbook on Insurance and Liability, to be published. Ferejohn/Ch. Shipan, Congressional Influence on Bureaucracy, Journal of Law, Economics and Organization, 1990, 6, 1-20. Fontaine, M. and Bourgoignie, T., Le droit de la consommation en Belgique et au Luxembourg, Van Nostrand Reinhold, 1981, 130.

246


BIBLIOGRAPHY

Frank, “Lessons from the Great Battle: Health Care Reform, 1992-1994”, 1997, Arch. Phys. Med. Rehabil, 122-123. Frenk, N., Kollektieve akties in het privaatrecht, diss. RUU, Kluwer, 1994. Frenk, N., “Toerekening naar kansbepaling”, Nederlands Juristenblad, 1995, 482-491. Frenk, N., “De directe schadeverzekering als vervanging van aansprakelijkheid”, Nederlands Juristenblad, 1999, 15-47. Fuchs, “From Bismarck to Woodcock: the “irrational” Pursuit of National Health Insurance”, Journal of Law and Economics, 1976, 347-359. Gardner, M., “Results of a case-control study of leukaemia and lymphoma among young people near Sellafield nuclear plant in West Cumbria”, British Medical Journal, 1990, 423-434. Giesen, I., “Werkgeversaansprakelijkheid, blootstelling aan asbest en bewijslastverdeling op basis van de strekking van de materiele norm”, Nederlands Tijdschrift voor Burgerlijk Recht, 2001, 242-247. Gilead, I., Tort Law and Internalization. The Gap between Private Loss and Social Cost, [1997] 17 International Review of Law and Economics (IRLE), 589-608. Gilhuis, P. and Verschuuren, J., “Een Milieuschadefonds in Nederland; een Onderzoek naar de Mogelijkheden”, Publicatiereeks Milieubeheer, 1994/3. Gilles, S., “Rule-Based Negligence and the Regulation of Activity Levels”, [1992] 22 JLS, 319. Gimpel-Hinteregger, M., Grundfragen des Umwelthaftung, (1994). Gordon Tullock. See e.g. his Trials on trial, Columbia University Press, 1980. Grimeaud, D.J.E., “An overview of the policy and legal aspects of the international climate change regime”, Environmental Liability, 2001, 39-52. H a a z e n , O. A . a n d S p i e r, J. , “ A m e r i k a a n s e t o e s t a n d e n e n d e n i e u w e aansprakelijkheidsverzekering voor bedrijven en beroepen”, Nederlands Juristenblad, 1996, 45-50. Haazen, O.A. and Spier, J., in De uitdijende reikwijdte van de aansprakelijkheid uit onrechtmatige daad, 56-79. Hamburger, M., “De aansprakelijkheidsverzekering in Nederland voor schade door kernongevallen in West-Europa”, NJB, 1996, 1340. Hankey, S., “Claims Made Policies and Choice of Law in the European Union”, International Insurance Law Review, 1994, 267. Hartlief, T. and Tjittes, R.P.J.L., “De aansprakelijkheid voor bedrijfsongevallen en – ziekten – recente ontwikkelingen met betrekking tot tewerkstelling en bewijslast”, Sociaal Recht, 1990, 282-288. Hartlief, T. and Tjittes, R.P.J.L., “Invloed van verzekering op de civiele aansprakelijkheid”, Preadvies voor de Vereniging voor Burgerlijk Recht, Lelystad, Vermande, 1990. Hartlief, T. and Spier, J., “Verzekeringen en aansprakelijkheid met “terugwerkende kracht””, Aansprakelijkheid en Verzekering, 1994, 27-33. Hartlief, T. and Tjittes, R.P.J.L., Verzekering en Aansprakelijkheid, Deventer, Kluwer, 1994. Hartlief, T., Ieder draagt zijn eigen schade, 1997, p. 28.


247

BIBLIOGRAPHY

Hartlief, T. and Mendel, M.M. (eds.), Juridische beschouwingen over de maatschappelijke rol van verzekeringen en verzekeringsmaatschappijen, 2000, pp. 315-342. Hartlief, T., “Open normen in het schadevergoedingsrecht; de artikelen 6:98, 99 en 101 BW”, Weekblad voor Privaatrecht, Notariaat en Registratie, 2002, 8-19. Hausmann, H. and Kraakman, R.H. “Toward unlimited shareholder liability for corporate torts”, Yale Law Journal, 1991, 1879. Havens, C.W. III and Theisen, R.M., “The Application of United States and ECC Antitrust Laws to Reinsurance and Insurance Pooling Arrangements”, The Antitrust Bulletin, 1986, 1301. Heimann, F., “The US Liability Protection System for Nuclear Power Plants”, paper presented at the symposium “Nuclear accidents, liability and guarantees”, Helsinki, Finland, September 1992. Hellingman, K., “An economic analysis of the regulation of lawyers in the Netherlands” in Faure, M. et al, Regulation of Professions, 167-169. Heyes, A. and Liston-Heyes, C., “Capping Environmental Liability: the Case of North American Nuclear Power”, The Geneva Papers on Risk and Insurance, 2000, 196. Hinteregger, M., “La nouvelle loi autrichienne sur la responsabilité civile pour les dommages nucléaires”, Bulletin de Droit Nucléaire, vol. 62, 1998, 27-34. Hogarth, R. and Kunreuther, H., “Ambiguity and Insurance Decisions”, American Economic Review, vol. 75, 1985, 386-390. Hondius, E.H., “Produktenaansprakelijkheid: de voordelen van een dualistische rechtsorde”, Ars Aequi 1996, 38. Hugenholtz, W. and Heemskerk, W.H., Hoofdlijnen van Nederlands Burgerlijk Procesrecht, Lemma, 1996, 113-118. Hulst, E., Grondslagen van Milieuaansprakelijkheid, diss., 1993. Hulst, E., “De werkelijkheid rondom een algemeen milieuschadefonds, een commentaar”, Tijdschrift voor Milieuaansprakelijkheid, 1995, 167-173. Janssen, C.A., “Aansprakelijkheid voor milieuschade en financiële zekerheid naar toekomstig recht: nieuwe oplossingen. Nederlands Recht” in Wiggers-Rust, L.F., en Deketelaere, K. (eds.), Aansprakelijkheid voor milieuschade en financiële zekerheid, Die Keure – Vermande, 1998, 97-120. Johnson, “Choice of Compulsory Isurance under Adverse Selection”, Public Choice, 1977, 23-25. Jones, B., “The Identification and Remediation of Contaminated Sites: The United Kingdom’s Environment Act 1995”, Tijdschrift voor Milieuaansprakelijkheid, 1995, (159), 163. Jones, B., “Deterring, Compensating and Remedying Environmental Damage: the Contribution of Tort Liability”, in: P. Wetterstein (ed.), Harm to the Environment: The right to compensation and the assessment of damages (1997), pp. 11-27. Jost, P.J., “Limited liability and the requirement to purchase insurance”, International Review of Law and Economics, 1996, 259-276. Katzman, M., “Pollution Liability Insurance and Catastrophic Environmental Risk”, Journal of Risk and Insurance, 1988, 75-100.

