Sebastian Oberthur

Institutional Interaction
in Global Environmental
Governance
Synergy and Conflict among
International and EU Policies
edited by
Sebastian Oberthür
and Thomas Gehring
foreword by Oran R. Young

Institutional Interaction in Global Environmental
Governance
Global Environmental Accord: Strategies for Sustainability and Institutional
Innovation
Nazli Choucri, series editor
A complete list of books published in this series appears at the back of the book.
Institutional Interaction in Global Environmental
Governance
Synergy and Conflict among International and EU Policies
edited by Sebastian Oberthür and Thomas Gehring

with a Foreword by Oran R. Young
The MIT Press

Cambridge, Massachusetts
London, England
( 2006 Massachusetts Institute of Technology
All rights reserved. No part of this book may be reproduced in any form by any electronic or
mechanical means (including photocopying, recording, or information storage and retrieval)
without permission in writing from the publisher.
MIT Press books may be purchased at special quantity discounts for business or sales promo-
tional use. For information, please e-mail special_sales@mitpress.mit.edu or write to Special
Sales Department, The MIT Press, 55 Hayward Street, Cambridge, MA 02142.
This book was set in Sabon on 3B2 by Asco Typesetters, Hong Kong and was printed and
bound in the United States of America. Printed on recycled paper.
Library of Congress Cataloging-in-Publication Data
Institutional interaction in global environmental governance : synergy and conflict among

international and EU policies / edited by Sebastian Oberthür and Thomas Gehring ; with a
foreword by Oran R. Young.
p. cm.
Includes bibliographical references and index.
ISBN 0-262-05115-X (alk. paper)—ISBN 0-262-65110-6 (pbk. : alk. paper)
1. Environmental policy—International cooperation. 2. Environmental policy—European
Union countries. 3. Environmental law, International. I. Oberthür, Sebastian. II. Gehring,
Thomas, 1957–
GE170.I53 2006
363.7 0 0526—dc22 2005052121
10 9 8 7 6 5 4 3 2 1
Contents
Series Foreword vii

Foreword by Oran R. Young ix
Acknowledgments xiii
List of Contributors xv
List of Abbreviations xvii
1 Introduction 1
Thomas Gehring and Sebastian Oberthür
2 Conceptual Foundations of Institutional Interaction 19
Sebastian Oberthür and Thomas Gehring
3 The Climate Change Regime: Interactions with ICAO, IMO, and the EU
Burden-Sharing Agreement 53
Sebastian Oberthür
4 The Convention on Biological Diversity: Tensions with the WTO TRIPS
Agreement over Access to Genetic Resources and the Sharing of Benefits 79
G. Kristin Rosendal
5 Protecting the Northeast Atlantic: One Problem, Three Institutions 103
Jon Birger Skjærseth
6 Institutional Interplay and Responsible Fisheries: Combating Subsidies,
Developing Precaution 127
Olav Schram Stokke and Clare Coffey
7 The Convention on International Trade in Endangered Species of Wild Fauna
and Flora (CITES): Responding to Calls for Action from Other Nature
Conservation Regimes 157
John Lanchbery
vi Contents
8 Interactions between the World Trade Organization and International

Environmental Regimes 181
Alice Palmer, Beatrice Chaytor, and Jacob Werksman
9 Interactions of EU Legal Instruments Establishing Broad Principles of
Environmental Management: The Water Framework Directive and the IPPC
Directive 205
Andrew Farmer
10 The EU Habitats Directive: Enhancing Synergy with Pan-European Nature
Conservation and with the EU Structural Funds 233
Clare Coffey
11 The EU Deliberate Release Directive: Environmental Precaution versus Trade
and Product Regulation 259
Ingmar von Homeyer
12 The EU Air Quality Framework Directive: Shaped and Saved by Interaction?
285
Jørgen Wettestad
13 Comparative Empirical Analysis and Ideal Types of Institutional Interaction
307
Appendix: Overview of Identified Cases of Institutional Interaction 373
Index 399
Series Foreword
A new recognition of profound interconnections between social and natural sys-

tems is challenging conventional constructs and the policy predispositions informed
by them. Our current intellectual challenge is to develop the analytical and theoreti-
cal underpinnings of an understanding of the relationship between the social and
the natural systems. Our policy challenge is to identify and implement effective
decision-making approaches to managing the global environment.
The series Global Environmental Accord: Strategies for Sustainability and Institu-
tional Innovation adopts an integrated perspective on national, international, cross-
border, and cross-jurisdictional problems, priorities, and purposes. It examines the
sources and the consequences of social transactions as these relate to environmental
conditions and concerns. Our goal is to make a contribution to both intellectual and
policy endeavors.
Nazli Choucri
Foreword
First published in 1999, the Science Plan of the long-term research project on the
Institutional Dimensions of Global Environmental Change (IDGEC) states that
‘‘the effectiveness of specific institutions often depends not only on their own fea-
tures but also on their interactions with other institutions’’ (Young et al. 1999, 49).
This observation, which gave rise to the development of an analytic theme known to
the IDGEC community as the problem of interplay, struck a responsive chord and
launched what has become an important stream of research on interactions between
and among distinct institutions that influence the course of human-environment
relations. We can say with some certainty at this stage that the results of this stream
of research will make up a significant component of IDGEC’s scientific legacy.
No one has made a larger contribution to this line of inquiry than Thomas Gehr-
ing and Sebastian Oberthür, two German political scientists who have addressed the
problem of interplay in a number of individual papers and who have now joined
forces in producing this major contribution to the literature on institutional inter-
play. Funded by the European Commission and endorsed by IDGEC, Institutional
Interaction in Global Environmental Governance brings together a sizable collec-
tion of case studies of interplay occurring both at the international level and at
the European Union level and subjects the findings from the cases to rigorous com-
parative analysis. The result is a book that sets the standard for all those seeking to
produce new insights pertaining to the dynamics of institutional interactions.
Three things make this volume especially noteworthy. First, Oberthür and Gehr-
ing adopt what amounts to a reductionist approach to the study of institutional in-
terplay. Thus, they disaggregate interplay to focus on a single source institution, a
single target institution, and a unidirectional causal pathway as their basic unit of
analysis. They argue that the route to understanding institutional interactions lies
in this reductionist approach. Once we understand interplay in its simplest form,
x Oran Young
we can proceed to build up an understanding of more complex interactions featur-

ing two-way flows and more than two institutions. Others have argued that we need
to focus directly on the more complex forms of interaction on the assumption that
some of the things we want to understand arise as emergent properties of institu-
tional complexes. We do not know at this stage what approach to institutional
interplay will ultimately prove most fruitful. But Oberthür and Gehring have done
us all a distinct service by flagging this issue and laying out their own approach in
a clear and rigorous manner.
A second distinctive feature of the volume is its sustained examination of causal
mechanisms. What is at stake here is an effort to identify the pathways and mecha-
nisms through which interactions between distinct institutions can affect the course
of human-environment relations. Using the familiar distinction among outputs, out-
comes, and impacts, Oberthür and Gehring began by differentiating among several
causal mechanisms in constructing the analytic framework for this project. They de-
voted particular attention to cognitive interaction, which highlights interinstitutional
learning; interaction through commitment, which features the effects of commit-
ments on the part of members of the source institution on the preferences of those
associated with the target institution; and behavioral interaction, which centers on
ways that behavior taking place within the source institution affects the operation
of the target institution. Of course, exploring the nature of these causal mechanisms
is a priority concern. But equally important in the context of institutional analysis is
the fact that they spelled out these mechanisms in considerable detail in advance of
the project’s empirical phase. As a result, the mechanisms function as theoretically
derived expectations to be tested or at least explored systematically through the
case studies conducted by all the participants in the project.
Third, as Gehring and Oberthür point out, there is a need to think more carefully
about the circumstances under which institutional interplay produces synergistic
outcomes in contrast to disruptive or conflict-ridden outcomes. Without doubt,
many of those who have become interested in institutional interplay have been con-
cerned about the potential for interference associated with institutional interaction.
The contributors to this volume do not deny that interference does occur or that it
may sometimes be serious in terms of its consequences. But their studies provide
evidence that synergistic interactions are common, perhaps more common than dis-
ruptive interactions. This is especially true when the relevant institutions operate
within the same policy field. The results are certainly not all in with regard to the
Foreword xi
incidence of synergistic and disruptive interactions. But the findings reported in this
volume regarding the importance of synergy certainly constitute food for thought.
The study of institutional interplay remains an infant industry. The findings we
have been able to generate so far are only first steps toward an understanding of
this pervasive phenomenon; much work remains to be done. Nevertheless, we have
already learned some important things about institutional interplay. Equally impor-
tant, the research conducted so far has established this subject as a fruitful area for
institutional analysis. No one deserves more of the credit for achieving these results
than the leaders of the project whose findings are presented in this volume.
Oran R. Young
Chair, Scientific Steering Committee
Institutional Dimensions of Global Environmental Change
Reference
Young, Oran R., Leslie A. King, Arun Aggarval, Arild Underdal, Peter H. Sand, and Merrilyn
Wasson. 1999. Institutional Dimensions of Global Environmental Change (IDGEC): Science
Plan. Bonn: International Human Dimensions Programme on Global Environmental Change.
Acknowledgments
This book is the result of a collaborative European research project that involved
four institutes from three countries: Ecologic—Institute for International and Euro-
pean Environmental Policy in Berlin, Germany; the Foundation for International
Environmental Law and Development (FIELD) in London, UK; the Fridtjof Nansen
Institute in Lysaker, Norway; and the Institute for European Environmental Policy
(IEEP) in London, UK. The project, titled ‘‘Institutional Interaction—How to Pre-
vent Conflicts and Enhance Synergies between International and EU Environmen-
tal Institutions,’’ was conducted between the end of 2000 and the beginning of
2003. More information about the project and its results can be found at http://
www.ecologic.de/projekte/interaction.
We are particularly grateful to the European Community for providing funding
for our project under the Specific Research and Technological Development Pro-
gramme ‘‘Energy, Environment and Sustainable Development’’ of the Fifth Frame-
work Programme for Research (Contract no. EVK2-CT2000-00079). We are also
thankful for the endorsement of the project by the Institutional Dimensions of
Global Environmental Change (IDGEC). This volume is intended to make a contri-
bution to this long-term international research project, which operates under the
auspices of the International Human Dimensions Programme on Global Environ-
mental Change (IHDP). The editors would like to extend special thanks to
Ecologic—Institute for International and European Environmental Policy and the
Otto-Friedrich University Bamberg for enabling us to continue working on the
book after the official end of the research project cofunded by the European
Community.
We also owe special thanks to the members of the project team for patiently
responding to recurring requests for revisions. Furthermore, we are indebted to
a number of individuals who have helped bring about this book in a number of
xiv Acknowledgments
ways. Oran R. Young and two anonymous reviewers provided helpful comments on
the manuscript. The series editor, Nazli Choucri, gave valuable and encouraging
guidance for the revision and finalization of the manuscript. Kristina Vesper
provided invaluable assistance in the conduct of the project throughout its lifetime.
Philipp Bleninger and Sebastian Krapohl have earned our gratitude by providing
important assistance in and advice on the handling of our database of cases of insti-
tutional interaction reflected in the appendix to this volume and its evaluation con-
tained in chapter 13. Thanks are also due to Ulrich Pilster for his tireless assistance
in producing the final manuscript.
Finally, we would like to express our appreciation to The MIT Press for their
excellent cooperation in producing this volume. Needless to say, responsibility for
the contents remains solely with the authors.

Bamberg, Germany
List of Contributors
Clare Coffey Alice Palmer

Institute for European Environmental Policy Foundation for International Environmental
(IEEP) Law and Development (FIELD)
Brussels, Belgium London, UK
Beatrice Chaytor G. Kristin Rosendal
Foundation for International Environmental Fridtjof Nansen Institute
Law and Development (FIELD) Lysaker, Norway
London, UK
Andrew Farmer Fridtjof Nansen Institute
Institute for European Environmental Policy Lysaker, Norway
(IEEP)
Olav Schram Stokke
London, UK
Fridtjof Nansen Institute
Thomas Gehring Lysaker, Norway
Otto-Friedrich University Bamberg
Jacob Werksman
Bamberg, Germany
Foundation for International Environmental
Ingmar von Homeyer Law and Development (FIELD)
Ecologic-Institute for International and London, UK
European Environmental Policy
Jørgen Wettestad
Berlin, Germany
Fridtjof Nansen Institute
John Lanchbery Lysaker, Norway
Royal Society for the Protection of Birds/
BirdLife International
Sandy, UK
Ecologic-Institute for International and
European Environmental Policy and Otto-
Friedrich University Bamberg
Berlin and Bamberg, Germany
List of Abbreviations
APEC Asia-Pacific Economic Cooperation

BAT Best available technique/technology
BREF BAT Reference Document
CAC Codex Alimentarius Commission
CAEP Committee on Aviation Environmental Protection
CAFÉ Clean Air for Europe
CAP Common Agricultural Policy (of the European Union)
CBD Convention on Biological Diversity
CCD Convention to Combat Desertification
CCAMLR Convention on the Conservation of Antarctic Marine Living
Resources
CCMV Convention for the Conservation and Management of the
Vicuña
CFP Common Fisheries Policy (of the European Union)
CITES Convention on International Trade in Endangered Species of
Wild Fauna and Flora
CLRTAP Convention on Long-Range Transboundary Air Pollution
CMS Convention on Migratory Species of Wild Animals
COP Conference of the Parties
CTE Committee on Trade and Environment
DRD Deliberate Release Directive
EC European Community
ECJ European Court of Justice
xviii Abbreviations
EU European Union
FAO Food and Agriculture Organization of the United Nations
FD Framework Directive
GATT General Agreement on Tariffs and Trade
GEF Global Environment Facility
GHG Greenhouse gas
GMO Genetically modified organism
ICAO International Civil Aviation Organization
ICCAT International Commission for the Conservation of Atlantic
Tunas
ICES International Council for the Exploration of the Sea
IMO International Maritime Organization
INSC International North Sea Conference
Interpol, ICPO International Criminal Police Organization
IPCC Intergovernmental Panel on Climate Change
IPOA International Plan of Action
IPPC Directive Integrated Pollution Prevention and Control Directive
IPPC International Plant Protection Convention
IPR Intellectual property rights
IUCN World Conservation Union (previously: International Union for
the Conservation of Nature and Natural Resources)
LCP Large combustion plants
LMO Living modified organism
MEA Multilateral Environmental Agreement
MEPC Marine Environment Protection Committee
NAFO Northwest Atlantic Fisheries Organization
NATO North Atlantic Treaty Organization
NEC National Emission Ceiling
NGO Nongovernmental organization
OECD Organization for Economic Cooperation and Development
OSPAR Oslo and Paris Conventions for the protection of the Northeast
Atlantic
Abbreviations xix
OSPARCOM OSPAR Commissions

PCB Polychlorinated biphenyl
PD Pesticides Directive
PPM Process and production method
SAC Special Area of Conservation
SCM Agreement Subsidies and Countervailing Measures Agreement
SPA Special Protection Area
SPS Agreement Agreement on Sanitary and Phytosanitary Measures
TAC Total allowable catch
TBT Agreement Agreement on Technical Barriers to Trade
TRAFFIC Trade Records Analysis of Fauna and Flora in Commerce
TRIPS Agreement Agreement on Trade-Related Aspects of Intellectual Property
Rights
UNCED United Nations Conference on Environment and Development
UNECE United Nations Economic Commission for Europe
UNEP United Nations Environment Program
UNFCCC United Nations Framework Convention on Climate Change
UPOV International Union for the Protection of New Varieties of
Plants
WCMC World Conservation Monitoring Centre
WCO World Customs Organization
WHO World Health Organization
WIPO World Intellectual Property Organization
WTO World Trade Organization
WWF World Wide Fund for Nature
1
Introduction
International and EU environmental governance is affected by the interaction of

numerous sectoral legal instruments. International environmental regimes and EU
environmental instruments do not exist in isolation from each other or from institu-
tions in other policy fields. While they are usually targeted at specific issue areas,
their effects are not limited to their own domains. Frequently, they exert influence
on the development and effectiveness of other policy instruments, and are them-
selves influenced by other such instruments, both within the same policy field
and beyond (‘‘institutional interaction’’). This influence may create synergy by sup-
porting the policy of the affected institution or it may undermine and disrupt its
effectiveness. Hence, interinstitutional influence may be employed to enhance inter-
national and EU environmental governance, but it also aggravates the difficulties of
governance.
This book seeks to understand how institutional interaction can occur and what
its governance effects are. It is based on a coherent analytic approach to the rela-
tively novel subject of institutional interaction and explores interaction phenomena
in international and EU environmental governance from a comparative perspective.
We focus on the investigation of the causal mechanisms that drive individual cases
of interaction in which one international institution or EU legal instrument affects
another’s effectiveness or institutional development. Our empirical investigation
covers a broad range of interaction phenomena in EU and international environ-
mental governance and reaches beyond the ad hoc study of individual cases or lim-
ited ‘‘nests’’ of interacting institutions prevailing in the field to date.
The Relevance of Institutional Interaction
The heavily fragmented institutional structure of international environmental gover-

nance has contributed to an increasing political salience of issues of interinstitutional
2 Thomas Gehring and Sebastian Oberthür
coordination and cooperation. Over the last decades, states have entered into a
growing number of international environmental agreements, and they have tended
to establish them separately from each other. To date, more than two hundred
agreements have been concluded. While on average one treaty was adopted per
year until the 1970s, this number has grown to five since the 1980s (Beisheim et al.
1999). Whenever a new international treaty is adopted, it enters an institutional
setting that is already densely populated. The growing number of separately estab-
lished international environmental institutions suggests the rapidly increasing rele-
vance of institutional interaction. Accordingly, conferences of parties of multilateral
environmental agreements increasingly address issues of interinstitutional coordi-
nation and cooperation. Problems of institutional interaction support suggestions
to ‘‘cluster’’ multilateral environmental agreements—that is, to integrate groups of
such agreements or certain of their parts (Oberthür 2002), as discussed within the
framework of the UN Environment Program (UNEP). Calls for the creation of a
‘‘world environment organization’’ have partly been justified with reference to an
increasing demand for interinstitutional coordination within the field of the envi-
ronment and beyond, and to a growing potential for duplication of work (German
Advisory Council on Global Change 2001; Biermann and Bauer 2005).
Likewise, EU environmental policy consists of a patchwork (Héritier 1996) of nu-
merous instruments with diverse regulatory approaches, which has reinforced efforts
at improved policy integration. To date, the EU has produced more than two hun-
dred environmental legal instruments, primarily directives and regulations (Krämer
1999; Haigh 2003). Some of them set quality standards, while leaving the mode of
implementation to the member states. Others envisage emission control limits and
detailed technical regulations. Yet others prescribe particular procedures for the as-
sessment of environmental risks and impacts, or establish crosscutting mechanisms
such as environmental liability. In any given problem area, environmental gover-
nance rests on several of these instruments and is also influenced by instruments
from other policy fields. Enhancing synergies and coherence between different policy
instruments has therefore been a central element in the debate launched by the Eu-
ropean Commission’s White Paper on European Governance (European Commis-
sion 2001) as well as in discussions on an EU Sustainable Development Strategy
started in 2001. Since 2003, the European Commission is required to examine all
significant economic, social, and environmental impacts of a proposed measure,
both within and beyond the EU’s borders (European Commission 2002; in general,
Haigh 2003; Wilkinson 1998; Lenschow 2002).
Introduction 3
We find both synergistic and disruptive cases of institutional interaction in inter-

national and EU environmental governance. The global regime on the trans-
boundary movement of hazardous wastes has been strengthened as a result of the
establishment of a number of regional regimes addressing the same environmental
problem (Meinke 2002). And the EU Nitrates Directive has contributed to the im-
plementation of the North Sea Declarations. This ‘‘surprisingly effective’’ relation-
ship has allowed the EU to act as an international leader in this area (Vogler 1999,
24). In contrast, incentives for establishing fast-growing monocultural tree planta-
tions provided by the 1997 Kyoto Protocol to the UN Framework Convention on
Climate Change in order to maximize carbon sequestration from the atmosphere
are potentially at odds with the 1992 Convention on Biological Diversity (CBD)
that aims at preserving the biological diversity of forest ecosystems (Pontecorvo
1999).
Institutional interaction also occurs across the boundaries of policy fields. Perhaps
the most prominent example in international environmental governance concerns
the relationship between the World Trade Organization (WTO) that promotes free
international trade and several multilateral environmental agreements that estab-
lish new trade restrictions, such as the 1973 Convention on International Trade
in Endangered Species of Wild Fauna and Flora (CITES) and the 1987 Montreal
Protocol on Substances That Deplete the Ozone Layer. The WTO has constrained
the options available to environmental policymakers in these cases and may have
limited the effectiveness of the environmental agreements concerned (chapter 8).
Also, the EU Structural Funds that provide financial support to economic develop-
ment projects have undermined the effectiveness of EU nature conservation policies
enshrined, in particular, in the EU Habitats Directive. In this case, Structural Fund
rules have been revised so as to provide incentives for the implementation of the
Habitats Directive (chapter 10).
Political actors increasingly recognize the constraints, and employ the opportu-
nities, arising from the growing institutional density and interdependence at both
the international and EU levels. On the one hand, institutional interaction provides
opportunities for forum shopping and purposive policy development (Raustiala and
Victor 2004). It may also be employed to overcome obstacles that hinder policy
development within another institution. The International North Sea Conferences
were exclusively established to enhance the momentum of pollution-abatement ac-
tivities in an area already governed by an existing international institution with very
similar membership (chapter 5). On the other hand, environmental policymakers are
utterly aware of the constraints imposed by WTO rules and have adapted relevant
multilateral environmental agreements to make them WTO-compatible. The ten-
sions between the world trade system and various multilateral environmental agree-
ments are addressed in the Doha Round of trade negotiations launched in 2001
(chapter 8).
Issues of institutional interaction have increasingly attracted the attention of the
scientific community. Whereas the analysis of international institutions has for a
long time started from the fiction that institutions exists in isolation from and do
not significantly interfere with the performance of each other (Keohane 1984; Ritt-
berger 1993), research on the broader consequences of international institutions has
intensified more recently (Gehring 2004). As part of this shift in perspective, initial
steps have been made to examine the side effects of international institutions beyond
their own issue areas. Contributions have particularly built on the analytic frame-
work established to assess the effectiveness of international institutions (Young et al.
1999; Young 2002; Underdal and Young 2004; Stokke 2001a). Legal scholars and
political scientists have identified a risk of ‘‘treaty congestion’’ (Brown Weiss 1993,
679) and a growing ‘‘regime density’’ (Young 1996). Empirically, much of the
increasing literature on the phenomena of institutional interaction has focused on
instances of interinstitutional conflict that had raised political interest, while cases
resulting in synergy have received far less attention (e.g., Rosendal 2000, 2001;
Andersen 2002; Chambers 1998, 2001; Oberthür 2001; Stokke 2001b). Institu-
tional interaction has also been identified as a key issue for future research—for ex-
ample, by the Institutional Dimensions of Global Environmental Change (IDGEC)
project of the International Human Dimensions Programme on Global Environ-
mental Change (IHDP) (Young et al. 1999) and by the Concerted Action on the
Effectiveness of International Environmental Agreements sponsored by the EU
(Breitmeier 2000).
A diversity of terms is employed in the literature to denote phenomena of inter-
institutional influence, including interplay, linkage, interlinkage, overlap, and inter-
connection (e.g., Herr and Chia 1995; King 1997; Young 1996, 2002; Young et al.
1999; Chambers 1998; Stokke 2001b). The term interaction appears to us particu-
larly suitable because it emphasizes that interinstitutional influence is rooted in deci-
sions taken by the members of one of the institutions involved. It is thus action that
triggers interaction.
The present volume attempts to generate a more comprehensive picture of the
largely uncharted territory of institutional interaction in international and EU envi-
Introduction 5
ronmental governance by presenting the results of a large comparative exploration

of relevant interaction phenomena. It intends to advance our knowledge about insti-
tutional interaction by focusing on the causal influence of governance institutions
on each others’ normative development and performance. Empirically, it analyzes a
wide range of interaction incidents of varying political salience, different outcomes,
and distinct causal pathways in different areas of environmental governance beyond
the nation-state.
Conceptual Foundations
The development of the conceptual foundations for the investigation of institutional
interaction is at an early stage. Existing approaches mainly constitute typologies and
attempts to categorize phenomena of institutional interaction, which differ signifi-
cantly in form and substance (see Stokke 2001a, 1–8). They do not provide a suffi-
cient basis for the systematic analysis of the causal mechanisms and driving forces of
institutional interaction. The concept of institutional interaction on which the pres-
ent volume rests, relies on the following three components. Accompanied by an
overview of existing approaches, it is fully elaborated in chapter 2.
The Notion of International and EU Institutions The inquiry of the present vol-
ume focuses exclusively on negotiated sectoral legal systems because we are gen-
erally interested in capturing institutions that are established for the purpose of
governance. Only negotiated institutions may be used instrumentally to bring about
collectively desired change in the international system. Scholars exploring the effec-
tiveness of international institutions have focused their attention on such negotiated
institutions (e.g., Haas, Keohane, and Levy 1993; Young 1999; Miles et al. 2002).
International institutions can be defined as ‘‘persistent and connected sets of rules
and practices that prescribe behavioral roles, constrain activity, and shape expec-
tations’’ (Keohane 1989, 3). Usually they include a separate communication and
decision-making process from which their norms and rules emerge (Gehring 1994).
Both international regimes based on international treaties and international organi-
zations qualify as specific international institutions.
We identify EU legal instruments, in particular directives and regulations, as
the suitable functional equivalent of specific international institutions at the EU
level. Like international institutions, they constitute distinct systems of norms
negotiated to balance the interests of the member states and other actors involved.
They also focus on limited functionally defined issue areas and possess separate
communication processes. EU member states and the European Commission decide

on their establishment and development separately from other instruments. Fre-
quently, so-called comitology committees related to a particular instrument prepare
amendments of that instrument or take implementing decisions. This is not to deny
that the EU has a particularly sophisticated overarching institutional framework for
lawmaking and implementation, which includes the EU and EC Treaties as well as
the European Court of Justice, the European Commission, and the European Parlia-
ment. Whereas this framework has to be taken into account, EU legal instruments
are sufficiently similar to international institutions so as to serve as functional equiv-
alents in our study of institutional interaction.
Specific international institutions (international regimes and specific international
organizations) and EU legal instruments (directives, regulations, decisions) consti-
tute the prime governance instruments in international and European affairs. The
choice of these institutions as the principal units of institutional interaction relates
the inquiry of this volume to the extensive literature on international and European
governance. It also provides a suitable foundation for investigating the ramifications
of institutional interaction for the system of international and EU governance.
The Meaning of Institutional Interaction Generally, institutional interaction will

exist, if one institution affects the institutional development or the effectiveness
(performance) of another institution. In essence, institutional interaction refers to a
causal relationship between two institutions, with one of these institutions (‘‘the
source institution’’) exerting influence on the other (‘‘the target institution’’). In the
absence of causal influence, we would be faced with mere coexistence of two or
more institutions. In studies on the emergence and development of institutions, the
institution constitutes the dependent variable to be explained. In studies on institu-
tional effectiveness, it constitutes the independent variable that explains observed
effects. In the case of institutional interaction, both the independent and the depen-
dent variables are institutions.
Our concept of institutional interaction does not imply that influence runs back
and forth between the institutions involved. On the contrary, causal influence
implies that influence runs unidirectionally from the source to the target. Accord-
ingly, a causal relationship between the institutions involved will be established,
if we identify (1) the source institution and, more specifically, the relevant rules/
decision(s) from which influence originates; (2) the target institution and, more spe-
cifically, the relevant parts of the institution itself or the issue area governed by it
Introduction 7
that are subject to the influence of the source institution; and (3) a unidirectional
causal pathway connecting the two institutions.
This concept of institutional interaction requires that complex interaction situa-
tions are analytically disaggregated into a suitable number of individual cases so
that clear causal relationships between pairs of institutions can be identified. In
real-world situations, a clear-cut causal relationship between two institutions may
be difficult to identify—be it because interaction involves more than two institu-
tions, or because influence runs back and forth between two institutions, or because
two institutions influence each other in various ways. Complex situations are diffi-
cult to analyze rigorously unless we disaggregate them into a suitable number of
cases comprising a single source institution, a single target institution, and a single,
clearly identifiable causal pathway. Emergent properties of more complex situations
are then expected to result from particular forms of the coexistence of, and interplay
between, several cases of interaction. They may be examined by carefully recombin-
ing the individual cases.
This concept expands the study of the effectiveness of environmental institutions
to the investigation of institutional interaction. We share with the established re-
search on the effectiveness of international institutions the interest in cases in which
the ‘‘output’’ of an institution (i.e., its norms and decisions) results at least poten-
tially in behavioral changes of relevant actors (‘‘outcome’’)—changes that have
actual or potential effects on the environment or another target of governance
(‘‘impact’’; on these categories of effectiveness, see Underdal 2004). In the present
volume, we do not examine cases of institutional interaction with little or no impli-
cations for the performance of the institutions involved, such as attempts to increase
the bureaucratic efficiency of institutions, for example through streamlining or coor-
dinating reporting requirements.
Causal Mechanisms of Institutional Interaction Causal mechanisms help us under-

stand how, and under which conditions, governance institutions are capable of
exerting influence on each other. They elucidate the driving forces of institutional
interaction beyond a mere description of coevolution processes or of the density of
institutional settings in particular areas such as marine pollution or nature conserva-
tion. Causal mechanisms not only structure the multifaceted realm of institutional
interaction and explain variations in cases, but also demonstrate which actors are
indispensable for the emergence of interinstitutional influence. The empirical analy-
sis of the present study relies on three of four causal mechanisms that were derived
from theories of international institutions, cooperation theory, and negotiation

theory and that we believe are exhaustive. The way the target institution is affected
by a decision of the source is particularly important for the identification of the
causal mechanisms, because no interaction occurs without a noticeable effect on
the target institution or the issue area governed by it. Causal mechanisms of interac-
tion identified on this basis differ in particular with respect to how source and target
are linked.
In two of the four causal mechanisms, interaction affects the decision-making pro-
cess of the target institution. First, a case of interaction may be based on a transfer
of knowledge (Cognitive Interaction). This causal mechanism follows from the com-
plexity of the world and the fact that actors have limited information and informa-
tion processing capacities at their disposal. Cognitive Interaction may occur between
any two institutions, since ‘‘learning’’ can take place without any overlap in issues
or in the memberships of the institutions involved. Moreover, it can be triggered by
states and nonstate actors, such as nongovernmental organizations and secretariats
of international institutions. Second, a case of interaction may also be based on
commitments agreed on within the source institution that affect the constellation of
interests and the decision-making process within the target institution by influenc-
ing the payoffs of available options (Interaction through Commitment). In this case,
conditions are much more restrictive. The issues dealt with by the institutions
involved must overlap somewhat, because otherwise commitments under one insti-
tution could not affect options considered in the other institution. Memberships of
the institutions must also overlap somewhat, because otherwise no participant in the
policymaking process within the target institution would be subject to the commit-
ments of the source institution. It follows that member states are the most important
actors for Interaction through Commitment.
In the other two causal mechanisms, interaction affects the effectiveness of the
target institution within its own domain. First, the source institution may induce be-
havioral changes of actors within its own issue area that are relevant for the effec-
tiveness of the target institution within its issue area (Behavioral Interaction). Such
interaction occurs outside the decision-making processes of the two interacting insti-
tutions and usually involves nonstate actors such as companies and citizens that are
active within the issue areas concerned. It affects the performance of the target insti-
tution directly without requiring a decision from the latter. Finally, a case of interac-
tion may be based on effects on an institution’s ultimate target of governance (such
as international trade or the ozone layer) induced by the source institution at the im-
Introduction 9
pact level (Impact-Level Interaction). In this case, the effect on the target is a direct
spillover of the effects of the source institution on its target of governance that may
occur due to the ‘‘functional interdependence’’ (Young 2002, 23) of the issue areas
concerned.
Apart from their value as a basis for systematic and meaningful research and ac-
cumulation of knowledge, the causal mechanisms of institutional interaction help
distinguish between different conditions of governance existing in the realm of insti-
tutional interaction. They differ with respect to the actors who might initiate them
and the purposes for which they might be employed, as well as the forums in which
options to enhance synergy or mitigate conflict might primarily be pursued. A
careful analysis of the underlying causal mechanism of a case of interaction will
therefore also facilitate systematic thinking about effective policy options so as to
enhance international and European governance.
The Design of the Empirical Investigation

Our empirical analysis of institutional interaction starts from eleven environmental
or environmentally relevant international regimes and environmental EU directives.
The analysis explores the cases of interaction in which these ‘‘core institutions’’ are
involved and investigates how each core institution interacts with other interna-
tional institutions and pieces of EU legislation. Three criteria guided the selection
of the core institutions. First, we aimed at covering varying environmental media
or policy areas at both the international and the EU level (i.e., atmosphere, protec-
tion of biodiversity, marine pollution/water policy, management of living resources).
Second, we focused on institutions with a demonstrated political relevance for inter-
national or EU environmental governance. Third, we selected institutions that could,
according to preliminary expert judgment, reasonably be expected to interact with
other international and EU institutions in significant ways. We believe that our se-
lection of core institutions covers a broad range of international and EU environ-
mental governance.
Chapters 3–8 each start from an environmental or environmentally relevant inter-
national institution. The climate change regime based on the UN Framework Con-
vention on Climate Change and its Kyoto Protocol serves as the core institution of
institutional interaction explored by Sebastian Oberthür in chapter 3. In chapter 4,
G. Kristin Rosendal takes the Convention on Biological Diversity as the point of
departure for her investigation. Jon Birger Skjærseth analyzes in chapter 5 the insti-
tutional interactions in which the international regime for the protection of the
Northeast Atlantic has been involved. Olav Schram Stokke and Clare Coffey exam-
ine in chapter 6 the global fisheries regime composed, in particular, of the UN Fish
Stocks Agreement and a number of FAO regulations. Chapter 7, authored by John
Lanchbery, is devoted to the examination of institutional interactions involving the
Convention on International Trade in Endangered Species of Wild Fauna and Flora
(CITES). In chapter 8, Alice Palmer, Beatrice Chaytor, and Jacob Werksman explore
the interactions between the World Trade Organization and a number of multilat-
eral environmental agreements.
Environmental EU directives serve as core institutions of chapters 9–12. In chap-
ter 9, Andrew Farmer covers the institutional interactions of the EU Water Frame-
work Directive and the Directive on Integrated Pollution Prevention and Control
(IPPC Directive) related to industrial plants. Chapter 10, authored by Clare Coffey,
explores the manifold interactions in which the EU Habitats Directive on nature
conservation is involved. Ingmar von Homeyer discusses in chapter 11 interactions
with the EU Deliberate Release Directive on genetically modified organisms. Finally,
in chapter 12, Jørgen Wettestad investigates institutional interactions of the EU Air
Quality Framework Directive.
Reflecting the first step of the empirical analysis in our project, each chapter
attempts to provide a comprehensive overview of the most significant cases of
institutional interaction in which each of the core institutions has been involved.
Authors searched for empirical cases of interaction irrespective of their political sa-
lience and were thus able to ascertain cases that may not have caught the attention
of policymakers. This approach promises to elucidate the network of institutional
interactions in which the core institutions are involved. Networks of interaction
relate both to horizontal interaction between international institutions or between
EU legal instruments (depending on the core institution) and vertical interaction
between international institutions and EU legal instruments.
Of particular relevance for the identification of cases of institutional interaction is
the problem of remote causation and long causal chains (Underdal 2004). In our
project, we prioritized obvious cases of interaction with short causal chains over
less obvious ones with longer causal chains. Thus, we focused on cases in which in-
fluence runs directly from the source institution to the target institution, not on con-
stellations in which it passes through numerous intermediate steps. Moreover, we
concentrated on identifying cases driven by three of the four general causal mecha-
nisms mentioned above, namely, Cognitive Interaction, Interaction through Com-
mitment, and Behavioral Interaction. The empirical analysis does not consider
Introduction 11
Impact-Level Interaction because it is frequently based on complicated natural

science links and requires investigation of long causal chains with many intervening
variables.
In the second step, each chapter examines a smaller number of cases of interaction
in more detail. The in-depth analysis especially focuses on elucidating the causal in-
fluence exerted by the source institution on the target institution, and on investigat-
ing the causal mechanism driving the case of interaction in question and examining
its momentum and effects. Cases for in-depth analysis were selected so as to ensure
that horizontal interaction at the international level, horizontal interaction between
EU legal instruments, and vertical interaction between international institutions and
EU instruments were roughly equally represented.
In the third step, we engage in a comparative analysis of institutional interaction.
Case-study authors were asked to codify the cases according to a number of criteria
presented in chapter 13. The resulting database covering all identified 163 cases of
interaction is reflected in the appendix to this volume. It provides a comparatively
broad picture of the diversity and variety of institutional interaction in interna-
tional and EU environmental governance. The database enabled us to aggregate
data on individual cases and to identify dominant patterns of institutional interac-
tion. Inductively derived patterns provided the basis for elaborating Weberian ideal
types of institutional interaction that follow distinct rationales. These ideal types
further differentiate our general causal mechanisms and may provide a kit of stan-
dard forms of interaction (see also Gehring and Oberthür 2004). Chapter 13 pro-
vides a comparative evaluation of our codification data and elaborates ideal types
of institutional interaction.
Empirical Findings and Conceptual Development

Our analysis of institutional interaction leads to a wealth of general and case-
specific findings. In this section, we highlight findings relating to two complexes,
namely, the empirical analysis of our sample of more than 150 cases of institutional
interaction, and the development of Weberian ideal types of institutional interaction
designed to fine-tune the general causal mechanisms. The general results are elabo-
rated in more detail in chapter 13.
Empirical Findings The comparative examination of a larger set of cases of inter-

action enables us to derive aggregate insights about the patterns of institutional
interaction and thus supports the generation of policy-relevant knowledge. Given
the methodology of case selection, we are confident that our results at least roughly
reflect the overall situation in international and EU environmental governance, al-
though figures may be expected to differ to some extent for other samples. The
empirical conclusions constitute inductively generated hypotheses, which might be
tested against other samples of cases. However, we caution that the sample is not
statistically representative and therefore does not allow for the generalization of
insights to other populations of cases. In particular, we do not claim that the empir-
ical results hold for interaction phenomena beyond international and EU environ-
mental governance.
All three general causal mechanisms on which the empirical inquiry was based
were represented in our sample, but distribution varies considerably. Cognitive
Interaction was comparatively rare, whereas Behavioral Interaction accounted for
about half and Interaction through Commitment for about 40 percent of all cases.
Cases of Cognitive Interaction may be underrepresented in our sample because
‘‘learning’’ may be a tacit process, which is not easy to detect. However, we see
that institutional interaction is a multifaceted phenomenon that cannot be reduced
to a single causal mechanism.
A clear majority of the cases of interaction identified by us created synergy, while
only about one-quarter resulted in disruption. Whereas disruption was some-
what more frequent at the international level, synergy dominates at all levels,
namely, in horizontal interaction between international institutions, in horizontal
interaction between EU legal instruments, and in vertical interaction between inter-
national and EU instruments. This finding contrasts with conventional wisdom.
Much of the existing literature has focused on the problems arising from institu-
tional interaction. According to our sample, this focus does not provide a full pic-
ture of the interaction phenomenon. It may be a consequence of the fact that
conflict attracts significantly higher political and scientific attention than harmoni-
ous or synergistic situations, because people react more strongly to the risk of losses
(conflict) than to the promise of additional benefits. This finding further suggests
that institutional interaction may not primarily be a bad thing that ought to be
diminished as far as possible. The prevailing institutional fragmentation of inter-
national and EU environmental governance as well as substantive overlap do not
predominantly result in conflict or undesirable ‘‘duplication of work.’’ They may
provide a valuable asset for skillful policymaking to enhance environmental gover-
nance. Policies to minimize allegedly undesirably interaction could risk sacrificing
this asset.
Introduction 13
Disruption prevails in interaction across the boundaries of policy fields, while syn-
ergy dominates within the field of environmental policy. Interaction staying within
environmental affairs has supported the effectiveness of governance in more than
80 percent of the relevant cases in our sample, whereas conflicts prevailed in inter-
action with institutions from other policy fields. Most cases of disruption in our
sample related to interaction across policy fields, and a majority of cases of interac-
tion across policy fields resulted in disruption. Once again, this pattern holds true
for horizontal interaction between international institutions and between EU instru-
ments as well as for vertical interaction. This finding might not come as a surprise
because institutions belonging to different policy fields will frequently have consider-
ably diverging objectives and may be supported by different constituencies.
Whereas more than a third of the unintentionally triggered cases of interaction in
our sample resulted in disruption, intentionally triggered cases of disruption appear
to be particularly rare. It may not be surprising that disruptive interaction is virtu-
ally absent from the environmental policy field; we may expect that it plays a more
prominent role in more competitive policy fields such as security affairs. It is more
noteworthy that disruptive interaction is occasionally employed intentionally even
in environmental governance to bring about change within other institutions, in par-
ticular those belonging to other policy fields. It is also remarkable that roughly half
of our synergistic cases were unintentionally triggered. Moreover, a majority of the
disruptive cases have been responded to, whereas roughly 80 percent of the syner-
gistic cases have not drawn a collective political response. This may be explained
by the fact that conflicts leave some actors aggrieved who may then struggle for im-
provement, whereas synergy tends to be simply ‘‘consumed.’’ Overall, significant
opportunities exist for enhancing international and EU environmental governance
by an intensified political management and use of institutional interaction.
Weberian Ideal Types of Institutional Interaction Our sample of cases also pro-
vided a solid basis for the development of Weberian ideal types of institutional inter-
action, which subdivide and specify the general causal mechanisms and elaborate
their distinctive features. Thus, we move beyond the three basic causal mechanisms
and develop a more sophisticated framework for the analysis of individual cases of
interaction so as to be able to better explain and understand the strikingly different
properties of cases of interaction driven by the same causal mechanism. Ideal types
are abstract and deductively generated models, which reflect mutually exclusive
rationales inherent in different social-interaction phenomena, to which real-world
cases can be compared. We identified two ideal types of Cognitive Interaction and
three types of Interaction through Commitment, while we were unable to identify
ideal types of Behavioral Interaction.
Intentionality is the crucial distinction between the two types of Cognitive Interac-
tion. While ‘‘learning’’ cannot be imposed, it may or may not be intentionally trig-
gered by the source institution. If Cognitive Interaction is not intended, members of
the target institution use an institutional arrangement or policy idea of the source
institution as a policy model. For example, the compliance system under the Mon-
treal Protocol for the protection of the ozone layer influenced the negotiations on
the compliance system under the Kyoto Protocol on climate change because it pro-
vided a model of how to supervise implementation and deal with cases of possible
noncompliance. If Cognitive Interaction is intentionally triggered, the source institu-
tion largely frames the learning process by requesting assistance from the target,
ultimately in order to trigger a feedback case of Behavioral Interaction furthering
its own effectiveness. For example, the Convention on International Trade in Endan-
gered Species of Wild Fauna and Flora (CITES) requested assistance from spe-
cialized international institutions such as the World Customs Organization (WCO)
and Interpol because it expected this assistance to facilitate the effective implementa-
tion of CITES obligations (chapter 7).
The three ideal types of Interaction through Commitment are characterized by a
key difference in the objectives or memberships or means of governance of the insti-
tutions involved. Cases of Interaction through Commitment that are driven by dif-
ferences in objectives create a demand for jurisdictional delimitation. Due to their
underlying rationale, they will usually cause disruption and restrain the effectiveness
of both institutions involved. Consider that international trade is regulated within
the WTO with the purpose of liberalizing trade and thus removing obstacles to in-
ternational trade. At the same time, the Cartagena Biosafety Protocol to the Con-
vention on Biological Diversity (CBD) governs international trade in genetically
modified organisms (GMOs) predominantly with the purpose of protecting the
environment of the importing countries. In this situation of contentious interde-
pendence, the governance challenge consists in arriving at a delimitation of juris-
dictions. However, it might occasionally prove useful as a political strategy to
deliberately raise (potential) jurisdictional conflict.
Interaction through Commitment may also take place between two institutions
that differ exclusively with respect to their membership, while pursuing identical
objectives and employing the same means. Under these circumstances, interested
Introduction 15
actors may promote governance by establishing a smaller ‘‘pilot’’ institution in order

to affect decision making in a larger institution addressing a similar range of issues.
For example, the establishment of the Natura 2000 system of nature conservation
sites under the EU Habitats Directive led to the creation of the similar Emerald net-
work under the pan-European Bern Convention. In such cases, interaction relies on
the fact that agreement reached within the smaller institution significantly changes
the situation and interest constellation facing actors in the larger institution. If rules
are applied regionally, the leading coalition within the broader institution will be
strengthened and acceptance costs will decrease so that opposition against equiva-
lent measures may wane—a mechanism that may be purposefully exploited by po-
litical actors.
Skillful reinforcement of international or European governance through institu-
tional interaction will also be promising, if interested actors manage to activate ad-
ditional means to realize their desired objectives. Frequently, international and EU
institutions do not control the full spectrum of possible governance instruments
but differ in the means available to them. Interaction will regularly raise the effec-
tiveness of both institutions involved if the diffusion of an obligation from one in-
stitution to another one with identical objectives and memberships activates
an additional means of implementation. For example, the ministerial North Sea
Conferences were established in the 1980s in order to reinforce the existing OSPAR
Convention for the protection of the Northeast Atlantic. They raised the political
salience of the issues at stake, but resulted in ‘‘soft-law’’ agreements. Transforma-
tion into binding international law under the OSPAR Convention and into EU
supranational law subsequently increased the originally low degree of obligation
and mobilized additional enforcement mechanisms that enhanced the effectiveness
of all institutions involved (chapter 5).
Obviously, this volume does not resolve all issues of institutional interaction; it
may even raise more questions than it was able to answer. An important area for
future research is the systematic analysis of more complex settings of institutional
interaction. Eventually, we will not be content with knowing how and why a
particular case of institutional interaction matters. The conceptual framework of
this volume can be employed and further developed to systematically explore more
complex settings and their emergent properties by recombining individual cases. We
identify two particular ways individual cases might be recombined so as to account
for more complex interaction situations. Cases of interaction may form sequential
chains so that an individual case gives rise to a subsequent case that feeds back on
the original source institution or influences a third institution. Cases of interaction

may also cluster around certain issues and institutions, so that a number of institu-
tions jointly address a particular problem and contribute to the effectiveness of gov-
ernance of a certain area. Whereas our study has hardly been able to delve into
pertinent aspects of these more complex interactional situations, it has made a start
that demonstrates the potential of the effort. On this basis, we may over time be able
to gain a clearer picture of the interlocking structure of international governance
institutions and EU legal instruments.
References
Andersen, Regine. 2002. The Time Dimension in International Regime Interplay. Global En-
vironmental Politics 2 (3): 98–117.
Beisheim, Marianne, Sabine Dreher, Gregor Walter, Bernhard Zangl, and Michael Zürn.
1999. Im Zeitalter der Globalisierung? Thesen und Daten zur gesellschaftlichen und politi-
schen Denationalisierung. Baden-Baden: Nomos.
Biermann, Frank, and Steffen Bauer, eds. 2005. A World Environment Organization: Solu-
tion or Threat for Effective International Environmental Governance? Aldershot: Ashgate.
Breitmeier, Helmut. 2000. ‘‘Complex Effectiveness’’: Regime Externalities and Interaction
(Working Group III). In Jørgen Wettestad, ed., Proceedings of the 1999 Oslo Workshop of
the Concerted Action Network on the Effectiveness of International Environmental Regimes,
45–48. Oslo: Fridtjof Nansen Institute.
Brown Weiss, Edith. 1993. International Environmental Issues and the Emergence of a New
World Order. Georgetown Law Journal 81 (3): 675–710.
Chambers, Bradnee W., ed. 1998. Global Climate Governance: Inter-Linkages between the
Kyoto Protocol and Other Multilateral Regimes. Tokyo: United Nations University.
Chambers, Bradnee W., ed. 2001. Inter-Linkages: The Kyoto Protocol and the International
Trade and Investment Regimes. Tokyo: United Nations University Press.
European Commission. 2001. European Governance: A White Paper. Doc. COM(2001) 428
final, 25.7.2001. Brussels: European Commission.
European Commission. 2002. Communication from the Commission on Impact Assessment.
Doc. COM(2002)276 final, 5.6.2002. Brussels: European Commission.
Gehring, Thomas. 1994. Dynamic International Regimes: Institutions for International Envi-
ronmental Governance. Frankfurt/Main: Peter Lang.
Gehring, Thomas. 2004. Methodological Issues in the Study of Broader Consequences. In
Arild Underdal and Oran R. Young, eds., Regime Consequences: Methodological Challenges
and Research Strategies, 219–246. Dordrecht: Kluwer.
Gehring, Thomas, and Sebastian Oberthür. 2004. Exploring Regime Interaction: A Frame-
work for Analysis. In Arild Underdal and Oran R. Young, eds., Regime Consequences: Meth-
odological Challenges and Research Strategies, 247–269. Dordrecht: Kluwer.
Introduction 17
German Advisory Council on Global Change. 2001. World in Transition: New Structures for
Global Environmental Policy. London: Earthscan.
Haas, Peter M., Robert O. Keohane, and Marc A. Levy, eds. 1993. Institutions for the Earth:
Sources of Effective International Environmental Protection. Cambridge, MA: MIT Press.
Haigh, Nigel, ed. 2003. Manual of Environmental Policy: The EU and Britain. Leeds: Maney
Publishing and Institute for European Environmental Policy.
Héritier, Adrienne. 1996. The Accommodation of Diversity in the European Policy-Making
Process and Its Outcomes: Regulatory Policy as a Patchwork. Journal of European Public
Policy 3 (2): 149–167.
Herr, Richard A., and Edmund Chia. 1995. The Concept of Regime Overlap: Toward Identi-
fication and Assessment. In B. Davis, ed., Overlapping Maritime Regimes: An Initial Recon-
naissance, 11–26. Hobart: Antarctic CRC and Institute of Antarctic and Southern Ocean
Studies.
Keohane, Robert O. 1984. After Hegemony: Cooperation and Discord in the World Political
Economy. Princeton, NJ: Princeton University Press.
Keohane, Robert O. 1989. International Institutions and State Power: Essays in International
Relations Theory. Boulder, CO: Westview.
King, Leslie A. 1997. Institutional Interplay: Research Questions. Paper Commissioned by
Institutional Dimensions of Global Environmental Change and International Human Dimen-
sions Programme on Global Environmental Change. Burlington VT: Environmental Studies
Programme, School of Natural Resources, University of Vermont.
Krämer, Ludwig. 1999. Bilanz und Perspektiven der europäischen Umweltpolitik unter beson-
derer Berücksichtigung der Erweiterung, Presentation at the Forum Constitutionis Europae
of the Humboldt University Berlin on July 15, 1999 (FCE 6/1999). Available at http://
www.rewi.hu-berlin.de/WHI/deutsch/fce/fce699/index.htm.
Lenschow, Andrea, ed. 2002. Environmental Policy Integration: Greening Sectoral Policies in
Europe. London: Earthscan.
Meinke, Britta. 2002. Multi-Regime-Regulierung: Wechselwirkungen zwischen globalen und
regionalen Umweltregimen. Darmstadt: Deutscher Universitäts-Verlag.
Miles, Edward L., Arild Underdal, Steinar Andresen, Jørgen Wettestad, Jon Birger Skjærseth,
and Elaine M. Carlin. 2002. Environmental Regime Effectiveness: Confronting Theory with
Evidence. Cambridge, MA: MIT Press.
Oberthür, Sebastian. 2001. Linkages between the Montreal and Kyoto Protocols: Enhancing
Synergies between Protecting the Ozone Layer and the Global Climate. International Envi-
ronmental Agreements: Politics, Law and Economics 1 (3): 357–377.
Oberthür, Sebastian. 2002. Clustering of Multilateral Environmental Agreements: Potentials
and Limitations. International Environmental Agreements: Politics, Law and Economics 2
(4): 317–340.
Pontecorvo, Concetta Maria. 1999. Interdependence between Global Environmental Regimes:
The Kyoto Protocol on Climate Change and Forest Protection. Zeitschrift für ausländisches
öffentliches Recht und Völkerrecht 59 (3): 709–749.
Raustiala, Kal, and David G. Victor. 2004. The Regime Complex for Plant Genetic Re-
sources. International Organization 58 (2): 277–309.
Rittberger, Volker, ed. 1993. Regime Theory and International Relations. Oxford: Clarendon
Press.
Rosendal, G. Kristin. 2000. The Convention on Biological Diversity and Developing Coun-
tries. Dordrecht: Kluwer Academic.
Rosendal, G. Kristin. 2001. Impacts of Overlapping International Regimes: The Case of Bio-
diversity. Global Governance 7 (1): 95–117.
Stokke, Olav Schram. 2001a. The Interplay of International Regimes: Putting Effectiveness
Theory to Work. FNI Report 14/2001. Lysaker, Norway: Fridtjof Nansen Institute.
Stokke, Olav Schram, ed. 2001b. Governing High Seas Fisheries: The Interplay of Global and
Regional Regimes. Oxford: Oxford University Press.
Underdal, Arild. 2004. Methodological Challenges in the Study of Regime Effectiveness. In
Underdal, Arild, and Oran R. Young, eds. 2004. Regime Consequences: Methodological
Challenges and Research Strategies. Dordrecht: Kluwer.
Vogler, John. 1999. The European Union as an Actor in International Environmental Politics.
Environmental Politics 8 (3): 24–48.
Wilkinson, David. 1998. Steps towards Integrating the Environment into Other EU Policy
Sectors. In Timothy O’Riordan and Heather Voisey, eds., The Transition to Sustainability:
The Politics of Agenda 21 in Europe, 113–129. London: Earthscan.
Young, Oran R. 1996. Institutional Linkages in International Society: Polar Perspectives.
Global Governance 2 (1): 1–24.
Young, Oran R., ed. 1999. The Effectiveness of International Environmental Regimes: Causal
Connections and Behavioral Mechanisms. Cambridge, MA: MIT Press.
Young, Oran R. 2002. The Institutional Dimensions of Environmental Change: Fit, Inter-
play, and Scale. Cambridge, MA: MIT Press.
2
Conceptual Foundations of Institutional Interaction
This chapter introduces the shared conceptual framework of the contributions to

this volume in five steps. First, a brief review of existing approaches to institutional
interaction demonstrates that the conceptual development of the systematic study of
institutional interaction is still at an early stage. Existing approaches cannot readily
guide the empirical study of interaction between international and EU environ-
mental institutions, but provide a starting point for developing a suitable conceptual
foundation to that end.
Second, we identify international and EU institutions as the units of interaction.
In contrast to research on European integration, scholars of international relations
have developed a particular concept of issue-area specific governing institutions
that has been fruitfully applied to the study of international regimes. We identify
single legal instruments such as directives and regulations as the functional equiva-
lents to these institutions at the EU level that are most appropriate for the empirical
analysis of institutional interaction.
Third, we develop a notion of institutional interaction that is based on the identi-
fication of a causal relationship between two interacting institutions. A case of inter-
action thus comprises a source institution from which influence originates, a target
institution that is affected, and a causal pathway through which influence runs from
the source to the target. Complex interaction situations are analytically disaggre-
gated into individual cases with a single source institution, a single target institution,
and an identifiable causal pathway.
Fourth, we derive deductively from various theories of institutions, as well as from
negotiation theory and from cooperation theory, four general causal mechanisms
that may drive institutional interaction. A source institution may exert influence
directly on the rule-making process of the target institution in two different ways. It
may either trigger a learning process that leads to purely voluntary adaptation by
20 Sebastian Oberthür and Thomas Gehring
the target institution (Cognitive Interaction). Or it may commit its members to

an obligation that changes their preferences on matters negotiated within the target
institution (Interaction through Commitment). A source institution may also affect
the effectiveness of the target institution within its issue area in two different ways.
It may either exert influence on the behavior of states and nonstate actors that is
relevant for the implementation of the target institution (Behavioral Interaction).
Or it may directly affect the ultimate target of protection of the target institution
(Impact-Level Interaction). Three of these mechanisms serve as theoretical points of
reference for the empirical case studies contained in this volume.
Fifth, we spell out the tasks faced in the ensuing empirical investigation.
Existing Approaches to the Study of Institutional Interaction
Much of the empirically founded research on institutional interaction has so far

been motivated by concern about the detrimental impact of this interaction on the
effectiveness of the institutions involved, in particular in the field of environmental
protection. This literature constitutes an offspring of the research on the effective-
ness of international institutions that flourished in the 1990s (Haas, Keohane, and
Levy 1993; Victor, Raustiala, and Skolnikoff 1998; Young 1999; Miles et al.
2002). Like effectiveness research, it is interested in identifying successes and failures
of deliberate policymaking in order to draw lessons for global governance. Most
prominent is the empirical analysis of single problematic cases of institutional inter-
action (e.g., Chambers 1998; Stokke 1999; Rosendal 2000, 2001; Oberthür 2001;
Andersen 2002). Raustiala and Victor (2004) expanded the research focus and
explored the ‘‘regime complex’’ in the area of plant genetic resources composed of
five ‘‘elementary regimes.’’ They put forward conjectures about how actors can
use different institutions in the same policy field to pursue their interests, as well as
about the impact of institutional interaction on the evolution of regulatory ap-
proaches and legalization in international relations.
In a series of contributions, Oran Young has put forward a number of analytic
concepts and categories that constitute the single most important attempt to provide
a basis for systematic research on institutional interaction. In an influential article,
he put forward a taxonomy of four different types of interaction (Young 1996).
Embeddedness refers to the relationship of a governance institution to overarching
principles and practices such as sovereignty. Nestedness denotes the relationship of a
smaller institution to a functionally or geographically broader institution such as the
Conceptual Foundations of Institutional Interaction 21
nesting of the Multi-Fiber Agreement within the General Agreement on Tariffs

and Trade (GATT; now incorporated into the World Trade Organization, WTO)
(Aggarval 1983). Clustering refers to the deliberate combination of different institu-
tions such as the linkage of a wide variety of issues in the law of the sea convention.
Finally, overlap constitutes ‘‘a separate category of linkages in which individual
regimes that were formed for different purposes and largely without reference to
one another intersect on a de facto basis, producing substantial impacts on each
other in the process’’ (Young 1996, 6). Later, Young proposed to distinguish be-
tween horizontal interaction between institutions at the same level of social organi-
zation and vertical interaction between hierarchically ordered units at different levels
of social organization from the local to the international (Young et al. 1999; Young
2002, 113–132). This approach to the study of institutional interaction provides an
idea of the wide variety of possible subjects of inquiry and is useful in identifying
areas of particular interest. However, its overall research focus is so broad that the
categories and distinctions are too unspecific to guide a multicase empirical investi-
gation of interaction involving international and EU institutions.
Young and a program group of the Institutional Dimensions of Global Environ-
mental Change (IDGEC) project in the framework of the International Human
Dimensions Program (IHDP) also put forward a distinction between two different
drivers of institutional interaction (Young 2002, 23; Young et al. 1999, 50). Func-
tional linkages will exist ‘‘when substantive problems that two or more institutions
address are linked in biogeophysical or socioeconomic terms’’ (Young 2002, 23,
also 83–109). Based on this particular form of interdependence, they reflect ‘‘facts
of life’’ because ‘‘the operation of one institution directly influences the effectiveness
of another through some substantive connection of the activities involved’’ (Young
et al. 1999, 50). For example, chlorofluorocarbons (CFCs) functionally link the
international regime for the protection of the ozone layer and the regime on global
climate change because they have ozone-depleting properties and are at the same
time potent greenhouse gases. Action taken within the ozone regime is immediately
relevant for the climate change regime. In contrast, political linkages ‘‘arise when
actors decide to consider two or more arrangements as parts of a larger institutional
complex’’ (Young et al. 1999, 50). In this case, actors deliberately design the rela-
tionship between different institutions.
This classification is compelling at first glance, but at closer inspection it creates
considerable analytic difficulty. In particular, the two categories do not denote mu-
tually exclusive types. For example, Young et al. (1999, 53) take the protocols on
SO2 , NOX , and VOCs of the international regime on transboundary air pollution
as an example for a functional linkage. However, these protocols are undoubtedly
parts of a larger institutional complex, since they all belong to one convention man-
aged under the UN Economic Commission for Europe (Levy 1993). They may be
linked ‘‘functionally’’ and ‘‘politically’’ at the same time, with no clear cause-effect
relationship apparent between both types of linkage.
Based on a series of studies on international resource management (Stokke 1999,
2000, 2001a), Stocke (2001b) distinguishes between four different forms of institu-
tional interaction. Ideational interplay (previously referred to as diffusive interplay)
relates to ‘‘processes of learning’’ (Stokke 2001a, 10) and implies that the substan-
tive or operational rules of one institution serve as models for those negotiating an-
other regime. This may, for example, help us understand the rapid spread of general
normative principles such as sustainability, precaution, and ecosystem management.
Normative interplay refers to situations where the substantive or operational norms
of one institution either contradict or validate those of another institution (e.g., in
the case of the relationship of the WTO and multilateral environmental agreements).
Utilitarian interplay relates to situations where decisions taken within one institu-
tion alter the costs and benefits of options available in another. Interplay man-
agement, finally, relates to the political management of interinstitutional influence,
including the deliberate coordination of activities under separate institutions in
order to avoid normative conflict or wasteful duplication of programmatic efforts.
Stokke’s taxonomy constitutes an attempt to derive causal mechanisms of institu-
tional interaction from theoretical approaches such as organizational learning (diffu-
sion), legitimacy (normative interaction), and utilitarian cost-benefit analysis rather
than inductively from empirical cases. It has provided valuable input for the concep-
tual framework presented in this chapter.
To summarize, conceptual work on interaction between international and EU
institutions is still at an early stage. While research on European integration and
policymaking does not have concepts at its disposal to systematically analyze inter-
action phenomena, the limited conceptual work by international relations scholars
has not yet produced an encompassing framework (for an overview see Stokke
2001b, 1–8). Existing approaches mainly constitute attempts to categorize and sys-
tematize phenomena of institutional interaction, which differ significantly in form
and substance. Rather than providing a ready conceptual foundation for a compar-
ative empirical investigation of interaction phenomena, they provide a starting point
and basis for our efforts to develop such a foundation in the following pages.
International and EU Institutions as the Interacting Units
In the present volume, we focus exclusively on deliberately established (‘‘negoti-

ated’’) systems of norms, because they may be used to bring about collectively
desired change within the international system. This delineation of the research
area is in line with much of the literature on institutional institutions and their effec-
tiveness (e.g., Keohane 1993; Victor, Raustiala, and Skolnikoff 1998; Miles et al.
2002). From a governance perspective, deliberately established institutions receive
primary attention in the research, because they are employed instrumentally to bring
about change. Likewise, deliberately established governance arrangements attract
much interest in the literature on policymaking within the European Union (see gen-
erally, Scharpf 1999; Jachtenfuchs 2001). We recognize that systems of norms can
also emerge spontaneously as the result of uncoordinated behavior of actors in the
international system (Young 1982; Axelrod 1984). However, spontaneous institu-
tions such as international customary law (Hurrel 1993) and sovereignty are not part
of our research focus because they are merely suited to stabilize existing behavior and
cannot be created and developed purposefully as governance instruments. We would
also like to point out upfront that our use of the term institution diverges from
usage in the literature on European integration, where it predominantly denotes the
supranational bodies of the EU such as the European Commission, the European
Parliament, and the European Court of Justice. This is further discussed below.
Relevant international institutions have two components. First, they encompass
substantive rules and obligations that indicate socially desirable behavior. These
norms are the principal instruments of governance that may affect the behavior of
addressees and have an impact within the issue area governed. Second, ‘‘negotiated’’
institutions are distinguished from other types of institutions, such as spontaneous
institutions, by the particular decision-making processes from which their norms
and behavioral guidelines emerge. Generally, they include their own decision-
making apparatuses (see Levy, Young, and Zürn 1995). The importance of negotia-
tions and collective decision making has been increasingly acknowledged (Young
1994; Morrow 1994; Fearon 1998; Victor, Raustiala, and Skolnikoff 1998). The
procedural arrangements for the making of collective decisions enable actors to
adapt and develop international regimes dynamically and use them as flexible
instruments of international governance (Gehring 1994). Where the same problem
area is governed by several systems of norms, institutions are best identified accord-
ing to their distinct decision-making processes.
Both formal international organizations and more loosely institutionalized regimes

qualify as relevant international institutions (Simmons and Martin 2002). Issue-area
specific international regimes that are usually based on one or several international
treaties have been a major subject of inquiry for more than two decades (Keohane
1984; Rittberger 1993; Miles et al. 2002). In contrast, formal international organi-
zations, such as the Food and Agriculture Organization of the UN (FAO), the WTO,
and the International Maritime Organization (IMO), have attracted far less scientific
attention (but see Abbott and Snidal 1998; Barnett and Finnemore 1999). They are
usually defined by reference to their secretariats and their ability to enter into legal
contracts (e.g., Young 1994, 163–183; Keohane 1989, 3–4). Frequently, interna-
tional organizations are issue-area specific arrangements that closely resemble inter-
national regimes (e.g., the WTO). At times, they fulfill functions within regimes
and thus become part of them. In other cases, international regimes are embedded
in international organizations. Regularly, international organizations develop proce-
dures for elaborating collectively binding decisions, and many have elaborate legis-
lative programs (Keohane 1989, 5; Abbott and Snidal 1998, 15–16).
Because our research subject includes institutional interaction at the EU level, we
must identify EU equivalents of international regimes and organizations. The orga-
nizational actors that are usually denoted as ‘‘European institutions’’ such as the Eu-
ropean Commission, the European Court of Justice, and the European Parliament
do not qualify. Although they play an important role in EU policymaking, they do
not represent systems of norms, rules, and related decision-making processes that
are deliberately established to govern a given area of European relations. In particu-
lar they lack substantive rules designed to guide actors’ behavior. Given that the EU
at large is occasionally conceptualized as a highly developed international institution
(Keohane and Hoffmann 1991; Moravcsik 1998; Gehring 2002), it might be consid-
ered a suitable equivalent of relevant international institutions. Alternatively, one
could take the different EU policies, such as environmental, agricultural, or single-
market policy as EU institutions. However, both the Union at large and the policies
themselves consist of numerous structured and institutionalized communication
processes, which characterize international regimes and organizations. They display
considerable internal differentiation of instruments, policy approaches, and pro-
cesses. Neither the EU at large nor its policy areas thus seem to constitute functional
equivalents of international regimes and organizations.
EU legal instruments such as directives and regulations (in some cases also deci-
sions) can be considered functional equivalents of specific international institutions.
Despite the specific conditions of EU policymaking, such EU legal instruments share

the fundamental characteristics of specific international regimes and organizations.
Their substantive norms and obligations are also designed and adopted to guide
and influence the behavior of relevant addressees. Like international obligations,
they are decided on with strong intergovernmental participation, and they are usu-
ally addressed to the respective member states. Moreover, EU legal instruments are
usually negotiated separately from each other rather than in comprehensive pack-
ages. We even find decision-making processes that are specific to particular EU
instruments in the form of so-called comitology committees attended by representa-
tives of the member states and the European Commission (Joerges and Vos 1999;
Pollack 1997). These committees monitor the implementation of the instruments
and collaborate with the Commission on the development and adoption of second-
ary legislation.
However, we must take into account that international institutions and EU legal
instruments differ in important respects. While the decision-making processes of
separately established international institutions are formally independent from each
other, the decision-making processes of EU directives and regulations are embedded
in an overarching and integrated institutional framework. EU legal instruments are
the result of the regular EU lawmaking processes involving the European Com-
mission, the Council of Ministers, and the European Parliament as supranational
bodies. The Commission enjoys the right to initiate legislation and deals with issues
of implementation, secondary decision making, and enforcement, and it generally
occupies a much stronger role than the secretariats of international institutions.
The apparatus of regular EU decision making fulfills functions similar to the ‘‘ma-
chinery’’ of international regimes and organizations, but it does so in specific ways.
Moreover, EU legislation has a supranational character. It does not require national
ratification to take effect and the EU has a particularly strong enforcement system at
its disposal, which includes the Commission as a ‘‘prosecutor’’ and the European
Court of Justice, whose decisions are binding and can be enforced through a system
of penalty payments. Due to the comprehensive institutional framework, it will fre-
quently be much easier to adopt an EU directive than to establish an international
regime. Moreover, EU policymaking is embedded in, and relies on, broader pro-
grams and policy processes more often than policymaking at the international level
(on EU policymaking see, for example, Hix 2005; Nugent 2003).
To conclude, we identify as relevant institutions for the purpose of this volume
international regimes and organizations as well as EU legal instruments, in particular
directives and regulations, always including the associated decision-making pro-

cesses. Without denying important differences, we argue that the particularities
of the EU decision-making process do not render EU directives and regulations fun-
damentally different from international governance institutions. This delineation of
our area of research excludes two sorts of phenomena from our inquiry in institu-
tional interaction. We do not focus on interaction between a specific international
or EU institution and broad or unspecific entities such as ‘‘the UN system.’’ The
interplay between different instruments of a single regime or organization is also
beyond our research focus. For example, interaction between the UN Framework
Convention on Climate Change and its Kyoto Protocol is considered part of the
development within an international regime rather than interaction between institu-
tions, even though the analytic framework developed here may easily be employed
to investigate such intraregime influence (Gehring 2004).
The Concept of Institutional Interaction
In this section, we develop a concept of institutional interaction (see also Gehring

and Oberthür 2004). The analysis of institutional interaction requires establishing
a clear-cut cause-effect relationship between the institutions involved and disaggre-
gating complex situations into a suitable number of cases of interaction.
Establishing a Cause-Effect Relationship between Institutions

Institutional interaction relies on a cause-effect relationship between two institu-
tions. Interaction will occur if one institution affects the development or performance
of another institution (Breitmeier 2000). Thus defined, institutional interaction is
clearly distinguished from developments that occur in the presence of other institu-
tions but without their causal influence.
Conceptualizing institutional interaction as a matter of causal influence between
institutions relates our concept to existing research on the establishment and effec-
tiveness of international institutions. Exploration of causal influence constitutes the
core of this branch of research that has been fruitful for many years. In traditional
regime analysis, the emergence and design of international regimes have been attrib-
uted to different constellations of interests (Oye 1985; Martin 1993). Research on
the effectiveness of institutions is ultimately interested in whether, how, and to
what extent institutions affect the state of the environment or other targets of gover-
nance (Underdal 1992).
To establish a case of institutional interaction, we must explore the influence

exerted by one international or EU institution on another international or EU insti-
tution. For the exploration of cause-effect relationships, it is generally indispensable
to identify factors from which influence originates, and factors affected by such in-
fluence. In the case of institutional interaction, both the independent variable and
the dependent variable are institutions. Accordingly, establishing a case of institu-
tional interaction requires careful identification of (1) the source institution and,
more specifically, the relevant rules/decision(s) from which influence originates; (2)
the target institution and, more specifically, the relevant parts of the institution itself
or the issue area governed by it that are subject to the influence of the source insti-
tution; (3) a unidirectional causal pathway connecting the two institutions. Causa-
tion requires that influence runs unidirectionally from the source to the target.
Hence, our understanding of the term interaction does not imply that influence
runs back and forth between the institutions involved.
This approach implies that there cannot be a case of interaction without an ob-
servable effect in the target institution (or the issue area governed by it). Without
any effect on the target side, there would be no detectable influence, and thus no
case of interaction. However, observable changes on the target side of the interac-
tion may be the result of action of the source institution that is only anticipated.
An institution may adapt its own rules in reaction to the development (rather than
the adoption) of new rules by another institution. For example, the Convention on
Biological Diversity reacted to the negotiations on the WTO Agreement on Trade-
Related Aspects of Intellectual Property Rights (TRIPS Agreement) by adapting its
provisions on equitable sharing of benefits from genetic resources (chapter 4).
The effect in the target institution that forms a necessary part of any case of inter-
action must be carefully distinguished from additional response action, which is not
a necessary part of a case of interaction. Consider that the Kyoto Protocol provides
incentives for carbon sequestration, which might induce states and nonstate actors
to plant fast-growing trees and encroach on established habitats protected under
the Convention on Biological Diversity (CBD) (Pontecorvo 1999). Ideally, the case
will be complete, if there are observable—in this case detrimental—effects on habi-
tats protected by the CBD. In response, the actors operating within either institution
might adopt new rules to mitigate this undesired effect. This response action would
be additional to the case of interaction, because interinstitutional influence would be
observable even in its absence. This even holds true for cases in which response
action occurs in response to interinstitutional effects that are only anticipated. Thus,
actors may anticipate adverse effects of the climate change regime on biodiversity
and may respond to them before they materialize because the climate change regime
establishes incentives that can be expected to lead to relevant behavioral changes.
Response action would then indicate effects that are not yet empirically observable,
but sufficiently well established to provide the foundation for additional decisions.
Identifying a causal relationship between two international or EU institutions is,
at least implicitly, based on counterfactual arguments. Generally, institutional inter-
action requires that observed changes in the target institution are caused by the
source institution. Establishing causal influence in the social sciences must not be
confused with accounting for all possible factors contributing to the occurrence of
an observable event. In the case of institutional interaction, it (merely) requires prov-
ing that the observed changes within the target institution or the issue area governed
by it could not be expected to have occurred in the absence of the source institution
or its relevant parts (King, Keohane, and Verba 1994, 75–85). The construction of
counterfactual scenarios addresses the hypothetical question of how the target insti-
tution and the issue area governed by it would have developed in the absence of the
source institution. This is an important and well-known method for establishing
causality in the social sciences (Tetlock and Belkin 1996; Bierstecker 1993).
However, the usefulness and reliability of counterfactual scenarios decrease
sharply with the length of the causal chains in question because of the increasing
number of intervening factors that have to be taken into consideration. Consider
that the creation of the North Atlantic Treaty Organization (NATO) arguably con-
tributed to the bipolar stability of the world order between 1950 and 1990, facilitat-
ing prosperous economic development in particular in the OECD part of the world.
As a side effect, this economic development aggravated a number of environmental
problems, including global climate change, which has since been responded to by
the creation of an international regime. Does it follow that NATO has contributed
to the emergence of the international climate change regime? As a first consequence
of this so-called problem of ‘‘Cleopatra’s nose,’’ we might prioritize obvious cases
of interaction with short causal chains over less obvious ones with longer causal
chains. Second, counterfactual analysis may be complemented by the exclusion
of alternative explanations (Bernauer 1995), that is, by exploring the question of
whether factors other than the source regime might convincingly explain the ob-
served result in the target institution. These methods are well established and widely
employed in the literature on the simple effectiveness of international regimes
(Underdal 2004).
Dealing with Complexity: Identifying ‘‘Cases’’ of Interaction

Serious causal analysis of institutional interaction requires disaggregation of a
complex situation into an appropriate number of cases of interaction. Frequently,
real-world situations comprise more than two institutions and develop over a longer
period of time; influence may run back and forth between the institutions involved
(Young 2002, 111–138; Raustiala and Victor 2004). Such complex interaction sit-
uations must be disaggregated into a suitable number of ‘‘cases’’ to allow for identi-
fying clear causal relationships of the institutions involved encompassing a single
source institution, a single target institution, and a unidirectional causal pathway
connecting the two. In particular, three types of complex interaction phenomena re-
quire disaggregation.
Disaggregation will be necessary, if two institutions interact in more than one way
at the same time. This may in particular be expected to be the case for complex in-
ternational and EU institutions such as the WTO, the regime for the protection of
the Baltic Sea, or the EU Directive on Integrated Pollution Prevention and Control
(IPPC Directive), all of which govern broad issue areas. However, even an allegedly
single-purpose institution such as the regime for the protection of the ozone layer
comprises a number of different components and includes auxiliary arrangements
such as a funding mechanism (DeSombre and Kauffman 1996) and a system for
implementation review (Victor 1998). As a consequence, two institutions can be
involved in numerous cases of interaction at the same time. For example, the Mon-
treal Protocol for the protection of the ozone layer indirectly promotes the use of
certain greenhouse gases (hydrofluorocarbons, HFCs), thus undermining the objec-
tive of the Kyoto Protocol to the UN Framework Convention on Climate Change.
At the same time, it mandates the phaseout of CFCs that are also potent greenhouse
gases, thus reinforcing efforts under the international climate change regime. More-
over, the Montreal Protocol’s noncompliance procedure provided a precedent for
the elaboration of a similar component within the climate change regime (Oberthür
2001). Although the situation involves only two institutions and influence runs
exclusively from the ozone regime to the climate change regime, the three instances
of interaction follow different causal pathways and have different properties. Each is
best analyzed as a separate case of interaction.
Disaggregation will also be useful, if an interaction situation involves more than
two institutions. For example, EU policy on industrial installations is made up of a
number of different directives. While the IPPC Directive provides an integrative
framework, other directives, such as the Air Quality Framework Directive and
related daughter directives, contain relevant quality standards, and yet others—for
example, the Large Combustion Plants Directive—set relevant emission standards
(see chapter 9). To investigate causal influence, such complex situations must be dis-
aggregated into pairs of interacting institutions with clear-cut cause-effect relation-
ships that can be analyzed separately. Accordingly, we would have to investigate
whether, and how, the development and performance of the IPPC Directive was
affected by the Air Quality Framework Directive, and separately whether, and
how, it was influenced by the Large Combustion Plants Directive, and so forth. In
addition, we would have to investigate separately how the IPPC Directive influenced
the development and performance of each of these instruments.
Disaggregation will also be appropriate, if two or more institutions coevolve
over time. If coevolution involves feedback processes, neither of the institutions in
question would exist in its current state without the existence of the other (Meinke
2002). Moreover, influence will be bidirectional. However, the observation that
two or more coevolving institutions are mutually constitutive does not explain how
these regimes exert influence on each other. Analytic disaggregation of the complex
coevolution process into a suitable number of cases of interaction with a single di-
rection of influence enables us to examine when, how, and why influence actually
runs back and forth (rather than repeatedly in the same direction). The principal
strategy is to analytically divide a coevolution process into different phases over
time (Archer 1985; Carlsnaes 1992). In doing so, we assume for the purposes of
our analysis that there was a point in time when neither of the institutions in ques-
tion was influenced by the other. While it may not be possible to identify any such
moment empirically, we will not expect that a particular decision, or set of deci-
sions, may cause effects prior to its adoption (or at least prior to the anticipation of
its adoption). For example, the EU Habitats Directive and the 1979 Bern Conven-
tion on the Conservation of European Wildlife and Natural Habitats codeveloped
for almost twenty-five years and appear to have been mutually reinforcing (chapter
10). The adoption of the Bern Convention can then be seen as triggering the first
case of interaction that caused EU member states to develop and adopt the Habitats
Directive. Incidentally, the EU Habitats Directive installed a more thorough protec-
tion scheme than the Bern Convention. In the second phase, the EU scheme thus
exerted influence on the Bern Convention that subsequently adopted the more strin-
gent EU model. Influence in this second case is also clearly directed, but it runs from
the EU Habitats Directive toward the Bern Convention. Hence, ‘‘coevolution’’ of the
two institutions involved a feedback loop and may be analyzed as a sequence of two
cases with reverse direction of influence.
The strategy of disaggregating complex situations into a suitable number of clear-

cut cases is based on the assumption that complex patterns of interaction can be
reduced to their component cases. While it might miss properties that are not inher-
ent in the individual cases but emerge only from the complexity of the situation or
from the combination of cases, this approach allows for a clear causal analysis and
appears to cover all major incidents of institutional interaction discussed in the
existing literature. Moreover, it does not preclude that we recombine cases to gain a
more complex picture. Coevolution processes such as the one between the Bern Con-
vention and the Habitats Directive can be analyzed and understood as a causal chain
in which one case of interaction triggers the next. Other complex situations appear to
be made up of clusters of parallel cases of interaction. For example, the IPPC Direc-
tive demonstrates that an institution may be concurrently affected by different source
institutions in similar ways, while it may generate, as a source institution, parallel
effects on a number of other institutions (chapter 9). Given the limited conceptual ma-
turity of the analysis of institutional interaction at present, emerging properties of
complex situations do not constitute the focus of the present volume. The possibilities
of recombining connected cases of institutional interaction and of analyzing complex
interaction situations on the basis of our concept are further explored in chapter 13.
Causal Mechanisms of Institutional Interaction
At the current stage of research on institutional interaction, it is too early for the
development of a full-fledged, deductively derived theory of its driving forces and
effects. However, the rigor of empirical analysis as well as the significance of its
results will be greatly enhanced by a reliable theoretical conception of the causal
mechanisms that might drive particular cases of institutional interaction. In this
section, we first explore the general concept of causal mechanisms of institutional
interaction. Subsequently, we develop two causal mechanisms in which the source
institution directly affects the rule-making process of the target institution through
its influence on the preferences of relevant actors within the target institution. More-
over, we present two causal mechanisms in which the source institution affects the
performance of the target institution within the latter’s own domain. We claim that
each of these causal mechanisms derived from various theories of international insti-
tutions as well as from negotiation theory and cooperation theory has its own logic
that clearly distinguishes it from the other three mechanisms. Whereas other mecha-
nisms might be constructed, they are of a rather hypothetical nature. It is highly im-
probable that they gain relevance in practice.
The Concept of Causal Mechanisms

Exploration of the causal mechanisms of institutional interaction is intended to
elucidate how causal influence travels from the source institution to the target insti-
tution. A causal mechanism is a set of statements that are logically connected
and provide a plausible account of how a given cause creates an observed effect
(Schelling 1998). It opens the black box of the cause-effect relationship between the
institutions involved in a case of institutional interaction (Elster 1989, 3–10; King,
Keohane, and Verba 1994, 85–87; Hedström and Swedberg 1998).
Causal mechanisms of institutional interaction reveal how actors matter. The
search for causal mechanisms raises the question of how developments located
at the systemic level of the institutions are related to developments located at the
actors’ level. Like any other theory in the social sciences, a concept of institutional
interaction requires a reliable micro-macro link (Buzan, Jones, and Little 1993, 104;
Alexander and Giesen 1987). In the bulk of traditional analyses of international and
EU institutions, the micro-macro link is quite obvious, because causes and effects are
located at different levels (on the discussion of the agent-structure relationship, see
Wendt 1987; Carlsnaes 1992). If we want to explain the establishment and design
of an international or EU institution, the dependent variable is located at the macro-
level, while the explanatory variable—namely, the relevant group of actors, their
perceptions, interests, and so on—is located at the microlevel. Most effectiveness
research examines the reverse direction. A given institution as the independent vari-
able is located at the macrolevel, while the dependent variable, namely effects caused
by the institution such as changes of behavior of relevant actors, is located at the
microlevel. In cases of institutional interaction, the micro-macro link is not as obvi-
ous because both the independent and the dependent variables (the source institu-
tion and the target institution) are located at the macrolevel. Yet it is difficult to
imagine that an institution could influence another institution directly without inter-
mediate (changes of) action by relevant actors.
A causal mechanism of institutional interaction consists of three separate stages.
In the first stage, the source institution, or a relevant component of it, will affect
the preferences or behavior of relevant actors within its own domain. Influence orig-
inates from the structural level and is directed at the actors’ level. This stage is thus
driven by a logic of the situation. In the second stage, this effect must lead to a
change of preferences or of individual behavior of actors relevant to the target insti-
tution. Hence, the causal mechanism comprises a theory of action that tells us why
actors behave as they do. In the third stage, individual action must produce the
Figure 2.1
The logic of causal mechanisms
effect observed within the target institution or its issue area. The causal mechanism
therefore also comprises a logic of aggregation. The analytic concept of the causal
mechanism, as illustrated in figure 2.1, is well established in the social sciences
(Coleman 1990, 1–23; Hedström and Swedberg 1998, 21–23).
The approach may be illustrated by the interaction regarding forest management
between the two international regimes on global climate change and on biodiversity
(Pontecorvo 1999; Jacquemont and Caparrós 2002). To establish the causal rela-
tionship between the two regimes, one would have to demonstrate (1) that relevant
actors react to the adoption of the Kyoto Protocol in the form of increasing carbon
sequestration in forests (or can be expected to react in this way); (2) that these be-
havioral changes encroach on biodiversity-rich habitats such as tropical rainforests;
and (3) that the Convention on Biological Diversity is (at least potentially) affected
by this change because its performance is undermined.
A causal mechanism provides an abstract model of the actual causal pathway that
a case of institutional interaction follows. Being deductively derived, such a model
must be theoretically coherent, but it cannot be empirically right or wrong (Snidal
1985). It may or may not fit a given case of interaction. A causal mechanism points
to the steps of the causal chain leading from the source institution to the target insti-
tution and highlights how institutional interaction can ideally take place. It draws
attention to the actors and their behavior that are indispensable or possibly relevant
for cases of interaction that fit their inherent rationale.
Whereas the concept of causal mechanism points to the fact that actors are indis-
pensable transmitters of influence, the precise role of actors is far from obvious. The
same actors might be active in both institutions, but this is not a necessary condi-
tion. Moreover, different types of actors and behavior may be involved. Effects
observed within the target institution will frequently be attributable to changes in
the behavior of key states, because states are the members of both international
and EU institutions and institutional obligations are primarily addressed at states.
However, other types of actors, like nongovernmental organizations, industry, the
secretariats of international institutions, or the European Commission, may play an
important role (see also Selin and VanDeveer 2003).
To identify relevant causal mechanisms of institutional interaction, we follow the
distinction of three levels of effectiveness of governance institutions, namely output,
outcome, and impact. This distinction has been introduced and widely employed in
research on the effectiveness of international regimes (Underdal 2004). No interna-
tional or EU institution may directly affect the state of the environment or another
ultimate target of governance. It merely produces collectively agreed-on knowledge
or norms prescribing, proscribing, or permitting behavior, as its immediate output.
To become effective, this output of an institution must generate some form of behav-
ioral outcome. It must result in an observable influence on the behavior of relevant
states and/or substate actors such as affected industries or private households. Fi-
nally, a behavioral outcome may or may not result in an impact on the targeted
part of the environment or other ultimate target of governance. The three levels
of effectiveness are hierarchically ordered: impact requires outcome and outcome
requires output. It should be noted that the output is a property of the institution
in the narrow sense, whereas outcome and impact occur within the issue area gov-
erned by the institution.
Institutional interaction can occur at various levels. If an institution exerts influ-
ence on another institution, such influence must originate from its output (norms,
including institutional arrangements and decisions as well as knowledge), from the
outcome within its own domain (influence on behavior of relevant actors), or from
the impact on its ultimate target of governance (e.g., influence on the global cli-
mate). And vice versa: the target institution may be affected by the source institution
at the output level (influence on its decision-making process), at the outcome level
(influence on the behavior of relevant actors within its domain), or at the impact
level (direct influence on its ultimate target of governance).
A causal mechanism of institutional interaction has to clarify in particular how a
given event within a source institution or its issue area can affect the target institu-
tion. The effect on the target is not only constitutive of a case of interaction, but it is
also more selective than the causal event originating from the source institution.
While an institution generates numerous effects that might potentially trigger inter-
action, only a minority of them results in actual institutional interaction. The key to
deriving causal mechanisms of institutional interaction lies, therefore, in answering
the question of how, and under which circumstances, the output, outcome, or im-
pact of the target institution can be influenced by a source institution. Hence, to
identify relevant causal mechanisms, we start reasoning from the target side.
Interaction Influencing the Decision-Making Process of the Target Institution