248


BIBLIOGRAPHY

Kerkhoff, “The Physicians’ Fraternity: Safeguard for Quality or Cartel?”, in: L.J. Gunning-Schepers/G.J. Kronjee/R.A. Spasoff (eds.), Fundamental Questions About the Future of Health Care, 1996, p. 117. Kerremans, H., “Aansprakelijkheid voor Milieuschade en verzekeringsmogelijkheden in Milieuzorg in de Onderneming”, I., Juridische, fiscale en organisatorische aspecten, Antwerpen, Standaard, pp. 537-583. J. van Kessel, “Schade bij werknemers: aandachtspunten voor deze eeuw. Betere werking van het huidige stelsel door nauwere afstemming op schadeleer, meer flexibiliteit en verdergaande integratie”, in S. Klosse (ed.) Sociale zekerheid: een ander gezichtspunt. Toekomstperspectief vanuit vier disciplines, 2000, pp. 257-260. Kidner, R., Casebook on Torts, Oxford University Press, 2002 7th.ed. Klik, P., “Het wetsvoorstel vorderingsrecht van belangenorganisaties bij de Eerste Kamer. Terug naar af”, Tijdschrift voor Milieuaansprakelijkheid, 1994, 34. Klosse, S., “Bedrijfsongevallen en beroepsziekten: individuele of collectieve verantwoordelijkheid”, Nederlands Juristenblad, 1993, 1608-1613. S. Klosse, “Schadeleer: geschikt fundament voor sociale zekerheid in de 21e eeuw?”, in S. Klosse (ed.) Sociale zekerheid: een ander gezichtspunt. Toekomstperspectief vanuit vier disciplines, 2000, p. 12. Koch, B.A. and Koziol, H. (eds.), Unification of Tort Law: Strict Liability, London, Kluwer Law International, 2002. Kolstad, Ch.D., Ulen, Th.S. and Johnson, G.V., “Ex Post Liability for Harm vs. Ex Ante Safety Regulation: Substitutes or Compliments?”, American Economic Review, 1990, Vol. 80, 888-901. Köndgen, J., “Multiple causation and joint tortfeasors in pollution cases according to German law”, in Van Dunné, J.M. (ed.), Transboundary Pollution and Liability, the Case of the River Rhine, Lelystad, Vermande, 1991, 99-106 Kornhauser, L. and Revesz, R., “Sharing Damages among multiple Tortfeasors”, Yale Law Journal, 1989, 831-884. Kornhauser, L. and Revesz, R., “Apportioning damages among potentially insolvent actors”, Journal of Legal Studies, 1990, 617-651. Kottenhagen-Edzes, P.A., “Onrechtmatige daad en milieu. Het gebruik van het privaatrecht bij het voorkomen van milieu-aantasting en het verhaal van milieuschade”, diss. EUR, Arnhem, Gouda Quint, 1992, 167-168. Kötz, A., “A common private law for Europe: perspectives for the reform of European legal eduction” in De Koziol, H., Die Arzthaftung im geltenden und künftigen Recht in: Haftungsrechtliche Perspektiven der ärztlichen Behandlung, 1997, pp. 21-35. Kremer, F.Th., “Moeten alle buiten-gerechtelijke kosten vergoed worden; en zo ja, waarom niet?” in Wansink, J.H. (ed.), Een Salomons Oordeel, Tjeenk Willink, 1991, 25-31. Kunreuther, H. and Freeman, P. in A. Heyer, (ed.), The Law and Economics of the Environment, 316. Kunreuther, H., Hogarth, R. and Meszaros, J., “Insurer Ambiguity and Market Failure”, Journal of Risk and Uncertainty, 1993, 71-87. Landes, W. and Posner, R., “The Positive Economic Theory of Tort Law”, [1981] Georgia Law Review (GLR), 851-924


249

BIBLIOGRAPHY

Landes, W. and Posner, R., “Tort law as a regulatory regime for catastrophic personal injuries”, Journal of Legal Studies, 1984, 417-434. Lavrysen, L., “Judicial responses in the nineties to Dutch (and German) shipments of waste to Belgium in the eighties”, Maastricht Journal of European and Comparative Law, 1995, 219-243. Le Blanc, Bismarck als benchmark. Enkele rechtseconomische beschouwingen over publieke sector en sociale zekerheid in Nederland, Ars Aequi, 1990, 654-659. Levie, G. and Cousy, H. (eds.), La politique Européenne de concurrence en matière d’assurance, Bruylant, Brussel, 1994. Lindenbergh, Arbeidsongevallen en beroepsziekten, 2000. Mackaay, E., Economics of information and the law, Boston, Kluwer, 1982. Margereson and Hancock v. J.W. Roberts, “Queens Bench Division, Leeds, 27 October 1995”, Product Liability International, 1995, 179. Markesinis, B.S., The German Law of Torts, A Comparative Introduction, 880-881. Marrone, J., “Closing the Circle of Protection for the Public-the Evolution of the System in the United States”, paper presented at the symposium “Nuclear accidentsliability and guarantees”, Helsinki, Finland, September 1992. Marshall, J. “Moral Hazard”, American Economic Review, 1976, 880-890. Messer, E.A., Risicoaansprakelijkheid voor milieuverontreiniging in het WW, diss., RUU, Gouda Quint, 1994. Mitchell/S. Zeldes, Social Security Privatisation: A Structure for Analysis, 1996, p. 11. Mölenberg, L.J.H., Het collectief actierecht voor consumentenorganisaties op het terrein van de algemene voorwaarden, Gouda Quint, 1995. Monti, A., “Environmental Risk: A comparative law and economics approach to liability and insurance”, European Review of Private Law, forthcoming. Morris, J., (eds.), Rethinking Risk and the Precautionary Principle, Oxford, Butterworth, 2000. Müller, W., “The role of insurance industry in covering nuclear third party liability risks”, in: Nuclear Third Party Liability and Insurance, Munich Symposium, Status and Prospects, Parijs, OECD, 1985, 166-171. Mullis, A. and Oliphant, K., Torts, MacMillan, 1993, 132-142. Nagareda, R.A., “Autonomy, Peace and Put-Options in the Mass Tort Class Action”, Harvard Law Review, 2002, 749-829. Newhouse, J., “Reimbursing Health Plans and Health Providers: Selection versus Efficiency in Production”, Journal of Economic Literature, 1996, 34, 1236-1263. Nieuwenhuis, J.H., “De dag verga, waarop ik geboren werd”, RM Themis, 2001, 97-98. Niezen, J., “Nieuwe milieuschadeverzekering – geen panacee”, Milieu en Recht, 1998, 114. Niezen, G.J., “Aansprakelijkheid voor milieuschade in de Europese Unie” in Ongebonden Recht Bedrijven, Kluwer, 2000, 165-169. O’Riordan, T., Cameron, J. and Jordan, A., (eds.), Reinterpreting the precautionary principle, London, Cameron, May, 2001.

250


BIBLIOGRAPHY

OECD, Liability and Compensation for Nuclear Damage, An International Overview, Paris, OECD, 1994. Ogus, A.I., “Quantitative Rules and Judicial Decision-Making”, in Burrows, P. and Veljanovski, C. (eds.), The Economic Approach to Law, London, Butterworth, 1981, 210-225. Ogus, A.I., “Standard Setting for Environmental Protection: Principles and Processes”, in Faure, M., Vervaele, J. and Weale, A. (eds.), Environmental Standards in the European Union, 1994, 25-37. Olson, M., The logic of collective action, Harvard University Press, Cambridge, 1971. Pagh, P., “The New Danish Act on Strict Liability for Environmental Damage”, Environmental Liability, 1995, 15. Pauly, M., “The Economics of Moral Hazard: Comment”, American Economic Review, 1968, 531-545. Pauly, M., “Overinsurance and the Public Provision of Insurance: The Roles of Moral Hazard and Adverse Selection”, Quarterly Journal of Economics, 1974, 44-62. Pauly, M., “Taxation, Health Insurance and Market Failure in the Medical Economy”, Journal of Economic Literature, 1986, 629-675. Pauly, M., “Competition in Health Insurance Markets”, Law and Contemporary Problems, 1988, 237-271. Pieters, D. (ed.), Introduction into the social security law of the member States of the European community, 1993. Pieters, D., Introduction in the basic principles of social security, 1993. Pieters (ed.), Introduction into the social security law of the member States of the European community, 1993. Polborn, M., “Mandatory Insurance and the Judgement-Proof Problem”, International Review of Law and Economics, 1998, 141-146. Polinsky, A.M., Introduction to Law and Economics, (1983). Popp, A. “Liability and compensation for pollution damage caused by ships revisited – report on an international conference”, Loyds Maritime and Commercial Law Quarterly, 1985, 118-131. Posner, R., “An Economic Approach to Legal Procedure and Judicial Administration”, Journal of Legal Studies, 1973, 399. Posner, R., “Theories of Economic Regulation”, Bell Journal of Economics, 1974, 335-358. Posner, R., Economic Analysis of Law, 5th edition, New York, Aspen Law and Business, 1998. Pozzo Zanchetta, B., The Liability Problem in Modern Environmental Statutes, [1996] 4 ERPL, 112-129. Pratt, J., “Risk Aversion in the Small and in the Large”, Econometrica, 1964, 122-136. Priest, G., “The Current Insurance Crisis and Modern Tort Law”, Yale Law Journal, 1987, 1521-1590. Radetzki, M., “Private arrangements to cover large-scale liabilities caused by nuclear and other industrial catastrophes”, The Geneva Papers on Risk and Insurance, 2000, 180-195.