Interaction influencing the decision-making process of the target institution can be
driven by two distinct causal mechanisms. One is related to the knowledge of actors
operating within the target institution, while the other is related to the commitments
of actors entered into within the source institution. An institution will influence the
output of another institution if it affects the preferences of relevant actors within
the target institution. It is evident that interaction of this sort cannot occur against
the will of the members of the target institution. No institution, however powerful in
comparison, has a direct grip on the decision-making process of another institution.
While the source institution might affect the preferences of the members of the target
institution, the latter must decide and can thwart institutional interaction by decid-
ing not to adapt. Hence, establishing a causal relationship between source and tar-
get requires demonstrating that the members of the target institution would have
decided differently in the absence of the relevant decision of the source institution.
Cognitive Interaction The decision-making process of an international institution

will be influenced if information, knowledge, or ideas (Haas 1992; Risse-Kappen
1994; Yee 1996) produced within another institution modify the perception of rele-
vant decision makers. For example, the adoption of an innovative noncompliance
procedure under the Montreal Protocol influenced the deliberations on a similar ar-
rangement within the climate change regime. It provided a policy model that actors
to some extent followed and that served as a starting point for their deliberations
(Werksman 2005), whereas no actor was in any way committed or obliged to adopt
the model.
Cognitive Interaction is based on the premise that actors must aim at reducing
‘‘analytic uncertainty’’ (Keisuke 1993) and will be prepared to adapt their percep-
tions to new information. These perceptions then shape their interests (Checkel
1998; Risse 2000). In real-world situations, the rationality of actors is usually
‘‘bounded,’’ either because the actors do not have all relevant information available
or because their information processing capacity is limited (Keohane 1984, 100–
115; Simon 1972). If relevant actors operating within the target institution were
fully informed rational utility maximizers, as is frequently assumed in rational-
choice cooperation and regime theory (see Keohane 1984; Hasenclever, Mayer,
and Rittberger 1997), Cognitive Interaction could not be expected to occur.
Cognitive Interaction passes through the following steps. First, the source institu-
tion must produce some new information such as a report revealing new insights
on a scientific or technological problem or an institutional arrangement solving a
particular regulatory problem. Note that new information has to emerge from the
collective decision-making process of the source institution. If it were merely pre-
sented by some actors individually, it would not qualify as an output of the institu-
tion. Second, the information must be picked up by some actors capable of feeding
it into the decision-making process of the target institution. This may be done by
member states, nongovernmental organizations, or the secretariats of the institutions
involved. Third, the information obtained from the source institution must change
the order of preferences of actors relevant to the target institution, be they member
states and their representatives or nongovernmental organizations capable of influ-
encing member states. Finally, the modification of the preferences of some actors
must have an impact on the collective negotiation process of the target institution
and its output.
Cognitive Interaction is purely based on persuasion and may be conceived of as
a particular form of interinstitutional learning. It is similar to the ‘‘ideational inter-
play’’ referred to by Stokke (2001b, 10) and can occur between any two institutions
whether or not their memberships or the subjects regulated overlap, as long as there
is a similarity of problems that allows for learning across institutional borders. The
source institution does not exert any pressure on the decision makers of the target
institution. However, once sufficiently relevant actors adapt their preferences to
new information, the decision-making process of the target institution will be af-
fected. This effect will be felt even by participants in the process that have not been
convinced.
Interaction through Commitment Institutional interaction directed at the decision-

making process of the target institution can also occur if commitments entered into
by some members of the source institution affect the preferences of actors related to
the target institution. For example, the EU distributed its joint commitment to re-
duce the emission of greenhouse gases among its member states by assigning differ-
entiated targets to them in a ‘‘Burden-Sharing Agreement.’’ Since emission increases
were granted to EU laggard countries, these states lost their potential interest in
securing less demanding obligations internationally and joined the whole of the
EU in requesting stringent emission targets in the international negotiations. The
Burden-Sharing Agreement thus enabled the European Union as a bloc to pursue
and eventually secure comparatively stringent emission reductions internationally
in the Kyoto Protocol (chapter 3). This case of interaction does not touch on the
outcome or impact levels of either of the two institutions, but is confined to the out-
put level.
Interaction through Commitment is based on the desire of member states to avoid
mutually incompatible obligations, or on their desire to broaden the geographic
scope of such obligations. Cooperative arrangements regularly promise cooperation
gains to their members in exchange for their commitment to a particular way of
action, so that noncompliance with the commitments endangers the gains from co-
operation. In a Prisoner’s Dilemma situation, for example, actors frequently set up
institutional arrangements to supervise, and possibly enforce, cooperative behavior
in order to avoid free riding and achieve cooperation (Martin 1993). Hence, once
sincere cooperators have entered into an obligation within one institution, they be-
come interested in avoiding incompatible decisions in other forums, because other-
wise they might not be able to comply with their commitments. They will endeavor
to preserve a reputation of keeping their promises because possible future coopera-
tors might otherwise be less inclined to enter into agreements with them (Keohane
1984, 105–106; Young 1992, 75–76). Moreover, a commitment already subscribed
to within one institution can easily be accepted within another institution because it
does not produce additional costs of adaptation. Actors may also be expected to
actively promote the transfer of a commitment to another institution, if this results
in additional benefits such as the extension of the commitment to potential compet-
itors. Being aware of the binding force of obligations, actors may also wish to adopt
commitments in one institution in order to frame the policy choices available in
another institution.
Members of the target institution may less easily avoid Interaction through Com-
mitment than Cognitive Interaction. While the latter is purely based on persuasion,
the former is based on a modification of preferences of relevant actors motivated
by substantive costs and benefits. The mechanism relates to Stokke’s ‘‘normative’’
and ‘‘utilitarian’’ interplay (Stokke 2001b). Interaction through Commitment adds
a new dimension to the insight of negotiation analysis that the order of preferences
of the participants of negotiations within the framework of institutions as well as
the resulting constellation of preferences may be influenced by adding or subtracting
issues and/or parties, even if no single actor has changed its general interests (Sebe-
nius 1983, 1992). Other international institutions may also influence the preferences
of relevant actors and the ensuing preference constellation. They can constitute an
important determinant of negotiations, if they commit actors in the target institution
in ways that influence the range of options acceptable to them.
Interaction through Commitment evolves in the following steps. First, members
of the source institution agree on an obligation that might be relevant for the target
institution. Second, this obligation actually commits one or more states that are
members of both institutions. Third, the commitment accepted by these member
states induces one or more of them to modify their preferences related to the target
institution. Fourth, the modified preferences influence the collective decision-making
process of the target institution and its output. Ideally, Interaction through Commit-
ment will take place when actors who are already bound to an obligation originat-
ing from the source institution participate in a subsequent decision-making process
of the target institution on a related subject. However, anticipated commitments to
be entered into within the source institution may trigger the mechanism, if an actor
participates in concurrent decision-making processes, so that coevolution of the
norms of two institutions may pertain to this category.
Interaction through Commitment requires a certain overlap of both the member-
ships and the issue areas of the interacting institutions. Without a jurisdictional
overlap of issue areas, neither inconsistent commitments nor side benefits of ex-
tending commitments to potential competitors could occur. Without overlapping
memberships, the target institution would remain unaffected because none of its
members would be subject to relevant commitments. Except for rare cases in which
nonstate actors enter, formally or informally, into commitments within the frame-
work of international and EU institutions, this mechanism commonly relies on state
action because only states decide on, and are directly bound by, obligations in the
framework of these institutions.
Possible Further Causal Mechanisms It is difficult to imagine that an international

or EU institution affects another institution by means other than knowledge and
commitment. Hypothetically, a potential source institution might employ other
means at its disposal such as financial transfers or the threat of sanctions. However,
to influence the decision-making process of the target institution directly, actors

operating within the latter would have to be bribed or forced to change their voting
or negotiating behavior according to the policy of the former. While this cannot be
excluded, we are not aware of anything coming close to it in contemporary interna-
tional relations. To our knowledge, nothing of this kind has been reported even in
the case of the rather tense relations between international environmental regimes
and the WTO, or between adverse military alliances such as NATO and the
Warsaw Pact.
Furthermore, we can hardly imagine that the decision-making process of an inter-
national or EU institution is directly affected by events occurring at the outcome or
impact levels of another institution. We cannot expect, for example, that the mem-
bers of the regime for the protection of the ozone layer will adapt the rules of their
institution because the WTO triggers behavioral changes within its own issue area
(e.g., the reduction of trade barriers by member states), if there are no immediate
or anticipated effects of these behavioral changes on the performance of the ozone
regime. If the performance of the ozone regime was affected, however, we would
be faced with a case of Behavioral Interaction influencing the outcome level of the
target institution (see below). We also cannot expect the members of the ozone
regime to react to the impact of the WTO on its ultimate target, namely increased
international trade, if there is no immediate impact on the state of the ozone layer.
If the ozone layer were affected, however, we would be faced with a case of Impact-
Level Interaction (see below).
Interaction Influencing the Target Institution at the Outcome Level: Behavioral

Interaction
The target institution may also be influenced by the source institution at the out-
come level, if the latter triggers behavioral effects within its own domain that be-
come relevant for the former. All international governance institutions are designed
to influence the behavior of relevant actors in order to achieve their objectives such
as protecting the environment or liberalizing international trade (Levy, Young, and
Zürn 1995; Young 1992). In some cases, behavioral effects in one issue area affect
the implementation of another institution. Consider again the interaction between
the Kyoto Protocol on climate change and the Convention on Biological Diversity
(CBD). The Kyoto Protocol creates incentives for interested states and nonstate
actors to engage in forestry activities that threaten to undermine the performance
of the CBD. Interaction of this type influences the target institution through changes
of behavior of states and nonstate actors relevant to its implementation. The rules of
the target institution may remain totally unaffected, and if they are modified, this
adaptation occurs as an additional response to the original effect.
The causal mechanism of Behavioral Interaction is composed of the following
steps. First, the source institution produces an output with a potential effect on the
behavior of relevant actors outside the decision-making process, such as behavioral
prescriptions or proscriptions, behaviorally relevant knowledge, an offer of financial
assistance, or a decision to impose sanctions for noncompliance. Second, relevant
actors such as the member states of the source institution, other states, or nonstate
actors (e.g., companies polluting the environment or nongovernmental organiza-
tions advocating human rights) adapt their behavior significantly in response to the
output by acting differently from what we would have expected in its absence. Rel-
evant behavioral changes include unforeseen side effects and deviating behavior,
such as increased smuggling in response to trade restrictions to protect endangered
species or to stabilize the coffee price at the world market. Third, the behavioral
changes triggered by the source institution within its own issue area are either also
directly relevant for the performance of the target institution or prompt further be-
havioral changes that affect the target’s outcome level. Fourth, the behavioral effect
within the issue area of the target institution affects the performance and effective-
ness of the target institution.
Behavioral Interaction is characterized by a high ability of the source institution
to influence the target institution unilaterally. In contrast to interaction directed
at the decision-making process of the target institution, Behavioral Interaction does
not depend on a decision within the target institution, because it occurs as the aggre-
gate result of the uncoordinated behavior of actors operating within the two issue
areas involved. The effect on the target institution might even come about unnoticed
by the members of the institutions involved. A collective decision by the target
institution or the source institution, or a ‘‘political linkage’’ (Young et al. 1999, 50)
between them, may occur in response to the behavioral interaction, but such inter-
action ‘‘management’’ (Stokke 2001b) is not an essential element of this causal
mechanism.
Interaction influencing the behavioral performance of the target institution will
always originate from the behavioral effects of the source institution. Effects of an
institution on the behavior of actors outside its issue area are always a secondary
effect of behavioral effects within its own domain, irrespective of whether the inter-
action is intentionally created or not and whether it is anticipated or not. Other
hypothetical causal pathways are hardly relevant in reality. In particular, we might

speculate, and cannot exclude, that the members of an institution make decisions
that directly induce behavioral effects in the issue area of the target institution with-
out prior behavioral effects within its own domain. However, it seems a rather
remote possibility that, for example, the international ozone regime would pass
international trade rules unrelated to the protection of the ozone layer. Nothing of
that kind has been reported in the literature so far. On the one hand, all international
and EU sectoral institutions are predominantly established to govern their own issue
areas. They do not enact rules, norms, or other decisions that are directed at issue
areas exclusively governed by other institutions. On the other hand, it is difficult to
see why actors involved in international trade should adapt their behavior in re-
sponse to rules created within the ozone regime, if these rules do not address ozone
protection and thus first trigger behavioral changes within the domain of the latter.
Likewise, effects at the impact level within one institution cannot directly affect the
behavior of actors governed by another institution. If they were also relevant for the
target institution, we would be faced with a case of Impact-Level Interaction.
Interaction Influencing the Impact Level of the Target Institution

Principally, an institution may also be affected by another institution at the impact
level. In this case, its ultimate target of governance, such as international trade or
the ozone layer, is directly influenced by side effects originating from the ultimate
target of governance of another institution. A stylized example that we owe to
Arild Underdal may illustrate this least intuitive causal mechanism. Consider that
protection of the stocks of cod and herring are the ultimate targets of two separate
international institutions. Because cod eats herring, successful protection of cod,
resulting in a growing population of this species, will unintentionally decrease the
population of herring. In this case, the two institutions involved are not linked
at the level of output (neither the norms nor knowledge produced within the cod
regime influence the norms protecting herring), nor through behavioral changes
(decreased fishing of cod does not directly influence the fishing activities related to
herring). They are ‘‘functionally linked’’ (Young et al. 1999; Young 2002) at the im-
pact level because the effect of the source institution on its ultimate regulatory target
(increased population of cod) directly influences the ultimate regulatory target of the
target institution (decreasing the population of herring).
Impact-Level Interaction is characterized by the following causal chain. First,
the source institution produces an output, which might trigger behavioral effects.
Second, states and nonstate actors operating within the issue area governed by the
source institution adapt their behavior in response to this signal. Third, these behav-
ioral changes have an impact on the ultimate target of governance of the source
institution. Fourth, this impact affects the target institution’s ultimate target of gov-
ernance. Fifth, this impact on the target institution’s ultimate target of governance is
relevant for the performance and effectiveness of the target institution. For the same
reasons as in the case of Behavioral Interaction, Impact-level Interaction is charac-
terized by a high ability of the source institution to influence the target institution
unilaterally. Possible interaction ‘‘management’’ (Stokke 2001b) by the institutions
involved—be it separate or in the form of a ‘‘political linkage’’ (Young et al. 1999,
50) between them—is not an essential part of the causal mechanism.
Interaction exerting influence on the impact level of the target institution has to
run through the impact level of the source institution. The output of the source in-
stitution cannot directly affect the target institution’s ultimate target of governance,
because neither norms nor knowledge nor financial assistance nor sanctions can di-
rectly affect any ultimate target of governance such as the state of the ozone layer
or free trade without intermediate behavioral changes of relevant actors. It is also
difficult to see how behavioral effects of an institution within its own domain could
directly affect another institution’s ultimate target of governance. They must first
create behavioral effects within the issue area of the target institution that subse-
quently affect the target’s ultimate target of governance—as is the case in the exam-
ple of biodiversity-relevant forestry activities induced by the Kyoto Protocol. In this
case, the effect on the target institution’s ultimate target of governance (the earth’s
biological diversity) follows from a case of Behavioral Interaction.
Concluding Remarks
Figure 2.2 illustrates the four general causal mechanisms of institutional interaction.
They differ from each other by their underlying rationale. The bold arrows indicate
the principal course of influence. In cases of Cognitive Interaction (1) and of Interac-
tion through Commitment (2), influence originates at the output level of the source
institution and directly affects the output level of the target institution. In cases of
Behavioral Interaction (3), an output of the source institution changes the behavior
of relevant actors within the issue area of the source institution, before it can exert
influence on the behavior of actors relevant for the effectiveness of the target institu-
tion. In cases of Impact-Level Interaction (4), an output of the source institution cre-
ates effects at the behavioral and impact levels of that institution, before it directly
Figure 2.2
Four general causal mechanisms of institutional interaction
affects the impact level of the target institution. Hence, all causal mechanisms ulti-
mately start with a significant output of the source institution, but the link between
the domains of the source institution and the target institution is located at different
levels of effectiveness. The relatively modest number of causal mechanisms is due
to the fact that all causal mechanisms are located at one level of effectiveness,
whereas cross-level interaction is either theoretically impossible or empirically highly
improbable.
The thin arrows at the right side of figure 2.2 indicate possible secondary effects
of institutional interaction in the target institution. They may follow from the
respective causal mechanisms, but they are not part of these mechanisms, because
interaction will take place even in their absence. Rules of the target institution that
are modified upon Cognitive Interaction or Interaction through Commitment may

subsequently affect the outcome and, thereby, ultimately the impact of the target in-
stitution. Behavioral Interaction may be responded to at the output level of the tar-
get institution and may lead to subsequent effects at the impact level. If Impact-Level
Interaction is recognized, actors whose behavior is relevant to the effectiveness of the
target institution may adjust their behavior, and the rules of this institution may be
adapted at the output level. Possible response action occurring within the domain of
the source institution is not illustrated.
Consequences for the Empirical Investigation of Institutional Interaction
Three of the four general causal mechanisms developed in the preceding section pro-
vide the theoretical point of reference for the empirical investigation of institutional
interaction in this volume. We focus on Cognitive Interaction, Interaction through
Commitment, and Behavioral Interaction. Empirical case studies do not consider
Impact-Level Interaction because it is rarely as limited and focused as in the hypo-
thetical example of the regulation of cod and herring given above. In the real world,
Impact-Level Interaction is diffuse and difficult to analyze. Frequently, it is based on
complicated natural scientific links and requires investigation of long causal chains
with many intervening variables. For example, establishing Impact-Level Interaction
between the international regime for the protection of biodiversity and the interna-
tional regime on climate change would require demonstrating both that the interna-
tional climate change regime has actually slowed down climate change and that this
effect has led to an increase in, decrease in, or stabilization of biological diversity
(for illustrations of the complexities involved, see IPCC 2002; CBD 2003). It should
be noted that the difficulties in exploring Impact-Level Interaction are exclusively of
an empirical, not of a conceptual nature.
In line with the literature on the effectiveness of institutions, we exclusively exam-
ine cases of interaction that affect, or have the potential to affect, the issue areas
governed by the institutions involved. Generally, research on the effectiveness of
institutions is interested in whether, how, and to what extent institutions affect the
state of the environment or other ultimate targets of governance (e.g., Young 1999;
Haas, Keohane, and Levy 1993; Wettestad 1999). Thus, we are not interested in in-
teraction cases that are limited to efficiency gains without significant effects in the
domains of the institutions involved—for example, minor changes in reporting or
organizational structure.
The first major task of the following empirical investigation of institutional inter-
action is to explore whether, and to what extent, the three general causal mecha-
nisms are relevant in practice in the field of international and EU environmental
policy. So far, we do not know, for example, whether actual interaction cases are
evenly distributed, or whether they concentrate on one or two of these mechanisms,
and if yes, on which ones and why. For this purpose, case-study authors compiled
inventories of identifiable cases of institutional interaction, in which their respective
core institutions have been involved either as the source institution or as the target
institution. The following chapters thus provide an overview of the network of
interinstitutional relations in which the respective core institutions are embedded.
The second major task is to explore empirically how exactly institutional interac-
tion operates in particular cases. We wish to know whether each of the cases exam-
ined fits one of our three general causal mechanisms, and vice versa, whether our
three models reflect the basic properties of the empirical cases. For this reason,
case-study authors explore in depth selected cases of interaction involving their
core institutions. Together, cases were selected so as to cover a broad variety of
phenomena of institutional interaction. The case studies demonstrate both the use-
fulness of our theoretically derived causal mechanisms and the empirical variety of
actual cases of interaction. Examples of all three causal mechanisms are examined.
Moreover, case studies address the three different dimensions in which international
and EU institutions can interact. An international institution may interact vertically
with an EU institution. It may also interact horizontally with another international
institution. Finally, an EU institution may interact horizontally with another EU
institution. Because of the relatively sophisticated overarching institutional frame-
work of the EU (see above), we may expect the two forms of horizontal interaction
to differ significantly.
The third major task of the following empirical investigation is to examine the
effects of cases of interaction for international governance. Effects are assessed in
terms of their compatibility with the policy direction of the target institution. The
policy direction indicates the direction of collectively desired change or the objective
of maintaining a desired status quo against some collectively undesired change
(Gehring 1994, 433–449). It has generally been the major yardstick in the literature
on regime effectiveness. Note that the policy direction is a property of the institution
that may not be fully supported by all of its members. While the climate change re-
gime is aimed at halting global climate change, some of its members may reject the
adoption of measures to implement this objective. In contrast, conformity with the
policy direction of the source institution is conceptually less relevant. The primary
effects of institutional interaction occur in the target institution and thus outside
the source institution. We may also assume that interaction effects will to a large ex-
tent be in line with the objectives of the source institution, because they ultimately
resulted from the source institution’s decisions.
The effects of a case of institutional interaction may be beneficial, adverse, or neu-
tral for the target institution. Beneficial effects will create synergy between the two
institutions because the policy direction of the target institution is supported by
measures originating from the source institution. For example, the EU IPPC Direc-
tive creates synergy with other environmental EU directives, including those on
waste and water management, because it supports the achievement of their objec-
tives (chapter 9). Adverse effects will result in disruption of target-institution policies
because measures originating from the source institution thwart or undermine the
effectiveness of the target institution’s own measures, or they force the target institu-
tion to adopt unwanted rules. Consider that the objective of the WTO to promote
free world trade and reduce trade obstacles creates the potential for disruption of
the policies of several environmental regimes that restrict trade in certain goods in
order to achieve their objectives (chapter 8). Finally, effects on the target institution
may also be indeterminate or neutral. If Interpol and the World Customs Organiza-
tion adapt to the needs of the Convention on International Trade in Endangered
Species of Wild Fauna and Flora (CITES) on the latter’s request, this effect is neither
detrimental to, nor supportive of, the policy directions of the target institutions
(chapter 7).
Based on our theoretical approach, the empirical investigation of institutional in-
teraction creates a wealth of empirical insights about single cases of interaction and
enables us to further advance our understanding of the operation of the causal
mechanisms of institutional interaction. In chapter 13, we further differentiate our
general causal mechanisms by developing a number of Weberian ideal types of insti-
tutional interaction based on distinct rationales and deriving hypotheses about their
effects on the target institution.
References
Abbott, Kenneth W., and Duncan Snidal. 1998. Why States Act through Formal International
Organizations. Journal of Conflict Resolution 42 (1): 3–32.
Aggarwal, Vinod K. 1983. The Unraveling of the Multi-Fiber Arrangement, 1981: An Exam-
ination of International Regime Change. International Organization 37 (4): 617–646.
Alexander, Jeffrey, and Bernhard Giesen. 1987. From Reduction to Linkage: The Long View
of the Micro-Macro Link. In Jeffrey Alexander, Bernhard Giesen, Richard Münch, and Neil J.
Smelser, eds., The Micro-Macro Link, 1–42. Berkeley: University of California Press.
Andersen, Regine. 2002. The Time Dimension in International Regime Interplay. Global
Environmental Politics 2 (3): 98–117.
Archer, Margaret S. 1985. Structuration versus Morphogenesis. In Shmuel N. Eisenstadt and
Horst-Jürgen Helle, eds., Macro-Sociological Theory: Perspectives on Sociological Theory
Vol. 1, 58–88. London: Sage.
Axelrod, Robert. 1984. The Evolution of Cooperation. New York: Basic Books.
Barnett, Michael N., and Martha Finnemore. 1999. The Politics, Power, and Pathologies of
International Organizations. International Organization 53 (4): 699–732.
Bernauer, Thomas. 1995. The Effect of International Environmental Institutions: How We
Might Learn More. International Organization 49 (2): 351–377.
Bierstecker, Thomas J. 1993. Constructing Historical Counterfactuals to Assess the Conse-
quences of International Regimes: The Global Debt Regime and the Debt Crisis of the
1980s. In Volker Rittberger, ed., Regime Theory and International Relations, 315–338.
Oxford: Clarendon.
Breitmeier, Helmut. 2000. ‘‘Complex Effectiveness’’: Regime Externalities and Interaction
(Working Group III). In Jørgen Wettestad, ed., Proceedings of the 1999 Oslo Workshop of
the Concerted Action Network on the Effectiveness of International Environmental Regimes,
45–48. Oslo: Fridtjof Nansen Institute.
Buzan, Barry, Charles Jones, and Richard Little. 1993. The Logic of Anarchy: From Neoreal-
ism to Structural Realism. New York: Columbia University Press.
Carlsnaes, Walter. 1992. The Agent-Structure Problem in Foreign Policy Analysis. Interna-
tional Studies Quarterly 36 (3): 245–270.
CBD. 2003, October. Ad hoc Technical Expert Group on Biological Diversity and Climate
Change. Interlinkages between Biological Diversity and Climate Change: Advice on the Inte-
gration of Biodiversity Considerations into the Implementation of the United Nations Frame-
work Convention on Climate Change and Its Kyoto Protocol. Montreal: Secretariat of the
Convention on Biological Diversity.
Chambers, Bradnee W. ed. 1998. Global Climate Governance: Inter-Linkages between the
Kyoto Protocol and Other Multilateral Regimes. Tokyo: United Nations University.
Checkel, Jeffrey T. 1998. The Constructivist Turn in International Relations Theory. World
Politics 50 (3): 324–348.
Coleman, James S. 1990. Foundations of Social Theory. Cambridge, MA: Belknap Press of
Harvard University.
DeSombre, Elisabeth R., and Joanne Kauffman. 1996. The Montreal Protocol Multilateral
Fund: Partial Success. In Robert O. Keohane and Mark A. Levy, eds., Institutions for Envi-
ronmental Aid: Pitfalls and Promise, 89–126. Cambridge, MA: MIT Press.
Elster, Jon. 1989. The Cement of Society: A Study of Social Order. Cambridge: Cambridge
University Press.
Fearon, James D. 1998. Bargaining, Enforcement, and International Cooperation. Interna-

tional Organization 52 (2): 269–305.
Gehring, Thomas. 2002. Die Europäische Union als komplexe internationale Organisa-
tion: Wie durch Kommunikation und Entscheidung soziale Ordnung entsteht. Baden-Baden:
Nomos.
Haas, Peter M. 1992. Introduction: Epistemic Communities and International Policy Coordi-
nation. International Organization 46 (1): 1–35.
Haas, Peter M., Robert O. Keohane, and Marc A. Levy, eds. 1993. Institutions for the Earth:
Sources of Effective International Environmental Protection. Cambridge, MA: MIT Press.
Hasenclever, Andreas, Peter Mayer, and Volker Rittberger. 1997. Theories of International
Regimes. Cambridge: Cambridge University Press.
Hedström, Peter, and Richard Swedberg. 1998. Social Mechanisms: An Introductory Essay.
In Peter Hedström and Richard Swedberg, eds., Social Mechanisms: An Analytical Approach
to Social Theory, 1–31. Cambridge: Cambridge University Press.
Hix, Simon. 2005. The Political System of the European Union. 2nd edition. London: St.
Martin’s.
Hurrel, Andrew. 1993. International Society and the Study of Regimes. A Reflective
Approach. In Volker Rittberger, ed., Regime Theory and International Relations, 49–72.
Oxford: Clarendon Press.
IPCC. 2002, April. Climate Change and Biodiversity. IPCC Technical Paper V. Geneva: Inter-
governmental Panel on Climate Change.
Jachtenfuchs, Markus. 2001. The Governance Approach to European Integration. Journal of
Common Market Studies 39 (2): 245–264.
Jacquemont, Frédéric, and Alejandro Caparrós. 2002. The Convention on Biological Diver-
sity and the Climate Change Convention 10 Years After Rio: Towards a Synergy of the Two
Regimes? Review of European Community and International Environmental Law 11 (2):
139–180.
Joerges, Christian, and Ellen Vos, eds. 1999. EU Committees: Social Regulation, Law and
Politics. Oxford: Hart.
Keisuke, Iida. 1993. Analytic Uncertainty and International Cooperation: Theory and Appli-
cation to International Economic Policy Coordination. International Studies Quarterly 37 (4):
431–457.
Keohane, Robert O. 1989. International Institutions and State Power: Essays in International
Relations Theory. Boulder, CO: Westview.
Keohane, Robert O. 1993. The Analysis of International Regimes: Towards a European-
American Research Programme. In Volker Rittberger, ed., Regime Theory and International
Relations, 23–45. Oxford: Clarendon Press.
Keohane, Robert O., and Stanley Hoffmann. 1991. Institutional Change in Europe in the
1980s. In Robert O. Keohane and Stanley Hoffmann, eds., The New European Community:
Decisionmaking and Institutional Change, 1–39. Boulder, CO: Westview.
King, Gary, Robert O. Keohane, and Sidney Verba. 1994. Designing Social Inquiry: Scientific
Inference in Qualitative Research. Princeton, NJ: Princeton University Press.
Levy, Mark A. 1993. European Acid Rain: The Power of Tote-Board Diplomacy. In Peter
M. Haas, Robert O. Keohane, and Mark A. Levy, eds., Institutions for the Earth: Sources of
Effective International Environmental Protection, 75–132. Cambridge, MA: MIT Press.
Levy, Marc A., Oran R. Young, and Michael Zürn. 1995. The Study of International
Regimes. European Journal of International Relations 1 (3): 267–330.
Martin, Lisa L. 1993. The Rational State Choice of Multilateralism. In John Ruggie, ed., Mul-
tilateralism Matters: The Theory and Praxis of an Institutional Form, 91–121. New York:
Columbia University Press.
regionalen Umweltregimen. Darmstadt: Deutscher Universitäts-Verlag.
Moravcsik, Andrew. 1998. The Choice for Europe: Social Purpose and State Power from
Messina to Maastricht. Ithaca, NY: Cornell University Press.
Morrow, James D. 1994. Modelling the Forms of International Cooperation: Distribution
versus Information. International Organization 48 (3): 387–423.
Nugent, Neill. 2003. Government and Politics of the European Union. 5th edition. Durham,
NC: Duke University Press.
Oye, Kenneth A. 1985. Explaining Cooperation under Anarchy: Hypotheses and Strategies.
World Politics 38 (1): 1–24.
Pollack, Mark A. 1997. Delegation, Agency, and Agenda Setting in the European Commu-
nity. International Organization 51 (1): 99–134.
Raustalia, Kal, and David G. Victor. 2004. The Regime Complex for Plant Genetic Re-
Risse, Thomas. 2000. ‘‘Let’s Argue!’’ Communicative Action in World Politics. International
Organization 54 (1): 1–39.
Risse-Kappen, Thomas. 1994. Ideas Do Not Float Freely: Transnational Coalitions, Domestic
Structures, and the End of the Cold War. International Organization 48 (2): 185–214.
Rittberger, Volker, ed. 1993. Regime Theory and International Relations. Oxford: Clarendon
Press.
Rosendal, Kristin. 2001. Impacts of Overlapping International Regimes: The Case of Biodi-
versity. Global Governance 7 (1): 95–117.
Scharpf, Fritz W. 1999. Governing in Europe: Effective and Democratic? Oxford: Oxford
University Press.
Schelling, Thomas. 1998. Social Mechanisms and Social Dynamics. In Peter Hedström and
Richard Swedberg, eds., Social Mechanisms: An Analytical Approach to Social Theory, 32–
44. Cambridge: Cambridge University Press.
Sebenius, James K. 1983. Negotiation Arithmetics: Adding and Subtracting Issues and Parties.
International Organization 37 (2): 281–316.
Sebenius, James K. 1992. Challenging Conventional Explanations of International Coopera-
tion: Negotiation Analysis and the Case of Epistemic Communities. International Organiza-
tion 46 (1): 323–365.
Selin, Henrik, and Stacy D. VanDeveer. 2003. Mapping Institutional Linkages in European
Air Pollution Politics. Global Environmental Politics 3 (3): 14–46.
Simmons, Beth A., and Lisa L. Martin. 2002. International Organizations and Institutions.
In Walter Carlsnaes, Thomas Risse, and Beth Simmons, eds., Handbook of International
Relations, 192–211. London: Sage.
Simon, Herbert A. 1972. Theories of Bounded Rationality. In Charles B. McGuire and Roy
Radner, eds., Decision and Organization, 161–176. Amsterdam: North Holland.
Snidal, Duncan. 1985. The Game Theory of International Politics. World Politics 38 (1): 25–
57.
Stokke, Olav Schram. 1999. Governance of High Seas Fisheries: The Role of Regime Link-
ages. In Davor Vidas and Willy Østreng, eds., Order for the Oceans at the Turn of the Cen-
tury, 157–172. The Hague: Kluwer Law International.
Stokke, Olav Schram. 2000. Managing Straddling Stocks: The Interplay of Global and Re-
gional Regimes. Ocean and Coastal Management 43 (2–3): 205–234.
Stokke, Olav Schram. 2001a. Conclusions. In Olav Schram Stokke, ed., Governing High Seas
Fisheries: The Interplay of Global and Regional Regimes, 329–360. Oxford: Oxford Univer-
sity Press.
Stokke, Olav Schram. 2001b. The Interplay of International Regimes: Putting Effectiveness
Tetlock, Philip E., and Aaron Belkin, eds. 1996. Counterfactual Thought Experiments in
World Politics: Logical, Methodological, and Psychological Perspectives. Princeton, NJ:
Princeton University Press.
Underdal, Arild. 1992. The Concept of Regime ‘‘Effectiveness.’’ Cooperation and Conflict 27
(3): 227–240.
Victor, David G. 1998. The Operation and Effectiveness of the Montreal Protocol’s Non-
Compliance Procedure. In David G. Victor, Kal Raustiala, and Eugene B. Skolnikoff, eds.,
The Implementation and Effectiveness of International Environmental Commitments: Theory
and Practice, 137–176. Cambridge, MA: MIT Press.
Victor, David G., Kal Raustiala, and Eugene B. Skolnikoff, eds. 1998. The Implementation
and Effectiveness of International Environmental Commitments: Theory and Practice. Cam-
bridge, MA: MIT Press.
Wendt, Alexander. 1987. The Agent-Structure Problem in International Relations Theory.
Werksman, Jacob. 2005. The Negotiation of a Kyoto Compliance System. In Olav Schram
Stokke, Jon Hovi, and Geir Ulfstein, eds., Implementing the Climate Regime: International
Compliance. 17–38. London: Earthscan.
Wettestad, Jørgen. 1999. Designing Effective Environmental Regimes: The Key Conditions.
Cheltenham: Edward Elgar.
Yee, Albert S. 1996. The Causal Effects of Ideas on Politics. International Organization 50
(1): 69–108.
Young, Oran R. 1982. Regime Dynamics: The Rise and Fall of International Regimes. Inter-
national Organization 36 (2): 277–297.
Young, Oran R. 1992. The Effectiveness of International Institutions: Hard Cases and Critical
Variables. In James N. Rosenau and Ernst-Otto Czempiel, eds., Governance without Govern-
ment: Order and Change in World Politics, 160–194. Cambridge: Cambridge University
Press.
Young, Oran R. 1994. International Governance: Protecting the Environment in a Stateless
Society. Ithaca, NY: Cornell University Press.
Young, Oran R. ed. 1999. The Effectiveness of International Environmental Regimes: Causal
Connections and Behavioral Mechanisms. Cambridge, MA: MIT Press.
3
The Climate Change Regime: Interactions with
ICAO, IMO, and the EU Burden-Sharing Agreement
Reflecting the vast scope and complexity of the climate change challenge, the inter-
national regime on climate change is one of the broadest and most complex interna-
tional governance systems in the field of the environment and beyond. Representing
the biggest environmental challenge at the beginning of the twenty-first century, cli-
mate change has a variety of impacts on the natural environment and on human so-
ciety. Various human activities and sectors of society contribute to the problem and
will, therefore, be influenced by any effective policy response (IPCC 2001a, 2001b).
Consequently, the climate change regime is one of the politically most important in-
ternational environmental institutions and spans an enormous scope. Since interna-
tional negotiations on a UN Framework Convention on Climate Change began in
1991, the growth in the number, detail, and complexity of the relevant international
rules has become particularly apparent with the adoption of the 1997 Kyoto Proto-
col and the subsequent elaboration of its provisions, including a number of innova-
tive elements such as emissions trading and opportunities to take credit for forestry
activities (e.g., Oberthür and Ott 1999; Yamin and Depledge 2004).
Given its enormous scope, it is hardly surprising that the climate change regime
interacts with a great number of other international institutions and EU legal instru-
ments, as further detailed in this chapter. The chapter first briefly introduces the
main elements of the international regime on climate change. This is followed by
an overview of the major interactions of the climate change regime with other inter-
national institutions and EU legal instruments. The chapter then focuses on the
interaction with three other institutions in more detail. The interaction with the
International Maritime Organization (IMO) and the International Civil Aviation
Organization (ICAO) exemplifies the at times problematic relationship of the cli-
mate change regime with institutions from other policy fields. The climate change
regime’s request to the IMO and ICAO to restrict greenhouse gas emissions from
54 Sebastian Oberthür
international transport raised the issue of which of the institutions involved should
possess regulatory authority in this respect. The request has largely failed to draw an
effective response to date, because the requested restrictions are not in the immediate
interest of the target institutions. Similar issues arise more frequently especially be-
tween environmental and economic institutions. Subsequently, the chapter analyzes
the climate change regime’s interaction with the agreement on differentiated emis-
sion limitation and reduction commitments of EU member states (‘‘Burden-Sharing
Agreement’’). This interaction provides an example of how EU legal instruments can
facilitate and strengthen international environmental governance. This rather posi-
tive perspective on the EU Burden-Sharing Agreement contrasts with the harsh criti-
cism by several non-EU countries. The concluding section summarizes the findings.
The International Regime on Climate Change
The international regime on climate change is built on two international treaties, the
UN Framework Convention on Climate Change (UNFCCC) of 1992 and its Kyoto
Protocol adopted in 1997 (Bodansky 1993; Oberthür and Ott 1999). The rules
under the Kyoto Protocol were further specified in agreements reached in 2001
(Bail, Marr, and Oberthür 2003). As of mid-2005, the Convention had 189 parties
and the Protocol had been ratified by 150 countries and the EU. The EU and its
member states are all parties to both the Convention and the Protocol.1 The Kyoto
Protocol entered into force in February 2005. However, the new U.S. President
George Bush in March 2001 decided not to ratify the Kyoto Protocol.
The Convention established the regime by defining the principles that guide its de-
velopment (Art. 3) and its ultimate objective: to stabilize atmospheric concentrations
of greenhouse gases (GHGs) ‘‘at a level that would prevent dangerous anthropo-
genic interference with the climate system’’ (Art. 2). It also established the soft aim
that industrialized countries would strive to return their GHG emissions to 1990
levels by 2000. It covers all GHGs ‘‘not controlled by the Montreal Protocol’’ for
the protection of the ozone layer and establishes that removals by sinks such as for-
ests are to be taken into account.
The Kyoto Protocol for the first time establishes legally binding emission-
reduction commitments for industrialized countries. These differentiated commit-
ments must amount to an overall reduction of at least 5 percent from 1990 levels
by 2008–2012 (the ‘‘commitment period’’). The commitments cover carbon dioxide
(CO2 ), methane (CH 4 ), nitrous oxide (N2 O), and three (groups of) fluorinated
Climate Change Regime 55
gases, namely hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), and sulfur

hexafluoride (SF6 ). Removals and emissions of GHGs from afforestation, reforesta-
tion, and deforestation are to be accounted for (Art. 3.3). In addition, parties agreed
in 2001 that forest management and agricultural activities (cropland management,
grazing-land management, and revegetation) could be taken into account as addi-
tional sink categories under Article 3.4 of the Protocol (Bail, Marr, and Oberthür
2003).
The Protocol furthermore establishes three innovative ‘‘Kyoto Mechanisms’’ that
allow countries to meet their emission obligations by acquiring emission credits
from abroad. An emissions-trading system allows industrialized countries with ex-
cess emission allowances to transfer them to other countries in need of such allow-
ances (Art. 17). Under the ‘‘Joint Implementation’’ (JI) scheme according to Article
6 of the Protocol, an investor and a host industrialized country can generate addi-
tional emission reductions by implementing a suitable project jointly, with the in-
vestor receiving (part of) the resulting emission credits. Similarly, industrialized
countries can invest in emission-reduction projects (including sinks projects) in
developing countries to earn additional emission credits under a Clean Development
Mechanism (CDM, Art. 12). Further rules and guidelines on the operation of the
Kyoto Mechanisms form part of the agreements reached in 2001 (Bail, Marr, and
Oberthür 2003).
The institutional structure of the UNFCCC and the Kyoto Protocol are closely
related. The Conference of the Parties (COP), which usually meets once a year, is
the supreme decision-making body of the Convention. It is assisted by two stand-
ing subsidiary bodies, the Subsidiary Body for Scientific and Technological Advice
(SBSTA) and the Subsidiary Body for Implementation (SBI). The Convention fur-
thermore establishes a financial mechanism to assist developing countries in their
implementation, which is operated by the Global Environment Facility (GEF), and
a secretariat (located in Bonn). It also acknowledges the role of the Intergovern-
mental Panel on Climate Change (IPCC) established by the WMO and UNEP in 1988
to provide scientific advice to its parties. Since no rule on voting could be agreed, all
decisions under the Convention have so far required consensus. Until the entry into
force of the Protocol in February 2005, the development of the regime occurred in
the framework of the Convention. Whereas the Subsidiary Bodies and the financial
mechanism of the Convention as well as the secretariat are also adapted to serve
under the Protocol, the COP sessions concurrently serve as the meeting of the
parties to the Protocol (COP/MOP). In addition, parties to the Protocol have
elaborated detailed rules on reporting, monitoring, and review of information as

well as a compliance system to determine and address cases of noncompliance
(Bail, Marr, and Oberthür 2003; Yamin and Depledge 2004).
The Climate Change Regime as Source and Target of Institutional Interaction
The climate change regime interacts with many other environmental and nonenvi-
ronmental international institutions and EU legal instruments. Twenty-four specific
cases of horizontal and vertical institutional interaction are listed in table 3.1. This
list is not necessarily exhaustive. Not included are numerous cases in which a spe-
cific interaction has not occurred yet and/or the causal pathway leading from one
institution to the other is rather long. For example, trade liberalization advanced
by the World Trade Organization (WTO) may lead to rising GHG emissions due
to induced growth in international trade, as may the EU Single Market. Further-
more, effective climate protection may prevent the spreading of health diseases (rel-
evant to the World Health Organization, WHO), help efforts to preserve biological
diversity (Convention on Biological Diversity, EU Habitats Directive) and wetlands
(Ramsar Convention) and combat desertification (Convention to Combat Desertifi-
cation), and so on (see IPCC 2001a). In other instances, a potential for the emer-
gence of future interaction exists, for example between the Kyoto Mechanisms and
the WTO (e.g., Chambers 1998, 2001; Charnovitz 2003). In these cases, the inter-
action has so far remained rather indirect and unspecific. Furthermore, more EU
legal instruments affect GHG emissions, but including them would have been be-
yond the scope of this study.
All major causal mechanisms of institutional interaction are represented in the
twenty-four cases. Cognitive Interaction is apparent from the model function that
the compliance procedure of the Montreal Protocol has performed in the elabora-
tion of the compliance system of the Kyoto Protocol. The Montreal Protocol has
also served as a model that was not accepted (due to a blocking minority) with
respect to the establishment of technology and economic assessment panels. Other
international institutions have also served as templates of various elements of the
climate change regime, but including them would have been beyond the scope of
this chapter.
In other instances, the commitments entered into under the climate change regime
have affected the rules of other international institutions and EU legal instruments
(Interaction through Commitment). Thus, the GEF operates the financial mechanism
Table 3.1
Interactions of the Climate Change Regime
Montreal Protocol on Has served as a model in several respects (e.g.,
Substances That Deplete the compliance procedure)
Ozone Layer Has served as a model that was blocked by a minority
with respect to the establishment of technology and
economic assessment panels
Has helped phase out ozone-depleting substances that
are also potent GHGs

Has supported use of fluorinated GHGs regulated
under the Kyoto Protocol (while the latter has provided

a disincentive for such use to replace ozone-depleting
substances)
Convention on Biological May suffer from establishment of monocultural tree
Diversity plantations induced by climate change regime
Ramsar Convention on May benefit from additional resources for wetland
Wetlands management or suffer from conversion of wetlands for
carbon sequestration induced by climate change regime
Convention to Combat May benefit from forestry activities promoted under
Desertification the climate change regime that help combat
desertification
International Civil Aviation Was asked by climate change regime to act on GHG
Organization emissions from international aviation
International Maritime Was asked by climate change regime to act on GHG
Organization emissions from international shipping
World Trade Organization Is used as a major argument against elaboration of
trade-relevant climate-protection measures (‘‘chill effect’’)
World Bank Has greened its policies to some extent in response to
the climate change regime
Global Environment Facility Has been asked to operate the financial mechanism of
the climate change regime
EU Landfill Directive Results in reductions of methane emissions and thus
helps implement the Kyoto Protocol
EU Renewable Energy Directive Is to result in increasing use of non-GHG-emitting
energy sources and thus helps implement the Kyoto
Protocol
EU Directive on the Internal Is expected to result, inter alia, in lower energy prices
Market for Electricity counteracting efforts to save energy and reduce GHG
emissions
EU Directives on car emission Require cars to be equipped with catalytic converters,
standards leading to increases of GHG emissions
Table 3.1
(continued)
EU GHG monitoring Responds to international reporting and monitoring
requirements under the Kyoto Protocol
EU Burden-Sharing Agreement Facilitated agreement on and strengthened targets
under Kyoto Protocol
Was codified in supranational EU law in response to
Kyoto Protocol
Helps implement the Kyoto Protocol by strengthening
enforcement in the EU
EU Regulation and Directive Was triggered by the Kyoto Protocol
on fluorinated greenhouse gases Is expected to lead to reductions of emissions of
fluorinated GHGs
EU Emissions Trading Directive Was triggered by the Kyoto Protocol
Is expected to result in reductions of GHG emissions
of the climate change regime and the World Bank has to some extent made its
policies more climate-friendly. As discussed further in the next section, ICAO and
IMO have initiated some activities to address GHG emissions from international
transport in response to a request by the Kyoto Protocol. Furthermore, the climate
change regime has shaped the EU’s legislation implementing the Kyoto Protocol,
including the EU GHG monitoring mechanism, the EU regulatory framework on
fluorinated GHGs, the codification of the EU Burden-Sharing Agreement, and the
EU Emissions Trading Directive. The climate change regime has also been the target
of Interaction through Commitment. For example, free-trade commitments under
the WTO (chapter 8) have contributed to preventing elaboration of trade-related cli-
mate protection measures. In contrast, the EU’s commitment to its Burden-Sharing
Agreement facilitated and strengthened the Kyoto Protocol, as analyzed further in
this chapter.
The climate change regime has also served as a source and a target of Behavioral
Interaction. The EU Landfill Directive, the Renewable Energy Directive, the Emis-
sions Trading Directive, EU rules on the internal market for electricity, the directives
on car emission standards, the EU Burden-Sharing Agreement, the EU regulatory
framework on fluorinated GHGs, the IPPC Directive (chapter 9), and other EU legal
instruments affect the level of GHG emissions within the EU. The Montreal Protocol
has had synergistic and disruptive effects by phasing out ozone-depleting substances
such as chlorofluorocarbons (CFCs) that are also potent GHGs, while encouraging
the use of other fluorinated GHGs (Oberthür 2001). The Kyoto Protocol, in turn,
provides incentives for forestry activities that are expected to support the objectives
of the Convention to Combat Desertification (CCD). In contrast, the Protocol is
likely to have a disruptive effect on the Convention on Biological Diversity (CBD)
by providing incentives for investments in fast-growing monocultural forest planta-
tions (Pontecorvo 1999; see also chapter 4). It may also violate the prohibition of
dumping at sea under the OSPAR Convention for the protection of the Northeast
Atlantic by providing an incentive to sequester CO2 in North Sea oil fields (chapter
5). Due to scientific uncertainties, uncertain behavioral effects, and unknown appli-
cation of rules in practice, the behavioral effects of the climate change regime are not
always unambiguously synergistic or disruptive. For example, whether and to what
extent wetland conservation and management regulated under the Ramsar Conven-
tion on wetlands will benefit from additional resources made available through the
climate change regime or may be harmed by conversion of wetlands for carbon se-
questration depends heavily on the future development and application of relevant
rules.
The relations of the climate change regime with nonenvironmental institutions
have been disruptive more frequently than those with other environmental institu-
tions. Of the five identified environmental-economic interactions with ICAO, IMO,
the WTO, the World Bank, and the EU electricity market, all except the one con-
cerning the World Bank have been disruptive. In contrast, only five of the about
twenty interactions with other environmental institutions have resulted in disrup-
tions (including the interaction with the OSPAR Convention; see chapter 5).
Political decision making can lead to improvements. For example, some decisions
have been made in the framework of the climate change regime to mitigate the
disruptive effect on the CBD (Jacquemont and Caparrós 2002). The relationship be-
tween the climate change regime and the CBD as well as the Convention to Combat
Desertification and the Ramsar Convention and others are also actively managed to
enhance synergy. Since most cases have a potential for further improvement, the
situation may change in the future.
Requesting Change from Unfriendly Institutions: Regulatory Competition between

the Kyoto Protocol and ICAO and IMO
Although climate change is not among their main concerns, ICAO and IMO started
to address GHG emissions from international transport in response to a request
contained in the 1997 Kyoto Protocol. Both organizations have, however, been far
from enthusiastic about the newly acquired task. The limited action they have taken
has mainly been driven by the threat of regulation by the climate change regime and
of unilateral action by major players. A more elaborate analysis of this case of Inter-
action through Commitment can be found in Oberthür 2003.
Structure and Objectives of ICAO and IMO

ICAO and IMO are the prime international organizations responsible for interna-
tional aviation and shipping, respectively. Their major objectives are the promotion
and enhancement of these modes of international transport. Shipping and aviation
interests (owners, builders, operators) are their main stakeholders. As of mid-2005,
ICAO had 188 and IMO 165 member states (http://www.icao.int; http://www.imo
.org).
IMO possesses an Assembly of all parties, a Council with a limited membership
elected by the Assembly, and various committees. Its supreme governing body is
the Assembly. In between its biennial meetings, the Assembly’s functions are largely
performed by the Council. However, with respect to the central task of IMO—the
elaboration of international agreements (Art. 3(b), IMO Convention)—the Council
may not recommend adoption of regulations or amendments to such regulations on
behalf of the Assembly. IMO agreements become binding on parties subject to their
ratification (http://www.imo.org).
While ICAO also possesses an Assembly, a limited-membership Council, and
various committees, its supreme governing body is the Council. The Assembly meets
once every three years and provides general policy guidelines for the work of the
other ICAO bodies framed in ‘‘Assembly Resolutions.’’ The Council governs the or-
ganization in the interim. In addition to passing resolutions and recommendations,
it adopts legally binding standards and recommended practices that are included in
annexes to the ICAO Convention. An international standard adopted by the Coun-
cil immediately binds all ICAO members that do not explicitly decide to deviate
from the standard. Member states undertake to comply with the organization’s reg-
ulations, which also apply over the high seas (Art. 12, ICAO Convention; see Buer-
genthal 1969; http://www.icao.int).
Both organizations have assumed at least partial competence for regulating envi-
ronmental matters relating to their mode of international transport. In the case of
IMO, such authority is an explicit part of its mandate. According to Article 1(a) of
its Convention, the purposes of IMO include ‘‘to encourage the general adoption
of the highest practicable standards in matters concerning maritime safety, efficiency

of navigation and the prevention and control of marine pollution from ships.’’
Among the five open-ended committees of the organization is a Marine Environ-
ment Protection Committee (MEPC) that is primarily concerned with adopting and
amending the organization’s environmental conventions and reports to the Council
and the Assembly. IMO has adopted a number of conventions addressing marine
pollution and oil spills, most importantly the International Convention for the Pro-
tection of the Marine Environment from Pollution by Ships of 1973/78 (MARPOL
1973/78).
In contrast, environmental protection is not among the explicit objectives of
ICAO (compare Art. 3, IMO Convention, and Art. 44, ICAO Convention). How-
ever, according to Article 44(d) of the ICAO Convention, the organization aims at,
inter alia, meeting ‘‘the needs of the peoples of the world for safe, regular, efficient
and economical air transport.’’ The Convention allows for the establishment of
committees as appropriate (Diederiks-Verschoor 1993, 36–40). The Committee on
Aviation Environmental Protection (CAEP) that was established by the ICAO Coun-
cil in 1983—superseding two committees on aircraft noise and aircraft engine emis-
sions created in the 1970s—prepares the Council’s decisions on environmental
matters. ICAO has elaborated a limited number of environmental standards, most
importantly regarding nitrogen oxide emissions of aircraft.
The Trigger of the Interaction: The Request by the Kyoto Protocol

A request of the climate change regime to ICAO and IMO marks the beginning of
the interaction (figure 3.1). Article 2.2 of the Kyoto Protocol implicitly contains this
request by committing industrialized countries to ‘‘pursue limitation or reduction
of emissions of greenhouse gases not controlled by the Montreal Protocol from
aviation and marine bunker fuels [i.e., the fuel sold to and burned by aircraft and
ships in international transport], working through the International Civil Aviation
Organization and the International Maritime Organization, respectively.’’ It was
the result of a political deadlock during the elaboration of the Kyoto Protocol. In
protracted discussions, the parties to the UNFCCC were unable to reach agreement
on how to deal with GHG emissions from international transport. Consequently,
such emissions are not subject to the emission targets agreed on in Kyoto, and the
parties decided to turn to ICAO and IMO (Oberthür 2003, 193).
International transport contributes a significant and growing share to global GHG
emissions. According to the available data, international aviation and shipping
Figure 3.1
Kyoto Protocol triggers action by IMO and ICAO
account for about 4 percent of global CO2 emissions in total. This is in the range
of German CO2 emissions in the 1990s. The IPCC has estimated the share of inter-
national aviation to have amounted to about 2 percent in 1992. Aviation’s overall
contribution to radiative forcing even amounted to 3.5 percent due to other factors
(buildup of ozone, contrails, and so on). International shipping has been found to
have been responsible for about 1.8 percent of global CO2 emissions in 1996 (while
accounting for a much larger volume of freight than air transport). While CO2 emis-
sions from aviation were set to increase dynamically by about 3 percent per year be-
tween 1990 and 2015, shipping emissions are to increase by at least 1–2 percent per
year (IPCC 1999; IMO 2000; WBGU 2002; UNFCCC 2002). On the basis of these
growth rates, emissions from international transport would double around 2020.
The request of the Kyoto Protocol increased the pressure on IMO and ICAO to
address GHG emissions from international transport. The Protocol committed its
parties (in particular industrialized-country parties) to take action on GHG emis-
sions from international transport. Because of the large overlap in membership, this
commitment essentially concurrently extended to most member states of IMO and
ICAO. Furthermore, the commitment contained the implicit threat that restrictions
on GHG emissions from aviation and marine bunker fuels could in principle be
imposed under the climate change regime, which would be of immediate relevance
to air and sea transport fostered by ICAO and IMO. The regulatory competition
with the climate change regime has been an important motivation for both organi-
zations’ efforts to deal with climate change. Thus, the ICAO Assembly called on the
ICAO Council not to leave the initiative on aviation matters related to the environ-
ment ‘‘to other organizations’’ (Abeyratne 2001, 38). Less strongly, the IMO As-
sembly declared that IMO ‘‘should take the lead in developing GHG limitation
and reduction strategies and mechanisms for international shipping’’ (IMO 2003).
Overall, the case for taking action on climate change within both organizations
was strengthened.
The threat of regulatory competition has, however, remained weak. First of all,
the Kyoto Protocol only entered into force in 2005. In the interim, regulation of
GHG emissions from international transport had basically fallen off the agenda of
the UNFCCC for several years after Kyoto. Activities within the UNFCCC focused
on, and remained confined to, improving the informational basis. Initiatives by the
EU and others regularly failed to significantly advance the issue due to resistance in
particular by the United States and oil-producing countries (Oberthür 2003, 199).
Another potential driving force, the threat of unilateral action, has also remained
weak. Transnational aviation and shipping interests at times prefer uniform inter-
national regulation to a disparate regulatory environment with widely varying na-
tional standards. Norway introduced taxation of kerosene in spring 1999 but was
forced to abandon the tax when international airlines complained and refused to
pay (Oberthür and Ott 1999, 112). The EU has also considered introducing an
emission charge/levy for (international) air transport for several years (European
Commission 1999). In 2001, the EU Environment Council declared that the EU
should take action if no concrete measures were agreed on within ICAO by 2002
(Council of the European Union 2001, para. 5). However, no specific action was in
sight as of the end of 2004. Shipping has received less attention mainly due to the
fact that it is considered relatively environmentally friendly as compared with air
transport. Although the EU announced that the European Commission would iden-
tify and undertake specific actions to reduce GHG emissions from shipping if no
such action was agreed on within IMO (ECON 2003, 19), no initiative has resulted
yet.
The Response by ICAO and IMO: Slow with Uncertain Results

Although they had recognized the problem of climate change before, both ICAO
and IMO started to consider effective action on GHG emissions primarily in re-
sponse to the Kyoto Protocol. ICAO had emphasized the need for further study of
the problem in the early 1990s (Crayston 1993, 53) and requested the afore-
mentioned Intergovernmental Panel on Climate Change (IPCC) in 1996 to prepare
what became the IPCC Special Report on Aviation and the Global Atmosphere
(IPCC 1999). Referring to the Kyoto Protocol, the thirty-second ICAO Assembly in
1998 then asked the CAEP ‘‘to study policy options to limit or reduce the green-
house gas emissions from civil aviation, taking into account the findings of the
IPCC special report and the requirements of the Kyoto Protocol’’ (ICAO 1998, Ap-
pendix F; see also Crayston and Hupe 2000, 32). IMO first addressed the issue
in September 1997 when the Kyoto Protocol was already looming. An IMO confer-
ence called on the organization to undertake a study of CO2 emissions from ships
and the MEPC to identify feasible CO2 -reduction strategies. In November 1998,
the MEPC decided to commission a study on GHG emissions from ships, noting
explicitly that IMO had the mandate from the Kyoto conference to address the issue
(Fayette 2001, 204–208). The study was presented in mid-2000 (IMO 2000).
Both organizations originally considered a similar range of measures (including
levies and charges, voluntary measures, technical and operational measures, emis-
sion standards, and emissions trading), and they have reached similar conclusions
on a number of them. The potential of voluntary measures is rather limited in both
international aviation and shipping, given that governments lack a stick to move in-
dustry beyond ‘‘business as usual’’ (ECON 2003, 26–27, 36; IMO 2000; Bode et al.
2002, 175–176). Realization of technical and operational improvements is further
considered and promoted by both ICAO and IMO (ICAO 2004, Appendix H;
IMO 2003), but is either expected to occur regardless of further action in the fore-
seeable future (IPCC 1999) or hinges on provision of appropriate incentives for
shipbuilders and shipowners (IMO 2000). Finally, both organizations have in effect
dismissed emission standards and internationally coordinated levies or charges as
impractical or unwarranted (despite continuing proposals for their introduction:
e.g., WBGU 2002). The abandonment of emission standards is particularly note-
worthy in the case of the IMO, because of the organization’s experience with such
standards. In particular, it had been considered that GHG emission standards could
become part of Annex VI of the IMO-administered MARPOL Convention on
air pollution from ships that was elaborated in the 1990s (Fayette 2001) and cur-
rently contains standards for emissions of sulfur dioxide and nitrogen oxides (Pisani
2002).
ICAO has been particular in its discouragement of the unilateral introduction of
levies by individual countries. An ICAO recommendation on reciprocal tax exemp-
tions for foreign aircraft has become the norm in international air transport by its
incorporation into most bilateral air transport agreements between states. While
introducing an emission charge might in principle still be possible, it is difficult to
design such a charge so that it would not be considered taxation. In addition,
ICAO has defined rather restrictive guidelines for emission-related levies. Accord-
ingly, ‘‘The funds collected should be applied in the first instance to mitigating
the environmental impact of aircraft engine emissions’’ (addressing specific damage,
funding research). Furthermore, such charges should not serve any fiscal aims,
should be related to costs, and ‘‘should not discriminate against air transport
compared with other modes of transport’’ (ICAO 1996; see also Abeyratne 2001).
While this policy is not legally binding on members and leaves some room for
interpretation, the ICAO Assembly mandated further work on the issue by 2007
and, in the interim, urged countries to refrain from unilateral action (ICAO 2004,
Appendix I).
Despite the similarities mentioned above, ICAO and IMO have headed off in
different directions. The ICAO Assembly assigned priority to the development of
‘‘open emissions trading for international aviation’’ by the Council (ICAO 2001,
Appendix I). An ‘‘open’’ emissions-trading system could be connected to the
emissions-trading system under the Kyoto Protocol and would thus allow aviation
to trade emission permits with other sectors. To implement such a system, a cap on
emissions from aviation would need to be defined and the resulting amount of emis-
sion allowances allocated to the aviation industry. Given the inconclusiveness of
many years of discussions on the allocation of emissions from international trans-
port under the UNFCCC, resolving this issue will represent a major challenge for
ICAO. The ICAO schedule originally aimed at finalizing related proposals to the
UNFCCC by 2003 (Abeyratne 2001). In 2004, however, the ICAO Assembly
endorsed the further development of an open emissions-trading system for interna-
tional aviation and repeated its previous instruction to the ICAO Council ‘‘to de-
velop concrete proposals and provide advice as soon as possible to the Conference
of the Parties of the UNFCCC’’ (ICAO 2004, Appendix I).
While it had originally also put emphasis on emission standards and emissions
trading (MEPC 2002; UNFCCC 2002; ECON 2003, 12–13), IMO has shifted its
focus toward ‘‘GHG emission indexing.’’ GHG emission indexing refers to the
determination of a set of environmental criteria (emission standards, technological
and operational measures) that can be used to give an index to each vessel indicating
its GHG emission performance. It can provide a basis for differentiating taxes, port
dues, and charges or insurance rates, but had not received a particular blessing in
the aforementioned IMO study of 2000 (IMO 2000, 150–151). GHG emission
indexing grants particular flexibility to shipowners/operators, since they can choose
between different components of the index for achieving any required improvement.
At the end of 2003, the IMO Assembly adopted a resolution on ‘‘IMO Policies and
Practices Related to the Reduction of Greenhouse Gas Emissions from Ships’’ that
had been prepared by a Correspondence Group established by the MEPC (MEPC
2002; UNFCCC 2002). The resolution in particular mandates the MEPC to de-
velop GHG emission indexing further. The resolution also allows further work on
emissions trading by calling for the evaluation of ‘‘market-based solutions’’ (IMO
2003). In 2004, the MEPC further developed a CO2 -indexing scheme and asked
members to apply it in a trial period (http://www.imo.org). The work has, however,
not resulted in any binding measures yet.
Conclusions and Outlook

ICAO and IMO have started consideration of action on GHG emissions from avia-
tion and maritime transport, but have not gone beyond ‘‘symbolic’’ action yet. More
than seven years after the adoption of the Kyoto Protocol, neither of the organiza-
tions has agreed on any tangible measures, and it is doubtful whether this situation
will change in the foreseeable future. On the one hand, climate change does not be-
long to their core concerns, contributing further to an already heavy workload, and
mitigating emissions from international aviation and shipping may even be consid-
ered incompatible with the organizations’ main objective of furthering these sectors.
On the other hand, the threat of regulatory competition by the climate change re-
gime, the EU, and individual countries has remained rather weak. Furthermore, dis-
agreement over whether any measures would have global coverage or should only
apply to industrialized countries has delayed progress.
In the case of IMO, the slow progress is also due to two other factors. First, IMO
has perceived sea transport as part of the solution rather than as part of the prob-
lem. Shipping is seen as a comparatively environmentally friendly transport mode
and its contribution to climate change as ‘‘relatively small’’ (MEPC 2002; see also
IMO 2000, 169; UNFCCC 2002). Second, IMO has stressed that placing an addi-
tional burden on shipping requires similar measures to be taken with respect to
other modes of transport (i.e., aviation). Otherwise, shipping might become uncom-
petitive, which would lead to a modal shift to less environmentally friendly modes of
transport (IMO 2000; UNFCCC 2002).
The lack of coordination between ICAO, IMO, and the climate change regime
provides a further indication of the current stalemate. The demand for coordination
is apparent not only because there is a need to ensure that international aviation and
shipping contribute their fair share to the overall endeavor. A particular demand for
coordination exists with respect to an open emissions-trading system envisaged by
ICAO because it requires compatibility with the system of emissions trading under
the Kyoto Protocol. To date, however, members of the three institutions have
responded to this demand for coordination primarily by exchanging information
through mutual participation in meetings and reporting on relevant developments
and decisions by the respective secretariats. In reality, reports have triggered little
substantive debate and have resulted in very limited follow-up. As a result, members
of the climate change regime may identify insufficiencies and incompatibilities of any
measures only after ICAO and IMO have elaborated them (see in more detail Ober-
thür 2003, 200–202).
On the basis of the preceding analysis, we can identify in particular three options
for enhancing the willingness and ability of ICAO and IMO to take effective action
in the future:
1. Since the potential regulatory competition by the climate change regime has al-
ready been a significant driving force in the past, continuing work on measures to
limit and reduce GHG emissions from international transport within the climate
change regime could help keep up the pressure on ICAO and IMO. The entry into
force of the Kyoto Protocol in early 2005 may improve this prospect.
2. The implementation of domestic action by individual states could enhance the
willingness of aviation and shipping interests as well as state governments to accept
effective international regulation. Because the EU is the biggest contributor to
bunker-fuel emissions by contributing a good third of reported emissions of this
source from industrialized countries, it is less constrained by considerations of com-
petitive disadvantages than others and appears particularly suited to taking such
action. Other OECD countries in favor of effective action to address, in particular,
GHG emissions from aviation (e.g. Norway, Switzerland, New Zealand) could be
expected to get on the EU ‘‘bandwagon’’ by taking equivalent action.
3. Through their deliberation, ICAO and IMO may ‘‘learn’’ that effective action
on climate change is compatible with and may even be supportive of their general
objectives. Controlling GHG emissions may not appear to be immediately and di-
rectly supportive of the orderly development of international shipping and air
transport. However, GHG emission control may well increase its legitimacy and
acceptance, and can thus contribute to achieving the core objectives of ICAO and
IMO. Public-awareness campaigns about the environmental impacts of interna-
tional transport may further such a learning process.
Potential for creating synergy between the climate change regime and IMO and
ICAO thus exists. Should ICAO and/or IMO fail in their efforts, however, GHG
emissions from international transport may have to be addressed by the climate
change regime. Even if the targeted organizations took action, measures under the
UNFCCC and its Kyoto Protocol could complement such regulation.
Facilitating and Strengthening International Cooperation on Climate Change: The

EU Burden-Sharing Agreement
On the basis of an agreement on differentiated targets of the then fifteen EU member

states reached in March 1997, the EU constituted the major leader in the Kyoto
negotiations. Without this Burden-Sharing Agreement, a similar leadership coalition
could not have emerged. Consequently, the Agreement facilitated and strengthened
the commitments agreed to in Kyoto (Interaction through Commitment). Subse-
quently, the Kyoto Protocol prompted the codification of the burden sharing in EU
law, which strengthened the Protocol’s implementation by subjecting compliance
of EU member states with their quantitative emission commitments to the special
enforcement powers of the EU.
Strengthening and Facilitating Agreement in Kyoto

The EU member states reached a first Burden-Sharing Agreement about nine months
prior to the Kyoto conference in March 1997. It foresaw differentiated targets for
the individual member states ranging from þ40 percent for Portugal to 30 percent
for Luxembourg and amounted to an overall GHG emission reduction of 9.2 per-
cent. The Agreement remained conditional on an acceptable outcome of the interna-
tional negotiations. It was a consequence of the EU’s long-established objective
to act jointly in international climate policy, because competence in this area is
shared between the EU and its member states (Oberthür and Ott 1999, 141–142).
Table 3.2 provides the differentiated targets under the Burden-Sharing Agreement
of 1997 together with the figures as adapted to the outcome of the Kyoto negotia-
tions in 1998 and subsequently codified in EU law in 2002.
The Agreement of 1997 committed the EU member states to a common position
and thus established the EU as a powerful leading coalition favoring stringent emis-
Table 3.2
The EU Burden-Sharing Agreements of 1997 and 1998/2002
1997: emission 1998/2002:
reduction by emission reduction
Member state 2010 by 2008–2012
Luxembourg 30% 28%
Denmark 25% 21%
Germany 25% 21%
Austria 25% 13%
United Kingdom 10% 12.5%
Belgium 10% 7.5%
Netherlands 10% 6%
Italy 7% 6.5%
Finland 0% 0%
France 0% 0%
Sweden þ5% þ4%
Ireland þ15% þ13%
Spain þ17% þ15%
Greece þ30% þ25%
Portugal þ40% þ27%
EU-Total C9.2% C8%
Note: While targets of 1997 relate to CO2 , CH4 and N2 O, targets of 1998/2002 relate to all
GHGs regulated under the Kyoto Protocol.
sion reductions in the negotiations on the Kyoto Protocol (figure 3.2). It took the
form of Council conclusions that do not bind member states legally but entail a
strong political commitment. The differentiated targets of EU member states under
the Agreement of 1997 are indicative of the range of positions of individual member
states. However, several EU member states had to make concessions so that the
Agreement went significantly beyond the original aggregate of the positions of indi-
vidual EU member states (e.g., Ringius 1999). In the absence of the Burden-Sharing
Agreement, member states would thus have pursued widely diverging interests (table
3.2), with some of them probably favoring even less stringent targets. Overall, the
Agreement created an otherwise unlikely coalition of fifteen industrialized countries
in the Kyoto negotiations.
First of all, acting as a united coalition, the EU facilitated reaching agreement
in Kyoto by reducing the number of negotiating parties. The trilateral negotiations
between the United States (with an emission share of 36.1 percent), Japan (8.5
Figure 3.2
EU Burden-Sharing Agreement facilitates and strengthens Kyoto Protocol
percent), and the EU (24.2 percent) in Kyoto covered more than 68 percent of indus-
trialized countries’ CO2 emissions in 1990. Reaching a similar coverage without the
EU acting in unity would have meant increasing the number of negotiating parties
at least to eight (United States, Japan, Germany, United Kingdom, Canada, Italy,
Poland, France)2 or, alternatively, seventeen (if all EU member states had been
involved separately). Such an increase would have placed a considerable burden
on the negotiations and would have increased the likelihood of failure, because it is
‘‘almost axiomatic that the more parties (and issues), the higher the costs, the longer
the time, and the greater the informational requirements for a negotiated settle-
ment’’ (Sebenius 1983, 308–309). It would have added to an already very high com-
plexity and great time pressure (on the negotiations in Kyoto, see Oberthür and Ott
1999, chap. 7). In addition, it would have been difficult to establish which countries
were to participate in the core negotiations without the EU acting in a unitary fash-
ion, since there is no clear line between countries such as France (2.7 percent), the
United Kingdom (4.3 percent), Canada (3.3 percent), and Italy (3.1 percent) (emis-
sion shares according to the Appendix of the Kyoto Protocol).
The EU acting in unity was also instrumental in achieving concessions from Japan
and the United States and thus in strengthening the targets agreed on in Kyoto.
In the political bargaining process with the United States and Japan—which was
complemented by pledges of other industrialized countries—the EU, based on its
Burden-Sharing Agreement, supported the deepest GHG emission cut of 15 percent.