251

BIBLIOGRAPHY

Ranson, D., “Verzekering van milieuaansprakelijkheid”, Milieu- en Energierecht, 2000, 66-73. Rehbinder, E., “Towards a community environmental liability regime: the Commission’s White Paper on environmental liability”, Environmental Liability, 85-96. Reitsma, S.M.S., “Nuclear Insurance Pools: History and Development”, paper presented at the OECD Symposium on Nuclear Accidents, Liabilities and Guarantees, Helsinki, Finland, September 1992. Rémard-Gouillond, M., “Faute ou Risque?” in Van Dunné, J.M. (ed.), Transboundary Pollution and Liability, Vermande, 1991, 67. Restatement of the Law. Torts: Liability for Physical Harm (Basic Principles), Tentative Draft No. 1, The American Law Institute, March 28, 2001. Revesz, R. and Stewart, R. (eds.), Analysing Superfund, Economics, Science and Law, Washington, Resources for the Future, 1995. Richardson, B.J., “Financial Institutions for Sustainability”, Environmental Liability, 2000, 52-64. Rice, P., “From Lugano to Brussels via Arhus: environmental liability White Paper published”, Environmental Liability, 2000, 39-45. Robesin, M., “Aansprakelijkheid voor kernongevallen, ruim een jaar na Tsjernobyl”, Milieu en Recht, 1987, 228-229. Robinson, G., “Probabilistic causation and compensation for tortious risk”, Journal of Legal Studies, 1985, 798. Rogers, W.V.H., Spier, J. and Viney, G., in Spier, J. (ed.), The Limits of Liability, 13. Rogers, W.V.H., “Liability for Environmental Pollution in the Common Law: The Cambridge Water Case”, Aansprakelijkheid en Verzekering, 1994, 64. Rogers, W.V.H., Spier, J. and Viney, G., “Preliminary Observations” in Spier, J., (ed.), The limits of liability. Keeping the Floodgates Shut, Kluwer law, 1996, 1-15. Rogge, J., Les assurances en matière d’environnement, Loose-Leaf, Kluwer, 1997. Rose-Ackerman, S., “Public Law versus Private Law in Environmental Regulation: European Union Proposals in the Light of United States Experience”, Review of European Community and International Environmental Law, RECIEL, vol. 4, 312-32. Rose-Ackerman, S., “Regulation and the Law of Torts”, American Economic Review, Papers and Proceedings, 1991, 54-58. Rose-Ackerman, S., “Environmental Liability Law”, in Tietenberg, T.H., (ed.), Innovation in Environmental Policy, Economic and Legal Aspects of Recent Developments in Environmental Enforcement and Liability, Brookfield, Edward Elgar, 1992, 223-243. Rose-Ackerman, S., Rethinking the Progressive Agenda, the Reform of the American Regulatory State, New York, The Free Press, 1992. Rose-Ackerman, S., Controlling Environmental Policy: the Limits of Public Law in Germany and the United States, 1995, Yale University Press, New Haven and London. Rosenberg, D., “The causal connection in Mass Exposure cases: a public law vision of the tort system”, Harvard Law Review, 1984, 851-929.

252


BIBLIOGRAPHY

Rothschild and J. Stiglitz, “Equilibrium in Competitive Insurance Markets: an Essay on the Economics of Imperfect Information”, Quarterly Journal of Economics, 1976, 629-649. Rowden, M., Kraemer, J. and Cuoco, L., “La révision de 1988 de la loi Price-Anderson ou mieux vaut tard que jamais”, Bulletin de droit nucléaire, Nr. 42, 79-101. Schmidt-Salzer, J., “Unternehmens- und Mitarbeiterhaftung im deutschen und europäischen Produkt- und Umwelthaftungsrecht” in Ahrens, M. and Simon, J. (eds.), Umwelthaftung, Risikosteuerung und Versicherung, Erich Schmidt Verlag, 1996, 59. Schmitz, On the joint use of liability and safety regulation, IRLE, 2000, 371-382. Schneider, H., Die Anerkennung von Diplomen in der Europäischen Gemeinschaft, Maklu, 1995, 265-350. Schoordijk, H.C.F., “Wrongful life mede vanuit rechtsvergelijkend perspectief”, Nederlands Tijdschrift voor Burgerlijk Recht, 2001, 212-218. Schoorens, “Segmentering en discriminatie”, in H. Cousy/H. Claassens/C. van Schoubroek (eds.), Competitiviteit, ethiek en verzekering, 1998, p. 217-277. Schwartz, A. and Wilde, L., “Intervening in markets on the basis of imperfect information: a legal and economic analysis”, University of Pennsylvania Law Review, 1979, 630-682. Schwartz, A., “Statutory Interpretation, Capture, and Tort Law: The regulatory compliance defense”, American Law and Economics Review, 2000, 1-57. Schwartz, G., “The Ethics and the Economics of Tort Liability Insurance”, Cornell Law Review (CLR), 1990, 313-365. Schwartz, G., “Mixed theories of tort law: affirming both deterrence and corrective justice”, Texas Law Review, 1997, 95, 1804-1834. Shavell, S., “On Moral Hazard and Insurance”, Quarterly Journal of Economics, 1979, 541-562. Shavell, S., “Strict Liability versus Negligence”, Journal of Legal Studies, 1980, 1-25. Shavell, S., “Suit, Settlement and Trial: A Theoretical Analysis under Alternative Methods for the Allocation of legal Costs”, Journal of Legal Studies, 1982, 55. Shavell, S., “Liability for Harm versus Regulation of Safety”, Journal of Legal Studies, 1984, 357-374. Shavell, S., “A Model of the Optimal Use of Liability and Safety Regulation”, Rand Journal of Economics, 1984, 271-280. Shavell, S., “Criminal law and the optimal use of non-monetary sanctions as a deterrent”, Columbia Law Review, 1985, 1232-1262. Shavell, S., “Uncertainty over causation and the determination of civil liability”, Journal of Law and Economics, 1985, 587-609. Shavell, S., “The judgement proof problem”, International Review of Law and Economics, 1986, 43-58. Shavell, S., Economic Analysis of Accident Law, Cambridge, Harvard University Press, 1987. Sieswerda, P., “(On)voorwaardelijke toevoeging en de vergoeding van pre-processuele kosten”, Advocatenblad, 1987, 513.


253

BIBLIOGRAPHY

Simon, J., “Das Umwelthaftungsgesetz” in Ahrens, M. and Simon, J. (eds.), Umwelthaftung, Risikosteuerung und Versicherung, Erich Schmidt Verlag, 1996, 13. Skogh, G., “Public insurance and accident prevention”, The International Review of Law and Economics, 1982, 2, 67-80. Skogh, G., “The Transactions Cost Theory of Insurance: Contracting Impediments and Costs”, Journal of Risk and Insurance, 1989, 726-732. Skogh, G., “The combination of private and public regulation of safety”, in M. Faure/ R. Van den Bergh (eds.), Essays in law and economics. Corporations, accident prevention and compensation for losses, 1989, pp. 87-101. Skogh, G., “Development Risks, Strict Liability and the Insurability of Industrial Hazards”, The Geneva Papers on Risk and Insurance, 1998, 247-265. Skogh, G., “Mandatory insurance: transaction costs analysis of insurance”, in Bouckaert, B. and De Geest, G. (eds.), Encyclopedia of Law and Economics, II, Civil Law and Economics, Cheltenham, Edward Elgar, 2000, 521-537. Smets, H., “Pour une indemnisation garantie des victimes de pollution accidentelle”, in Bocken, H. and Ryckbost, D. (eds.), Insurance of Environmental Damage, Brussels, Story-Scientia, 1991, 397-423. Smets, H., “COSCA: A complementary system for compensation of accidental pollution damage”, in Wettenstein, P, (ed.), Harm to the environment: The right to compensation and the assessment of damage, Oxford, Clarendon Press, 1997, 223-248. Smith, R.E., Canelo, E.A. and Di Dio, A., “Reinventing Reinsurance using the Capital Markets”, Geneva papers on Risk and Insurance, 1997, 26-37. Snoep, M., “NMa mist kans om eind te maken aan discussie”, Advocatenblad, 2002, 194-198. Spence, M. and Zeckhauser, R., “Insurance, Information, and Individual Action”, American Economic Review, 1971, 380-391. Spier, J., Sluipende schade, Deventer, Kluwer, 1990. Spier, J., “Wederom: De EEG en gebrekkige diensten”, Nederlands Juristenblad, 1991, 663. Spier, J., De maalstroom van het aansprakelijkheidsrecht, Houthoff-lecture, 24 March 1992. Spier, J. and Sterk, C.H.W.M., “The draft convention on civil liability for damage resulting from activities dangerous to the environment”, Tijdschrift voor milieu en recht, 1992, 591. Spier, J., “Green boek on remedying environmental damage”, Aansprakelijkheid en Verzekering, 1993, 13. Spier, J., “De Novelle inzake art. 47 lid 5 ontwerp inbouwwet bodemsanering”, Aansprakelijkheid en Verzekering, 1993, 43-4. Spier, J. and Wansink, J.H., “Joint and Several Liability of DES Manufacturers: A Dutch Tort Crisis”, International Insurance Law Review, 1993, 176-181. Spier, J., “De grenzen der buitengerechtelijke kosten”, Aansprakelijkheid en Verzekering, 1995, 55. Spier, J. and Sterk, C.H.W.M., Aansprakelijkheid voor gevaarlijke stoffen, Kluwer, 1995. Spier/O.A. Haazen, Aansprakelijkheidsverzekeringen op claims’ made grondslag, 1996, p. 125.

254


BIBLIOGRAPHY

Spier, J. (ed.), The Limits of Liability, Keeping the Floodgates Shut, The Hague, Kluwer, 1996, 123-125. Spier, J. and Haazen, O.A., Aansprakelijkheidsverzekeringen op claims’ made grondslag, Deventer, Kluwer, 1996. Spier, J., “How to keep liability within reasonable limits? A brief outline of Dutch law” in Spier, J. (ed.), The Limits of Liability, 98-99. Spier, J., Een nieuwe dageraad voor het aansprakelijkheidsrecht?, Deventer, Tjeenk Willink, 1999. Spier, J., De uitdijende reikwijdte van het aansprakelijkheidsrecht, Preadvies Nederlandse Juristenvereniging, Chapter 2. Sterk, C.H.W.M., Verhoogd gevaar in het aansprakelijkheidsrecht, diss. KUB, Kluwer, 1994. Stapleton, J., Product liability, Butterworths, 1994. Stolker, C.J.J.M., “Aansprakelijkheid voor beroepsfouten; van droomcarrière naar nachtmerrie”, Ars Aequi, 1995, 13. Storm, P., “Een gebrekkig produkt”, Tijdschrift voor Vennootschappen, Verenigingen en Stichtingen, 1985, 245. Swuste, P., “Van Individuele ’Asbestprocessen’ naar een Asbestfonds?”, Arbeidsomstandigheden, 1996, 119-122. Taams, R. and Uithoorn, M., “The Dutch Soil Pollution Cases: Shell, Sovay, Duphar and Fasson/Van den Brink”, Environmental Liability, 1995, 39. Teunissen, J.M.H.F., Het Burgerlijk kleed van de Staat, diss. RL, Tjeenk Willink, 1996, critically discussing the new provisions in the Civil Code. Tietenberg, T.H., “Indivisible toxic torts: the economics of joint and several liability”, Land Economics, vol. 65, 1989, 305-319. Tillemann, W.A., Lucas, A.R. and Hughes, E., Environmental law and policy, Toronto: Edmond Montgomery Publications, 1998. Tjittes, Hollandse toestanden, 1999, RMThemis, 141-142. Trauberman, J., “Statutory reform of toxic torts. Relieving legal, scientific and economic burdens on the chemical victim”, in Harvard Environmental Law Review, 1983, 177-296. Trebilcock, M.J., “The Social Insurance-Deterrence Dilemma of Modern North American Tort Law: A Canadian Perspective on the liability Insurance Crisis”, San Diego Law Review, Vol. 24, 1987, 929-1002. Trebilcock, M. and Winter, R., “The Economics of Nuclear Accident Law”, International Review of Law and Economics, 1997, 215-243. Tyran, J.R. and Zweifel, P., “Environmental risk internalization through capital markets (Ericam): the case of nuclear power”, International Review of Law and Economics, 1993, 431-444. Van, A., “Statistisch bewijs van causaal verband”, Discussion of High Court of Justice 8 October 1993 (Sellafield Claims), Tijdschrift voor Milieuaansprakelijkheid, 1994, 109-118. Van, A., Onzekerheid over daderschap en causaliteit, Arnhem, Gouda Quint, 1995, 145-154.