In the absence of a common EU position, stringent targets would have received far
less international support because only some member states would have supported
them. The United States (stabilization) and Japan (small reduction) would have
found it much easier to defend their positions because they would have been neatly
in the middle of other countries. Accordingly, Australia, which demanded a growth
target for itself, unsuccessfully tried to establish the differentiated targets of EU
member states as the reference point in the international negotiations by propos-
ing that industrialized countries’ targets should be within a range of 30 percent
to þ40 percent (Oberthür and Ott 1999, 144). But compared with the common
position of the EU, the other major players came under pressure. They eventually
accepted targets close to the EU’s (EU: 8 percent; United States: 7 percent; Japan:
6 percent) (on the negotiations see Oberthür and Ott 1999, chaps. 4–7). Accord-
ingly, there is broad agreement in the literature that without the EU, the commit-
ments by the United States and Japan would have been lower.
The Revision of the Burden-Sharing Agreement: Supporting Effective Climate

Protection
The first interaction between the EU Burden-Sharing Agreement and the Kyoto
Protocol resulted in two follow-up cases. First, the Kyoto Protocol triggered the re-
vision of the Burden-Sharing Agreement and its codification in supranational EU
law (Interaction through Commitment). Second, as a result of this codification, the
supranational enforcement mechanisms of the Union provide a particular incentive
to EU member states to comply with their Kyoto targets and thus support the effec-
tive implementation of the Protocol (Behavioral Interaction).
While there was no time in Kyoto to fix targets for each EU member state, Article
4 of the Kyoto Protocol allowed any group of countries to fulfill their commitments
under the Protocol jointly and, to this end, to redistribute their emission allowances
among them.3 After notification of the secretariat at the time of ratification, the re-
distribution cannot be further modified. EU member states had an obvious interest
in using Article 4, which had been included at the EU’s request. Given the internal
differences in starting points and positions, the then fifteen EU member states and
the EU itself could only become parties to the Kyoto Protocol if they redistributed
their common target of 8 percent under Article 4. It is worth highlighting that, by
means of Article 4, the Kyoto Protocol in fact delegated the task of fixing targets for
individual EU member states to the EU itself. The Kyoto Protocol thus indirectly
made use of the comparatively sophisticated framework of decision making of the

EU to reach binding agreement between member states, which is the daily bread of
the Union. As a result, the international negotiations were relieved of the burden of
establishing targets for fifteen states.
EU member states agreed on a revised Burden-Sharing Agreement in mid-1998
and codified it under supranational EU law in 2002. The Agreement of 1997 needed
adaptation in light of the outcome of the Kyoto negotiations. While the 1997 Agree-
ment had been related to three GHGs (CO2 , CH 4 , N2 O), the Kyoto targets also
included fluorinated GHGs. The Kyoto Protocol established a commitment period
of 2008–2012, whereas the 1997 Agreement referred to 2010 as a single target
year. Finally, the latter amounted to overall reductions of 9.2 percent, while the
common EU target under the Kyoto Protocol was 8 percent. Targets of individual
member states under the revised Burden-Sharing Agreement range from þ27 percent
for Portugal to 28 percent for Luxembourg (table 3.2). The 1998 Agreement be-
came legally binding in spring 2002 as part of the Council Decision to ratify the
Kyoto Protocol (European Union 2002). The UNFCCC Secretariat was notified on
ratification of the Protocol on May 31, 2002.
The codification of the Burden-Sharing Agreement in supranational law has cre-
ated an additional incentive for EU member states to comply with their commit-
ments under the Protocol and thus supports climate protection. It hardened the
international commitments of EU member states to limit and reduce GHG emis-
sions. By means of the Council Decision, the Agreement became part of the supra-
national law of the EU and is thus subject to the supranational adjudication and
enforcement mechanisms of the Union. In particular, the European Commission
will monitor EU member states’ compliance with their targets and may initiate in-
fringement proceedings, if required. As a result, noncompliant member states may
be brought before the European Court of Justice (ECJ), which issues binding rulings
and may even authorize financial penalties to be imposed by the Commission. While
a similar enforcement mechanism does not exist for any other party to the Kyoto
Protocol, it provides a powerful additional incentive for EU member states to
comply with their Kyoto targets and thus enhances climate protection.
Conclusions
Reflecting climate change’s manifold causes and consequences, the international

regime on climate change influences and is influenced by a great number of other
international institutions and EU legal instruments. A considerable potential exists

for further interactions that may materialize in the future. Institutional interactions
of the climate change regime cover all three major causal mechanisms. On some
occasions, they have resulted in synergy, while leading to tensions on others, in par-
ticular if involving institutions from other policy fields. In many cases, there is po-
tential for creating or enhancing synergy.
The horizontal interaction with ICAO and IMO followed the causal mechanism
of Interaction through Commitment and exemplifies the at times problematic rela-
tionship between the climate change regime and economic institutions. The Kyoto
Protocol’s request to ICAO and IMO committed the members of the climate change
regime (in particular industrialized countries) to addressing GHG emissions from
international transport. The request created pressure on both organizations and
empowered the proponents of action on GHG emissions among their members be-
cause it carried the implicit threat of regulatory action under the climate change re-
gime, if ICAO and IMO failed to take action. In response, both organizations have
begun to address the issue. However, coordination between them and with the cli-
mate change regime has barely occurred yet. Furthermore, the objectives of ICAO
and IMO to enhance international air and sea transport hardly led them to advance
their efforts of their own accord. In addition, the threat of regulatory competition by
the climate change regime and individual actors has remained weak. As a result, nei-
ther of the organizations has gone beyond ‘‘symbolic’’ action, and little progress in
addressing GHG emissions from international transport has been achieved to date.
Under the circumstances, the future success of the interaction is also in doubt. Prog-
ress may in particular be driven by (1) a strengthened threat of regulatory action
within the climate change regime, (2) domestic action by the EU and other countries,
and (3) a learning process within ICAO and IMO. In any event, the request of the
Kyoto Protocol begs the question of which of the institutions involved will regulate
GHG emissions from international aviation and shipping (and to what extent).
The vertical interaction with the EU Burden-Sharing Agreements of 1997 and
1998/2002 demonstrates the potential of the EU to promote the development and
implementation of international institutions. The Agreement of 1997 triggered a
case of Interaction through Commitment, facilitating and strengthening agreement
on the Kyoto Protocol. By committing EU member states to a stringent joint target,
the Agreement united the EU to form a leadership coalition pushing for strong
emission limitation and reduction targets in the Protocol negotiations. This coalition
even included EU member states that would not have requested strong commitments
on their own. As a result, the EU was able to secure more stringent commitments
from its negotiating partners in Kyoto than would otherwise have been the case.
Furthermore, getting to agreement in Kyoto was significantly facilitated primarily
because the number of core negotiating partners was reduced to three.
As a follow-up, the Kyoto Protocol in effect delegated to the EU itself the dis-
tribution of the EU target to individual member states. The international process
was thus relieved of this task and the advanced decision-making capacity of the EU
employed to this end. The resulting Burden-Sharing Agreement of 1998/2002 in
turn supported the implementation of the Protocol by subjecting EU member states
to the particular enforcement powers of the European Union. In case of noncompli-
ance, EU member states may face financial penalties authorized and enforced by the
European Court of Justice. In this way, the Burden-Sharing Agreement significantly
hardens the quantitative emission commitments for EU member states. This find-
ing contrasts starkly with persistent allegations by other OECD countries and the
United States in particular that the allowance to fulfill their commitments jointly
represents an unjustified preferential treatment of EU member states (Oberthür and
Ott 1999, chap. 12).
Under certain circumstances, the EU can thus help advance international negotia-
tions decisively and can strengthen the implementation of international commit-
ments. EU member states could credibly commit to the targets of the Burden-Sharing
Agreement because the supranational structure of the EU facilitates concluding
binding agreements (which is the daily bread of EU policymaking). The EU provides
a forum for twenty-five countries at present to coordinate their position and to share
and implement their international commitments by employing the supranational
powers of the EU. It may also be possible to take advantage of the particularly
high ‘‘problem-solving capacity’’ of the EU more frequently in other contexts in
which it can be left to the EU to share/implement a joint international commitment.
Notes
1. Legally, the European Community (EC), not the EU, is a party to both the UNFCCC and
the Kyoto Protocol, in addition to the member states. I nevertheless refer to the EU through-
out this chapter for ease of reference.
2. Based on the assumption that Russia would also not have participated under these circum-
stances; for the percentage figures see the Appendix of the Kyoto Protocol.
3. While the Agreement facilitated the negotiations on targets, the issue of how to design the
resulting Article 4 of the Kyoto Protocol placed an additional burden on negotiators; see
Oberthür and Ott 1999, chap. 12.
References
Abeyratne, Ruwantissa. 2001. ICAO: Some Recent Developments in Aviation and Environ-
mental Protection Regulation. Environmental Policy and Law 32 (1): 32–40.
Bail, Christoph, Simon Marr, and Sebastian Oberthür. 2003. Klimaschutz und Recht. In
Hans-Werner Rengeling, ed., Handbuch zum europäischen und deutschen Umweltrecht, vol.
2, 254–304. Cologne: Carl Heymanns.
Bodansky, Daniel. 1993. The United Nations Framework Convention on Climate Change: A
Commentary. Yale Journal of International Law 18 (2): 451–558.
Bode, Sven, Jürgen Isensee, Karsten Krause, and Axel Michaelowa. 2002. Climate Policy:
Analysis of Ecological, Technical and Economic Implications for International Maritime
Transport. International Journal of Maritime Economics 4 (2): 164–184.
Buergenthal, Thomas. 1969. Law-Making in the International Civil Aviation Organization.
New York: Syracuse University Press.
Chambers, W. Bradnee. 1998. International Trade Law and the Kyoto Protocol. In W.
Bradnee Chambers, ed., Global Climate Governance: Inter-Linkages between the Kyoto Pro-
tocol and Other Multilateral Regimes, 39–58. Tokyo: United Nations University, Institute of
Advanced Studies.
Chambers, W. Bradnee, ed. 2001. Inter-Linkages: The Kyoto Protocol and the International
Trade and Investment Regimes. Tokyo: United Nations University Press.
Charnovitz, Steve. 2003. Trade and Climate: Potential Conflict and Synergies. Washington,
DC: Pew Center on Global Climate Change.
Council of the European Union. 2001. Climate Change: Preparation of COP 7 in Marrakesh
from 29 October to 9 November 2001. Council Conclusions. Available at http://register
.consilium.eu.int/pdf/en/01/st13/13439en1.pdf.
Crayston, John. 1993, January–June. Civil Aviation and the Environment. UNEP Industry
and Environment: 51–53.
Crayston, John, and Jane Hupe. 2000, October–December. Civil Aviation and the Environ-
ment. UNEP Industry and Environment: 30–32.
Diederiks-Verschoor, Isabella H. P. 1993. An Introduction to Air Law. Deventer: Kluwer
Law and Taxation.
ECON. 2003, January. GHG Emissions from International Shipping and Aviation. Report
01/03. Oslo: ECON Centre for Economic Analysis.
European Commission. 1999, December 1. Air Transport and the Environment: Towards
Meeting the Challenges of Sustainable Development. Communication from the Commission
to the Council, the European Parliament, the Economic and Social Committee, and the Com-
mittee of the Regions. Doc. COM (1999) 640 final. Brussels: European Commission.
European Union. 2002. Council Decision 2002/358/EC of 25 April 2002 Concerning the Ap-
proval, on Behalf of the European Community, of the Kyoto Protocol to the United Nations
Framework Convention on Climate Change and the Joint Fulfilment of Commitments there-
under. Official Journal of the European Communities L 130/1, 15 May 2002.
Fayette, Louise de La. 2001. The Marine Environment Protection Committee: The Conjunc-
tion of the Law of the Sea and International Environmental Law. International Journal of
Marine and Coastal Law 16 (2): 155–238.
ICAO. 1996, December 9. Council Resolution on Environmental Charges and Taxes. Resolu-
tion adopted by the Council at the 16th Meeting of its 149th session.
ICAO. 1998. Resolution A32-8: Consolidated Statement of Continuing ICAO Policies and
Practices Related to Environmental Protection. Resolutions adopted at the 32nd session of
the Assembly, Provisional Edition.
ICAO. 2001. Resolution A33-7: Consolidated Statement of Continuing Policies and Practices
Related to Environmental Protection. Resolutions adopted at the 33rd session of the Assem-
bly, Provisional Edition.
ICAO. 2004. Resolution A35-5: Consolidated Statement of Continuing ICAO Policies and
Practices Related to Environmental Protection. Resolutions adopted at the 35th session of
the Assembly, Provisional Edition.
IMO. 2000, March 31. Study of Greenhouse Gas Emissions from Ships. Issue no. 1. London:
International Maritime Organization.
IMO. 2003. Resolution A.963(23) adopted on December 5, 2003: IMO Policies and Practices
Related to the Reduction of Greenhouse Gas Emissions from Ships. IMO Doc. A 23/Res.963,
March 4, 2004. London: International Maritime Organization.
IPCC. 1999. Aviation and the Global Atmosphere: A Special Report of IPCC Working
Groups I and III. Cambridge: Cambridge University Press.
IPPC. 2001a. Climate Change 2001: Impacts, Adaptation, and Vulnerability. A Report of
Working Group II of the Intergovernmental Panel on Climate Change. Geneva: Intergovern-
mental Panel on Climate Change.
IPCC. 2001b. Climate Change 2001: The Scientific Basis. Cambridge: Intergovernmental
Panel on Climate Change.
Jacquemont, Frédéric, and Alejandro Caparrós. 2002. The Convention on Biological Diver-
sity and the Climate Change Convention 10 Years After Rio: Towards a Synergy of the Two
Regimes? Review of European Community and International Environmental Law 11 (2):
139–180.
MEPC. 2002. MEPC, 47th session, March 4–8, 2002. Marine Environmental Protection
Committee. Available at http://www.imo.org/Newsroom/mainframe.asp?topic_id=109&doc_
id=175.
Oberthür, Sebastian. 2003. Institutional Interaction to Address Greenhouse Gas Emissions
from International Transport: ICAO, IMO and the Kyoto Protocol. Climate Policy 3 (3):
191–205.
Oberthür, Sebastian, and Hermann E. Ott (in collaboration with Richard G. Tarasofsky).
1999. The Kyoto Protocol: International Climate Policy for the 21st Century. Berlin:
Springer.
Pisani, Christian. 2002. Fair at Sea: The Design of a Future Legal Instrument on Marine
Bunker Fuels Emissions within the Climate Change Regime. Ocean Development and Inter-
national Law 33 (1): 57–76.
Ringius, Lasse. 1999. Differentiation, Leaders, and Fairness: Negotiating Climate Commit-
ments in the European Community. International Negotiation 4 (2): 133–166.
Sebenius, James K. 1983. Adding and Subtracting Issues and Parties. International Organiza-
tion 37 (1): 281–316.
UNFCCC. 2002. Subsidiary Body for Scientific and Technological Advice. Emissions Result-
ing from Fuel Used for International Transportation: Activities of IMO on Prevention of Air
Pollution from Ships; Information by the IMO Secretariat. Sixteenth session, Bonn, June 3–
14 (on file with author).
WBGU. 2002. Entgelte für die Nutzung globaler Gemeinschaftsgüter. Politikpapier 2. Berlin:
Wissenschaftlicher Beirat der Bundesregierung Globale Umweltveränderungen.
Yamin, Farhana, and Joanna Depledge. 2004. The International Climate Change Regime: A
Guide to Rules, Institutions and Procedures. Cambridge: Cambridge University Press.
4
The Convention on Biological Diversity: Tensions
with the WTO TRIPS Agreement over Access to
Genetic Resources and the Sharing of Benefits
G. Kristin Rosendal
Biological diversity is a broad concept that has been used to embody the variability
among all living organisms, including diversity within species (genetic diversity),
among species, and among ecosystems. The issue of biological diversity constitutes
one of today’s greatest challenges, for several reasons. There is an increased aware-
ness and scientific agreement that the current rate of species extinction is extremely
high compared to the natural average rate (Wilson 1988; Heywood 1995, 232). As
the new biotechnologies greatly enhance the potential utility areas of the world’s
genetic resources, the economic interests linked to these resources are soaring.1 One
combined effect of these two trends has been a greatly enhanced interest in intellec-
tual property rights and in access to genetic resources and associated technologies.
Among the international instruments that deal with the various facets of biodiver-
sity management, the central treaty is the Convention on Biodiversity (CBD) that
was signed in Rio de Janeiro in 1992 and provides an overall legal framework for
the issue area. The CBD is not the first international treaty to address species or hab-
itat conservation, but it is the first to comprehensively address conservation, sustain-
able use, and equity issues related to biological diversity worldwide. Because of the
CBD’s broad scope, it is hardly surprising that it interacts with a great many other
international institutions and EU legal instruments relating to other aspects of na-
ture conservation but also to economic issues (in particular patenting and intellec-
tual property rights).
This chapter explores the institutional interactions of the CBD in general and
analyses in detail the way the CBD and the Agreement on Trade-Related Aspects
of Intellectual Property Rights (TRIPS Agreement) under the World Trade Organi-
zation (WTO) have influenced each other’s implementation and performance. To
this end, first the general features of the CBD are briefly introduced. Then, an over-
view of pertinent interactions between the CBD and other international institutions
80 G. Kristin Rosendal
as well as one EU legal instrument is provided. The main analysis of the chapter
concerns diverging incentives for behavioral activities stemming from the CBD and
the TRIPS Agreement. It examines the claim that both regimes may undermine each
other’s implementation by inducing behavioral changes that run counter to the
objectives of the other institution. This case of interaction is not only of interest be-
cause of the salience of the underlying political conflict. It also provides an interest-
ing example in which influence runs both ways and the two institutions involved
concurrently undermine each other in an ‘‘arms race’’ fueled by different actors pur-
suing their interests based on the diverging norms emanating from the two regimes.
The Convention on Biological Diversity
The CBD entered the international negotiation arena in 1989, was signed at
the 1992 UN Conference on Environment and Development (UNCED) in Rio de
Janeiro, and entered into force in December 1993. The CBD is governed by a Con-
ference of the Parties (COP) that is supported by a Subsidiary Body on Scientific,
Technical, and Technological Advice (SBSTTA). The CBD secretariat is located in
Montreal and has 36 professionals and 26 support staff (2003-4). As of mid-2005,
188 states had ratified the CBD.
The Convention on Biological Diversity is built on a threefold, interacting objec-
tive: ‘‘the conservation of biological diversity, the sustainable use of its components
and the fair and equitable sharing of the benefits arising out of the utilization of ge-
netic resources’’ (Art. 1). To this end, the CBD introduces a number of important
commitments and instruments. Parties shall develop national biodiversity strategies,
integrate biodiversity conservation in all policy levels and sectors, identify and mon-
itor biodiversity, establish systems of protected areas, and identify activities that are
likely to have adverse effects on biodiversity. Moreover, the parties shall adopt eco-
nomically and socially sound measures to act as incentives for conservation and sus-
tainable use; establish programs for scientific and technical education and training
for identification and conservation; and provide support for such training in devel-
oping countries. The CBD is equipped with a monitoring mechanism in the form of
national reporting and an incentive mechanism in the form of the Global Environ-
ment Facility (GEF).
In several respects, the CBD constitutes a framework agreement that needs to be
further developed and specified. For example, Article 19.3 commits parties to the
CBD to consider the elaboration of a protocol addressing the safe transfer, handling,
Convention on Biological Diversity 81
and use of genetically modified organisms. On this basis, the Cartagena Protocol on
Biosafety was adopted in 2000 and entered into force in 2003. While this chapter
does not address the Cartagena Protocol, the interaction between the Protocol and
the WTO is analyzed in chapter 8.
The provisions on access to genetic resources, including the equitable sharing of
the benefits of their utilization, form a central element of the CBD. The CBD defines
genetic resources as genetic material of actual or potential value (Art. 2). They may
be categorized as plant, animal, and microbiological genetic resources and are of
fundamental importance for agriculture and a number of industry sectors, including
the pharmaceutical sector. The bulk of the world’s terrestrial species are found in
tropical forests in the South (UNEP 1995, 749). In contrast, it is primarily the devel-
oped countries of the North that possess the technological and economic capacity to
reap—assisted by intellectual property rights—the ever larger benefits from the
genetic variability employed in the agribusiness and pharmaceutical industries
(Kate and Laird 1999). Because these industrial actors were pushing hard for a
strengthening of intellectual property rights systems within the TRIPS Agreement of
the WTO (Rosendal 2000), the issue of access to and sharing of benefits from the
utilization of genetic resources became central in the CBD negotiations and was
linked to the responsibility for costly biodiversity conservation.
The essential agreement within the CBD is that equitable sharing of the benefits
from the use of genetic resources is a precondition for their conservation and sus-
tainable use. Largely as a response to the general developments in patent legislation,
of which the TRIPS negotiations constituted a significant part, the CBD reconfirmed
national sovereign rights to genetic resources (Art. 15.1) and equitable sharing of
benefits from use of those resources (Art. 15.7). Access to the resources shall be
based on mutually agreed terms and be subject to prior informed consent (Art.
15.4 and 15.5). This is to ensure that the providers of genetic resources get their
fair share of the benefits derived from their use. Article 15.2 declares that the parties
shall facilitate access to the same resources. In the last phases of negotiations, and
as a direct response to the TRIPS Agreement that was being negotiated in the Uru-
guay Round of international trade talks, the CBD was equipped with Article 16.5,
which establishes that intellectual property rights (IPR) systems should ‘‘not run
counter to the objectives in the CBD.’’ Here, the diverging objectives constituting
the two regimes are explicitly referred to. At the same time, Article 16.2 states
that the technology-transfer process is to be consistent with ‘‘the adequate and effec-
tive protection of intellectual property rights.’’ Article 16.2 implies that the CBD
sanctions IPR, but only on certain conditions that are further specified in Articles
15.4, 15.5, and 16.5. In sum, the CBD aims directly at the skewed distribution of
biological resources and biotechnology between the North and the South by provid-
ing a legal basis for developing countries to demand a share of the benefits from the
utilization of genetic resources.2
Institutional Interactions in Conservation and Management of Biodiversity
The CBD is a broad convention, encompassing and transcending a large number

of sectors and scales. Conservation and sustainable use pertain to a large array of
interrelated sectors, because the main threats to biodiversity—habitat destruction
and fragmentation, overexploitation, and displacement by introduced species—are
associated with extensive human activities in industry, agriculture, and forestry. As
a consequence, the CBD touches on a great number of international and regional
agreements. Investigating the interactions with all of them is beyond the scope of
this chapter. For example, the CBD is likely to interact with international institu-
tions aimed at fisheries management, which are not investigated here. In the follow-
ing, I provide an overview of the institutional interactions with three clusters of
instruments pertaining to biodiversity conservation, forest management, and access
and benefit sharing relating to genetic resources. These cases of interaction are sum-
marized in table 4.1. Interactions with the EU Habitats Directive are addressed in
chapter 10.
First, the CBD interacts with a number of international institutions that share the
objective of conserving biological diversity. In particular, the Convention on Inter-
national Trade in Endangered Species of Wild Fauna and Flora (CITES; see chapter
7), the Ramsar Convention on Wetlands, and the Convention on Migratory Species
of Wild Animals (CMS) belong to this ‘‘biodiversity conservation cluster.’’ The CBD
differs from most of the other institutions within this cluster in its strong, additional
foci on sustainable use and equitable sharing (Rosendal 2001b). However, the im-
plementation of the more specific agreements just mentioned generally contributes
to the objective of conserving biological diversity enshrined in the broader CBD.
To the extent that they lead to an effective implementation, CITES, the Ramsar
Convention, and the CMS therefore quasi-automatically support the CBD at the
outcome level (Behavioral Interaction). The institutions in this cluster attempt to fur-
ther enhance synergy by developing joint working plans and memoranda of cooper-
ation as well as cooperating on streamlining their national reporting, scientific data
Table 4.1
Interactions of the Convention on Biological Diversity
Biodiversity conservation cluster
Convention on International Implementation of CITES helps CBD achieve its
Trade in Endangered Species of targets.
Wild Fauna and Flora (CITES)
Ramsar Convention on Implementation of Ramsar Convention helps CBD
Wetlands achieve its targets.
Convention on Migratory Implementation of CMS helps CBD achieve its targets.
Species (CMS)
Forest-management cluster
UN Forum on Forests (and CBD has facilitated international forest talks that could
predecessors) build on agreements reached under it.
Kyoto Protocol Protocol provides incentives for fast-growing
monocultural forest plantations endangering biological
diversity protected under the CBD.
Access and benefit-sharing cluster
WTO TRIPS Agreement TRIPS Agreement affected negotiations of the CBD on
equitable sharing/access to genetic resources.
TRIPS rules on patenting engender behavior at odds
with CBD objectives on equitable sharing/access.

CBD objectives on equitable sharing/access to genetic
resources lead to cumbersome patent protection at odds

with TRIPS Agreement.
FAO International Undertaking FAO Undertaking influenced negotiations on CBD
on Plant Genetic Resources for regarding property rights and access to genetic resources.
Food and Agriculture (1983)
FAO Treaty on Plant Genetic CBD influenced FAO Treaty regarding access to
Resources for Food and genetic resources.
Agriculture (2001)
EU Patent Directive CBD objectives on equitable sharing/access to genetic
resources influenced preparation of EU Patent Directive.
Patent Directive may undermine the implementation of
the CBD objectives on equitable sharing/access.

collection, and information exchange (Rosendal and Andresen 2003; Stokke and
Thommessen 2003).
A second cluster involves the international institutions that have a specific bearing
on forest management, including in particular the UN Forum on Forests (and its
predecessors) and the global climate change regime. Forest management has been
part of the CBD agenda and has been discussed separately by governments, first in
the context of the preparations for UNCED, then in forums under the Commission
on Sustainable Development and the UN Economic and Social Council, and finally
in the UN Forum on Forests. Because of the overlap of agendas, the separate forest
discussions were able to benefit from agreements reached under the CBD on relevant
issues, which could easily be transferred and used as a basis of the forest discussions
(Interaction through Commitment). At the same time, a certain competition exists
between both processes because both the UN Forum on Forests and the CBD claim
authority over regulating forest management. This competition would become par-
ticularly prevalent, if the UN Forum were to result in the elaboration of a forest con-
vention, as some actors demand. Furthermore, the regulations emanating from the
Kyoto Protocol to the United Nations Framework Convention on Climate Change
(UNFCCC) may have a significant negative impact on forest biodiversity because
they provide incentives for establishing fast-growing monocultural forest plantations
(Behavioral Interaction). From the point of view of biological diversity, however,
such massive uniformity is synonymous with genetic erosion (Gillespie 1999; Rosen-
dal 2001c; see chapter 3).
Third, several regimes are engaged in a delicate balance between the concerns for
protecting inventions and encouraging innovations in agriculture and pharmaceuti-
cals on the one hand, and environmental and distributional concerns on the other
(Rosendal 2001a; Raustiala and Victor 2004). Both the International Undertaking
on Plant Genetic Resources for Food and Agriculture of 1983 of the Food and Agri-
culture Organization of the United Nations (FAO) and the FAO Treaty on Plant
Genetic Resources for Food and Agriculture of 2001 belong to this access and equi-
table sharing cluster. The FAO Undertaking that predates the CBD influenced its
provisions regarding property rights and access to genetic resources. Subsequently,
the CBD influenced the relevant provisions of the FAO Treaty (Interaction through
Commitment). The objective of the WTO TRIPS Agreement is to strengthen and ex-
pand IPR systems worldwide. The negotiation processes of the CBD and the TRIPS
Agreement coincided in time—between 1988 and 1993—so that what went on in
one forum affected what went on in the other. In particular, the parallel TRIPS
negotiations enhanced the interest especially of developing countries in protecting

their claims regarding equitable sharing and access to genetic resources within the
CBD negotiations (Interaction through Commitment). As a result, both agreements
have been working at cross-purposes at the implementation level (Behavioral Inter-
action), as further detailed in the next section. It is worth mentioning that other
international institutions such as the Union for the Protection of New Varieties of
Plants (UPOV) and the World Intellectual Property Organization (WIPO) are also
active in the field. At the regional level, the EU Directive on Patents in Biotechnology
was partly influenced by the EU obligations under the CBD (Interaction through
Commitment), but nevertheless draws behavior in the same direction as the TRIPS
Agreement (Behavioral Interaction).
Intellectual Property Rights and Genetic Resources
Background
Some of the major ingredients in this interaction take us back to 1980. Originally,
international transactions of plant genetic resources were based on the principle of
common heritage of mankind. The more widely used definition of this principle
implies that the resources should be freely available and accessible to all, regardless
of economic and technological strength, and should hence be outside the reach of
intellectual property rights (Bilder 1980). This principle was gradually undermined
as patent legislation was reinterpreted to cover biological material (Mooney 1983).
This was, among other things, an effect of the rapid development of new biotechnol-
ogies, which made it possible to overcome what had previously been legal and bio-
logical barriers to patenting in biotechnology (Bent et al. 1987; Crespi 1988). In the
FAO Undertaking of 1983, developing countries pushed through an agreement that
all categories of plant genetic resources should be regarded as a common heritage of
mankind. They thus aimed to keep all types of breeding material within the pub-
lic domain and outside the scope of patents. This was in line with the basic principle
of the international gene banks—that seeds should be freely available as a source of
plant breeding and food security. International gene banks were stocked with seeds
from the most commonly used food plants and these seeds were primarily collected
from the extensive variation found in the South. In an official interpretation agreed
on at a FAO conference in 1989, however, developed countries succeeded in estab-
lishing intellectual property rights as compatible with the FAO Undertaking. This
reopened the Pandora’s Box of property rights to genetic resources.
At that time, the United States among other countries was becoming exasperated
by the FAO negotiations on plant genetic resources. It wanted a fresh start by nego-
tiating biodiversity conservation under the auspices of the United Nations Environ-
ment Program (UNEP). However, developing countries soon succeeded in adding
sustainable use and equitable sharing to that of conservation on the CBD agenda
(Pistorius and van Wijk 1999; Rosendal 1991; Svensson 1993; Koester 1997). The
South abandoned the common-heritage strategy and successfully demanded recon-
firmation of national sovereign rights over genetic resources. A sovereign right is
not the same as a property right, because it implies that the state has the prerogative
to regulate the area—for example, by establishing property-rights regimes for the
resources in question. Simultaneously, in the Uruguay Round of trade negotiations
leading up to the WTO and the TRIPS Agreement, the United States, Japan, and less
adamantly, the EU successfully demanded that all countries should provide and re-
spect intellectual property protection in all technical fields, including biotechnology.
National sovereign rights were hence used as a compromise for accepting, on certain
conditions, the expanding use of patents in biotechnology (Rosendal 2000).
If the contentious issue of property rights had been restricted to the agricultural
sector, the controversial issue of access and benefit sharing relating to genetic
resources might have been left to the non–legally binding FAO instruments—a
Commission for plant genetic resources, a never functioning Fund, and the Under-
taking. The new biotechnologies, however, also instigated rapid developments
within the pharmaceutical sector. In this sector it is primarily the wild genetic
resources—with tropical areas of the South as a primary source—that provide raw
material for the products of biotechnology. In effect, the benefit-sharing issue came
to stay with the CBD (which the United States—concerned with its large biotechnol-
ogy sector—has persistently refused to ratify).
The developing countries had high expectations with regard to the CBD. After the
UNCED Rio Conference in 1992 a central G77 spokesperson proclaimed: ‘‘Climate
change was theirs (the developed countries)—biodiversity was ours!’’3 Ten years
later, twelve countries (Brazil, China, Colombia, Costa Rica, Ecuador, India, Indo-
nesia, Kenya, Mexico, Peru, South Africa, and Venezuela) formed the Group of
Allied Mega-Biodiverse Nations to press for rules protecting their rights to genetic
resources found on their land. Since equitable sharing is still far off, the Group
now aims for more equal trade rules on patenting (Stevenson 2002). What caused
the great expectations of the South to plummet? Is part of the answer that the TRIPS
Agreement is obstructing the implementation of the CBD’s objectives? The follow-
ing examination of the relationship between the CBD and the TRIPS Agreement is
meant to provide some answers. To this end, the TRIPS Agreement is first intro-
duced in more detail in the following, before the largely disruptive relationship be-
tween the CBD and the TRIPS Agreement at the implementation level is examined.
The WTO Agreement on Trade-Related Aspects of Intellectual Property Rights

The TRIPS Agreement forms part of the WTO agreements concluded in 1994 as the
result of the Uruguay Round of trade negotiations that had commenced in 1986. As
of April 2004, the WTO had 147 member states. The TRIPS Agreement represents
the third pillar of the WTO, along with the agreements on trade in goods and ser-
vices (General Agreement on Tariffs and Trade, GATT, and General Agreement on
Trade in Services, GATS). As such, it must be adhered to by all states wishing to join
the WTO. The TRIPS Agreement embodies the basic WTO principles of nondiscri-
mination committing WTO parties to ‘‘national treatment’’ (i.e., treating one’s own
nationals and foreigners equally) and ‘‘most-favored-nation treatment’’ (i.e., equal
treatment of all trading partners in the WTO) (Chaytor and Cameron 2000; see
also chapter 8). The TRIPS Agreement covers questions about giving adequate pro-
tection to intellectual property rights, how countries should enforce those rights,
and how to settle disputes on intellectual property between members of the WTO.
Most importantly, the TRIPS Agreement is an attempt to standardize the way intel-
lectual property rights are protected around the world and to strengthen this harmo-
nization process in all technological fields—including biotechnology. Intellectual
property includes copyright, trademarks, geographic indications, industrial designs,
patents, layout designs (topographies) of integrated circuits, and undisclosed infor-
mation, including trade secrets (http://www.wto.org).
As has been noted, negotiations of the TRIPS Agreement coincided in time with
the negotiations of the CBD. Since both agreements aimed to regulate partially over-
lapping issues, it would have been surprising if they had not already influenced each
other during their elaboration. For example, the discussions on intellectual property
rights in the Uruguay Round had significant impact on the issues of access and ben-
efit sharing pertaining to genetic resources as debated in the negotiations on the
CBD. In turn, the principles elaborated within the CBD negotiations also affected
the TRIPS agenda.
With respect to intellectual property rights to varieties of animals and plants, the
final compromise embodied in Article 27 of the TRIPS Agreement closely reflects the
relevant provisions of Article 53(b) of the European Patent Convention of 1973,
which allowed for plants and animals to be excluded from patentability. The Euro-
pean Patent Convention applies primarily in the European Communities and is
administered by a European Patent Office (EPO). Article 27 of the TRIPS Agreement
on the one hand grants parties the right to exclude from patentability diagnostic,
therapeutic, and surgical methods for the treatment of humans and animals (Art.
27.3(a)), and plants and animals other than microorganisms (Art. 27.3(b)). On the
other hand, Article 27.3(b) obligates parties to also provide for protection of intel-
lectual property rights for plant varieties, either by patents or by an effective sui gen-
eris system (a legal system of its own kind).
Article 27.3(b) of the TRIPS Agreement therefore is of particular relevance for the
interaction with the CBD. In this respect, the central question is whether the TRIPS
Agreement leaves it up to the parties to design their own sui generis systems, or
whether they should preferably choose the breeder’s rights system provided by the
International Union for the Protection of New Varieties of Plants (UPOV). UPOV
was established by the Convention for the Protection of New Varieties of Plants of
1961, which has been revised three times since then. Its purpose is to provide uni-
form and clearly defined principles for the protection of plant breeders’ rights and
bring them more in line with patent protection. Before the last revision in 1991, the
Convention granted protection only to the final product, the variety, and not to sub-
sequent varieties bred on the basis of the protected one. Farmers were thus free to
use such seeds for next year’s sowing (‘‘farmers’ privilege’’) and breeders and scien-
tists could use UPOV-protected material for developing new products (‘‘breeders’
exemption’’), without paying royalties. As a result of the revision of 1991, however,
protection under the UPOV Convention has come to resemble patent protection on
a number of dimensions, most importantly by restricting the former provisions for
farmers and breeders (Walden 1995).
Implementation of the TRIPS Agreement Undermining the CBD

The TRIPS Agreement lays down the ground rules for what must be protected
by national patent legislation, including plant varieties and pharmaceuticals. The
changes in property-rights regimes pertaining to genetic resources originated from
developments in biotechnology. They led to the successful demand by developed
countries for all WTO members to provide and respect IPR protection, including in
the field of biotechnology. While the TRIPS Agreement does not create a single, uni-
versal patent system, it seeks to strengthen, expand the scope for, and harmonize the
domestic patent legislation in each of the member countries. Patent legislation is of a
national character, and patent protection is applicable only in the country where it
has been granted. National patent legislation is largely drawn from international
conventions, administered by the World Intellectual Property Organization (WIPO).
As multinational corporations seek patent protection around the world, they depend
on each country’s patent office to grant those rights. Any effect of the TRIPS Agree-
ment on the CBD at the outcome level would thus have to result from strengthened
and expanded national patent systems.
The potential for disruptive behavioral effects in the implementation of the two
regimes is primarily a result of the fact that the TRIPS Agreement and the CBD pur-
sue divergent policy objectives. While equitable sharing and conservation constitute
the core norms and principles of the CBD, the TRIPS Agreement is geared to pro-
mote a time-limited exclusive right to genetic resources. The existing system of intel-
lectual property rights and patents does not accommodate nonwestern systems of
knowledge and ownership, such as community or farmers’ rights. Hence, they can
hardly contribute to enhancing equitable sharing and, in effect, they provide little in-
centive for biodiversity conservation in poor but gene-rich countries (figure 4.1).
In principle, patent protection of genetic resources, as demanded by the TRIPS
Agreement, results in restricted access to these genetic resources—while such access
is, subject to certain conditions, to be facilitated under the CBD. Access to a rich
variety of genetic resources is essential for plant breeding and food security in all
parts of the world (Kloppenburg 1988; Kloppenburg and Kleinman 1987; Berg
Figure 4.1
Implementation of WTO TRIPS Agreement undermines CBD
1995; Harlan 1995). Applying patents to seeds from the international genebanks is
thus seen as representing a threat to the basic principle of free exchange of germ-
plasm, on which these genebanks are building (Keystone 1990; Barton and Siebeck
1992). Barton and Berger (2001, 5) conclude that ‘‘it may be very expensive or dif-
ficult for the public sector to gain access to patented technologies or to use protected
varieties for research in developing new applications for the smaller crops or subsis-
tence farmers.’’ On the same note, the FAO Commission on Plant Genetic Resources
has warned that ‘‘if the patent system is applied universally to living matter, includ-
ing plants and animals, and their genetic resources, then the principle of unrestricted
access will be severely eroded’’ (FAO 1993, 23). Developing countries maintain that
without the IPRs the question of benefit sharing would not arise and all genetic mate-
rial could be freely available in the same way as it was historically (Borring 2001).
Patenting may also be incompatible with the CBD objective of equitable sharing
of benefits because it grants exclusive rights to only some of those who have contrib-
uted to generating the benefit (Hendrickx, Koester, and Prip 1993; Crucible Group
1994; Correa 1999; Egziabher 1999). Patenting is a long and costly business that is
primarily employed by large corporations. Transnational corporations hold 90 per-
cent of all technology and product patents (Gleckman 1995). The developing world
holds no more than 1–3 percent of all patents worldwide (WCED 1987; UNDP
2000, chap. 2). Patenting is hardly a tool for indigenous and local communities,
even though these groups often harbor much knowledge about the use of biological
resources. Biotechnological products in agriculture often build on local breeders’
lines, which represent the end result of the work of generations of farmers. Likewise,
pharmaceutical products of biotechnology often build on traditional knowledge
about the medicinal traits of biological resources, such as wild plants.4 There are,
as yet, no applicable models for access and benefit sharing or alternative models to
IPR laws for protection of traditional knowledge. The knowledge in its traditional
form and the nonsystematically bred cultivars hardly fulfill general patent criteria,
such as reproducibility, or the UPOV criteria of being ‘‘new, distinct, uniform and
stable’’ (Art. 5.1 of the UPOV Convention).
Moreover, patenting may work indirectly to reduce genetic diversity, because
patenting is largely an asset of the developed world and a handful of multinational
corporations. Protected by patents, these actors dominate the seed industry and
pursue their interest in promoting their products. In contrast, developing countries,
where much of that diversity is found in situ, have few incentives to conserve their
genetic heritage (NORD 1992; Swanson and Johnston 1999; Fauchald 2001).
A central example for the disruptive influence of the TRIPS Agreement is the
EU Directive on Patents in Biotechnology that was adopted in 1998 after more
than ten years of negotiations in which the CBD and TRIPS-related principles
played a prominent role. The negotiation process saw the intense combat between
norms associated with the CBD and the TRIPS Agreement respectively. Several EU
member states, environmental and farmers’ groups, as well as the European Parlia-
ment strove to bring the Patent Directive more into line with obligations under the
CBD. In the end, however, the Patent Directive gave priority to patent protection for
plant varieties and pharmaceuticals without ensuring equitable sharing of benefits
from the utilization of genetic resources for patented products.5 As of September
2004, eleven member states (including France, Austria, Belgium, Germany, and the
Netherlands) are lagging behind in their implementation of the Patent Directive.6
Belgium and Denmark have made specific amendments in their domestic patent leg-
islation in order to bring it more into line with the CBD objectives.
There are also a number of related processes that support the same normative
elements as the TRIPS Agreement. The so-called TRIPS plus agreements (Dutfield
2001) are bilateral trade agreements primarily between the United States or EU and
a developing country or region, which include requirements for higher patent stan-
dards than what the TRIPS Agreement demands (Morin 2003). By 2004, about
twenty of these agreements had been concluded. For instance, the trade agreements
between the European Union and South Africa and Mexico assert that the parties
‘‘shall ensure adequate and effective protection of IPR in conformity with the highest
international standards.’’7 These agreements ensure implementation in advance of
the timetable set up in the TRIPS Agreement and they often include the condition
to implement the UPOV Convention of 1991. Although building onto the TRIPS
Agreement, these ‘‘TRIPS plus’’ agreements represent bilateralism in contrast to the
multilateral system offered by the TRIPS Agreement.
CBD Implementation Undermining the TRIPS Agreement

In contrast to the TRIPS Agreement, the CBD has spurred a great deal of domestic
access legislation, in particular in the South. Several developing countries have added
as preconditions for access to their genetic resources that any ensuing patent appli-
cations should include and identify the source of the genetic material and the tradi-
tional knowledge used, evidence of fair and equitable benefit sharing, and prior
informed consent. For example, additional disclosure measures are found in the
common regime on access to genetic resources established under the Andean Pact by
Bolivia, Colombia, Ecuador, Peru, and Venezuela in 1996. This agreement requires
that patent applications shall contain authorization to use traditional knowledge

(Decision 391). Another example is the 1998 African Model Law for the protection
of the rights of local communities, of farmers and breeders, and for the regulation of
access to biological resources of the Organization of African Unity (OAU; now the
African Union). This model legislation requires a permit and prior informed consent
of communities, sharing of benefits from commercial products (50 percent), and
community rights to control access to biological resources and knowledge. Cur-
rently, close to one hundred developing countries are in the process of making legis-
lation and institutional arrangements to control access to genetic resources. The
CBD lists twenty-eight such domestic legal frameworks.8
Does this legislation implementing the CBD run counter to the TRIPS Agreement?
Article 27.1 of the TRIPS Agreement lists the prevalent patent criteria of novelty, in-
ventive step, industrial application, and reproducibility. Any disclosure requirement
such as prior informed consent or proof of fair and equitable benefit sharing con-
tained in the aforementioned legislation comes on top of the TRIPS criteria. It is
not clear whether the TRIPS Agreement may prohibit members from including other
such requirements or if such activity would be incompatible with the TRIPS Agree-
ment. If this legislation were to be brought to the dispute-settlement body of the
WTO, it would have to address the WTO’s competence to rule on domestic legisla-
tion passed to implement another international instrument—the CBD (Cannabrava
2001). So far, there has been no legal testing of the two regimes.
The proliferation of access legislation in the South has brought up questions
about the effects on innovation. Some argue that the emerging legal regimes to reg-
ulate access to genetic resources in the South will be an obstacle to technological
innovations based on genetic materials (Grajal 1999). To the extent that this is the
case, it runs counter to the TRIPS Agreement, because the core objective of the IPR
systems established under it is to create incentives for innovation (see figure 4.2). On
the same note, in the eyes of the U.S. negotiation team, it was the developing coun-
tries that were trying to ‘‘hollow out’’ the TRIPS Agreement by using the CBD
(Raustiala 1997, 47).
Assessment and Outlook

The divergent approaches of the CBD and the TRIPS Agreement to access and ben-
efit sharing relating to genetic resources have led to the implementation of legisla-
tion in the North and the South that works at cross-purposes, and have resulted in
an ‘‘arms race.’’ The CBD rests on the principle that equitable sharing of biotechno-
Figure 4.2
Influence of the CBD on the implementation of the WTO TRIPS Agreement
logical use of genetic resources is an essential element of biodiversity conservation

and hence encourages such international benefit sharing. In contrast, the TRIPS
Agreement has been set up to strengthen and harmonize intellectual property rights
in all technological fields, including biotechnology. It thus is hardly compatible with
the CBD as regards both equitable benefit sharing and, indirectly, the objective of
conservation of biological diversity. Whereas the CBD does not embody legal and
economic enforcement mechanisms, patents and intellectual property rights in gen-
eral are legally enforceable under the WTO dispute-settlement procedure. At the na-
tional level, most developed countries have complied with the TRIPS obligation of a
strengthened patent system, but have not established legislation for access and ben-
efit sharing in accordance with the CBD. The EU Patent Directive is a case in point.
Several newly industrializing countries have also complied with the TRIPS Agree-
ment as part of the package of joining the WTO (Sell 2002). In contrast, the least
developed countries have not yet complied with the TRIPS Agreement and many
developing countries have put in place legislation to implement the requirements of
the CBD concerning benefit sharing. Both sides can thus be seen to have engaged in
an arms race at the implementation level (Behavioral Interaction).
There are indications that this arms race will continue and is escalating, fueled by
the diverging incentives emanating from the CBD and the TRIPS Agreement. The
‘‘TRIPS plus’’ agreements promoted by the North provide a case in point, as does
the Substantive Patent Law Treaty pursued by developed countries in the WIPO be-
cause it makes no reference to related processes dealing with traditional knowledge
or with access and benefit sharing (Tvedt 2005). On the other hand, national mea-
sures to promote technology transfer under Article 16 of the CBD may raise WTO
‘‘most-favored-nation’’ issues if Convention parties and nonparties were treated dif-
ferently. It might also raise TRIPS issues if owners of proprietary technology were
compelled to license technologies on grounds other than those prescribed in the
TRIPS Agreement. The regulatory controversy over whether TRIPS Article 27.3(b)
allows parties to design their own sui generis systems or implies the use of the plant
breeders’ rights system provided by UPOV also remains unresolved. Moreover,
institutional cooperation remains hampered because the CBD has still not been
granted observer status under the TRIPS Agreement, in spite of repeated calls from
its Conference of the Parties (COP). In the WTO TRIPS Council, the United States
has repeatedly turned down these efforts on the grounds that it does not see the need
for them.9
In this arms race, the TRIPS Agreement may be better equipped than the CBD.
The WTO and its TRIPS Agreement are stronger than the CBD in terms of compli-
ance mechanisms and in being controlled by the more powerful states. The dispute-
settlement mechanism of the WTO rates among the strongest of such international
instruments. In particular, if WTO members do not comply with the TRIPS Agree-
ment, they may become liable to economic sanctions. Also, patenting in the biotech-
nology sector is a small part of the larger issue area of international trade and
patenting in all sectors. Hence, even though the conflicts are explicitly admitted
and attended to, implementation of the CBD objectives may be hampered by the
stronger regulatory force of the TRIPS Agreement.
Nevertheless, there is ample evidence that the issue of the two regimes’ diverging
objectives is recognized in the WTO and the CBD alike, and efforts to mitigate the
conflict and increase synergy have been initiated in both forums as well as in third
institutions. Within the CBD, the COP has called for examination of the relation-
ship between the CBD objectives and the strengthened IPR systems under the WTO.
Moreover, the CBD established a working group on access and benefit sharing,
which prepared the Bonn Guidelines adopted at its sixth COP in The Hague in
2002 and further elaborated at the seventh COP in Kuala Lumpur in 2004. The
Bonn Guidelines encourage prior informed consent and mutually agreed-on terms
by making concrete suggestions for how these principles could be included in bilat-
eral bioprospecting agreements. Significantly, the Bonn Guidelines encourage dis-
closure of country of origin in patent applications. The Guidelines also propose a

certification system for trade in genetic resources and the establishment of an inter-
national Ombudsman for monitoring infringements on bioprospecting deals. An
unresolved question is how to draw the borders with respect to other international
regimes, such as the TRIPS Agreement.
Efforts are also underway in the WTO. As a response to the high level of conflict
associated with TRIPS Article 27.3(b), plans for its revision have been in the pipeline
several times. Only the deadline for least-developed-country compliance has so far
been altered by extending it from 2005 to 2016 at the WTO Doha ministerial meet-
ing in 2001. However, an increasing number of developed and developing countries
agree on the need to modify Article 27.3(b) of the TRIPS Agreement (Cannabrava
2001). The Doha Ministerial Declaration of 2001 instructed the WTO members to
examine the relationship with the CBD. This has focused on the issue of disclosing
the origin of genetic resources used as basis for inventions when applying for pa-
tents. In 2000, Brazil demanded that Article 27.3(b) should be amended to include
as conditions for patentability (1) the identification of the source of the genetic
material; (2) the related traditional knowledge used to obtain that material; (3) evi-
dence of fair and equitable benefit sharing; and (4) evidence of prior informed con-
sent from the government or the traditional community for the exploitation of the
subject matter of the patent.10 In addition, the EU made a turnaround on the issue
by tabling a plan to the TRIPS Council in September 2002 with the key proposal
to make disclosures (in a multilateral system) of geographic origin of biological
material used in biotechnology inventions. Patent applications would be required
to provide such disclosure, although this obligation would not constitute a ‘‘formal
patentability criterion’’ (European Commission 2002). The EU also addressed the
need to provide better protection of traditional knowledge and recognized the right
of subsistence farmers in developing countries to reuse and exchange seeds, even if
these seeds are covered by intellectual property rights.
Finally, a number of responses by other international forums indicate a growing
legitimacy of the CBD objectives and have the potential of mitigating the tensions
between the CBD and the TRIPS Agreement. First, UNESCO has addressed tradi-
tional knowledge and intellectual property by developing model provisions on the
protection of folklore. Second, UNCTAD runs the Biotrade Initiative, which seeks
to assist developing countries in developing, at the national level, an institutional en-
vironment to facilitate trade and investment in products and services of biological
diversity, as a means to attain the objectives of the CBD (http://www.biotrade.org/).
Third, the WIPO Intergovernmental Committee on Intellectual Property and Genetic

Resources, Traditional Knowledge and Folklore has yet to decide on the interrela-
tion between access to genetic resources and disclosure requirements in intellectual
property applications (Baumüller and Vivas-Eugui 2004). Another WIPO develop-
ment is the General Assembly decision of October 5, 2004, to adopt a ‘‘development
agenda.’’ This includes the promotion of developing-country access to knowledge
and technology, such as the World Wide Web and the Human Genome Project
(Capdevila 2004). An open question is whether this may affect the WIPO negotia-
tion process on the Substantive Patent Law Treaty, which has been insulated from
links to the access and benefit-sharing debate, or whether WIPO will mimic the
WTO in limiting cooperation with multilateral environmental agreements to the
general forums, while not allowing them as observers on more specific topics.11
Fourth, the FAO International Treaty on Plant Genetic Resources for food and
agriculture of 2001 explicitly prohibits patenting of genetic material ‘‘in the public
domain’’ and ‘‘in the form received’’ from its multilateral system of genebanks.
This multilateral system is aimed at solving problems of access to plant genetic
resources in agriculture to ensure that the resources are available for research and
plant breeding.
Conclusion
Due to its broad scope, the Convention on Biological Diversity interacts with a great
number of other international and regional legal instruments. The link is most obvi-
ous in the case of the multitude of other agreements that also pursue the goal of na-
ture conservation, although in a more limited way than the CBD. These include in
particular CITES, the CMS, and the Ramsar Convention on wetlands. Based on this
commonality of objectives, the relationship with these agreements has been rather
harmonious. In contrast, the relationship with a number of instruments pertaining
to the use of genetic resources in biotechnology has been more problematic, includ-
ing in particular regulations to protect intellectual property related to genetic re-
sources such as the WTO TRIPS Agreement and the EU Patent Directive.
The interaction between the CBD and the TRIPS Agreement is based on different
objectives and an overlap of memberships and jurisdictional scope, and it has led to
disruptive effects on either side. Both regimes have a global membership and regu-
late the same natural resource, namely genetic resources. At the same time, they op-
erate in different policy fields and pursue different objectives (equitable sharing of
benefits versus protection of intellectual property). With both regimes ten years into
their implementation phase, their interaction has led to disruptive effects in particu-
lar with regard to conservation of, access to, and equitable sharing of benefits from
genetic resources. The influence is arguably unintentional, but clearly anticipated.
Within both arenas, the parties acknowledge that both regimes influence each other.
Most of the interaction has been handled through collective decision making in the
two regimes and through behavioral adaptation in their implementation. The result
of the interaction so far may to some extent be viewed as an arms race. Strengthened
measures within one arena have in turn led to heavier ‘‘weapons’’ being applied by
the other. The TRIPS plus agreements, the negotiations on the WIPO Substantial
Patent Law Treaty, and legislation implementing the TRIPS Agreement in particular
in the North on the one hand contrast with increasingly cumbersome access reg-
ulations emerging in many developing countries on the other. Because of the in-
equality of subsistence farmers and transnational corporations that has been further
enhanced by new patent systems, the situation may be more problematic for the bio-
diversity side.
The existing tensions can be traced back to the regulatory competition between
both regimes that has so far not led to a clear delimitation of their jurisdictions.
The CBD agenda and the issue linking taking place in the biodiversity negotiations
greatly reflected the developments in the WTO/TRIPS negotiations. This was seen
most particularly in the early decision to include the issue of access to genetic
resources on the CBD agenda, which established the link to domesticated biological
resources and emphasized the need for equitable sharing of benefits. In contrast, in
the early phases, the biodiversity issue was subordinate to the more general discus-
sions in the WTO. Lately, however, this issue has entered the WTO equation in
relation to access to raw material and technological innovation. The pressure to in-
clude links to these dimensions within the TRIPS agenda has been steadily growing.
The examination of activities suggests that the quest for enhanced synergies is
hard to separate from a debate on values when different objectives clash. Policy
measures may facilitate technological innovation, improve conservation, or enhance
international equity. However, any one measure will hardly be able to achieve a top
score on all dimensions.
Therefore a need exists to increase bridge building and restore mutual trust in this
issue area. Klaus Töpfer, head of UNEP, has stated that ‘‘the interrelation between
the WTO and CBD is extremely high on the agenda. We don’t want to wait until we
have conflicting cases. We want to do whatever is possible to solve the interrelations
between them.’’12 A number of relevant activities are underway, including the FAO
Treaty on Plant Genetic Resources and the CBD Bonn Guidelines on access and ben-
efit sharing that may turn out to have mitigating effects. Other constructive activities
include the EU proposal to the TRIPS Council to make disclosure of origin of ge-
netic material an obligation. This EU proposal constitutes one among the first few
steps toward synergy, designed to increase mutual trust among the parties involved
in transactions with genetic resources. These activities may indicate that the legiti-
macy of the CBD objective of equitable sharing is gaining ground. The question of
how regimes develop legitimacy is crucial for future transactions with genetic mate-
rial and will have implications for biodiversity conservation and biotechnological
innovation alike.
Notes
1. Biotechnology traditionally includes activities such as baking bread and brewing beer,
where different organisms (such as yeast and wheat) interact to create a new product. The
concept of ‘‘new biotechnologies’’ indicates more direct dependency on human intervention
(recombinant DNA techniques and genomics).
2. The concepts of developing and developed/industrialized countries, or ‘‘South’’ and
‘‘North,’’ tend to obscure significant political and economic differences within the two groups.
In this case, moreover, most countries are heavily interdependent in their use of genetic
resources. Because large parts of the international debate on biodiversity have been for-
mulated as part of a North-South conflict, and for simplicity, I will nevertheless keep to this
admittedly simplified dichotomy. The CBD treaty makes specific use of these concepts in
designating duties and rights to the parties.
3. Personal communication with Jan P. Borring, senior advisor, Norwegian Ministry of Envi-
ronment and Norwegian delegation member to the CBD negotiations.
4. Screening results from Shaman Pharmaceuticals revealed that 74 percent of the samples
that displayed promising chemical activity directly correlated with the original ethnobotanical
use in the context of traditional knowledge (Sheldon and Balick 1995, 58–59).
5. Peter Johan Schei, special advisor to UNEP, Norwegian delegation leader, and director
and chair of several working groups and scientific panels to CBD, comments: ‘‘There is indeed
a conflict and the EU Patent Directive is contrary to the intentions of the CBD’’ (March 6,
2002, Research Council of Norway, Conference on Sustainable Development).
6. ‘‘Biotechnology: EU Countries Still Dragging Feet over Directive on Legal Protection of
Inventions,’’ Europe Information Environment, 307/2.12-18.2004, p. III3.
7. http://europa.eu.int/comm/trade/bilateral/mercosur/pr240702_en.htm; see also http://www
.bilaterals.org.
8. See http://www.biodiv.org/programmes/socio-eco/benefit/measures.aspx, which also con-
tains the mentioned regional measures.
9. Jan P. Borring, special advisor to the Norwegian Ministry of Environment, personal com-
munication, April 2002.
10. Review of Article 27.3(b), Communication from Brazil, Permanent Mission of Brazil,
Geneva, WTO document IP/C/W/228, November 24, 2000.
11. Interview with Aimee Gonzales, senior policy advisor, WWF, Gland, Switzerland, Sep-
tember 2004.
12. ‘‘Interview—Tensions Mount over Gene Rights, Trade—UNEP,’’ Planet Arc, April 9,
2002.
References
Barton, John, and Peter Berger. 2001, Summer. Patenting Agriculture. Issues in Science and
Technology Online. Available at http://www.nap.edu/issues/17.4/p_barton.htm.
Barton, John, and Wolfgang E. Siebeck. 1992. Intellectual Property Issues for the Interna-
tional Agricultural Research Centres: What are the Options? Issues in Agriculture, no. 4.
Washington, DC: Consultative Group on International Agricultural Research.
Baumüller, Heike, and David Vivas-Eugui. 2004. Towards Effective Disclosure of Origin: The
Role of the International ABS Regime. Bridges 8 (4): 21–22.
Bent, Stephen A., Richard L. Schwaab, David G. Conlin, and Donald D. Jeffery. 1987. Intel-
lectual Property Rights in Biotechnology Worldwide. New York: Stockton Press.
Berg, Trygve. 1995. Kulturplanter og husdyr. In Afrika—natur, samfunn og bistand, edited
by Nils Christian Stenseth, Kjetil Paulsen, and Rolf Karlsen. Oslo: Gyldendal, ad Notam.
Bilder, Richard B. 1980. International Law and Natural Resources Policies. Natural Re-
sources Journal 20: 451–486.
Borring, Jan Petter. 2001. The International Undertaking on Plant Genetic Resources for
Food and Agriculture: Is It Now or Never? IPRGI Newsletter for Europe (17): 5.
Cannabrava, Francisco. 2001. TRIPs and the CBD: What Language for the Ministerial Dec-
laration? Bridges, Between Trade and Sustainable Development. Special Doha Ministerial
Issues 5 (8): 7.
Capdevila, Gustavo. 2004, October 15. South Gains Ground in Intellectual Property Debate.
Geneva: Inter Press Service News Agency. Available at http://www.ipsnews.net/africa/interna
.asp?idnews=25876.
Chaytor, Beatrice, and James Cameron. 2000. The Treatment of Environmental Considera-
tions in the World Trade Organisation. In Helge Ole Bergesen, Georg Parmann, and Øystein
B. Thommessen, eds., Yearbook of International Co-operation on Environment and Develop-
ment 1999/2000, 55–64. London: Earthscan.
Correa, Carlos M. 1999. Access to Plant Genetic Resources and Intellectual Property Rights.
Background Study Paper No. 8. Rome: FAO Commission on Genetic Resources for Food and
Agriculture.
Crespi, R. Stephen. 1988. Patents: A Basic Guide to Patenting in Biotechnology. Cambridge:
Cambridge University Press.
Crucible Group. 1994. People, Plants and Patents: The Impact of Intellectual Property on
Biodiversity, Conservation, Trade and Rural Society. Ottawa: International Development Re-
search Centre.
Dutfield, Graham. 2001. Biotechnology and Patents: What Are Developing Countries Doing
about Article 27.3(b)? Bridges. Post-Doha Ministerial Issue 5 (9): 17–18.
Egziabher, Tevolde B. G. 1999. The TRIPs Agreement of the WTO and the Convention on
Biological Diversity: The Need for Co-ordinated Action by the South. Third World Resur-
gence 106: 4–7.
European Commission, Directorate-General for Trade. 2002, September 12. Communication
by the European Communities and Their Member States to the TRIPs Council on Review of
Article 27.3(b) of the TRIPs Agreement, and the Relationship between the TRIPs Agreement
and the Convention on Biological Diversity (CBD) and the Protection of Traditional Knowl-
edge and Folklore. A Concept Paper. Brussels: European Commission.
FAO. 1993. Harvesting Nature’s Diversity. Rome: Food and Agriculture Organization.
Fauchald, Ole Kristian. 2001. Patenter og Allmenningens Tragedie (Patents and the Tragedy
of the Commons). Lov og Rett. Norsk juridisk tidsskrift (Norwegian Journal of Legal Affairs)
40: 399–412.
Gillespie, Alexander. 1999. Sinks, Biodiversity and Forests: The Implications of the Kyoto
Protocol upon the Other Primary UNCED Instruments. Paper presented at the International
Conference on Synergies and Coordination between Multilateral Environmental Agreements,
United Nations University, July 14–16, Tokyo.
Gleckman, Harris. 1995. Transnational Corporations’ Strategic Responses to Sustainable De-
velopment. In Helge Ole Bergesen, Georg Parmann, and Øystein B. Thommessen, eds., Green
Globe Yearbook 1995, 93–106. Oxford: Fridtjof Nansen Institute and Oxford University
Press.
Grajal, Alejandro. 1999. Biodiversity and the Nation State: Regulating Access to Genetic
Resources Limits Biodiversity Research in Developing Countries. Conservation Biology 13
(1): 6–10.
Harlan, Jack R. 1995. The Living Fields: Our Agricultural Heritage. Cambridge: Cambridge
University Press.
Hendrickx, Frederic, Veit Koester, and Christian Prip. 1993. Convention on Biological Diver-
sity: Access to Genetic Resources; A Legal Analysis. Environmental Policy and Law 23 (6):
250–258.
Heywood, Vernon H. 1995. Global Biodiversity Assessment. Cambridge: Cambridge Univer-
sity Press.
Kate, Keriten, and Sara A. Laird. 1999. The Commercial Use of Biodiversity: Access to Ge-
netic Resources and Benefit-Sharing. London: Earthscan.
Keystone International Dialogue on Plant Genetic Resources. 1990. Final Consensus Report.
Madras Plenary Session, February 14, Keystone, Colorado.
Kloppenburg, Jack R. 1988. First the Seed. Cambridge: Cambridge University Press.
Kloppenburg, Jack R., and Daniel Lee Kleinman. 1987. The Plant Germplasm Controversy.
BioScience 37 (3): 190–198.
Koester, Veit. 1997. The Biodiversity Convention Negotiation Process and Some Comments
on the Outcome. Environmental Policy and Law 27 (3): 175–192.
Mooney, Pat R. 1983. The Law of the Seed: Another Development and Plant Genetic Re-
sources. Development Dialogue 1983: 1–2. Uppsala: Dag Hammarskjøld Foundation.
Morin, Jean-Frédéric. 2003. Le Droit International des Brevets: Entre le Multilatéralisme et le
Bilatéralisme Américain. Études internationales 34 (3): 537–562.
NORD. 1992. Bioteknologiska Uppfinningar och Immaterialrätten i Norden—del II (Bio-
technological Inventions and Intellectual Property Rights in the Nordic Countries—Part II).
Copenhagen: Nordisk Ministerråd (Nordic Council of Ministers).
Pistorius, Robin, and Jeroen van Wijk. 1999. The Exploitation of Plant Genetic Information.
Doctoral dissertation. Amsterdam: Amsterdam University Press.
Raustiala, Kal. 1997. Global Biodiversity Protection in the United Kingdom and the United
States. In Miranda A. Schreurs and Elizabeth C. Economy, eds., The Internationalization of
Environmental Protection, 42–73. Cambridge: Cambridge University Press.
Raustiala, Kal, and David G. Victor. 2004. The Regime Complex for Plant Genetic Re-
Rosendal, G. Kristin. 1991. International Conservation of Biological Diversity: The Quest for
Effective Solutions. FNI Report: 012-1991. Lysaker, Norway: Fridtjof Nansen Institute.
Rosendal, G. Kristin. 2001a. Impacts of Overlapping International Regimes: The Case of Bio-
Rosendal, G. Kristin. 2001b. Institutional Interaction in Conservation and Management of
Biodiversity: An Inventory. Available at http://www.ecologic.de/projekte/interaction.
Rosendal, G. Kristin. 2001c. Overlapping International Regimes: The Forum on Forests (IFF)
between Climate Change and Biodiversity. International Environmental Agreements: Politics,
Law and Economics 1 (4): 447–468.
Rosendal, G. Kristin, and Steinar Andresen. 2003. UNEP’s Role in Enhancing Problem-
Solving Capacity in Multilateral Environmental Agreements: Co-ordination and Assistance
in the Biodiversity Conservation Cluster. FNI Report 10/2003. Lysaker, Norway: Fridtjof
Nansen Institute.
Sell, Susan K. 2002. Private Power, Public Law: The Globalization of Intellectual Property
Rights. Paper presented at the International Studies Association Annual Meeting, March 24–
27, New Orleans.
Sheldon, Jennie Wood, and Michael J. Balick. 1995. Ethnobotany and the Search for Balance
between Use and Conservation. In Timothy M. Swanson, ed., Intellectual Property Rights
and Biodiversity Conservation, 19–44. Cambridge: Cambridge University Press.
Stevenson, Mark. 2002, February 19. China, Brazil, India, 9 Other Nations Form Alliance
against Biopiracy. Associated Press news wire.
Stokke, Olav Schram, and Øystein B. Thommessen, eds. 2003. Yearbook of International Co-
operation on Environment and Development 2003/04. London: Fridtjof Nansen Institute and
Earthscan.
Svensson, Ulf. 1993. The Convention on Biodiversity—A New Approach. In Svedin Sjøstedt
and Hägerhäll Aniansson, eds., International Environmental Negotiations: Process, Issues
and Contexts, 164–191. Report 93:1. Stockholm: FRN Utrikespolitiska Institutet.
Swanson, Timothy, and Sam Johnston. 1999. Global Environmental Problems and Interna-
tional Environmental Agreements. Cheltenham: Edward Elgar.
Tvedt, Morten Walløe. 2005. How Will a Substantive Patent Law Treaty Affect the Public
Domain for Genetic Resources and Biological Material? Journal of World Intellectual Prop-
erty 8 (3): 311–344.
UNDP. 2000. Human Development Report. New York: United Nations Development
Programme.
UNEP. 1995. Global Biodiversity Assessment. Ed. Vernon H. Heywood and Robert T. Wat-
son. United Nations Environment Programme. Cambridge: Cambridge University Press.
Walden, Ian. 1995. Preserving Biodiversity: Role of Property Rights. In Timothy M. Swan-
son, ed., Intellectual Property Rights and Biodiversity Conservation, 176–198. Cambridge:
WCED. 1987. Our Common Future. World Commission for Environment and Development.
Oxford: Oxford University Press.
Wilson, Edward O., ed. 1988. Biodiversity. Washington, DC: National Academy Press.
5
Protecting the Northeast Atlantic: One Problem,
Three Institutions
The North Sea has been a core area of collaboration on the protection of the North-
east Atlantic marine environment since the early 1970s. Surrounded by densely
populated areas, the North Sea is an area of intense human activity. Land-based
(river input and direct discharge) and ocean-based discharges (dumping and inciner-
ation at sea) of hazardous substances and nutrients as well as atmospheric fallout
have been among the major sources of contaminants to the North Sea. Many of
these problems have been dealt with more or less concurrently by three different
types of international institutions: legal treaties on marine pollution, the ‘‘soft-law’’
International North Sea Conferences (INSCs), and the European Union (EU).1 These
institutions are a significant part of Europe’s marine-environment management, and
the thirty-year history of environmental collaboration in the Northeast Atlantic can
serve as a fascinating example of a transformation from inertia to action.
Much of the literature on regime interaction or linkages tends to emphasize
problems of institutional congestion and density (e.g., Weiss 1993; Rosendal 2001;
Oberthür 2001; see also chapter 1). We would thus expect that a case such as this
would represent a clear example of duplicated work and coordination problems
leading to low effectiveness. Contrary to conventional wisdom, however, the three
types of overlapping institutions covering the North Sea and the wider Northeast
Atlantic have proven mutually beneficial by fulfilling different functions, all of which
are needed to manage marine pollution effectively. Moreover, these functions would
be difficult to manage within the same institution due to internal contradictory
requirements. The synergistic result of institutional interaction in this issue area is
evident in the significant overall reductions achieved in the emissions of regulated
organic substances, pesticides, heavy metals, nutrients, and dumping and incinera-
tion at sea (Skjærseth 1999, 2000, 2002a, 2002b).
104 Jon Birger Skjærseth
The collective workings of the ‘‘soft-law’’ INSCs, the ‘‘hard-law’’ Oslo and Paris
Conventions (OSPAR), and the ‘‘supranational’’ EU have proven instrumental for
the achievements reached. In the next section of the chapter, a brief introduction to
the core institutions, the INSCs and OSPAR, is provided. These core institutions
have also performed well with a number of other international environmental in-
stitutions. Following their introduction, a broad view is thus taken on the most
relevant cases of institutional interaction in which they have been involved. Subse-
quently, the focus is put on the collaborative efforts of the INSCs, OSPAR, and
the EU. The main argument in this respect is that the ‘‘soft-law’’ INSC declarations
have speeded up the decision-making processes in OSPAR and the EU. Finally, the
chapter analyzes how implementation has been strengthened by ‘‘hard law.’’ The
argument here is that OSPAR and the EU in turn have facilitated domestic imple-
mentation of the original INSC Declarations.
The synergistic relationship between these institutions has been enhanced by
means of conscious institutional design. The cases show how the original design of
the institutions ‘‘trapped’’ the parties in a situation of inertia. However, the parties
were able to change their path through leadership by creating a new institution—the
INSCs—that became linked to OSPAR and the EU water policy. The INSCs were
deliberately designed to speed up the decision-making processes in these bodies. Fur-
ther lessons to be learned from these synergistic processes are discussed in the con-
cluding section of this chapter.
OSPAR and the International North Sea Conferences
The history of international cooperation among countries bordering the North Sea/
Northeast Atlantic is the story of evolution from a state of water and marine pollu-
tion ‘‘anarchy’’ to domestic and international ‘‘governance.’’ In 1972, the Conven-
tion for the Prevention of Marine Pollution by Dumping from Ships and Aircraft
(Oslo Convention) was established. Signed by all thirteen Western European mari-
time states, the Oslo Convention covers the entire Northeast Atlantic up to the
North Pole. In 1974, the Convention for the Prevention of Marine Pollution from
Land-Based Sources (Paris Convention) was signed in Paris by roughly the same
states as the Oslo Convention. The Paris Convention allowed the EU to join as a
contracting member, and water policy was the first subsector developed under EU
environmental policy. The Oslo and Paris Conventions were supported by a joint
secretariat, executive commissions (Oslo and Paris Commissions), and several stand-
Protecting the Northeast Atlantic 105
ing and ad hoc scientific/technical bodies. Together, both instruments are referred to
as ‘‘OSPAR.’’ In 1992, the Oslo and Paris Conventions were brought together to
form a single legal instrument for the protection of the Northeast Atlantic (OSPAR
Convention).
The main objective of the 1974 Paris Convention on land-based sources was to
take all possible steps to prevent pollution of the sea by individually and jointly
adopting measures to combat marine pollution and by harmonizing the parties’ pol-
icies in this regard. The Paris Commission was responsible for the supervision of the
Convention. In retrospect, the Paris Commission did not even come close to realiz-
ing the aims of the Paris Convention. On average, the Paris Commission produced
roughly only one legally binding decision and one recommendation each year from
1974 to 1987 (Skjærseth 2002a). Most of the recommendations and decisions
adopted concerned the blacklisted substances mercury and cadmium, about which
the parties were free to choose whether quality standards or emission standards
should apply. In practice, quality standards—defining the minimum quality of
water—gave the parties a considerable amount of leeway due to inadequate moni-
toring and scientific uncertainty. Moreover, the commitments were frequently soft-
ened by requiring the parties to do something ‘‘as soon as practicable’’ or ‘‘as soon
as possible.’’ Added to the poorly developed reporting routines, it is questionable
whether the parties were actually bound to do anything at all.
The situation on reporting was somewhat better under the 1972 Oslo Convention
on dumping and incineration at sea. The Convention established a permit system,
which required parties to submit to the Commission records of dumping permits
and approvals they had issued. Thanks to this procedure, the Commission obtained
an overview of who dumped what, where, and how much. Nevertheless, the perfor-
mance of the Oslo Commission was even worse than that of the Paris Commission
on substantial action. By 1987, the Oslo Commission had adopted two decisions,
three recommendations, and seven so-called agreements (Skjærseth 2001b). Most
of these agreements were directed at establishing cooperative procedures aimed at
controlling current behavior rather than changing it. In essence, the parties contin-
ued to use the North Sea as a trash can for hazardous industrial waste and sewage
sludge.
The Oslo and Paris Conventions and Commissions had two deficiencies. First, al-
though the most pressing ecological problems concerned the North Sea, the Conven-
tions covered the entire Northeast Atlantic, and included ‘‘laggard’’ countries such
as Spain and Portugal. The scope of the Convention hampered collective decision
making and led to decisions close to the lowest common denominator. Second, the
parties sought to establish a dynamic regime. Partly due to the legally binding nature
of the Conventions, however, it proved static and hard to change in practice.
Spurred by dissatisfaction with existing international institutions, Germany took
the initiative in arranging the first International North Sea Conference (INSC) in
Bremen in 1984. This initiative reflected a combination of entrepreneurial and intel-
lectual leadership (Underdal 1991; Young 1991). Germany showed entrepreneurial
leadership by choosing an option that was particularly conducive to solving existing
problems. It essentially had a choice between striving for a new convention on the
North Sea (as was proposed in the European Parliament in 1983) or generating po-
litical impetus in existing conventions by convening a North Sea Conference aimed
at producing ‘‘soft-law’’ declarations that could take immediate effect. Both options
would exclude states not bordering the North Sea. The second option in addition
avoided time-consuming new legal arrangements, including the need to dismantle
existing conventions. Intellectual leadership was evident in the introduction of the
precautionary principle (‘‘Vorsorgeprinzip’’) to guide protection of the North Sea.
Agreement on the precautionary principle was a precondition for the percentage-
reduction targets and the phasing out of dumping at sea agreed on later. In 1980,
the Council of Environmental Advisors, an independent body of experts appointed
by the German government, introduced the principle in a report on environmental
problems affecting the North Sea (Ministry of Environment and Energy 1995b).
The Bremen Declaration subsequently hinted at the precautionary principle and the
London Declaration adopted it.
While the Bremen Conference was originally conceived of as a onetime event, the
conferences evolved as a more permanent institution over time by the establishment
of the standing Committee of North Sea Senior Officials.2 The Bremen Conference
was followed by conferences in London in 1987, The Hague in 1990, Esbjerg (Den-
mark) in 1995, and Bergen (Norway) in 2002. Conference participants have been
the eight North Sea coastal states and the EU, which represent a subset of the origi-
nal Oslo and Paris Conventions parties. From 1990, Switzerland was also invited to
participate.
The London Declaration represented a turning point because of its ambition
to phase out dumping of industrial waste and incineration at sea and to achieve
reductions in inputs of nutrients to sensitive areas and in total inputs of hazardous
substances reaching the aquatic environment on the order of 50 percent between
1985 and 1995. The 1990 Hague Declaration clarified and strengthened the Lon-
don Declaration, particularly concerning land-based sources. The Oslo and Paris
Commissions (OSPARCOM) together with the EU took significant steps in the
same direction in the wake of the 1987 North Sea Conference.
Synergy in Practice
The list of interactions in table 5.1 is not exhaustive, but it provides the clearest
and most important cases within the environmental policy field. The interactions
between the INSCs and OSPAR as well as between these institutions and various
EU directives are further explored in subsequent sections of this chapter (see also
chapters 9 and 10). In addition, the INSCs and OSPAR have been influenced by
the Montreal Protocol for the protection of the ozone layer, the UN Framework
Table 5.1
Instances of interaction involving the INSCs and OSPAR
INSC Has facilitated and sped up development of OSPAR
OSPAR Has expanded the scope of INSCs and helped
implement INSC Declarations on marine pollution
Montreal Protocol on Substances Has helped implement the INSC commitments on
That Deplete the Ozone Layer carbon tetrachloride and methyl chloroform
UN Framework Convention on May provide incentive for CO2 sequestration in
Climate Change and Kyoto North Sea oil fields, which could violate OSPAR
Protocol prohibition on dumping at sea
London Dumping Convention Benefited from OSPAR that facilitated agreement on
global ban on dumping and incineration at sea within
London Convention
Rhine Convention Helps achieve objectives of OSPAR
Convention on Long-Range Contributes to achieving the objective of OSPAR on
Transboundary Air Pollution nutrients and eutrophication (NOx )
EU Nitrates Directive Was facilitated/triggered by INSC Declaration
Helps implementation of INSC objectives on nutrients
EU Urban Waste-Water Directive Was facilitated/triggered by INSC Declaration
Helps implementation of INSC objectives on nutrients
Includes an obligation to phase out sewage-sludge
dumping, which was facilitated/triggered by OSPAR

Source: Skjærseth 2001. Since both the INSCs and OSPAR address nutrients and hazardous
substances, interaction with other international institutions or EU directives will in most cases
affect both forums. For reasons of simplicity, however, only the institution (OSPAR or INSC)
most directly involved is listed.
Convention on Climate Change, the Rhine regime, and the regime on long-range
transboundary air pollution at the implementation level (Behavioral Interaction).
Most of these interactions have created synergy in that the source institutions have
led to the reduction of pollutants in support of the INSCs and OSPAR. Potential dis-
ruption occurred only with respect to the climate change regime because the latter
may provide an incentive to sequester CO2 in North Sea oil fields, which could vio-
late the OSPAR prohibition on dumping at sea. As a source, the INSCs and OSPAR
have in particular influenced the global London Convention on dumping. They
facilitated global agreement on a ban on dumping at sea that was reached in 1990
because the regional ban on dumping adopted in 1987–1988 also ensured the sup-
port at the global level of previous laggard states such as the United Kingdom (Inter-
action through Commitment).3
The main conclusion to be drawn from these twelve instances of regime inter-
action is that the institutions governing marine pollution in the Northeast Atlantic
live in harmony with each other in the sense that almost all instances of interaction
have triggered higher levels of effectiveness. While OSPAR and the INSCs have
proven mutually beneficial, they have also been able to benefit from the other insti-
tutions and have themselves influenced other international regimes and EU direc-
tives in a positive manner.
The positive relationships between international institutions in this issue area have
most likely been facilitated by two factors. First, all the institutions share roughly
the same environmental goals, in contrast to institutions dealing with, for example,
trade and the environment (see chapter 8). A second probable explanation for the
high level of synergy is the relatively long history of institutional cooperation on
marine-pollution control. Effectiveness tends to increase along with regime ‘‘age’’—
at least up to a certain point (Miles et al. 2002). Most international institutions need
a period of learning by doing before they mature. And most of the cases of interac-
tion mentioned in this chapter trace their beginnings to the 1970s and 1980s. The
institutions involved have thus had time to adapt and adjust.
Speeding up Decision Making by ‘‘Soft Laws’’
By the early 1980s there were growing indications that specific regions in the North
Sea were becoming severely polluted (Ehlers 1990). At the international level, nei-
ther the work of the Oslo and Paris Commissions nor that of the EU suggested
that stringent commitments could be initiated without additional political impetus.
Figure 5.1
INSCs strengthen OSPAR and the EU
Against this backdrop of inertia, Germany took the initiative to arrange the first
INSC at ministerial level in 1984. The INSCs turned out to have a profound impact
on OSPARCOM as well as on the water- and marine-pollution policy of the EU.
The causal mechanism identified is Interaction through Commitment. In this section
I analyze the outcome of these INSCs, explain how the INSCs succeeded in changing
OSPARCOM and EU policies, and explore the responses of OSPARCOM and the
EU (see figure 5.1).
The Breakthrough: International North Sea Conference Declarations
The 1984 Bremen Declaration The aim of the Bremen Conference was not to cre-
ate a new set of international agreements, but to provide political impetus for inten-
sifying the work of the existing international bodies. References to the Oslo and
Paris Commissions and the EU are sprinkled throughout the Declaration (Ministry
of Environment and Energy 1995b). Ministers of the eight North Sea coastal
states—Belgium, Germany, France, the Netherlands, Norway, Sweden, Denmark,
and the United Kingdom—met, together with representatives of the European Com-
mission. The Conference Declaration did not significantly strengthen international
marine-pollution commitments. As Pallemaerts (1992, 6) points out, the elasticity
of phrases such as ‘‘as far as possible,’’ ‘‘practicable,’’ and ‘‘economically feasible’’
meant that the Declaration contained few substantive commitments. However, the
Bremen Declaration was probably the first international text to hint at the precau-
tionary principle: ‘‘Coastal states and the EEC must not wait for proof of harmful
effects before taking action’’ (Ministry of Environment and Energy 1995b, 22). The
Bremen Conference had initially been envisaged as a onetime event, as noted, but
the ministers welcomed an invitation from the United Kingdom to host a second
INSC to review implementation and adopt further measures.
The 1987 London Declaration Particularly with respect to dumping at sea, the
London Declaration represented a turning point in stringency compared to the Bre-
men Declaration. For the first time, it was decided to impose significant targets on
dumping and incineration at sea within fixed time limits. For example, the parties
aimed at phasing out the dumping of industrial wastes in the North Sea by Decem-
ber 31, 1989. Commitments covering land-based sources were made subject to sim-
ilar targets. Eutrophication was included for the first time, and ambitious goals were
agreed to for phosphorus and nitrogen substances: a substantial reduction (on the
order of 50 percent) between 1985 and 1995 of inputs of phosphorus and nitrogen
to those areas of the North Sea where such inputs are likely, directly or indirectly, to
cause pollution. In effect, the agricultural sector was saddled with joint commit-
ments. The commitments made with regard to hazardous substances also appear
quite specific at first glance: a substantial reduction (on the order of 50 percent) be-
tween 1985 and 1995 in the total inputs to the North Sea via rivers and estuaries of
substances that are persistent, toxic, and liable to bioaccumulate. However, the min-
isters failed to agree on specific substances beyond those already covered by interna-
tional commitments. In contrast to the Bremen Declaration, the London Declaration
focused squarely on domestic implementation by requiring the preparation of na-
tional action plans on implementation. The 1987 London Declaration was one of
the first international environmental texts ever to explicitly incorporate the princi-
ples of precautionary action.
The 1990 Hague Declaration The Hague Conference clarified and strengthened
the London Declaration particularly concerning land-based sources. With regard to
hazardous substances, the aim of reducing discharges of such substances to levels
not ‘‘harmful to man or nature’’ was adopted for the first time, as a principle in Ar-
ticle 1 (Ministry of Environment and Energy 1995b). Against this backdrop, a list of
thirty-six hazardous substances was adopted and directly linked to the 50 percent
reduction target concerning hazardous substances. Moreover, the goal was changed
from ‘‘of the order of 50%’’ to ‘‘50% or more.’’ With regard to nutrients, measures
in the municipal, industrial, and agricultural sectors were agreed on, the most spe-
cific covering the municipal sector. Targets of a 70 percent reduction of land-based
and atmospheric inputs were adopted for the most dangerous substances—dioxins,
cadmium, mercury, and lead. Some new obligations were also adopted at the Hague
Conference. Agreement was reached on phasing out and destroying polychlorinated
biphenyls (PCBs) and hazardous PCB substitutes by 1999 at the latest.
The 1995 Esbjerg Declaration The commitments adopted on hazardous sub-

stances in 1995 could symbolize the significant gains made since the 1970s. In the
mid-1980s, only a few substances were under international regulation and even
fewer were made subject to elimination. Only ten years later, ministers agreed to
prevent the pollution of the North Sea by phasing out all hazardous substances ‘‘by
continuously reducing discharges, emissions and losses of hazardous substances
thereby moving towards the target of their cessation within one generation (25 years)
with the ultimate aim of concentrations in the environment near background values
for naturally occurring substances and close to zero concentrations for man-made
synthetic substances’’ (Ministry of Environment and Energy, 1995a, 18). The 2002
Bergen Declaration did not introduce any new significant commitments on haz-
ardous substances or eutrophication but focused on a number of new issue areas
such as climate change, biodiversity, renewable energy, and an integrated ecosystem
approach: it is too recent to have caused any significant change in target institutions.
The Causes: Changes in Membership and Institutional Setup

Why did the breakthrough on nutrients, hazardous substances, and dumping and
incineration at sea take place within the INSCs and not OSPARCOM or the EU?
The INSCs solved the two deficiencies of existing institutions. As noted, the Oslo
and Paris Conventions covered the entire Northeast Atlantic, although the pressing
ecological problems concerned the North Sea. Moreover, the Oslo and Paris Com-
missions and the EU lacked political momentum, though for different reasons.
OSPARCOM had developed into a stagnant institution that proved hard to change.
Collaborative efforts induced only low levels of participation; decision making was
incremental, bureaucratic, and based on unanimity. In addition, the green move-
ment was denied access, and actors worked together within a legally binding frame-
work that was difficult to amend and changed very little in practice (Skjærseth
2000). The EU also embraced unanimity decision making at the time—and the en-
vironment was not included in the EC Treaty until the adoption of the Single Euro-
pean Act in 1986.
The INSCs solved these problems by changing membership and institutional
setup. Southern European states such as Portugal and Spain were parties to OSPAR.
These states frequently allied themselves with the United Kingdom to form a strong
minority that was in a position to ensure that decisions reflected the lowest common
denominator. In essence, the INSCs excluded the non–North Sea states and left the
United Kingdom alone as the main ‘‘laggard’’ among the North Sea states. The po-
sition of the United Kingdom rested on its dedicated defense of environmental qual-
ity objectives, which in turn was closely linked to the fact the United Kingdom was
a net exporter of marine pollution due to the counterclockwise direction of the
North Sea currents. In contrast, the majority preferred uniform emission standards.
Crudely put, this majority emphasized that discharges of substances known to be
toxic, persistent, and bioaccumulative and listed on the blacklists should be limited
as far as possible at their source, whereas the defenders of environmental-quality
objectives maintained that standards set should be determined by observable nega-
tive effects in the marine environment for each particular substance.
Changes in the membership were not sufficient to reach a breakthrough since the
United Kingdom remained within the INSCs. However, thanks to their soft-law
qualities and political nature, the INSCs became a truly dynamic institution. First,
the INSCs were based on ministerial representation, a circumstance that paved the
way for political pressure to be put on the United Kingdom. Second, INSC Declara-
tions could take immediate effect since they were based on soft law, while proposed
amendments to OSPAR or EU legislation could take many years. Third, INSC Dec-
larations were specific and visible and verification procedures and practice improved
dramatically. In contrast to the Oslo and Paris Commissions, the INSCs systemati-
cally reviewed the achievements of the preceding declarations by preparing compre-
hensive progress reports on measures taken by each country as well as reductions in
inputs from each country. This raised the level of transparency and generated pres-
sure from environmental groups and more progressive states toward the United
Kingdom. As a consequence of the increasing political costs involved, the United
Kingdom accepted the precautionary principle and uniform emission standards.
The change in the UK position was made explicit in a 1988 position paper that
had far-reaching implications for both the United Kingdom and the North Sea/
Northeast Atlantic cooperation, particularly with regard to dumping and hazardous

substances (Skjærseth 2000, 124).
Several other factors contribute to explaining the patterns witnessed but cannot
fully account for them. First, there were ‘‘shocks and crises’’ in the North Sea in
the form of the exceptional 1988 and 1989 toxic algae blooms and seal epidemics.
They helped instigate the 1988 EU resolution on the North Sea (Prat 1990). But
they cannot explain the 1987 INSC breakthrough because they occurred later.
Moreover, their importance should not be exaggerated since their political impact
faded rapidly (Skjærseth 1999).4 Second, public opinion on environmental matters
changed significantly throughout Europe in the late 1980s (Hofrichter 1991). Min-
isterial representation at the INSCs became important as a result. Since ministers are
responsible to their domestic electorates, the ‘‘green wave’’ of the late 1980s could
be channeled more effectively into international negotiations than the former low-
level government representation in the Oslo and Paris Commissions had allowed
for. In this way, the impact of the changes in level of representation became closely
linked to the changes in public opinion. However, since public values and attitudes
evolved most significantly after 1987, they can hardly in themselves explain the
1987 INSC breakthrough, even in connection with the algae blooms (Skjærseth
1999). Moreover, public interest in (marine) environmental questions has been on
the wane in most central North Sea states since the early 1990s. Nevertheless, the
1995 INSC and OSPARCOM have continued to tighten up previous commitments,
not least on hazardous substances, which shows that the dynamic impact of the
institutions has continued despite fluctuations in public opinion.
While such exogenous factors contribute to our understanding of the influence of
one institution on another, other potential explanations can be ruled out. Phenom-
ena similar to the Antarctic ozone hole in 1985 were not uncovered. Uncertainty
relating to the causes and consequences of marine pollution of the North Sea has
receded gradually, so that the 1987 INSC Declaration cannot be traced back to a
breakthrough in scientific knowledge comparable with the discovery of the Antarc-
tic ozone hole. The North Sea Quality Status Reports presented in 1984 and 1987
both painted a picture of high uncertainty and ‘‘moderate’’ pollution levels. More-
over, the basic interests of the participating states did not change significantly, since
these interests were largely linked to the counterclockwise direction of the North Sea
currents that placed the states in a chainwise relationship of exporters and importers
of marine pollution.
The Effects: OSPAR and the EU

Although containing far-reaching commitments, the INSC Declarations represented
‘‘soft law’’ that was not legally binding for the parties. Since ministers come and go
as governments change, the political Declarations carried a risk of ending up as
‘‘paper tigers.’’ This was avoided by transforming the Declarations into legally bind-
ing commitments under OSPAR and the EU.
As a consequence of the INSCs, the decision procedures within the Oslo and Paris
Commissions changed from the requirement of unanimity to consensus linked to
various ‘‘fast-track options.’’ Previously, all thirteen states bordering the Northeast
Atlantic had to agree on the same international regulations. The INSCs led to new
procedures that allowed for differential obligations between North Sea and non–
North Sea states.
The London Declaration had its strongest impact on the Oslo Commission. The
first decision adopted by the Commission in 1988 stated that the riparian states of
the North Sea would apply the principles on the reduction and cessation of dumping
of polluting materials as set out in the INSC Declaration. This represented a sea
change in dumping policies, and the collective decisions of the Oslo Commission in
this period show a significant expansion in number, legal status, and content com-
pared to previous periods (Skjærseth 1999). The Oslo Commission achieved signifi-
cantly more from 1987 to 1990 than it had from 1974 to 1987. Even though the
total number of commitments under the Paris Commission did not increase signifi-
cantly, the parties adopted several ‘‘new’’ commitments, including a recommenda-
tion on the reduction of inputs of nutrients in 1988 and a coordinated program for
the reduction of nutrients in 1989. The Paris Commission did not, however, act
effectively to reduce discharges of hazardous substances.
EU responses to the 1987 conference related mainly to inputs of nutrients. In
the fourth EU Action Program on the Environment, adopted by the EU Council of
Ministers in 1987, the fight against seawater pollution from both point and diffuse
sources was considered to be a matter of priority. In 1988, the Council adopted
a resolution specifically related to the protection of the North Sea, requesting
the European Commission to combat nutrients from different sources, particularly
agriculture, and to present proposals on urban wastewater treatment. The resulting
directives were to become the major means of combating eutrophication in the
North Sea. Concerning hazardous substances, a proposed directive on the elimina-
tion of PCBs was directly inspired by the 1987 London Declaration (Prat 1990).
The Hague Declaration had its strongest impact on the Paris Commission. On the
basis of the list of hazardous substances adopted by the Hague Conference, the Paris
Commission started systematically addressing discharges from specific industrial
sectors. It took action on several fronts, including best environmental practices
(BEP) on diffuse sources and best available technology (BAT) on industrial point
sources. In addition to the two recommendations adopted in 1988 and 1989 on
nutrients, the Commission adopted Recommendation 92/7 on the Reduction of Nu-
trient Inputs from Agriculture.
The INSCs and OSPAR created a dumping policy for the EU. The EU had been
working on regulating dumping since the early 1970s. The European Commission
tabled the first proposal for a directive in 1976. However, the Council did not suc-
ceed in adopting any specific directives on dumping. The EU implemented the deci-
sion of the INSC and the Oslo Commission on phasing out sewage-sludge dumping
by including this obligation in the Urban Waste-Water Directive of 1991. Moreover,
the EU attended to OSPAR dumping policy by ratifying the 1992 OSPAR Conven-
tion in 1998 (Council Decision 98/249/EC (12)). Concerning nutrients, two impor-
tant EU directives were adopted in 1991 based on the initiatives of the late 1980s.
Besides their importance for the North Sea, the Nitrates and the Urban Waste-Water
Directives reflected a slightly different approach from previous directives. Like the
North Sea commitments, the new directives attacked the sources of pollution and
described clear goals within given time frames, while relying less on quality objec-
tives (Richardson 1994, 150). The Urban Waste-Water Directive set specific require-
ments on wastewater collecting systems to be implemented by the year 2000 or
2005 concerning nutrient discharges. The Nitrates Directive aims at supplementing
these efforts by specifically addressing nutrient emissions from the agricultural sec-
tor. The agricultural sector was also made subject to a regulation on environmen-
tally friendly production methods in 1992. These commitments overlap with both
the INSC Declarations and OSPAR commitments.
The Esbjerg Declaration on hazardous substances initiated actions both within
OSPAR and the EU. The political agreement to phase out hazardous substances
within twenty-five years with the ultimate aim of achieving concentrations in the
environment near background values for naturally occurring substances has been
viewed as a breakthrough. In 1998, OSPAR copied this agreement in the so-called
Sintra statement on a total phaseout of emissions of hazardous substances by
2020 (available at http://www.ospar.org/eng/html/md/sintra.htm). In 2000, the EU
adopted the Water Framework Directive (Directive 2000/60/EC), which sets out its
ambition ‘‘to cease or phase out discharges, emissions and losses of priority hazard-
ous substances, with the ultimate aim of achieving concentrations in the marine en-
vironment near background values for naturally occurring substances and close to
zero for man-made synthetic substances’’ (Art. 1; see also chapter 9).
Overall, the direct consequences of the OSPAR and EU responses to the North
Sea Declarations were thus twofold. First, the North Sea Declarations were trans-
formed into legally binding OSPAR commitments and/or EU regulations, directives,
or decisions. Second, the geographic coverage of the Declarations was extended to
the EU area and/or the Northeast Atlantic area.
Strengthening Implementation by ‘‘Hard Law’’
In this section, we will see how OSPAR and the EU subsequently facilitated the
implementation of the INSC Declarations. In particular, the hard-law nature of
OSPAR and EU commitments contributed decisively to behavioral changes that
improved the protection of the Northeast Atlantic and thus supported the achieve-
ment of the objectives of the INSC Declarations. This case is thus an example of Be-
havioral Interaction (figure 5.2).
Figure 5.2
OSPAR and the EU strengthen implementation of the INSCs
OSPAR and EU Legal Instruments: Implementing INSC Declarations

The 1995 Esbjerg Declaration identified previously adopted OSPAR decisions and
EU directives and regulations as the principal measures for goal attainment. While
this change in focus reflects important changes in EU marine and water policy, it
also reflects the fact that Norway and Switzerland were the only non-EU INSC
countries as of 1995. Concerning hazardous substances like pesticides, the Declara-
tion refers to Recommendation 94/7 of the Paris Commission on national plans for
reduction of pesticides from agriculture as well as to Directive 91/414/EEC concern-
ing the placing of plant-protection products on the market. On nutrients, the Decla-
ration links progress directly to national action plans based on a number of OSPAR
recommendations and EU directives, in particular the Urban Waste-Water and
Nitrates Directives. According to the Progress Report presented to the Esbjerg Con-
ference (Ministry of Environment and Energy 1995c, 33), ‘‘Although the North Sea
Conferences have provided the political skeleton, it has been left to the established
legal frameworks . . . to implement Ministers’ decisions by providing the necessary
detailed and binding (legal) framework for the North Sea States, in particular
through the medium of Decisions and Recommendations adopted by the Oslo and
Paris Commissions and by EU Directives.’’ This process of identifying previously
adopted EU and OSPAR commitments continued at the INSC in Bergen in 2002
(Ministry of the Environment 2002). The conference showed that there had been
progress toward achieving the targets on hazardous substances, and most North
Sea states had achieved a 70 percent reduction in mercury, lead, and cadmium. The
OSPAR strategy on hazardous substances and the EU Water Framework Directive
were regarded as effective means for coping with the remaining implementation defi-
ciencies and future challenges. According to the European Environment Agency
(EEA), the control of hazardous substances in marine waters has been a success
largely due to OSPAR (EEA 2001). The next INSC is scheduled to take place in
Stockholm in 2006.
There were more problems for the North Sea states with respect to nitrogen. Lack
of progress was directly related to delays in implementing the Nitrates and Urban
Waste-Water Directives. In Bergen, the North Sea ministers agreed on full imple-
mentation of these directives and the Water Framework Directive as well as on
reducing the use of fertilizer through Council Regulation EC/1257/1999 on support
for rural development from the European Agricultural Guidance and Guarantee
Fund (EAGGF).
The Causes: Institutional Authoritativeness and Enforcement

OSPAR has strengthened implementation mainly through the legally binding nature
of commitments. In general, there is a conviction in the legal literature on inter-
national obligations that joint commitments should preferably be legally binding
on states (Nollkaemper 1993). This view is based on the assumption that states will
be more inclined to comply with legal than with nonlegal obligations. The distinc-
tion between binding and nonbinding commitments has been perceived as impor-
tant by the OSPAR parties, and the legal nature of commitments has been discussed
repeatedly (Skjærseth 1999). Likewise, the INSC Declarations contain several refer-
ences indicating that the North Sea states considered it important that joint com-
mitments be implemented as binding decisions within OSPAR. When it comes
to enforcement, however, the distinction between binding and nonbinding obliga-
tions has little effect—if any. Enforcement has been regarded as the weakest part
of international efforts to regulate marine pollution: deterrence is considered un-
likely since few disputes are actually settled by international courts (Ijlstra 1986;
Nollkaemper 1993). The 1992 OSPAR Convention contains detailed dispute-
settlement procedures and places slightly more weight on enforcement, but it does
not set out which specific steps should be taken in cases of noncompliance. Thus,
it seems reasonable to assume that legally binding OSPARCOM decisions have
constrained state behavior due to their authoritativeness in addition to affecting
incentives.
The phasing out of dumping of industrial waste in the North Sea shows that by
transforming ‘‘soft law’’ into ‘‘hard law,’’ states that continued with dumping had
to follow demanding, legally binding procedures. The 1987 London Conference
took the decision to phase out such dumping by December 31, 1989, and the Oslo
Commission followed up by translating this political agreement among the North
Sea states into a legally binding decision. In this case, the EU had no competence
and the Community was not a party to the Oslo Convention. There are clear indica-
tions that the United Kingdom did not seriously plan to change its behavior in ac-
cordance with the international commitments (Skjærseth 2002b). In late 1989, the
UK Ministry of Agriculture, Food and Fisheries decided to support applications for
licensing 50,000 tons of toxic waste through the Oslo Commission’s Prior Justifica-
tion Procedure. This procedure was adopted in 1988 as a direct consequence of the
introduction of the precautionary principle by the 1987 INSC. It reversed the bur-
den of proof: potential dumpers were now obliged to prove through complicated
and expensive laboratory tests that substances they intended to dump could not
harm the marine environment. Several North Sea states protested the UK decision
and Greenpeace brought the case to the media’s attention. An extraordinary meet-
ing of the Oslo Commission ad hoc working group on dumping was convened.
The final decision to phase out dumping of industrial waste (and sewage sludge)
was made by Agriculture Minister John Gummer in 1990.
In cases of noncompliance, OSPAR has more competence than the INSCs due
to its legally binding properties and the EU has more competence than OSPAR
owing to its supranational attributes. When the EU acts it does so with significantly
more force than traditional regimes simply because the compliance instruments at
its disposal are more powerful (Tallberg 2002). Concerning enforcement, the main
formal difference between OSPAR and the EU is that EU directives impose legal
obligations directly on the member states (Skjærseth and Wettestad 2002). Failure
to comply with EU law can be relied on in national courts required to interpret
national laws in line with EU law (sympathetic interpretation). Such failure can
even result in awards of damages to individuals who have suffered loss as a
consequence.
The enforcement powers of the European Court of Justice (ECJ) are also unique.
The ECJ has developed a number of principles affecting national law and policies as
well as how EU policies apply (Haigh 2003). Since 1993, the ECJ may impose fines
against member states that have failed to comply with previous rulings of the court.
This provision was applied for the first time against Greece for its failure to observe
a court ruling on waste management.5 The United Kingdom was also threatened
with heavy daily fines for persistent failure to comply with EU bathing-water rules.6
Note that the threat has passed because the United Kingdom has come into compli-
ance with regard to the specified bathing waters—if a few decades behind schedule.
The ECJ has initiated legal action on a number of water directives linked to marine
pollution, including the Dangerous Substances Directive, the Urban Waste-Water
Directive, the Shellfish Directive, the Surface Waters Directive, the Nitrates Direc-
tive, and the Bathing Waters Directive.
The INSC obligation to reduce input of nitrogen substances in sensitive areas by
50 percent illustrates the causal mechanism of Behavioral Interaction (see figure
5.2). The INSC Declaration had scant effect on implementation. Implementation
problems in the case of nitrogen have mainly been related to strong farming lobbies
and conflicts of interest between environmental and agricultural authorities. The
United Kingdom, for example, argued that there were no sensitive areas around its
coast and did not take any significant steps to reduce nitrogen inputs.
However, the INSC obligation triggered the adoption of the EU Nitrates Direc-
tive. Even though the Nitrates Directive is an extremely poor example of ‘‘effective’’
EU implementation, it illustrates that the EU has more powerful tools at its disposal
than OSPAR and the INSCs when states do not comply. In October 1997, EU Envi-
ronment Commissioner Ritt Bjerregaard made a strong plea for better implementa-
tion of the Nitrates Directive. The EU Parliament issued a resolution in late 1998 in
which it said it was ‘‘shocked by the lack of progress’’ in implementing the nitrate
law and called for action from governments, the Commission, and farmers.7 In
2000, thirteen out of fifteen member states were facing legal proceedings in accor-
dance with the EU infringement procedure (table 5.2), which consists of three for-
mal stages. First, the Commission initiates a proceeding through a letter of ‘‘formal
notice.’’ Second, the Commission’s legal elaboration takes place through a ‘‘rea-
soned opinion’’ as a final warning. Third, the Commission refers a case to the ECJ
for a final decision (Tallberg 2002).
Again, the example of the United Kingdom illustrates that swift enforcement
action by the European Commission and the ECJ may indeed work. The United
Table 5.2
Status of Nitrates Directive infringement actions: stage of most advanced action as of April
2000
EU member state Status of Nitrates Directive infringement action
Austria Reasoned opinion
Belgium European Court
Denmark No action
Finland Formal notice
France European Court
Germany European Court
Greece European Court
Ireland Reasoned opinion
Italy European Court (Condemned 2001)
Luxembourg European Court
Netherlands European Court
Portugal Formal notice
Spain European Court
Sweden No action
United Kingdom European Court (Condemned 2000)
Source: ENDS, 737, 5 April 2000. ‘‘More EU countries in trouble over nitrates law’’. Bold
entries indicate North Sea states.
Kingdom was condemned in the European Court of Justice in December 2000.