255

BIBLIOGRAPHY

Van Boom, W.H., “Anticiperen op nieuwe gezondheidsrisico’s”, Aansprakelijkheid, Verzekering en Schade, 2001, 3-12. Van Dam, C.C., “De EG-richtlijn dienstenaansprakelijkheid”, Aansprakelijkheid en Verzekering, 1994, 113. Van Dam, C.C., in Opdracht en Dienstverlening, Series praktijkhandelingen, Tjeenk Willink, 1994. Van den Borre, T., “Transplantatie van kanalisatie van aansprakelijkheid van het kernenergierecht naar het milieu (aansprakelijkheids-)recht: een goede of een gebrekkige zaak?”, in Faure, M. and Deketelaere, K. (red.), Ius Commune en Milieurecht, Actualia in het Milieurecht in België en Nederland, Antwerpen, Intersentia, 1997, 329-382. Van den Borre, T., “Dekking van het nucleaire risico op nationaalrechtelijke basis of via internationale verdragen: de Verenigde Staten versus Europa”, in Deketelaere, K., Faure, M., en Verhoosel, G. (eds.), Grensoverschrijdende milieuproblemen: uitdagingen voor de nationale en internationale rechtsorde, Antwerpen, Intersentia, 1998, 443-490. Vanden Borre, T., Recente ontwikkelingen in het nucleaire aansprakelijkheidsrecht: innovatie of renovatie?, Milieu- en Energierecht (MER), 2000, pp. 101-111 and pp. 225-246. Vanden Borre, T., Efficiënte preventie en compensatie van catastroferisico’s. Het voorbeeld van schade door kernongevallen, 2001, pp. 323-326. Van de Griendt, “De Amerikaanse paradox”, Sociaal bestek, 1995, ⅞, 9-12. Van den Bergh, R., “The Subsidiarity Principle in European Community Law: Some Insights from Law and Economics”, Maastricht Journal of European and Comparative Law, 1994, 337. Van der Heijden, P.F., “Privatisering van de sociale zekerheid: begrensd door (internationale) sociaalrechtelijke normen” in: W.J.P.M. Fase (ed.), Sociale zekerheid: Privaat of Publiek, 1994, p. 58. Van der Schans, E., “Contingency fees”, Advocatenblad, 1994, 663. Van Dunné, J.M., “De rechtspraak inzake milieuaansprakelijkheid uit onrechtmatige daad: van schuldbeginsel naar risicobeginsel”, Tijdschrift voor Milieuaansprakelijkheid, 1988, 33. Van Dunné, J.M., “Een kamikazeactie op De Rotte. De visie van Vranken op de aansprakelijkheid uit art. 1401 BW, in het bijzonder bij bodemvervuiling uit het verleden”, Weekblad voor Privaatrecht Notariaat en Registratie, (not dated), 5976. Van Dunné, J.M., “Environmental liability: continental style”, Review of European Community and International Environmental Law, 1992, 396. Van Dunné, J.M., “Schending van veiligheidsnormen en causaliteit: conditio sine qua non als rudimentair vereiste”, Tijdschrift voor Milieuaansprakelijkheid, 2001, 1-13. Van Gerven, W., “Bridging the unbridgeable”, International Comparative Law Quarterly, 1996, forthcoming. Van Huizen, F., “Enkele begrenzingen van de (beroeps-)aansprakelijkheidsverzekering”, in Drie treden (De Ruiter-bundel), Zwolle, Tjeenk Willink, 1995, 325-339. Van Maanen, G.E., “Pleidooi voor verbetering van de rechtspositie van slachtoffers van kernongevallen”, Nederlands Juristenblad, 1986, 1342.