Whereas the Directive requires member states to identify all surface or groundwaters
polluted by nitrates or at risk of being so, and to designate all such areas as so-called
vulnerable zones enjoying particular protection, the United Kingdom had only iden-
tified surface and groundwaters used as sources of drinking water. The United King-
dom announced steps to broaden its definition to substantially increase the area of
land designated as nitrate vulnerable.8 However, in October 2001 the Commission
announced a repeat legal action stating that the United Kingdom still had too few
areas designated. In December, the environment ministry proposed measures to
bring England and Wales into compliance with the Nitrates Directive.9
EU members have been pulled before the court for different reasons. Germany,
for example, had prepared an action plan as required by the law, but the plan was
considered ‘‘insufficient’’ in its provisions for allowable storage capacities of live-
stock manure and failure to calculate maximum fertilizer application rates.10 During
2003, the Commission continued to take action over bad application of the Direc-
tive by a number of member states. In some cases, the Commission had to open
infringement procedures in order to make member states comply with earlier judg-
ments by the Court (European Commission 2004). Time will show whether these
actions prove sufficient for the North Sea ministers to conclude that the nitrogen tar-
get has been met at the upcoming North Sea Conferences.
Conclusion
Bureaucrats and lawyers defending their respective countries in the court cases re-
lated to the Nitrates Directive are probably not aware that the present problems
can be traced back to a ray of optimism among eight North Sea environmental min-
isters in 1987. Germany’s appearance in the ECJ over the Nitrates Directive can
actually be traced back to the ‘‘German’’ precautionary principle hinted at in the
first North Sea Declaration of 1984.
The interactive workings of the International North Sea Conferences, OSPAR,
and the EU have proven synergistic in two ways. First, the political ‘‘soft-law’’
INSCs have speeded up decision making within OSPAR and the EU by Interaction
through Commitment. Second, OSPAR and the EU have facilitated domestic imple-
mentation of the INSC Declarations through their institutional authoritativeness
and enforcement competence by means of Behavioral Interaction.
The most robust finding in this study is that overlap between institutions does not
necessarily imply duplication of work and low effectiveness. In essence, institutional
differences between the INSCs, OSPAR, and the EU account, at least partly, for the
progress in implementation witnessed. Cooperation on Northeast Atlantic environ-
mental management shows that different types of institutions can fulfill different
functions, all of which are needed to make international environmental cooperation
effective. Moreover, it would be difficult to fulfill these functions within one and the
same institution due to conflicting institutional requirements. For example, INSC
participants had to make political decisions within a couple of days every few years,
which were supposed to take immediate effect at the governmental level. This
secured swift action, but created political vulnerability since governments and polit-
ical priorities change. In contrast, even though the EU is in the process of expanding
its array of environmental policy instruments, EU directives are developed through
lengthy legislative processes and have to be transposed into national law. Most en-
vironmental directives now allow three years for their transposition, although many
are transposed late. In the case of OSPAR, the adoption of new legal instruments
can be even more protracted. For example, a new protocol on incineration at sea
was signed in 1983 but did not come into force until 1989.
OSPAR and the EU could not match the decision-making speed of the INSCs,
but were needed to keep up the pressure on implementation and compliance. Due
to its legal infringement procedure, the EU possesses more power to act when the
going gets tough. EU directives and regulations are more commanding than INSC
or OSPAR commitments due to the ‘‘supranational’’ nature of the EU. EU action
on the 1991 Nitrates Directive shows that EU enforcement tools in cases of noncom-
pliance are significantly more powerful than those ‘‘traditional’’ regimes possess.
The legally binding OSPAR also provided a legal and stable basis for the INSCs
and gave authoritative force in the crucial implementation phase. Such qualities
proved particularly important concerning dumping at sea where the EU had no
competence.
Rather than introducing new commitments on hazardous substances or eutrophi-
cation, the 2002 Bergen Declaration set out a number of new issue areas such as cli-
mate change, biodiversity, renewable energy, and an integrated ecosystem approach.
The search for new challenges indicates that the INSCs have ‘‘solved’’ most of the
problems related to hazardous substances, nutrients, dumping, and incineration at
sea through international cooperation. Further achievements now depend largely
on domestic political will and the capacity to follow through. The EU will continue
to have an important role to play in facilitating implementation and enforcing com-
pliance in the future.
Notes
1. The term European Union will be used broadly throughout this chapter, to include the
period before the Treaty of Maastricht.
2. Accordingly, OSPAR and INSC can be treated as two separate although closely related
institutions. OSPAR and INSC have separate agendas, and differ in membership as well as in
norms, rules, and communication processes.
3. Instances of less than harmonious collaboration can probably be found at the intersection
with other issue areas, such as EU agricultural policy. Instances involving the INSCs, OSPAR,
and EU directives on hazardous substances are not included. These interactions are probably
not that different from cases involving nutrients, but any interaction is extremely difficult to
pinpoint due to high causal complexity.
4. Important for the North Sea was that the Rhine Commission, motivated in part by the
Sandoz accident, adopted the Rhine Action Program in 1987. This program addressed a num-
ber of issues, including marine issues, and included some institutional strengthening to aid
compliance, thus ensuring continued action after public interest declined.
5. ‘‘EU Commission Back on the Compliance Warpath,’’ ENDS, 831, September 13, 2000.
6. ‘‘UK ‘to Escape Bathing Water Fines,’ ’’ ENDS, 1074, October 4, 2001.
7. ‘‘EU Parliament ‘Shocked’ by Nitrate Law Delays,’’ ENDS, 401, October 21, 1998.
8. ‘‘UK Condemned over EU Water Pollution Law,’’ ENDS, 891, December 7, 2000.
9. ‘‘UK Races to Comply with Nitrates Directive,’’ ENDS, 1128, December 20, 2001.
10. ‘‘EU Takes Further Action on Water Law Breaches,’’ ENDS, 341, July 8, 1998.
References
Cameron, James. 1994. The Status of the Precautionary Principle in International Law. In
Timothy O’Riordan and James Cameron, eds., Interpreting the Precautionary Principle. Lon-
don: Earthscan.
EEA. 2001. Environmental Signals 2001. Copenhagen: European Environment Agency.
Ehlers, Peter. 1990. The History of the International North Sea Conferences. In David
Freestone and Ton Ijlstra, eds., The North Sea: Perspectives on Regional Environmental Co-
operation (special issue of the International Journal of Estuarine and Coastal Law), 3–15.
London: Graham & Trotman.
European Commission. 2004. Fifth Annual Survey on the Implementation and Enforcement
of Community Environmental Law, 2003. Brussels: European Commission.
Haigh, Nigel. 2003. Manual of Environmental Policy: The EU and Britain. Leeds: Janey Pub-
lishing and Institute for European Environmental Policy.
Hayward, Peter. 1990. The Oslo and Paris Commissions. In David Freestone and Ton Ijlstra,
eds., The North Sea: Perspectives on Regional Environmental Co-operation (special issue
of the International Journal of Estuarine and Coastal Law), 91–101. London: Graham &
Trotman.
Hofrichter, Jürgen. 1991. Evolution of Attitudes towards Environmental Issues 1974–1991.

Mannheim: Zentrum für Europäische Umfragenanalysen und Studien, Universität Mannheim.
Ijlstra, Ton. 1986. Enforcement of International Environmental Instruments in the North
Sea: The Missing Link. Netherlands Institute for the Law of the Sea (NILOS). Utrecht: Uni-
versity of Utrecht.
Ministry of the Environment. 2002. Ministerial Declaration of the Fifth International Confer-
ence on the Protection of the North Sea. (Bergen Declaration.) Bergen, Norway, March 20–
21. Oslo: Ministry of the Environment.
Ministry of Environment and Energy. 1995a. Esbjerg Declaration. Ministerial Conference on
the Protection of the North Sea, Esbjerg, June 8–9. Copenhagen: Ministry of Environment
and Energy, Danish Environmental Protection Agency.
Ministry of Environment and Energy. 1995b. Ministerial Declarations. International Confer-
ence on the Protection of the North Sea. Bremen, 1984, London, 1987, The Hague, 1990.
Copenhagen: Ministry of Environment and Energy, Danish Environmental Protection
Agency.
Ministry of Environment and Energy. 1995c. Progress Report. Fourth Ministerial Conference
on the Protection of the North Sea, Esbjerg, June 8–9. Copenhagen: Ministry of Environment
and Energy, Danish Environmental Protection Agency.
Nollkaemper, Andrè. 1993. The Legal Regime for Transboundary Water Pollution: Between
Discretion and Constraint. London: Graham & Trotman.
Pallemaerts, Marc. 1992. The North Sea Ministerial Declarations from Bremen to Hague:
Does the Process Generate Any Substance? International Journal of Estuarine and Coastal
Law 7 (1): 1–26.
Prat, Jean-Luc. 1990. The Role and Activities of the European Communities in the Protection
and the Preservation of the Marine Environment of the North Sea. In David Freestone and
Ton Ijlstra, eds., The North Sea: Perspectives on Regional Environmental Co-operation (spe-
cial issue of the International Journal of Estuarine and Coastal Law), 101–111. London: Gra-
ham & Trotman.
Richardson, Jeremy. 1994. EU Water Policy: Uncertain Agendas, Shifting Networks and
Complex Coalitions. Environmental Politics 3 (4): 139–167.
Rosendal, G. Kristin. 2001. Impacts of Overlapping International Regimes: The Case of Bio-
Skjærseth, Jon Birger. 1999. The Making and Implementation of North Sea Pollution Com-
mitments: Institutions, Rationality and Norms. Oslo: University of Oslo and Akademika, AS.
Skjærseth, Jon Birger. 2000. North Sea Cooperation: Linking International and Domestic
Pollution Control. Manchester: Manchester University Press.
Skjærseth, Jon Birger. 2001. The International Regime on the Protection of the North-East
Atlantic: A Good Neighbour. Oslo: Fridtjof Nansen Institute.
Skjærseth, Jon Birger. 2002a. Cleaning up the North Sea: The Case of Land-Based Pollution
Control. In Edward L. Miles, Arild Underdal, Steinar Andresen, Jørgen Wettestad, Jon Birger
Skjærseth, and Elaine M. Carlin, eds., Environmental Regime Effectiveness: Confronting
Theory with Evidence, 175–197. Cambridge: MIT Press.
Skjærseth, Jon Birger. 2002b. Towards the End of Dumping in the North Sea: The Case of the
Oslo Commission. In Edward L. Miles, Arild Underdal, Steinar Andresen, Jørgen Wettestad,
Jon Birger Skjærseth, and Elaine M. Carlin, eds., Environmental Regime Effectiveness: Con-
fronting Theory with Evidence, 65–87. Cambridge, MA: MIT Press.
Skjærseth, Jon Birger, and Jørgen Wettestad. 2002. Understanding the Effectiveness of EU En-
vironmental Policy: How Can Regime Analysis Contribute? Environmental Politics 11 (3):
99–120.
Tallberg, Jonas. 2002. Paths to Compliance: Enforcement, Management and the European
Union. International Organization 56 (3): 609–643.
Underdal, Arild. 1991. Solving Collective Problems: Notes on Three Modes of Leadership. In
Challenges of a Changing World. Festschrift for Willy Østreng. Lysaker, Norway: Fridtjof
Nansen Institute.
Weiss, Edith Brown. 1993. International Environmental Law: Contemporary Issues and the
Emergence of a New World Order. Georgetown Law Journal 81 (3): 675–710.
Young, Oran. 1991. Political Leadership and Regime Formation: On the Development of
Institutions in International Society. International Organization 45 (3): 281–308.
6
Institutional Interplay and Responsible Fisheries:
Combating Subsidies, Developing Precaution
Olav Schram Stokke and Clare Coffey
The world’s oceans have struggled for decades to sustain large-scale and increas-
ingly invasive fishing activities. The response of governments to this situation
has ranged from neglect to the establishment of regional institutions for concerted
fisheries management. Where international institutions exist, they have often pro-
vided a mechanism for allocating fishing rights among members, and in some cases
also cooperative means for scientific activities and compliance control (Stokke 2001).
The effectiveness of such regional regimes varies considerably, however, and the
overall performance of the global fisheries regime leaves much to be desired. This
is partly because the underlying problem being addressed has regained much of the
severity it had prior to the introduction in the mid-1970s of exclusive economic
zones. Despite almost universal claims by coastal states to 200-mile fishing zones,
and the concomitant ‘‘nationalization’’ of some 90 percent of the world’s commer-
cial fisheries, coastal states have largely failed to manage their resources sustainably.
At the same time, distant-water fishing fleets have further enhanced the harvesting
capacity they deploy on the high seas and in foreign fishing zones. Pushed by capital
and modernization subsidies, rising competition, higher operating costs, and steadily
lower-value yields, fishing companies have introduced a range of new technologies
enabling them to profit from fish located in concentrations or at depths that would
previously have been beyond economic or technical reach. While technology and
harvesting capacity have increased, for most oceans, catches are now well below his-
toric peak levels (FAO 2001b). Importantly, the process has also involved the grad-
ual decline of larger, long-lived, and more valuable predator species such as tuna,
cod, and haddock in the oceans—known as ‘‘fishing down marine food webs’’—
with significant ramifications for marine ecosystems (Pauly et al. 2002).
Efforts to strengthen international fisheries rules, regional and global, during
the past decade occurred in response to these challenges. The main elements of the
128 Olav Schram Stokke and Clare Coffey
present global fisheries regime are introduced in the next section. This is followed
by a brief overview of important cases of interaction between the global fisheries
regime and other international institutions as well as the EU Common Fisheries
Policy. The chapter then focuses on the interaction between the global fisheries re-
gime and (1) the global trade regime with regard to rules on fisheries subsidies; (2)
the International Council for the Exploration of the Sea (ICES) with respect to the
formulation of scientific advice on fisheries management, notably the implementa-
tion of the precautionary approach; and in turn, (3) the interaction between ICES
precautionary advice and the EU Common Fisheries Policy. Our focus on subsidies
permits examination of cross-issue institutional interplay, whereas the precautionary
approach connects activities within institutions that focus on different aspects of re-
source management: research and decision making. Both themes are high up on the
political agenda of international environmental governance. Following a brief pre-
sentation of the broader issues involved in each case of interaction, assessments are
provided of the causal relationship between source and target institutions and the
adequacy of the policy responses to the set of interactions. In the last section, we
draw conclusions regarding important factors that can help explain the emergence
of institutional interaction and its impact on the effectiveness of the institutions
involved.
The Global Regime for Fisheries
The 1990s were highly dynamic as regards international fisheries rules. The 1995
UN Fish Stocks Agreement1 specifies the 1982 UN Convention on the Law of the
Sea with regard to straddling stocks and highly migratory stocks and influences de-
cision making within institutions that are narrower in geographic or functional
terms. It strengthens and specifies the duty under international law to cooperate on
all aspects of high-seas fisheries management. It provides that only states that are
members of, or adhere to, regional regimes shall have access to the fishery (Art. 8)
and elaborates certain basic conservation principles, including the precautionary
approach to fisheries management (Art. 6, Annex II). Although the principal focus of
the Fish Stocks Agreement is on high seas fisheries management, Article 7 requires
that national measures concerning straddling stocks and highly migratory stocks be
compatible with high seas measures, and Article 3 requires that the provisions on
precautionary management also apply in national waters. As regards compliance
control, the Agreement breaks new ground by creating global minimum standards
Institutional Interplay and Responsible Fisheries 129
that permit a broader range of compliance mechanisms than was previously the
norm within regional high seas management regimes. This includes strengthened
flag-state responsibilities; procedures for non-flag-state inspection, detention, and
arrest on the high seas; and elaboration of certain port-state measures to enhance
adherence to regional conservation and management measures (Art. 19–23). The
Fish Stocks Agreement entered into force in 2001 and forms, in conjunction with
the UN Convention on the Law of the Sea that came into force in 1994, the basis
for the global fisheries regime.
In parallel to the Fish Stocks Agreement, a set of international instruments was
negotiated under the auspices of the UN Food and Agriculture Organization (FAO).
The FAO Compliance Agreement, which strengthens flag-state responsibilities with
respect to vessels fishing on the high seas, was adopted in 1993. The FAO Code of
Conduct for Responsible Fisheries was agreed to in 1995 and has been followed by
four International Plans of Action (IPOAs) on fishing capacity, shark management,
seabird protection, and illegal, unreported, and unregulated fishing.2 The FAO Code
of Conduct and its plans of action are voluntary, directed at members as well as
nonmembers of the FAO, and reflect the active role of the FAO Committee on Fish-
eries in seeking to shape and support international fisheries rules. Alongside the UN
General Assembly, which annually reviews progress under the UN Convention on
the Law of the Sea, the FAO Committee on Fisheries is the only permanent interna-
tional forum that, periodically and on a worldwide basis, examines major fisheries
concerns and provides recommendations to governments, regional management
bodies, and other stakeholders.
Institutional Interaction and International Fisheries Management
We have provided an inventory of interactions between the global fisheries regime,

the EU Common Fisheries Policy (CFP), and other institutions, covering the full
functional scope of resource management: science, regulation, and compliance
control (Stokke and Coffey 2001). An overview of these interactions is given in
table 6.1.
All three causal mechanisms investigated in this volume are represented in the in-
ventory. The cases involving the precautionary approach (ICES and EU CFP), fish-
eries subsidies (World Trade Organization, the global fisheries regime, and EU), and
the transparency of decision making (NEAFC) revolve around Interaction through
Commitment. Here, the fact that states have assumed commitments in the context
Table 6.1
Cases of interaction of the Global Fisheries Regime
Scientific research and advice
International Council for the Has implemented the precautionary approach following its
Exploration of the Sea formal adoption in the global Fish Stocks Agreement (FSA)
(ICES) Introduced greater safety margins in its catch
recommendations and requested rapid recovery of troubled

stocks and thus contributed to improved conservation in
line with FSA
Is modifying its communication and terminology in
response to criticism by FAO and regional regimes

Regulation
North-East Atlantic Fisheries Responded to the transparency rules of the FSA by
Commission (NEAFC) adopting new provisions on access to meetings and reports
EU Common Fisheries Policy FSA supports implementation of the precautionary
(CFP) approach in the CFP (through ICES), which requires
preagreed decision rules within recovery/management plans
FSA rules support inclusion of management considerations
in EU fish-import regulations
FAO International Program of Action on fishing capacity
(IPOA-Capacity) contributes to reform of related EU rules

aiming at a reduction of subsidies
IPOA-Capacity contributes to reduction of fisheries
subsidies in the EU and its member states

EU fisheries subsidies still undermine conservation and
management of fish stocks required by global fisheries regime

World Trade Organization WTO subsidies rules improve information flows and
(WTO) bindingness and may therefore help reduce or limit fisheries
subsidies and capacity
The global fisheries regime helped place fisheries subsidies
on the formal WTO agenda (Doha Round)

Convention on International Has contributed to emergence of FAO International
Trade in Endangered Species Program of Action on the Conservation and Management
(CITES) of Sharks (IPOA-Sharks)
Benefited in its implementation from improved national
reporting procedures and records of shark catches resulting

from IPOA-Sharks
Compliance control
Northwest Atlantic Fisheries Facilitated adoption of FSA provisions on inspection,
Organization detention, and arrest
Central Bering Sea Doughnut Facilitated adoption of FSA provisions on inspection,
Hole Agreement detention, and arrest
of one regime affects decision making under another regime. The FAO-CITES case
displays Cognitive Interaction: the ensuing International Plan of Action on Sharks
Management (IPOA-Sharks) by the FAO did not result from commitments but
from learning that was triggered by CITES activities, notably its direct requests for
FAO inputs. Cognitive Interaction is also evident in the relationship between the
global fisheries regime and ICES practices with respect to the involvement of and
communication with stakeholders. The same is true for the two regional cases on
compliance-control procedures: the more intrusive detention and arrest provisions
stood out as salient in the global negotiations by having been applied successfully
in earlier processes. With respect to most interacting institutions listed in table 6.1,
Behavioral Interaction can also be observed: modified ICES advice has influenced
fisheries management ‘‘on the ground,’’ the EU subsidies rules have undermined sus-
tainable use and management of fish stocks, and FAO’s IPOA-Sharks help the effec-
tive implementation of CITES.
Some of the cases involve institutions with significantly different objectives, which
could imply that the source undermines the target. For instance, the EU’s fisheries
subsidies regime has historically aimed at industrial development rather than sustain-
ability pursued under the global fisheries regime. But whereas much of the debate on
institutional interplay has focused on potentially disruptive effects, most of our cases
are synergistic or have at least led to responses that reduced the level of disruption.
Nevertheless, they also typically reveal ample room for further improvement.
While the general approach in this book is to identify certain dyads of institu-
tional interaction, it is important to also consider the effects of other relevant insti-
tutions and processes when tracing the causal connection between source and target.
Reduction of fisheries subsidies, for instance, is an objective of work undertaken
also by the Organization for Economic Cooperation and Development (OECD),
the Asia-Pacific Economic Cooperation (APEC), and the UN Commission on Sus-
tainable Development. Similarly, in the European context, the North Sea Confer-
ences intervened in the translation of the precautionary approach to fisheries from
a global principle to regional scientific and policy approaches. Civil-society activities
around these forums can also serve to link the source or target institutions and
thereby influence their response action.
Capacity Control and Checks on Subsidies
This section reviews the interaction between the global fisheries regime and provi-
sions of the World Trade Organization (WTO) relevant to fisheries subsidies, with
an emphasis on the 2001 decision to place this issue on the agenda for the Doha
Round of multilateral trade negotiations.
Figures on the amount of subsidies provided to the fisheries sector vary widely, a
reflection partly of scattered knowledge and partly of different definitions or opera-
tionalizations of accepted definitions (Milazzo 1998; Stone 1997). Recent estimates
suggest a level somewhere in the range of U.S.$7–14 billion each year (Ruckes
2000). Fisheries subsidies are politically contested. On the one hand, governments
have a number of worthwhile reasons for providing them, including employment in
shipbuilding, harvesting, or processing sectors, food security, or protection of settle-
ments in sparsely inhabited or economically disadvantaged coastal regions.3 Also,
not all government financial transfers for fishing vessels or equipment are problem-
atic from a capacity perspective. For instance, investments that improve the effi-
ciency of vessel operations may be neutral in capacity terms if combined with
buyback or scrapping schemes. On the other hand, too many vessels chasing too
few fish is a fundamental impediment to responsible harvesting. Subsidies can be
an important factor in generating excessive fishing capacity, especially where man-
agement policies are unsatisfactory (Hannesson 2001, 17–19), including in many
high-seas areas and developing-country zones harboring distant-water fishing activ-
ity.4 In recent years, a number of states have reduced their financial contributions to
the fisheries industry (Gréboval 2000), but subsidy reduction remains an important
means of controlling the buildup of vessel capacity.
Part of the Problem: Inadequate WTO Subsidies Rules

Whereas the global fisheries regime aims to ensure the sustainable use of fish stocks,
the objective of the global trade regime is first and foremost to remove restrictions
on international trade between its members (see also chapter 8). Potential trade dis-
tortion associated with fisheries subsidies was therefore the main rationale for rais-
ing the issue within trade-oriented international institutions in the 1960s and 1970s,
especially the OECD and WTO’s predecessor—the General Agreement on Tariffs
and Trade (GATT) (Steenblik 1999). According to the preamble of the agreement
establishing the WTO, however, the WTO is also to allow for ‘‘the optimal use of
the world’s resources in accordance with the objective of sustainable development,’’
and much of the WTO debate on fisheries subsidies has therefore considered envi-
ronmental as well as trade aspects.
The global trade regime, specifically the 1994 Subsidies and Countervailing Mea-
sures (SCM) Agreement negotiated under the WTO, provides rules on subsidies that
are detailed, legally binding, and supported by an elaborate compliance system that
includes compulsory and binding dispute-settlement procedures and authorization
of countervailing trade sanctions. In spite of this no fisheries subsidy has so far
been challenged under WTO rules.
Notifications to the WTO of fisheries subsidies—an obligation under Article 25
of the SCM Agreement—have been very limited in terms of the amount of subsidies
reported, the range of subsidies covered, and the quality of information provided
(Schorr 1998, 154–155). Part of the reason is that several key concepts in the SCM
Agreement are defined in ways that make it difficult to determine whether govern-
ment expenditures and other interventions in the fisheries sector fall within the
domain of the agreement (Stone 1997). The definition of a subsidy as a ‘‘financial
contribution’’ has generated a lively debate on whether public investment in fisheries
management and enforcement should count as subsidies. Research, monitoring, and
control activities make up a significant proportion of government financial trans-
fers to the fisheries sector worldwide (OECD 2000). Similar questions arise where
governments fail to charge for access to resources or indeed purchase access to
resources in foreign exclusive economic zones on behalf of the fishing industry.
Furthermore, only ‘‘specific’’ subsidies (i.e., those limited to an enterprise, industry,
or region) are covered by the agreement, which makes it unclear whether govern-
ment provision of important infrastructure, such as quays and lighthouses, should
be notified.
Among subsidies that are to be reported, only those contingent on export perfor-
mance or the use of domestic rather than imported goods are prohibited. Other sub-
sidies are actionable under the SCM Agreement only if they can be shown to have
adverse effects on the interests of another party.5
Accordingly, only a limited subset of direct or indirect financial transfers to the
fisheries industry is clearly disciplined under present rules, and conceptual unclarity
contributes to a lack of information on the extent, nature, and objective of subsidies.
Many states have considered this situation inadequate and have requested clarifica-
tion of which part of a large gray area should be disciplined under WTO rules. Key
members of the so-called Friends of Fish group of countries pressing for reform of
fisheries subsidies rules in the WTO (Australia, Chile, Ecuador, Iceland, New Zea-
land, Peru, the Philippines, and the United States), especially the United States and
New Zealand, have a long track record of trying to strengthen international restric-
tions on subsidies in primary industries (Steenblik 1999). On fisheries subsidies, they
have been heavily supported by transnational environmental organizations, especially
the World Wide Fund for Nature (WWF), which sought to harness the free-trade
agenda in the interest of conservation. Compared to their most outspoken oppo-
nents on the fisheries subsidies issue, including Japan, the Republic of Korea, and
the European Union, the Friends of Fish countries have had relatively low levels of
fisheries subsidies and would therefore be less affected by stronger rules.
Cognitive Interaction: Programmatic Work under the Global Fisheries Regime

A FAO study published more than a decade ago (FAO 1992) is widely seen as cen-
tral to the comeback of fisheries subsidies on the international diplomatic agenda,
after a decade of moderate interest (Stone 1997; Steenblik 1999). The new twist
introduced by this report was to emphasize much more than before the linkage be-
tween subsidies and problems of resource sustainability. While subsequent examina-
tions by international organizations such as the World Bank (Milazzo 1998), OECD
(1997, 2000), and APEC (PricewaterhouseCoopers 2000) have concluded that
the U.S.$54 billion estimate of fisheries subsidies—or 70 percent of world catch
value—that many read into the FAO report was much too high,6 they have never-
theless documented that current levels of subsidies are substantial and a cause of
overcapacity in the fisheries sector.
In 1998, at a time when the Friends of Fish group was pushing for more work on
fisheries within the WTO and other international institutions, the Sub-Committee
on Fish Trade of the FAO’s Committee on Fisheries decided that the FAO had a
role to play in compiling and disseminating information on the impacts of subsidies
(FAO 1998a, para. 17). It is notable that from the outset, the FAO opted for a sup-
portive and complementary role to other international agencies addressing fisheries
subsidies, such as the WTO and OECD, and was careful to avoid any overlap with
activities taking place there (see in particular FAO, 1998a, paras. 6 and 7). This
stance was not favorable to those, including the EU and the Republic of Korea,
who objected to the WTO taking a more prominent role in international governance
of fisheries subsidies.7
In its International Plan of Action for the Management of Fishing Capacity
(IPOA-Capacity), the FAO pledged to ‘‘collect all relevant information and data
which might serve as a basis for further analysis aimed at identifying . . . subsidies
which contribute to overcapacity’’ (Art. 45). The organization set out to prepare
technical guidelines for the management of fishing capacity and organized a series
of technical and expert consultations on issues related to capacity. A Task Force on
Fishery Subsidies was established in the FAO Fisheries Department to serve as a
focal point in subsidy matters, monitor and review discussions and contributions
from other organizations, and map the various forms of fisheries subsidies.
While a first FAO expert consultation held in 2000 failed to contribute substan-
tially to the debate on whether and how WTO disciplines should be reformed, later
efforts proved somewhat more successful. The report of the Expert Consultation on
Economic Incentives and Responsible Fisheries was criticized by some within the
FAO Committee on Fisheries as having ‘‘raised more questions than answers,’’ and
the Committee decided that a second consultation on the issue would comprise a
broader range of experts with relevant practical and multidisciplinary experience in
fisheries-management and trade issues.8 At least the expert consultation identified
forms of government transfers to be prioritized in research, which were all compat-
ible with the conventional definition of subsidies espoused by the WTO: capital ex-
pansion such as vessel purchase or modernization grants, tax waivers and deferrals,
and price support.9 Nevertheless, the inadequacy of information was presented as a
key conclusion of the consultation in the WTO, and there is no indication that the
subsequent discussion on fisheries subsidies in the WTO Committee on Trade and
Environment (CTE) paid much attention to this particular input.10 In the subse-
quent years, however, the FAO organized information-exchange meetings with
other international agencies with ongoing work programs on fisheries subsidies,
including the OECD, the United Nations Environment Program (UNEP), and the
WTO (FAO 2001a, 2002).
On balance, programmatic efforts under the global fisheries regime have to some
extent, but hardly decisively, strengthened the hand of those who favor reform of
the WTO subsidies regime. Throughout the 1990s, both proponents and opponents
of WTO reform have referred to the FAO as a particularly relevant source of in-
formation on the matter.11 By attempting to develop clearer and more consensual
knowledge on which types of subsidies are capacity-driving and which can be
supportive of sustainable fisheries, the FAO sought to render obsolete a key argu-
ment against WTO reform in the area, namely the difficulty of separating ‘‘good’’
from ‘‘bad’’ fisheries subsidies. However, the FAO’s influence may have been
greatest in what it did not do. By not opting for a front-runner position in regula-
tory efforts, as it had with regard to high-seas compliance measures, but instead sup-
porting and facilitating the fisheries subsidies initiative within the global trade
regime, the FAO did not add fuel to those who argued that the WTO was a poor
arena for developing criteria to separate sustainable from nonsustainable fisheries
subsidies.
Interaction through Commitment: Global Fisheries Norms and the Strength of the
Friends of Fish Coalition
The global fisheries regime includes legal and political commitments that are sup-
portive of those in favor of reform of WTO subsidies. Under the Fish Stocks Agree-
ment, coastal states and states whose vessels fish on the high seas are to ‘‘take
measures to prevent or eliminate overfishing and excess fishing capacity and to en-
sure that levels of fishing efforts do not exceed those commensurate with the sustain-
able use of fishery resources’’ (Art. 5). There is no explicit mention of subsidies,
however.
The FAO Code of Conduct encourages states to ensure that ‘‘policies, programs
and practices related to trade in fish and fishery products do not result in . . . envi-
ronmental degradation’’ (Art. 6.14), and that ‘‘excess fishing capacity is avoided and
. . . the economic conditions under which fishing industries operate promote respon-
sible fisheries’’ (Art. 7.2.2). In addition, the Code provides that ‘‘States, aid agencies,
multilateral development banks and other relevant international organizations
should ensure that their policies and practices . . . do not result in environmental
degradation’’ (Art. 11.2.15).
The equally nonbinding FAO IPOA-Capacity calls on states to achieve, ‘‘prefera-
bly by 2003 but not later than 2005, an efficient, equitable and transparent manage-
ment of fishing capacity’’ (Art. 7). If excess capacity is undermining the achievement
of long-term management, states should take measures to limit and progressively re-
duce relevant fishing capacity. National plans for the management of fishing capac-
ity are important means for this objective and shall include assessments of ‘‘all
factors, including subsidies, contributing to overcapacity’’ (Art. 25). Finally, ‘‘States
should reduce and progressively eliminate all factors, including subsidies . . . which
contribute, directly or indirectly, to the build-up of excessive fishing capacity there-
by undermining the sustainability of marine living resources’’ (Art. 26).
It is difficult to measure the influence of these fisheries-regime provisions on the
process of regulating subsidies within the WTO. It is indicative of such influence,
however, that those in favor of reforming the trade rules on subsidies have consis-
tently emphasized the existence of global fisheries norms in this area and especially
the Fish Stocks Agreement and the IPOA-Capacity.12 The latter was presented to the
WTO Committee on Trade and Environment (CTE) in June 1999, at a time when
support for reform of fisheries-subsidies rules was growing in the Committee. FAO
was invited to report to the next CTE meeting ‘‘on the main elements of an indica-
tive work program aimed at addressing the impact of subsidies and other factors
which contribute to overcapacity and unsustainability in fisheries.’’13 Since then,

FAO representatives have regularly attended CTE meetings, providing updates on
FAO activities related to fisheries subsidies.14
The Outcome: Fisheries Subsidies at the Forefront of Global Negotiations

The 2001 Doha Ministerial Declaration, which provides the mandate for the new
round of trade negotiations, states that ‘‘participants shall also aim to clarify and
improve WTO disciplines on fisheries subsidies, taking into account the importance
of this sector to developing countries.’’ The Declaration also highlights fisheries sub-
sidies when sketching the broader trade-environment agenda.15
Programmatic work under the global fisheries regime and the latter’s norms con-
tributed to this outcome by making it more difficult for opponents to reject it. Right
from the outset, discussions within the WTO had focused on the need for more in-
formation on the impact of subsidies on sustainability and trade, and we have seen
that activities under the global fisheries regime were significant in this regard.16 By
summer 1999, following a high-level WTO symposium on trade and development,17
states favoring fisheries-subsidy reform presented several substantial papers at a
CTE meeting. These papers pointed to the ‘‘loss-loss-loss’’ relationship between fish-
eries subsidies on the one hand, and trade, environment, and development on the
other,18 and so support for placing the issue on the Doha Round agenda grew. Al-
ready at the WTO Seattle Ministerial Conference in late 1999, the EU as one major
opponent accepted that fisheries subsidies would be addressed in the next world
trade round. In Doha, the EU finally also agreed that the discussions could take
place under the SCM Agreement (rather than as part of the environment agenda).
More important than the interaction with the global fisheries regime, however,
was probably that fisheries subsidies increasingly became a development issue. Early
on in the negotiations, fisheries subsidies had been framed primarily as a trade and
environmental-sustainability issue addressed, since 1997, within the CTE (Porter
2002). However, development issues gained in prominence and proved decisive in
moving the subsidies issue onto the main negotiating agenda in 2001.19 From a
North-South perspective, fishing interests in developing countries had to compete
at sea and on international markets with heavily subsidized counterparts from
wealthier nations. There is not much to suggest that norms or activities undertaken
under the global fisheries regime played any important role in the mobilization of
developing countries. The work of UNEP was probably more salient in this regard.
Combining research work, country studies, and policy dialogue, it convened a
number of expert meetings or consultations with key officials and international insti-
tutions with particular attention to the impact of fisheries subsidies on developing
countries—notably in relation to third-country fishing agreements. Several leading
civil-society organizations, including WWF, have also emphasized the development
aspect of fisheries subsidies.
Summary and Assessment

Fisheries subsidies surfaced on the Doha Round agenda because the interests of a
coalition of countries with relatively small domestic subsidies programs coincided
with the interests of developing fishing nations, as well as with those of environmen-
tal organizations in industrialized countries. As summarized in figure 6.1, this sec-
tion has presented two cases of interaction.
The first causal impact of the global fisheries regime on WTO subsidies rules is
cognitive in nature: the FAO has raised concern over the issue and, along with a
few other international institutions, ensured for supporters and opponents of sub-
sidies reform alike a common source of legitimate expertise on fisheries subsidies.
The second case of interaction between the global fisheries regime and the subsidy
segment of the WTO regime is one of commitment. The subsidy-related provisions
in the Fish Stocks Agreement, the FAO Code of Conduct, and the FAO IPOA-
Figure 6.1
Global Fisheries Regime affects World Trade Round
Capacity supported those in the WTO who argued that fisheries subsidies are a bar-
rier to sustainability, as well as a barrier to free trade. The provisions therefore
played a part in bringing this particular issue onto the agenda for the Doha Round
of global trade negotiations. Moreover, attempts to define the WTO as peripheral to
the governance of fisheries subsidies and to place this issue within a forum with
more fisheries expertise but less ‘‘clout’’—that is, the FAO—were rejected. While
the FAO has a well-established practice of mobilizing relevant fisheries expertise in
the development of consensual guidelines on implementation at national levels, FAO
rules, unlike WTO provisions, tend to be soft and are rarely backed up by intrusive
enforcement mechanisms.
The instances of interaction described here have occurred at the level of institu-
tional output, although this does not rule out effects at the levels of outcome (sub-
sidies practices) or impact (on harvesting capacity). Both within the source and the
target regime, the interaction has been pursued intentionally. Appreciating the sig-
nificance that trade rules might assume with regard to controlling fisheries subsidies,
the FAO sought to coordinate fact-finding and analytic activities relevant to the
question of how subsidies influence sustainability. That said, in both regimes many
of the activities relevant to fisheries subsidies were generated by internal processes
and priorities, and by external pressures such as that from environmental organiza-
tions, rather than from the other regime.
Responses to the interaction have been of several kinds, although to date neither
of the institutions has adapted its rules. There has been moderate interinstitutional
coordination: the WTO invited the FAO to provide more specific information on
available knowledge about the relationship between fisheries subsidies and sustain-
ability problems. The FAO input provided in response has largely supported those
questioning fisheries subsidies within the WTO, but it has failed to provide more
than preliminary answers. Since the issue was highly controversial and FAO work
is based on consensus, this is not surprising and it is doubtful whether more cross-
institutional coordination could have changed the situation substantially. Adapta-
tion to rising international attention to environmentally harmful fisheries subsidies
is reflected in the fact that many governments chose to reduce their subsidies pro-
grams during the 1990s (Gréboval 2000).
Precaution, ICES, and the Common Fisheries Policy
The UN Fish Stocks Agreement contains one of the most explicit definitions pro-
vided in an international treaty of how the precautionary principle should apply in
practice. Under Article 6, and applying the guidelines set out in Annex II, states are
to decide on two types of precautionary reference points. First, a ‘‘limit reference
point’’ is associated with danger: if a stock falls below this level, preagreed conser-
vation and management action should be initiated to support stock recovery. The
overall aim of management strategies is to ensure that the risk of exceeding a limit
reference point is very low. Second, with respect to ‘‘target reference points,’’ man-
agement strategies should ensure that these are not, on average, exceeded.
Based on a FAO Technical Consultation that involved fisheries experts from a
wide range of organizations, a set of guidelines for implementing the precautionary
approach was developed. These emphasized (1) the development of operational and
measurable targets and constraints—that is, biological reference points that should
(2) reflect management objectives based not only on biological but also on socioeco-
nomic considerations and adequate stakeholder involvement, including the involve-
ment of fishing-industry and conservation groups, and be accompanied by (3)
preagreed decision rules that define what action should be taken when reference
points are exceeded (FAO 1996, especially paras. 20–35).
The following assessment of how the global fisheries regime has influenced the
emergence of a precautionary approach to fisheries management in the Northeast
Atlantic is related to these guidelines. The first two subsections examine how the
global fisheries regime has interacted with, respectively, ICES’ development and
communication of biological reference points. The third subsection analyzes the sig-
nificance of ICES for the emergence of preagreed decision rules in the EU Common
Fisheries Policy.
Interaction through Commitment: Biological Reference Points

Founded in 1902 under the Convention for the International Council for the Explo-
ration of the Sea (last amended in 1964), ICES is an intergovernmental organization
that coordinates scientific research and provides advice on fisheries management in
the Northeast Atlantic. While the Fish Stocks Agreement introduces general provi-
sions regarding the formulation and use of scientific advice at the global level, ICES
has developed and implemented a specific procedure for generating advice, tailored
to the relatively advanced level of knowledge about the various Northeast Atlantic
fish stocks. ICES advice is not binding on governments and its influence on decision
making depends largely on the organization’s reputation for scientific excellence
and neutrality toward the often competing claims of various stakeholders—which
is widely perceived as high (Gullestad 1998). Most ICES parties are politically,
and increasingly also legally, bound by the Fish Stocks Agreement and this is
an important driver of the interaction between the global fisheries regime and
ICES.
The precautionary approach to fisheries management was not among the
‘‘themes’’ addressed at the ICES Annual Science Conferences in the years preceding
the adoption of the Fish Stocks Agreement and the FAO Code of Conduct, but it
has been featured repeatedly since 1997.20 A Study Group on the Precautionary
Approach to Fisheries Management was convened that year, charged with develop-
ing a form of advice that would be consistent with the precautionary approach as
elaborated under the global fisheries regime.
Responding to the recommendations of the study group, the ICES Advisory Com-
mittee on Fisheries Management (ACFM) has since 1998 formulated its advice to
management agencies within a framework for implementation of the precautionary
approach (ICES 2000, 2001a). The advice is based on an estimate of current
stock status and usually occurs in the form of catch options that should maintain
the stock status inside ‘‘safe biological limits.’’ The latter term was introduced as
early as 1981 and refers to the level of the spawning stock below which there is an
unacceptable probability that recruitment will be impaired (see ICES 2000, 4; 1992,
5–6). Similarly, the precautionary management strategy—or the maximum harvest-
ing level that ICES would consider as precautionary—is identified by calculating a
buffer that generates a very low probability of reaching a stock level at which
recruitment will be impaired. The latter is ICES’ operationalization of a limit (dan-
ger) reference point. The size of the buffer depends on the natural variability of the
stock, the precision of the assessment, and the risk that management agencies are
willing to accept (ICES 2000, 2).
Two specific changes are associated with the implementation of the precautionary
approach: clearer articulation of the impairment risk that ICES considered acceptable
(for most stocks set as low as 5 percent), and greater commitment to advise forceful
rebuilding plans when stocks are depleted or overfished (compare ICES 2001b, iv
with ICES 1992, 12). Since ICES used the same actual reference points after its in-
troduction of a precautionary approach, one observer has stated that in effect, ‘‘the
earlier management regime . . . often remained unchanged, although the language
has been given a precautionary gloss’’ (MacGarvin 2002, 20). That statement would
surprise many nonscientist stakeholders in the region, and it underestimates the sig-
nificance of the ICES decision to restrict the term precautionary, in cases where
stocks are troubled, to conservation measures believed to rapidly rebuild the stock.
There is much to suggest that the elaboration of a precautionary approach in

the Fish Stocks Agreement has galvanized scientists involved in the ICES advisory
process, with their ‘‘long history of unsuccessfully trying to agree on targets with
management agencies’’ (Hilborn et al. 2001, 100) whenever their biology-based
recommendations face stakeholder complaints. A Russian request in 2000 for a
change in the estimate of a precautionary reference point for Barents Sea cod, for
instance, was rejected by ICES on the grounds that the data ‘‘available at present
give no firm basis for revision of reference points’’ (ICES 2001b, 12). There is broad
agreement that ICES has taken an early lead among regional scientific advisory
bodies in implementing the precautionary approach (Hilborn et al. 2001, 100), par-
ticularly regarding methodologies for establishing biological reference points
(Garcia 2000, 22), and the concept of precautionary management now permeates
ICES activities.
While the global fisheries regime is not the only institution that has influenced
ICES practices in this area, it has had the most specific effect. For instance, the pre-
cautionary approach to fisheries management was also discussed in depth during the
1997 Intermediate Ministerial Meeting on the Integration of Fisheries and Environ-
mental Issues under the International North Sea Conferences, which involved many
of the key participants in ICES. While this parallel process provided an occasion for
political discussion of the subject and may have bolstered commitment to the pre-
cautionary approach and increased involvement by environmental interests, it did
not advance the specification of the precautionary approach, which has largely
occurred within specialized fisheries forums.21
Cognitive Interaction: Socioeconomic Aspects and Stakeholder Involvement

The objective pursued by ICES in the fisheries area (promotion of research and pro-
vision of scientific advice) is much narrower than that of the global fisheries regime
(long-term conservation and sustainable use of fish stocks). Associated with this nar-
rower functional scope is a significant difference in the mix of expertise mobilized
under these institutions. In ICES fisheries bodies, marine biological expertise rele-
vant to stock assessment and marine ecosystems predominates and the organization
has a long tradition of advancing the research frontier in these particular areas
(Stewart 1991, 2549). Such expertise is also found in the UN fisheries bureaucracy,
primarily the FAO Fisheries Department, but here social scientists, economists, law-
yers, and technologists are also well represented.22 Whereas ICES is widely seen as
being in the forefront of the development of biological reference points, the FAO
(1996, 16–21) has emphasized that the precautionary approach also requires mobi-
lization of socioeconomic, technical, and institutional knowledge.
On the basis of these differences, complaints have been articulated—both within
the FAO and by the regional management regimes ICES serves—that the precau-
tionary procedure generates advice that is insufficiently sensitive to the socioeco-
nomic costs associated with quota cuts. It has been argued from within FAO that,
ideally, limit and target reference points should also be developed for socioeconomic
and institutional impacts of conservation measures (Garcia 2000, 34). In the context
of the EU Common Fisheries Policy, the sudden introduction of precautionary refer-
ence points and calls for recovery plans in the 1998 ICES advice left little time for
discussions with industry before decisions had to be taken, while making it difficult
for managers to gauge the costs and benefits of action (Brown 2000; Deas 2000).
Similarly, in the years following the introduction of ICES advice based on limit and
target reference points, the Norwegian-Russian Fisheries Commission chose to set
quotas well above the level ICES considered precautionary (ICES 2000, sec. 3.1.2).
As these examples demonstrate, complaints have easily translated into lack of accep-
tance of ICES recommendations.
The problem may have been exacerbated by the ways in ICES has communicated
its precautionary advice, which has been criticized as difficult to understand for
other stakeholders. Excessively technical language may reduce the ability of scien-
tists to convey to industry and managers a clear picture of the biological conse-
quences of management proposals, which is usually a requirement for obtaining
their acceptance of costly restrictions on harvesting. Obscure or variant terminology
may even engender suspicions that scientists ‘‘add extra (non transparent) conserva-
tism or precaution into the estimation process’’ (Mace and Gabriel 1999, 69), thus
trespassing into the domain of management agencies. Significant differences between
the various scientific advisory bodies to North Atlantic management regimes in op-
erational definitions of precautionary reference points (ICES 2000, 4–5) have fur-
ther contributed to the problem. For instance, in contrast to the ICES practice of
defining limit (danger) reference points in terms of impaired recruitment and then
adding a buffer, the Northwest Atlantic Fisheries Organization (NAFO) has linked
the limit reference point to the maximum sustainable yield (Garcia 2000, 24).
While ICES has never aspired to cover all relevant aspects of fisheries manage-
ment, it has responded to these complaints. It is ultimately the responsibility of man-
agement agencies to ensure that they are advised by available expertise on the social,
economic, and institutional effects of conservation measures—and not only on the
biological impacts. Nevertheless, ICES has stepped up its efforts to improve commu-
nication with managers and industry representatives, not least about the costs and
benefits of stock recovery plans.23 To achieve a more harmonized terminology,
ICES convened in 2000 an interagency meeting involving the FAO and a number
of North Atlantic management bodies with a view to identifying differences and
similarities in terminology and conceptual definitions, and exploring their conse-
quences for the provision of precautionary scientific advice.24 The ICES Advisory
Committee on Fisheries Management noted that these interagency discussions ‘‘can
be expected to result in further development and clarification of concepts and
changes in terminology’’ (ICES 2001b, v), and has pledged to harmonize its use of
the term limit reference point to that of NAFO (ICES 2000, 5). This reflects a gen-
eral change in attitude within ICES toward the concept of maximum sustainable
yield. Referring to the fact that the Fish Stocks Agreement includes this concept in
its definition of the precautionary approach, ICES now pledges to ‘‘develop a bal-
anced view on how best to interpret this reference point in a fish stock assessment
context.’’25
Hence, inadequate attention to communication issues may have weakened ICES’
role as an ‘‘interlocutor’’ between the global fisheries regime and regional manage-
ment regimes. Related complaints have led to institutional learning within ICES
induced, among other things, by the global fisheries regime. In the context of EU
fisheries policy, some observers report that the communication between scientists
and the industry has improved over time and has helped to convince parts of the
fisheries sector of the benefits of the precautionary approach, not least as a means
of introducing greater stability within the sector (Deas 2000). The adaptations initi-
ated by ICES hold the same promise, although concrete results such as those in the
EU have not yet materialized.
ICES and Precautionary Decision Making in EU’s Common Fisheries Policy

The framework for fisheries management in the EU is set out in Regulation 2371/
2002 on the Conservation and Sustainable Exploitation of Fisheries Resources
under the Common Fisheries Policy (OJ L358, 31.12.2002) and its daughter regula-
tions. Since the 1970s, a variety of instruments have been employed to curb over-
fishing or to prohibit damaging practices, yet a growing number of commercial fish
stocks in European waters are considered to be outside safe biological limits. This is
the case for some two-thirds of commercial stocks in the North Sea and the Medi-
terranean Sea, almost half of those in the Northeast Atlantic, and 20 percent of
commercial stocks in the Baltic Sea (FAO 2001b); highly migratory species of tuna
and swordfish are also overexploited.
The failure of the Common Fisheries Policy (CFP) to manage EU fish stocks is fre-
quently blamed on the central role given to total allowable catch limits (TACs) as a
management tool, and the process for setting these limits (e.g., European Commis-
sion 2000). The Council of Ministers adopts TACs each December, primarily cover-
ing commercial stocks in the Northeast Atlantic and the Baltic Sea. The Council
decision follows a proposal from the European Commission. The proposals are, in
turn, heavily derived from ICES advice. The annual Council negotiations over the
Commission proposals are invariably politically charged, with individual fisheries
ministers coming under considerable pressure to secure the best deal for their fisher-
men at home. Fisheries management is particularly vulnerable to such pressure pre-
cisely because of the uncertainty and ignorance about important bioecological as
well as socioeconomic processes involved in fisheries (Garcia 2000). As a result, as
one academic has aptly put it, ‘‘When the scientific advice has been refracted
through the [EU] political process it may appear to shed little light on the final deci-
sions’’ (Symes 1998, 12).
That said, the EU and several of its member states have been at the forefront of
discussions on the precautionary principle. Since the Maastricht Treaty entered into
force in 1993, Article 174 of the Treaty establishing the European Community
states that Community environmental policy ‘‘shall be based on the precautionary
principle.’’ The EU (legally: the European Community, EC) participated actively in
the negotiations on the UN Fish Stocks Agreement. And the EU is a major ‘‘client’’
of ICES and has benefited from, as well as contributed to, the leading role played by
ICES scientists in elaborating and adapting scientific advice in line with the precau-
tionary approach.
The challenge presented by the precautionary approach in the Fish Stocks Agree-
ment and the FAO guidelines is to develop alternative multiannual management
approaches, where decisions are based on preagreed decision rules. In other words,
the precautionary approach requires adaptations not only in the generation of scien-
tific advice but also in the decision-making process (Hanna 1999; Mace and Gabriel
1999). As Garcia (2000, 40) noted, ‘‘Alone or limited to its scientific aspects, [the
precautionary approach] will only represent yet another step towards scientific
sophistication along a 50-year old track which has produced large amounts of excel-
lent science against a background of inexorably growing overexploitation, eco-
system degradation, economic dysfunction, and social stress.’’
As early as 1992, the CFP had provided for the adoption of ‘‘management objec-
tives and strategies,’’ setting the scene for a Commission proposal to this effect.26
The proposal was the subject of some debate in the Council but discussions eventu-
ally stalled in 1995. Decisions on EU TACs were destined, for the time being, to
continue to be made on an annual basis and in the absence of predetermined deci-
sion rules.
Steps were nevertheless taken to move toward the use of multiannual manage-
ment plans, on a case-by-case basis, by the introduction of multiannual management
strategies under bilateral agreements with Norway (several stocks, including cod)
and other Northeast Atlantic states (herring and mackerel). These agreements speci-
fied that management action would be triggered when mortality or spawning stock
biomass passed the precautionary reference points. Further plans were adopted
within the context of the International Baltic Sea Fishery Commission (European
Commission 2000, 13). These prescribed that management action should ensure a
safe and rapid recovery, but without specifying exactly what this meant, or what
measures were to be used to achieve it. Managers had thus agreed when to take
action and certain management objectives, but had left open what action would be
taken (ICES 2001c).
As detailed above, ICES presented precautionary reference points for the first time
in its 1998 advice. In so doing, ICES introduced a precautionary approach ‘‘test’’ by
stating that if a stock is regarded as depleted, or if overfishing is taking place, only
an effective implementation of a rebuilding plan within a ‘‘reasonable’’ period
would satisfy the condition for a precautionary approach. By not developing effec-
tive recovery plans in the case of depleted stocks, the EU would be failing the test.
ICES applied this approach in 1999 when it stated that ‘‘fishing mortality on cod
should be reduced to the lowest level possible in 2000’’ (ICES 2000, part 2, 4) and
should be accompanied by a recovery plan to rebuild the spawning stock. Cod is
one of the EU’s most significant stocks, culturally if not economically, and the
ICES advice generated intense discussions in the Fisheries Council in December
1999. This led to a first set of emergency measures being adopted for Irish Sea cod
in 2000, followed by longer-term measures to rebuild the stock.
Following these initial developments, efforts were redoubled to develop multian-
nual management plans more widely for EU fisheries. Importantly, a meeting in Sep-
tember 2000 showed that the need to lay down multiannual procedures that took
the precautionary approach into account was now widely accepted by the member
states (European Commission 2000, 13). A more concerted focus on management
plans was consequently provided by a Communication of the Commission with pre-

determined decision rules presented as part of such an approach. The use of such
rules, according to the Commission, would end the practice that had resulted in pri-
ority being given to avoiding restrictions on fishing that were politically unpopular
in the short term (European Commission 2000, 11). In December of that year, min-
isters agreed that conservation measures were also needed for other stocks of cod,
and for hake, to be formulated within the context of multiannual recovery plans of
at least five years.
These developments coincided with the Commission’s preparations for the 2002
reform of the fisheries policy. In its Green Paper outlining the future direction of
the CFP, the Commission referred to the measures agreed on for cod and hake as
‘‘a test case’’ for introducing multiannual management strategies across a range of
commercial fisheries. It again supported the definition of multiannual strategies
compatible with the precautionary principle (European Commission, 2001, 23).
The final agreement, contained in the new basic CFP Regulation 2371/2002, states
that the Council ‘‘shall adopt’’ multiannual plans for stocks outside safe biological
limits and for other stocks, as far as necessary. Apart from biological reference
points, an explicit reference is made to establishing harvesting rules laying down
how annual catch or effort limits are to be arrived at. The new approach to manage-
ment was accompanied by efforts by the EU to improve stakeholder communication
and involvement through the introduction of regional advisory councils.
While political rather than legal, EU member states’ commitment to ICES’ advice
regarding the management of certain critical EU fish stocks effectively put pressure
on the EU to introduce recovery plans, which in turn were to provide the platform
for introducing preagreed decision rules. A key provision of the Fish Stocks Agree-
ment had thereby been translated into an EU context even before the EC and its
then fifteen member states became parties to the Agreement in 2003. ICES’ role as
interlocutor between the UN Fish Stocks Agreement and the CFP is all but explicit
in several ICES texts, including annual management advice to the EU that refers
directly to provisions of the Agreement.
Summary and Assessment

This section has examined a sequence of three cases of institutional interaction on
precautionary fisheries management. First, interaction between the global fisheries
regime and ICES has influenced the way scientists in the Northeast Atlantic generate
advice to decision makers. Partially overlapping membership is the main driver: the
scientists who participate in the more specialized regional institution were bound
by new global rules and responded by modifying the terms used in their advice and
by placing greater emphasis on rapid recovery whenever stocks are outside safe bio-
logical limits. The effect of this interaction could be deemed synergistic, since the
global fisheries regime induced ICES members to focus on a highly contentious
issue—how to respond to uncertainty regarding the effects of various harvesting
patterns—without disrupting their ability to come up with agreed-on recommenda-
tions. From another perspective, however, the largely natural-scientist driven imple-
mentation of the precautionary approach had tipped the fine balance between
biological and social concerns and was therefore in need of remedy (Hilborn et al.
2001).
This is where the second case of interaction becomes relevant since the global fish-
eries regime has also influenced the way ICES advice is communicated to other
stakeholders. Here, differences in the scope of objectives are the most important fac-
tor and the interaction is cognitive in nature. The FAO, which has a significant role
in the implementation of global fisheries rules, has a much broader mandate than
ICES. It has consistently and with some success advocated greater emphasis on non-
biological aspects of precautionary management, including socioeconomic impacts,
and has also pushed for greater emphasis on the interface between scientists and
other stakeholders. ICES has voluntarily responded to these considerations by
attempting to improve communication and harmonize terminology, because the
effectiveness of its advice depends on the voluntary acceptance by the institutions
it serves.
Third, these changes in ICES practice have intervened with the reform of the EU’s
Common Fisheries Policy and contributed to the provisions requiring multiannual
management plans and preagreed harvesting rules. The causal mechanism at work
in this case is Interaction through Commitment. Although some effect flows directly
from the EU’s involvement in the negotiation of the Fish Stocks Agreement and its
general commitment to the precautionary principle, the most specific causal signal
has been the scientific advice articulated by ICES ‘‘in the shadow’’ of the global fish-
eries regime. Notably, ICES’ dire advice regarding the state of EU cod and hake
stocks has acted as a motor behind the emergence of rebuilding plans.
Conclusions
The global fisheries regime has developed considerably during the past decade,
notably by the elaboration of a precautionary approach to fisheries management,
stronger and more specific requirements for cooperation on high-seas fisheries,

and more intrusive compliance-control procedures beyond national waters. Interac-
tion with international regimes in other issue areas, and with certain particularly
advanced regional fisheries regimes, has been important to this development (Stokke
2001; Stokke and Coffey 2001). This chapter has examined in detail two sets of
cases where the global fisheries regime has influenced problem-solving activities
under other institutions, specifically WTO regulation of fisheries subsidies, ICES’ pro-
vision of scientific advice, and regulatory decision making under the EU Common
Fisheries Policy (CFP).
For all of these cases, differences in means are an important driver of institutional
interaction. In the subsidies case, provisions of the fisheries regime relating to subsi-
dies are nonbinding and backed up by weaker compliance mechanisms than those of
the target regime, the WTO. The fisheries regime has raised awareness of partici-
pants in the trade regime to fisheries subsidies and lent credibility to a negative fram-
ing that highlights threats to sustainability and development rather than the positive
impacts that fisheries subsidies may have. Commitments of countries under the Fish
Stocks Agreement and more explicitly in IPOA-Capacity worked in the same direc-
tion. The precautionary cases also display institutional Interaction through Commit-
ment: the new binding commitments of the global fisheries regime have enhanced
the compelling force of precautionary procedures in ICES and, subsequently, the
EU CFP, resulting in modified scientific advice and more precautionary regulation
of fisheries supported by supranational enforcement.
Differences in membership are significant to the extent that they explain differ-
ences in means. With respect to precaution, the overall format provided by the
global fisheries regime has been specified and made operational within the narrower
Northeast Atlantic context where membership differs notably. Furthermore, differ-
ences in the objectives of the interacting institutions have led to differences as to
the bureaucratic sectors involved and the associated expertise, which has given rise
to Cognitive Interaction. One reason why the WTO has been so eager to maintain
cooperative links with the FAO on subsidies reform is the latter’s recognized exper-
tise in fisheries matters and its access to additional expertise at national management
levels. Similarly, the broader vocational blend characteristic of the FAO as com-
pared to ICES, especially with regard to the technological and social sciences,
explains in part why this organization has given relatively more emphasis to the
nonbiological aspects of precautionary management.
The overall effects of the cases of interaction examined here have been synergistic,
though not always overwhelmingly so. With regard to fisheries subsidies, the global
fisheries regime helped place this issue on the agenda of the new round of WTO ne-
gotiation, but it is yet unclear whether rules will actually be changed in a way that
will address overcapacity in the fisheries sector. Similarly, the precautionary provi-
sions of the Fish Stocks Agreement strengthened the hand of those within ICES and
subsequently EU fisheries bodies who favored greater safety margins, long-term
planning, and preagreement on recovery plans for endangered stocks. That said,
the actual impacts on management are unclear, partly because precautionary advice
has only recently been accompanied by regulatory decision making in favor of long-
term and precautionary management.
All cases display awareness among participants in source and target regimes of
the fact of interaction and also preparedness to respond to it, if necessary. Most of
the response has occurred within the respective regimes, while active interinsti-
tutional coordination has played only a moderate part. In the subsidies case, the
FAO was asked by the WTO to help clarify the causal relationships between subsi-
dies and responsible fisheries management but was unable, at least in the short term,
to provide specific findings that would facilitate agreement on the issue within the
WTO. There is little to suggest that more extensive cross-regime coordination would
have improved this interaction, due to the high level of political controversy that
surrounds the issue. The current WTO negotiations on fisheries subsidies that could
generate stronger and more enforceable rules create new opportunities for the FAO
and others to provide more specific inputs. As regards precautionary management,
some interagency coordination on how to improve implementation has also
occurred, in the form of broad expert meetings, but the higher level of conflict
among various stakeholders that accompanied introduction of the precautionary
approach has largely been addressed within each of the respective institutions.
The findings in this chapter confirm that institutional interaction at global and re-
gional levels can be significant for the ability of international and European regula-
tions to address environmental management effectively. Cross-institutional learning,
by flows of concepts and ideas, is an important way such interaction occurs, as are
processes of obligation in cases where the source regime is binding and the member-
ships partially overlap. Such impacts should not be expected to be strong in the
short term because inputs from other institutions are typically filtered through the
existing practices of the target regime. Accordingly, when examining how partici-
pants in the respective institutions respond to the interaction, it is vital that sufficient
attention be paid to the autonomous collective decision making within the source or
the target institutions (instead of focusing exclusively on interagency coordination
efforts).
Notes
The authors gratefully acknowledge constructive comments from Serge Garcia, Oran Young,
and our fellow contributors to this book, especially Thomas Gehring, Sebastian Oberthür,
and Alice Palmer. Helpful information has been provided by Olle Hagström, Carl-Christian
Schmidt, and Ronald Steenblik. The work has been supported by the European Community
under the Specific Research and Technological Development Programme ‘‘Energy, Environ-
ment and Sustainable Development’’ (contract no. EVK2-CT2000-00079); by the Research
Council of Norway (grant no. 128925/520/ KS); and by the Esmee Fairbairn Foundation,
UK. Parts of the material in the section ‘‘Precaution, ICES, and the Common Fisheries Policy’’
appeared in Stokke and Coffey 2004.
1. Agreement for the Implementation of the Provisions of the United Nations Convention on
the Law of the Sea of December 10, 1982, relating to the Conservation and Management of
Straddling Fish Stocks and Highly Migratory Fish Stocks, available at http://www.un.org/
Depts/los/index.htm.
2. The first three were adopted in 1999, the fourth in 2001; texts are available at http://
www.fao.org.
3. See, for example, WTO doc. WT/CTE/W/175, October 24, 2000, available at http://
www.docsonline.wto.org/gen_search.asp.
4. Access conditions are generally believed to be the most important factor explaining cross-
state variation in excess capacity (Cunningham and Gréboval 2001).
5. SCM Agreement, Articles 3, 5, and 6. Until December 31, 1999, even subsidies shown to
have adverse effects were nonactionable if they related to research activities, disadvantaged
regions, or adaptation of existing facilities in response to new environmental rules; see Arti-
cle 8.
6. Note that the FAO report did not actually make this claim, but much of its influence is due
to the fact that many read this finding into it (Milazzo 1998; also Stone 1997); see, for exam-
ple, WT/CTE/W/51, May 19, 1997, 2; also FAO 1998b, para. 8.
7. See WTO doc. WT/CTE/W/173, October 23, 2000.
8. WT/CTE/W/189, June 18, 2001, p. 1.
9. WTO doc. WT/CTE/W189, June 18, 2001.
10. Trade and Environment News Bulletins, TE/036, July 6, 2001.
11. WTO docs. WT/CTE/W/51 (United States), WT/GC/W/303 (Friends of Fish), and WT/
CTE/W/173 (Japan).
12. See, for example, WTO docs. WT/CTE/W/121, June 28, 1999; WT/GC/W/303, August
6, 1999; WT/CTE/W/154, July 4, 2000; TN/RL/W/3, April 24, 2002.
13. See FAO 1998; WTO doc. WT/CTE/W/126, October 12, 1999.
14. See, for example, WTO docs. WT/CTE/W/135, February 29, 2000; WT/CTE/W/189,
June 18, 2001.
15. Doha Declaration, paras. 28 and 31; available at http://www.wto.org.
16. See Trade and Environment News Bulletins, TE/018, July 1997; TE/021, December 16,
1997; TE/023, May 14, 1998; TE/028, March 31, 1999.
17. International Institute for Sustainable Development Report on the WTO’s High-Level
Symposium on Trade and Environment, March 15–16, 1999, available at http://www.wto.org.
18. See Trade and Environment News Bulletins, TE/029, July 30, 1999.
19. Compare the two 1997 documents on the issue—WT/CTE/W/51 (May 19, 1997) and
WT/CTE/W/67 (November 7, 1997, pp. 33–34)—with WT/CTE/W/121 (June 28, 1999);
also the report of the subsequent CTE meeting in WTO Trade and Environment News Bulle-
tins, TE/031 (November 8, 1999).
20. Detailed information on the topics discussed at these conferences is available at http://
www.ices.dk.
21. Compare ‘‘Statement of Conclusions from the Intermediate Ministerial Meeting on the
Integration of Fisheries and Environmental Issues’’ (http://www.dep.no/md/nsc/ Intermediate_
meeting/023021-990005/index-dok000-b-n-a.html), especially sections on Guiding Principles
and Strategies, with FAO 1996).
22. ‘‘FAO Fisheries Department—Staff List and Capabilities’’ (available at http://www.dec
.ctu.edu.vn/cdrom/cd6/projects/faofish_1197/staffirm.htm).
23. ‘‘Resolutions Adopted at the 2001 Annual Science Conference,’’ p. 14, available at http://
www.ices.dk.
www.ices.dk.
www.ices.dk.
26. Proposal for a Council Regulation fixing management objectives and strategies for cer-
tain fisheries or groups of fisheries for the period 1994 to 1997, Doc. COM(93)663, in Offi-
cial Journal C 017, 20/01/1994.
References
Brown, Sue. 2000. The Precautionary Approach: A User’s View. Paper presented at ICES
2000 Annual Science Conference, September 27–30, Brugge. Copenhagen: International
Council for the Exploration of the Sea.
Caddy, John. 1998. A Short Review of Precautionary Reference Points and Some Proposals
for Their Use in Data-Poor Situations. FAO Fisheries Technical Paper 379: 30. Rome: Food
and Agriculture Organization.
Cunningham, Steve, and Dominique Gréboval. 2001. Managing Fishing Capacity: A Review
of Policy and Technical Issues. FAO Fisheries Technical Paper 409. Rome: Food and Agricul-
ture Organization.
Deas, Barrie. 2000. Fishermen and Scientists: Collaboration as the Basis for Stock Recovery.
Paper presented at ICES 2000 Annual Science Conference, September 27–30, Brugge. Copen-
hagen: International Council for the Exploration of the Sea.
European Commission. 2000, December 1. Communication from the Commission to
the Council and European Parliament—Application of the Precautionary Principle and
Multi-annual Arrangements for Setting TACs. Doc. COM(2000)803. Brussels: European

Commission.
European Commission. 2001, March 20. Green Paper on a Future Common Fisheries Policy.
Doc. COM(2001)135. Brussels: European Commission.
FAO. 1992. Special Chapter: Marine Fisheries and the Law of the Sea—A Decade of Change.
The State of Food and Agriculture 1992. Rome: Food and Agriculture Organization. Avail-
able at http://verweile.com/445_html/los-1/.
FAO. 1996. Precautionary Approach to Capture Fisheries and Species Introductions. FAO
Technical Guidelines for Responsible Fisheries 2. Available at http://www.fao.org/DOCREP/
003/W3592E/W3592E00.htm.
FAO. 1997. Review of the State of World Fishery Resources: Marine Fisheries. FAO Fisheries
Circular 920. Available at http://www.fao.org/docrep/003/w4248e/w4248e00.htm.
FAO. 1998a. COFI Report of the Sixth Session of the Sub-Committee on Fish Trade. FAO
Fisheries Reports R589. Rome: Food and Agriculture Organization.
FAO. 1998b. Committee on Fisheries, Sub-Committee on Fish Trade. Fisheries Management,
Subsidies and International Fish Trade. Sixth Session, Bremen, June 3–6. Doc. COFI: FT/VI/
98/4. Rome: Food and Agriculture Organization.
FAO. 2001a. Report of the Ad Hoc Meeting with Intergovernmental Organizations on Work
Programmes Related to Subsidies in Fisheries. FAO Fisheries Reports R649. Available at
http://www.fao.org/DOCREP/003/ Y0999E/ Y0999E00.HTM.
FAO. 2001b. The State of the World Fisheries and Aquaculture 2000. Available at http://
www.fao.org/DOCREP/003/ X8002E/ X8002E00.htm.
FAO. 2002. Report of the Second Ad Hoc Meeting with Intergovernmental Organizations on
Work Programmes Related to Subsidies in Fisheries. FAO Fisheries Reports R688. Rome:
Food and Agriculture Organization.
Garcia, Serge M. 2000. The Precautionary Approach to Fisheries 1995–2000: Progress Re-
view and Main Issues. In Report of the CWP Intersessional Meeting. Advisory Committee
on Fisheries Management. ICES doc. CM 2000/ACFM:17, 19–46. Copenhagen: Interna-
tional Council for the Exploration of the Sea. Available at http://www.ices.dk/reports/acfm/
2000/cwp/cwp00.pdf.
Gréboval, Dominique. 2000. The FAO International Plan of Action for the Management of
Fishing Capacity. Paper presented at the Conference IIFET 2000, International Institute of
Fisheries Economics and Trade, Oregon State University, July 10–15, 2000, Corvallis, Ore-
gon. Available at http://oregonstate.edu/Dept/ IIFET/2000/abstracts/greboval.html.
Gullestad, Petter. 1998. The Scope for Research in Practical Fishery Management. Fisheries
Research 37 (1–3): 251–258.
Hanna, Susan S. 1999. Strengthening Governance of Ocean Fishery Resources. Ecological
Economics 31 (2): 275–286.
Hannesson, Rögnvaldur. 2001. Effects of Liberalizing Trade in Fish, Fishing Services and In-
vestment in Fishing Vessels. OECD Papers, 1. Available at http://www.oecd.org/dataoecd/1/
11/1917250.pdf.
Hilborn, Ray, Jean-Jacques Maguire, Ana M. Parma, and Andrew A. Rosenberg. 2001. The
Precautionary Approach and Risk Management: Can They Increase the Probability of Suc-
cesses in Fishery Management? Canadian Journal of Fisheries and Aquatic Sciences 58 (1):
99–107.
ICES. 1992. Report of the ICES Advisory Committee on Fisheries Management. ICES Coop-
erative Research Report 179. Copenhagen: International Council for the Exploration of the
Sea.
ICES. 2000. Report of the ICES Advisory Committee on Fisheries Management, 1999. ICES
Cooperative Research Report 236. Copenhagen: International Council for the Exploration of
the Sea.
ICES. 2001a. Advisory Committee on Fishery Management. Report of the Study Group on
the Further Development of the Precautionary Approach to Fisheries Management. ICES
Doc. CM 2001/ACFM:11. Copenhagen: International Council for the Exploration of the Sea.
ICES. 2001b. Report of the ICES Advisory Committee on Fisheries Management. ICES Co-
operative Research Report 242. Copenhagen: International Council for the Exploration of
the Sea.
ICES. 2001c. Report of the ICES Advisory Committee on Fisheries Management. ICES Coop-
erative Research Report 246. Copenhagen: International Council for the Exploration of the
Sea.
Mace, Pamela M., and Wendy L. Gabriel. 1999. Evolution, Scope and Current Applications
of the Precautionary Approach in Fisheries. Paper presented at the Fifth National NMFS
Stock Assessment Workshop, February 24–26, Key Largo, Florida. NOAA Technical Memo-
randum, NMFS-F/SPO-40. Available at http://www.st.nmfs.gov/st2/nsaw5.html.
MacGarvin, Malcolm. 2002. Fisheries: Taking Stock. In Poul Harremoës, David Gee, Mal-
colm MacGarvin, Andy Stirling, Jane Keys, Brian Wynne, and Sofia Guedes Vaz, eds., The
Precautionary Principle 1896–2000: Late Lessons from Early Warnings, 10–25. London:
Earthscan.
Milazzo, Matteo. 1998. Subsidies in World Fisheries: A Reexamination. World Bank Techni-
cal Paper: Fisheries Series. Washington, DC: World Bank.
OECD. 1997. Towards Sustainable Fisheries: Economic Aspects of the Management of Liv-
ing Marine Resources. Paris: OECD.
OECD. 2000. Transition to Responsible Fisheries: Economic and Policy Implications. Paris:
OECD.
Pauly, Daniel, Villy Christensen, Sylvie Guénette, Tony J. Pitcher, U. Rashid Sumaila, Carl J.
Walters, Reg Watson, and Dirk Zeller. 2002, August 8. Towards Sustainability in World
Fisheries. Nature 418: 689–695.
Porter, Gareth. 2002. Fisheries Subsidies and Overfishing: Towards a Structured Discussion.
Volume 1 in the UNEP/ETB Fisheries and the Environment Series. Geneva: United Nations
Environment Program. Available at http://www.unep.ch/etu/etp/acts/capbld/rdtwo/FE_vol_1
.pdf.
PricewaterhouseCoopers. 2000. Study into the Nature and Extent of Subsidies in the Fisheries
Sector of APEC Members Economies. Prepared for Fisheries Working Group Asia Pacific
Economic Co-operation. Available at http://www.apec.org/apec/apec_groups/working_

groups/fisheries.downloadlinks.0004.LinkURL.Download.ver5.1.9.
Ruckes, Erhard. 2000. International Trade in Fishery Products and the New Global Trading
Environment. In Multilateral Trade Negotiations on Agriculture: A Resource Manual. Rome:
Food and Agriculture Organization. Available at www.fao.org/DOCREP/003/X7351E/
X7351E00.HTM.
Schorr, David. 1998. Towards Rational Disciplines on Subsidies to the Fishery Sector: A Call
for New International Rules and Mechanisms. The Footprint of Distant Water Fleets on
World Fisheries (a series of papers published by WWF). Godalming, UK: World Wide Fund
for Nature. Available at http://www.panda.org/news_facts/publications/search.cfm?uNC=
54358643.
Steenblik, Ronald P. 1999. Previous Multilateral Efforts to Discipline Subsidies to Natural
Resource Based Industries. In Michael Riepen, ed., Report of Proceedings: Workshop on the
Impact of Government Financial Transfers on Fisheries Management, Resource Sustainabil-
ity, and International Trade, August 17–19, 1998, Manila, The Philippines, 39–80. Singa-
pore: Pacific Economic Co-operation Council.
Stewart, James E. 1991. A Brief Review of the International Council for the Exploration of
the Sea (ICES) on the Occasion of the Formation of the North Pacific Marine Science Organi-
zation. Canadian Journal of Fisheries and Aquatic Sciences 48 (12): 2543–2550.
Stokke, Olav Schram, ed. 2001. Governing High Seas Fisheries: The Interplay of Global and
Regional Regimes. Oxford: Oxford University Press.
Stokke, Olav Schram, and Clare Coffey. 2001. Institutional Interaction in High Seas Fisheries
Management: An Inventory. Project Deliverable No. D8 of the Project: Institutional Interac-
tion: How to Prevent Conflicts and Enhance Synergies between International and EU Envi-
ronmental Institutions. Available at http://www.ecologic.de/projekte/interaction/results.htm.
Stokke, Olav Schram, and Clare Coffey. 2004. Precaution, ICES and the Common Fisheries
Policy: A Study of Regime Interplay. Marine Policy 28 (2): 117–126.
Stone, Christopher D. 1997. Too Many Fishing Boats, Too Few Fish: Can Trade Laws Trim
Subsidies and Restore the Balance in Global Fisheries? Ecology Law Quarterly 24 (3): 505–
544.
Symes, David. 1998. The Integration of Fisheries Management and Marine Wildlife Conser-
vation. JNCC Report No 287. Peterborough: Joint Nature Conservation Committee. Avail-
able at http://www.jncc.gov.uk/marine/fisheries/reports/rpt_integFishMan.htm.
7
The Convention on International Trade in
Endangered Species of Wild Fauna and Flora
(CITES): Responding to Calls for Action from
Other Nature Conservation Regimes
John Lanchbery
The Convention on International Trade in Endangered Species of Wild Fauna and

Flora (CITES) is unusual among international environmental regimes in that its
main mode of operation, the regulation of international trade, tends to drive it to
interact and cooperate with a number of nonenvironmental institutions, in particu-
lar concerning trade. From its entry into force in 1975, it has thus worked not only
via its parties and their domestic police and customs organizations but also directly
with the World Customs Organization (WCO), the International Criminal Police
Organization (Interpol), and the General Agreement on Tariffs and Trade (GATT)
that later became part of the World Trade Organization (WTO).
Furthermore, CITES interacts with a number of other international regimes con-
cerning the conservation of wildlife,1 of which CITES is probably the leading exam-
ple and has in some ways formed the center. Wildlife conservation regimes often
work together as a group of treaties, assisting each other to meet common objec-
tives. This is because they share a common, overall goal (the conservation of wild
animals or plants), but differ either in the species they cover or in the way they pur-
sue the goal. Thus, while CITES regulates international trade in endangered species,
the Convention on Migratory Species of Wild Animals (CMS) provides for conser-
vation measures ‘‘on the ground,’’ the International Convention on the Regulation
of Whaling restricts the taking of whales, and so on.
The fact that the conservation agreements are interrelated in this way is no acci-
dent. Many of them have common origins, often in the form of calls for their estab-
lishment by the World Conservation Union (IUCN), and they were often specifically
designed to operate in different but complementary ways. The IUCN was the initial
instigator of most of the major wildlife treaties concluded in the second half of
the twentieth century, including both CITES and the CMS. Cooperation between
the wildlife treaties is actively promoted by the IUCN, both directly and via the
158 John Lanchbery
secretariats to the treaties, which it commonly provides with the United Nations
Environment Program (UNEP).
This chapter focuses mainly on the interactions between CITES and two other na-
ture conservation regimes: the CMS and the Convention for the Conservation and
Management of the Vicuña (Convenio para la Conservación y Manejo de la Vicuña,
CCMV). These cases of interaction were selected chiefly because they are examples
of how the nature conservation treaties often act in concert and of the central role of
CITES in this. In particular, they demonstrate how regimes with limited member-
ships have used CITES, with its larger, global membership, to assist them in meeting
common nature conservation goals. CITES responded to a call for action by the
CMS to protect the Asiatic subspecies of the houbara bustard, a bird recognized as
endangered by both CITES and the CMS. The CCMV asked CITES for assistance in
protecting the vicuña, which was again recognized as threatened by both regimes.
The chapter begins with a description of CITES, its origins, and its mode of oper-
ation. It then provides an overview of the interactions of CITES with a broad range
of other international institutions and EU legislative instruments. Subsequently,
the two cases of interaction with the CMS concerning the houbara bustard and the
CCMV concerning the vicuña are investigated in greater detail. In each case, the
other treaty is described and an account of the concrete case of interaction is pro-
vided. The chapter ends with some general conclusions.
The Convention on International Trade in Endangered Species of Wild Fauna and

Flora (CITES)
CITES seeks to protect wild species by regulating trade in both the species them-
selves and products made from them. Its aim is to conserve endangered species, but
it does so by attempting to control trade. In doing so, it implicitly recognizes that it
is often hard for international regimes to effectively prescribe what their parties
should do at home, whereas it can be comparatively easy to regulate an interna-
tional activity, such as trade.
CITES’ focus on attempting to limit trade appears to be well justified. Interna-
tional trade in wildlife is enormous, and controlling it to sustainable levels would
undoubtedly do much to conserve many species. The Directorate General Environ-
ment of the European Commission estimates that international wildlife trade, both
legal and illegal, is worth at least U.S.$10–20 billion annually. From 1995 to 1999,
legal trade in CITES-listed species alone involved 1.5 million live birds, 640,000 live
CITES 159
reptiles, about 3 million reptile skins, 150,000 furs, almost 300 tons of caviar, over
1 million pieces of coral, and 21,000 hunting trophies (Mulliken 2002).
The idea of limiting international trade in endangered species is not new. It dates
back to the first decade of the twentieth century, when there was a call for a treaty
limiting trade in the exotic bird feathers then used in ladies’ hats (Lyster 1985).
Later, the 1933 London Convention on the preservation of fauna and flora in Africa
included provisions for restricting trade and also introduced the concept of having
easily changeable annexes or appendixes listing endangered species (Lyster 1985;
Sand 1997). Many later regional fauna and flora treaties followed the example of
the London Convention, paving the way for the global CITES agreement that limits
trade in species listed in exactly the same way (Lanchbery 1995).
Substantive political moves for a global agreement on trade in endangered species
began in the 1950s, together with other wildlife agreements, notably the Ramsar
Convention on Wetlands (Lyster 1985; Burns 1990). In 1963, the Governing Coun-
cil of the IUCN called for ‘‘an international convention on regulating the export,
transit and import of rare or threatened fauna and flora species or their skins or tro-
phies.’’2 Subsequent progress on negotiating a treaty was, however, slow until the
1972 Stockholm United Nations Conference on the Human Environment reempha-
sized the need for such an agreement. By March 1973 CITES had been negotiated
and signed by its first twenty-one parties (Sand 1997). By spring 2005 it had 167
parties (http://www.cites.org), the largest membership of all wildlife conservation
treaties, of which it is very much the ‘‘flagship’’ (Wijnstekers 2003).3
The main features of the mode of operation of CITES are three appendixes that
are reviewed and can be changed at its Conference of the Parties (COP), which has
met roughly every two years. The first appendix lists species in which trade is
banned in all but the most exceptional circumstances. Appendix II includes species
not necessarily threatened with extinction, but in which trade must be controlled in
order to ‘‘avoid utilization incompatible with their survival’’ (Article II.2.a of
CITES). Appendix III contains species that are protected in at least one country,
which has asked other CITES parties for assistance in controlling the trade (Article
II).
The treaty lays down detailed rules governing the import and export of species or
products made from them. It also requires parties to establish national Management
Authorities and Scientific Authorities. It further determines that the COP to CITES
should meet every two years. Some operational features of the treaty have evolved
considerably over time, extending and strengthening the remit of CITES over and
160 John Lanchbery
above the provisions in the original text of the treaty (Wijnstekers 2003). It has, for
example, set up a Standing Committee to oversee the operation of the agreement be-
tween COPs as well as other committees that assist the parties in making decisions
on the classification of species. It has also developed a compliance and enforcement
mechanism operated by the Secretariat and the Standing Committee. This mecha-
nism includes the possibility of banning all wildlife trade with recalcitrant states.
Throughout the 1990s, for example, the Standing Committee banned such trade
with a number of countries, including Italy, Thailand, and Greece, until they came
into compliance (Lanchbery 1995; Reeve 2002).
No description of CITES, or almost any wildlife treaty, is complete without a de-
scription of the role of the IUCN. Founded on October 5, 1948, as the International
Union for the Protection of Nature (IUPN), the organization changed its name to the
International Union for the Conservation of Nature and Natural Resources (IUCN)
in 1956. In 1990 this was shortened to IUCN–The World Conservation Union. It
called for many of the conservation agreements and helps considerably in their op-
eration, by providing information concerning endangered species and by providing
services such as secretariats, often with UNEP (Lanchbery 1995). Indeed, the IUCN’s
‘‘Red Lists’’ of endangered species of wild animals and plants that are produced by
the Species Survival Commission of the IUNC (http://www.redlist.org) largely drive
the listing of species in CITES appendixes.
The IUCN is a complex and unusual international organization in that it has
a large and varied membership of both states and nongovernmental organiza-
tions (NGOs). In 2002, its membership comprised 675 national NGOs, 68 inter-
national NGOs, 72 states (i.e., governments), and 107 government agencies. This
unique combination of members makes it both a source of excellent, reliable infor-
mation and, because of its governmental membership alone, a powerful force to be
reckoned with. It has a large permanent staff, of in excess of 1,000, with 10,000 ex-
pert volunteers (http://www.iucn.org).
CITES and the other wildlife conservation treaties also have strong links with the
more conventional wildlife NGOs, almost all of which are members of the IUCN.
Notable among these are the World Wide Fund for Nature (WWF), the World Con-
servation Monitoring Centre (WCMC), Trade Records Analysis of Fauna and Flora
in Commerce (TRAFFIC), and, especially in the case of treaties with a significant
wild-bird interest (such as the CMS and Ramsar), BirdLife International. Both the
WCMC and TRAFFIC were originally set up by WWF but are now independent of
it. The WCMC has in the meantime been incorporated in UNEP.
CITES 161
These organizations play a key part in the operation of CITES and the CMS. In-
deed, the IUCN, WCMC, and TRAFFIC are formally recognized as the ‘‘technical
partners’’ in CITES. They provide much of the information about science and about
infractions that the regime needs in order to operate effectively. The IUCN’s ‘‘Red
Lists’’ of endangered species are based heavily on information from organizations
such as BirdLife International and WWF and misbehavior by parties is often
reported to the CITES Secretariat by TRAFFIC (Lanchbery 1995; Reeve 2002).
Interactions and Synergies with Other Institutions
CITES has been involved in many interactions with other international institutions
and EU legal instruments. Fourteen such interactions between CITES, involving
seven other international institutions and the EU, are summarized in table 7.1. This
list is not exhaustive but covers many pertinent cases. In most cases, extensive and
lasting cooperation and exchanges between CITES and the other institutions have
developed in response to the initial interinstitutional influence indicated in the table.
Horizontal Interactions with Other International Institutions

While it is hard to generalize about all CITES interactions from what is not an ex-
haustive list, all the horizontal interactions between international institutions were
synergistic or neutral, or they were managed successfully so that tensions were pre-
vented from turning into open conflict. The interaction with the WTO is the only
one in which the underlying relationship was disruptive because the WTO promotes
free trade, which is restricted by CITES. The commitments of both institutions
are potentially at odds (Interaction through Commitment). This is not to say that
political conflict did not occur about the political responses to interinstitutional
influence. For example, a minority of parties (the remaining nations that conduct
whaling) vigorously opposed the CITES decision to restrict trade in whale products
in response to a request by the International Convention on the Regulation of Whal-
ing and lodged reservations to it (for an account of the changes in the whaling re-
gime and its impacts elsewhere, see Andresen 1998). The request as such, however,
was fully in line with the conservation objective of CITES.
With the exception of the interaction with the WTO, all other cases of interaction
followed the causal mechanisms of Cognitive Interaction and Behavioral Inter-
action. All cases in which CITES asked other institutions for support, or vice
versa, constituted Cognitive Interaction because the decision on whether to respond
162 John Lanchbery
Table 7.1
Interactions of CITES
Convention for the Conservation CCMV has asked CITES for help in limiting trade in
and Management of the Vicuña vicuña products and CITES has responded positively
(CCMV) with a COP Resolution.
International Convention on the The ICRW has asked for help in restricting trade in
Regulation of Whaling (ICRW) whale products and CITES has responded positively
with a COP Resolution.
Convention on Migratory In response to a call for action by CMS, CITES
Species of Wild Animals (CMS) passed a COP Resolution in support of CMS
concerning the houbara bustard.
Convention on Biological CITES supports implementation of the CBD; several
Diversity (CBD) decisions and resolutions in both institutions promote
cooperation and synergy.
World Trade Organization CITES restricts free trade and thus is in potential
(WTO) conflict with the WTO; it allows for trade with
nonparties complying with its obligations.
World Customs Organization CITES asked the WCO for help in implementation.
(WCO) In response, WCO supports implementation of
CITES by helping coordinate CITES enforcement and
training via its members (national customs
organizations). There is a CITES/WCO memorandum
of understanding on cooperation.
CITES asked WCO to change customs codes (e.g.,
for shark products).