256


BIBLIOGRAPHY

Van Maanen, G.E., Spier, J. and Sterk, C.H.W.M., “Van schuld naar risico? Enkele opmerkingen over ontwikkelingen in het aansprakelijkheidsrecht” in Raaymakers, M.J.G.C. (eds.), Aansprakelijkheden, Kluwer, 1990, 65. Van Maanen, G.E., “De civielrechtelijke aansprakelijkheid voor kernongevallen naar Nederlands Recht”, in Faure, M. (ed.), Aansprakelijkheid voor het Nucleaire Risico, Antwerp, Maklu, 1993, 19-36. Van Maanen, G.E., “Onrechtmatig leven”, Nederlands Tijdschrift voor Burgerlijk Recht, 2001, 57. Van Manen, “De paradox van individualisering en collectivisering” in: N.F. van Manen/ R.H. Stutterheim (eds.), Wie draagt de schade?, 1998, pp. 95-113. Van Mierlo, “Hervorming van het Nederlandse stelsel van sociale zekerheid: financiële consequenties en politieke haalbaarheid”, in: F.A.J. van den Bosch/C. Peterson (eds.), Economie en arbeidsongeschiktheid: analyse en beleid, 1983, pp. 215-241. Van Mierlo, “Een alternatief voor het Plan-Simons? Overwegingen vanuit de welvaartstheorie”, ESB, 1991, 1164. Van Schoubroek, C. and Schoorens, G., “De aansprakelijkheidsverzekering: a never ending story”, Tijdschrift voor Belgisch Handelsrecht, 1995, 644-660. Van Steenberge/S. Klosse/L.J.M. de Leede (eds.), Preventie: een solide basis voor sociale zekerheid?, 1994. Van Wijngaarden/F.G. van den Heuvel/J.C. Vrooman (eds.), Preventie in de sociale zekerheid, 1992. Viscusi, W.K., “The Dimensions of the Product Liability Crisis”, Journal of Legal Studies, 1991, 147-177. Von Bar, C., The Common European Law of Torts, (vol.2), Oxford: Clarendon Press, 2000. Voorontwerp Decreet Milieubeleid, 1995, pp. 943-985. Vranken, J.B.M., “Zorgvuldigheidsnorm en aansprakelijkheid voor bodemverontreiniging uit het verleden”, Weekblad voor Privaatrecht Notariaat en Registratie (not dated), 5953. Vranken, J.B.M., “Wrongful life in Frankrijk: drie recente uitspraken”, Weekblad voor Privaatrecht, Notariaat en Registratie, 2001, 752-755. Wagner, G., “Umwelthaftung und Versicherung”, Versicherungsrecht, 1991, 249-260. Wagner, G., “Die Aufgaben des Haftungsrechts – eine Untersuchung am Beispiel des Umwelthaftungsrechts – Reform”, Juristenzeitung, 1991, 175. Wagner, G., “Die Zukunft der Umwelthaftpflichtversicherung”, Versicherungsrecht, 1992, 261-272. Wagner, G, “Versicherungsfragen der Umwelthaftung” in Ahrens, M. And Simon, J. (eds.) Umwelthaftung, Risikosteuerung und Versicherung, Berlin, Erich Schmidt Verlag, 1996, 104-105. Wagner, F., “Risk Securitization. An Alternative Risk Transfer of Insurance Companies”, Geneva Papers on Risk and Insurance, 1998, 575. Wansink, J.H., “De nieuwe milieuaansprakelijkheidsverzekering”, Milieu en Recht, 1985, 98. Wa n s i n k , J. H . , “ H e t D E S - a r r e s t i n h e t p e r s p e c t i e f v a n v e r z e k e r b a r e slachtofferbescherming”, Aansprakelijkheid en Verzekering, 1993, 7-12. Wansink, J.H., De algemene aansprakelijkheidsverzekering, second edition, 1994.


257

BIBLIOGRAPHY

Wansink, J.H., “De aansprakelijkheidsverzekering en de dekking van ’long tail risico’s’”, Aansprakelijkheid en Verzekering, 1995. Wansink, J.H., “De Belgische wet op de verzekeringsovereenkomst”, Verzekeringsrechtelijke Berichten, 1995, 38. Wansink, J.H., “Het polismodel AVB 96 en de dekking voor long-tail risicos”, A&V, 1996, 120-122. Wansink, J.H., “Hoe plotseling en onzeker is de verzekeringsdekking voor milieuaansprakelijkheidsrisico’s?” in Miscellanea. Juris Consulto vero dedicata, Essays offered to Prof.Mr. J.M. van Dunné, Deventer, Kluwer, 1997, 451-460. Wansink, J.H., “Verzekering en milieuschade als gevolg van vervoer/opslag van gevaarlijke stoffen”, Tijdschrift voor Milieuaansprakelijkheid, 1999, 77-82. Wessels, B., “Beroepsaansprakelijkheid van de notaris”, Weekblad voor Privaatrecht Notariaat en Registratie, (not dated), 5992. Wessels, B., “Zorgen om morgen: beheersing van beroepsaansprakelijkheid in discussie”, NJB, 1995, 235. Wessels, B., “Risicobeheer bij preadvisering”, Preadvies Ne Voa, Deventer, Law and Practice Publishers, 1995, 54. Whelan, C. and McBarnet, D., “The 'Crisis' in Professional Liability Insurance”, The Geneva Papers on Risk and Insurance, 1989, 296. Whinston, “Moral Hazard, Adverse Selection and the Optimal Provision of Social Insurance”, Journal of Public Economics, 1983, 49-71. Wijmen, P.C.E. van, “Bescherming van natuur- en milieuwaarden in rechte; de verdediging van collectieve goederen”, Tijdschrift voor Milieu en Recht, 1994, 234. Wils, W.P.J., “Insurance Risk Classifications in the EC: Regulatory Outlook”, Oxford Journal of Legal Studies, 1994, 449-467. Wilthagen, “Normativiteit en effectiviteit van de regulering van arbeidsomstandigheden” in: P. de Jong and others (eds.), Arbeid, recht en risico, 1995, pp. 85-111. Witte, B. and Forder, C. (eds.), The common law of Europe and the future of legal education, Kluwer, 1992, 31. Wittman, D., First come, first served: an economic analysis of “coming to nuisance”, [1980] JLS, 557-568. Zech, J., “Will the international insurance market replace traditional insurance products”, Geneva Papers on Risk and Insurance, 1998, 494. Zeckhauser, R., “Public Finance Principles and National Health Care Reform”, Journal of Economic Perspective, 1994, 55-60. Zeckhauser, “Medical Insurance: a Case Study of the Trade-off Between Risk Spreading and Appropriate Incentives”, Journal of Economic Theory, 1970, 2, 10-26. Zweigert, K. and Kotz, H., Introduction to Comparative Law, Oxford: Clarendon Press, 1998, 3rd ed.