International Criminal Police CITES asked Interpol for help in implementation and
Organization (ICPO, Interpol) enforcement.
In response, Interpol supports implementation of
CITES. There is a CITES/ICPO memorandum of

understanding on cooperation, a CITES/ICPO Wildlife
Crime Working Group, and work on joint training.
EU Regulation on trade in The EU (then: the EEC) adopted a regulation to
endangered species control wildlife trade in 1982 in response to CITES,
although it was not a party to CITES.
CITES concerns about abolition of internal border
controls caused the EU to strengthen its CITES

Regulation in 1997 (revised in 2001).
CITES Regulation of the EU supports the
implementation of CITES.
EU Single Market Program Abolition of intra-EU border controls for goods and
persons endangers effective implementation of CITES
trade restrictions in the EU.
CITES 163
favorably to the request for help was made completely voluntarily by the targets.
The request as such included no substantive carrots or sticks that could have moti-
vated the target, but drew the latter’s attention to the needs of the requesting institu-
tion. The positive response of the WCO and Interpol as well as the UN Food and
Agriculture Organization (FAO; see chapter 6) to the request by CITES then con-
tributed to a more effective implementation of CITES (Behavioral Interaction). The
implementation of CITES itself contributes to achieving the objective of the Conven-
tion on Biological Diversity (CBD) to protect biological diversity (see also chapter 4).
CITES has ‘‘natural’’ synergies with many other environmental institutions, par-
ticularly those concerning wildlife conservation. These arise from the fact that these
treaties often have either the same or overlapping aims, basically to conserve wildlife
and, more generally, biological diversity. These treaties include global agreements
such as the Convention on Migratory Species, the CBD, and a host of regional
agreements, such as that on the conservation of the vicuña. This synergy tends to
be reinforced by the fact that they typically employ different means to achieve their
ends: CITES restricts wildlife trade, the CMS provides for protection measures ‘‘on
the ground,’’ and so on. Rather than compete, the regimes therefore usually comple-
ment each other, often with one regime asking another to support it in attaining a
particular goal, as in the case of the interactions recounted later.
That CITES has been a prominent target for requests for help from other regimes
may be due to the fact that it has the largest membership, possibly the greatest influ-
ence, and certainly the highest public profile of the wildlife regimes.
The interactions of CITES with nonenvironmental institutions are quite different
from those with wildlife treaties. Because CITES seeks to influence international
trade it clearly should, in order to be effective, attempt not to come into conflict
with and, if possible, benefit from other bodies concerning trade, notably the World
Customs Organization (WCO), the International Criminal Police Organization
(ICPO, Interpol), and the WTO.
Because of the institutions’ different modes of operation, the potential for conflict
between CITES and the WTO is greater than with the WCO and Interpol. The WCO
and Interpol tend to act in many ways as trade associations whereas the WTO, like
CITES, aims to regulate and control behavior—that is, international trade. The
WCO and Interpol mainly aim to foster cooperation among their members, for
example by providing education and training programs. The request for help in
implementation originally issued by CITES to both WCO and Interpol was thus
compatible with their overall aims, although it also did not promise to facilitate
164 John Lanchbery
achieving their primary aims. Both organizations responded positively and have
since supported the effective implementation of CITES, for instance by providing
focused training to national customs and police officers (Reeve 2002). CITES has
also long had memoranda of understanding with the WCO and Interpol, which
were updated in 1999 in the drive for greater effectiveness by CITES. The CITES
Secretariat sits on Interpol’s Wildlife Crime Working Group.
As indicated above, the relationship between CITES and the WTO has been more
problematic but has been managed successfully so as to avoid open conflict so far.
Like other trade-related multilateral environmental agreements (chapter 8), CITES
restricts international trade, but it promotes compatibility of its provisions with in-
ternational trade rules by allowing parties to trade products regulated by CITES
with nonparties that essentially comply with CITES requirements. While not all ten-
sions may have been removed, no dispute related to CITES has yet arisen under the
WTO. A cooperative and amicable relationship between both institutions is further
promoted by the CITES Secretariat that sits on the WTO’s Committee on Trade and
Environment (CTE).
CITES has actively promoted synergy with other international agreements in
order to maximize its effectiveness, and has had a formal policy of doing so since
the early 1990s. The need to promote synergy, in general, was first formally raised
at the ninth meeting of the Conference of the Parties (Fort Lauderdale 1994) in the
Strategic Plan of the Secretariat (CITES 1994). At the tenth Conference of the Par-
ties, this need was reiterated in the context of a review of the effectiveness of the
Convention, and a decision was adopted calling for intensified and extended coop-
eration with other conventions (CITES Decision 10.110). This led the Secretariat to
produce a document for the CITES Standing Committee at its forty-second meeting
in 1999 titled Synergy Between the Biodiversity-Related Conventions and Relations
with Other Organizations (CITES 1999), which contains various recommendations
spanning a wide scope. In practice, members of the secretariats of wildlife agree-
ments regularly attend others meetings. Also, CITES and the CMS concluded a
memorandum of understanding concerning the need to work more closely together
in September 2002.
Nevertheless, cooperation is often far from perfect, as the CITES Secretariat
pointed out in the 1999 synergies paper:
The need now to develop synergy and provide better policy coordination among existing and
future agreements is obvious. This particularly applies to the so-called biodiversity-related
MEAs: CITES, the Convention on Biological Diversity (CBD), the Convention to Combat
CITES 165
Desertification (CCD), the Convention on Migratory Species (CMS) and the Convention on
Wetlands (Ramsar). Although these Conventions address different aspects of the same issue,
the risk of some overlap and duplication of effort is evident. (CITES 1999, 1)
Vertical Interactions with EU Legal Instruments

Although the EU is not a party to CITES, its legislation has interacted ‘‘vertically’’
with CITES in several ways. The interaction has relied on the fact that most EU
member states have always been parties to CITES and have in large part relied on
the EU for the implementation of their commitments. Even though CITES was offi-
cially amended in 1984 to allow for membership of regional economic integration
organizations (i.e., the EU), the EU has not been in a position to join CITES because
the amendment still awaits sufficient ratifications for entry into force. In the mean-
time, Ireland was the last EU member state to join CITES in 2000. Since all ten
states that joined the EU in 2004 are also parties to CITES, all EU member states
are now CITES parties.
Facilitated by the fact that (most) EU member states were committed to imple-
menting CITES controls (Interaction through Commitment), the EU acted, synergis-
tically, in many ways as though it were a member by implementing CITES in EU
legislation. As early as 1982 it developed a Regulation (EEC No. 3626/82) for
implementing CITES within the EU, which came into force in 1984. As elaborated
below, it has since revised and updated this legislation to ensure compliance with
CITES requirements. It set up a CITES Committee and an Enforcement Working
Group and established a common reporting format. It also imposed trade sanctions
against states for noncompliance with CITES, for example by banning trade in
endangered species with Indonesia in 1992 (Reeve 2002, 126).
The significance of this EU implementation of CITES is that it activates the partic-
ularly effective supranational enforcement powers of the EU that subsequently
support an effective implementation of CITES in the EU (Behavioral Interaction).
Evidence for this synergistic effect of the EU implementation is found in enforcement
activities of the European Court of Justice. In 1990, for example, the Court found
that France had unlawfully issued import permits for 6,000 wildcat skins from Bo-
livia (Reeve 2002, 113).
Another part of the EU’s ‘‘acquis communautaire,’’ however, increasingly endan-
gered the effective implementation of CITES since the 1980s and thus had a disrup-
tive effect at the behavioral level. The Single European Act of 1987 introduced a
legislative program that aimed at the complete abolition of internal border controls.
This added to concerns raised by other parties because it would endanger an effective
166 John Lanchbery
implementation of CITES in the EU. The effectiveness of CITES enforcement within

the EU would essentially be determined by the member state with the most lax ex-
ternal import and export controls. CITES responded to this challenge by adopting
a number of resolutions calling on the EU and its member states to ensure a more
effective implementation of CITES (Reeve 2002, 112–120).
The EU’s eventual response (after fifteen years) was new Council Regulation (EC)
No. 338/97 and Commission Regulation (EC) No. 939/97, which came into effect
on June 1, 1997. These conceded that the new regulations were needed to ‘‘ade-
quately reflect the current structure of trade’’ and ‘‘cope with the abolition of inter-
nal border controls which resulted from the Single Market. The abolition of internal
borders has made necessary the adoption of stricter trade control measures at the
Community’s external borders.’’ In 2001 Commission Regulation (EC) No. 939/97
was replaced by Commission Regulation (EC) No. 1808/2001.
The Interaction between CITES and the CMS Concerning the Houbara Bustard
The Origins and Operation of the CMS

In the 1960s, the IUCN General Council drew international attention to the plight
of many migratory animals and called for a treaty to protect them. As in the case of
CITES, this call was reinforced by the 1972 Stockholm Conference, in its Recom-
mendation 32 (CMS 2002). After a long gestation period, the treaty was concluded
in Bonn in 1979 and entered into force in November 1983.
As in the case of many postwar wildlife treaties, the CMS was constructed as
a framework agreement that allows its commitments to be expanded or changed
over time. Like CITES, it achieves this by having two revisable appendixes that list
species needing protection. Species can be listed on both appendixes. The first lists
species having ‘‘unfavorable conservation status,’’ which countries within the natu-
ral range of a migratory animal (range states) are obliged to protect. The second lists
species for which completely new subagreements are required. In other words, the
CMS is deliberately designed to spawn new agreements. There are now agreements
on African-Eurasian migratory waterbirds, small cetaceans of the Baltic and
North Seas, European bats, and the Great Bustard and six other species or sets of
species (see the CMS website at http://www.cms.int for the latest information on
agreements).
Originally, it was intended that only full, legally binding new subagreements
(Agreements with a capital A) would be set up, but it soon became apparent that
CITES 167
nonbinding memoranda of understanding could be usefully employed as well, and

so both are now used. Which is employed is largely a practical decision. Where
few parties with a record of cooperation are involved, memoranda are commonly
used, because they can be agreed on more quickly and avoid the need for formal
ratification.4
Both Agreements and memoranda are usually developed with one party taking a
lead in drafting it. The party selected should be both a range state for the species
involved and have knowledge of and expertise in it. In the case of the interaction
covered here, the lead on the houbara bustard was taken by Saudi Arabia.5
In the context of this study, it is significant that the CMS has more limited partic-
ipation (by states) than some other wildlife conservation agreements that are also,
nominally, global in scope. Notable absentees from the list of CMS parties are the
United States, Russia, Canada, Mexico, and Brazil. The reasons for lack of partici-
pation vary. The United States, for example, claims that existing arrangements with
its neighbors make participation unnecessary. However, many states that are not
parties to the CMS itself still participate in its Agreements and memoranda. The
United States and Russia are, for example, parties to such agreements. Nevertheless,
it can still assist the CMS in achieving its aims if it can elicit the support from other
institutions with wider participation, such as, in this case, CITES.
The Interaction
The houbara bustard (chlamydotis undulata) is a rare and endangered species of
large terrestrial bird (IUCN 2002). There are two subspecies: the Asiatic, which is
migratory, and the North African, which is not (BirdLife International 2000). Only
the former, migratory population (chlamydotis undulata macqueenii) therefore po-
tentially qualifies for inclusion in the CMS although both populations are eligible
for inclusion in, and are included in, the CITES appendixes. The Asiatic subspecies
of the houbara bustard is included in Appendix I of CITES.
Chlamydotis undulata macqueenii breeds mainly in Central Asia and migrates to
the Arabian peninsula during the period October to March. Throughout most of
their vast range the houbara’s numbers have long been in decline. The reasons for
this decline appear to be habitat destruction, through overgrazing and intensive
farming in their breeding areas, coupled with overhunting, human disturbance, and
overtrapping in countries through which they migrate (BirdLife International 2001).
The Asiatic subspecies of the houbara bustard has long been on the IUCN ‘‘Red
Lists’’ and therefore, within the CMS, it was recognized as having ‘‘unfavorable
168 John Lanchbery
conservation status’’ and requiring either an Agreement or a memorandum of un-

derstanding between range states. The first institutional call for concerted action
was made at the third Conference of the Parties of the CMS (COP Resolution 3.2)
in 1991. The call was repeated at the fourth COP in 1994 (Resolution 4.4). CMS
Resolutions are less targeted than its Decisions, and generally intended to provide
long-standing guidance. They are politically important because they indicate a
strong desire by the COP for parties to take action.
As a result of these Resolutions, Saudi Arabia offered to take the lead in drafting
an agreement, either binding or nonbinding. Saudi Arabia is both a range state for
the houbara bustard and has considerable knowledge of it. It was also the Asia rep-
resentative on the Standing Committee of the CMS, which provides policy and ad-
ministrative guidance between regular COPs. The Standing Committee consists of
seven representatives of the main regions and includes the depositary (Germany)
and the host of the next COP (http://www.cms.int). It was generally accepted that
drafting an action plan would probably not be hard, given Saudi Arabia’s expertise
on the bustard.6
The status of the bustard was then raised, again, at the fifth COP of the CMS
in April 1997 (Recommendation 5.4), by which time Saudi Arabia had prepared
a draft agreement. Without specifically mentioning CITES, the Recommendation
merely reiterated the need to conserve the bustard and for range states to assist
Saudi Arabia with the agreement. In addition, IUCN’s World Conservation Con-
gress had highlighted the poor conservation status of the houbara bustard in 1996
(Recommendations 1.27 and 1.28). Recommendation 1.27 was a general call for the
protection of two species of the houbara bustard and mentions the CMS and CITES
in the context of many countries being obliged to protect it under these agreements
(it being listed in the appendixes of both). Recommendation 1.28 specifically calls
on range states to conclude an agreement under the CMS quickly and to assist Saudi
Arabia in doing so.
Shortly after the CMS COP 5, CITES met for its tenth COP in June 1997. The
parties to CITES were already familiar with the plight of the Asiatic species of the
bustard: it had long been included in the IUCN’s Red Lists and had been listed in
CITES Appendix I. A number of countries that were parties to both CITES and the
CMS (and the IUCN) now used the call for action passed by the CMS COP, as rein-
forced by the IUCN’s World Conservation Congress, to raise the issue within CITES
and bring it to the attention of other CITES parties. Realerted of the need for urgent
action, the majority of CITES parties had no direct stake in the Asiatic houbara bus-
CITES 169
tard and were thus sympathetic to further supporting its protection, as suggested by
the initiators.
As a result, the CITES COP passed Resolution 10.11 on the houbara bustard that
responded to the call for action by the CMS, as reinforced by the IUCN, without
having been specifically asked to do so. The resolution mentioned all three of the
CMS and IUCN resolutions and called on range states to take domestic action to
protect the bustard. This was not particularly unusual, because CITES often echoes
calls both from other nature conservation conventions and the IUCN. However,
in its Resolution 10.11, CITES also ‘‘calls upon all range states of the Asiatic sub-
species of the houbara bustard (chlamydotis undulata macqueenii) to review the
Draft Agreement officially circulated by the Government of Saudi Arabia and com-
municate their comments to the National Commission for Wildlife Conservation
and Development (NCWCD), Riyad, Saudi Arabia.’’ This was a remarkable action
for CITES to take because it called on CITES parties to take a highly specific action
in support of a decision taken by another institution.
Following the CMS and CITES resolutions in 1997, there was a long delay and
little progress was made on the CMS agreement. Indeed, in 1999, COP 6 of the
CMS passed a further recommendation (6.4) on the subject. However, by COP 7,
in 2002, matters were moving along more satisfactorily, as Resolution 7.7 states:
i. [The COP] Takes note of the information provided by the representative of the Kingdom of
Saudi Arabia that an updated text of an Agreement and Action Plan on the Asiatic popula-
tions of the Houbara Bustard is ready for official dissemination and comment;
ii. Takes further note that an informal meeting to review the updated text will be held some
time in early 2003; and
iii. Welcomes the information that the Kingdom of Saudi Arabia will hold a meeting of the
Range States to conclude the Agreement and Action Plan in late 2003.
Discussion and Conclusions

Although the CMS had not issued a formal request for help to CITES, it was still
influential in bringing about the action by CITES supporting the protection of the
houbara bustard under the CMS. Having been alerted by several parties and the
IUCN who employed the related CMS call for action (as reinforced by IUCN),
CITES was cognizant of the CMS decisions about the houbara bustard, mentioning
the most recent one in the preamble to its resolution. Even without a formal request,
the knowledge about the need for urgent action as expressed in the CMS decisions
changed the situation within CITES by raising awareness and enhancing support for
170 John Lanchbery
Figure 7.1
Convention on Migratory Species triggers CITES action on Houbara Bustard
action by CITES. The ensuing CITES resolution was thus a result of Cognitive Inter-
action (figure 7.1).
Interestingly, the CITES resolution calls on both CMS parties and nonparties (in
the form of range states) to take action under the CMS, rather than CITES. In most
cases of one institution assisting another, the supporting institution would call for
action within its own sphere of competence or influence. CITES might, for example,
call for trade sanctions in support of another regime. In this case, the houbara bus-
tard was already subject to trade regulation (being listed in CITES Appendix I) and
CITES was calling for action under the auspices of the CMS, rather than itself.
Furthermore, the interaction demonstrates the way nature conservation treaties
act in support of one another, with the IUCN facilitating such support both with
scientific information and via its access to both parties and institutions. It was the
IUCN Red Listing that first alerted CITES and the CMS of the need for action on
the houbara bustard. The recommendations of the IUCN World Conservation Con-
gress then reinforced the calls for action of the CMS and helped transmit them to
CITES, which responded by asking for action in the CMS.
The question remains, however, why CITES should bother to help the CMS in this
case, when the bustard was already protected by CITES. The answer, most proba-
bly, lies in the fact that CITES has a far larger membership than the CMS. In partic-
ular, more of the range states for the Asiatic houbara bustard are parties to CITES
CITES 171
than are parties to the CMS. Given that it is common for states that are not parties
to the CMS to join Agreements or memoranda of understanding concluded under it,
CITES asking for help from range states for the bustard makes considerable sense.
As can be seen from table 7.2, only four of the twenty-four range states for the hou-
bara were parties to the CMS at the end of 1997 (although several range states par-
ticipate in CMS subagreements without becoming parties to the CMS itself, such as
Iran, Russia, China, Oman, and Yemen). CITES membership in 1997 included more
than three times as many range states (thirteen). The fact that there are so many
range states for the bustard, many of which are parties to neither CITES nor the
CMS and many of which are poorer developing countries with more pressing prior-
ities than the environment, may also help to explain why it is taking so long to con-
clude an Agreement.
The Interaction with the Convention for the Conservation and Management of the
Vicuña (CCMV)
The Vicuña and the Convention Covering It

The vicuña (vicugna vicugna) is the smallest species of the South American Camel-
idae, which include the llama, guanaco, and alpaca (Torres 1987). It lives on the
high Andean plateaus. At the time of the Incas, vicuñas were an important source
of wool and meat. The main product was, however, their wool and so they were
normally captured, shorn, and released into the wild again, maintaining the popula-
tion at an estimated 1.5 million (Grizmek 1990). Vicuña numbers dropped with the
coming of the Spanish. During the nineteenth and twentieth centuries, there was a
huge commercial demand for the wool, which is very soft and can be woven into
delicate fabrics. Because the vicuña was then killed for its wool, rather than being
‘‘farmed’’ as in the time of the Incas, by 1960, vicuña numbers had fallen to only
6,000 (Torres 1987; Mendoza 1987).
Since the conclusion of the vicuña protection treaties and increased domestic
efforts to conserve them in their range states, the vicuña population had risen to
125,000 by 1990 (Nowak 1991). By 2002 numbers stood at about 150,000 (http://
www.iucn.org). The vicuña was until recently classified as vulnerable by the IUCN
in its Red Lists but has recently been downgraded to lower risk. It was originally
listed in Appendix I (endangered) of CITES although, as numbers have risen, popu-
lations have increasingly been transferred to Appendix II (threatened).
172 John Lanchbery
Table 7.2
Membership of Range States of the Houbara Bustard in CITES and CMS at the end of 1997
Range States of chlamydotis
undulata macqueenii Whether party Whether party
(from IUCN Red List) to CMS to CITES
Afghanistan Not a party Party
Armenia Not a party Not a party
Azerbaijan Not a party Not a party
Bahrain Not a party Not a party
China Not a party Party
India Party Party
Iran Not a party Party
Iraq Not a party Not a party
Israel Party Party
Jordan Not a Party Party
Kazakhstan Not a party Not a Party
Kuwait Not a party Not a party
Lebanon Not a party Not a party
Mongolia Not a Party Party
Oman Not a party Not a party
Pakistan Party Party
Qatar Not a party Not a Party
Russian Federation Not a party Party
Saudi Arabia Party Party
Syria Not a party Not a party
Tajikistan Not a Party Not a party
Turkmenistan Not a party Not a party
United Arab Emirates Not a party Party
Uzbekistan Not a Party Party
Yemen Not a party Party
Source: http://www.cites.org and http://www.cms.int.
CITES 173
The origins of the vicuña convention are in a bilateral treaty concluded in 1969
between Bolivia and Peru, in La Paz. The agreement declared that the vicuña was a
‘‘species on the way to extinction’’ (Preamble) and all commercial exploitation of
the species or products made from it was prohibited for a period of ten years. Dur-
ing the following decade, Chile and Argentina joined the agreement. In October
1979 a meeting was held in Lima at which the four parties to the agreement were
joined by Ecuador. On December 20, 1979, the treaty was amended to include the
five countries at the meeting, whose territories include all of the current natural
range of the vicuña. While the vicuña does not occur naturally in Ecuador, it was
introduced later on. The agreement was also extended indefinitely and named the
Convention for the Conservation and Management of the Vicuña (Torres 1987).
The treaty prohibits hunting of the vicuña and their live export, with exception of
those used for scientific purposes or for display in legally established zoological gar-
dens. It also bans trade in vicuña wool, hair, skins, and items manufactured from
them and the manufacturing itself within the territories of the parties, except under
special license. Licensed products nowadays bear a special CCMV logo. The parties
are obliged to establish and maintain reserves and centers for raising the vicuña. In
addition, they are obliged to conducting awareness raising and training activities.
As in the case of other wildlife agreements, the CCMV works closely both with
environmental regimes (including CITES) and with environmental groups. Coopera-
tion with the IUCN’s Species Survival Commission Specialist Group on South Amer-
ican Camelids—which serves CITES as well—is particularly close.7 The CCMV also
works closely with WWF and with UNEP, which funds some of the IUCN work
on vicuñas. Indeed, a compendium of resolutions of the Conference of the Parties
to the CCMV has an entire section devoted to relationships with the IUCN, WWF,
and TRAFFIC (Government of Argentina 2004). Over the years, the CCMV has
cooperated closely with CITES. Indeed, between 1980 and 2000 the CCMV passed
twenty-four resolutions concerning and involving CITES. Many of them concerned
listings in the CITES appendixes, but the CCMV has also repeatedly tried to im-
prove its effectiveness by having CITES ask its parties to restrict trade in vicuña
products, particularly cloth made from vicuña wool (Government of Argentina
2004).
The Interaction
Although territories of the parties to the CCMV cover the entire natural range of the
vicuña, most demand for the valuable vicuña wool, and products made from it,
174 John Lanchbery
comes from highly developed countries that are not parties to the CCMV. The very
long, often extremely rugged borders of the CCMV parties, coupled with the fact
that these countries are not wealthy, makes it hard for them to effectively control
the illegal export of vicuña wool and cloth. Controlling their import into highly
developed countries may provide effective complementary protection. Since the
1980s, the CCMV has thus repeatedly called on CITES to ask its members, which
include all relevant importing developed countries, to help it to restrict trade in
vicuña products (Government of Argentina 2004).
Interaction between the CCMV and CITES concerning vicuña wool and cloth
began in 1987 when the CCMV passed Resolution 56/87. It asked the CITES Secre-
tariat to recommend to all CITES parties that had stocks of vicuña cloth and wool
to submit a list of those stocks, as soon as possible, to the CITES Secretariat, which
would forward them to the CCMV. The CITES Secretariat acted accordingly by
issuing a notification (number 472) to CITES parties asking them to respond. The
idea was to enable an accurate assessment of globally, and legitimately, held stocks
so as to better be able to track which trades were of legally held stocks and which
were likely to be illegal.
This call was apparently not effective because, at its twelfth meeting in 1990, the
CCMV issued a reminder to CITES about declaring stocks, CCMV Resolution 97/
90. This also pointed out that all legally exported wool and cloth should bear
CCMV official markings (logos). The logo was to enable a clear distinction between
legal and illegal trades in wool and cloth. CITES responded by passing Resolution
8.11 concerning notification of stocks of wool and cloth, and markings, at its eighth
COP in Kyoto in 1992.
This too was apparently not completely effective because in 1994 the CCMV fired
off two more resolutions (133/94 and 137/94) to CITES, reminding it of CCMV
resolutions 56/87 and 97/90, again concerning stocks of wool or cloth and their
marking. The CCMV COP also passed two additional resolutions on listings in
CITES appendixes and another on wool. The former two resulted in an amendment
to CITES Resolution 8.11 at the tenth CITES COP in Harare in 1997.
Eventually, the CCMV’s point about declaring stocks of wool and cloth seem to
have been heeded, but the point about using the official CCMV logo on all cloth
would appear not to have been acted on by all parties. At CITES COP 11 in Kenya
in 2000, yet another resolution (11.6) was passed, essentially reiterating the pre-
vious resolutions. After first reminding the CITES parties of previous CCMV and
CITES resolutions, it recommends that parties should ‘‘only authorize the import
CITES 175
of vicuña cloth if the reverse bears the logotype corresponding to the country of
origin and the trade mark VICUÑA—COUNTRY OF ORIGIN or if it is cloth
containing pre-Convention wool of vicuña.’’
Discussion and Conclusion

This case shows how wildlife conservation treaties often interact so as to support
each other in pursuit of common goals. The goal of the CCMV is the conservation
of the vicuña and, to the extent that the vicuña is endangered or threatened as a spe-
cies, this is also part of the objective of CITES. The main reason for the request by
the CCMV was that the developed-country parties to CITES are the importers of
vicuña products, both legal and illegal. Their controlling imports thus helps consid-
erably in the enforcement of the CCMV, particularly because tight control over
exports from the countries of origin is hard. Improving information on stocks of
vicuña wool and cloth helped to clarify in which countries they had ended up after
export, thereby facilitating the tracking of further trade. Gaining recognition of the
CCMV logo helped in the practical enforcement of restrictions on both imports and
exports.
The interaction between the CCMV and CITES followed the causal mechanism of
Cognitive Interaction (figure 7.2). The requests of the CCMV were not supported by
any carrot or stick so that CITES was completely free in its choice of whether and
Figure 7.2
Vicuña Convention requests CITES assistance
176 John Lanchbery
how to respond. The request was easily brought to the attention of CITES parties
because it was explicitly directed at them. The secretariats and CCMV members
that were also parties to CITES transmitted the relevant CCMV decisions. These
decisions drew the attention of CITES parties to the enforcement problem concern-
ing the protection of the vicuña, which they may otherwise have ignored. Once
alerted, most CITES parties had little reason to oppose the request that implied
action only by a limited number of developed countries.
Ostensibly, the interaction between the CCMV and CITES appears to have been
effective because assistance by CITES occurred as requested and vicuña numbers
rose. However, numbers were rising prior to the interaction, and the resurgence of
the vicuña appears to have been primarily due to domestic measures to protect and
manage them. Having said this, the trade measures pursued by CITES in response to
the CCMV have probably made a limited positive contribution, but it is hard to
quantify their practical effect.
Conclusion
CITES actively and systematically strives to improve synergy with other institutions.
The Convention works particularly closely with other international nature conser-
vation regimes, as facilitated by several international and nongovernmental organi-
zations. This is partly because the nature conservation regimes were designed to
complement each other and share the common goal to conserve nature, and partly
because they have evolved closer links over time. The role of the IUCN in forging
these links has been considerable by, for example, providing information on which
species are endangered and threatened by means of its Red Lists, which are used by
all wildlife treaties.
In this study, horizontal interactions between CITES and other international insti-
tutions were either found to be synergistic or managed successfully so as to prevent
tensions from turning into open conflicts. There have notably been tensions between
CITES and the WTO, but CITES carefully manages its relationships so as to avoid
conflict and promote synergy. This is facilitated by the fact that the overlap in mem-
bership between such large global institutions is considerable, and it seems unlikely
that states would knowingly pursue one course of action in one forum and an
opposing course of action in another.
There have been political conflicts between CITES and the EU, mainly as a result
of CITES trying to improve enforcement. However, these conflicts have been
CITES 177
handled in productive ways so as to eventually create synergistic effects. As early as

1982, the EU passed legislation that enhanced the effective implementation of
CITES in the Union by subjecting it to the supranational EU enforcement powers.
When the EU abolished internal border controls in the 1980s and 1990s, however,
this potentially endangered effective implementation and enforcement of CITES
restrictions on wildlife trade, leading to rising concerns of CITES and considerable
political conflict. Eventually, the EU responded by strengthening its internal legisla-
tion and other implementation measures. In the end, the interaction thus produced
synergistic results.
A common cause of the synergistic interactions between CITES and other wildlife
treaties is that the latter try to use the greater membership of CITES to increase the
effectiveness of their enforcement, as was at least partially the case in both interac-
tions detailed here. However, in spite of the efforts by the institutions concerned, the
interactions of CITES with the CMS and the CCMV demonstrate that interaction is
often less successful in achieving tangible results ‘‘on the ground.’’ It usually results
in institutional action in the form of a COP decision or recommendation, but these
do not necessarily have the desired practical effect.
This is because there is, perhaps, too much belief in the extent to which inter-
actions between international regimes, or action by the regimes themselves, can
achieve concrete, practical outcomes, especially regarding enforcement. This is par-
ticularly so in the case of organizations such as the WCO, which is designed to
foster cooperation rather than enforcement. Ultimately, enforcement must mainly
occur domestically, as seems to be demonstrated by the CITES/CCMV case concern-
ing vicuña wool and cloth, where it appears to be domestic action—rather than re-
gime interactions—that has caused vicuña numbers to increase.
Notes
1. Throughout this chapter, the term wildlife includes both wild fauna and wild flora.
2. Personal communication from Wolfgang Burhenne of the IUCN’s International Law
Centre in Bonn.
3. The World Heritage Convention had 177 parties in spring 2004 but is not exclusively con-
cerned with wildlife conservation since it covers both cultural and natural heritage.
4. CMS 2002, and personal communications with Arnulf Muller-Helmbrecht, CMS
Coordinator.
5. Personal communications within the BirdLife Partnership.
178 John Lanchbery
6. Personal communications from the BirdLife Partnership. The Saudi Arabian government
representative in the CMS—the National Commission for Wildlife Conservation and Devel-
opment, NCWCD—is the Saudi Arabian BirdLife Partner.
7. For more details on the group see the IUCN Species Survival Commission website at http://
www.iucn.org/themes/ssc/sgs/sgs.htm#SACSR.
References
Andresen, Steinar. 1998. The Making and Implementation of Whaling Policies: Does Partici-
pation Make a Difference? In David G. Victor, Kal Raustiala, and Eugene B. Skolnikoff, eds.,
The Implementation and Effectiveness of International Environmental Commitments: Theory
and Practice, 431–474. Cambridge, MA: IIASA/MIT Press.
BirldLife International. 2000. Threatened Birds of the World. Barcelona and Cambridge:
Lynx Edicions/BirdLife International.
BirdLife International. 2001: BirdLife’s Online World Bird Database: The Site for Bird Con-
servation. Version 1.0. Cambridge: BirdLife International.
Burns, William C. 1990. CITES and the International Regulation of International Trade in
Endangered Species of Flora: A Critical Appraisal. Dickinson Journal of International Law 8
(2): 203–223.
CITES. 1994. CITES Secretariat. Strategic Plan of the Secretariat. 9th Conference of Parties,
Fort Lauderdale, Florida, 1994. CITES Doc. 9.17.
CITES. 1999. CITES Standing Committee. Synergy Between the Biodiversity-Related Con-
ventions and Relations with Other Organizations. Forty-second meeting of the Standing
Committee, Lisbon (Portugal), September 28–October 1, 1999. CITES Doc. SC.42.17.
CMS. 2002. CMS Secretariat. Guide to the Convention on the Conservation of Migratory
Species of Wild Animals, January 2002. Available at http://www.unep-wcmc.org/cms/pdf/
CMS_Guide_ Jan02_en.pdf.
Government of Argentina. 2004. Resoluciones de la Comisión Técnico-Administradora
del Convenio de la Vicuña, Reuniones Ordinarias (Analizadas y Extractadas) 1980–2000.
Available at http://www.medioambiente.gov.ar/fauna/programas/manejo/proyecto_vicuna/
convenio/default.htm.
Grizmek, Bernhard. 1990. Grizmek’s Encyclopedia of Mammals. Volume 5. New York:
McGraw-Hill.
IUCN. 2002. The IUCN Red List of Threatened Species. IUCN Species Survival Commission,
Gland, Switzerland: IUCN–The World Conservation Union. Available at http://www.redlist
.org/search/details.php?species=4702.
Lanchbery, John. 1995. Reviewing the Implementation of Biodiversity Agreements. In John B.
Poole and Richard Guthrie, eds., Verification 1995, 327–350. Boulder, CO: Westview Press/
VERTIC.
Lanchbery, John. 1998. Long-Term Trends in Systems for Implementation Review in Interna-
tional Agreements on Fauna and Flora. In David G. Victor, Kal Raustiala, and Eugene B.
CITES 179
Skolnikoff, eds., The Implementation and Effectiveness of International Environmental Com-

mitments: Theory and Practice, 57–87. Cambridge, MA: IIASA/MIT Press.
Lyster, Simon. 1985. International Wildlife Law. Cambridge: Grotius.
de Mendoza, Luis Hurtado. 1987. Notas Arqeológicas y Etnohistoricas acerca de la Vicuña
en el Antiguo Perú (Notes about the Vicuna in Old Peru). In Hernàn Torres, ed., Tecnicas
para el Manejo de la Vicuna, 13–23. IUCN Species Survival Commission, South American
Camelid Specialist Group, UNEP. Cambridge: IUCN.
Mulliken, Teresa. 2002. Wildlife and Livelihoods. World Conservation 3 (special issue on
‘‘The Species Trade—CITES in the New Millennium’’), 14–15.
Nowak, Ronald M. 1991. Walker’s Mammals of the World. 5th ed. Baltimore: Johns Hop-
kins University Press.
Reeve, Rosalind. 2002. Policing International Trade in Endangered Species: The CITES
Treaty and Compliance. London: Earthscan.
Sand, Peter H. 1997. Commodity or Taboo? International Regulation of Trade in Endan-
gered Species. In Helge Ole Bergesen, Georg Parmann, and Øystein B. Thommessen, eds.,
Green Globe Yearbook 1997, 19–36. Oxford: Fridtjof Nansen Institute and Oxford Univer-
sity Press.
Torres, Hernàn. 1987. Preface. In Hernàn Torres, ed., Tecnicas para el Manejo de la Vicuna,
5–7. IUCN Species Survival Commission, South American Camelid Specialist Group, UNEP.
Cambridge: IUCN.
Wijnstekers, Willem. 2003. The Evolution of CITES: A Reference to the Convention on In-
ternational Trade in Endangered Species of Wild Fauna and Flora. 7th ed. Geneva: CITES
Secretariat.
8
Interactions between the World Trade Organization
and International Environmental Regimes
Alice Palmer, Beatrice Chaytor, and Jacob Werksman
The international trading regime governed by the World Trade Organization

(WTO) interacts with many international environmental regimes, which also regu-
late international trade. The WTO is often a source of the interaction, invoking
reactions from international environmental regimes in the design and implementa-
tion of rules that are responsive to WTO prescriptions. The vast number of WTO
members, the institution’s economic significance, and its unparalleled ability to en-
force its rules through its rigorous dispute-settlement mechanism, contribute to the
WTO’s effectiveness as a source of interaction. Nevertheless, the WTO is also a tar-
get of influence by international environmental regimes, which are typically more
proactive in seeking to inform and cooperate with the WTO. Where the WTO con-
stitutes the source of interinstitutional influence, its primary objective of facilitating
free trade has conflicted with the principal objectives of environmental regimes,
which are leading to compromises in the design or implementation of the environ-
mental regimes aimed at accommodating WTO rules. The mere possibility of a
WTO challenge can inhibit negotiations and the implementation of measures under
the international environmental regimes.
This chapter examines the nature and effects of interaction between the WTO and
two international environmental regimes in particular: the Cartagena Protocol on
Biosafety (‘‘Biosafety Protocol’’) and the International Commission for the Conser-
vation of Atlantic Tunas (ICCAT). Both the Biosafety Protocol and the ICCAT serve
as valuable illustrations of the dynamics of interaction of international environmen-
tal regimes with the WTO from negotiation of commitments (Biosafety Protocol)
through to their implementation (ICCAT). The chapter commences with a descrip-
tion of the WTO, followed by a summary of the experience of interaction between
the WTO and environmental regimes in general. Subsequently, the interactions be-
tween the WTO and the Biosafety Protocol and ICCAT are studied in depth. The
182 Alice Palmer, Beatrice Chaytor, and Jacob Werksman
chapter concludes with general observations about the interaction between the
WTO and the two environmental regimes.
Introduction to the World Trade Organization
The WTO is an intergovernmental organization established in 1995 to provide a

common institutional framework for the conduct of trade relations formerly gov-
erned by the 1947 General Agreement on Tariffs and Trade (GATT) and related
instruments (WTO 1994; Jackson 1997; Trebilock and Howse 1999). The WTO
aims to liberalize markets, recognizing the need to make ‘‘use of the world’s re-
sources in accordance with the objective of sustainable development’’ and to ‘‘pro-
tect and preserve the environment . . . in a manner consistent with [the members’]
respective needs and concerns at different levels of economic development’’ (WTO
1994, preamble). As of early 2005, the WTO had a membership of 148 countries
and customs territories (http://www.wto.org). The WTO administers the multilateral
agreements regulating the international trade in goods and services and the protec-
tion of intellectual property rights and provides a forum for the negotiation of new
trade rules. The WTO agreements are backed by a compulsory dispute-settlement
system—whereby disputes can be referred to an ad hoc arbitration panel of trade
experts and, on appeal, to a permanent Appellate Body of seven independent
trade jurists—with the WTO membership having the ability to authorize bilateral
trade sanctions (known as suspension of concessions).
The WTO agreements will interact with any environmental regulation that has an
impact on the international trade in goods and services among its members, includ-
ing those regulations enacted pursuant to multilateral environmental agreements
(MEAs). The three main WTO agreements that have been of particular relevance
to international environmental regimes are the General Agreement on Tariffs and
Trade 1994 (GATT), the Agreement on Technical Barriers to Trade (TBT Agree-
ment), and the Agreement on Sanitary and Phytosanitary Measures (SPS Agreement)
(WTO 1994, Annex 1A). At the most basic level, all three agreements share the
common purpose of ensuring that measures that affect trade in products do not dis-
criminate on the basis of a product’s country of origin (known as ‘‘national’’ and
‘‘most-favored-nation’’ treatment), and that these measures are no more trade-
restrictive than is necessary to achieve the purpose for which they were designed
(‘‘proportionality,’’ e.g., GATT Art. XX, SPS Art. 2.2, TBT Art. 2.2). Each agree-
ment has detailed rules and a growing body of practice that develops these disci-
plines further.
WTO and International Environmental Regimes 183
The WTO agreements also anticipate the need to take into account other existing
international agreements, such as MEAs, and other relevant state practice. Both the
SPS and the TBT Agreements make reference to international standards developed by
competent international organizations operating outside the WTO system. Under the
SPS and the TBT Agreements, a WTO member is generally required to base its mea-
sures on international standards (SPS Art. 3.1; TBT Art. 2.4). To date, no MEA has
been recognized as an ‘‘international standard’’ under the SPS or the TBT Agreements.
The WTO’s institutional framework comprises its governing body, the General
Council, and several other councils and committees that are supported by the
WTO Secretariat in Geneva. The principal organ responsible for trade and environ-
ment issues at the WTO is the Committee on Trade and Environment (CTE). Since
the WTO’s establishment, the CTE has had the mandate to explore the relationship
between the WTO and MEAs. In the CTE, and other WTO organs dealing with
environmental matters, members have discussed a range of trade and environment
issues. These include the application of the WTO rules to trade measures taken pur-
suant to MEAs (Abdel Motaal 2001); the application of WTO rules to measures
based on nonproduct-related process and production methods (PPMs) (traditionally
viewed as WTO-inconsistent) (Charnovitz 2002); environmental labeling (especially
with respect to genetically modified organisms) (Abdel Motaal 1999); the relevance
of the precautionary principle to risk assessments based on scientific evidence (par-
ticularly in the context of the SPS Agreement) (Bohanes 2002); and the environmen-
tal impacts of certain subsidies, especially fisheries subsidies (Chang 2003; see also
chapter 6).
Most observers acknowledge the usefulness of the CTE’s work in promoting a
better understanding of the WTO-MEA relationship, and in acknowledging the le-
gitimate role of MEAs in promoting environmental objectives. However, the CTE’s
work has thus far been general and inconclusive, other than recognizing that inter-
national trade rules and international environmental rules should be designed and
implemented in a manner that is ‘‘mutually supportive’’ (WTO, 1996, para. 167).
The CTE has been widely criticized for failing to produce any conclusions or recom-
mendations of a substantive nature that would, for example, instruct the WTO’s
dispute-settlement system on how to deal with a conflict should one arise (Charno-
vitz 1997).
WTO members convene a ministerial conference approximately every two years
(WTO 1994, para. 6). At the fourth WTO Ministerial Conference in Doha in No-
vember 2001, the WTO membership agreed to a new round of trade negotiations
in a number of areas, including on ‘‘the relationship between existing WTO rules
and specific trade obligations set out in multilateral environmental agreements

(MEAs)’’ (WTO 2001c, para. 31(i)). The mandate is both vague and restrictive,
with the negotiations being limited in scope to ‘‘the applicability of such existing
WTO rules as among parties to the MEA in question’’ (WTO, 2001c, para. 31(i)).
Nevertheless, the mandate does suggest that for the first time the WTO may produce
substantive rules aimed directly and intentionally at trade-related measures con-
tained in MEAs to which its members are also parties.
The fourth WTO Ministerial Conference also encouraged ‘‘efforts to promote
cooperation between the WTO and relevant international environmental organiza-
tions’’ (WTO 2001c, para. 6) and launched negotiations between the members on
‘‘procedures for regular information exchange between MEA secretariats and the
relevant WTO committees, and the criteria for the granting of observer status’’
(WTO 2001c, para. 31(ii)). The Doha mandate stems from existing practice in the
CTE, which has already granted observer status to some intergovernmental environ-
mental organizations (WTO 2004) and hosts meetings with MEA secretariats to dis-
cuss issues relevant to the WTO and MEAs (Shaw and Schwartz 2002).
WTO Interactions with International Environmental Regimes
There is a wary coexistence between the WTO and MEAs (international environ-
mental regimes), which, like the WTO, regulate international trade (Sands 2003;
Schoenbaum 2002; Nordstrøm and Vaughan 1999). The interaction between the
WTO and environmental regimes is generated by differences in regime objectives
and institutional features designed to achieve those objectives. The WTO is designed
to promote free trade; the environmental regimes in varying degrees require or au-
thorize trade restrictions in order to discourage the production and consumption of
specific products with negative environmental consequences. The WTO agreements
are backed by a compulsory dispute-settlement system with the ability to autho-
rize bilateral trade sanctions, while the arrangements for enforcement within most
MEAs are looser and less binding (WTO 2001a). Membership of the WTO and
MEAs substantially overlaps since each regime aims for universal membership.
Table 8.1 contains a summary of the interactions between the WTO and five
international environmental regimes: the Montreal Protocol (regulating trade in
ozone-depleting substances), the Biosafety Protocol (regulating trade in ‘‘living’’
genetically modified organisms), the Basel Convention (regulating trade in haz-
ardous waste), ICCAT (regulating trade in Atlantic tuna and other fish), and the
Table 8.1
Interactions between the WTO and international environmental regimes
Montreal Protocol on Substances Granted exemptions from trade restrictions to
That Deplete the Ozone Layer nonparties complying with the Montreal Protocol,
thereby avoiding WTO prohibitions on discrimination
on the basis of a product’s country of origin (national
and most-favored-nation treatment)
Did not apply planned restrictions on trade in
products produced with ozone-depleting substances

partly because WTO rules have traditionally been
understood to prohibit trade restrictions based on
nonproduct-related process and production methods
(PPMs)
Has provided a model for WTO of admissible trade
restrictions for environmental purposes in a

multilateral framework
Did not restrict trade in products that are obsolete
because of bans on ozone-depleting substances due, to

a large extent, to WTO disciplines
Was limited in its effectiveness because WTO
disciplines limited Montreal Protocol parties’ use of

trade measures with respect to products that are
obsolete because of bans on ozone-depleting substances
Cartagena Protocol on Biosafety Was used to block attempts to regulate biosafety
under the WTO
Includes preambular language attempting to reconcile
the Biosafety Protocol with the WTO Agreements

Adapts its risk-assessment procedure from WTO’s
SPS Agreement, which potentially constrains the use of

trade measures under the Biosafety Protocol
Basel Convention on the Control Granted exemptions from trade restrictions to
of Transboundary Movements nonparties complying with the Basel Convention,
of Hazardous Wastes and Their thereby avoiding WTO prohibitions on discrimination
Disposal on the basis of a product’s country of origin (national
and most-favored-nation treatment)
International Commission for Granted exemptions from trade restrictions to
the Conservation of Atlantic nonparties complying with ICCAT, thereby avoiding
Tunas (ICCAT) WTO prohibitions on discrimination on the basis of a
product’s country of origin (national and most-
favored-nation treatment)
Convention on the Conservation Granted exemptions from trade restrictions to
of Antarctic Marine Living nonparties complying with CCAMLR, thereby
Resources (CCAMLR) avoiding WTO prohibitions on discrimination on the
basis of a product’s country of origin (national and
most-favored-nation treatment)
Convention on the Conservation of Antarctic Marine Living Resources (CCAMLR)

(regulating trade in Antarctic marine living resources). Chapters 3, 4, 6, and 7 of
this volume address further interactions between the WTO and international envi-
ronmental regimes. The WTO and the Conferences of the Parties of the various
MEAs each have the mandate to act in areas that lie in the other’s jurisdiction.
Thus their ‘‘influence’’ over each other, though implicit, is as powerful as if it were
expressly stated. Vertical interactions between the WTO and EU legal instruments
are beyond the scope of this chapter. Some such interactions are analyzed in chapter
11 of this volume.
The GATT/WTO consistency of trade restrictions has been a concern that has
constrained the respective rules and regulations of some environmental regimes (Bio-
safety, Montreal, ICCAT). The effect has been viewed as ‘‘chilling,’’ disrupting or
slowing negotiation processes (Montreal, Biosafety), and limiting the composition
and reach of trade measures (Biosafety, Basel), and their further development and
application (Montreal) (Interaction through Commitment). Although a challenge to
a MEA measure has never been decided by the WTO dispute-settlement system,
the threat of such a WTO challenge further influences the design of rules under the
environmental regimes. The parties to MEAs remain acutely conscious of this inter-
action and have at times refrained from taking trade-related measures when imple-
menting their commitments (Montreal) (Behavioral Interaction).
The response from the environmental regimes has been twofold. The first has
been to deliberately avoid the possibility of conflict. Some MEAs stipulate that trade
restrictions should not be imposed on nonparties where they can demonstrate com-
pliance with the relevant MEA rules (ICCAT, Basel, CCAMLR, Montreal). Indeed,
both the ICCAT and CCAMLR regimes go out of their way to provide a multistep
process for nonparties to comply with their regulations before action is taken
against them. On their side, some WTO members (parties or nonparties of MEAs)
have chosen not to enforce their rights under the WTO where their products have
been the subject of MEA-based trade restrictions (ICCAT, Montreal), implying tacit
acceptance of the MEA strategies. The second response has been to initiate and
maintain institutional coordination between respective secretariats. MEA secretar-
iats (CBD, Montreal, ICCAT) have been granted observer status at meetings of the
WTO’s CTE, and in turn, the WTO Secretariat has attended sessions of the Confer-
ences of the Parties of various MEAs. Since 1997, MEA secretariat representatives
have participated in information exchanges in the CTE, tabling papers about trade
measures used in MEAs. The process of information exchange has also involved
requests by respective MEA membership for expert advice from the GATT/WTO
Secretariat (Montreal) (see, in general, Shaw and Schwartz 2002).
The response from the WTO has been reflected in its ‘‘judicial’’ decisions, through
the settlement of disputes, and in its political agenda. While there has never been a
WTO decision based on a challenge to a MEA-related trade measure, the WTO’s
‘‘judicial’’ organ—the Appellate Body—has taken into account existing interna-
tional agreements and state practice when clarifying relevant provisions of the
GATT. Indeed, the Appellate Body made reference to a number of MEAs when
clarifying the meaning of ‘‘exhaustible natural resources’’ under one of the ‘‘envi-
ronmental’’ exceptions in GATT Article XX. When analyzing this exception, the
Appellate Body looked to MEAs as one source of evidence of the ‘‘contemporary
concerns of the community of nations about the protection and conservation of the
environment’’ (WTO 1998b, para. 129). The WTO’s active management of the re-
lationship with MEAs in the context of dispute settlement has been in contrast to
the historically passive response of its political organs. While some environmental
regimes have been cited in the WTO as examples of properly functioning, multilat-
erally negotiated, and narrowly drawn exceptions to free-trade rules (CCAMLR,
Montreal: Cognitive Interaction) (e.g., WTO 2002, para. 5; 2000b, para. 22), no
political decisions have yet been made. The new round of WTO negotiations on
the relationship between MEAs and WTO rules, however, could lead to a more con-
structive relationship between the WTO and international environmental regimes.
WTO Interactions with the Cartagena Protocol on Biosafety
Cartagena Protocol on Biosafety

The Cartagena Protocol on Biosafety to the Convention on Biological Diversity (the
‘‘Biosafety Protocol’’) was adopted in 2000 and entered into force on September 11,
2003. The Protocol governs the international trade between parties in living organ-
isms that have been genetically modified through the use of modern biotechnology
(‘‘living modified organisms’’ or ‘‘LMOs’’)—as opposed to the use of traditional or
conventional techniques. LMOs are commonly referred to as genetically modified
organisms, or GMOs, although GMOs could include both living and dead geneti-
cally modified organisms (Glowka, Burhenne-Guilmin, and Synge 1994, 45). The
Biosafety Protocol was negotiated under the auspices of the Convention on Biologi-
cal Diversity (CBD) and is served by the secretariat to the Convention in Montreal
(http://www.biodiv.org/biosafety).
The Biosafety Protocol’s primary objective is protection from the risks that LMOs
may pose to biological diversity, and to human health (Art. 1), providing a basis for
policymakers in a country of import to regulate against such risks. It puts in place
standards of treatment for import, export, and shipment of LMOs that require
exporting countries to ensure that importing countries are informed of a proposed
transfer of LMOs, through a system of ‘‘Advance Informed Agreement.’’ The
Biosafety Protocol authorizes importers to impose greater constraints on LMOs
destined for release into the environment, primarily seeds, but is less stringent with
regards to LMOs destined for food or feed, or for processing (‘‘LMO-FFPs’’ or
‘‘GM commodities’’). With regard to LMOs destined for release into the environ-
ment, parties of proposed import may, on the basis of a risk assessment, approve
the import, with or without conditions, or impose import bans (Art. 10). With re-
gard to GM commodities, exporters are required to inform other parties of decisions
to place GM commodities on their domestic markets; other parties may make deci-
sions on the import of GM commodities in accordance with the Biosafety Protocol’s
objective (Art. 11).
Shipments of LMOs are subject to certain handling, transport, packaging, and
identification requirements, which vary depending on the category of LMO. In
particular, shipments of LMOs intended to be introduced into the environment or
subject to contained use must be accompanied by documentation identifying them
as LMOs and, in the case of GM commodities, accompanying documentation is
required to state that the shipment ‘‘may contain’’ LMOs and that they are not in-
tended for introduction into the environment (Art. 18).
A significant part of the Biosafety Protocol’s rules provide standards for the con-
duct of risk assessment on LMOs (Art. 15 and Annex III). The Biosafety Protocol
also contains numerous references to the ‘‘precautionary approach’’ (e.g., Art. 1)
and to the use of socioeconomic data in the risk-assessment process (e.g., Art. 26).
Finally, the Biosafety Protocol requires the proponent of export to pay for any risk
assessment that may be demanded by the importing country if the latter so requests
(Art. 15).
As an instrument that provides a legal basis for the regulation of the trade in prod-
ucts, the Biosafety Protocol necessarily has the potential to interact with the WTO
rules. The Biosafety Protocol’s Advance Informed Agreement and risk-assessment
procedures and identification requirements could create trade restrictions subject to
the WTO agreements (Eggers and Mackenzie 2000, 539–540; see also Charnovitz
2000; Phillips and Kerr 2000). While it is as yet unclear which of the WTO agree-
ments might apply to a Protocol trade measure, the GATT and the TBT and SPS
Agreements contain the same basic provisions that require trade measures to be
nondiscriminatory, and to be no more trade restrictive than necessary to achieve
their objective (‘‘proportionate’’). The TBT and SPS Agreements also require harmo-
nization to international standards and the SPS Agreement requires trade measures
to have a scientific basis.
The Interaction with the WTO

The interaction between the Biosafety Protocol and the WTO results from an over-
lap in means and a divergence of objectives. Both regimes seek to create interna-
tional standards for the regulation of the international trade in products that are
associated with risks to the environment of the importer. The regimes diverge funda-
mentally in their principal objectives. The Biosafety Protocol is primarily concerned
with protecting the rights of the importer to be informed of and to regulate LMOs,
while the WTO agreements are primarily concerned with the rights of the exporter
to ensure that its products are treated in a nondiscriminatory and rational (scientifi-
cally based) manner. Interaction also occurs by virtue of overlapping member-
ships. As of January 2003, 98 of the 117 signatories and parties to the Biosafety
Protocol were also WTO members. In addition, 10 further signatories and parties
to the Protocol were also WTO observers, awaiting WTO membership (http://
www.biodiv.org/biosafety and http://www.wto.org).
The relationship with the WTO was at the heart of the Biosafety Protocol nego-
tiations and was largely responsible for the collapse of the negotiations in Cartagena
during the first meeting that was intended to complete and adopt a text in 1999
(Pomerance 2000; see also Eggers and Mackenzie 2000). WTO rules were relevant
to many of the areas of disagreement in the Biosafety Protocol negotiations, such
as scope and GM commodities, and were considered specifically in the context of
the discussions on the Protocol’s relationship to other international agreements and
other trade-related matters.
With respect to scope and GM commodities, there was significant discussion
about the extent to which GM commodities would be covered by the Biosafety Pro-
tocol and the manner in which they should be identified (Pomerance 2000). The
Miami Group (including Argentina, Australia, Canada, Chile, the United States,
and Uruguay) and other LMO-exporting countries fought to exclude GM com-
modities from the Advance Informed Agreement procedure, arguing that the proce-
dure should apply only to those LMOs that were intended to be introduced into the
environment and therefore posed a risk to biological diversity. The Miami Group
also argued that the risk-assessment requirements underlying the Advance Informed
Agreement procedure should be based on ‘‘sound science’’ and in accordance with
WTO rules, while the European Union (EU) pushed for risk assessment based on
the ‘‘precautionary principle.’’ As for identification of GM commodities, the Miami
Group insisted that the identification requirement be limited to a statement that the
shipment ‘‘may contain’’ LMOs, allowing exporters to mix LMO and non-LMO
products in the same shipment (Newell and Mackenzie 2000, 315–316).
Up until and during the negotiating meeting in Cartagena, it had been proposed
that the Biosafety Protocol include an article on the relationship with other interna-
tional agreements (draft Art. 31). The Miami Group wanted a ‘‘savings clause’’ that
would preserve WTO obligations and effectively subordinate the Biosafety Protocol
to the WTO agreements, a position unacceptable to most developing countries and
the European Union. In the course of the negotiations, the discussion of the relation-
ship between the Biosafety Protocol and other international agreements was merged
with the discussion of other ‘‘trade-related’’ matters concerning nonparties and
nondiscrimination (UNEP 1999b). The implications of these provisions for the inter-
national trading regime established by the WTO were the principal basis for dis-
agreement between the negotiators in Cartagena. Trade-related matters served to
stall the adoption of the Biosafety Protocol as some delegations sought full debate
and consideration of the implications for the WTO (Pomerance 2000, 618–619;
Falkner 2002).
The influence of the trade regime on the development of the Biosafety Protocol
has been felt through the direct intervention of the LMO-exporting countries, repre-
sented in large part by the Miami Group, in the negotiating process. However, it
should be noted that most delegations in the Biosafety Protocol negotiations were
dominated by their environment ministries, which were reluctant to defer to their
trade counterparts. The influence common to both regimes was industry, which par-
ticipated directly in some delegations and contributed independently to the Biosafety
Protocol negotiations (personal communication with Biosafety Protocol delegate).
A WTO Secretariat representative attended negotiation meetings of the Biosafety
Protocol (e.g., UNEP 2000, 6, 21). In some instances, a representative from the
WTO addressed the Biosafety Protocol negotiators and informal consultations be-
tween the CBD and WTO were initiated. However, this interaction had little or no
impact on the negotiations, and WTO bodies and trade-related organizations deal-
ing with issues specific to the Biosafety Protocol were notably absent (personal com-
munication with Biosafety Protocol delegate).
The WTO’s influence on the Biosafety Protocol is apparent in the adaptation of
WTO principles and conflict avoidance. The main effect is reflected in the text of
the Biosafety Protocol in two main ways: (1) the risk-assessment procedures set out
in Annex III of the Biosafety Protocol to a large extent follow the approach in the
WTO SPS Agreement; (2) the negotiators sought, as far as possible, to direct policy-
makers through preambular paragraphs to avoid conflict by interpreting the
Biosafety Protocol and the WTO agreements in a ‘‘mutually supportive’’ manner.
Perhaps the most significant outcome of the interaction was the narrowing of the
scope of the Biosafety Protocol’s main regulatory instrument (Advance Informed
Agreement procedures) from LMOs generally to only those other than GM com-
modities (Pomerance 2000). It appears that this has effectively carved out from the
Biosafety Protocol the commodities that are most significant from an economic or
trade perspective: those destined for use as food or feed, or for processing (Newell
and Mackenzie 2000, 315).

The influence of the WTO on the Biosafety Protocol constituted Interaction through
Commitment. WTO members were bound by WTO rules on nondiscrimination,
proportionality, scientific basis, harmonization, and transparency. As a consequence,
free-trade interests in the negotiations on the Biosafety Protocol were strengthened
because they could refer to the WTO and claim consistency. As a result, the rules
of the Biosafety Protocol regarding Advance Informed Agreement, risk assessment,
and identification requirements were designed so as to minimize the potential for
conflict with WTO rules (figure 8.1).
The potential for future interaction between the Biosafety Protocol and the WTO
in the administration, implementation, and development of rules is difficult to ana-
lyze in the abstract: the Biosafety Protocol does not specify the precise products it
seeks to regulate, what specific risks it is intended to guard against, and what mea-
sures it will authorize parties to take against those risks. Countries on both sides of
the debate have made statements praising the Biosafety Protocol for having dove-
tailed neatly into existing trade rules (Phillips and Kerr 2000, 68; Falkner 2000,
311). However, recognizing that the adoption of the Biosafety Protocol is a signifi-
cant step toward reconciling trade and environmental interests, many commentators
Figure 8.1
WTO influence on the design of the Biosafety Protocol
acknowledge that the preambular language and other compromises made on trade-
related matters are imperfect solutions (Falkner 2000, 312–313; Eggers and Mack-
enzie 2000, 542–543; see also Qureshi 2000, 835, 853; Phillips and Kerr 2000, 69,
74).
Ultimately, it would appear that the terms of the Biosafety Protocol and WTO
agreements do not, in and of themselves, conflict and that, with or without savings
language, an international adjudicator could interpret their provisions to comple-
mentary effect (Safrin 2002). One important indicator of whether the Biosafety Pro-
tocol has achieved a successful accommodation of WTO rules will be the rate of
ratification by major LMO exporters, represented largely by the Miami Group. It
should be noted that the United States cannot become a party to the Protocol with-
out first becoming a party to the CBD (CBD Art. 32(1)), although it is likely to
voluntarily comply with the Protocol until then (Phillips and Kerr 2000, 65). The
optimism and skepticism surrounding the compromise language of the Biosafety
Protocol will be tested if measures justified under the Biosafety Protocol are chal-
lenged under the WTO (Eggers and Mackenzie 2000). In this respect, the outcome
of the WTO complaint over the EU de facto moratorium on GMO imports initiated
by the United States and other countries in 2003 may prove important—although
the moratorium is technically not a Biosafety Protocol measure (see also chapter
11).
In the future, the Parties to the Biosafety Protocol are expected to consider
whether and how it should cooperate with other intergovernmental organizations.
There has already been a growing degree of institutional coordination, through re-
ciprocal representation at meetings, and coincident representation at meetings in
third ‘‘standard-setting’’ institutions relevant to LMOs (such as the Codex Alimen-
tarius Commission, organs of the International Plant Protection Convention, and
the Office International des Epizooties) (Buckingham and Phillips 2001). The
WTO’s mandate from the fourth Ministerial Conference to negotiate on the WTO-
MEA relationship might present further opportunities for clarifying the relationship
between the WTO and the Biosafety Protocol.
Finally, it is worth noting in passing that, in addition to being a source of interac-
tion with the Biosafety Protocol, the WTO was also a target. The negotiation of the
Biosafety Protocol overlapped with the WTO’s third Ministerial Conference in Seat-
tle, where some WTO members sought to create a working group on biotechnology
in part to allow the WTO to seize exclusive jurisdiction over the issue (Buckingham
and Phillips 2001; Falkner 2000, 305). Supporters of the Biosafety Protocol rejected
the initiative, citing the ongoing negotiations under the CBD as the appropriate
forum for resolving these issues (Tapper 2000; UNEP 1999a, para. 9). From the
point of view of those within the WTO with expansionist ambitions, the result was
disruptive.
WTO Interactions with the International Commission for the Conservation of

Atlantic Tunas
International Commission for the Conservation of Atlantic Tunas

The International Commission for the Conservation of Atlantic Tunas (ICCAT) is an
intergovernmental organization, consisting of over thirty contracting parties (includ-
ing Canada, China, the EC, Japan, and the United States), governed by an interna-
tional convention signed in 1966 (the Convention) (http://www.iccat.es). Established
in Madrid in 1969 when the Convention entered into force, the Commission is re-
sponsible for the conservation of tunas and tunalike fish stocks in the Atlantic
Ocean and adjacent seas, including the Mediterranean Sea. The Commission collects
statistics from all tuna fisheries and conducts stock assessments, which it uses in
adopting regulatory measures for the management of various stocks of tunas. Such
regulatory measures include catch quotas, time/area closures, size limits, and import
bans. In recent years, the Commission has focused its regulatory regime on the
management of Atlantic bluefin tuna (Thunnus thynnus) and swordfish (Xiphias

gladius), both of whose stocks have been overexploited.
The Commission first adopted regulatory measures (minimum size and catch lim-
its) for the entire Atlantic in November 1975. As the depletion of the western Atlan-
tic stock became apparent, the Commission recommended at its 1981 meeting that
fishing for western Atlantic bluefin tuna be prohibited, except for a minor quantity
for scientific monitoring purposes. This allowance was shared by three bluefin tuna
fishing countries (Canada, Japan, and the United States) and according to ICCAT
has been well monitored (UNEP 1998). In 1993, the Commission adopted a recom-
mendation that prohibits bluefin tuna fishing by large pelagic longline fishing vessels
(over 24 meters in length) during June and July (the bluefin spawning season) in
the Mediterranean. In 1996, it adopted a recommendation to prohibit purse seine
fishing in August in the Mediterranean Sea to protect small fish. In 1994, it recom-
mended, for the first time, a catch reduction for the eastern stock, which entered
into effect in 1995 (WTO 1998a).
Another of the Commission’s priority objectives has been to stop illegal, unregu-
lated, and unreported fishing carried out by vessels flagged by both parties and
nonparties. ICCAT is particularly concerned about the fishing activities of vessels
registered in countries operating ‘‘open registries’’ (‘‘flags-of-convenience’’ states),
since such vessels have traditionally had little or no control exercised by their re-
spective flag states over their fishing activities.
The regulatory measures recommended by ICCAT are binding only on its parties
(ICCAT Art. VIII.2). However, since 1995, ICCAT measures may bind a category
known as ‘‘cooperating parties.’’ These are nonparties that ‘‘voluntarily fish in con-
formity with the conservation decisions of ICCAT’’ (Resolution 94-6 and Recom-
mendation 03-20). Such status is granted for a period of one year and is subject to
annual review. In 1999, Mexico and Chinese Taipei were granted status as cooper-
ating parties, and Mexico subsequently became a party (http://www.iccat.es).
ICCAT has passed trade measures because even stringent regulations for sus-
tainable management of fisheries may be undermined in the absence of effective
enforcement measures. International experience has shown that lack of appropri-
ate disincentives cause irresponsible vessel operators to overfish and thereby deplete
fisheries resources (Stone 1997; Downes and Van Dyke 1998). Overfishing may re-
sult in the collapse of entire fisheries. ICCAT is therefore concerned to enforce the
implementation of its fisheries-management regime by both parties and nonparties,
through a package of specific actions. ICCAT has recorded that the unreported
catches of flags-of-convenience vessels of nonparties make its scientific studies pro-

gressively more inaccurate and difficult to carry out. In addition, the increase in
such fishing operations may end up discouraging parties from effectively implement-
ing the regulations (WTO 1998a, 3).
The Interaction with the WTO

The interaction between ICCAT and the WTO is, again, based on different objec-
tives of the institutions involved and their overlapping memberships. ICCAT’s pri-
mary aim is the conservation and sustainable management of species of tuna and
other similar fish stocks in a particular geographic area. Enforcement of its manage-
ment measures to serve this goal necessarily implies limits on the fishing efforts of
the countries under its jurisdiction. The WTO, on the other hand, aims to remove
restrictions and limits on the trade and economic activities of its members. It there-
fore disciplines trade restrictions such as import bans or quotas. As of January
2003, twenty-six of the thirty-four ICCAT parties were WTO members. These are
particularly concerned about respecting WTO principles in the design of ICCAT
trade measures. For example, both the EU and the United States, in enacting their
import bans, comment on the WTO compatibility of those measures (EU Council
2000). There are also some nonparties (which are WTO members) that ICCAT con-
siders to be key players because of their involvement in tuna fisheries. These are
Turkey, Cyprus, Malta, Belize, Denmark (Faroe Islands), the Philippines, and Thai-
land (WTO 2001b). Turkey partially complies with ICCAT management regula-
tions. Other countries such as Taiwan, being ‘‘cooperating parties,’’ comply with
ICCAT regulations to the same extent as parties.
While the Convention itself does not contain any specific trade measures, resolu-
tions adopted by the Commission do contain trade restrictions. For example,
ICCAT adopted two action plans for bluefin tuna, which came into force in 1995,
and for swordfish, which entered into force in 1996 and were subsequently revised
(Resolutions 94-3 and 95-13 replaced by 03-15). Pursuant to these action plans, the
Commission will recommend that parties prohibit imports of bluefin tuna and
swordfish and their products in any form, from parties or nonparties, whose vessels
it is determined are fishing the species in a manner inconsistent with ICCAT’s regu-
latory regime. Paragraph (f) in each action plan provides that ‘‘the Commission will
recommend that Contracting Parties take nondiscriminatory trade restrictive mea-
sures, consistent with their international obligations.’’ For instance, import bans
with respect to bluefin tuna took effect regarding three nonparties in 1997 (Belize
and Honduras: Recommendation 96-11 and Panama: Recommendations 96-12);

and regarding one party in 2000 (Equatorial Guinea: Recommendation 99-10). Sim-
ilarly, to make effective its Resolution Concerning Illegal, Unreported and Unregu-
lated Catches of Tuna by Large-Scale Longline Vessels (Resolution 98-18 replaced
by 03-15), ICCAT has recommended a ban on imports of tuna from several coun-
tries (e.g., Belize, Cambodia, Honduras and St. Vincent and the Grenadines: Recom-
mendation 00-15). Furthermore, where the vessel of a nonparty has onboard species
subject to ICCAT conservation measures, landings and transshipments of all its fish
are subject to inspection when it voluntarily enters a port of a party. Unless the ves-
sel establishes that the fish were caught outside the Convention Area or in compli-
ance with the relevant ICCAT conservation measures and requirements under the
Convention, its landings and transshipments of relevant species will be prohibited
(Recommendation 98-11).
Under ICCAT, all parties must also require all imported bluefin tuna to be accom-
panied by an ICCAT statistical document that is defined by the Commission. The
program has applied to frozen fish since 1993 and to fresh bluefin tuna since 1994.
Document details include the name of the country issuing the document, the name
of the exporter and importer, and the area of the harvest. In principle, all bluefin
tuna products (not only from the Atlantic) must now be accompanied by such a
document validated by a government official. This applies to products from parties
and nonparties (WTO 1998a, 4). Only products accompanied by documents con-
taining seals and signatures registered with the Commission are accepted for import.
Both the import bans implemented by ICCAT parties and the ICCAT documenta-
tion requirements are in potential conflict with WTO rules. Import bans against
nonparties such as Belize or St. Vincent and the Grenadines could violate GATT
and TBT nondiscrimination provisions or the GATT Article XI prohibition on the
imposition of quantitative restrictions against the products of other WTO members.
Technical documentation requirements might raise issues relevant to the TBT Agree-
ment. Furthermore, the documentation requirements implicitly rest on an essential
distinction between bluefin tuna fish products. Differences in the manner in which
the imported and domestic products have been harvested or produced are a central
concern of ICCAT’s conservation measures. However, some WTO members might
argue that such nonproduct-related process and production methods (PPMs) are not
a legitimate basis for distinguishing between products under the WTO rules. Also,
the de facto application of the discriminating measures to areas outside ICCAT’s
jurisdiction (tuna caught outside the Atlantic) raises further questions about their
WTO compatibility, particularly where the measures are taken in relation to non-
parties. Unless those countries have expressly waived their WTO rights, they can-
not be said to have forgone them through their participation in the documentation
program.
Finally ICCAT import bans appear to authorize a blanket ban on all target spe-
cies imported from a particular country. Such import bans do not, therefore, distin-
guish fish caught by a vessel complying with ICCAT conservation measures from
fish caught by a noncomplying vessel flying the same flag. This could be viewed as
arbitrary and discriminatory under GATT rules. In the U.S.–Shrimp case, the WTO
Appellate Body noted that the U.S. measure may have been better tailored to pro-
vide for shipment-by-shipment certification (WTO 1998b). However, ICCAT im-
port bans are generally targeted at promoting the enforcement by a government of
rules applicable to all vessels flying its flag, and in particular appear to have been
designed to address the ‘‘flags-of-convenience’’ problem. So far the target countries
for such import bans have been small developing economies that are less likely to
invoke WTO dispute-settlement procedures. According to ICCAT, it will continue
to use such measures ‘‘to combat the undermining of its regulatory measures in
order to ensure continuous conservation of the stocks under its mandate’’ (WTO
2000a, para. 6).
The Effects of the Interaction

In response to the potential tensions with the WTO, ICCAT has designed its
measures so as to make them WTO-consistent, which has also made them more
resource-intensive to administer. For instance, ICCAT has developed a staged ap-
proach to the adoption of trade measures. With respect to its tuna and swordfish
action plans, ICCAT first informs the flag state concerned of any illegal fishing car-
ried out by its registered vessels. ICCAT then requests the cooperation of the flag
state in bringing its vessels’ activities into compliance with ICCAT measures. Lastly,
ICCAT warns the relevant flag state that nondiscriminatory, trade-restrictive mea-
sures may be taken against it in the event of continued noncompliance. For instance,
warning letters were sent to Guinea-Bissau concerning bluefin tuna catches, Sierra
Leone concerning its bluefin tuna and swordfish fishing activities, and Equatorial
Guinea concerning its fishing practices (WT 2001b, 12). At the Commission meeting
following the issuance of a warning, where the flag state has continued to fish in
contravention of ICCAT regulations, ICCAT generally recommends that parties
take nondiscriminatory, trade-restrictive measures against the offending state, in-
cluding prohibiting imports of its relevant tuna or swordfish products. Significant
resources are also employed in identification of infractions, which involves numerous
legal studies, biological analyses, gathering of trade statistics and landings data, and
sophisticated monitoring of vessels’ activities (WTO 1998a, 5).
ICCAT has made further efforts to enhance the chances that its import bans,
quotas, and technical restrictions could be saved by the ‘‘environmental’’ exceptions
in GATT Article XX or be found to comply with the provisions on trade-restrictive
measures under the TBT Agreement. Most of ICCAT’s management measures have
a scientific basis, and it is careful to ensure that the trade measures it recommends to
its parties are nondiscriminatory. As mentioned before, ICCAT also stresses that the
trade measures implemented by its parties should be ‘‘consistent with their interna-
tional obligations,’’ which include obligations under the WTO.
Furthermore, the Commission makes conscientious efforts to achieve the coopera-
tion of nonparties. It specifically encourages countries that are active in tuna and
swordfish fisheries to accede to the Convention. In 1994, ICCAT urged nonparties
fishing in the Convention area for species under ICCAT competence to become
parties or ‘‘cooperating parties,’’ and requested their observance of ICCAT’s conser-
vation measures (Resolution 94-6 and Recommendation 03-20). Also, ICCAT’s use
of trade measures acts as a powerful incentive to nonparties (including WTO mem-
bers) to accede to the Convention. For example, after trade measures were imposed
against Panama, it acceded to ICCAT and implemented stringent domestic mea-
sures, such as cancellation of all open registries of tuna fishing vessels, and started
a licensing system with a strict satellite vessel-monitoring system. Based on those
measures, in 1999 ICCAT recommended the lifting of the ban against Panamanian
imports of tuna (Recommendation 99-9). According to the ICCAT Secretariat, as
soon as strict quotas were introduced, membership of ICCAT increased (WTO
2001b, 12). Incentives to join include the right to participate in the process of estab-
lishing regulatory measures, taking part in the allocation of quotas, and obtaining a
quota.
The consistency with WTO rules of ICCAT measures against nonparties (that are
WTO members) is further enhanced by direct cooperation between the two institu-
tions. ICCAT has observer status in the WTO Committee on Trade and Environ-
ment, and the WTO is notified of all its recommendations and resolutions at the
same time that they are officially transmitted to ICCAT parties (WTO, 1998a,
2000a). When they enter into force, notification is again given to the WTO. In con-
trast, although ICCAT has consistently invited the WTO to attend its meetings, the
WTO has not yet participated in any ICCAT activities. Nevertheless, ICCAT
receives the reports of the CTE meetings from the WTO Secretariat.