258


LIST OF ABBREVIATIONS

List of Abbreviations AER AERPP ALER A&V AVB CBA CEA CERCLA CLR D EELR EJLE EL ERPL FAR GIELR GLR GPRI IILR IRLE JLE JLS JPE JRE JRI JZ MAS MenR MJ MSV

American Economic Review American Economic Review, Papers and Proceedings American Law and Economic Review Aansprakelijkheid en Verzekering Aansprakelijkheidsverzekering voor bedrijven (Liability Insurance for Companies in The Netherlands) Cost-Benefit Analysis Comité Européen des Assurances Comprehensive Environmental Response, Compensation and Liability Act Cornell Law Review Dalloz (France) European Environmental Law Review European Journal of Law and Economics Environmental Liability European Review of Private Law Financial Assurance Requirements Georgetown International Environmental Law Review Georgia Law Review Geneva Paper on Risk and Insurance International Insurance Law Review International Review of Law and Economics Journal of Law and Economics Journal of Legal Studies Journal of Political Economics Journal of Regulatory Economics Journal of Risk and Insurance Juristenzeitung Milieuaansprakelijkheidsverzekering samenwerkingsverband (Environmental Liability insurance pool in The Netherlands) Milieu en Recht (The Netherlands) Maastricht Journal of European and Comparative Law Milieuschadeverzekering (Environmental Damage Insurance in the Netherlands)


259

LIST OF ABBREVIATIONS

NJB NRD OCSLA OJLS OPA QJE RCRA RECIEL RJE RW SDLR SR SMCRA TBH TLR TMA TMR TPR UPLR VR YLJ

260

Nederlands Juristenblad Natural Resource Damages Outer Continental Shelf Lands Act Oxford Journal of Legal Studies Oil Pollution Act Quarterly Journal of Economics Resource Conservation and Recovery Act Review of European Community and International Environmental Law Rand Journal of Economics Rechtskundig Weekblad San Diego Law Review Sociaal Recht Surface Mining Control and Reclamation Act Tijdschrift voor Belgisch Handelsrecht Texas Law Review Tijdschrift voor Milieuaansprakelijkheid (The Netherlands) Tijdschrift voor Milieurecht (Belgium) Tijdschrift voor Privaatrecht University of Pennsylvania Law Review Versicherungsrecht Yale Law Journal


OECD PUBLICATIONS, 2, rue André-Pascal, 75775 PARIS CEDEX 16 PRINTED IN FRANCE (21 2003 04 1P) ISBN 92-64-10289-2 – No. 53089 2003

Insurance and Expanding Systemic Risks (Policy Issues in Insurance, No. 5)

Catastrophic Risks And Insurance: Policy Issues in Insurance No.8

Catastrophic Risks And Insurance (Policy Issues in Insurance)

Policy Issues in Insurance Assessing the Solvency of Insurance Companies (Policy Issues in Insurance)

Policy Issues in Insurance Financial Management of Large-Scale Catastrophes (Policy Issues in Insurance)

Terrorism Risk Insurance in Oecd Countries: Policy Issues in Insurance No.9

Risks in the Making: Travels in Life Insurance and Genetics

Love Insurance

Endpoint Insurance

Health Insurance

Opportunities in Insurance Careers

Powdersmoke Insurance

Insurance Economics

Opportunities in Insurance Careers

Risk and insurance in construction

Opportunities in Insurance Careers

Love Insurance

Endpoint Insurance

Environmental Risks and Insurance: A Comparative Analysis of the Role of Insurance in the Management of Environment-Related Risks

Insurance risk and ruin

Financial and Insurance Formulas

Tort Law and Liability Insurance (Tort and Insurance Law)

Risks at Sea: Amsterdam Insurance and Maritime Europe, 1766-1780

Risks at Sea: Amsterdam Insurance and Maritime Europe, 1766-1780

Topics in Unemployment Insurance Financing

Stochastic Models in Life Insurance

Introduction to insurance mathematics

Law of Marine Insurance

Cliffsnotes Understanding Health Insurance

Dictionary of Insurance Terms

Careers in Insurance and Risk Management

Insurance and Expanding Systemic Risks (Policy Issues in Insurance, No. 5)

Catastrophic Risks And Insurance: Policy Issues in Insurance No.8

Catastrophic Risks And Insurance (Policy Issues in Insurance)

Policy Issues in Insurance Assessing the Solvency of Insurance Companies (Policy Issues in Insurance)

Policy Issues in Insurance Financial Management of Large-Scale Catastrophes (Policy Issues in Insurance)

Terrorism Risk Insurance in Oecd Countries: Policy Issues in Insurance No.9

Risks in the Making: Travels in Life Insurance and Genetics

Love Insurance

Endpoint Insurance

Health Insurance

Opportunities in Insurance Careers

Powdersmoke Insurance

Insurance Economics

Opportunities in Insurance Careers

Risk and insurance in construction

Opportunities in Insurance Careers

Love Insurance

Endpoint Insurance

Environmental Risks and Insurance: A Comparative Analysis of the Role of Insurance in the Management of Environment-Related Risks

Insurance risk and ruin

Financial and Insurance Formulas

Tort Law and Liability Insurance (Tort and Insurance Law)

Risks at Sea: Amsterdam Insurance and Maritime Europe, 1766-1780

Risks at Sea: Amsterdam Insurance and Maritime Europe, 1766-1780

Topics in Unemployment Insurance Financing

Stochastic Models in Life Insurance

Introduction to insurance mathematics

Law of Marine Insurance

Cliffsnotes Understanding Health Insurance

Dictionary of Insurance Terms

Careers in Insurance and Risk Management

Recommend Documents