There appears to be general agreement by both the WTO and ICCAT that ICCAT’s
trade measures do not raise conflicts with WTO rules (personal communication with
ICCAT representative; WTO 2001b, 12–13). Where trade measures are considered
necessary, ICCAT is careful to ensure that it advocates measures that are nondiscri-
minatory in effect and stresses that the measures taken by its parties should be ‘‘con-
sistent with their international obligations.’’ ICCAT has tried generally to
incorporate WTO principles in a manner that improves the fairness and the legiti-
macy of its trade measures, and to date, no formal WTO dispute has arisen or
been initiated. In seeking to be nondiscriminatory with regard to nonparties, and in
providing repeated opportunities—particularly to developing countries—to remedy
noncompliant behavior, ICCAT has made its trade restrictions more WTO compat-
ible and generally more acceptable.
The interaction has followed the causal mechanism of Interaction through Com-
mitment. WTO members are bound by WTO rules on nondiscrimination and pro-
portionality. The majority of ICCAT parties have been WTO members and have
thus had a significant interest in avoiding commitments that would be inconsistent
with WTO disciplines. As a result, ICCAT has made particular efforts to minimize
and prevent potential tensions between both regimes by designing WTO-consistent
trade measures (see figure 8.2).
Figure 8.2
Influence of WTO rules on the design of ICCAT trade rules
The continued interaction with WTO rules may involve deeper scrutiny of
ICCAT trade measures, and eventually require them to be tailored more precisely.
The question could be raised as to whether trade-restrictive measures are the appro-
priate response to infractions by countries at low levels of economic development.
Nevertheless, it is fair to say that ICCAT’s trade measures are combined with sup-
portive measures such as technical assistance to establish statistical systems; train-
ing; access to scientific research, statistical databases, and other information; and
special funding arrangements.
Finally, it should be noted that the 1995 UN Fish Stocks Agreement provides
added support to port-state enforcement under ICCAT. Article 21 of this agreement
creates an elaborate system of regional cooperation for the enforcement of region-
ally agreed-on measures against vessels that are suspected of violating them. Not
only may a port state (party to the Fish Stocks Agreement and ICCAT) inspect
documents, fishing gear, and catch onboard a vessel that is voluntarily in its port,
but pursuant to Article 23.3 of the Fish Stocks Agreement, it may also prohibit the
landings and transshipment where it is established that the catch has been taken in a
manner that undermines the effectiveness of ICCAT’s management measures on the
high seas. This type of enforcement measure is allowed regardless of whether the
flag state is a member of ICCAT or not.
Concluding Observations: WTO, Biosafety, and ICCAT
The cause of interaction for both the Biosafety Protocol and ICCAT arises from
overlapping memberships and jurisdictions but potentially divergent objectives with
those of the WTO. The Biosafety Protocol and ICCAT, together with the WTO, reg-
ulate trade in goods but the objective of the environmental regimes—for example to
protect biological diversity—is potentially at odds with the WTO’s aim to liberalize
markets. Measures taken under the Biosafety Protocol or pursuant to ICCAT deci-
sions could restrict trade on a discriminatory basis—a ban on imports of living
modified organisms or tuna from a specific country or countries—and the extent to
which the measures restrict trade would need to be proportionate to their aim of
protecting biological diversity if they are to survive a WTO challenge.
The influence of the WTO regime on the Biosafety Protocol and ICCAT has been
felt at different stages in the cycle of their development. The Biosafety Protocol has
experienced the weight of interaction to date in its negotiation, where the interaction
was arguably disruptive. However, the resulting rules appear to largely accommo-
date both objectives of protecting biological diversity and promoting free trade. In
contrast, ICCAT was established during GATT’s lifetime but before the WTO and
its agreements came into being. The opportunities for interaction have arisen in the
design and implementation of measures under ICCAT instruments, where interac-
tion seems to have been managed successfully.
The interactions of the Biosafety Protocol and ICCAT with the WTO appear to
be Interactions through Commitment. Ultimately, both the Biosafety Protocol and
ICCAT have adapted to and sought to avoid conflict with the WTO regime. In the
case of the Biosafety Protocol, it is widely perceived as having achieved a basis on
which the two regimes can be mutually supportive. However, it will be necessary
to await actual implementation of the Protocol in order to assess whether this re-
sponse will ensure mutual supportiveness of the two regimes. The absence of any
challenge to measures taken under the auspices of ICCAT attests, to some extent,
to its success in adapting to the WTO rules. However, the absence of a WTO chal-
lenge could also be due to poor compliance with ICCAT disciplines, combined with
poor enforcement, leading to ineffective measures with little or no trade impact to
give rise to a complaint.
Both the Biosafety Protocol and ICCAT have used observership in certain WTO
bodies to promote awareness and understanding of their respective regimes within
the WTO. However, this means of communication is largely one-sided. Also, the
CBD and ICCAT secretariats generally only communicate information to the CTE
rather than engaging in an active dialogue that results in concrete outcomes. More-
over, issues relevant to the Biosafety Protocol and ICCAT are often considered
by other WTO bodies to which their responsible secretariats are not necessarily
observers (e.g., TRIPS Council). Ultimately, effective responses are most likely to
come in the form of specialized technical discussion and information exchange (as
is the case with the Biosafety Protocol and other international institutions relevant
to LMOs), as opposed to communications made to the political organs of the re-
spective institutions.
References
Abdel Motaal, Doaa. 1999. The Agreement on Technical Barriers to Trade, the Committee
on Trade and Environment, and Eco-Labelling. In Gary P. Sampson and W. Bradnee
Chambers, eds., Trade, Environment and the Millennium, 223–239. Tokyo: United Nations
University.
Abdel Motaal, Doaa. 2001. Multilateral Environmental Agreements (MEAs) and WTO
Rules: Why the ‘‘Burden of Accommodation’’ Should Shift to MEAs. Journal of World Trade
35 (6): 1215–1233.
Bohanes, Jan. 2002. Risk Regulation in WTO Law: A Procedure-Based Approach to the
Precautionary Principle. Columbia Journal of Transnational Law 40 (2): 323–389.
Buckingham, Donald E., and Peter W. B. Phillips. 2001. Hot Potato, Hot Potato: Regulating
Products of Biotechnology by the International Community. Journal of World Trade 35 (1):
1–31.
Chang, Seung Wha. 2003. WTO Disciplines on Fisheries Subsidies: A Historic Step towards
Sustainability? Journal of International Economic Law 6 (4): 879–921.
Charnovitz, Steve. 1997. Critical Guide to the WTO’s Report on Trade and Environment:
Symposium on NAFTA and the Expansion of Free Trade: Current Issues and Future Pros-
pects. Arizona Journal of International and Comparative Law 14 (2): 341–379.
Charnovitz, Steve. 2000. The Supervision of Health and Biosafety Regulation by World
Trade Rules. Tulane Environmental Law Journal 13 (2): 271–301.
Charnovitz, Steve. 2002. The Law of Environmental ‘‘PPMs’’ in the WTO: Debunking the
Myth of Illegality. Yale Journal of International Law 27 (1): 59–110.
Downes, David R., and Brennan Van Dyke. 1998. Fisheries Conservation and Trade Rules:
Ensuring That Trade Law Promotes Sustainable Fisheries. Washington: Center for Interna-
tional Environmental Law and Greenpeace.
Eggers, Barbara, and Ruth Mackenzie. 2000. The Cartagena Protocol on Biosafety. Journal
of International Economic Law 3 (3): 525–543.
EU Council. 2000. Council Regulation (EC) No 2092/2000 of 28 September 2000 Prohibit-
ing Imports of Atlantic Blue-Fin Tuna (Thunnus thynnus) Originating in Belize, Honduras
and Equatorial Guinea. Official Journal L 249, 4.10.2000, 1.
Falkner, Robert. 2000. Regulating Biotech Trade: The Cartagena Protocol on Biosafety. In-
ternational Affairs 76 (2): 299–313.
Falkner, Robert. 2002. Negotiating the Biosafety Protocol: The International Process. In
Christoph Bail, Robert Falkner, and Helen Marquard, eds., The Cartagena Protocol on
Biosafety: Reconciling Trade in Biotechnology with Environment and Development?, 3–22.
London: RIIA/Earthscan.
Glowka, Lyle, Françoise Burhenne-Guilmin, and Hugh Synge (in collaboration with Jeffrey A.
McNeely and Lothar Gündling). 1994. A Guide to the Convention on Biological Diversity.
Gland, Switzerland, and Cambridge, UK: IUCN.
Jackson, John H. 1997. The World Trading System: Law and Policy of International Eco-
nomic Relations. 2nd ed. Cambridge, MA: MIT Press.
Newell, Peter, and Ruth Mackenzie. 2000. The 2000 Cartagena Protocol on Biosafety: Legal
and Political Dimensions. Global Environmental Change 10 (4): 313–317.
Nordstrøm, Hakan, and Scott Vaughan. 1999. Special Studies 4: Trade and Environment.
Geneva: WTO Publications.
Phillips, Peter W. B., and William A. Kerr. 2000. The WTO versus the Biosafety Protocol for
Trade in Genetically Modified Organisms. Journal of World Trade 34 (4): 63–75.
Pomerance, Rafe. 2000. The Biosafety Protocol: Cartagena and Beyond. NYU Environmental
Law Journal 8 (3): 614–621.
Qureshi, Asif H. 2000. The Cartagena Protocol on Biosafety and the WTO: Coexistence or
Incoherence? International and Comparative Law Quarterly 49 (4): 835–855.
Safrin, Sabrina. 2002. Treaties in Collision? The Biosafety Protocol and the World Trade
Organization Agreements. American Journal of International Law 96 (3): 606–628.
Sands, Philippe. 2003. Principles of International Environmental Law. 2nd ed. Cambridge:
Schoenbaum, Thomas. 2002. International Trade and Environmental Protection. In Patricia
Birnie and Alan Boyle, eds., International Law and the Environment, 2nd ed., 697–750.
Oxford: Oxford University Press.
Shaw, Sabina, and Risa Schwartz. 2002. Trade and Environment in the WTO: State of Play.
Journal of World Trade 36 (1): 129–154.
Stone, Christopher D. 1997. Too Many Fishing Boats, Too Few Fish: Can Trade Laws Trim
Subsidies and Restore the Balance in Global Fisheries? Ecology Law Quarterly 24 (3): 505–
544.
Tapper, Richard. 2000. Biosafety Protocol—the Outlook for Renewed Negotiations. Avail-
able at http://www.ukabc.org/cartagena.htm#c2.
Trebilcock, Michael J., and Howse Robert. 1999. The Regulation of International Trade. 2nd
ed. London: Routledge.
UNEP. 1998, April 3. Conference of the Parties to the Convention on Biological Diversity.
Administration of the Convention on Biological Diversity. Report by the Executive Secretary.
Doc. UNEP/CBD/COP/4/24.
UNEP. 1999a, February 15. Conference of the Parties to the Convention on Biological Diver-
sity. Report of the Sixth Meeting of the Open-Ended Ad Hoc Working Group on Biosafety.
Doc. UNEP/CBD/ExCOP/1/2.
UNEP. 1999b, February 21. Open-ended Ad Hoc Working Group on Biosafety. Protocol on
Biosafety—Draft Text Submitted by the Chair of the Working Group. Doc. UNEP/CBD/
BSWG/6/L.2/Rev.2.
UNEP. 2000, February 20. Conference of the Parties to the Convention on Biological Diver-
sity. Report of the Extraordinary Meeting of the Conference of the Parties for the Adoption
of the Protocol on Biosafety to the Convention on Biological Diversity. Doc. UNEP/CBD/
ExCOP/1/3.
WTO. 1994. Marrakesh Agreement Establishing the World Trade Organization, 15 April
1994, in force 1 January 1995, reprinted in: International Legal Materials 33 (1994): 1144–
1153. Also available at http://www.wto.org/english/docs_e/legal_e/legal_e.htm.
WTO. 1996, November 12. Committee on Trade and Environment. Report. World Trade
Organization. Doc. WT/CTE/1.
WTO. 1998a, July 16. Committee on Trade and Environment. Communication from the
Secretariat of the International Commission for the Conservation of Atlantic Tunas. World
Trade Organization. Doc. WT/CTE/W/87.
WTO. 1998b, October 12. Report of the Appellate Body. United States—Import Prohibition
of Certain Shrimp and Shrimp Products. World Trade Organization. Doc. WT/DS58/AB/R.
WTO. 2000a, June 29. Committee on Trade and Environment. Conservation Measures
Taken by the International Commission for the Conservation of Atlantic Tunas (ICCAT).
Communication from the ICCAT Secretariat. World Trade Organization. Doc. WT/CTE/W/
152.
WTO. 2000b, September 19. Committee on Trade and Environment. Report of the Meeting
Held on 5–6 July 2000. World Trade Organization. Doc. WT/CTE/M/24.
WTO. 2001a, June 6. Committee on Trade and Environment. Compliance and Dispute Set-
tlement Provisions in the WTO and in Multilateral Environmental Agreements. Note by the
WTO and UNEP Secretariats, World Trade Organization. Doc. WT/CTE/W/191.
WTO. 2001b, June 14. Committee on Trade and Environment. Matrix on Trade Measures
Pursuant to Selected MEAs. World Trade Organization. Doc. WT/CTE/W/160/Rev.1.
WTO. 2001c, November 20. Ministerial Declaration, Doha, Qatar, adopted 14 November
2001, World Trade Organization. Doc. WT/MIN(01)/DEC/1.
WTO. 2002, October 3. Committee on Trade and Environment. Submission by Japan. Doc.
TN/TE/W/10.
WTO. 2004, August 31. Committee on Trade and Environment. International Intergovern-
mental Organizations—Observer Status in the Committee on Trade and Environment—
Revision. Doc. WT/CTE/INF/6/Rev.1.
9
Interactions of EU Legal Instruments Establishing
Broad Principles of Environmental Management:
The Water Framework Directive and the IPPC
Directive
Andrew Farmer
Both the EU Water Framework Directive of 2000 and the EU Integrated Pollution
Prevention and Control (IPPC) Directive of 1996 establish broad principles for envi-
ronmental management, while detailed rules of implementation are developed by the
member states.1 The Water Framework Directive establishes an extensive system for
the management of fresh and coastal water resources in the EU. The IPPC Directive
establishes broad requirements for the regulation of industrial installations. They
contrast with many other types of EU environmental legislation that set out a
limited number of specific requirements, such as emission standards or quality stan-
dards relating to particular activities or pollutants, which must be implemented in
the member states. Given their broad scope, both directives have extensive interac-
tions with other EU legal instruments and with a range of international conventions.
The first section of the chapter addresses interactions related to the Water Frame-
work Directive. It considers the range of interactions associated with the Directive
and then focuses on a case of interaction with the Convention on the Protection
and Use of Transboundary Watercourses and International Lakes adopted within
the United Nations Economic Commission for Europe (UNECE). This interaction
is of particular interest within the framework of this book, because it elucidates
how EU instruments interact with multilateral environmental agreements. The
UNECE area is the region with the most comprehensive regional environmental co-
operative framework. Currently there are six regional environmental conventions
and related protocols adopted on transboundary watercourses, water-related dis-
eases, prevention and response to industrial accidents, air-pollution control, envi-
ronmental impact assessment, and public information and participation in decision
making. These instruments are supported by around 100 bilateral or multilateral
agreements and by supranational law adopted within the EU. Additionally, a num-
ber of soft-law instruments (e.g., recommendations and guidelines) are developed
206 Andrew Farmer
under the UNECE to assist implementation. The Convention on Transboundary

Watercourses and Lakes and its interaction with the Water Framework Directive
is only one of many interactions between similar instruments (see Tanzi 2000)
and may highlight processes common to other interinstitutional interactions in the
region.
The second section of the chapter examines interactions related to the IPPC Direc-
tive. Following an introduction to the Directive and an overview over the range of
interactions associated with it, the analysis focuses on the case of interaction with
the EU Large Combustion Plant Directive. This case stands as a model for a class
of interactions of the IPPC Directive with more specific EU legal instruments estab-
lishing emission limits for industrial activities. These instruments influence the IPPC
Directive irrespective of whether they existed prior to its adoption and coming into
force. The case study illuminates how decision makers are to take the specific
requirements of such directives into account when implementing the IPPC Directive.
The interaction is driven by concern over whether the IPPC Directive will properly
be implemented in the member states.
Finally, the chapter draws some general conclusions from these interactions.
The Water Framework Directive
The EU Directive Establishing a Framework for Community Action in the Field of

Water Policy (2000/60/EC, Official Journal L327, 22.12.2000), commonly known
as the Water Framework Directive, represents a major departure for EU water pol-
icy. The Directive developed out of a much earlier discussion concerning the limita-
tions of existing EU water legislation. Beginning in the 1980s, debates took place in
various forums on the development of new EU water legislation (Stern 1995). This
resulted in the adoption of some issue-specific directives in the early 1990s (e.g., on
urban wastewater treatment and on nitrates). However, legislation remained largely
restricted to a series of directives establishing environmental quality standards for
specific water bodies (e.g., bathing waters) or emission limits for specific activities
or for substances (e.g., nitrates). EU water law was criticized for being too frag-
mented, concentrating on specific aspects of environmental quality or specific threats
to that quality. In 1994, the European Commission submitted a proposal for a di-
rective on the ecological quality of surface waters, which, despite its broad scope,
still did not provide a sufficiently comprehensive framework. The Commission,
therefore, submitted a new proposal in early 1997 (European Commission 1997),
EU Legal Instruments and Principles of Environmental Management 207
which evolved into the Water Framework Directive, adopted in December 2000
(Farmer 2000, 2001a).
The scope of the Directive is extensive. It covers all freshwater resources as well as
seawaters up to one nautical mile from the coast. It pursues ecological objectives,
supported by chemical and hydromorphological objectives. There is a strong em-
phasis on developing controls of activities that have an impact on water objectives.
Finally, detailed requirements on reporting, public participation, and so on are
included.
The Directive requires that surface water (i.e., rivers, lakes, and coastal waters)
and groundwater are to be managed within the context of River Basin Management
Plans (Lanz and Scheuer 2001; WWF 2001). All waters are to be characterized
according to their biological, chemical, and hydromorphological characteristics.
These parameters are to be compared with an assessment of waters unmodified by
human activity and classified according to different categories of ecological status.
All waters are required to meet ‘‘good status,’’ except where specific derogations
apply. Member states are to achieve these objectives through the development of
the River Basin Management Plans, which shall integrate existing EU measures to
protect the water environment and identify all remaining human pressures, which
may result in a failure to achieve ‘‘good status’’ (Griffiths 2002). Member states are
required to establish their own programs for each river basin that spell out the mea-
sures by which they intend to achieve the objectives of the Directive (Chave 2001).
Hence, the Water Framework Directive allows significant flexibility on the detailed
regulations that member states might adopt.
The Directive (Bloch 2001) seeks to cluster water-protection measures within a
common rule framework. It is intended not to affect adversely rules that are already
established at either the EU or the international level (e.g., on nitrates or urban
wastewater treatment). Conversely, it seeks to set them in a broader, integrated con-
text to enhance implementation results, not only for those issues newly addressed
in the Framework Directive, but also in older legislation (Farmer and Wilkinson
2001). Potentially, interaction between broad framework rules and specific rule
approaches will become more apparent as implementation proceeds. This might
in the future lead member states to seek rule changes in existing directives not nec-
essarily to alter environmental outcomes and impact, but to alter the ways such out-
comes are to be achieved. Therefore, the Water Framework Directive poses some
interesting possible future interactions in relation to governance issues within the
EU.
208 Andrew Farmer
Overview of Cases of Interaction Involving the Water Framework Directive
Given its extended scope, the Directive interacts with a considerable number of
other institutions. Table 9.1 identifies the most important cases of horizontal and
vertical interaction in which the Water Framework Directive is involved. Indeed,
the Directive itself lists a range of existing EU instruments, which are to be taken
into account when implementing its obligations. These include many items of water
and pollution-control legislation. Other cases will only become obvious as the com-
plexities of implementation become clear.
Among the horizontal interactions between the Water Framework Directive and
other EU legislative instruments we find several cases of Behavioral Interaction.
Generally, effects of these cases are synergistic. For example, active management of
water quality in water basins will almost automatically enhance the quality of bath-
ing waters and thus help implement the Bathing Waters Directive (76/160/EEC),
which defines specific quality requirements of bathing waters. In turn, more locally
concentrated action to improve bathing waters will automatically contribute to
implementing the overall obligations of the Water Framework Directive. Similarly,
implementation of the Nitrates Directive (91/676/EEC) providing for specific limits
of nitrates pollution from agricultural sources, in particular manure, helps imple-
ment the Water Framework Directive, and vice versa. In other cases of Behavioral
Interaction, influence runs predominantly one way. Improvement of water quality
after implementation of the Water Framework Directive will support the conserva-
tion of habitats protected under the Habitats Directive (see also chapter 10) that de-
pend on freshwater resources or coastal waters. Finally, the Environmental Liability
Directive introduces financial compensation for environmental damage and will
thus add a new instrument supporting the implementation of the Water Framework
Directive.
Other horizontal interactions follow the causal mechanism of Interaction through
Commitment. For example, the revised Bathing Waters Directive provides stronger
links with the Water Framework Directive. This will include rule changes attempt-
ing to integrate its objectives more closely with the Framework Directive, in partic-
ular requiring that monitoring and assessments are done within the system of river-
basin planning and programs of measures under the Framework Directive. Some
other EU environmental directives explicitly refer to the Water Framework Directive
and thus provide explicit links between these instruments. Hence, the Environmental
Liability Directive defines the concept of environmental damage with reference to
Table 9.1
Institutional interactions of the EU Water Framework Directive
EU Nitrates Directive Nitrates Directive supports implementation of Water
Framework Directive.
Water Framework Directive is expected to support
implementation of Nitrates Directive.

EU Bathing Waters Directive Water Framework Directive influences the amendment
of Bathing Waters Directive.
Bathing Waters Directive supports implementation of
Water Framework Directive.

Water Framework Directive is expected to support
implementation of Bathing Waters Directive.

EU IPPC Directive IPPC Directive allows the Water Framework Directive
to establish explicit link by defining environmental
quality standards.
Later Water Framework Directive changes meaning of
environmental quality standards protected under earlier

IPPC Directive.
IPPC Directive supports implementation of Water
Framework Directive by requiring that permits do not

lead to breaches of established environmental quality
standards.
EU Habitats Directive Reference to Natura 2000 included in Water
Water Framework Directive supports achievement of
objectives of Habitats Directive.

EU Directive on Environmental Water Framework Directive is used for defining
Liability damage under Liability Directive.
Liability Directive is to support implementation of

Convention on Transboundary Convention triggers certain implementing provisions
Watercourses and Lakes in Water Framework Directive.
Water Framework Directive contributes to achieving
the objectives of the Convention.

Oslo and Paris Convention OSPAR triggers certain implementing provisions in
(OSPAR) for the protection of Water Framework Directive.
the Northeast Atlantic Water Framework Directive supports achieving
objectives of/helps implement OSPAR.

International North Sea International North Sea Conferences facilitated/
Conferences triggered Water Framework Directive.
Aarhus Convention on public Aarhus Convention led to enhanced public
participation and access to participation requirements in Water Framework
information/justice Directive.
Ramsar Convention for the Water Framework Directive supports achieving
protection of wetlands objectives of/helps implement Ramsar Convention.
210 Andrew Farmer
other directives, among them the Water Framework Directive. The Water Frame-
work Directive also includes reference to Natura 2000 of the Habitats Directive
(see also chapter 10).
Finally, there are several interactions between the Water Framework Directive
and the IPPC Directive, which establishes broad requirements for the regulation of
industrial installations and is dealt with in more detail in the second half of this
chapter. Whereas the IPPC Directive does not explicitly refer to the Water Frame-
work Directive, it requires that regulators should not issue permits for industrial
activities that would lead to exceedance of existing EU environmental standards.
This obligation was not directly influenced by the Water Framework Directive, be-
cause it was negotiated and adopted previous to the latter instrument. However, it
constitutes an automatic reference that allows the Water Framework Directive to
modify the substance of member states’ obligations under the IPPC Directive even
after its adoption. Hence, the commitments under the Water Framework Directive
automatically influenced the commitments under the IPPC Directive (Interaction
through Commitment). Subsequently, implementation of the IPPC Directive helps
implement the Water Framework Directive (Behavioral Interaction).
Informal mechanisms, such as the Water Directors’ Group (comprising senior
officials from member states and the European Commission), have been established
to coordinate EU water legislation in detail (Water Directors’ Group 2000). This
may facilitate rule change, or at least rule adaptation, clarify implementation issues,
and stimulate proposals to amend existing institutions if this is seen as necessary.
Other informal (nonstatutory) developments, for example integrated coastal-zone
management, will also affect the rule development under the Framework Direc-
tive (Farmer 1999). However, only detailed analysis following implementation
will reveal the full nature of these possible future interactions and responses to
interaction.
There are also a number of cases of vertical interaction between the Water Frame-
work Directive and international conventions. In most cases, the conventions were
adopted prior to the adoption of the Water Framework Directive, so that the com-
mitments of the former influenced the content of the Directive (Interaction through
Commitment). The most significant of these cases are those relating to the regional
seas conventions and the Convention on Transboundary Watercourses and Lakes.
This Convention has introduced issues such as ecological status and integrated
river-basin management, which are central to the Directive. The Oslo-Paris Conven-
tion for the protection of the Northeast Atlantic (OSPAR) and the International
North Sea Conferences (see chapter 5), as well as the Helsinki Convention for the
protection of the Baltic Sea, have led to the development of specific rule agreements,
such as on the discharge of dangerous substances, and significantly influenced the
details of the subsequent decision on priority substances that forms Annex X of the
Directive. According to the European Commission, the Aarhus Convention on Ac-
cess to Information, Public Participation in Decision-Making and Access to Justice
in Environmental Matters led to enhanced public participation requirements in the
Water Framework Directive (Water Directors’ Group 2002).
The Water Framework Directive also supports the implementation of several
international conventions by providing a stronger legal framework for parties to
meet objectives (Behavioral Interaction). This applies in particular to OSPAR and
the Convention on Transboundary Watercourses and Lakes. The effect is synergis-
tic, intentional (there being clear reference to these objectives in the Directive), and
outcome-based. A similar, though less strong, case can be made for interaction with
the Ramsar Convention on the protection of wetlands. Within the Convention on
Transboundary Watercourses and Lakes, attempts have been made to enhance these
synergistic effects through the development of ‘‘soft’’ implementation rules (as dis-
cussed in more detail in the subsequent case study).
Interaction between the Water Framework Directive and the Convention on

Transboundary Watercourses and Lakes
This case study focuses on a particular case of vertical interaction between the
UNECE Convention on Transboundary Watercourses and Lakes and the Water
Framework Directive. It illustrates how the Directive supports the implementation
of the Convention (Behavioral Interaction). It highlights the importance of the differ-
ent nature of institutional development and rules at the international and EU level
in this process. It will also be seen that as implementation of the Water Framework
Directive has taken place, parties to the Convention have responded to this interac-
tion by elaborating further rules for nonmembers of the EU.
It should be noted that this case is part of a more complex set of vertical inter-
actions between the Water Framework Directive and the Convention on Trans-
boundary Watercourses and Lakes. Initially, the Convention has had a strong and
synergistic influence on the development of the Directive (table 9.1). Only in the
second stage is the Directive now capable of assisting implementation of the
Convention.
212 Andrew Farmer
The Convention on Transboundary Watercourses and Lakes

Cooperation on water issues has taken place within the framework of the UNECE
for around thirty years. In its early stages this led to a series of recommendations
relating to issues having a transboundary component, for example recommenda-
tions on agriculture in the 1970s. Concerted activity on transboundary waters began
with the UNECE decision on principles of transboundary water cooperation in
1987 (Enderlein 2001). It contained the major elements found in the Convention,
which was adopted in 1992 and entered into force on October 6, 1996. However,
the development of the Convention was not just a transition from ‘‘soft law’’ at the
UNECE level to a binding agreement in order to ensure implementation. It also
developed out of soft-law agreements at a bi- or multilateral level between countries
tackling specific transboundary water problems. The experience of the implementa-
tion of such agreements between countries was essential in developing workable
solutions within the Convention. The debate leading up to the adoption of the Con-
vention overlapped with the debate on the future of water policy within the EU. All
EU member states as well as the EU (legally: the European Community) itself have
since ratified the Convention.
The Convention has the aim of strengthening local, national, and regional actions
to protect transboundary surface and groundwaters based on maintaining the func-
tion of ecosystems. Parties to the Convention are, particularly, required to prevent,
control, and reduce pollution of transboundary waters by hazardous substances,
nutrients, and microorganisms. The Convention recognizes the precautionary and
polluter-pays principles as guiding principles in the implementation of measures as
well as the need to sustain water systems for future generations. It establishes
the river basin as the unit of management and requests that parties identify the
catchments or subcatchments, which are subject to cooperation. Because previous
agreements between riparian countries of shared rivers and lakes often resulted in
incomplete resolution of the transboundary problems that they sought to address,
the Convention specifically obliges parties ‘‘to enter into bilateral or multilateral
agreement or other arrangements, where these do not yet exist, or adapt the existing
ones, where necessary, to eliminate the contradictions with the basic principles of
this Convention, in order to define their mutual relations and conduct regarding
the prevention, control and reduction of transboundary impact’’ (Art. 9 of the Con-
vention). The Convention, therefore, represents an important milestone in interna-
tional cooperation, which is taken further by the Water Framework Directive.
The Case
While the Water Framework Directive is the principal instrument to implement the
Convention within the European Union, it changes the situation of the member
states significantly. The Directive is not simply an instrument at EU level acting in a
‘‘parallel’’ context to the Convention, it is the legal response of the EU (legally: the
European Community) to ratification of the Convention, as a member of that Con-
vention. For the EU member states, however, the Directive has a very different char-
acter from the Convention, mainly for two reasons.
First, the Directive is significantly broader in scope than the Convention. While it
draws extensively on the general requirements of the Convention, such as ecosystem
objectives and integrated river-basin management, in two areas in particular it
reaches beyond the Convention. Whereas the Convention covers merely transbound-
ary waters and lakes and relevant parts of a catchment, the Directive covers all
waters, not just those of a transboundary nature. The reality of water management
is that a full catchment-based approach is the most effective means of achieving
environmental objectives. The treatment of transboundary areas on their own may
lead to poor decision making and failures to achieve objectives. Thus the Directive
clearly has the potential to enhance the effects of the Convention. Moreover, the Di-
rective obliges member states to take account of a wide scope of issues in achieving
its objectives. Hence, member states are subject to additional obligations. The role
of EU legislation in this case is particularly important. The Convention requires
action to be undertaken within each member country. However, its focus is on
transboundary cooperation through joint planning and action. Thus, even if the
Convention is translated into ‘‘hard’’ national law (as could be done with country-
specific obligations under conventions relating to pollutant reduction), these trans-
boundary elements are difficult to incorporate. In contrast, EU legislation is the
perfect vehicle for this.
Second, the Directive reflects particularly ‘‘hard’’ supranational law to be imple-
mented by the member states. The European Commission monitors compliance
and, where it considers compliance to be inadequate, it can seek legal action against
a member state through the European Court of Justice. None of these ‘‘encourage-
ments’’ to compliance is available under the Convention. While the implementation
of the Convention itself in the member states is thus not subject to the detailed scru-
tiny of the European Commission, its provisions incorporated into the Framework
Directive will be.
214 Andrew Farmer
The EU member states have reacted to the stricter set of obligations enshrined in
the Directive in a number of ways. Although it is difficult at this stage to be precise
about the additional support that the Directive provides because the Directive still
awaits detailed implementation in the member states, the influence of the Directive
on transboundary cooperation may be significant. For example, the Rhine Commis-
sion has indicated that, in response to the Directive, it will extend its cooperative
efforts beyond existing countries (Switzerland and countries downstream) to include
all countries in the catchment (e.g., Austria and Italy) (ICPR 2001). There have also
been a series of activities within member states and, collectively, by the member
states at the EU level. A range of working documents illustrate this, examples being
Ireland (Environmental Protection Agency 2000) and the guidance on planning be-
ing developed under the Common Implementation Strategy (2002) and by EU non-
governmental organizations (NGOs) (WWF 2001).
Whereas the EU member states are predominantly faced with implementation of
the Water Framework Directive, they simultaneously boost implementation of the
Convention. The examples of transboundary cooperation just mentioned could, or
even should, have taken place through the implementation of the Convention. How-
ever, they have not been apparent prior to the Directive. It is the Directive that has
taken them forward because its provisions are enforceable as ‘‘hard’’ law. The need
for integrated planning is highlighted within the Convention, yet activity has only
taken off since the adoption of the Directive. This is also true in other areas where
the two institutions overlap, such as in setting ecological objectives. This synergistic
vertical interaction and the enforceability of EU legislation will thus result in greater
implementation of the Convention.
The case follows the causal mechanism of Behavioral Interaction as illustrated
by figure 9.1. It does not affect the rules of either of the two institutions involved,
but their performance within the issue areas governed. The Directive implements
the Convention at EU level, based on the fact that both the European Union and
its member states are parties to the Convention. The Directive incorporates all pro-
visions of the Convention, and it expands its scope significantly. Even more im-
portant, it turns them into ‘‘hard’’ supranational EU law. Early indications are
that member states are responding to the harder nature of the Directive. Thus,
the Convention’s new supervisory instrument regarding implementation by EU
member states enhances the implementation of the Convention. Given the early
stages of the implementation of the Directive, the practical consequences of
this interaction at the outcome level are not yet fully known. However, response
Figure 9.1
Water Framework Directive helps implement Convention on Transboundary Watercourses
and Lakes
action within the Convention to enhance and fully exploit existing synergies is under
way.
Response Action
The actors operating within the Convention have attempted to enhance the synergis-
tic interaction with the Water Framework Directive as far as possible through delib-
erate response action and intensified collaboration of the two institutions. This is
seen through the direct influence that the Directive is having on pushing forward
the development of ‘‘soft’’ rules such as guidelines under the Convention to elabo-
rate its rules. Most particularly, the Convention has taken advantage of this activity
to enhance its own guidance for implementation and to seek to improve coordina-
tion between the instruments and exploit synergies. While these guidelines would be
unlikely to have any additional effect on parties that will implement the Directive,
they will influence other parties in the UNECE region.
The implementation of the Convention is supported by the work of a series of
working groups addressing issues such as legal and administrative aspects, monitor-
ing and assessment, water and health, and so on. Their responsibility is to draft
recommendations, codes of practice and other soft-law instruments, and assist the
Conference of the Parties to the Convention and promote the harmonization of
rules. A specific task for the working groups is to ‘‘avoid, to the extent possible,
216 Andrew Farmer
duplication of effort with other water-related activities of other United Nations

bodies and other international organizations’’ (UNECE 2001). Within the UNECE
region, this clearly includes initiatives from the EU.
Guidelines on monitoring of waters under the Convention were published in 1996
(UNECE 1996). In 2000 a review of these guidelines was undertaken that began to
establish links with developments at the EU level. The review highlighted that ‘‘spe-
cial attention’’ (UNECE 2000) had been given to recent developments in EU legisla-
tion, especially developments on the Water Framework Directive. The review makes
specific recommendations on objectives. These specifically include that ‘‘the manage-
ment of the water environment should aim at sustaining or restoring the good eco-
logical quality of river basins’’ (UNECE 2000). This objective explicitly refers to the
Water Framework Directive and the definitions contained in it.
The links between the two institutions are strongly promoted under the Conven-
tion. An important milestone in this regard was a workshop in Bratislava in October
2001. Among its conclusions and recommendations, it stated that ‘‘it is recom-
mended that the Bureau of the Meeting of the Parties to the UNECE Water Conven-
tion makes arrangements, as appropriate, with the European Commission so that
activities under both the UNECE Water Convention and the EC water framework
Directive are brought even closer together.’’ Moreover, ‘‘the working groups estab-
lished under the Convention may encourage the water framework regime to be used
by non-EU parties.’’ Finally, it is ‘‘recommended that Parties to the Convention that
are not EU Member States or candidate countries should make appropriate use of
the guidance on the economic analysis that is under development in the framework
of the EU Common Strategy. A program element aiming at testing this guidance
document in countries in transition and adapting it, if required, to the specific con-
ditions of these countries, could be considered for inclusion in the work plan for
2003 to 2006’’ (UNECE 2001).
The work plan for the Convention for the period 2000–2003 also included
coordinating elements between the two institutions. The Meeting of the Parties
should provide guidance on the prevention of deterioration of ecosystems and
their protection, and the promotion of sustainable water use as stipulated both
in the Convention and the (proposal for the) EU Directive. It should also assist
countries in environmental management and sustainable development of inter-
national lakes, taking account of legal frameworks such as the Convention and the
Directive.
The Integrated Pollution Prevention and Control (IPPC) Directive
The Integrated Pollution Prevention and Control (IPPC) Directive (96/61/EC) of

1996 provides a comprehensive approach to the regulation of a wide range of indus-
trial installations in the EU. Prior to the adoption of the Directive, EU legislation
provided only a limited framework for industrial regulation. However, the IPPC Di-
rective not only broadens the scope of such regulation, it also makes explicit links to
the requirements in other EU legislation in order to provide an integrated approach
to delivering environmental objectives. This sets an interesting framework for the
examination of interactions between institutions.
The Directive requires regulation of industrial activities grouped into six cate-
gories: energy industries, production and processing of metals, mineral industries,
chemical industries, waste management, and other activities such as pulp and paper,
tanning, food processing, and certain agricultural activities like intensive animal
units. The Directive is aimed mostly at large installations, and for the majority of
sectors there are production-capacity thresholds that exclude the smaller installa-
tions (Emmott and Haigh 1996; Emmott 1999; Gislev 2001).
The objective of the Directive is to attain ‘‘a high level of protection for the envi-
ronment taken as a whole’’ (Art. 1). In contrast to previous legislation, which dealt
with emissions only to air or water, this is to be achieved by preventing or reducing
emissions to air, water, and land, and it includes measures concerning waste.
Manufacturers and authorities thus have to think about all emissions and their envi-
ronmental impact in the design of the whole plant (‘‘clean technology’’) rather than
relying on ‘‘end-of-pipe’’ techniques, which frequently solve one environmental
problem by creating a new one. Europe’s traditional environmental regulation
has emphasized end-of-pipe abatement techniques (Gouldson and Murphy 1998;
Farmer 2001b). In contrast to most end-of-pipe measures, pollution prevention pro-
moted by the IPPC Directive is not only of environmental benefit. It also represents a
significant economic benefit, because the generation of pollution and waste, includ-
ing heat, reveals inefficiencies within the production process. The basic technology
requirement to be reflected in IPPC permits is ‘‘best available techniques’’ (BAT).
Within the definition of BAT, ‘‘available’’ is specified as meaning economically and
technically viable, taking into consideration costs and advantages. In determining
BAT, special consideration must be given to certain factors listed in an annex,
including the use of low-waste technology, the use of less hazardous substances,
218 Andrew Farmer
the furthering of recovery and recycling, the consumption of raw materials and
water, and energy efficiency. An obligation is placed on member states to ensure
that their competent authorities remain informed of BAT developments (Art. 11).
Member states’ competent authorities must ensure that installations are operated
in such a way that certain general principles are followed. These are: to take all ap-
propriate preventive measures against pollution; to ensure that no significant pollu-
tion is caused; to avoid waste production and to recover or safely dispose of waste
produced; to use energy efficiently; to take the necessary measures to prevent acci-
dents; and to protect and clean up the site on cessation of the industrial activity.
Member states are to ensure that installations covered by the Directive are not
operated without a permit (Bohne 2000; Ten Brink and Farmer 2004). Permit
holders must be required to advise the competent authorities of any changes in their
operations, and any substantial modifications must also be made subject to prior
authorization. Furthermore, competent authorities must reconsider and, if neces-
sary, update permit conditions periodically. Reconsideration must be undertaken
inter alia when excessive pollution occurs or when developments in BAT allow sig-
nificant emission reductions without excessive cost (Art. 13).
From November 1999, EU member states have to ensure that no new installation
is operated, and that no substantial change is made to the operation of an exist-
ing installation, without an IPPC permit. The rules do not immediately apply to
unmodified existing installations that have been given an eight-year transition
period until 2007, with longer exemptions for certain installations in the new mem-
ber states (Farmer 2003). Nevertheless, some member states have chosen to require
mandatory IPPC permits for existing installations before this deadline.
Overview of Cases of Interaction for the IPPC Directive
The IPPC Directive has interacted both with institutions established prior and sub-
sequent to its adoption. Virtually all these interactions produce synergistic effects.
Likewise, interactions across the boundaries of the policy field are virtually absent.
Although the Directive regulates some agricultural activities like intensive pig and
poultry units, for example, there is no interaction with the Common Agricultural
Policy (CAP), because these activities receive no CAP subsidies. The Directive on
Emissions Trading for Carbon Dioxide (2003/87/EC) might prove disruptive in pro-
cedural terms in its interaction with the IPPC Directive, since carbon dioxide con-
trols from selected industries are no longer subject to BAT assessments. However,
it might be synergistic in achieving overall environmental objectives. Precise effects

can only be determined as implementation occurs. Table 9.2 provides an overview
of important horizontal and vertical interactions.
The IPPC Directive is the target institution in a number of cases of Interaction
through Commitment by which its rules are directly or indirectly shaped and molded.
The rules of the Directive were explicitly developed to take account of the principles
of waste management already established in the Waste Framework Directive (2000/
61/EC), such as that of waste avoidance. This was a clear intention of both the
Commission and member states. In several other cases, the Directive is deliberately
linked to directives that set out their own environmental objectives. The IPPC Direc-
tive requires that regulators do not issue permits for industrial activities leading to
an exceedence of any EU environmental standard (Art. 9). The Air Quality Frame-
work Directive has developed air-limit values (chapter 12). The Water Framework
Directive will identify good ecological status for surface waters (see above in this
chapter and table 9.1). The Habitats Directive and the Birds Directive identify favor-
able conservation status for designated sites (chapter 10). Even without an explicit
reference to these instruments, the IPPC Directive adopts their substantive environ-
mental standards. Commitments under the instruments mentioned also become
commitments under the IPPC Directive, because the latter allows other instruments
to define its substantive environmental standards. Although this particular type of
interaction becomes apparent only as individual permit decisions are made, it there-
fore constitutes a form of Interaction through Commitment.
A further example of such Interaction through Commitment is seen in the inter-
action between the IPPC Directive and those EU directives that establish standard
emission limits for processes across the EU. Examples include the Solvent Emissions
Directive (1999/13/EC), which establishes limits for emissions of volatile organic
compounds to air, and the Large Combustion Plants Directive (the latter interaction
is examined in detail below). Such limits act as minimum standards to be applied by
regulators, and the commitments under the instruments mentioned thus affect the
material content of the obligations of the IPPC Directive. However, they do not mit-
igate the necessity to determine BAT, which could result in stricter requirements.
The IPPC Directive is also the source, and occasionally the target, of a number of
cases of Behavioral Interaction. Given the absence of independent environmental
objectives and strict standards, the crosscutting IPPC Directive is designed to facili-
tate achievement of the more concrete objectives of the sector-specific instruments.
Its implementation will almost automatically assist in achieving the objectives of all
220 Andrew Farmer
Table 9.2
Institutional interactions of the IPPC Directive
EU Waste Framework Has led to the inclusion of the principle of waste
Directive avoidance in the IPPC Directive
Benefits from implementation of IPPC Directive that
requires waste avoidance in line with the Waste

Framework Directive
EU Air Quality Framework Is linked to IPPC Directive that requires permits do not
Directive lead to breaches of established environmental quality
standards
requires that permits do not lead to breaches of established

environmental quality standards
EU Habitats Directive Is linked to IPPC Directive that requires that certified
installations do not affect adversely ‘‘favorable
conservation status’’
Benefits from IPPC Directive that requires that certified

EU Birds Directive Is linked to IPPC Directive that requires that certified
requires that certified installations do not affect adversely

‘‘favorable conservation status’’
EU Large Combustion Plants Ensures minimum standards, thus limiting flexibility in
(LCP) Directive issuing permits under IPPC Directive
EU Solvents Directive Ensures minimum standards, thus limiting flexibility in
issuing permits under IPPC Directive
EU Environmental Facilitates implementation of IPPC Directive in that
Management and Auditing EMAS certificates can be used to simplify permitting
Scheme (EMAS Regulation) procedures
provides incentive to establish EMAS

EU Environmental Impact Is linked to IPPC Directive that allows use of
Assessment (EIA) Directive Environmental Impact Assessment under EIA Directive in
permitting procedures
Facilitates implementation of IPPC Directive (in that
Environmental Impact Assessment can be used under IPPC)

EU Seveso II Directive on Benefits from IPPC Directive that promotes an integrated
industrial accidents approach and contributes to a greening of installations so
that the impact of any accident would be minimized
Table 9.2
(continued)
Aarhus Convention on public Has led to amendments of rules on public participation
participation and access to under IPPC Directive
information/justice
OSPAR Convention for the Benefits from implementation of IPPC Directive
protection of the Northeast (reduction of pollution of Northeast Atlantic)
Atlantic
Helsinki Convention for the Benefits from implementation of IPPC Directive
protection of the Baltic Sea (reduction of pollution of Baltic Sea)
ECE Convention on Long- Benefits from implementation of IPPC Directive
Range Transboundary Air (reduction of air pollution)
Pollution
UN Framework Convention Benefits from implementation of IPPC Directive
on Climate Change (reduction of emissions of greenhouse gases)
EU environmental institutions that regulate matters also regulated under the IPPC
Directive. For example, if an IPPC permit requires reduction of industrial wastes,
or the limitation of emissions of air and water pollutants, or the protection of a rec-
ognized habitat, it simultaneously helps implement the relevant instruments, namely
the Waste Framework Directive, the Air Quality Framework Directive (chapter 12),
the Habitats Directive (chapter 10), the Birds Directive, or the Water Framework
Directive (see earlier in this chapter and table 9.1). The integrated approach pursued
by the IPPC Directive also reduces the danger of industrial accidents and thus sup-
ports implementation of the objectives of the Seveso II Directive (96/82/EC). In turn,
environmental impact assessments required under the Environmental Impact Assess-
ment Directive of 1985 and certificates issued under the Environmental Manage-
ment and Auditing Scheme (EMAS) Regulation of 2001 may be introduced into the
IPPC licensing process and thus help implement the IPPC Directive procedurally.
Implementation of the IPPC Directive also assists implementation of several inter-
national environmental regimes. These include global institutions, such as the UN
Framework Convention on Climate Change (chapter 3), those at the European level,
such as the UNECE Convention on Long-Range Transboundary Air Pollution, and
those at a subregional level, such as the regional seas (OSPAR and Helsinki Conven-
tions). Prevention or reduction of pollution to the atmosphere or to surface waters
occurs within the issue areas governed, and thus at the outcome level (Behavioral
Interaction). There is no interaction at the rule level. The only relevant instance of
222 Andrew Farmer
vertical Interaction through Commitment resulted in the only amendment of the

IPPC Directive to date. The Aarhus Convention on public participation and access
to environmental information identified a series of occasions on which the public
should have access to information and have a right to contribute to decision making
(Jendroska and Stec 2001). The IPPC Directive required public consultation on per-
mit applications, but the Convention introduced the requirement for public informa-
tion of the permit decisions that are made. EU member states and the Commission,
in signing the Convention, recognized in advance that this would require an amend-
ment of the IPPC Directive.
Automatic Rule Adaptation: How the Large Combustion Plants Directive Provides
a Safety Net for the IPPC Directive
The IPPC Directive and the Large Combustion Plants (LCP) Directive interact
according to the causal mechanism of Interaction through Commitment. This
revolves around the incorporation, by the IPPC Directive, of emission-limit values
set in the framework of the LCP Directive. This type of interaction relates to any
EU directive that will set emission limits for activities covered by the IPPC Directive.
The Large Combustion Plants Directive

The fundamental aim of the original Large Combustion Plants Directive (Directive
88/609/EEC; Official Journal L336, 7.12.1988) was to reduce emissions of the gases
that form the major contributors to acid rain from their main sources (mostly fossil-
fuel power stations). The 1988 Directive was the first EU law to significantly tackle
these sources of air pollution. Although these sources were also regulated by other
instruments of EU legislation, not least the IPPC Directive, the Directive was
renewed in 2001.
The 1988 Large Combustion Plants Directive required that member states take
appropriate measures to ensure that all licenses for the existing plants contain con-
ditions that comply with the emission-limit values for SO2 , NOx , and dust set in
Annexes III to VII for new plants (post-1988). Alternatively they could subject
them to a national emission-reduction plan,2 which ought to lead to the same emis-
sion level that would have been achieved by applying the emission-limit values indi-
vidually to the existing plants in operation. In other words, the plan would allow
some plants to exceed the limits provided that emissions from other plants were
below them. The reduction targets for emissions of SO2 and NOx from existing
plants were set in Annexes I and II, with which the plan had to comply. The na-
tional emission-reduction plan was to include objectives, related targets, and the
measures and timetables to reach them.
Directive 88/609 was repealed by Directive 2001/81/EC on November 27, 2002,
which has its roots in the Commission’s 1997 acidification strategy. It sets stricter
emission limits in line with the technological progress that has been achieved in this
sector. The 2001 Directive applies to all combustion plants, not just those coming
into operation since 1988. During adoption, Germany and the United Kingdom
had disagreed over the role of older plants.3 Germany was unwilling to accept a di-
rective that would not include old plants, whereas the United Kingdom was equally
adamant not to include them. According to a compromise reached in June 2000,
existing plants were included in the Directive only with a residual life of 20,000
hours after the 2008 deadline. The European Parliament adopted the compromise
agreement reached with the Council.4
The Case
Whereas implementation of the LCP Directive has resulted in significant reductions
of both the emissions of acidifying substances and acid deposition, the IPPC Direc-
tive is designed so as to replace sector-specific instruments regulating particular
forms of pollution from industrial plants. The clue to this case of interaction is pro-
vided by the answer to the question of why the member states and EU decision-
making bodies chose to retain the LCP Directive and even bothered to thoroughly
revise the instrument.
Retention, or change, of the LCP Directive has the immediate effect of establishing
strict and comparatively clear-cut emission-limit values under the IPPC Directive.
While the IPPC Directive does not itself set strict emission limits, it incorporates
limits established under other EU environmental instruments (Art. 18 of the IPPC
Directive). The LCP Directive constitutes legislation setting emission-limit values
as under Article 18 of the IPPC Directive (consideration 7 of the renewed LCP
Directive). Due to the automatic reference clause in Article 18 of the IPPC Direc-
tive, adoption or renewal of the LCP Directive thus results automatically in a simul-
taneous change of the rules of the IPPC Directive. Practically, it introduces strict
emission limits into the murky discussion on the appropriate application of best
available techniques. The interaction is illustrated in figure 9.2.
How strong this interaction is depends on exactly how the member states would
implement the IPPC Directive and guidelines associated with it in the absence of the
224 Andrew Farmer
Figure 9.2
Automatic incorporation of rules of Large Combustion Plants Directive into IPPC Directive
LCP Directive. To understand this, we need to consider the basis for implementing
the IPPC Directive and for setting emission-limit values under this instrument.
Implementation of the IPPC Directive is closely related to the application of the
‘‘best available techniques’’ (BAT) and immediately depends on the interpretation
of this term. BAT is defined in Article 2(11) as follows:
‘‘Best available techniques’’ shall mean the most effective and advanced stage in the develop-
ment of activities and their methods of operation which indicate the practical suitability of
particular techniques for providing in principle the basis for emission limit values designed
to prevent and, where that is not practicable, generally to reduce emissions and the impact
on the environment as a whole.
The Directive points out that BAT is not a static concept, but develops over time.
As mentioned before, it requires a competent authority to be informed of develop-
ments in best available techniques and periodically to reconsider and, under certain
circumstances (e.g., improvements in BAT), update permit conditions.
The Directive requires the Commission to organize an exchange of information
on BAT (Art 16(2)), which has resulted in the development of BAT reference docu-
ments (BREFs; Bär, Kraemer, and Emmott 2000). A BREF explores the available
techniques for a particular type of process and sets out options, including emission
limits, that should be used by member states in determining BAT on an individual
basis. The documents are not binding and require interpretation on a case-by-case
basis. The BREF outline and guide under the IPPC Directive states that ‘‘a BREF
does not interpret the Directive itself, nor does it remove the obligations on opera-
tors and Member States under the Directive to make decisions at national, regional
or local level including the necessary balanced decisions required by the Directive.
BREFs do not prescribe techniques or emission limit values’’ (Institute for Prospec-
tive Technology Studies 2000). The section on BAT of this outline and guide states
that ‘‘while the BAT reference documents do not set legally binding standards, they
are meant to give information for the guidance of industry, Member States and the
public on achievable emission and consumption levels when using specified tech-
niques. The appropriate limit values for any specific case will need to be determined
taking into account the objectives of the IPPC Directive and the local considera-
tions.’’ It is critical to understand how the ‘‘appropriateness’’ of limit values is
determined.
The development of BREFs has not resulted in a single definition of BAT for iden-
tical or for different categories of installation. For example, the BREF for refineries
discusses various options, listing emission-limit values for NOx ranging from 70
to 500 mg/m 3 . This is a very wide range. Similarly, the 400-page BREF on LCPs
describes a wide range of techniques that can be applied, but does not prescribe
emission-limit values for member states to use as BAT. The information contained
in the BREFs indicates that while following their recommendations would require
stricter conditions on installations in a number of member states, the current re-
quirements on some installations may be likely to meet BAT requirements (although
this will only become clear once full BAT determinations are carried out). Member
states have to report to the European Commission on the application of BAT. It is
likely that these reports will show a wide range of interpretations across the EU,
reflecting the variation found in the BREFs (see also Schnutenhaus 1994; Haigh
2000).
The BREF on LCPs has taken account of a wide range of technical developments.
It presents a range of options that might be considered BAT. However, not being
prescriptive, it has not been subject to the political arguments that took place during
the adoption of the 2001 Large Combustion Plants Directive. There are different
views on the consequences of this instrument for the IPPC Directive. According to
an NGO position, the emission limits of the renewed LCP Directive do not reflect
BAT and implementation of the IPPC Directive requires stricter emission limits
(James 2001). However, others view the LCP Directive as providing a very strong
lead on what is expected at the EU level. Thus, interpretations on what emission-
limit values might be required under IPPC on a case-by-case basis vary significantly.
226 Andrew Farmer
There are likely to be a number of instances where the LCP limits will be set in IPPC
permits.
Therefore, the rationale not only for the survival of the LCP Directive, but for its
renewal, rests in the concern of actors within the EU over the potential for full im-
plementation of the IPPC Directive. The interpretation of BAT, in terms of which
specific emission limits will be applied under the IPPC Directive for large combus-
tion plants, remains uncertain. Only if in the future BREFs are found to be perfect
and all member states follow them perfectly should there be no need for specific
directives (Haigh 2000). The development of the LCP Directive of 2001 suggests
that ‘‘imperfection’’ was expected at some point in the implementation of BAT in
the member states. The position of the United Kingdom with regard to existing
installations during the negotiations of the LCP Directive, as described above, sug-
gests that concerns about the practical implementation of the IPPC Directive might
have been justified.
Implementation of the IPPC Directive (on this issue) is now much easier to moni-
tor than it would have been in the absence of the LCP Directive. The latter requires
implementation early in 2008, a few months after the deadline for compliance under
the IPPC Directive. Assessing compliance with the specific emission-limit values set
under the LCP Directive is relatively straightforward. It is, therefore, relatively easy
for the Commission, regulators, or the public to identify whether these limits are
being complied with. In the absence of the LCP Directive, member states would
(eventually) have had to report on emission-limit values derived from BAT assess-
ments. Given that these were not prescribed, lengthy debates would almost surely
ensue over whether the best available techniques have been adequately determined.
A challenge of IPPC permits for large combustion plants by the Commission would
necessarily take considerable time because a technical assessment would be required.
It could, therefore, delay compliance for years.
Whereas the LCP Directive clearly supports the achievement of objectives of the
IPPC Directive for protection of the environment in the beginning, the direction of
influence, as well as the causal mechanism by which it is driven, might change over
time. Requirements of the IPPC Directive include observation of the emission-limit
values originating from the LCP Directive, but this is merely a necessary, not a suf-
ficient condition for compliance. After all, the IPPC Directive requires application of
best available techniques. ‘‘Such compliance may involve more stringent emission
limit values, emission limit values for other substances and other media, and other
appropriate conditions’’ (consideration 8). In particular, ‘‘best available techniques’’
is a dynamic concept that envisages continuous adaptation of requirements. Hence,

the LCP emission limits constitute minimum standards under the IPPC Direc-
tive. As soon as these limits are exceeded, the IPPC Directive will begin to assist
(over-)achievement of the objectives of the LCP Directive. This effect will not be a
matter of direct or indirect rule change, but purely one of implementation. It will
thus constitute Behavioral Interaction.
Overall, the LCP Directive can be viewed as a ‘‘safety net’’ so as to prevent the
member states from using the looser framework of the IPPC Directive to undermine
current standards of environmental protection. Through the automatic reference
clause within the IPPC Directive it is capable of modifying the obligations under
this Directive. Thus, the interaction produces, on the surface, a straightforwardly
synergistic effect on the IPPC Directive by ensuring that specified emission limits
for large combustion plants are complied with. The strict limits in the LCP Directive
help achieve the environmental objectives of the IPPC Directive because it also regu-
lates all activities covered by the LCP Directive. However, the synergistic effect is
only as strong as the likelihood that, given the chance, member states would have
sought to justify weaker emission limits when interpreting the IPPC Directive alone.
And it will last only until the European Commission manages to challenge any indi-
vidual determinations of BAT, which could take many years. Over time, the direc-
tion of influence will be gradually reversed, if the obligation to apply BAT under the
IPPC Directive results in stronger emission limits than those provided for by the
LCP Directive.
Conclusions
The IPPC Directive and the Water Framework Directive represent a type of frame-
work legislation that seeks to trigger interaction with many other legal instruments
to increase the effectiveness of EU environmental governance. They form the pin-
nacles of EU lawmaking in their respective areas, namely water management and
industrial pollution control. They are designed to be holistic, to be relatively com-
prehensive, and to act as frameworks within which other regulatory activities can
reside. Thus, both directives not only have many interactions with other EU legal
instruments and international institutions, but are specifically designed to do so.
Although such an objective could also be developed at the international level, the
comprehensive and supranational institutional framework of the European Union
is particularly suited for its realization. Having said this, the integration with other
228 Andrew Farmer
EU institutions is not perfect and those implementing the rules can be frustrated by
inconsistencies and lack of coherence between specific rules of the different instru-
ments (see Farmer, Skinner, and Beyer 2003; Farmer and ten Brink 2004).
The inventories of interactions show that almost all of the identified horizontal
interactions with other EU instruments are synergistic, and that the two directives
considered in this chapter constitute attempts to enhance synergy as far as possible.
The virtually exclusively synergistic quality of effects may be attributed to the fact
that interaction occurs predominantly with instruments pursuing broadly similar or
at least easily compatible objectives. Generally, interaction occurs in two ways. On
the one hand, both the IPPC Directive and the Water Framework Directive draw on
particular concepts and precise emission-limit values contained in the more specific
instruments. Hence, their rules are influenced by the commitments of these specific
instruments (Interaction through Commitment). As the case of the Large Combus-
tion Plants Directive and the IPPC Directive demonstrates, influence may originate
even from instruments adopted later than the relevant target institution, if the latter
contains a general reference clause. On the other hand, the two directives discussed
in this chapter are the source of synergy from Behavioral Interaction assisting the
implementation of various other instruments.
Interactions in which the two directives are involved demonstrate that some ‘‘dou-
ble work’’ has been deliberately developed at the EU level. Acting as ‘‘frameworks’’
or ‘‘umbrellas,’’ the two directives result in ‘‘double work’’ for EU decision makers
and implementing parties. In most cases both directives seek to minimize this effect
by encouraging the practical integration of the rules of other institutions within the
rule framework of the relevant framework directive. However, the case of the inter-
action between the IPPC Directive and the Large Combustion Plants Directive illus-
trates that double work may be desired. The concern of some actors within the EU
decision-making process about how well member states will implement the IPPC
Directive has led them to adopt a seemingly ‘‘unnecessary’’ directive to ensure that
environmental outcomes are met—and to resort to a safety-net approach.
The vertical interaction between the Water Framework Directive and the Conven-
tion on Transboundary Watercourses and Lakes demonstrates that interaction is
frequently a two-way process composed of two—or more—separate cases. Basic
concepts of ecological quality were elaborated initially within the Convention,
which influenced the elaboration of the Directive. This case was driven by Interac-
tion through Commitment. After its adoption, the Directive with its encompassing
approach and its hard supranational law facilitates implementation of the Conven-
tion through Behavioral Interaction. This second interaction is responded to by co-

ordination between the two instruments and the adoption of secondary measures
within the Convention. Hence, each of the two instruments has pushed the other
during one crucial phase of the governance process.
Notes
1. The term European Union (EU) will be used broadly throughout this chapter, which
includes the period before the Treaty of Maastricht.
2. ‘‘Large Combustion Plants Directive-Limits Also for Existing Plants,’’ Acid News 3,
October 2000, 8–9.
3. ‘‘Large Combustion Plants Directive-Limits Also for Existing Plants,’’ Acid News 3,
October 2000, 8–9.
4. ‘‘Ministers and Parliament Agree Air Pollution Directives,’’ ENDS Report 319, August
2001, 52; European Environmental Bureau, ‘‘National Emission Ceilings and Large Combus-
tion Plants Directives: Conciliation Could Have Been Better,’’ press release, Brussels, July 7,
2001; ‘‘Parliament to Press for Tougher Limits on Emissions by Large Combustion Plants,’’
International Environmental Reporter, March 28, 2001, 236–237.
References
Bär, Stefanie, Andreas R. Kraemer, and Neil Emmott. 2000. Der Sevilla-Prozess: Motor für
Umweltschutz in der Industrie. Wasser und Abfall 10/2000: 64–66.
Bloch, Helmut. 2001. EU Policy on Nutrients Emissions: Legislation and Implementation.
Water, Science and Technology 44 (1): 1–6.
Bohne, Eberhard. 2000. The Evolution of Integrated Permitting and Inspections of Industrial
Installations in the European Union: Report to the European Commission. Brussels: DG
Environment.
Chave, Peter. 2001. The EU Water Framework Directive—An Introduction. London: Inter-
national Water Association.
Common Implementation Strategy. 2002. Best Practices in River Basin Planning: First Draft.
Common Strategy on the Implementation of the Water Framework Directive. Brussels: Water
Directors’ Group, DG Environment.
Emmott, Neil. 1999. IPPC and Beyond—Developing a Strategic Approach to Industry for
European Environmental Policy. Journal of Environmental Policy and Planning 1 (2): 77–92.
Emmott, Neil, and Nigel Haigh. 1996. Integrated Pollution Prevention and Control: UK and
EC Approaches and Possible Next Steps. Journal of Environmental Law 8 (2): 301–311.
Enderlein, Rainer E. 2001. The ECE Water Convention: A Powerful Tool to Prevent Con-
flicts over Transboundary Waters. Geneva: UNECE.
Environmental Protection Agency. 2000. Preparation for the Proposed EU Water Framework
Directive. Dublin: Environmental Protection Agency.
230 Andrew Farmer
European Commission. 1997. Proposal for a Directive Establishing a Framework for Com-
munity Action in the Field of Water Policy. Doc. COM(1997)49. Brussels: European
Commission.
Farmer, Andrew M. 1999. The Proposed EU Water Framework Directive with Particular Ref-
erence to the Coastal Zone. Marine Environmental Management 6 (1): 41–48.
Farmer, Andrew M. 2000. Integrating the EU’s Water Policies. Inland, Coastal and Estuarine
Waters 2 (1): 6–7.
Farmer, Andrew M. 2001a. The EC Water Framework Directive: An Introduction. Water
Law 12 (1): 40–46.
Farmer, Andrew M. 2001b. Industrial Regulation and Sustainable Development. Brighton:
National Society for Clean Air.
Farmer, Andrew M. 2003. The Environmental and Sustainability Implications of EU Enlarge-
ment. Elements 18 (1): 14–17.
Farmer, Andrew M., Ian S. Skinner, and Peter Beyer. 2003. Effective Enforcement Needs a
Good Legal Base. Brussels: European Commission, IMPEL.
Farmer, Andrew M., and Patrick ten Brink. 2004. ‘‘Simple is Better’’: Effective Regulation for
a More Competitive Europe: The 1996 EU Integrated Pollution Prevention and Control
(IPPC) Directive. The Hague: The Netherlands Ministry of Housing, Spatial Planning and
the Environment.
Farmer, Andrew M., and David Wilkinson. 2001. The EU Sixth Environmental Action Pro-
gramme: Implications for Water Management. Water Law 12 (4): 236–239.
Gislev, Magnus. 2001. The EU IPPC Directive and Its Links to Other Relevant Instruments.
Paper presented at a conference on Prevention and Control of Industrial Pollution: Interna-
tional Conference on Policy Approaches. Brussels: DG Environment.
Gouldson, Andrew, and Joseph Murphy. 1998. Regulatory Realities: The Implementation
and Impact of Industrial Environmental Regulation. London: Earthscan.
Griffiths, Martin. 2002. The European Water Framework Directive: An Approach to Inte-
grated River Basin Management. European Water Management Online, 2002. Available at
http://www.ewaonline.de/journal/2002_05.pdf.
Haigh, Nigel. 2000. The IPPC Directive and BAT in a Wider Context. Proceedings of Euro-
pean Conference. In Umweltbundesamt, ed., European Conference—The Sevilla Process: A
Driver for Environmental Performance in Industry, Stuttgart, 6–7 April 2000, Proceedings,
57–60. Berlin: German Federal Environmental Protection Agency.
ICPR. 2001. Rhine 2020—Program on the Sustainable Development of the Rhine. Koblenz,
Germany: International Commission on the Protection of the Rhine.
Institute for Prospective Technology Studies. 2000. Guidance on BREFs. Seville: Institute for
Prospective Technology Studies. Available at http://eippcb.jrc.es/pages/Boutline.htm.
James, Lesley. 2001. The NGO Perspective. Paper presented at an international conference
titled ‘‘Prevention and Control of Industrial Pollution: International Conference on Policy
Approaches,’’ Seville, April 25–26, 2002.
Jendroska, Jerzy, and Stephen Stec. 2001. The Aarhus Convention: Towards a New Era in
Environmental Democracy. Environmental Liability 9 (2): 140–151.
Lanz, Klaus, and Stefan Scheuer. 2001. EEB Handbook on EU Water Policy under the Water
Framework Directive. Brussels: European Environmental Bureau.
Schnutenhaus, Jörn. 1994. Integrated Pollution Control: New German Initiatives in the Euro-
pean Environment Council. European Environmental Law Review 3 (12): 207–220.
Stern, Alex. 1995. Water in Europe: What to Expect from the EU Policy Review. Brussels:
Club de Bruxelles.
Tanzi, Attila. 2000. The Relationship between the 1992 UNECE Convention on the Protec-
tion and Use of Transboundary Watercourses and International Lakes and the 1997 U.N.
Convention on the Law of the Non Navigational Use of International Watercourses: Report
of the UNECE Task Force on Legal and Administrative Aspects. Geneva: United Nations
Economic Commission for Europe.
Ten Brink, Patrick, and Andrew M. Farmer. 2004. Analysing Views, Policies and Practical
Experience in the EU of Permitting Installations under IPPC. The Hague: The Nether-
lands Ministry of Housing, Spatial Planning and the Environment. Available at http://
www2.vrom.nl/docs/internationaal/04%20ENAP%20background%20report.pdf.
UNECE. 1996. Guidelines on Water-Quality Monitoring and Assessment of Transboundary
Rivers. Geneva: United Nations Economic Commission for Europe.
UNECE. 2000. Guidelines on Monitoring and Assessment of Transboundary Rivers. Geneva:
United Nations Economic Commission for Europe.
UNECE. 2001. Working Group on Water Management. Report on the Workshop on
Approaches and Tools for River Basin Management: Experience Drawn from the Imple-
mentation of the EC Water Framework Directive (Bratislava, 15–16 October 2001). Doc.
MP.WAT/WG.1/2001/8, December 24, 2001. Geneva: United Nations Economic Commis-
sion for Europe.
Water Directors’ Group. 2000. Common Strategy on the Implementation of the Water
Framework Directive. Brussels: DG Environment.
Water Directors’ Group. 2002. Guidance on Public Participation in Relation to the Water
Framework Directive. Brussels: DG Environment.
WWF. 2001. Elements of Good Practice in Integrated River Basin Management. Brussels:
World Wide Fund for Nature.
10
The EU Habitats Directive: Enhancing Synergy with
Pan-European Nature Conservation and with the
EU Structural Funds
Clare Coffey
The 1992 EU Directive on the Conservation of Natural Habitats and of Wild Fauna
and Flora (Directive 92/43/EEC) represents the cornerstone of EU nature conserva-
tion policy.1 Proposed by the European Commission in the late 1980s, the Habitats
Directive sought to respond to the continuing deterioration of European natural
habitats and an increasing number of seriously threatened wild species. More than
a decade after its adoption, the Habitats Directive remains the single most important
EU instrument for safeguarding biodiversity across the EU.
In addition to its importance as a policy instrument, the Directive also provides a
rich seam for anyone interested in institutional interaction. The territorial reach of
the Directive’s network of protected areas and the level of protection to be afforded
to these areas make interactions with other EU institutions almost inevitable. In ad-
dition, the Directive serves as the EU’s main instrument for implementing the 1979
Convention on the Conservation of European Wildlife and Natural Habitats (Bern
Convention), as well as key provisions of the Convention on Biological Diversity
(CBD). In general, several interactions of the Habitats Directive have generated
varied and innovative responses.
This chapter opens with a more extensive introduction of the Habitats Directive,
followed by a summary of the main interactions between the Directive and other EU
and international institutions. The core of the chapter centers on two of these inter-
actions. The first interaction is between the Habitats Directive and the Bern Conven-
tion and focuses on the two ecological networks under the respective regimes—
Natura 2000 and the Emerald Network. This interaction is particularly interesting
with respect to the way it has evolved over time. The second interaction between
the Habitats Directive and one of the EU’s main funding mechanisms—the Struc-
tural Funds—emphasizes the way attempts have been made to harness the financial
muscle of the Funds (at the heart of an initially disruptive interaction) to support
234 Clare Coffey
better implementation of the Habitats Directive. The chapter closes by drawing out
key lessons to inform more general efforts to secure less disruptive and more syner-
gistic institutional interaction.
The Habitats Directive and Natura 2000
The principal aim of the 1992 Habitats Directive is to ‘‘contribute towards ensuring
bio-diversity through the conservation of natural habitats and of wild fauna and
flora in the European territory of the Member States’’ (Article 2). This is to be
achieved by restoring or maintaining natural and seminatural habitats and wild
species of EU-wide interest at a ‘‘favorable conservation status.’’
Although the Directive contains important provisions governing the protection of
listed species, monitoring, research, and so on, it is probably best known for intro-
ducing the ‘‘Natura 2000’’ network. Natura 2000 brings together both Special
Protection Areas (SPAs) of bird habitats classified under the 1979 Birds Directive
(Directive 79/409/EEC) and Special Areas of Conservation (SACs) designated under
the Habitats Directive, which apply to the protection of nonbird species and their
habitats. The vision for Natura 2000 is one of an ecologically coherent network of
SPAs and SACs, spanning the entire territory of the EU.
The process for establishing and managing Natura 2000 sites is at the heart of the
Habitats Directive. As regards the creation of SACs, member states are to propose
sites on the basis of their relative national, European, and global significance, most
notably by taking into account their biogeographic specificities, their rarity, as well
as associated threat levels (Art. 4(1)). Suitable sites were to be included in a Euro-
pean list of ‘‘Sites of Community Importance’’ that was to be adopted by the Euro-
pean Commission by mid-1998 (Art. 4(2)). Subsequently, member states have up to
six years to formally designate the sites as SACs (Art. 4(4)). Annex III of the Direc-
tive sets out the criteria for the selection of Sites of Community Importance and the
subsequent designation of SACs. In contrast to the SAC process, the procedure for
the selection and classification of SPAs under the Birds Directive is largely left to the
member states (Art. 4).
In most respects, the Habitats Directive establishes a common level of protection
for the Natura 2000 network as a whole. The deterioration of sites is to be avoided
(Art. 6(2)), in some cases requiring the active management of sites (Art. 6(1)). Any
plans or projects likely to have a significant effect on a site’s conservation objectives
have to be subject to ‘‘appropriate assessment’’ (Art. 6(3)). Such plans and projects
EU Habitats Directive 235
should only be granted approval after ascertaining that activities will not ‘‘adversely
affect the integrity of the site concerned,’’ or if, in the light of a negative assessment
and in the absence of alternative solutions, there are ‘‘imperative reasons of overrid-
ing public interest’’ (Art. 6(4)). If damaging projects are to go ahead, member states
have to take all compensatory measures necessary to ensure the overall coherence
of Natura 2000 (Art. 6(4)), for example, by designating suitable alternative areas.
Apart from Article 6(1), these provisions apply to all SPAs and all Habitats Direc-
tive sites included in the Commission’s lists of Sites of Community Importance—in
other words, before these sites are formally designated as SACs.
Natura 2000 was described by the former Environment Commissioner Margot
Wallström as ‘‘the most ambitious initiative ever undertaken at European level,’’2
but practical implementation of the network has ‘‘been plagued by difficulties and
delays’’ (Jen 2002). By 2003, the Commission’s ‘‘Natura Barometer’’ showed that
significant progress had been made, resulting in more than 16 percent of the
territory of the then fifteen EU member states being identified for Natura 2000.
However, the process of proposing sites under the Habitats Directive was still in-
complete in all fifteen member states, seven years after the legal deadline for submit-
ting site proposals under the Habitats Directive. Moreover, just three member states
were considered to have completed the classification of bird sites required under the
1979 Birds Directive.3
Many different factors contributed to Natura 2000’s poor record. The Directive’s
scope and stringency, local resistance to Natura 2000 (Paavola 2002), the lack of
financial and human resources to support site identification, designation and man-
agement costs (Jen 2002; European Commission 2002), as well as agricultural and
other land-use pressures (European Commission 2004b) are all believed to have
contributed to implementation failures. Some of the insufficiencies have resulted
from negative interactions with other institutions, of which the one with the EU
Structural Funds is analyzed in more detail below.
Living with Other EU and International Agreements
The number of institutional interactions between the Habitats Directive and other
EU instruments or international institutions is significant. The territorial focus and
reach of the Natura 2000 network, coupled with the more general and non-site-
specific species-protection requirements, means that the Habitats Directive has
implications for the whole of the EU territory. At the same time, the successful
236 Clare Coffey
delivery of the Directive’s provisions will be influenced by a wide range of human

activities and related policies. Based on an initial inventory by Coffey and Shaw
(2001), table 10.1 provides an overview of key cases of ‘‘horizontal’’ interaction
between the Directive and four other EU environmental and nonenvironmental
policies, and ‘‘vertical’’ interaction between the Directive and five international envi-
ronmental agreements. It should be noted, however, that neither the inventory nor
table 10.1 contains all interactions between the Habitats Directive and other instru-
ments. The Habitats Directive interacts with numerous other EU sectoral institutions,
including fisheries, agricultural, and transport policies, for instance. Similarly, the
type of interaction between the Habitats Directive and OSPAR is to an extent repli-
cated in relation to other European regional seas conventions, covering the Baltic
and Mediterranean Seas.
The majority of interactions identified involved EU policy or international agree-
ments in the environmental policy field. These interactions involved all three causal
mechanisms distinguished in this volume. The Habitats Directive has drawn on con-
cepts, knowledge, and ideas from the Birds Directive and also reflects aspects of the
Convention on Biological Diversity (CBD) that was negotiated concurrently with
the Directive (Verschuuren 2002) (Cognitive Interaction). Interaction through Com-
mitment is visible from the fact that commitments under the Habitats Directive have
led to direct references in the EU Water Framework Directive (2000/60/EC; see also
chapter 9), the IPPC Directive (96/61/EC; see chapter 9), and the Environmental Li-
ability Directive (2004/35/EC). Furthermore, the European Court of Justice’s inter-
pretation of EU member states’ commitments under the Birds Directive eventually
resulted in a weakening of the related provisions in the Habitats Directive (Fairbrass
and Jordan 2001). The Birds Directive was directly altered by the Habitats Direc-
tive, something that is possible in the EU system where one piece of legislation can
amend another piece of legislation. In the case of the Bern Convention, the commit-
ment to its implementation in the EU provided an important driving force in the
preparation of the Habitats Directive. In turn, the commitment to Natura 2000
contained in the Habitats Directive largely drove the development of the Bern Con-
vention’s Emerald program, as is further detailed later on in this chapter. With re-
spect to Behavioral Interaction, the implementation of the Habitats Directive has
been supported by the implementation of the Water Framework Directive, the
IPPC Directive (see chapter 9), OSPAR, the Agreement on Small Cetaceans of
the Baltic and North Seas (ASCOBANS), and the Pan-European Biodiversity and
Landscape Diversity Strategy. It should also be supported by implementation of the
Table 10.1
Institutional interactions of the Habitats Directive
EU Birds Directive Birds Directive has served as a model for Habitats
Directive (in parts).
Habitats Directive has supported achievement of
objectives of Birds Directive, including by strengthening

certain aspects of protection of bird sites.
Strict interpretation of site protection under the Birds
Directive by European Court of Justice results in

Habitats Directive being weaker than initially proposed.
Habitats Directive includes provisions that directly
weaken level of protection afforded to sites designated

under Birds Directive.
EU Water Framework Directive A reference to Natura 2000 sites was included in the
(WFD) WFD.
WFD supports achievement of objectives of Habitats
Directive, in general and specifically by introducing

legal deadline for reaching good ecological status in
water-dependent Natura 2000 sites.
EU Environmental Liability Habitats Directive is used in part for defining damage
Directive covered under Liability Directive.
Liability Directive introduces additional incentive for
behavior that is compatible with the Habitats Directive.

EU Structural Funds Structural Fund projects that conflict with nature
conservation objectives undermine implementation of
the Habitats Directive.
Cross-compliance rule in the Structural Funds is used
to encourage member states to submit site lists under

the Habitats Directive, and to avoid damage to sites.
OSPAR Convention for the OSPAR supports implementation of Habitats Directive
protection of the Northeast in the marine environment.
Atlantic
Bern Convention on the Necessity to implement Bern Convention in the
Conservation of European Community drives preparation of Habitats Directive.
Wildlife and Natural Habitats Habitats Directive supports achievement of objectives
of Bern Convention.
Habitats Directive Natura 2000 program provides
model/reference point for development of Bern

Convention Emerald program.
Convention on Biological Habitats Directive has adopted concepts also
Diversity (CBD) contained in the CBD, which was negotiated
simultaneously.
Habitats Directive supports achievement of objectives
of CBD in the EU.

238 Clare Coffey
Table 10.1
(continued)
Convention on Migratory CMS Agreement on Small Cetaceans of the Baltic and
Species of Wild Animals (CMS) North Seas (ASCOBANS) supports achievement of
objectives of Habitats Directive at the national level.
Pan-European Biodiversity and PEBLDS supports implementation of Habitats
Landscape Diversity Strategy Directive, particularly through its focus on ecological
(PEBLDS) corridors between sites.
Source: Coffey and Shaw 2001.
Liability Directive, which requires member states to introduce liability for damage
caused to Natura 2000 sites. The Habitats Directive has supported the implementa-
tion of the Birds Directive, the Bern Convention, and the CBD.
While disruption was the exception rather than the rule in the case of interactions
between the Directive and other environmental institutions, it would appear to be
more commonplace in interactions with legal instruments related to other, nonen-
vironmental policy fields. Among the interactions with other environmental instru-
ments, disruption only occurred with the Birds Directive because the Habitats
Directive lowered the level of protection afforded to sites classified under the Birds
Directive. Table 10.1 contains one exemplary case of disruptive interaction with
nonenvironmental instruments concerning the EU Structural Funds that have funded
various projects that conflict with the nature conservation objectives of the Habitats
Directive (Behavioral Interaction). Although not contained in table 10.1, the rela-
tionship between the Habitats Directive and EU policies such as the Common
Agricultural Policy (CAP), the Common Fisheries Policy (CFP), and the Common
Transport Policy is also problematic. In particular in response to these disruptive
interactions, however, nonenvironmental policies have also at times supported the
Habitats Directive. The Structural Funds, for instance, have progressively been
amended, to allow greater scope to use funds for nature conservation purposes. As
discussed later in this chapter, a form of ‘‘cross-compliance’’ has in addition been
used in an effort to improve the compatibility between the Habitats Directive and
the Funds. Similar examples exist with respect to the CAP (Baldock, Dwyer, and
Sumpsi Vinas 2002) and the CFP (e.g., Regulation 602/2004 protecting deep water
corals around the Darwin Mounds from deepwater trawling). It has not been the
case, however, that disruptive interactions have ceased altogether.
In seeking to promote synergies and reduce disruptions, one question that arises
is whether interactions are actually intended or simply incidental. In most of the
identified cases of synergistic interaction with the Habitats Directive, whether this
involved environmental policies or not, the positive influence was intended either
by the source or target of the interaction, or both. The opposite can on the whole
be said of disruptive interactions, which were generally found to be unintended at
the level of the policy, even if the interaction was predictable. This suggests that po-
tential disruptions, once identified, may be prevented or remedied more easily than
would be the case if disruptions were intended by one or other of the institutions. In
all cases (including the initially disruptive interactions), a response has been gener-
ated that has improved synergies and strengthened the overall coherence of policies
in favor of the Habitats Directive and Natura 2000. There is little doubt, however,
that efforts to improve synergy can and should be strengthened further.
The Habitats Directive and the Bern Convention: An Evolving Relationship
The Bern Convention came into force on June 1, 1982, six years before the Habitats
Directive was formally proposed (European Commission 1988) and ten years prior
to the Directive’s adoption. This pan-European Convention acknowledges in its pre-
amble ‘‘that wild flora and fauna constitute a natural heritage of aesthetic, scientific,
cultural, recreational, economic and intrinsic value that needs to be preserved and
handed on to future generations.’’ It has been ratified by forty-five European and
African states, as well as the European Community (http://www.coe.int).
The Bern Convention was drafted by the Council of Europe, Europe’s oldest
political organization unrelated to the EU, founded in 1949. Work under the Con-
vention is coordinated by a Standing Committee that is open to all parties to the
Convention and meets every year. The Standing Committee, among other things,
makes recommendations to the parties and reviews the implementation of the Con-
vention (http://www.coe.int).
As will be demonstrated below, the coevolution of the Habitats Directive and
the Bern Convention largely followed the causal mechanism of Interaction through
Commitment. First, the commitments accepted by the EU and its member states
under the Bern Convention contributed to the emergence of the Habitats Directive.
The Directive in large part implements the Convention, but also goes beyond the
Convention’s provisions, in particular by establishing a coherent network of pro-
tected sites (Natura 2000). In a second phase, the commitment to Natura 2000
under the Habitats Directive was a driving force in the establishment of the Conven-
tion’s Emerald Network. Moreover, the Directive is now driving habitat protection
240 Clare Coffey
in a pan-European context. While the first phase of the interaction is briefly outlined
below, the main focus of this section is on the second phase.
Initial Interaction: A ‘‘Simple’’ Case of EU Implementation?

In 1982, when the Bern Convention entered into force, the 1979 Birds Directive
was the only EU (legally then: European Economic Community) directive explicitly
addressing nature conservation issues. However, more general ideas on nature
conservation had been set out in the early Environmental Action Programs (Haigh
2003), with the second Action Program (1977–1981) including among its objectives
to ‘‘maintain a satisfactory ecological balance and ensure the protection of the bio-
sphere’’ (European Commission 1977). The Commission’s broader ambitions in the
area of nature conservation had thus been outlined.
By having adopted the Birds Directive, the EU had assumed competence to act
externally on those issues covered by the Directive. Three months after adoption of
the Birds Directive by the Council, the Commission came forward with a proposal
to ratify the Bern Convention (European Commission 1979). The proposal was
adopted in 1981 (European Commission 1982), and the EU came under pressure
from the European Parliament and members of the nongovernmental community
to go beyond the Birds Directive and to adopt additional EU implementing legisla-
tion (Haigh 2003).
However, opposition among certain member states to further nature conservation
legislation, while weakening, persisted. At the time of ratification of the Bern Con-
vention, opponents had been assured that the Convention would not lead to addi-
tional EU legislation in those areas not covered by the Birds Directive (personal
communication with Commission official). Now, these member states expressed
their doubts over the effectiveness, practicality, and desirability of a directive that
covered all of the issues provided for under the Convention. It has even been sug-
gested by some that some member states were actively trying to deflect initiatives
for an EU instrument on habitat protection, by instead proposing the reinforcement
of relevant provisions under the Bern Convention (personal communication with
Council of Europe official). This enhanced resistance was, arguably, a sign that pres-
sure on the opponents to a directive was increasing and support for their case weak-
ening. In particular, to the extent that new EU legislation on nature conservation
was merely to implement the Bern Convention, it would not introduce new sub-
stantive commitments because all EU member states and the EU itself were already
bound by the Convention. It thus became increasingly difficult to argue against
adopting implementing EU legislation.
Figure 10.1
EU Habitats Directive implements Bern Convention
Following the establishment of a clear legal basis for EU nature conservation leg-
islation through the 1987 Single European Act, the Commission came forward with
a proposal for a new directive in 1988 (European Commission 1988). The Habitats
Directive was adopted four years later (figure 10.1).
In many respects, the Directive implements the Convention in the EU and thus
contains many of the Convention provisions. The Directive, for instance, adopted
relevant definitions used by the Convention and more importantly, reinforced the
emphasis given to habitat conservation by Recommendations 14 and 16 of the
Convention’s Standing Committee. A reference to ‘‘species [and endangered natural
habitats] requiring specific habitat conservation measures’’ in Recommendation 14
is echoed in Annexes I and II of the Directive, which list natural habitat types as
well as animal and plant species of Community interest that require the designation
of Special Areas of Conservation (SACs). The phrase ‘‘Special Areas of Conserva-
tion’’ was itself inspired by the terminology employed by Recommendation 16 on
‘‘Areas of Special Conservation Interest’’ (Council of Europe 2001). Other provi-
sions of the Directive were also influenced by the Convention’s provisions on habitat
protection4 (Council of Europe 2001).
Once in place, the Directive provided an additional means, besides the Birds
Directive, of implementing and enforcing the Bern Convention in the EU. EU law,
unlike international law, imposes legal obligations that are directly applicable in the
member states and must be invoked by national courts even if national implementing
242 Clare Coffey
legislation is lacking or conflicting (Haigh 2003). The provisions relating to Natura

2000 site protection and management are also believed, at least by some EU mem-
ber states, to have direct effect (Verschuuren 2002). This means that they can confer
rights on individuals that can be relied on against the state and other individuals
or companies (Haigh 2003). Moreover, the European Commission has powers,
resources, and duties to supervise, support, and enforce EU legislation, reaching far
beyond those available to the Council of Europe. In accordance with Article 226 of
the Treaty of Rome (as amended), the Commission can, for example, start legal pro-
ceedings against any member state failing to comply with EU legislation, potentially
resulting in member states being referred to the European Court of Justice. Persistent
noncompliance can even result in fines being imposed on the guilty member state
(Art. 228). Compliance with the Bern Convention, by contrast, is followed up by
the opening of case files and ‘‘on-the-spot’’ appraisals (Council of Europe 1999b),
a procedure that has no formal legal implications.
Apart from serving as an instrument to implement the Bern Convention in the EU,
the Habitats Directive went beyond the Convention’s legal provisions and reflected
some of the criticisms of the Convention (Haigh 2003). In particular, it was more
explicit than the Bern Convention about the need for site protection and a coherent
European ecological network, and set a precedent for transboundary coordination
in habitat protection by establishing Natura 2000 as a core instrument. In addition,
the Directive set a clear time frame for the transposition and implementation of its
provisions and clear reporting requirements. The Habitats Directive thus repre-
sented more than just a simple case of transposition, if such a thing exists.
Changing the Direction of Influence: The Bern Convention Becomes the New
Target
The elements of the Habitats Directive that went beyond a simple implementation of
the Bern Convention mark the beginning of a reversal in the direction of influence
between the Directive and the Convention, resulting in the Natura 2000 concept
being ‘‘exported’’ to the Bern Convention (figure 10.2). While the Habitats Directive
was intended to improve the effectiveness of the Convention but not to influence the
Convention itself, the Bern Convention secretariat and Standing Committee
responded to the newly adopted Habitats Directive. In effect, the Emerald Network
emulated and developed in parallel to Natura 2000, extending the network of pro-
tected areas beyond the EU and into the territories of the other parties to the Bern
Convention.
Figure 10.2
Habitats Directive affects regulatory approach of Bern Convention
The change in roles between the two institutions emerged from converging inter-
ests among three groups of actors: the EU member states and the European Com-
mission; the countries of Central and Eastern Europe waiting to join the EU; and
the Convention secretariat. Being committed to the Habitats Directive and Natura
2000, the EU member states and the European Commission had an interest in ensur-
ing that pan-European commitments would follow a similar route (and certainly
not go beyond EU requirements). First, the Council of Europe’s endorsement of
Natura 2000 meant that EU member states would be considered to be automatically
complying with the Bern Convention’s site provisions. Second, the effectiveness
of Natura 2000 would be improved, if neighboring countries adopted similar ap-
proaches. Third, looking ahead to enlargement, the Emerald Network could help
pave the way for the quick adoption of the Natura 2000 approach by the accession
countries.
For the accession countries of Central and Eastern Europe, the prospect of joining
the EU and the related need to implement Natura 2000 was an important motiva-
tion. Since they had to comply with the Habitats Directive in due course, establish-
ing the Emerald Network imposed relatively limited additional costs and instead
promised to help with the preparations for joining the EU. By 1996, delegates from
the Central and Eastern European states were therefore stating their keen interest
244 Clare Coffey
in the development of an ecological network and calls for an increase in priority

awarded to habitat protection under the Bern Convention were being heard (Coun-
cil of Europe 1996).
The secretariat of the Convention also recognized the potential benefits of syn-
chronizing developments under the Habitats Directive and the Bern Convention,
using and building on the approach adopted for Natura 2000. It would not only fa-
cilitate preparations for Natura 2000 among accession countries but also provide a
comparable ecological network to those countries unable, reluctant, or otherwise
unsuitable to join the EU.
Work on the Convention’s Areas of Special Conservation Interest had already
started in 1989 (Standing Committee Recommendation 16), but was subsequently
suspended to await finalization of Natura 2000. The changes in the political map
of Europe in the late 1980s and early 1990s contributed to this suspension as the
secretariat of the Convention was redirecting its attention toward extending its
scope to include the new democracies of Central and Eastern Europe (Council of
Europe 2001). By 1996, Natura 2000 was starting to be implemented and prepara-
tions for accession to the EU by Central and Eastern European countries were
advancing. The political priority awarded to the accession process and associated
momentum gained at the national and international level effectively promoted the
creation of a preaccession tool, which helped applicant countries meet the require-
ments for EU accession. The secretariat consequently ‘‘revived’’ the idea of a pan-
European network (Council of Europe 2001), and the Standing Committee adopted
Resolution No. 3 concerning the setting up of a Pan-European Ecological Net-
work in 1996 (available at http://www.coe.int). In doing this, the Habitats Directive
served as the source of inspiration. The regulatory provisions for the Emerald Net-
work were deliberately drafted in such a way that Natura 2000 would effectively
constitute a regional unit within the Emerald Network.
The importance of EU enlargement as a driving force for the interaction has been
confirmed by efforts to align implementation within the accession countries through
the Emerald Network pilot-project scheme (Council of Europe 1999a), offering
practical and financial support for the establishment of the network in non-EU
member states. Pilot countries benefited from participating in training workshops
and obtaining access to the Emerald Network software, which was technically com-
patible with the respective software used for Natura 2000. Successful completion of
the pilot phase required, among other things, the submission of a list of proposed
Areas of Special Conservation Interest (ASCIs) to the Standing Committee to the
Bern Convention (Council of Europe 1999a). By 2002, eighteen countries had par-
ticipated in the pilot-project scheme, including the ten new EU member states and
three further accession countries (Council of Europe 2001). The focus had thus
been on the countries preparing to join the EU, emphasizing the preparatory role
of the Emerald Network and the motivating force of EU membership.
While the Emerald Network might also have been established in the absence of
the Habitats Directive and the accession process, it is unlikely to have occurred at
the same pace and in the same form. The ‘‘network concept’’ was already a well-
established conservation tool within the Council of Europe, which had already
launched the European Network of Biogenetic Reserves in 1976. Also, the need for
greater emphasis on habitat conservation had been acknowledged previously. How-
ever, the operationalization of the network concept in the context of the Emerald
program was heavily influenced by Natura 2000. The influence of Natura 2000 is
also obvious from the implementation of both networks. For instance, when differ-
ences in the lists of habitat types used by the two institutions caused problems, it
was informally agreed that accession countries were, for the time being, to concen-
trate on those habitat types listed in the Habitats Directive.
Continuing Efforts to Coordinate

Beyond the decision to establish the Emerald Network, coordination efforts between
the Habitats Directive and the Bern Convention regarding their networks of conser-
vation sites have continued, supported in particular by the European Environment
Agency (EEA) and its Topic Centre on Nature Protection and Biodiversity (ETC/
NPB), the Convention secretariat, and the European Commission.
The EEA and its Topic Centre have provided scientific and technical support in the
development of an integrated regime for nature conservation at the Community and
pan-European level. While established under EU law, the EEA serves an extended
membership of the Council of Europe. A memorandum of cooperation was signed
between the Convention’s secretariat and the EEA to ensure the ‘‘mutual compati-
bility of data, information and approaches for measurement, analysis and assess-
ment in the environment field . . . and avoid duplication of activities’’ (Council of
Europe 2000). On this basis, the EEA has monitored conservation action across Eu-
rope and provided a common database of designated areas. The EEA’s role in estab-
lishing greater synchrony between various European habitat classification systems,
and its input in the development of compatible software, has significantly improved
the synergistic development of the two institutions. In particular, it has helped
246 Clare Coffey
ensure that the scientific and technical aspects of developing Natura 2000 and the
Emerald Network are taken forward in a parallel and mutually reinforcing process.
Important, although less central, has been the ongoing contact between the Bern
Convention secretariat and the European Commission. The former has generally
been, with the help of representatives of certain contracting parties, the driving force
in facilitating better coordination between Natura 2000 and the Emerald Network.
The European Commission’s involvement appears to have fluctuated with personal
initiative and individual responsibility. According to officials from the Council of
Europe, a change in staff responsible for overseeing work on the Bern Convention
led to an improvement of cooperation in the mid-1990s. In 2002, there was nobody
directly responsible for the Bern Convention within the Directorate General Envi-
ronment of the Commission, suggesting that coordinative action may have been
given a lower priority.
Perfect Interaction or Room for Improvement?

The relationship between the Habitats Directive and the Bern Convention has been
synergistic in both directions. The Bern Convention was a key driver behind the
adoption of the Directive that served to strengthen the effectiveness of the Conven-
tion. Subsequently, the Directive’s concept of Natura 2000 has driven, and contin-
ues to drive, the development of the Emerald Network under the Convention. The
Habitats Directive generated a significant impetus and incentive for countries pre-
paring to join the EU to participate in the Emerald Network. Non-EU countries
have also adapted their habitat site protection to Natura 2000. In consequence, a
more effective pan-European system of habitat protection has emerged, facilitated
by scientific and technical assistance of the EEA and its Topic Centre as well as co-
ordination of the Convention secretariat and the European Commission.
While this may seem to represent an example of a perfect institutional interac-
tion, there is room for further improvement. Interinstitutional coordination could
be strengthened, not least because it has often remained dependent on personal
initiative, individual responsibility, and good timing. Improvements could be made,
for example, to the harmonization of habitat lists and the establishment of site-
management standards. Both the Directive in its Article 11 and the Convention
(para. 4(e) of Standing Committee Recommendation 16 of 1989) require their mem-
bers to monitor the conservation status of sites under protection. Yet guidelines for
appropriate monitoring standards at the European level are lacking, and monitoring
schemes at the national level diverge considerably. The European Commission was
expected to address the issue of ecological monitoring under the Habitats Directive,
and was to work in collaboration with the EEA to draw up a set of guidelines for
EU member states. Full coordination on such issues with the Council of Europe
would appear advisable.
In conclusion, the synergistic development of the two institutions has so far pre-
dominantly been successful and ongoing work suggests that this trend should con-
tinue. Future progress in creating spatial as well as institutional coherence between
the two ecological networks depends on the continued interinstitutional coopera-
tion, and on technical support. This is likely to be enhanced by ongoing activities.
Making the Most of Differing Objectives: Harnessing the Power of the Structural
Funds
The long-standing interaction between the Habitats Directive and the EU Structural
Funds has been the result of the territorial overlap of the two institutions, combined
with differences in objectives: whereas the main objective of the Habitats Directive is
to contribute to the conservation of nature and biodiversity, the Structural Funds’
objective is to secure social and economic development by supporting regional de-
velopment and cohesion within the EU. The Structural Funds represent the EU’s
main expenditure item, aside from the Common Agricultural Policy.
While not necessarily at odds, the two institutions have not generally been imple-
mented in a consistent or mutually supportive way. In particular, projects supported
by the Structural Funds have damaged Natura 2000 sites. The response has been
twofold. First, reforms of the Structural Funds required EU member states to com-
ply with EU nature conservation objectives. Second, on that basis, the European
Commission intervened in a rather innovative way in the late 1990s to ‘‘force’’ the
member states to effectively implement the Habitats Directive. As a result, the
Natura 2000 site-proposal process was speeded up and the disruptive interaction
between Natura 2000 and the Structural Funds improved, although problems per-
sisted. This section outlines the different stages and characteristics of the interaction,
with a particular focus on the Commission’s initiative to secure a more synergistic
coexistence between the two instruments.
The EU Structural Funds

The Structural Funds are one of the EU’s two main funding instruments and aim to
promote the economic and social development of disadvantaged regions, sectors,
248 Clare Coffey
and social groups in the EU. There are four Structural Funds. The European Re-
gional Development Fund provides support for productive investment, basic infra-
structure, and local development as well as for employment initiatives and small
and medium enterprises (Regulation 1783/1999). The European Social Fund is to
develop the labor market and human resources (Regulation 1784/1999). The Euro-
pean Agriculture Guidance and Guarantee Fund (Guidance Section) supports agri-
culture and rural development (Regulation 1257/1999). The Financial Instrument
for Fisheries Guidance provides funding for the structural adjustment of the fisheries
sector (Regulation 1263/1999).
The overall aim of assistance is to reduce social and economic disparities in Eu-
rope, with a significant proportion of funding targeted at Europe’s most disad-
vantaged (‘‘Objective 1’’) areas, which often also happen to be the EU’s most
important areas in nature conservation terms. Together the Funds account for ap-
proximately one-third of the EU’s annual budget, totaling EUR 195 billion over the
period 2000–2006. Moreover, since almost all assistance is released only if matched
with national and private funds, the Structural Funds have the effect of mobilizing
significant additional expenditure beyond the EUR 195 billion (e.g., see European
Commission 2004b).
The way assistance is provided has changed considerably over the last thirty
years, originally involving the European Commission directly in the process of
selecting and funding individual projects. The Commission’s remit has since receded,
with funding distributed on the basis of multiannual programs. The Funds now rep-
resent one of the most devolved policies of the EU. Aid is distributed to the regions
and sectors on the basis of seven-year rolling programs. These are drawn up and
subsequently implemented by the member states. Within the basic regulatory frame-
work agreed by the EU, the member states decide where to direct aid and how to
make it available (Haigh 2003). Member states and regional authorities conse-
quently play a major role in determining the content, use, and eventual impact of
assistance. This provides significant scope for tailoring funding to suit specific re-
gional and national priorities and needs. It also increases the chances of funding
being used in ways that conflict with EU nature conservation requirements.
Structural Funds Undermine Implementation of Habitats Directive

Even before the Habitats Directive had been adopted, there was evidence of Struc-
tural Fund projects contributing to the deterioration of important natural areas,
including EU bird sites (SPAs) and other sites of national and international impor-
tance (see Baldock and Long 1988). Until the 1990s, Structural Fund regulations did
not contain environmental conditions for funding, nor did they explicitly require
member states to observe all EU laws or provide for specific safeguards to help en-
sure that funding was not used in ways that conflicted with other EU laws. More-
over, those responsible for overseeing the implementation of funding programs
rarely involved environmental authorities in the programming or project-selection
phases (Coffey 1998). The result was that Structural Funds assistance was being
used to stimulate infrastructure projects in different parts of the EU, some of which
conflicted with the Community’s nature policy (Long 1995, after Lenschow 2002).
At the beginning of the 1990s, pressure mounted in favor of a general greening of
the Funds, resulting in the incorporation of more effective environmental safeguards
in the Structural Funds. The conservation organizations World Wild Fund for Na-
ture (WWF) and the Royal Society for the Protection of Birds (RSPB) organized a
well-orchestrated campaign to this end. In July 1993, a year after the adoption of
the Habitats Directive, a major revision of the Structural Funds covering the
period 1994 to 1999 led to agreement on more extensive environmental provisions,
including a requirement for member states to produce environmental profiles of
the regions for which they were seeking support. These were to contain state-of-
the-environment reports, an evaluation of the environmental impact of programs
‘‘in terms of sustainable development, in agreement with the provisions of Commu-
nity law in force,’’ and a description of arrangements to ensure compliance with EU
environmental law (Wilkinson 1994).
Despite these strengthened provisions, the Structural Funds continued to trigger
Behavioral Interaction that undermined the effective implementation of the Habi-
tats Directive (see figure 10.3). The Funds created an incentive for regions to attract
financial assistance irrespective of the requirements of the Habitats Directive. Ac-
cordingly, the member states continued to not always respect agreed environmental
commitments and to pursue environmentally problematic projects. Poor compliance
with the Habitats and Birds Directives was linked to fear, on behalf of economic
authorities, that development or growth objectives would be compromised (Len-
schow 2002). In effect, the financial ‘‘muscle’’ of the Structural Funds gave them
greater weight when decisions were made at the national, regional, and local level,
compared to environmental policies that were supported only with a relatively
minor budget at the EU level.5
As a consequence, the Habitats Directive was often given second place by
member states or regional authorities in Structural Funds programs and in the
250 Clare Coffey
Figure 10.3
EU Structural Funds undermine implementation of Habitats Directive
project-selection process (Lenschow 2002). The higher-profile cases included projects

to develop a visitor centre in the Burren in Ireland6 and to divert the river Acheloos
in Greece.7
In addition, the process of preparing Structural Funds programs and selecting
projects arguably provided an incentive for regions to delay the submission of lists
of protected sites. Because designation of sites was associated with the obligation to
protect them, nondesignation provided more freedom in the choice of fundable de-
velopment projects. Hence, designation might have jeopardized chances of securing
Structural Funds support. Thus, by June 1998 very limited progress had been made
in identifying sites, although member states had been legally required to submit lists
of proposed sites to the European Commission by mid-1995 (Art. 4(1) Habitats
Directive).
Reforming the Structural Funds and Ensuring Compliance with Their Rules
The failure of member states to implement the environmental safeguards and the
disincentive provided by the Structural Funds with regard to compliance with envi-
ronmental legislation, including the Habitats Directive, reinforced efforts to reform
the Structural Funds in the mid-1990s. Following discussions between the Director-
ate General Environment of the Commission and European Parliamentarians, the
European Parliament’s considerable powers in relation to the EU’s annual budget

were used to underpin calls for a reform of the Funds. In October 1995, the Parlia-
ment threatened to freeze 50 percent of the EU’s Structural and Cohesion Funds
budget unless the Commission produced an environmental code of conduct govern-
ing the future use of the Funds (Wilkinson 1998). The Commission responded rap-
idly by adopting a communication on ‘‘Cohesion Policy and the Environment’’ that
set out a ten-point plan for tightening environmental requirements in the current
Structural Funds regulations and promised to incorporate strengthened provisions
in the 1999 Structural Funds revisions (European Commission 1995).
The main rules governing all four Structural Funds were strengthened in 1999.
The new rules provided the legal framework for the Funds for the period 2000–
2006. According to Article 12 of Regulation 1260/1999 laying down the general
provisions governing the Structural Funds, it became an explicit requirement
that ‘‘operations financed by the Funds . . . be in conformity with the provisions of
the Treaty, with instruments adopted under it and with Community policies and
actions, including the rules on . . . environmental protection and improvement.’’
Plans and programs were to be drawn up and monitored by representative ‘‘partner-
ships’’ of national and local stakeholders, ‘‘taking account of the need to promote
. . . sustainable development through the integration of environmental protection
and improvement requirements’’ (Art. 8). The implementation of Structural Fund
assistance was also to be subject to ex ante, midterm, and ex post evaluations
designed to assess their impact and effectiveness (Art. 40).
Despite the requirement for all funding to be in compliance with other EU legisla-
tion, problems between the Habitats Directive and the Structural Funds continued.
In particular, member states were putting forward new draft EU regional funding
programs outlining intended expenditures over the period 2000–2006, without first
having complied with the requirement under the Habitats Directive to submit com-
pleted lists of proposed Natura 2000 sites. In the absence of proposed site lists, it
was virtually impossible for the Commission to assess whether funding programs
would or would not have damaging implications for areas suitable for nomination
under the Habitats Directive.
Under these circumstances, the new 1999 Structural Funds rules provided a plat-
form for different parts of the Commission to jointly force member states to come
forward with Natura 2000 site proposals and details of how they were going to
ensure sites were not damaged by Structural Funds expenditure. The Commission
initiative took shape in July 1999 when an agreement was made between the two
252 Clare Coffey
Commissioners responsible for regional policy and environment, respectively. Fol-

lowing earlier commitments regarding greening the Structural Funds, and con-
tinued pressure from the European Parliament and nongovernmental organizations
(NGOs), the Commission decided to exercise its powers to withhold Structural
Funds unless proposed lists of sites under the Habitats Directive were submitted to
it. Each member state’s Structural Funds plans and programs were to be examined
for their environmental merit, and applications would not be processed ‘‘if elements
concerning respect for Natura 2000 are insufficient’’ (Haigh 2001). The Commis-
sion was stating that it was unable to assess the environmental impact of new Struc-
tural Funds programs unless it had lists of proposed sites at its disposal (European
Commission 2000).
In March 2000, the Commission adopted guidelines ‘‘concerning the relationship
between implementation of the Structural Funds programming for 2000–2006 and
the respect of Community environmental policy (Natura 2000 Directives)’’ (Fischler
2000). The guidelines stated that
programming documents . . . must contain clear and irrevocable commitments to guarantee
consistency of their programmes with the protection of sites as provided for under Natura
2000. An explicit part of such a commitment should be to send in proposed lists under the
‘‘Habitats’’ Directive (Article 4(1)) together with the related scientific information. . . .
States concerned must also give a formal guarantee that they will not allow sites to be
protected under Natura 2000 to deteriorate during operations part-financed by the Structural
Funds. They should also commit themselves to providing the Commission, in their program-
ming documents or at least when presenting their programming complement for each opera-
tional programme, with information about the steps they have taken to prevent the
deterioration of sites to be protected under Natura 2000.8
The Commission action was based on the existing provisions of both the Habitats
Directive and the relevant Structural Funds Regulation but did not require changes
of these instruments themselves. Rule adaptation was limited to interpreting stan-
dards. The existence of explicit EU rules requiring ‘‘cross-compliance’’ between
Structural Funds expenditure and EU environmental legislation provided a clear
legal basis for this way of encouraging member states to adapt their behavior. With-
out such a legal basis, the Commission would not have been able to take the initia-
tive and threaten withholding EU funding.
The Commission initiative relied on the common interest of the two Commission
Directorates General on regional policy and the environment, personal initiative of
Commission officials responsible for the Habitats Directive, as well as the Commis-
sioners for Regional Policy and Environment. The environment side saw it as a
Figure 10.4
Cross-Compliance requirement makes Structural Funds support implementation of Habitats
Directive
means of advancing the Natura 2000 site-selection process, while the regional policy
side had an interest in applying its earlier commitments regarding greening of the
Structural Funds so as to avoid accusations and threats from the European Parlia-
ment and NGOs. Despite the pivotal role played by officials and Commissioners,
the initiative survived changes in staff and the departure of the two Commissioners
in September 1999, when a new Commission was agreed.
The threat of the Commission to withhold financial assistance changed the effect
of the Behavioral Interaction on the Habitats Directive quite dramatically (see figure
10.4). The financial strength of the Structural Funds that contributed to the initial
disruptive interaction was now used as a powerful lever to secure progress in the
Natura 2000 site-designation process. The Structural Funds provided an incentive
to member states and their regions to speed up designation of habitats sites and
to ensure proper protection of those sites. Specifically, the European Commission
warned five countries (Ireland, Germany, France, the Netherlands, and Portugal)
that it might not be able to assess certain draft Structural Funds programs covering
the 2000–2006 period, unless a sufficient number of proposed Natura 2000 sites
had been put forward to the Commission (IEEP 1999).
Since significant funds were at stake, the member states concerned took the threat
very seriously and adapted their behavior. For example, following criticism of the
254 Clare Coffey
UK list of proposed sites at a meeting in September 1999, and facing the prospect of
funding being withheld in certain regions, the United Kingdom announced in sum-
mer 2000 that it was forwarding a further list of proposed sites to the Commission
(Defra 2001). In addition, further safeguards were put in place to ensure that future
Structural Funds expenditure would not undermine sites themselves. This included a
requirement for authorities responsible for monitoring Natura 2000 sites to submit
a signed declaration as to whether or not projects were likely to have significant neg-
ative effects on sites (personal communication with European Commission official).
However, the Commission was somewhat reticent to carry through its threat of
withholding funds in the case of member states not submitting site lists. Thus, while
insufficiencies continued to be noted in relation to all fifteen member states’ pro-
posed site lists, funding was actually only blocked in relation to the Wallonia region
of Belgium (personal communication with European Commission official).
Overall, as a result of changes of the rules governing the Structural Funds agreed
to in the 1990s, the European Commission has been empowered to require cross-
compliance by member states with the Habitats Directive’s requirements. This pro-
vided significant incentives for member states to submit lists of proposed Natura
2000 sites and ensure their proper protection. The Commission thus harnessed
the initial cause of the disruptive interaction—the Structural Funds’ financial
‘‘muscle’’—to induce compliance with the Habitats Directive and to ensure compat-
ibility between both institutions. The threat of withholding funding indeed sup-
ported the implementation of the Habitats Directive by leading member states to
submit lists more rapidly than would otherwise have been expected (figure 10.4).
However, an unwillingness to carry through the threat and actually withhold fund-
ing may have contributed to the fact that national site lists were still incomplete in
all cases by the end of 2003.9
Conclusions
The cases of interaction analyzed follow the causal mechanisms of Interaction

through Commitment and Behavioral Interaction. The Bern Convention cases cen-
ter on the changing relationship between the two institutions. A rather ‘‘simple’’
case of regional implementation of the international commitments of the Bern Con-
vention is followed by a spread of Habitats Directive commitments across Europe,
well beyond the boundaries of the former fifteen EU member states (Interaction
through Commitment). The interaction between the Structural Funds and the Hab-
itats Directive, in contrast, involved modifying the driver of an initially disruptive

interaction—funding—so as to make it contribute to nature conservation objectives
in a constructive and mutually supportive manner. The existing provisions and tools
of the regimes were used together as a means of changing the behavior of actors,
thus reducing the level of disruption between the regimes, even if problems persist
(Behavioral Interaction).
As demonstrated by the Structural Funds cases, the cause of an interaction does
not necessarily tell us anything about the effect. The political (or economic) weight
given to each institution can be much more important. It is quite feasible that,
endowed with billions of euros per annum, the Habitats Directive would have
impeded delivery of the Structural Funds objectives so that the direction of influ-
ence between both institutions would have been different. Furthermore, the role of
broader political processes and the influence of nongovernmental actors can be crit-
ical in generating changes in interactions, in this case redirecting the power of the
Funds to work in favor (at least in part) of the Habitats Directive.
Membership issues and associated commitments have been central to the Bern
Convention case. This involved not only differences in membership of the two insti-
tutions at any given time, but also expected future changes in membership. The
prospect of EU enlargement, and its political and legal ramifications, has been par-
ticularly significant in generating a strong response to the Habitats Directive, as the
new EU member states prepared to take on new commitments under EU law.
In the interactions that were studied in detail, responses to the interactions have
always led to improvements in synergy. The strength of responses was influenced
by a number of particular factors, including the work of supporting agents (the Eu-
ropean Environment Agency) and personal initiative (Commission and Council of
Europe officials). Moreover, the interactions analyzed resulted in particularly strong
and rather unexpected responses, primarily in the form of behavioral adaptation
and adjustments made within the confines of the institution itself.
Several reasons are suggested for the relative lack of responses involving changes
to the institutions directly—that is, by collective decision making. First, the collec-
tive decision-making process is slow and is not guaranteed to result in an ‘‘optimal’’
outcome. Alternative solutions are therefore often preferable, at least in the short or
medium term. Second, the policy framework may not be directly responsible for the
disruptive interaction even if policy changes could help to mitigate conflicts. The
real problem may lie with governments, agencies, or other actors that are unable to
coordinate multiple objectives sufficiently, instead prioritizing one set of objectives
256 Clare Coffey
over another. This may be particularly problematic in relation to devolved or frame-

work legislation where member states have more scope to decide how objectives are
delivered. The challenge in these and other cases is to help or indeed force actors to
address trade-offs between objectives. The best way of doing that in the short term
may not involve adopting new or adapting existing institutions.
Notes
This chapter is based on an earlier draft by Coffey and Richartz (2002).

1. The term European Union (EU) will be used broadly throughout this chapter, to include
the period before the Treaty of Maastricht.
2. Margot Wallström, Commissioner for the Environment, Tenth Anniversary of the Hab-
itats Directive: there is much to celebrate! Available at http://www.natura2000benefits.org/
ireland/carta.htm.
3. Natura Barometer. Brussels: European Commission. Available at http://www.europa.eu
.int/comm/environment/nature/barometer/barometer.htm.
4. Resolution No. 1 (1989) of the Standing Committee on the provisions relating to the
conservation of habitats; Recommendation No. 15 (1989) of the Standing Committee on
the conservation of endangered natural habitat types; Recommendation No. 25 (1991) of
the Standing Committee on the conservation of natural areas outside protected areas proper;
all available at http://www.coe.int.
5. Regulation 1655/2000 Concerning the Financial Instrument for the Environment, Official
Journal L192, 28.7.2000.
6. See http://www.iol.ie/~burrenag/hist32000.html for a history of the case.
7. See http://www-penelope.drec.unilim.fr/penelope/cases/greek/leaf1.htm.
8. Quote translated from letter in French with attached guidelines from Commissioners
Monika Wulf-Mathies and Ritt Bjerregaard to the Permanent Representative of Ireland.
Brussels, June 23, 1999.
9. See note 2.
References
Baldock, David, Janet Dwyer, and José M. Sumpsi Vinas. 2002. Environmental Integration
and the CAP. Report to DG Agriculture, European Commission. London: Institute for Euro-
pean Environmental Policy.
Baldock, David, and Tony Long. 1988. The Mediterranean under Pressure: The Influence of
the CAP on Spain and Portugal and the IMPs in France, Greece and Italy. Gland, Switzer-
land: WWF International.
Ballesteros, Marta. 1996. Implementation of Recommendation No. 16 on Areas of Special
Conservation Interest: Part I Legal Basis for a European Ecological Network. Doc. T-PVS
(96) 5. Strasbourg.
BirdLife International. 1995. The Structural Funds and Biodiversity Conservation, PR27/L37.
Cambridge: BirdLife International.
Coffey, Clare. 1998. European Community Funding and the Sustainability Transition. In
Timothy O’Riordan and Heather Voisey, eds., The Transition to Sustainability: The Politics
of Agenda 21 in Europe, 130–150. London: Earthscan.
Coffey, Clare, and Saskia Richartz. 2002. The EU Habitats Directive: Generating Strong
Responses. Unpublished manuscript.
Coffey, Clare, and Karen Shaw. 2001. The Habitats Directive 92/43: Inventory of Interac-
tions with Other Institutions. London: Institute for European Environmental Policy.
Council of Europe. 1996. Report on the 16th Standing Committee Meeting. Strasbourg,
December 2–6, 1996. Doc. T-PVS (96) 102.
Council of Europe. 1999a. Group of Experts for the Setting up of the Emerald Network of
Areas of Special Conservation Interest. Building up the Emerald Network: A Guide for
Emerald Network Country Team Leaders. Strasbourg, September 18–19, 2000. Doc. T-
PVS/Emerald (99) 1 rev.
Council of Europe. 1999b. Opening and Closing of Files—and Follow-up to Recommenda-
tions. Doc. T-PVS (99) 16, Strasbourg.
Council of Europe. 2000. Memorandum of Co-operation between the Council of Europe and
the European Environment Agency. Doc. T-PVS (2000) 14. Copenhagen, July 3, 2000.
Council of Europe. 2001. Meeting of the Group of Experts for the Setting up of the Em-
erald Network. The Emerald Network. Istanbul, October 4–6, 2001. Doc. T-PVS (2001) 51,
Strasbourg.
Defra. 2001. First Report by the UK under Article 17 on the Implementation of the EC Hab-
itats Directive. London: Department of the Environment, Food and Rural Affairs.
European Commission. 1977. Second Environmental Action Programme. Official Journal
C139, 13.6.77.
European Commission. 1979. Proposal for a Council Decision Concerning the Conclusion of
the Convention on the Conservation of European Wildlife and Natural Habitats. Official
Journal C210, 22.8.1979.
European Commission. 1982. Council Decision 82/72/EEC Concerning the Conclusion of
the Convention on the Conservation of European Wildlife and Natural Habitats. Official
Journal L38, 10.2.1982.
European Commission. 1988. Proposal for a Directive on the Protection of Natural and
Semi-natural Habitats and of Wild Fauna and Flora. Doc. COM(88)381, Official Journal
C247, 21.9.1988.
European Commission. 1995. Cohesion Policy and the Environment. Doc. COM(95)509
final, 22.11.1995. Brussels: European Commission.
European Commission. 2002. Article 8 Working Group. Final Report. Brussels: European
Commission.
European Commission. 2004a. A New Partnership for Cohesion: Convergence, Competitive-
ness, Cooperation: Third Report on Economic and Social Cohesion. Luxembourg: European
Communities.
258 Clare Coffey
European Commission. 2004b. Report on Implementation of the Birds Directive: Draft Re-
port. Brussels: European Commission.
European Commission. n.d. Composite Report from the Commission on the Implementation
of the Directive 92/43/EEC on the Conservation of Natural Habitats and of Wild Fauna and
Flora. Brussels: European Commission.
Fairbrass, Jenny, and Andrew Jordan. 2001. Protecting Biodiversity in the European Union:
National Barriers and European Opportunities? Journal of European Public Policy 8 (4):
499–518.
Fischler, Franz. 2000. Letter regarding Commission Guidelines on the Relationship between
Rural Development Programming and Respect for the Community Environmental Policy.
Brussels, May 25, 1999.
IEEP. 1999. Nature Conservation: Commission Threatens Action over Habitats Directive.
Briefing note. London: Institute for European Environmental Policy.
Jen, Sandra. 1999. The Convention on the Conservation of European Wildlife and Natural
Habitats (Bern, 1979): Procedures of Application in Practice. Journal of International Wildlife
Law and Policy 2 (2): 239–258.
Jen, Sandra. 2002. WWF European Policy Office. WWF Contribution to Minutes of
Natura 2000 Conferences. Brussels: World Wide Fund. Available at http://greens-efa.org/
pdf/documents/Natura2000/WWF_EN.pdf.
Lenschow, Andrea. 2002. Dynamics in a Multilevel Polity: Greening the European Union Re-
gional and Cohesion Funds. In Andrea Lenschow, ed., Environmental Policy Integration:
Greening Sectoral Policies in Europe, 193–215. London: Earthscan.
Long, Tony. 1995. Shaping Public Policy in the European Union: A Case Study of the Struc-
tural Funds. Journal of European Public Policy 2 (4): 672–679.
Paavola, Juoni. 2002. Protected Areas Governance and Justice: Theory and the European
Union’s Habitats Directive. Environmental Sciences 1 (1): 59–77.
Ten Brink, Patrick, Claire Monkhouse, and Saskia Richartz. 2002. Promoting the Socio-
Economic Benefits of Natura 2000. London: Institute for European Environmental Policy.
Verschuuren, Jonathan M. 2002. Implementation of the Convention on Biodiversity in Eu-
rope: 10 Years of Experience with the Habitats Directive (Comments). Journal of Interna-
tional Wildlife Law and Policy 5 (3): 235–250.
Wilkinson, David. 1994. Using the European Union’s Structural Funds and Cohesion Funds
for the Protection of the Environment. Review of European Community and International
Environmental Law 3 (2/3): 119–126.
Wilkinson, David. 1998. Steps Towards Integrating the Environment into Other EU Policy
Sectors. In Timothy O’Riordan and Heather Voisey, eds., The Transition to Sustainability:
The Politics of Agenda 21 in Europe, 113–129. London: Earthscan.
11
The EU Deliberate Release Directive: Environmental
Precaution versus Trade and Product Regulation
Ingmar von Homeyer
The first generation of genetically modified (GM) agricultural products, such as GM

seeds, crops, and food, has been commercially available since the mid-1990s. GM
maize and soy are the most common products, followed by oilseed rape. A particu-
larly common characteristic of these products is genetically engineered pesticide re-
sistance, which is associated with several possible environmental risks. For example,
resistance may be transferred to related wild species through crossbreeding or gene
transfer and GM crops may have undesirable side effects on nontarget organisms,
such as beneficial insects. In addition to these primary effects, there are secondary
effects caused by the complementary pesticide, rather than the crops themselves.
For instance, pesticide use may cause resistance in target weeds or have a negative
impact on biodiversity (Sauter and Meyer 2000, 64–75, 221–226).
In 1990 the EU adopted the Directive on the Deliberate Release of Genetically
Modified Organisms into the Environment (Deliberate Release Directive, DRD) to
address the potential environmental risks of GM products. The DRD is one of the
most controversial pieces of EU environmental legislation. Its alleged negative effects
on the commercialization of, and trade in, genetically modified organisms (GMOs)
have been heavily criticized by the United States and other main producers as well as
parts of industry and some scientists working in the field. The Directive interacts
with several environmental and nonenvironmental institutions, in particular in the
agricultural and food sectors and concerning trade. Politically significant interac-
tions are with EU product-sector legislation, such as rules governing the author-
ization for Communitywide sale of food, feed, seeds, and pesticides, and various
international institutions, in particular the Sanitary and Phytosanitary (SPS) Agree-
ment and the Technical Barriers to Trade (TBT) Agreement of the World Trade Or-
ganization (WTO).
260 Ingmar von Homeyer
This chapter starts with a short overview of the DRD. The next section summa-
rizes the Directive’s main interactions. This is followed by more detailed analyses of
the interactions with the EU Pesticides Directive (PD) and the WTO SPS Agreement.
Both interactions focus on the DRD’s precautionary approach. The final section
draws some conclusions from the main findings.
The Deliberate Release Directive
The DRD (Directive 90/220/EEC) pursues the twofold aim of European harmoniza-
tion and protecting human health and the environment (Art. 1). It covers deliberate
releases of GMOs for research purposes (Part B) and marketing of GM products
(Part C). Releases for research purposes are authorized at the national level in ac-
cordance with the Directive. Involvement of EU-level institutions is significantly
stronger in product approval than with respect to releases for research purposes.
An application is submitted to a National Competent Authority (NCA), but author-
ization depends on other EU member states that can raise objections in the so-called
Article 21 Committee, which oversees the implementation of the DRD. This EU reg-
ulatory committee decides on the basis of a Commission proposal, the adoption of
which requires a qualified majority of votes held by the member states. Otherwise
the matter is referred to the EU Council of (Environment) Ministers. If the Council
does not decide by a qualified majority within three months, the Commission may
adopt its original proposal. However, Article 16 of the DRD allows an EU member
state to provisionally suspend the authorization of a GM product if there are ‘‘justi-
fiable reasons to consider that a product . . . constitutes a risk.’’
The DRD is based on a precautionary regulatory approach, which assumes that—
given potentially serious environmental or health risks—protective measures may be
justified even if there is no clear scientific evidence of harm. Precaution constitutes
the DRD’s raison d’être and justifies the Directive’s technology-specific character.
More specifically, the DRD uses characteristics of the production process—that is,
use of genetic engineering—as ‘‘regulatory trigger.’’ This approach reflects the as-
sumption that GMOs pose potential risks requiring specific regulation because of
the way they are produced. The DRD also uses a precautionary approach at the
operational level because risk assessment focuses on the evaluation of potential
environmental risks.
In addition to the default authorization procedure involving the Article 21
Committee, the DRD provides for an alternative procedure based on conventional
EU Deliberate Release Directive 261
product-sector legislation. Under sectoral legislation, regulatory oversight is trig-

gered by the kind of product as defined by its intended use (product groups), rather
than by technology. Reflecting the DRD’s precautionary approach, Article 10(2) of
the DRD allows for exclusive authorization under applicable sectoral legislation
only if such legislation ‘‘provides for a specific environmental risk assessment similar
to that laid down in this Directive’’ (emphasis added).
The EU adopted a heavily revised DRD (Directive 2001/18/EC) in March 2001
that is significantly stricter than its predecessor. Among other things, the revised
DRD explicitly incorporates the precautionary principle as a rationale for its adop-
tion (Recital 8 and Art. 1) and as a principle of risk assessment (Art. 4 and Annex
II). Furthermore, the revised Directive clarifies that precautionary risk assessment
covers indirect and long-term effects. It also introduces new requirements for clear
labeling and traceability of GMOs (Art. 12 and 14) and makes authorization under
sectoral legislation contingent on the application of ‘‘equivalent’’ (rather than ‘‘sim-
ilar’’) measures, including risk assessment (Art. 12(3)).
Overview of Interactions
Because of the wide range of possible applications of GMOs, there is a large poten-
tial for interactions involving the DRD. For example, GMOs are an issue for the
Food and Agriculture Organization of the UN (FAO) because they may affect food
security (Becker and Hanrahan 2004, 3). GMOs may also aid environmental pro-
tection by dissolving toxic chemicals or mitigating climate change by increasing
the capacity of plants to absorb CO2 . This may affect international institutions for
the protection of water resources and the climate change regime. More worrying are
applications in biological warfare—a possibility that could lead to interactions with
security and defense regimes. However, for the time being, many of these interac-
tions remain hypothetical, not least because only relatively few nonmedicinal GM
products have so far been developed.
The DRD’s most intensive and politically important interactions with other EU
legal instruments and international institutions are summarized in table 11.1.
These interactions focus on three main issues: the DRD’s precautionary regulatory
approach, and its labeling and traceability provisions. Four clusters of horizontal
interactions with other EU legislative instruments can be distinguished. First of all,
the DRD’s precautionary risk assessment influenced EU sectoral legislation in two
ways. First, the DRD added a second regulatory hurdle for GM products that must
Table 11.1
Main interactions of the Deliberate Release Directive
EU sectoral legislation (risk DRD established second regulatory hurdle requiring
assessment) precautionary risk assessment of GMOs, thus
disrupting authorization under sectoral legislation.
DRD requires sectoral legislation to provide for
‘‘similar’’ (original DRD) or ‘‘equivalent’’ (revised

DRD) environmental risk assessment.
EU Novel Food Regulation DRD required labeling only for certain GMOs and
thus undermined the Novel Food Regulation that aimed
to introduce more general labeling of GM food.
EU sectoral legislation (labeling) Amended DRD allowed for labeling irrespective of
actual GMO content, thereby undermining sectoral
legislation that aims to introduce unambiguous labeling
of GMOs according to certain thresholds of GM
content.
EU Pesticides Directive Unclear scope of DRD’s precautionary risk assessment
led to EU decision to evaluate secondary effects under
PD, rather than DRD.
As an increasing number of member states pursued
evaluation of secondary effects under DRD,

authorization of products conforming to PD was
negatively affected.
EU Environmental Liability Liability of GMOs was not included in revised DRD
Directive because it was to be dealt with in upcoming Liability
Directive.
Liability Directive refers to DRD and partially covers
GMOs.
Liability Directive is expected to have a positive effect
on the implementation of the DRD.

WTO SPS Agreement DRD’s precautionary approach leads to delays in GM
product approval or even bans, thereby restricting
international trade.
Partial ‘‘internationalization’’ of DRD’s precautionary
approach may lead to modification of SPS Agreement’s

concept of precaution.
WTO dispute-settlement mechanism provides SPS
Agreement with potentially effective instrument to

impose rudimentary precaution on DRD.
WTO TBT Agreement DRD’s labeling and traceability provisions restrict
trade, thereby undermining objective of TBT
Agreement.
WTO dispute-settlement mechanism provides TBT
Agreement with potentially effective instrument to

enforce compliance, possibly undermining the DRD.
UNEP Technical Guidelines on DRD constituted input for formulation of UNEP
Biosafety Guidelines.
be approved under both sectoral legislation and the DRD (Barling 1997, 1044). The
additional hurdle increases regulatory uncertainty, because it may cause delays and
result in additional authorization requirements or outright product bans. These be-
havioral effects disrupt the traditional approval process under sectoral legislation
(Behavioral Interaction). Second, by allowing for exclusive approval of GM prod-
ucts under sectoral legislation only if such legislation incorporates a ‘‘similar’’ (orig-
inal DRD) or even ‘‘equivalent’’ (revised DRD) risk assessment, the DRD committed
EU legislators to a precautionary risk assessment and thus put pressure on them to
amend sectoral legislation accordingly (Interaction through Commitment).
The second cluster of horizontal interactions concerns the EU Pesticides Directive
(PD). It is analyzed in more detail in the next section. Reacting to differing interpre-
tations, the Article 21 Committee decided in 1994 that the evaluation of secondary
effects of GM products was to be considered under the Pesticides Directive, rather
than the DRD. This interpretation, which was not supported by all member states,
was driven by the desire to avoid conflicting interpretations of their commitments by
individual member states (Interaction through Commitment). This attempt to settle
the dispute was, however, unsuccessful because a growing number of member states
defied the Article 21 Committee’s decision and requested evaluation of secondary
effects under the DRD. They employed Article 16 of the DRD to suspend authoriza-
tions and blocked authorizations in the Article 21 Committee. This development
contributed to delays in product approval and outright bans and thus undermined
the PD at the outcome level (Behavioral Interaction).
Third, the DRD interacted with EU rules on labeling. From 1997 on, the Novel
Food Regulation required labeling of food consisting of GMOs. However, the
DRD only required labeling of GM products associated with specific health or envi-
ronmental risks. Therefore, some GM crops that may be used as food ingredients
and had already been authorized under the DRD, such as certain GM maize and
soy varieties, were not subject to labeling requirements, which undermined the label-
ing provisions of the Novel Food Regulation (Behavioral Interaction). Reacting to
the Novel Food Regulation, the EU amended the DRD to require more general la-
beling of GMOs. However, for products sold in mixtures that may or may not con-
tain GMOs, the amendment allowed a ‘‘may contain GMOs’’ label. This had the
effect of undermining the labeling provisions of several pieces of subsequent sectoral
legislation establishing thresholds for GMO content beyond which labeling was
required, because the ‘‘may contain GMOs’’ label occurred irrespective of actual
GMO content (Behavioral Interaction).
Finally, the DRD also interacted with the EU Environmental Liability Directive
(2004/35/EC). Initially, the commitment to elaborate a Liability Directive was suc-
cessfully employed as a major argument to prevent the inclusion of liability provi-
sions regarding GMOs in the revised DRD (Interaction through Commitment). In
its final form, the Liability Directive indeed refers to the DRD and covers GMOs,
although to a lesser extent than many supporters of including provisions on liability
in the revised DRD may have wished (Interaction through Commitment). Neverthe-
less, the Liability Directive is likely to assist in the effective implementation of the
DRD (Behavioral Interaction).
The DRD’s vertical interactions with international institutions frequently involve
the WTO SPS Agreement. First, the DRD’s precautionary risk assessment led to
delays, product bans, and moratoriums, which restricted U.S. exports of GM maize
and soy to the EU. These import restrictions undermined the SPS Agreement’s objec-
tive of trade liberalization (Behavioral Interaction). Second, the SPS Agreement is
backed up by WTO dispute settlement, which may reject the DRD’s precautionary
approach and authorize trade sanctions against the EU. Such trade sanctions would
undermine the implementation of the DRD (Behavioral Interaction). Third, the SPS
Agreement recognizes certain internationally agreed-on safety standards. Reflecting
its commitment to the DRD, the EU was partly successful in promoting the DRD’s
concept of precaution in several international institutions (Interaction through Com-
mitment). This may eventually lead to a reinterpretation or formal modification of
the SPS Agreement’s concept of precaution. The interactions between the DRD and
the SPS Agreement are examined in more detail below.
Furthermore, the DRD’s provisions for extensive labeling and traceability of
GMOs tend to restrict trade, in particular because compliance requires potentially
costly separation of GM products from conventional products (Becker and Hanra-
han 2004, 4–5). Therefore, these provisions undermine the WTO TBT Agreement’s
main objective of increasing trade flows (Wolff 2003, 7–9) (Behavioral Interaction).
Like the SPS Agreement, the TBT Agreement is backed up by the WTO dispute-
settlement mechanism. If the latter ruled against the DRD’s labeling and traceablity
provisions in a possible future case, the resulting trade sanctions would undermine
the implementation of the DRD (Behavioral Interaction).
Finally, the DRD also influenced the Technical Guidelines for Safety in Bio-
technology adopted by the United Nations Environment Program (UNEP) in 1995,
which are partly modeled on the DRD. The United Kingdom and the Netherlands
played key roles in preparing draft guidelines, which were subsequently discussed
by the Article 21 Committee. At that time, the DRD constituted the most compre-
hensive and advanced set of rules specifically designed to ensure the safe use and
commercialization of GMOs and thus served as an important informational input
in the preparation of the UNEP Guidelines (UNEP 1997, sec. III; Cantley 1995,
633–634) (Cognitive Interaction).
Interactions with the Pesticides Directive: Precaution Prevails
The interaction with the Pesticides Directive resulted from conflicting interpretations
of the scope of the DRD’s precautionary risk assessment. Against some opposing EU
member states, the Commission, supported by a majority of member states, decided
that secondary effects should be evaluated under the PD. However, once GM crops
had been authorized on this basis, several EU member states invoked Article 16 of the
DRD and enacted national bans citing, among other things, possible harmful sec-
ondary effects of the complementary pesticides. This undermined the PD, according
to which the bans were not justified. As a result of these conflicts, the revised DRD
explicitly provides for a precautionary risk assessment covering secondary effects.
The Pesticides Directive

The Pesticides Directive (Directive 91/414/EEC Concerning the Placing of Plant Pro-
tection Products on the Market) harmonizes authorization criteria and basic ap-
proval procedures for pesticides in the EU. First, the Directive establishes approval
criteria for national authorities. Second, it bans pesticides containing substances not
listed in its Annex I, while also laying down a procedure for adding substances to that
Annex. Third, it establishes the principle of ‘‘mutual recognition’’ according to which
a product that has been authorized in a member state in accordance with the Direc-
tive can be sold throughout the Union. National authorities retain responsibility for
product authorizations, but the procedure for adding substances to Annex I involves
the Standing Committee on Plant Health, in which member states and the Commis-
sion are represented. Decision making is similar to the Article 21 Committee, includ-
ing qualified majority voting and the possibility of referring decisions to the Council
(here: the Agriculture Council). If the Council fails to act, the Commission may
adopt its proposal. Scientific advisory committees, in particular the Scientific Com-
mittee for Pesticides (later: Scientific Committee on Plants), assist the Commission.
In contrast to the DRD, the Pesticides Directive is not based on the precautionary
principle. Its authorization requirements focus on clearly identifiable, actual effects:
pesticides may not have any ‘‘harmful effect on plants or plant products, directly or
indirectly (e.g. through drinking water, food or feed), or on groundwater and they
may not have any unacceptable influence on the environment, having particular re-
gard to the following considerations: its fate and distribution in the environment,
particularly contamination of water including drinking water and groundwater; its
impact on non-target species’’ (Art. 4(b(iv, v)), Pesticides Directive). These criteria
are less demanding than those of the DRD. In particular, the DRD uses a broader
definition of harmful effects, which includes potential effects (Sauter and Meyer
2000, 119). Article 1(3) of the PD acknowledges these differences, calling on the
Commission to submit a proposal for an amendment of the Pesticides Directive.
The amendment was to provide for a specific environmental risk assessment so as
to allow application of Article 10(2) of the DRD on product approval under sectoral
legislation. However, such an amendment was not adopted.
Deliberate Release Directive Undermines Pesticides Directive

When the first marketing applications for pesticide-resistant GM crops were dis-
cussed in the mid-1990s, secondary effects played a major role. These effects are
not caused by the GM crops themselves, but by the complementary pesticide (often
a common pesticide that can be used on additional kinds of crops if these crops have
been genetically modified to make them resistant to the pesticide). The effects in-
clude developing resistance in target weeds or a negative impact on biodiversity. If
certain GM crops are not authorized under the DRD due to the potential secondary
effects of the complementary pesticide, the DRD restricts the applicability of that
pesticide. Consequently, the DRD’s precautionary approach restricts the market for
that pesticide and undermines the conventional environmental evaluation criteria
for pesticides as set out by the PD.
The DRD neither defines the scope of ‘‘adverse effects on human health and the
environment’’ (Art. 4) nor provides sufficiently detailed procedures and criteria for
risk assessment (Schomberg 1998, 5). Against this background, a minority of mem-
ber states argued that the precautionary environmental risk assessment required
under the DRD covered secondary effects originating from the complementary
pesticide. The diverging interpretations primarily reflected differing national agricul-
tural policies and approaches to risk assessment. In 1994, the Commission attempted
to clarify the situation. It proposed that secondary effects were not covered by the
DRD but by sectoral legislation, specifically the Pesticides Directive (Commission
Decision 94/385/EC). However, some member states dissented, not least because
Figure 11.1
Deliberate Release Directive undermines Pesticides Directive
the PD did not provide for a precautionary environmental risk assessment. Despite
initial support by most member states for the Commission’s restrictive interpretation
of the DRD’s scope (Torgersen and Seifert 2000, 210–211), a growing number of
member states subsequently also included, or called for, an evaluation of secondary
effects under the DRD.
Figure 11.1 illustrates how the resulting conflict undermined the implementation
of the Pesticides Directive. First, the DRD’s vague provisions on risk assessment
invited conflicting interpretations, in particular concerning secondary effects.1 Sec-
ond, faced with in their view insufficient risk assessments, some member states
enacted national bans invoking the DRD’s Article 16 emergency clause. They also
used the Article 21 Committee procedure—which requires a qualified majority for
product approval—to grind the authorization process to a complete halt. Third, in
doing so, member states ignored assessments of secondary effects under the PD,
which had concluded that such effects were tolerable. Since pesticide-resistant GM
crops did not become available, the use of the complementary pesticides was also
restricted—for reasons that were invalid according to the criteria of the Pesticides
Directive. Thus, the effectiveness of this Directive was undermined.
The controversial approval of Ciba-Geigy’s Bt-176 maize had a particularly
strong impact on the escalation of conflict. This GM crop variety was engineered
to be resistant to certain herbicides and to produce a naturally occurring toxin

(‘‘Bt’’) with insecticidal effects. In 1995 Austria, Denmark, Sweden, and the United
Kingdom objected to the marketing application for the Bt-176 maize in the Article
21 Committee. Except for the United Kingdom, these countries argued that, among
other things, no assessment of secondary effects originating from the complementary
pesticide and the self-produced toxin had been made (Sauter and Meyer 2000, 228–
231; Toft 1999, 14). The Commission rejected this view, reiterating its position that
these issues were a matter for the Pesticides Directive (Levidow, Carr, and Wield
1999a, 51). After the Council had failed to achieve the qualified majority necessary
for approval, the Commission first referred the issue to its scientific advisory com-
mittees and then approved the Bt-176 maize in December 1996. This decision was
supported by the Scientific Committee on Pesticides, which had argued that second-
ary effects—for example, Bt-resistance in destructive insects—constituted an agro-
nomic rather than an environmental problem. As with conventional crops, the
Committee considered that such problems could be managed using alternative pesti-
cides. Additional effects that might be caused by a switch to alternative pesticides,
decreasing effectiveness of Bt as an ‘‘environmentally friendly’’ pesticide, and the
possibility of an overall increase in pesticide use were not addressed (Sauter and
Meyer 2000, 229–230, 233). Immediately following the Commission’s approval of
the Bt-176 maize, Austria, Luxembourg, and Italy invoked Article 16 of the DRD,
banning cultivation on their territories (Friends of the Earth 1997, 6).
The controversial approval of herbicide-resistant GM oilseed rape (MS1xRF1/
RF2) developed by the company Plant Genetics Systems (later: AgrEvo) illustrates
how support among EU member states for the prevailing interpretation of the divi-
sion of competencies between the DRD and the PD waned in the second half of the
1990s. France and the United Kingdom, in particular, reassessed their positions.
The company had applied for a broad authorization of its oilseed rape in France.
The Commission ignored calls by several member states for an evaluation of second-
ary effects under the DRD (Levidow, Carr, and Wield 1999a, 29; Schomberg 1998,
17; Fromwald and Strauss 1998, 15) and proposed approval. As with the Bt-176
maize, secondary effects were treated as an agronomic rather than environmental
problem (Levidow and Carr 2000, 263; Levidow, Carr, and Wield 1999a, 28;
1999b, 14, 18; Roy and Joly 1999, 19). However, following adoption of the Com-
mission’s proposal, France withdrew its support and imposed a general moratorium
on the commercialization of pesticide-resistant oilseed rape. This U-turn prevented
the final approval of the oilseed rape, which must be granted by the country where
the original application was submitted (in this case: France). More generally, it
implied that the French government had reassessed its position and now considered
secondary effects relevant for risk assessment under the DRD (Roy and Joly 1999,
19–20). France subsequently supported demands to explicitly require evaluation of
secondary effects under the DRD (Roy and Joly 2000, 253). British authorities had
often been the most outspoken and influential opponents of such demands. How-
ever, in 1998 the United Kingdom decided to require risk assessment to cover sec-
ondary effects (Levidow, Carr, and Wield 1999b, 14). Additional monitoring and
testing requirements eventually culminated in a new British approach combining
‘‘managed development’’ of GM crops with a three-year moratorium on commercial
cultivation (Levidow and Carr 2000, 264). Similar measures were subsequently
adopted in other member states, such as Germany (Sauter and Meyer 2000, 98).
These bans, moratoriums, and modifications of national risk-assessment practices
meant that an increasing number of EU member states ignored the evaluation crite-
ria for assessment of secondary effects under the PD.
Response Action to Solve the Conflict: Revision of the DRD

Whereas the early bans and moratoriums reflected the positions of individual mem-
ber states, a more coordinated approach emerged in 1997 when the Commission
proposed to lift national bans of the Bt-176 maize. When the proposal failed to
achieve a qualified majority in the Article 21 Committee and, subsequently, the
Council in 1998, the Commission refrained from taking legal action against the
bans for fear of a political backlash (Winkler 1999, 10). More importantly, the con-
flicts over the scope of the DRD’s risk assessment contributed significantly to the
adoption of the de facto moratorium on the approval of GM crops announced by
the Environment Council in June 1999 (Sauter and Meyer 2000, 91–93). The mor-
atorium halted new approvals for more than five years. Also in 1999, the Environ-
ment Council adopted a legislative Common Position on the revision of the DRD.
The Common Position provided the main substantive basis for the adoption of the
revised DRD in 2001.
The revised DRD clarifies that risk assessment covers secondary effects. According
to its Principles for Environmental Risk Assessment (Annex II, Part A), the risk as-
sessment aims to identify and evaluate ‘‘potential adverse effects of the GMO, either
direct or indirect, immediate or delayed, on human health and the environment.’’
Such adverse effects may, among other things, occur through ‘‘changes in manage-
ment, including, where applicable, in agricultural practices’’ (Annex II, Part C.2).
The introduction of market-stage precautionary measures, such as monitoring plans

to identify previously unknown or delayed adverse effects (Art. 20) and time-limited
authorizations (Art. 17), supports the new provisions on secondary effects.
The adoption of the provisions on secondary effects is remarkable, because the
initiative to revise the DRD originally responded to various complaints about exag-
gerated regulation of GMOs (Homeyer 2002). While this may to some extent be
characterized as the result of a learning process (Levidow, Carr, and Wield 2000,
204–206), two other factors help explain why the DRD’s risk assessment was
extended rather than reduced. First, as a result of the blocking of further commer-
cialization of GMOs in the EU (de facto moratorium) emanating from the insufficient
provisions of the original DRD, interested member states were able to put consider-
able pressure on industry, other member states, and the Commission to reverse the
thrust of the revision process and to accept their demands, including the explicit re-
quirement of an evaluation of secondary effects under the DRD. Second, and more
importantly, in 1996 the political climate for the revision of the DRD changed
markedly when the BSE (‘‘mad-cow’’ disease) crisis coincided with the first EU
imports of GM soy from the United States (Vogel 2001). This event reinvigorated
opposition by environmental nongovernmental organizations (NGOs), such as Green-
peace and Friends of the Earth, and others to GM crops and food. In the following
years the unfolding new political context hardened the position of member states
that had already evaluated secondary effects under the DRD and won over others,
such as the United Kingdom and France. Overall, it seems difficult to imagine that a
sufficient number of member states would have supported the de facto moratorium
in the absence of the heightened public opposition and protests (Homeyer 2003).
Given that risk assessment under the revised DRD clearly covers secondary
effects, the new provisions appear to resolve the disruptive interaction with the PD.
However, it is unlikely to resolve all issues over the evaluation of secondary effects.
For example, the revised DRD’s provisions on market-stage precautions (monitoring
and so on) may create new conflicts, because it is not clear which factors, including
certain secondary effects, need to be assessed before a marketing permission is given,
and which ones may be assessed later, in the framework of market-stage precau-
tions. To avoid conflicts over this question, it will be necessary to develop common
criteria and procedures (Sauter and Meyer 2000, 142–143). Furthermore, there is
no consensus among member states on the normative yardstick necessary to define
‘‘adverse effects’’ (Art. 4(1) of the revised DRD), including secondary effects. For ex-
ample, the reduction of pesticide use is an important agricultural-policy objective in
some member states, such as Austria (Torgersen and Seifert 2000, 213–214), but
less so in others. Because of this difference, evaluations of relevant secondary effects
are likely to continue to differ. A resolution of these conflicts would require agree-
ment on principles of good agricultural practice (Sauter and Meyer 2000, 114–
115). Ongoing discussions concerning the reform of the EU’s Common Agricultural
Policy (CAP), the integration of environmental concerns into the CAP, and sustain-
able development may offer opportunities to address relevant issues.
Interaction with the WTO SPS Agreement: Trade versus Precaution?
The Deliberate Release Directive subjects all GM products—including imports—

to a strict approval procedure. Products approved outside the EU may therefore
be subject to delayed EU authorization, additional requirements, or outright bans.
Reflecting an increasingly strict interpretation of the DRD’s precautionary risk as-
sessment by several EU member states, the 1999 de facto moratorium blocked
authorization of GM products that had so far not received EU clearance. As a result,
the DRD undermined the free-trade objectives of the World Trade Organization
(WTO), in particular of the WTO SPS Agreement. The Agreement seeks to restrict
the scope for national safety measures to increase trade in agricultural commodities
and food. However, opinions are divided on whether the precautionary approach of
the DRD constitutes a formal breach of WTO rules. Ultimately, this question will be
decided within ongoing and potential future WTO dispute-settlement cases, which
in turn are influenced by various other international forums and agreements.
The WTO Agreement on Sanitary and Phytosanitary Measures

The General Agreement on Tariffs and Trade (GATT) was established in 1947. In
1994 the Uruguay Round of GATT negotiations resulted in the creation of the
WTO and the adoption of the SPS Agreement. The main function of the WTO is to
ensure that trade flows as smoothly, predictably, and freely as possible. As part of
the WTO rules, the SPS Agreement aims to reduce trade barriers created by different
national standards to protect human, plant, and animal life and health. Therefore, it
introduces restrictions on these standards. In particular, safety measures should not
be more trade-restrictive than necessary and their application should not result in
arbitrarily different treatment. Measures should not create disguised trade barriers
and, with limited exceptions for interim measures, must be based on scientific prin-
ciples (Art. 2 and 5 of the SPS Agreement).
Rather than emphasizing the precautionary principle, the SPS Agreement not only
tends to treat science as a sufficient but—at least in practice—also as a necessary cri-
terion for establishing the legitimacy of relevant trade restrictions (Peel 2004; Wolff
2003, 1, 9–10; Scott 2000, 148–149). According to its Article 2(2), WTO members
have the obligation to ‘‘ensure that any sanitary or phytosanitary measure . . . is
based on scientific principles and is not maintained without sufficient scientific
evidence.’’ The rudimentary version of the precautionary principle contained in Ar-
ticle 5(7) provides the only exception. This article states that in ‘‘cases where rele-
vant scientific evidence is insufficient, a Member may provisionally adopt sanitary
or phytosanitary measures on the basis of available pertinent information. . . . In
such circumstances, Members shall seek to obtain the additional information neces-
sary for a more objective assessment of risk and review the sanitary or phytosani-
tary measure accordingly within a reasonable period of time [emphases added].’’
The underlying assumption seems to be that it will be relatively easy to provide con-
clusive scientific evidence to determine the legitimacy of measures in any given case.
The emphasis on science also influences the SPS Agreement’s provisions on risk as-
sessment: according to Article 5(1 and 2) of the Agreement, safety standards must
be based on an appropriate risk assessment that ‘‘shall take into account available
scientific evidence.’’
The SPS Agreement automatically recognizes national measures that conform to
international standards. In this context, the Agreement explicitly refers to three
international standard-setting bodies: the Codex Alimentarius Commission (CAC;
food safety), the International Plant Protection Convention (IPPC; plant health),
and the Office International des Epizooties (OIE; animal health). Essentially, except
for precautionary interim measures, countries may consequently not introduce or
maintain measures that are stricter than internationally agreed-on standards without
scientific justification (Art. 3).
Like other WTO rules, the SPS Agreement is supported by comparatively strong
enforcement mechanisms. The SPS Committee comprising WTO member countries
oversees implementation of the SPS Agreement. The Committee reviews notifica-
tions of relevant national measures, keeps contact with other international bodies,
and makes proposals for amendments (Art. 12). If member countries disagree on
the conformity of national measures with the SPS Agreement, they may take re-
course to the WTO dispute-settlement mechanism, which can authorize the dam-
aged party to use trade sanctions against an infringing country.
Deliberate Release Directive Undermines SPS Agreement

The interaction between the DRD’s provisions on precautionary risk assessment and
the WTO SPS Agreement occurred against the background of a fast rise in the
global production of GM agricultural products in the second half of the 1990s
and the simultaneous emergence of a highly uneven international pattern of cultiva-
tion of GM crops. Starting from a very modest 1.7 million hectares in 1996, six
years later GM crops were already grown on a total of 58.7 million hectares. In
2002, the United States—the leading country in the development of agricultural
biotechnology—accounted for 66 percent of total agricultural land planted with
GM crops worldwide. At 23 percent Argentina was also a significant producer.
Canada (6 percent) and China (4 percent) accounted for most of the remaining pro-
duction (Brack, Falkner, and Goll 2003, 2).
The gap between high levels of cultivation of GM crops in the United States
and other countries, and low production in the EU corresponds to some extent to
differences in the approach and stringency of safety legislation. First, EU and U.S.
laws are built on different assumptions: reflecting the precautionary principle, the
EU regulatory approach addresses (bio-)technology as such. In contrast, product
characteristics—for example, whether a GM crop is associated with a plant-pest
risk—trigger regulation in the United States. As a result, the EU had to elaborate
new, specific legislation for GMOs (i.e., the DRD), while U.S. regulatory oversight
could largely rely on preexisting laws. Second, following from these differences in
approach, product authorization under the revised DRD is generally given for a lim-
ited period of time and coupled with monitoring requirements because it is built on
the assumption that experience and scientific knowledge about the potential risks of
GM products are lacking. While the U.S. approach does not exclude such precau-
tionary measures, they are applied on a much less general basis (Levidow 2001, 8).
Because of the relative stringency of its precautionary approach, the DRD restricts
trade, thereby undermining the SPS Agreement (figure 11.2). EU imports originat-
ing from countries with less stringent regulations—mainly the United States—face
lengthy authorization procedures or may not be approved at all. This compromises
the SPS Agreement’s main objective of trade liberalization. As early as 1998, the
U.S. Department of Agriculture (USDA) estimated that the EU’s prolonged approval
of U.S. GM maize led to a loss of around U.S.$200 million for U.S. exporters
(Kelch, Simone, and Madell 1998, 1). The ‘‘American share of European maize . . .
imports dropped from 86% in 1995 to 12% in 1999,’’ largely because of delays in
Figure 11.2
Deliberate Release Directive undermines WTO SPS Agreement
EU product approval (Vogel 2001, 14). Regulatory hurdles also contributed to a

sharp decline in U.S. soybean exports to the EU between 1995 and 1999. Similar
U.S. losses in exports of maize and soybeans to the EU occurred in subsequent years.
However, the losses caused by delays in the EU approval process are likely to be
smaller than stated by U.S. authorities. First, because of the widespread rejection of
GM food by European consumers and campaigns by environmental NGOs, Euro-
pean retailers and food producers have mostly stopped storing or manufacturing
GM products (Levidow and Murphy 2003, 56; USDA 2003, 6). The resulting U.S.
export losses cannot be attributed to EU regulations. Second, long delays in the EU
authorization process were often caused by the EU de facto moratorium. Although
the moratorium was related to the DRD in that it reflected a precautionary ap-
proach,2 it also conflicted with the DRD’s approval procedure, which does not pro-
vide for a general moratorium but requires a case-by-case evaluation of applications.
Outlook
The conflict between the DRD and the SPS Agreement may be resolved by collective
decision making. The EU could adapt the DRD to the SPS Agreement’s rudimentary
concept of precaution, for example in response to a related finding of the WTO
dispute-settlement mechanism. Alternatively, the SPS Agreement could be made
compatible with a wider understanding of precaution either by formally changing
or by reinterpreting it. Finally, both institutions could be adapted so as to meet at

some middle ground. Whether and to what extent these solutions will be employed
depends on at least three partly interrelated factors: the outcome of ongoing and
future WTO dispute-settlement proceedings, negotiations in other relevant inter-
national organizations, and negotiations within the EU and the WTO. To the extent
that the solutions resulted from developments in the EU or WTO, they would con-
stitute new cases of interaction between both.
WTO dispute-settlement proceedings could have an important impact by either
increasing the pressure on the EU to adapt the DRD or interpreting WTO rules so
as to make them more compatible with the DRD’s precautionary approach. A U.S.
complaint filed at the WTO in 2003 against the EU de facto moratorium and vari-
ous national bans provides a first test case because precaution was a key rationale
for the moratorium. The United States had previously threatened to file a WTO
complaint (Yerkey 1999), but had not done so for a number of reasons, including
the risk of losing a WTO case, unfavorable political circumstances such as the final
negotiations of the Cartagena Protocol on Biosafety (U.S. Mission to the EU 1999;
Kelch, Simone, and Madell 1998), and, subsequently, the Afghanistan and Iraq
wars, as well as concerns that too much overt pressure on the EU would further in-
crease consumer and political resistance.3 Eventually, signs of a growing ‘‘interna-
tionalization’’ (Gupta 1999, 28) of the DRD’s precautionary approach contributed
to the filing of the WTO complaint. In particular, Zambia, supported by other Afri-
can countries, had rejected U.S. GM food aid in 2003, citing precaution and con-
cerns regarding potential negative implications for agricultural exports to the EU. It
has even been argued that the WTO complaint was motivated more by a wish to
deter third countries from following the DRD’s precautionary approach than by
the actual losses of U.S. export opportunities to the EU (Murphy and Chataway
2003, 12).4 Previous rulings by the WTO Dispute Settlement Body, in particular in
the Hormones case, do not provide clear indications as to the outcome of the new
case, with some elements seemingly supporting the DRD’s precautionary approach
and others appearing to speak against it (Peel 2004; Brack, Falkner, and Goll 2003,
9; Murphy and Chataway 2003, 11).
The outcome of this and of potential future WTO dispute-settlement proceedings
will be influenced by third international institutions. As mentioned above, the SPS
Agreement recognizes trade-restrictive measures conforming to standards originat-
ing from some international institutions. Standards developed by two organizations
appear particularly relevant with respect to the DRD. The International Plant Pro-
tection Convention (IPPC) focuses on plant pests, including environmental effects
on biodiversity. Its overall significance for the interaction might be relatively limited
because most trade in GMOs concerns GM feed and food. The Codex Alimentarius
Commission operating in this field has become increasingly active since the late
1990s. Although it deals with food safety rather than environmental issues, its
standards cover GM agricultural products and address relevant issues of scientific
uncertainty (Murphy and Chataway 2003; Brack, Falkner, and Goll 2003, 8–9;
Genewatch 2003, 6–7). In addition, the Cartagena Protocol on Biosafety adopted
in 2000 may influence WTO dispute settlement.
In 2004, the IPPC’s governing body, the Interim Commission on Phytosanitary
Measures (ICPM), adopted a supplement to the International Standard on Phytosa-
nitary Measures (ISPM) No. 11, which addresses the plant-pest risks of GMOs.5 As
a result of the decision to amend ISPM No. 11 instead of adopting a separate GMO
standard, the new IPPC provisions reflect, and build on, those for conventional
plants, rather than the SPS Agreement’s science-dependent approach. In fact, the
United States had called for a separate IPPC standard for GMOs, but had been will-
ing to compromise in order to ensure a quick adoption of IPPC standards that were
perceived as an alternative to, or precedent for the implementation of, the Cartagena
Biosafety Protocol (USDA 2001). Although the revised standard does not mention
precaution, it may support the DRD’s approach in several ways. Perhaps most
importantly, it is modest as to the role of science. ISPM No. 11 treats scientific un-
certainty as an inherent problem of risk assessment that regulation needs to address
(Sections 2.4, 3). Also, ISPM No. 11 risk assessment is based on economic, social,
and environmental considerations in addition to science (Section 2.3.2.4). Other
ISPM No. 11 provisions may also offer support for the DRD. For example, the po-
tential to be injurious may be sufficient for a GMO to qualify as a pest (Section 1.1).
Direct, indirect, and long-term effects may be assessed (Section 2.3.1). In addition to
comparison to similar organisms, other ways of establishing the acceptable level of
risk are possible (Section 3.1). Many factors may be indicative of the need to subject
a GMO to a complete risk assessment, including lack of knowledge and insufficient
information (Annex 3).
The CAC has adopted two particularly relevant standards in 2003. First, the
‘‘Working Principles for Risk Analysis for Application in the Framework of the
Codex Alimentarius’’ provide general guidelines. However, because of the implica-
tions for the SPS Agreement and the ongoing controversies over the DRD’s precau-
tionary approach (Murphy and Chataway 2003, 4, 13; Cosbey 2000), the CAC
failed to agree on guidance for cases involving scientific uncertainty. It therefore
decided that, rather than adopting an official standard, it would merely adopt a
‘‘related text, such as a code of practice’’ in possible future cases (CAC 2001, 12).
Second, the CAC adopted ‘‘Principles for the Risk Analysis of Food Derived from
Modern Biotechnology,’’ which tend to support the DRD’s precautionary approach.
First, the Principles are also technology-specific and stipulate that GM food must
generally undergo a risk assessment. Second, the standards support an interpretation
of ‘‘substantive equivalence’’ (of GM and conventional food) that seems compatible
with the DRD’s precautionary approach. Traditionally, substantive equivalence
used to be invoked to deny biotechnology-specific risk assessment of ‘‘substantively
equivalent’’ GM products. However, the Codex standard employs the concept
primarily as a heuristic tool to guide rather than to deny risk assessment of GM
products (Levidow and Murphy 2002, 11–12). Finally, scientific uncertainty is rec-
ognized as inherent in risk assessment, in particular with respect to allergenicity
(Murphy and Chataway 2003, 15–16).
The Cartagena Protocol on Biosafety to the Convention on Biological Diversity
(CBD) is also likely to influence the interaction. Like the DRD, the Protocol is
technology-specific (Brack, Falkner, and Goll 2003, 3–4). Going beyond IPPC and
Codex standards, it explicitly refers to precaution in its Preamble and Article 1
(objectives). In addition, the Protocol suggests a precautionary approach in its
operational clauses, stating that ‘‘lack of scientific certainty . . . shall not prevent . . .
a decision . . . in order to avoid or minimize . . . potential adverse effects’’ (Art. 10(6);
see also chapter 8). The Protocol is not recognized as a relevant international
standard-setting body under the SPS Agreement, and the contentiousness of the Pro-
tocol makes it unlikely that it will be recognized as such anytime soon. Even without
formal recognition, however, the Protocol forms part of the wider legal context in
which the SPS Agreement must be interpreted (Oliva 2004, 9) and may thus have
a—weaker—impact on WTO dispute-settlement proceedings.
The DRD interacts with the SPS Agreement as a result of commitment to the ex-
tent that, first, the DRD influenced the ISPM, CAC, and Biosafety negotiations and,
second, these institutions contribute to a corresponding reinterpretation of the SPS
Agreement. More specifically, the EU member states’ commitment to the DRD’s pre-
cautionary approach seems to have strengthened their resolve to promote similar
measures in relevant third institutions. In turn, these institutions affect the interpre-
tation of the SPS Agreement.
Furthermore, relevant decisions have already been made within the EU. The
EU in 2003 adopted Regulation 1830/2003 implementing the traceability and
labeling requirements of the revised DRD. To reduce negative trade impacts, the
Regulation introduces thresholds for accidental or technically unavoidable GM con-
tamination and even allows for contamination by certain GMOs that have not yet
been authorized in the EU. Following the adoption of additional pieces of imple-
menting legislation, the EU approved import of a GM maize variety in May 2004,
thereby ending the de facto moratorium. Although this decision increases the chan-
ces that the moratorium could be deemed a temporary precautionary measure under
the SPS Agreement (Brack, Falkner, and Goll 2003, 9–10), it seems unlikely to re-
solve the wider tensions between the DRD’s precautionary approach and the SPS
Agreement.
There have also been attempts to address the issue of precaution in WTO negotia-
tions. In the context of the 1999 WTO Ministerial Conference in Seattle, the United
States and others proposed unsuccessfully to form a working group to examine
GMO-related issues (Wolff 2003, 12; see also chapter 8). Subsequently, several
WTO members—including the EU, the United States, South Korea, and Japan—
have taken relevant initiatives within the WTO to pursue their varying objectives
(European Commission 2000; European Union 2001; WTO 2000; Wolff 2003,
12). Finally, the relationship between the WTO and multilateral environmental
agreements became part of the Doha Round of trade negotiations agreed to in
2001. The negotiations could clarify the relationship between the Biosafety Protocol
and the SPS Agreement. However, these initiatives have so far made little progress
(Brack, Falkner, and Goll 2003, 7).
The interaction may thus eventually be played out by means of the WTO dispute-
settlement procedure rather than political decision making. The WTO would, prima
facie, appear to be in a strong position to resolve the interaction in favor of the SPS
Agreement due to its powerful dispute-settlement mechanism. However, WTO dis-
pute settlement is embedded in other international institutions, in this case primarily
the Codex, the IPPC, and the Biosafety Protocol. These institutions approach the
role of science and scientific uncertainty in risk assessment in ways that tend to be
more similar to the DRD’s precautionary approach than to rudimentary precaution
enshrined in the SPS Agreement. In particular with respect to the Biosafety Protocol,
and to a lesser extent the Codex, this may partly be explained in terms of successful
EU efforts to internationalize the DRD’s precautionary approach.
Conclusions
Most of the DRD’s main institutional interactions involve either EU sectoral legisla-
tion or the WTO. This may help explain why most interactions have had the DRD
as a source and have resulted in disruption (at the outcome level). Interactions
with sectoral legislation often reflect the challenge of coordinating precautionary,
technology-specific legislation with more traditional approval procedures. In partic-
ular, this holds for the four cases involving the DRD’s precautionary environmental
risk assessment. The DRD’s risk assessment reflects a belief in potentially significant
risks of GM products combined with an actual or perceived lack of experience/
knowledge. In contrast, sectoral legislation builds on accumulated experience with
various product groups and focuses on assessing product characteristics from the
perspective of intended use. Because the DRD challenges the basic assumption un-
derlying sectoral legislation—that regulation on the basis of product groups is gen-
erally sufficient—it emerged as a source of disruptive Behavioral Interaction with
sectoral legislation. The interaction with the Pesticides Directive exemplifies the
tensions: although risk assessment under this instrument addressed the secondary
effects of pesticide-resistant GM crops, this was insufficient in terms of the DRD’s
precautionary approach, which covers a wider range of potential risks.
Interactions with the WTO, in particular the SPS Agreement, are related to similar
differences in regulatory approaches. Here, these differences translate into conflicts
over the role of science. More specifically, the SPS Agreement’s strong reliance on
science—which is essentially in harmony with the focus on use-related experience
at the base of sectoral legislation because such experience can often be transformed
into scientific knowledge—clashes with the core assumptions underlying precaution-
ary regulation: scientific uncertainty and insufficient knowledge. Prima facie, trade
restrictions caused by precautionary regulations are therefore more likely to dis-
rupt the implementation of the SPS Agreement than those caused by conventional
product-sector legislation.
Coordination to address the DRD’s disruptive effects is generally weak. This is re-
markable, in particular with respect to EU-level interactions, where coordination
could have benefited from a high degree of institutional integration. In fact, the
detailed analysis of the interactions with the Pesticides Directive and the SPS Agree-
ment suggests that coordination may have been more relevant at the less integrated
global level. Explicitly covering secondary effects, the revised DRD resolved the
interaction with the Pesticides Directive. But this solution resulted primarily from
the political and economic pressure generated by national bans and the de facto
moratorium. In addition, the underlying question of which normative yardstick
should be used to evaluate secondary effects has, if anything, only been addressed
in an insufficient and indirect way. In contrast, coordination appears to be more rel-
evant for efforts to address the interaction with the SPS Agreement. The standard-
setting bodies recognized by the SPS Agreement provide relevant forums. Although
negotiations take place ‘‘in the shadow of the WTO’’ (Cosbey 2000), links to other
institutions, such as the Biosafety Protocol, and entrenched regulatory traditions
seem to provide the CAC and the ISPM with sufficient autonomy to play a media-
ting role.
Notes
1. In addition to secondary effects, different interpretations regarding other issues—for exam-

ple, the use of antibiotic-resistant genes—contributed to the escalation of conflicts.
2. For instance, in June 1999 some EU member states supported the moratorium by
declaring their intention ‘‘to take a thoroughly precautionary approach . . . [and] not to
authorize the placing on the market of any GMOs until it is demonstrated that there
is no adverse effect on the environment and human health’’ (2194th Council Meeting—
ENVIRONMENT—Luxembourg, June 24–25, 1999; available at http://ue.eu.int/ueDocs/
cms_Data/docs/pressData/en/envir/ACF5B.htm).
3. ‘‘US Shifts Tactics in GMO Clash with EU,’’ Financial Times, October 15, 2002, 6.
4. ‘‘EU Plays for Time over US Threat on GMOs,’’ Agence France-Press (AFP), Brussels, Jan-
uary 12, 2003.
5. Using the terminology of the Cartagena Biosafety Protocol, the IPPC refers to living modi-
fied organisms (LMOs) rather than GMOs. For the present purpose and in everyday usage the
terms can be considered to be the same (see also chapter 8).
References
Barling, David. 1997. Regulatory Conflict and Marketing of Agricultural Biotechnology in

the European Community. In Jeffrey Stanyer and Gerrey Stoker, eds., Contemporary Political
Studies, 1040–1048. Nottingham: Political Studies Association of the U.K.
Becker, Geoffrey S., and Charles E. Hanrahan. 2004. Agricultural Biotechnology: The
U.S.-EU Dispute. CRS Report for Congress, Code RS21556, updated March 16, 2004.
Washington, DC: Congressional Research Service, Library of Congress.
Brack, Duncan, Robert Falkner, and Judith Goll. 2003. The Next Trade War? GM Products,
the Cartagena Protocol and the WTO. Briefing Paper No. 8. September 2003. London: Royal
Institute of International Affairs, Development Programme.
CAC. 2001. Report of the Twenty-Fourth Session of the Codex Alimentarius Commission,
Geneva, July 2–7, 2001. Doc. ALINORM 01/41.
Cantley, Mark F. 1995. The Regulation of Modern Biotechnology: A Historical and Euro-
pean Perspective. In Hans-Juergen Rehm and Gerald Reed, eds., Biotechnology Vol. 12:
Legal, Economic, and Ethical Dimensions, 505–579. Weinheim, Germany: VCH.
Cosbey, Aaron. 2000. A Forced Evolution: The Codex Alimentarius Commission, Scientific
Uncertainty and the Precautionary Principle. Winnipeg: International Institute for Sustainable
Development.
European Commission. 2000. Communication from the Commission on the Precautionary
Principle. Doc. COM (2000) 1, February 2, 2000. Brussels: European Commission.
European Union. 2001. Food Safety. Note by the European Communities. Doc. D(2001)
(DIVERS/500186 ur), Brussels, July 20, 2001; available at http://europa.eu.int/comm/
agriculture/external/wto/document/food_en.pdf.
Friends of the Earth. 1997. Austria Bans Genetech Maize. Mailout 3 (1): 6. Friends of the
Earth Europe Biotechnology Programme.
Fromwald, Susanne, and Sylvia Strauss. 1998, May. Genetically Engineered Oil Seed
Rape (AgrEvo/PGS): A Critical Assessment and Background Information. Amsterdam:
Greenpeace.
Genewatch. 2003, December. The GM Dispute at the WTO: Forcing GM Food on Europe?
GeneWatch UK. Briefing No. 25.
Gupta, Aarti. 1999. Framing ‘‘Biosafety’’ in an International Context: The Biosafety Protocol
Negotiations. ENRP Discussion Paper E-99-10. Cambridge, MA: Kennedy School of Govern-
ment, Harvard University.
von Homeyer, Ingmar. 2002. Institutional Change and Governance in the European Union:
The Case of Biotechnology Regulation. Doctoral dissertation, European University Institute,
Florence.
von Homeyer, Ingmar. 2003. The Revision of the Directive on the Deliberate Release of
Genetically Modified Organisms (GMOs) into the Environment. The European Public
Space Observatory (EUROPUB) Project. Case study report (WP 2). Available at ftp://ftp.iccr-
international.org/europub/europub-d2-annex2.pdf.
Kelch, David, Mark Simone, and Mary Lisa Madell. 1998. Biotechnology in Agriculture
Confronts WTO Agreements. Agriculture in the WTO/WRS-98-4. Washington, DC: Eco-
nomic Research Service, U.S. Department of Agriculture (USDA).
Levidow, Les. 2001. Unsound Science? Trans-Atlantic Regulatory Disputes over GM Crops.
National Europe Centre Paper No. 3. Canberra: The Australian National University.
Levidow, Les, and Susan Carr. 2000. UK: Precautionary Commercialisation? Journal of Risk
Research 3 (3): 261–270.
Levidow, Les, Susan Carr, and David Wield. 1999a. EU Level Report. In Safety Regula-
tion of Transgenic Crops: Completing the Internal Market? A Study on the Implementa-
tion of Directive 90/220. Centre for Technology Strategy, Faculty of Technology, The
Open University, Milton Keynes, UK. Available at http://technology.open.ac.uk/cts/srtc/EU-
Levelreport%20update.pdf.
Levidow, Les, Susan Carr, and David Wield. 1999b. United Kingdom: Precautionary Com-
mercialisation. In Safety Regulation of Transgenic Crops: Completing the Internal Market?
A Study on the Implementation of Directive 90/220. Centre for Technology Strategy, Faculty
of Technology, The Open University, Milton Keynes. Available at http://technology.open.ac
.uk/cts/srtc/UK-NATReport.pdf.
Levidow, Les, Susan Carr, and David Wield. 2000. Genetically Modified Crops in the Euro-
pean Union: Regulatory Conflicts as Precautionary Opportunities? Journal of Risk Research 3
(3): 198–208.
Levidow, Les, and Joseph Murphy. 2002. The Decline of Substantial Equivalence: How Civil
Society Demoted a Risky Concept. Paper presented at a conference at the Institute of Devel-
opment Studies titled ‘‘Science and Citizenship in a Global Context: Challenges from New
Technologies,’’ University of Sussex, December 12–13.
Levidow, Les, and Joseph Murphy. 2003. Reframing Regulatory Science: Trans-Atlantic Con-
flicts over GM Crops. Cahiers d’économie et sociologie rurales (68–69): 48–74.
Murphy, Joseph, and Joanna Chataway. 2003, October 1. Science and International Agree-
ments: The Case of Genetically Modified Organisms and Risk. Written Evidence to the
Science and Technology Select Committee of the House of Lords (Sub-Committee I). Avail-
able at http://www.innogen.ac.uk/ownImages/Lords%20SCST%20evidence.pdf.
Oliva, Maria Julia. 2004. Science and Precaution in the GMO Dispute: A Brief Analysis of the
U.S. Submission. Bridges 8 (5): 8–10.
Peel, Jacqueline. 2004. Risk Regulation under the WTO SPS Agreement: Science as an Inter-
national Normative Yardstick? Jean Monnet Working Paper 02/04. New York: New York
University School of Law.
Roy, Alexis, and Pierre-Benoit Joly. 1999. France: Broadening Precautionary Expertise. In
Safety Regulation of Transgenic Crops: Completing the Internal Market? A Study on the
Implementation of Directive 90/220. Centre for Technology Strategy, Faculty of Technology,
The Open University, Milton Keynes, UK. Available at http://technology.open.ac.uk/cts/srtc/
FR-NATReport.pdf.
Roy, Alexis, and Pierre-Benoit Joly. 2000. France: Broadening Precautionary Expertise? Jour-
nal of Risk Research 3 (3): 247–254.
Sauter, Arnold, and Rolf Meyer. 2000. Risikoabschätzung und Nachzulassungs-Monitoring
transgener Pflanzen. TAB-Arbeitsbericht Nr. 68. Berlin: Büro für Technikfolgen-Abschätzung
beim Deutschen Bundestag.
von Schomberg, René. 1998. An Appraisal of the Working in Practice of Directive 90/220 on
the Deliberate Release of Genetically Modified Organisms. Draft Final Report, Scientific and
Technological Options Assessment (STOA), European Parliament, Luxembourg.
Scott, Joanne. 2000. On Kith and Kine (and Crustaceans): Trade and Environment in the EU
and the WTO. In Joseph Weiler, ed., The EU, the WTO and NAFTA: Towards a Common
Law of International Trade, 125–167. Oxford: Oxford University Press.
Toft, Jesper. 1999. Denmark: Potential Polarisation or Consensus? In Safety Regulation of
Transgenic Crops: Completing the Internal Market? A Study on the Implementation of Direc-
tive 90/220. Centre for Technology Strategy, Faculty of Technology, The Open University,
Milton Keynes, UK. Available at http://technology.open.ac.uk/cts/srtc/DK-NATReport.pdf.
Torgersen, Helge, and Franz Seifert. 2000. Austria: Precautionary Blockage of Agricultural
Biotechnology. Journal of Risk Research 3 (3): 209–217.
UNEP. 1997, March 18. Background Document on Existing International Agreements Re-
lated to Biosafety. United Nations Environment Programme. Doc. UNEP/CBD/BSWG/2/3.
USDA. 2001, August 23. APHIS Public Meeting in the Matter of International Plant Protec-
tion Convention (IPPC) Standard on Plant Pest Risks Associated with Living Modified
Organisms (LMOs). Washington, DC: U.S. Department of Agriculture.
USDA. 2003. United Kingdom Exporter Guide Annual 2003. GAIN-Report-UK 3031. Wash-
ington, DC: U.S. Department of Agriculture Foreign Agricultural Service.
U.S. Mission to the EU. 1999, March 19. Rominger Press Conference at USEU on Biotech-
nology. United States Mission to the European Union. Available at http://www.useu.be/
archive/Roming322.html.
Vogel, David. 2001. Ships Passing in the Night: The Changing Politics of Risk Regulation in
Europe and the United States. EUI Working Paper RSC No. 2001/16. Florence: European
University Institute.
Winkler, Roland. 1999. Legal Aspects at the EU Level. In Safety Regulation of Transgenic
Crops: Completing the Internal Market? A Study on the Implementation of Directive 90/
220. Centre for Technology Strategy, Faculty of Technology, The Open University, Milton
Keynes, UK. Available at http://technology.open.ac.uk/cts/srtc/EU-legal-aspects.pdf.
Wolff, Christiane. 2003. Biosafety, Biotechnology and the WTO: The SPS and TBT Agree-
ments and GATT. Agriculture and Commodities Division, WTO Secretariat.
WTO. 2000, April 18. WTO Secretariat. Summary of the Meeting Held on 15–16 March
2000. Committee on Sanitary and Phytosanitary Measures. World Trade Organization. Doc.
G/SPS/R/18.
Yerkey, Gary G. 1999. U.S. Considers Filing WTO Complaint over EU Barriers to GMO
Trade, USTR Says. Available at http://www.biotech-info.net/filing_complaint.html.
12
The EU Air Quality Framework Directive: Shaped
and Saved by Interaction?
Jørgen Wettestad
Although the development of EU air-quality policy started in the 1980s, the 1996
Air Quality Framework Directive (FD) is an important legal milestone in the EU’s
fight against air pollution (Wettestad 2002). The Directive itself did not create pre-
cise air-quality objectives. The main reason for describing it as a milestone is found
in four subsequent daughter directives, which have both broadened the scope of EU
policy in this area in terms of substances covered and introduced far more stringent
air-quality limit values than previously applied. Because it makes sense to see the
1996 Directive and the subsequent daughter directives as different facets of the
same institutional entity, references in the rest of the chapter to the Air Quality FD
also encompass the daughter directives.
The FD has interacted and interacts with a number of institutions and processes
both outside and within the EU. A main thesis of this chapter is that to understand
both the shaping and future performance of the EU air-quality FD, it is crucial to
understand the FD’s interaction with external and internal institutions. Three cases
of (potential) interaction are investigated in detail. In terms of vertical interaction
with external institutions, the air-quality guidelines produced by the World Health
Organization (WHO) seem to have been especially crucial influences on the EU leg-
islation. It is interesting to scrutinize this interaction further for several reasons.
First, because there is considerable internal EU expertise in this field, this case of in-
teraction touches on the question of the extent to which EU environmental policy
should be a ‘‘home brew’’ and the extent to which it should rely on external exper-
tise. Moreover, the WHO guidelines are nonbinding international ‘‘soft law,’’ and
hence this case also touches on the question of the extent to which nonbinding
instruments can have a profound influence on the development of binding legisla-
tion in international (environmental) politics. In addition, because the WHO is very
much a scientific organization and the EU is primarily about politics, the case
286 Jørgen Wettestad
touches on the generally complex, but important, relationship between science and
politics in the international environmental arena (e.g., Andresen et al. 2000). Hence,
a first central puzzle in this chapter becomes: How can it be explained that the bind-
ing EU Air Quality FD has seemingly been heavily shaped by the nonbinding and
external WHO guidelines?
Horizontal interaction within the EU is distinctly different from the relationship
between EU legislation and the policies of other international institutions. Within
the EU, the different strands of (environmental) policy should ideally be compatible.
But separate decision making processes may in practice easily lead to uncoordinated
and fragmented decision making. Given the central role of cleaner fuel and lower
emissions from vehicles for the improvement of air quality, there is an obvious link
to the revised fuel standards and vehicle-emission limits emanating from the Auto-
Oil I process. After being founded in 1992, the Auto-Oil I Program led to directive
proposals in 1996 on stricter fuel standards and tighter emission limits, and these
directives were adopted in 1998. Hence, the air-quality directives and the Auto-Oil
I Program and directives were debated and adopted during roughly the same period.
They also partly targeted the same substances, such as nitrogen oxides (NOX ), ben-
zene, and carbon monoxide. But the fact that the EU launched an air-policy integra-
tion program in 2001—called the Clean Air for Europe (CAFÉ) Program—could be
a sign that integration and coordination of air-quality and vehicle-emissions reduc-
tion policies had been lacking so far. So, a second central question in this chapter
becomes: To what extent have the air-quality and Auto-Oil policymaking processes
interacted and how well have they been coordinated?
However, this is not necessarily the only or the most important process of interac-
tion between these strands of legislation. Given the obvious role of cleaner fuel and
lowered emissions for achieving better air quality, there is a distinct possibility that
the really important interaction effects take place at the stage of outcomes—that
is, when the focused directives are implemented. Hence, a third central question
addressed in this chapter becomes: Will the most important interaction between EU
Auto-Oil standards and air-quality standards take place when the former standards
are implemented? In other words, will the Air Quality FD’s fate and performance be
‘‘saved’’ by the Auto-Oil directives?
The next section sums up the main elements of the Air Quality FD and the four
subsequent daughter directives. This is followed by a summary of how the Air
Quality FD has interacted with various institutions and processes within and outside
of the EU. The major emphasis of the chapter is then on the interactions with the
EU Air Quality Framework Directive 287
WHO air-quality guidelines and the Auto-Oil directives, including an assessment of

the CAFÉ response. The chapter concludes with some more general observations
and lessons.
The Air Quality Framework Directive
The Framework Directive is part of the air-quality policy of the EU. Instruments of
this policy set binding or recommendatory air-quality standards for particular pollu-
tants, while they do not specify the pollution-abatement measures that member
states might apply to achieve the stipulated limits. The first EU air-quality directives
were adopted in the early 1980s. In 1980, Directive 80/779/EEC on air-quality limit
values and guide values for sulfur dioxide and suspended particulates was adopted.
This was the first piece of Communitywide legislation to lay down mandatory air-
quality standards. The next air-quality directive targeted lead and was adopted in
1982 (Directive 82/884/EEC on limit value for lead in the air). The lead directive
was followed by Directive 85/203/EEC on air-quality standards for nitrogen diox-
ide. The EU’s fourth Environmental Action Program, launched in 1987, placed
more emphasis on the problem of photochemical pollution and particularly ground-
level ozone. The Commission described this as ‘‘one of the major environmental
problems’’ of the century (Haigh 2003, sec. 6.15, 2). Moreover, in the early 1990s,
increasing concentrations of ground-level ozone were a source of concern for many
European countries. Against this background, Directive 92/72/EEC on air pollution
by ozone was adopted in 1992. The directive required member states to develop a
network for the collection of information on ozone levels.
According to Elsom (1999, 106), a revision of EU air-quality policy was spurred
by WHO’s decision in 1993 to review and revise its 1987 air-quality guideline
values for Europe. By the spring of 1993, early drafts of a framework directive on
air quality were discussed in meetings between the Commission and national offi-
cials. More than twenty substances were targeted1 and a system consisting of differ-
ent types of quality objectives was envisaged. ‘‘Alert thresholds’’ were related to the
top three pollutants; ‘‘guide values’’ recommended by international expert groups
such as WHO were to be established for all the pollutants; and ‘‘limit values’’
(i.e., obligatory environmental quality standards to be met) were to be set for the
pollutants in stages.
As indicated, the 1996 Air Quality FD (i.e., EU Council Directive 96/62/EC on
Ambient Air Quality Assessment and Management) itself did not create any precise
air-quality objectives, but rather set out a framework and basic principles for
ambient-air-quality monitoring and management. These were to go into effect once
daughter directives for specific pollutants had been adopted. At the same time that
ambient standards were set, criteria were to be established for the assessment of air
pollution. These criteria were to include details on the location, number, and type of
sampling sites, as well as on the use of other techniques such as modeling. Once
daughter directives had been adopted, member states would have to assess ambient
air quality in accordance with the provisions specified.2 A key ingredient in the
Directive was a requirement for all EU countries to adopt monitoring systems in
accordance with common standards (both in terms of the location of monitor-
ing sites and measurement techniques), and to report the results regularly to the
Commission.
As the next step, several groups of experts, consisting of representatives from
the European Commission, the European Environment Agency, the World Health
Organization, EU member states, industry, and environmental nongovernmental
organizations (NGOs), developed proposals for daughter directives. The first one
concerned standards for sulfur dioxide, nitrogen dioxide, particulate matter, and
lead. In June 1998, a common position on this daughter directive was obtained in
the EU Council of Ministers, signaling an overall tightening of standards, but also
with standards for particulates and NOX ‘‘significantly diluted’’ from those origi-
nally proposed by the Commission.3 The first daughter directive was then formally
adopted as EU Council Directive 99/30/EC (Directive Relating to Limit Values for
Sulfur Dioxide, Nitrogen Dioxide and Oxides of Nitrogen, Particulate Matter and
Lead in Ambient Air) in April 1999. A second daughter directive was proposed by
the Commission in December 1998, targeting benzene and carbon monoxide. It was
adopted in November 2000 (EU Council Directive 2000/69/EC Relating to Limit
Values for Benzene and Carbon Monoxide in Ambient Air).
A third daughter directive has targeted the issue of ground-level ozone. As indi-
cated, this issue was given specific emphasis in EU air-pollution policymaking from
the early 1990s on. The formal proposal for a directive on ozone in ambient air was
presented in June 1999. The proposed daughter directive contained aspirational,
nonbinding target values for ozone by 2010, which were ‘‘widely seen as ambi-
tious’’.4 Agreement was reached in November 2001 and the resulting Directive
2002/3/EC (Directive relating to ozone in ambient air) adopted in February 2002,
with member states being obliged to meet the 2010 targets ‘‘save where not achiev-
able through proportionate measures.’’
In December 2001, it was reported that the Commission had delayed proposing a
fourth daughter directive, which was supposed to target polyaromatic hydrocarbons
(PAHs), cadmium, arsenic, and nickel. This was due to vocal industry protests.5 In
October 2002, it was announced that the Commission was planning to launch a
directive with nonbinding instead of mandatory limits.6 The EU Council and the Eu-
ropean Parliament reached a final compromise in April 2004 on ‘‘stricter, though
still not strictly binding, targets’’.7
Main Horizontal and Vertical Interactions of the Air Quality Framework Directive:
An Overview
The FD has interacted and interacts with a number of other EU policy instruments
and international institutions. We have provided an inventory of such interaction
elsewhere and only provide a brief summary here (Wettestad and Farmer 2001).
The relevant interactions are summed up in table 12.1.
With regard to other EU policy instruments, the Air Quality FD has in particular
interacted with five processes and instruments at the level of implementation (Behav-
ioral Interaction). First of all, the establishment of the Single Market in the EU has
Table 12.1
Interactions of the EU Air Quality Framework Directive
EU Single Market Has amplified a trend of increasing transport
emissions and hence disruptively affected the process
of improving air quality
EU National Emission Ceilings Sets emission ceilings, which will help achieve air-
(NEC) Directive quality targets
EU Auto-Oil I Directives Set stricter fuel standards and vehicle-emission
limits, which will help achieve air-quality targets
EU SAVE Program Contributes to increased energy efficiency, which
helps reduce emissions and improve air quality
EU Renewable Energy Directive Contributes to enhanced use of renewable-energy
sources, which helps reduce emissions and improve
air quality
Convention on Long-Range Developed the critical-loads concept, which
Transboundary Air Pollution facilitated the EU process of setting air-quality
(CLRTAP) standards
World Health Organization Produces air-quality guidelines, which have heavily
(WHO) influenced EU air-quality standards
led to a growth of emissions of air pollutants in particular in the area of traffic,

which has negatively affected air quality. In contrast, other EU legislative instru-
ments have directly and indirectly contributed to emission reductions and thus have
supported the objectives of the Air Quality FD. These are in particular the Acidifica-
tion Strategy and the National Emission Ceilings (NEC) Directive, the EU’s SAVE
Program entailing directives on minimum energy-efficiency standards and related
labeling, the Renewable Energy Directive promoting the use of renewable sources
of energy, and the Auto-Oil I Program and related directives on fuel standards and
vehicle-emission limits. The EU’s Directive on Integrated Pollution Prevention and
Control (IPPC Directive) has had a similar effect (chapter 9).
Regarding vertical interaction with international institutions, two cases are partic-
ularly relevant. First, the so-called critical-loads concept developed in the context of
the Convention on Long-Range Transboundary Air Pollution (CLRTAP) under the
UN Economic Commission for Europe (UNECE) facilitated the setting of air-quality
standards in the framework of the Air Quality FD. A critical load can be defined as
a quantitative estimate of an exposure to one or more pollutants below which sig-
nificantly harmful effects on specified sensitive elements of the environment do not
occur according to present knowledge (Wettestad 2002). Second, the air-quality
guidelines produced by the WHO influenced the air-quality standards of the FD, as
further detailed in the next section. Both these interactions have operated at the
output level in that agreements reached under CLRTAP and the WHO influenced
the design and content of the Air Quality FD (Interaction through Commitment).
Vertical Interaction with the WHO: External Air-Quality Guidelines as Ambitious

Benchmarks for EU Policy Entrepreneurs?
The WHO Air-Quality Guidelines

The WHO recognized as early as 1958 that air pollution was a global threat to
health. In 1972, initial guidelines regarding the levels of ambient air pollutants that
constitute hazards to health were formulated for the ‘‘classic’’ compounds sulfur di-
oxide (SO2 ), solid particulate matter, carbon monoxide (CO), and photochemical
oxidants. The next major step in this process was the publication of the Air Quality
Guidelines for Europe for a much broader set of air pollutants in 1987. These
guidelines were based on evidence from the epidemiological and toxicological lit-
erature published in Europe and North America. Work on making the guidelines
globally applicable started in 1997. With regard to the health significance of air pol-
lution, a new database of time-series studies was developed, first in the United States
and later in Europe and other areas.8
The revision of the guidelines during the 1990s involved a more elaborate process
and several stages. First, a planning group composed of national experts outlined
the structure of the revised report. Then experts in various countries were asked to
produce drafts for various sections of the report. These drafts were reviewed by
WHO working groups, and subsequently sent out for a final round of peer review
by other experts.9 With regard to the specific role of the EU, a funding agreement
between the European Commission and the European Regional Centre of WHO,
resulting in WHO’s 2000 Air Quality Guidelines for Europe, should be noted (Eu-
ropean Commission 2001, 14). As guidelines for EU and national policymaking,
they constitute nonbinding soft law.
Negotiating the Daughter Directives: The Heyday of the WHO Guidelines

The EU’s fifth Environmental Action Program, adopted in 1992, called for the effec-
tive protection of everyone from recognized health risks caused by air pollution. As a
central ‘‘action,’’ proposals for amendments of existing legislation were pinpointed.
A central target for 2000 was that ‘‘WHO values [should] become mandatory at EU
level’’ (European Commission 1993, 8).
While the work on the Air Quality FD was spurred by the revision of the WHO
guidelines, the processes of producing the daughter directives have been character-
ized as ‘‘the heyday of the WHO guidelines’’ with regard to the use of expertise.10
In the explanatory memorandum to the proposal for the first daughter directive
(European Commission 1997), the Commission justified each of the proposed limit
values by reference to the most recent WHO guidelines. In the case of sulfur di-
oxide, WHO guidelines established a maximum concentration limit averaged over
ten minutes. The Commission then ‘‘translated’’ this guideline into suggestions for
hourly and daily limit values. The suggested hourly limit was 350 micrograms per
cubic meter (m/m 3 ), and the suggested 24-hour limit was 125 m/m 3 . As proposals
were developed in expert groups with member-state representation, the proposals
for limit values were accompanied by political modifications in the form of allowed
exceedances per year and deadlines for compliance. For instance, the 350 m/m 3
hourly limit was not to be exceeded more than twenty-four times a year. Because
Commission proposals already included central member-state opinions, they were
not much altered in the legislative process. For example, the Commission proposal
for SO2 was adopted unaltered by the Council.
Even though this was not always the case, the WHO values were generally main-
tained throughout. For example, WHO’s advised maximum hourly limit for NOX
was 200 m/m 3 , complemented by an annual 40 m/m 3 limit, and a vegetation limit of
30 m/m 3 . The Commission’s proposal for the first daughter directive of the Air Qual-
ity FD closely followed the WHO with regard to all these limit values, but included
an opening for eight exceedances of the hourly limit per year. The final directive
maintained all the WHO limit values (i.e., 200/40/30), but increased the allowed
exceedances of the hourly limit to eighteen times per year (to be met as of 2005).
In the case of particulate matter, the number of allowed exceedances was increased
from twenty-five to thirty-five. As indicated above, some saw this as a ‘‘significant
dilution’’ of environmental ambitiousness.11
In the case of the second daughter directive, targeting benzene and carbon mon-
oxide, the connection to the WHO guidelines was somewhat weaker. This had to do
in particular with the fact that WHO could not come up with a recommended safe
level of exposure for benzene (WHO 2000, chap. 5.2). However, in the case of car-
bon monoxide, the maximum daily eight-hour mean of 10 mg/m 3 recommended by
WHO (2000, chap. 5.5) was proposed by the Commission and adopted by the
Council.
To protect human health, the June 1999 proposal for a third daughter directive
on ground-level ozone took its lead from the WHO guidelines, which called for a
limit on ambient ozone concentrations of 120 m/m 3 . Based on stakeholder and ex-
pert input, the Commission then suggested that this limit could be breached up to
twenty days per calendar year (Amann and Lutz 2000, 9). According to Amann
and Lutz (2000, 18), a higher 160 m/m 3 limit value with fewer allowed exceedances
was discussed in the preparatory work. However, the Commission saw it as impor-
tant to maintain the connection to the WHO guidelines and hence the 120 m/m 3
limit value (with somewhat more allowed exceedances) was chosen. Cutting a long
story short, agreement was finally reached in November 2001, with member states
being obliged to meet the 120 m/m 3 target by 2010 ‘‘save where not achievable
through proportionate measures.’’ The final agreement also allowed twenty-five
exceedances per year, hence diluting the connection to the WHO guidelines a little
further.12 Overall, the agreement constituted a typical EU compromise by creating
more than aspirational, nonbinding targets, but still with a certain flexibility in
terms of bindingness.

The WHO guidelines have been the most important scientific reference point in EU
policymaking on air quality. As soon as WHO’s work in this area started in 1987, it
became an important benchmark in the EU context. Already in the 1992 ozone
directive, the threshold values were directly derived from the WHO guidelines. In
the same year, the EU’s fifth Environmental Action Program established as a target
that WHO values should become mandatory within the EU by 2000. This can
be interpreted through the lenses of Cognitive Interaction. The WHO guidelines of
1987 constituted information that influenced the perceptions of EU decision makers
and made them adopt these Guidelines as a template for EU policy. The broad-
based international character and participation of high-quality experts clearly gave
and continue to give WHO’s processes a high-quality stamp within the EU.
The interaction between WHO and the Air Quality FD can primarily be inter-
preted through the lenses of Interaction through Commitment. Actors within the
EU became committed to the WHO guidelines through the fifth Environmental
Action Program of 1992. Although the Action Programs did not become legally
binding until 1993, targets in the earlier Programs represented strong political sig-
nals for EU policymaking. As pointed out by Weale et al. (2000, 58), all five Action
Programs provided important indications of policy orientation. Hence, actors within
the EU have used the scientific and politically undiluted WHO guidelines as bench-
marks, from which it was difficult to diverge because of the existing general com-
mitment to them. Marking ‘‘the heyday of the WHO guidelines,’’ the Commission
utilized WHO target values as the foundation for its proposals for Air Quality
daughter directives whenever an appropriate WHO guideline existed. Even though
the WHO targets were always politically softened during the preparatory work on
the Commission proposals, mostly by introducing a number of allowed days of
exceedance, they still closely reflected the WHO guidelines. This also holds true
for the occasions on which the softening was increased further in the Council
(figure 12.1).
The interaction has been intentional and has had synergistic effects. The WHO
guidelines have been produced in order to influence other actors such as the EU.
There are strong reasons to assume that the EU has, as a result, adopted more am-
bitious policies than would have come about without the WHO benchmarks. For
instance, all the summaries of the preparatory processes published in the Commis-
sion’s directive proposals include information about some member states and indus-
trial stakeholders reacting negatively to the high level of policy ambition stemming
Figure 12.1
WHO guidelines strengthen EU air-quality legislation
from the WHO input. However, to measure this effect more precisely, more
detailed knowledge about the relevant negotiation processes and the exact degree
of member-state resistance overcome with the help of the WHO guidelines would
be required.
Because of the synergistic effects so far, it may not be surprising that the WHO
and its air-quality guidelines are likely to continue to play a significant role in EU
policymaking in the years ahead. As mentioned in the introduction, in 2001, the
Commission’s Environment Directorate launched an initiative to incorporate the
various aspects of EU air-quality policy into a unified framework under the banner
of Clean Air for Europe or CAFÉ (European Commission 2001). CAFÉ can hence
be seen as providing an important EU response to the interaction and (lack of)
coordination between subissues in the issue area of air-pollution control. So how
does the EU assess the interaction with WHO? Overall, the response is positive.
According to the 2001 CAFÉ Communication, during the consultation process
leading to the launch of CAFÉ, it became clear that a large majority of national rep-
resentatives and stakeholders supported the use of WHO guidelines as the funda-
mental advice on risk (European Commission 2001, 14). This seems also to some
extent to be the position of the Commission, although the picture is somewhat
ambiguous.13
Accordingly, the Commission signed a contract with WHO on a revision of the

‘‘Guidelines for Europe.’’14 Due to delays in getting the contract going, it was uncer-
tain whether these guidelines could be ready for the next main round of revision of
EU air-quality policy. Moreover, the process of producing the WHO guidelines has
been criticized for lack of transparency. Hence, there seems to be a need to rethink
the way the EU organizes its scientific advisory process in this issue area. In a re-
appraisal process, several options could be evaluated, including a possible lower-
cost option of producing the relevant knowledge ‘‘in-house’’ in the EU.
Horizontal Interaction with the EU Auto-Oil Legislation: Limited Policy

Coordination, but Crucial Implementation Effects?
EU internal interaction is distinctly different from the relationship between EU

legislation and international institutions. Within the EU, the different strands of (en-
vironmental) policy should ideally interact and be compatible. In the air-quality con-
text, measures to curb vehicle emissions are of central interest. In this field, the
Auto-Oil I process was the most important EU process in the 1990s. Since discus-
sions on the air-quality directives and the Auto-Oil I Program and related directives
proceeded roughly in parallel and partly targeted the same substances, an interesting
first question becomes: How closely have these ‘‘functionally’’ related processes and
directives interacted at the level of outputs? This issue is discussed in the next sub-
section. Subsequently, interaction at the level of outcomes is assessed since cleaner
fuel and stricter emission limits contribute to better air quality, and the most impor-
tant interaction effects may thus emanate from the implementation of the Auto-Oil
directives. This is followed by a discussion of the EU system’s main response to
the interaction, before a concluding summary and interpretation of this horizontal
interaction.
Output Interaction: The Evolution of Auto-Oil I

Two main phases in the Auto-Oil I process can be distinguished: a preparatory
phase, from 1992 to 1996, and a policymaking phase, from 1996 to 1998.
The Auto-Oil Preparatory Phase (1992–1996) The Auto-Oil process was initiated
by the Commission in 1992. The adoption of the strengthened 1990 U.S. Clean Air
Act, WHO’s air-quality guidelines for Europe, and growing frustration in industry
circles over a policymaking approach paying too little attention to costs in relation
to benefits, have all been identified as contributing factors (Weale et al. 2000; Young
and Wallace 2000, chap. 2; Haigh 2003, sec. 6.8, 8). There was thus an interest in
finding more cost-effective solutions.15 As a response to these various concerns, a
symposium on vehicle-emission standards for the year 2000 was organized and
attended by a wide range of participants (Friedrich, Tappe, and Wurzel 1998,
105). A central theme of this symposium was the development of cost-effective mea-
sures based on ambient-air-quality standards.
The Commissioners for Environment, Industry and Energy then invited the Euro-
pean Automotive Manufacturers Association (ACEA) and the European Petroleum
Industry Association (EUROPIA) to collaborate in a technical research program. It
was decided to launch three independent but interrelated projects:
Urban ambient-air-quality studies. The aim here was to predict the air quality of
seven European cities (Athens, Cologne, The Hague, London, Lyon, Madrid, and
Milan) and ground-level ozone across the EU for the year 2010, and on this basis
derive emission-reduction targets.
A ‘‘European Program on Emissions, Fuels and Engine Technologies,’’ focusing on
the effect of vehicle technology and fuel characteristics on emissions.
A cost-effectiveness study, calculating the costs and emission impact for different
emission-reduction measures.
While legislation proposals were originally to be ready by the end of 1994, this
was delayed,16 and the Commission did not put forward the first two proposals
for directives to take effect in the year 2000 until June 1996, when the main policy-
making phase of this process started.
To what extent did EU air-quality legislation and the Auto-Oil I process influence
each other? Given the early stage of development of EU air-quality legislation, its in-
fluence on the developing fuel-standard and vehicles-emission legislation was almost
inevitably weak. Instead, the Auto-Oil legislation was directly influenced by the
WHO air-quality guidelines that acquired additional force because of the expected
EU air-quality legislation. As noted by Goodwin (1999, 11), ‘‘The levels of air pol-
lution that were used as the target levels came from the emerging guidelines being
developed by the World Health Organization. These were stricter than the stan-
dards that applied in the Member States and the EU but were used because they
were expected to form the basis for future legislation.’’ Influence was also weak the
other way around. Given the general character of the FD and its main focus on
monitoring, it is not surprising to find no formal reference to the developing Auto-
Oil legislation in it. Article 7 only mentions the possible need for measures targeting
motor-vehicle traffic in the air-quality action plans to be drawn up by member
states.
The Auto-Oil Policymaking Phase (1996–1998) The Auto-Oil I study was ready
in the spring of 1996. Although the Commission had used target levels based on
WHO guidelines as yardsticks for assessing the acceptability of future levels of
pollution, it was still noted that ‘‘Article 4 of Directive 94/12/EC on Motor Vehicle
Air Pollution Control requires that measures to reduce emissions from road traffic
shall be designed to meet the requirements of the Community’s air quality criteria
and related objectives’’ (European Commission 1996, sec. 3).
In June 1996 the Commission then formally put forward the first two proposals
for directives arising from the Auto-Oil I Program. Central components in this initial
package of proposals consisted of tighter emission standards for passenger cars and
fuel specifications for petrol and diesel. First, by 2000 petrol was to contain no more
than 200 parts per million (ppm) of sulfur (down from the estimated market average
of 300 ppm), and diesel 350 ppm of sulfur (down from the estimated market aver-
age of 450 ppm). Second, a proposal to strengthen the existing emission limits for
passenger cars (these being based on Directive 70/220/EEC on Motor Vehicle Air
Pollution Control, last amended by Directive 94/12/EC) contained emission limits
for 2000 and lower ‘‘indicative’’ limits for 2005. These limits related to carbon
monoxide (CO), hydrocarbons (HC), NOX , and—only for diesel cars—particulates.
According to then Environment Commissioner Ritt Bjerregaard, the proposals were
‘‘extremely ambitious,’’ and she emphasized again that the proposed new standards
should help the EU to achieve WHO air-quality standards.17 The proposed mea-
sures were to be followed by an Auto-Oil II phase, specifying requirements for the
year 2005.
After several rounds of inputs from the European Parliament and the Council,
and a final round in a conciliation committee, somewhat tightened directives were
agreed to in June 1998. EU Council Directive 98/69/EC relating to passenger cars
and light commercial vehicles tightened existing emission limits in two stages (2000
and 2005). In EU Council Directive 98/70/EC relating to the quality of petrol and
diesel fuels, key targets for the year 2000 were petrol with 150 parts per million
(ppm) and diesel with 350 ppm of sulfur. For 2005, the petrol sulfur target was 50
ppm (see Haigh 2003, secs. 6.8 and 6.20, for more information about these direc-
tives). Since the conciliation agreement settled many of the 2005 standards intended
to be specified in an Auto-Oil II Program, this latter Program was redesigned as a

more analytic Program aiming at improved modeling and knowledge.
So what role did EU air-quality legislation in this field play in this phase of the
Auto-Oil I process? Air-quality considerations functioned as a central contextual
factor, while the specific links between EU air-quality legislation and the Auto-Oil I
legislation remained weak. Although the main controversies were fought over tech-
nological possibilities and the costs of different options, all sides in the debate con-
tinually referred to the implications of these possibilities and options for air quality
and health. The final agreement was hailed by members of the European Parlia-
ment, member states, and the Commission as a major breakthrough in the effort to
fight urban air pollution. For instance, according to Environment Commissioner
Ritt Bjerregaard, ‘‘We shall see cleaner air in our cities, and we shall have fewer
ozone episodes in the summer’’.18 However, EU air-quality legislation, including
the 1996 Framework Directive, was only referred to as a general parameter for
Auto-Oil legislation, while a specific influence is not detectable. As in the prepara-
tory phase, decision makers instead used WHO’s air-quality guidelines as central
benchmarks. This is hardly surprising because none of the daughter directives had
yet been adopted by June 1998 and hence Auto-Oil I decision makers had few re-
cent, specific EU air-quality standards available.
All in all, it is also hard to find clues indicating that the influence the other way
around from Auto-Oil I legislation on the preparation and negotiation of air-quality
legislation was very strong. The negotiation of the ground-level ozone directive was
closely related to the processes of producing an EU ozone strategy and a directive on
national emission ceilings (NECs), as further discussed elsewhere (Wettestad 2002).
Hence, interpolicy coordination was taking place, and it is reasonable to assume
that the relationship between Auto-Oil I and air-quality legislation was included as
a topic in the interpolicy coordination efforts that took place in the period from
1997 to 1999. This coordination, however, primarily involved processes other
than Auto-Oil I. Furthermore, as described by Amann and Lutz (2000), a central
analytic tool in the EU’s work on revising its ozone legislation was the RAINS
model developed by the International Institute for Applied Systems Analysis (IIASA)
in Laxenburg, Austria. Among the key inputs to this model was a summary of
emission-control policies within the transport sector regarding NOX and volatile or-
ganic compounds (VOCs), including recent policies adopted within the Auto-Oil I
context. Hence, this modeling work represented one way the Auto-Oil legislation
influenced air-quality legislation. However, given the multitude of relevant policies
to be included, this influence was rather indirect and probably not very strong.
Outcome Interaction: Auto-Oil Implementation Will Undoubtedly Help the

Achievement of Air-Quality Targets
We must first recall here that the Auto-Oil I directives were adopted relatively
recently. As noted above, revised fuel standards and vehicle-emission requirements
were adopted in June 1998. Moreover, for instance the United Kingdom did not
put the necessary fuel-standard implementing legislation into place until December
2001 (Haigh 2003, sec. 6.20, 4). This makes it hard to find clear effects of Auto-
Oil implementation on the achievement of the EU’s revised air-quality targets under
the FD.
Although it is not possible to find clear evidence so far of synergistic interaction
effects, it is clear that such synergistic effects will occur in the years ahead (figure
12.2). For instance, in 1996 the Commission estimated that emissions of NOX
from road traffic would decline by 65 percent by 2010 from 1995 levels, and those
of VOCs, carbon monoxide, and urban particulate matter by 70 percent. Between a
third and a half of these reductions would be attributable to the Auto-Oil Pro-
gram.19 Moreover, in the 2002 Environmental Signals report of the European Envi-
ronment Agency, transport measures from Auto-Oil I and II are one of the factors
expected to lead to reduced emissions of particulate matter (European Environment
Agency 2002, 77). The main available evidence is related to the implementation of
Auto-Oil I Directive 98/70/EC on the quality of petrol and diesel fuels. In April
2004, the first annual implementation report was published (European Commission
Figure 12.2
Auto-Oil legislation supports implementation of EU air-quality legislation
2004). Falling EU motor-fuel sulfur levels were reported. For instance, by 2002,
low-sulfur fuels had already attained a market share of 47 percent for petrol and
43 percent for diesel. The report noted that the specifications laid down in Directive
98/70/EC were generally met (European Commission 2004, 2). This development is
of course good news for European air quality, although it is hard to measure the
precise effect and the precise contribution of the Auto-Oil directive.
However, various worrisome trends and signals make it highly questionable
whether the synergy has been strong enough and the relationship between transport
and air quality has been handled optimally so far. For instance, in the Commission’s
review of Auto-Oil II, only modest reductions in exposure to particulate matter were
expected by 2010 (European Commission 2000, 11). Moreover, in September 2001,
on the basis of a preliminary assessment of the follow-up of the first air-quality
daughter directive, the Commission’s Environment Directorate issued clear warning
signals. Urban areas all over Europe would likely fail to meet the limits on NOX and
particulates set in the 1999 directive. A central reason pinpointed was the volume of
road transport growing out of control.20 This disturbing picture is reinforced in the
European Environment Agency’s Environmental Signals reports (European Environ-
ment Agency 2002, 2004). In these reports, the assessments of urban air-quality
exceedances for both ground-level ozone and particulates reveal unfavorable trends.
So, the time has come to take a closer look at the response of the EU system in the
form of the CAFÉ Program. How will the relationship between these activities and
strands of legislation be handled in the future?
Outlook: Improved Interaction through CAFÉ?

Two of the central, stated objectives of CAFÉ are to ensure that measures in differ-
ent sectors needed to achieve air-quality objectives are taken in a cost-effective man-
ner at the relevant policy level through the development of effective structural links
with the relevant policy areas; and to develop an overall integrated strategy to
achieve air-quality objectives in a cost-effective way.21 Particulate matter and ozone
are specifically targeted issues.
It makes sense to see CAFÉ as confirming inadequate EU integration of trans-
port and air-quality policies so far. On the one hand, the CAFÉ Communication
points out that considerable achievements have been made in terms of putting into
place various pieces of legislation to improve fuel standards and bring down emis-
sions from transport. Hence, the integration of vehicle and transport policies and
air-quality policies has not been a total failure. On the other hand, the two targeted
issues of particulate matter and ozone are closely linked to emissions from trans-
port. As mentioned in the previous section, various worrisome trends continue to
exist and serious air-quality problems remain. As a result, the Auto-Oil II Program
singled out particulate matter and ozone as remaining air-quality problems. So, de-
spite significant progress, there is still room for improvement.
In terms of addressing these integration challenges, CAFÉ has utilized several or-
ganizational devices.22 First, the program has been developed under the leadership
of a permanent secretariat housed within the Commission’s Environment Director-
ate, assisted by an interservice group composed of all relevant Commission depart-
ments. The mandate of this group, chaired by the Environment Directorate, is to
foster strategic discussion and consensus between the services and ensure that Com-
mission policy in this area is fully coordinated. Second, a steering group composed
of representatives of the member states, the European Parliament, stakeholders, and
relevant international organizations has met two or three times a year to advise the
Commission on the strategic direction of the program. This has meant the active
involvement of a broader range of actors than those involved in the Auto-Oil con-
text, especially Auto-Oil I.23 Third, in terms of more specific coordination between
sectors, a sectoral coordination group has been set up with the objective of ensuring
full communication between CAFÉ and sectoral policies, including the development
of source-based measures.
Are the measures announced in CAFÉ an adequate response? It is too early to
determine this conclusively. CAFÉ has been criticized for unclear financing of the
program,24 and for lacking specification of the work program that must be rapidly
established in order to meet the 2004 deadline.25 With regard to the coordination
efforts, it has been noted that ‘‘most if not all the key sectoral areas of policy making
required to build the strategy remain the responsibility of those same parts of the
Commission that ran them previously. They have not been brought into the grasp
of policy makers running the CAFÉ program, so it remains to be seen if the requisite
linkages can be made to work in practice as well as they may look in theory’’
(‘‘Commission Launches Clean Air for Europe Programme’’).26
A Summary and Interpretation of the Interaction

Although this study has not been able to reveal the complex coordination processes
within the Commission in detail, the distinct impression is one of a quite limited in-
teraction at the level of outputs. The Auto-Oil process and outputs mainly used
WHO guidelines as benchmarks. The air-quality process included Auto-Oil effects
in the modeling input, but more effort was seemingly made in coordinating with
other policies, such as the NEC Directive. The explanation may be simple: because
the Auto-Oil and air-quality directives inevitably interact when they are imple-
mented, the challenge for policymakers was much more to achieve the strongest out-
puts possible than to spend time and resources on coordination.
The strongest interaction between these two strands of legislation will take place
at the level of outcomes. In terms of interaction mechanisms, this is thus a case of
Behavioral Interaction. The Auto-Oil directives establish obligations of EU member
states to implement fuel standards and emission limits for vehicles. As the member
states implement these obligations, this is bound to lead to emission reductions that
improve air quality and thus support the implementation of the air-quality legisla-
tion (figure 12.2). The EU has estimated that without the Auto-Oil directives, the
emissions of NOX , carbon monoxide, and particulate matter from transport by
2010 would be almost twice as high. Hence, the implementation of these directives
and the reduced emissions will make it considerably easier for the EU to attain tar-
gets established in the air-quality daughter directives for concentrations of NOX and
the other pollutants mentioned. However, given the many processes both within and
outside the transport sector that contribute to air quality, it is difficult to be more
precise with regard to the specific synergistic effects of the Auto-Oil directives. This
interaction at the level of outcomes has been intentional, since a central aim for
the Auto-Oil directives was to improve air quality and hence the performance of
EU air-quality directives.
The overall management of the interaction so far must be characterized as moder-
ately positive. On the one hand, the Auto-Oil and air-quality directives mean that
the gap within the EU between transport activities and air-quality ambitions has
been narrowed. Therefore, a concern is now that sectors other than transport
are lagging behind in reducing emissions. On the other hand, the ‘‘transport-air-
quality’’ gap has in no way been closed, and the CAFÉ Program is launching
improved interinstitutional coordination as a central instrument to come up with
stricter policies to narrow the gap further in the years ahead.
Some Concluding Notes on the Role of Non-EU Expertise, Soft-Law Instruments,

and Timing
How much should EU environmental policy be a ‘‘home brew’’ and how much
should it rely on external expertise? With regard to the vertical, strongly synergistic
interaction with the WHO guidelines, it was noted that the overall response to the
interaction within the EU system has been positive, and the CAFÉ Program signals
a continuation of interaction practices. But in a policymaking climate where cost-
effectiveness is increasingly emphasized as a guiding principle for the work of the
Commission and the EU as a whole, an increasing interest in more low-cost, in-
house expertise was also noted. This forms the basis for a more general dilemma
for the EU, with far wider relevance than only the air-pollution context, and for a
possible lesson to be learned. On the one hand, the ability to point to highly author-
itative external expertise may provide EU actors with a sort of undisputable, ‘‘dis-
tant’’ authority. On the other hand, in line with the considerable development of
EU environmental policy, the EU has built up considerable in-house national and
EU-wide expertise. Relying on this expertise will often be cheaper than using exter-
nal expertise. However, the use of national expertise may also potentially be more
controversial if suspicions of a ‘‘politicization’’ of science arise. This case holds no
clear answer to this dilemma, but given the constructive role played by the WHO
guidelines in a number of EU processes, a shift toward more use of in-house exper-
tise should be carefully considered before a decision is made.
Another general lesson emanating from this vertical case has to do with the dis-
cussion of the role of nonbinding instruments in international environmental poli-
tics. Are they toothless paper tigers or are they flexible instruments for pushing up
environmental standards? This case offers further evidence that nonbinding instru-
ments can have a profound influence on the development of binding legislation
in international and European (environmental) politics. The crucial factor in this
case seems to be the high legitimacy surrounding the guideline-producing processes
within the WHO.
With regard to the horizontal, EU-internal interaction between the air-quality
directives and the stricter fuel standards and vehicle-emission limits being developed
under the banner of Auto-Oil I, there is clearly an interesting timing dimension
involved.27 When the Commission put forward the fuel-standard and emission-limit
proposals in June 1996, none of the air-quality daughter directives (which laid out
the more specific air-quality requirements) had been presented or adopted. It is
reasonable to assume that if they had been adopted prior to the fuel-standard and
vehicle-emission proposals, they could have functioned more as explicit reference
points in the debate on the need for stricter policies. However, in this specific con-
text, the lack of EU legislation was to a large extent compensated for by a direct ref-
erence to the WHO guidelines. But it may be more of an exception than a rule that
such external, authoritative reference points are available, and then the timing of in-
ternal EU policymaking processes becomes more critical.
Notes
The author would like to express special thanks to Andrew Farmer, IEEP, for valuable inputs
to the initial work on this chapter.
1. The top fourteen were sulfur dioxide, nitrogen oxides, ozone, particulates, black smoke,
lead, carbon monoxide, cadmium, acid deposition, toluene, benzene, benzopyrene, formalde-
hyde, and peroxyacetylnitrate (PAN) (European Commission 2001, 14).
2. Measurement is mandatory in so-called agglomerations—zones where the population
concentration exceeds 250,000 inhabitants—and in zones where pollution exceeds some pro-
portion of the limit values.
3. ‘‘Air Quality-Some Standards Ready,’’ Acid News 3, October 1998, 7; ‘‘Ministers Agree
Air Pollution Rules, Lengthy Deadline on Water Quality,’’ ENDS Report 281, June 1998, 47.
4. ‘‘Executive Finalizes EU Proposals on Emissions Ceilings,’’ Environment Watch, June 18,
1999, 7–9 (here: 7).
5. ‘‘Fourth EU Air Quality Directive Delayed,’’ ENDS Environment Daily, December 20,
2001.
6. ‘‘Commission Back-Pedals on Metals in Air Limits,’’ ENDS Environment Daily, October
30, 2002.
7. ‘‘Fourth EU Air Quality Directive Agreed,’’ ENDS Environment Daily, April 14, 2004;
‘‘Agreement on Air Quality Directive for Heavy Metals, PAHs,’’ ENDS Report 351, April
2004, 53–54.
8. The time-series approach takes a day as the unit of analysis and relates the daily occur-
rence of events such as deaths or admissions to hospital to daily average concentrations of
pollutants, while taking careful account of confounding factors such as season, temperature,
and day of the week. Powerful statistical techniques have been applied and coefficients have
been produced that relate the daily average concentrations of pollutants to their effects. For
more information, see http://www.who.int/environmental_information/Air/Guidelines.
9. Personal communication with Per Schwarze, Department Director within the Norwegian
Institute of Public Health and participant in the process of producing the revised WHO
guidelines.
10. Interview with Commission official, February 13, 2002.
11. See note 3.
12. ‘‘Council and MEPs Agree on Ozone in Ambient Air,’’ Europe Environment 601, De-
cember 4, 2001, 1.
13. Interviews with Commission official, February 13, 2002.
14. According to M. Krzyzanowski, WHO, the EU’s financial contribution to this revision
process is much more significant than the contribution in the mid-1990s. Personal communi-
cation, December 11, 2002.
15. Interview with Matthew Ferguson, Institute for European Environmental Policy, Novem-
ber 1999.
16. ‘‘Car Industry Lashes Out at Auto-Oil Proposals,’’ ENDS Report 257, June 1996, 41–43
(here: 41).
17. ‘‘Auto-Oil: Commission Proposes New Emission and Fuel Standards,’’ Europe Environ-
ment 480, June 27, 1996, 1.
18. ‘‘Legislation Setting Auto Emission, Fuel Quality Standards Approved by Union,’’ Inter-
national Environment Reporter, July 8, 1998, 671.
19. ‘‘Full Package Revealed,’’ Acid News 4: 1, October 1996, 3.
20. ‘‘EU Likely to Miss Air Quality Goals,’’ ENDS Environment Daily, September 24, 2001.
21. The other main objectives are to develop, collect, and validate scientific information con-
cerning air pollution, including projections, inventories, integrated assessment modeling, and
cost-effectiveness analysis studies, leading to the development of air-quality and deposition
objectives and indicators and identification of measures required to reduce emissions; to sup-
port the implementation of legislation and develop new legislation, especially the Air Quality
Framework Directive daughter directives, and contribute to the review of international proto-
cols; and to disseminate widely (including to the public) information and results from the pro-
gram (European Commission 2001).
22. This section relies heavily on information provided by the CAFÉ website. See http://
europa.eu.int/comm/environment/air/cafe.htm#Organisationalstructure.
23. This can be seen as a response to and balancing of Auto-Oil’s heavy industry involve-
ment. Interview with Commission official, February 13, 2002.
24. ‘‘Editorial: Should Have Learnt,’’ Acid News 2, June 2001, 2.
25. ‘‘Commission Launches Clean Air for Europe Programme,’’ Environment Watch, May
11, 2001, 5–6.
26. See note 25 (here: 6).
27. See Wettestad 2002 for a discussion of how unfortunate timing with other processes
influenced the process of developing a National Emission Ceilings Directive.
References
Ammann, Markus, and Martin Lutz. 2000. The Revision of the Air Quality Legislation in the
European Union Related to Ground-Level Ozone. Working paper. Laxenburg, Austria, and
Brussels: International Institute for Applied Systems Analysis (IIASA)/European Commission,
DG-ENV.
Andresen, Steinar, Arild Underdal, Tora Skodvin, and Jørgen Wettestad. 2000. Science and
Politics in International Environmental Regimes: Between Integrity and Involvement. Man-
chester: Manchester University Press.
Elsom, Derek M. 1999. Development and Implementation of Strategic Frameworks for Air
Quality Management in the UK and the European Community. Journal of Environmental
Planning and Management 42 (1): 103–121.
European Commission. 1993. Towards Sustainability—A European Community Programme

of Policy and Action in Relation to the Environment and Sustainable Development. Brussels:
Commission of the European Communities.
European Commission. 1996. Communication from the Commission to the European Parlia-
ment and the Council on a Future Strategy for the Control of Atmospheric Emissions from
Road Transport Taking into Account the Results from the Auto/Oil Programme. Doc.
COM (1996) 248, June 18, 1996. Brussels: European Commission.
European Commission. 1997. Explanatory Memorandum to the Proposed First Daughter
Directive. Doc. COM(1997) 500, October 8, 1997. Brussels: European Commission.
European Commission. 2000. Communication from the Commission: A Review of the Auto-
Oil II Programme. Doc. COM(2000) 626, October 5, 2000. Brussels: European Commission.
European Commission. 2001. The Clean Air for Europe (CAFÉ) Programme: Towards a
Thematic Strategy for Air Quality. Communication from the Commission. Doc. COM
(2001) 245, May 4, 2001. Brussels: European Commission.
European Commission. 2004. Quality of Gasoline and Diesel Fuel Used for Road Transport
in the European Union: First Annual Report. Doc. COM (2004) 310, April 27, 2004. Brus-
sels: European Commission.
European Environment Agency. 2002. Environmental Signals. Copenhagen: EEA.
European Environment Agency. 2004. EEA Signals 2004. Copenhagen: EEA.
Friedrich, Axel, Matthias Tappe, and Rüdiger Wurzel. 1998. The Auto-Oil Programme:
Missed Opportunity or Leap Forward? Research paper 1/98. Hull: University of Hull.
Goodwin, Frazer. 1999. Controlling Traffic Pollution and the Auto Oil Programme. Paper
99/8. Brussels: European Federation for Transport and Environment.
Weale, Albert, Geoffrey Pridham, Michelle Cini, Dimitrios Konstadakopulos, Martin Porter,
and Brendan Flynn. 2000. Environmental Governance in Europe. Oxford: Oxford University
Press.
Wettestad, Jørgen. 2002. Clearing the Air—European Advances in Tackling Acid Rain and
Atmospheric Pollution. Aldershot: Ashgate.
Wettestad, Jørgen, and Andrew Farmer. 2001. The EU Air Quality Framework Directive: EU
Pitcher; International Catcher. Inventory paper. Lysaker, Norway, and London: Fridtjof
Nansen Institute and Institute for European Environmental Policy.
WHO. 2000. WHO Air Quality Guidelines. 2nd ed. WHO Regional Publications, European
Series 91. Geneva: World Health Organization.
Young, Alasdair R., and Helen Wallace. 2000. Regulatory Politics in the Enlarging European
Union: Weighing Civic and Producer Interests. Manchester: Manchester University Press.
13
Comparative Empirical Analysis and Ideal Types of
The empirical chapters of this volume demonstrate that interaction between interna-
tional and EU environmental institutions matters for the development and the per-
formance of these institutions. The eleven core institutions included in our study
(i.e., six environmentally relevant international institutions and five environmental
EU directives) have significantly influenced the functioning and effectiveness of other
governance instruments, and have themselves been the subject of similar influence.
Policymakers are therefore well advised to take into account interaction effects
when designing international agreements and EU legal instruments. Analysts who
want to understand the development and performance of institutions will have to
take into account the network of international and EU institutions in which these
institutions are embedded.
The wealth of empirical information gathered in this study provides a unique
opportunity to push the exploration of institutional interaction a step further. The
inventories contained in chapters 3–12 comprise more than 160 cases of interaction
in which (at least) one of our eleven core institutions has been involved, either as the
source of influence or as its target. Cases were identified independently from their
political salience, from the degree of intentionality of interested actors, and from
their effects on the target. This sample of cases not only provides, for the first time,
a broader picture of the realm of institutional interaction; it also allows us to ex-
plore systematic patterns of institutional interaction in a comparative perspective.
To this end, this chapter proceeds in three steps. First, we examine the overall
contours of the phenomenon of institutional interaction as reflected in our sample
of cases. Cases are analyzed according to some variables that we believe to be im-
portant for advancing the understanding of institutional interaction. In this effort,
the effect of a case of interaction on the target institution constitutes the key de-
pendent variable. While our methodology of case selection does not ensure the
generalizability of our results, the empirical analysis produces a number of interest-

ing findings that may easily be tested against other samples of cases from environ-
mental affairs and beyond. Second, we develop a number of Weberian ideal types
of institutional interaction, which subdivide and specify the general causal mecha-
nisms developed in chapter 2, and elaborate their distinctive features. In doing so,
we develop a framework for a more sophisticated analysis of institutional interac-
tion so as to be able to better understand and explain the strikingly different proper-
ties of cases of interaction driven by the same causal mechanism. Cases of Cognitive
Interaction follow one of two distinct ideal types, whereas Interaction through Com-
mitment occurs in three variants. In contrast, it proved difficult to identify ideal
types of Behavioral Interaction. Since their inherent logics are deductively generated,
the Weberian ideal types of institutional interaction are independent from our em-
pirical sample. Finally, we examine certain more complex phenomena of institu-
tional interaction that reach beyond the occurrence of cases, and look ahead to the
future of the study of institutional interaction.
A Sample of 163 Cases of Institutional Interaction
In this section, we engage in a comparative analysis of institutional interaction based

on a database that contains the core characteristics of all cases of our sample. Data
gathering involved three steps. First, we selected eleven core institutions so as to en-
sure that relevant patterns of interaction are not accidentally excluded, and that the
core institutions address the full spectrum of international and EU environmental
governance and include well-known cases of institutional interaction. No relevant
area of governance is systematically excluded. Moreover, the institutions were to
be important in order to support the relevance of our findings. Second, case-study
authors identified the important cases of institutional interaction in which these
institutions have been involved, either as the source or as the target, and that have
a significant impact on the effectiveness of governance. They focused on cases with
shorter rather than longer causal chains to keep the analytic effort within manage-
able limits. The inventories that are part of chapters 3–12 of this volume contain a
total of 163 cases of institutional interaction. Third, case-study authors coded the
cases of our sample according to a number of variables assumed to be important
for institutional interaction. These variables refer to dimensions along which we
may expect actual cases of interaction to vary. The resulting data are reflected in
the database included in the appendix to this book.
Empirical Analysis and Ideal Types 309
In the following, we introduce the core characteristics of cases of institutional in-

teraction on which the empirical analysis of this chapter is founded. On this basis,
we subsequently examine the universe of the case-specific data comparatively and
identify a number of interesting findings about patterns of interaction in our sample.
Core Characteristics of a Case of Interaction

For the systematic analysis of institutional interaction, we need variables that allow
us to characterize and distinguish interaction cases in meaningful ways. Human
beings generally recognize and structure their living environment through self-
constructed distinctions, internal models, ‘‘paradigms,’’ or ‘‘theories.’’ For example,
‘‘cars’’ remains an amorphous category, unless we introduce further distinctions, for
example between vans, limousines, and station wagons.
In this subsection, we introduce six important variables that may serve to distin-
guish and characterize cases of institutional interaction. These variables are not
necessarily the only ones that could be explored fruitfully. However, they illuminate
core characteristics of cases of interaction, refer to categories well established in
standard institutional analysis, and embody critical factors relevant to the causal
pathways of institutional interaction (Gehring and Oberthür 2004). The variables
and their range of potential values are listed in table 13.1.
Table 13.1
Variables of institutional interaction and relevant distinctions
Variable Relevant Distinctions
Quality of effect (within the target institution) Synergy
Disruption
Neutral or unclear
Policy fields (of source and target institution) Same policy field
Different policy fields
Intentionality (of the triggering action of the Intentional
source institution) Unintentional
Key differences Objectives of source and target
Memberships of source and target
Means of source and target
Policy responses Collective response
No collective response
Potential for further improvement Significant potential
No significant potential
In line with our overall interest in the effectiveness of governance, the effects of,
and policy responses to, institutional interaction are in the center of our interest
as dependent variables, while the other core characteristics serve as independent
variables. As an offspring of research on the effectiveness of international and EU
institutions, the study of institutional interaction is particularly interested in the con-
sequences of interaction for the effectiveness of governance and in its political man-
agement. We want to know which kinds of interaction can be expected to produce
beneficial or detrimental effects on governance and how these effects may be influ-
enced through targeted policy responses. In introducing the variables, we point out
what we may learn from investigating them empirically and derive hypotheses about
the empirical results we may expect to find, focusing in particular on the effective-
ness of governance.
Quality of Effect: Synergy or Disruption The objective of the target institution rep-
resents the major yardstick for assessing the consequences of a case of institutional
interaction. An institution’s objective indicates the direction of collectively desired
change, or the aim of maintaining a desired status quo against some collectively
undesired change (Gehring 1994, 433–449). Every institution established for pur-
poses of governance has such an objective, even though some of its members might
not fully support its active pursuit. The objective of a given international institution
has also been the major yardstick for assessing its consequences in the established
research on the effectiveness of international institutions. The effects of a case of
institutional interaction are generally felt within the target institution, be it in its
decision-making process or in the issue area governed by it (chapter 2).
The effects of a case of institutional interaction may be synergistic, disruptive, or
neutral/unclear for the target institution. If the effects support the objectives of the
target institution, they create synergy between the two institutions involved. If they
contradict the target’s objective, they result in disruption. The effects of an interac-
tion may also be indeterminate or neutral, if they do not clearly hamper or reinforce
the target institution’s pursuit of its objective (chapter 2). The empirical chapters of
this volume provide various illustrations of the differing quality of effects.
Perhaps the most fundamental empirical issue concerning the effects of institu-
tional interaction relates to the question of whether synergy or disruption dominates
the realm of institutional interaction. Given that the existing literature on institu-
tional interaction has put particular emphasis on problematic interaction (chapters
1 and 2), conventional wisdom may expect disruption to prevail. However, policy-
makers may also be expected to strive to avoid disruption to the extent possible.
Moreover, it will be interesting to know whether the distribution of synergy and dis-
ruption varies between our general causal mechanisms.
Policy Fields Interaction may take place within a single policy field or between
institutions belonging to different policy fields. For example, an environmental insti-
tution may interact either with another environmental institution or with an institu-
tion belonging to the field of transportation or agriculture. In the policy debate on
the increasing potential for interinstitutional conflict and the growing need for inter-
institutional coordination, interaction between institutions belonging to different
policy fields plays a particularly prominent role—as witnessed by the debate on the
relationship between the World Trade Organization (WTO) and multilateral envi-
ronmental agreements (chapter 8).
The policy field to which an institution belongs is determined by ministerial com-
petences within countries. Our assignment of an institution to a policy field does not
depend on our own views or on the perspectives of scientific observers. Whether two
international and/or EU institutions belong to the same or different policy fields
depends on whether the same or different ministries are generally responsible for
representing state interests in these institutions. Hence, the 1995 UN Fish Stocks
Agreement aiming at the sustainable use of the world’s fish resources and their pre-
cautionary protection may be characterized as an environmental institution by some
scientific observers, while the EU Common Fisheries Policy might be considered eco-
nomic in nature because of the prevalence of economic, or regional, considerations
(chapter 6). However, in most countries both institutions belong to the portfolio of
fisheries ministers, so that they are assigned to the same policy field despite their
diverging objectives.
We may hypothesize that institutions from different policy fields are likely to dis-
rupt each other, while institutions from the same policy field may tend to create syn-
ergy. Ministries may be more inclined to avoid inconsistencies between institutions
belonging to their portfolio than with respect to other institutions, the performance
of which they will not be held accountable for. Environment ministers may, for
example, be expected to be less inclined to avoid inconsistencies with economic
institutions. We may also expect that policy responses to institutional interaction
within one policy field will tend to be more effective than with respect to interaction
between institutions in different policy fields. This would correlate with the view
that current interaction problems, to a large extent, reflect problems of policy con-
sistency and coordination at the national level (Victor 1999).
Intentionality Actors within the source institution may trigger a particular instance
of institutional interaction intentionally or not. For example, the EU Directive on
Integrated Pollution Prevention Control (IPPC Directive) was adopted in order to
improve the performance of a number of other EU environmental instruments
(chapter 9). In contrast, trade rules agreed within WTO were not intended to inter-
fere with a number of environmental regimes, but to liberalize international trade
(chapter 8).
For our purposes, intentionality refers to the action within the source institution
that triggers a case of institutional interaction. It does not refer to the target institu-
tion because it is not an appropriate variable to characterize effects. Only in the case
of Cognitive Interaction are actors within the target institution completely free to
decide on the interaction. In contrast, the effects of Behavioral Interaction are felt
within the target institution irrespective of the intentions and wishes of its members.
In the case of Interaction through Commitment, members of the target institution
have to decide under pressure.
Intentionality must be kept separate from anticipation (Martin and Simmons
1998). While all unanticipated cases of interaction are also unintended, not all
anticipated cases are intended. If interaction effects are known to the actors within
the source institution, they may still occur as unintentional side effects of the pursuit
of some other objective. If the effect is disruption, actors may choose not to avoid
unintended but anticipated interaction because they consider the costs of doing so
higher than the benefits. In contrast, the effects of unanticipated interaction come
about as a surprise, although it might have been possible to anticipate them on
closer inspection.
We may expect intentional disruption to occur rarely in the realm of institutional
interaction involving environmental institutions. It is difficult to imagine environ-
mental ministers intentionally making decisions in one forum that are at odds with
rules of another environmental institution. And even across the borders of different
policy fields, intentional disruption might be the exception rather than the rule
because it would undermine the functioning of the system of international and
EU institutions. Also, policy coordination within governments may be assumed to
work against the making of conflicting decisions in international and EU forums.
We may expect that intentional disruption will occur particularly rarely within the
EU because of the comparatively sophisticated institutional framework.
Key Differences There must be situation-specific drivers that provide a basis for
interinstitutional influence to occur. Such problem-specific drivers constitute an im-
portant part of what is occasionally referred to as problem structure (Miles et al.
2002; Hasenclever, Mayer, and Rittberger 1997). We seek to identify them in the
form of key differences between institutions. This effort is based on the following
consideration: If two institutions were very similar or even identical in all important
respects, it would be difficult to see how one of them could exert significant influ-
ence on the other. If two institutions were so different in all relevant aspects that
they did not touch on each other’s areas of activity, interaction would be improba-
ble. Therefore, interacting institutions will differ in some important respects, while
being similar in others.
We suggest that for interaction to occur two institutions must differ significantly
with respect to at least one of three key factors, namely their objectives, their mem-
berships, or their means of governance. Institutions may interact because they pur-
sue different objectives, while memberships and means of governance are similar. All
specific international and EU institutions are focusing on a limited issue area and a
limited set of objectives. States establish such issue-area-specific institutions to
enable them to focus on some relevant issues at one time instead of being over-
whelmed by complexity. From negotiation analysis, we know that actors will adjust
their preferences according to the delimitation of issue areas. Modification of issues
to be negotiated will influence the constellation of interests (Sebenius 1983, 1992).
Hence, an identical group of states may pursue different interests in institutions
governing different issue areas and may thus maintain institutions with starkly
diverging objectives. Furthermore, memberships of two institutions can be so clearly
distinct from each other that we may expect their interaction to rely on this varia-
tion. From negotiation analysis, we know that adding or subtracting actors will
change the constellation of interests. Hence, two institutions with similar objectives
and means may develop differently because their memberships vary significantly,
and their interaction may be a result of the difference in memberships. Finally, two
institutions may have largely similar objectives and identical memberships, while the
means employed to realize their objectives differ. Some institutions operate with
soft-law instruments, while others employ hard law (Abbott and Snidal 2000).
Some institutions have powerful sanctions at their disposal, while others comprise
financial mechanisms or lack sanctioning power. Hence, interaction may matter

even if the institutions involved are established and maintained by the same group
of actors for the same purpose.
The variable of key differences seems theoretically very promising, but it proved
difficult to codify. Codification difficulties originated from the fact that interacting
institutions frequently differ in two, occasionally even all three respects. Identifying
the most significant key difference in a given case is far from trivial. As a result, the
classification of some cases proved contentious, and for several cases it was not pos-
sible to clearly determine one key difference. For this reason, we do not analyze the
data of this category quantitatively. However, we found the three key differences
particularly useful as empirical points of reference for the theoretical develop-
ment of ideal types of Interaction through Commitment (see the next section). The
appendix contains data of this category only for cases of Interaction through
Commitment.
Policy Responses and Potential for Further Improvement From the perspective of
international and EU governance, we are interested in collective policy responses
in the framework of the interacting institutions. A case of interaction may be
responded to either within the source institution or within the target institution, or
within both institutions (or even within a third institution) in order to mitigate
disruption or to enhance synergy. Such ‘‘interplay management’’ (Stokke 2001;
also Stokke 2000) involves the ‘‘politics of institutional design and management’’
(Young 2002, 23). One of the institutions involved may take appropriate decisions
separately. A policy response will be coordinated, if it results from a communication
process overarching the two institutions, for example in the form of an exchange of
the relevant secretariats or of negotiations between the two groups of actors. In this
case, the interacting institutions create a ‘‘political linkage’’ that arises ‘‘when actors
decide to consider two or more arrangements as parts of a larger institutional com-
plex’’ (Young et al. 1999, 50).
Response action must be carefully distinguished from the original interaction
effect in the target institution that constitutes an essential element of every case of
institutional interaction. Response action is always additional and not a necessary
element of the interaction. It may or may not occur depending on whether relevant
actors recognize a case of interaction, how they assess the benefits of a response,
whether a suitable response is available, and so on. The original interaction effect
is particularly easily confused with response action in the case of output-level inter-
action, because both may consist of collective decision making within the target in-
stitution. For example, air-quality standards adopted by EU member states under
the Air Quality Framework Directive were heavily influenced by relevant guidelines
of the World Health Organization (WHO). The decision to rely on the guidelines
constituted the main effect of the interaction. In contrast, the subsequent decision
of the European Commission to provide funding for the revision of the WHO guide-
lines was a policy response. It did not form part of the original causal pathway and
was not a direct effect of the WHO guidelines (chapter 12).
The examination of policy responses is particularly interesting from a governance
perspective because policy responses enable actors to react to institutional interac-
tion and influence its effects. There are many aspects of policy responses that deserve
scientific attention. For example, few approaches to classifying and systematizing
responses to institutional interaction exist (see King 1997, 18, 23). We may, for in-
stance distinguish between collective decision making by the target institution and
by the source institution, coordinated decision making by the two institutions,
administrative coordination between their secretariats, and an exchange of informa-
tion. On the basis of an appropriate categorization, it might also be worth investi-
gating in more detail the level of success and effectiveness of different (kinds of)
policy responses in varying interaction situations. However, these questions are be-
yond the scope of our study. We limit ourselves to investigating (1) whether or not
collective policy responses have occurred and (2) whether or not a significant poten-
tial for further improvement through collective policy responses continues to exist.
The assessment of the potential for further improvement relies on the subjective
expert judgment of the case-study authors. Data reflect the situation in 2001 and
2002 that may have changed as a result of subsequent developments. Data do not
reflect cases of merely marginal potential for improvement. In cases of Behavioral
Interaction, we took the core obligations of the source institution as a given that
could not be altered in order to modify the effect on the target. The disruptive influ-
ence of the Convention on International Trade in Endangered Species of Wild Fauna
and Flora (CITES) on the WTO (chapter 7), for example, could in principle be
ended if CITES stopped restricting international trade. However, this would in effect
mean dissolving CITES.
We may expect that collective policy responses occur more frequently in cases of
disruptive effects than in cases producing synergy. Experimental psychology has
found that people react more strongly to the risk of losses (conflict) than to the
promise of additional benefits (Tversky and Kahnemann 1981, 1984). Moreover,
disruptive interaction creates aggrieved actors who will struggle to remedy the
situation, while synergistic interaction produces merely winners, even if the opportu-
nities for cooperation gains are not fully exploited. Accordingly, actors in both insti-
tutions involved may tend to consume the additional benefit without engaging in
further efforts, while disruptive effects will attract their attention more easily.
With respect to the potential for further improvement, it may be particularly in-
teresting to explore whether or not such a potential exists more frequently in cases
that have or have not led to a collective response yet. Given our expectation that in
particular a number of synergistic cases may not have been responded to at all, we
may on the one hand expect a potential for further improvement in particular in
cases that have not drawn a collective response yet. On the other hand, existing
responses may have easily left unexploited the potential for improvement.
Key Insights from the Analysis of 163 Cases of Institutional Interaction

In the following pages, we present key findings from the empirical analysis of the
163 cases of institutional interaction of our sample. After providing a brief overview
of the sample, we present and analyze key findings in three areas: synergy and dis-
ruption, policy responses, and potential for future improvement. Given the method-
ology of case selection, we are confident that our results at least approximate and
roughly reflect the overall situation in international and EU environmental gover-
nance, although the figures may be expected to differ to some extent for other sam-
ples drawn from this area. However, our insights are evidently tied to our specific
sample. We caution that the sample is not statistically representative and therefore
does not allow for the generalization of insights to other populations of cases. In
particular, we do not claim that the empirical results hold for interaction phenom-
ena beyond international and EU environmental governance.
The Universe of Cases of Our Sample Table 13.2 provides aggregate information
on our dataset in total and according to the three kinds of horizontal and vertical
interaction. The cases of the sample are roughly evenly distributed across the three
dimensions of interaction: our sample contains fifty-eight cases of horizontal in-
teraction between international institutions, forty-nine cases of horizontal interac-
tion between EU legal instruments, and fifty-six cases of vertical interaction between
international and EU instruments.
As can be seen in table 13.2, more than two-thirds of our cases of institutional
interaction occurred within one policy field, namely environmental protection. In
Table 13.2
The sample of cases of institutional interaction
Horizontal Horizontal
international EU Vertical Total
Variables No. % No. % No. % No. %
Totals 58 100 49 100 56 100 163 100
Synergy 27 46.6 34 69.4 41 73.2 102 62.6
Disruption 19 32.8 9 18.4 13 23.2 41 25.2
Neutral/unclear 12 20.7 6 12.2 2 3.6 20 12.3
Same policy field 27 46.6 38 77.6 45 80.4 110 67.5
Different policy fields 31 53.4 11 22.4 11 19.6 53 32.5
Intentional 18 31.0 18 36.7 25 44.6 61 37.4
Unintentional 39 67.2 28 57.1 30 53.6 97 59.5
Uncertain 1 1.7 3 6.1 1 1.8 5 3.1
Response 33 56.9 14 28.6 15 26.8 62 38.0
No response 25 43.1 35 71.4 41 73.2 101 62.0
No further potential 14 24.1 12 24.5 20 35.7 46 28.2
Further potential 21 36.2 13 26.5 19 33.9 53 32.5
Unclear 23 39.7 24 49.0 17 30.4 64 39.3
Source: Own compilation from database reflected in the appendix.
slightly less than one-third of the cases, the interaction crossed the border sepa-
rating two policy fields. Interaction between institutions belonging to different pol-
icy fields is more frequent at the international level than both at the EU level and
vertically.
Although the majority of cases occurred unintentionally, policymakers in the
source institution collectively employed institutional interaction as a means to pur-
sue their objectives in a significant share of cases. Nearly two-fifths of all identified
cases of interaction were triggered intentionally by the source institution. Intentional
interaction was somewhat more common at the EU level than internationally,
which may reflect the comparatively integrated framework of EU policymaking.
Vertical interaction was even more frequently triggered intentionally, partly because
implementation of international commitments at the EU level has been coded as
intentional throughout. In any event, policymakers in international and EU environ-
mental governance already exploit opportunities arising from institutional interac-
tion at least partially.
Data on the quality of effect, policy responses, and the potential for further im-
provement related to the aggregate of our sample are discussed in more detail in
the next subsection. More detailed information about the distribution of our sample
across the general causal mechanisms is presented in the context of the discussion of
ideal types of institutional interaction later in this chapter.
Quality of Effects: Synergy Dominates, but Disruption Occurs Institutional inter-

action was found to lead to synergy with the target institution and thus improve
the effectiveness of international and European governance in more than 60 percent
of our cases. Only about one-quarter of the cases identified resulted in clear disrup-
tion. The balance is made up of twenty cases in which the effect was neutral or in
which scientific uncertainty or uncertainty about the relationship between the effect
and the target institution’s objective prevented a clear-cut assessment. Disruption
appears to be somewhat more frequent at the international level (nearly one-third).
However, synergistic effects dominate at all three levels of interaction, namely inter-
action between international institutions, between EU legal instruments, and verti-
cally between international and EU institutions (table 13.2).
This finding suggests that institutional interaction may not primarily be a bad
thing that ought to be diminished as much as possible. It contrasts with conven-
tional wisdom, as reflected in the relevant literature that has so far emphasized the
problems rather than the opportunities arising from the interaction of interna-
tional and EU legal instruments. Higher political salience of disruptive cases may
be explained by the fact that people generally react more strongly to the risk of
losses (conflict) than to the promise of additional benefits (Tversky and Kahnemann
1981, 1984). The predominance of synergistic interaction also casts doubt on the
widespread belief that institutional interaction is a phenomenon of institutional
overflow (‘‘treaty congestion,’’ Brown Weiss 1993), which threatens to lead to
dysfunctional duplication of work and conflict in international and EU environ-
mental affairs. Our sample suggests that institutional fragmentation may also
provide a valuable asset for skilful policymaking to enhance environmental gover-
nance. Policies to minimize allegedly undesirably interaction could risk sacrificing
this asset.
Whereas our finding calls to mind that synergy must not be neglected, it does not
suggest that synergy is necessarily more important than disruption, or that every-
thing is all right with international and EU environmental governance. Although
we do not have supporting evidence, we cannot exclude that the aggregate effects
of the less numerous disruptive cases weigh more heavily than the aggregate effects
of the more numerous synergistic cases. To calculate the net effects of institutional
interaction, one would have to aggregate cases and assign a particular weight to
every single case (Underdal 2004; Gehring 2004), which is beyond the scope of
our analysis. There may also be more scope to reduce disruption than to enhance
synergy.
The distribution of synergy and disruption across our general causal mechanisms
provides few additional insights. For all three causal mechanisms, synergy by far
outweighs disruption. It is noteworthy that disruption appears to be comparatively
rare and neutral/unclear effects rather frequent in cases of Cognitive Interaction.
Overall, however, there is no strong correlation between the quality of effect and
the causal mechanisms of institutional interaction. Accordingly, Cramer’s V de-
scribing the strength of the correlation between the two nominal variables as a value
between 0 (no correlation) and 1 (complete correlation) is 0.255 and thus compara-
tively low (table 13.3).
Distinguishing interaction staying within the same policy field from interaction
involving different policy fields is more revealing (table 13.4). As expected, synergy
is much more frequent in the former case, accounting for more than 80 percent of
the cases of interaction within the same policy field. In contrast, a majority of about
57 percent of the cases of institutional interaction transcending the borders of a pol-
icy field have resulted in disruption. Also, 30 of the 41 cases of disruption occurred
between institutions located in different policy fields. In contrast, about 9 out of 10
synergistic cases (91 of 102) were found to be within one policy field. With some
Table 13.3
Quality of effect and causal mechanisms
Interaction
Cognitive through Behavioral
interaction commitment interaction Total
No. % No. % No. % No. %
Synergy 10 62.5 36 54.5 56 69.1 102 62.6
Disruption 1 6.3 16 24.2 24 29.6 41 25.2
Neutral/unclear 5 31.3 14 21.2 1 1.2 20 12.3
Total 16 100 66 100 81 100 163 100
Cramer’s V 0.255

Table 13.4
320
Quality of effect, policy fields, and intentionality

Horizontal international Horizontal EU Vertical Total
Different Different Different Different
Same PF PFs Same PF PFs Same PF PFs Same PF PFs
Quality of effect No. % No. % No. % No. % No. % No. % No. % No. %
Synergy 20 74.1 7 22.6 31 81.6 3 27.3 40 88.9 1 9.1 91 82.7 11 20.8
Disruption 4 14.8 15 48.4 2 5.3 7 63.6 5 11.1 8 72.7 11 10.0 30 56.6
Neutral/unclear 3 11.1 9 29.0 5 13.2 1 9.1 0 0.0 2 18.2 8 7.3 12 22.6
Total 27 100 31 100 38 100 11 100 45 100 11 100 110 100 53 100
Cramer’s V 0.516 0.632 0.736 0.605
Inten- Uninten- Inten- Uninten- Inten- Uninten- Inten- Uninten-
tional tional tional tional tional tional tional tional
No. % No. % No. % No. % No. % No. % No. % No. %

Synergy 11 61.1 15 38.5 15 83.3 17 60.7 21 84.0 20 66.7 47 77.0 52 53.6
Disruption 0 0.0 19 48.7 2 11.1 6 21.4 3 12.0 10 33.3 5 8.2 35 36.1
Neutral/unclear 7 38.9 5 12.8 1 5.6 5 17.9 1 4.0 0 0.0 9 14.8 10 10.3
Total 18 100 39 100 18 100 28 100 25 100 30 100 61 100 97 100
Cramer’s V 0.498 0.244 0.282 0.312
Source: Own compilation from database reflected in the appendix; assessment of intentionality excludes five cases in which inten-
tionality remained uncertain; PF ¼ policy field.
variation, the picture is similar for all three dimensions of interaction. Whether an
interaction occurs between institutions within the same or across different policy
fields is far more relevant for the quality of the effect. The value for Cramer’s V of
more than 0.5 for all dimensions provides a strong indication of the relevance of the
correlation (table 13.4).
Also confirming our expectations, disruption has in most cases occurred as an
unintended side effect. According to our sample, policymakers in international and
EU environmental governance apparently employ institutional interaction primarily
to create synergy. Intentional disruption appears to be the exception rather than the
rule. Only 1 out of 8 disruptive cases was triggered intentionally by the source insti-
tution (5 of 40) as compared with nearly half of the synergistic cases (47 of 99).
More than three-fourths of the cases of intentional interaction resulted in synergy,
while this was only the case for less than 55 percent of the unintentional cases (dis-
ruption: less than 10 percent and more than 35 percent, respectively). Although with
a Cramer’s V of 0.312 the correlation of the quality of effect with intentionality
is noticeable, it is less relevant than the correlation with the policy-field variable.
Whereas there is hardly any relevant correlation for horizontal interaction between
EU instruments and for vertical interaction, Cramer’s V of 0.498 indicates a com-
paratively strong correlation between quality of effect and intentionality at the
international level, which may be traced back to a comparatively high number of
disruptive cases that are exclusively unintentional (table 13.4).
Our figures do not confirm the expectation that intentionally created disruption is
particularly rare at the EU level (due to the comparatively integrated legislative pro-
cess). While the limited number of cases does not allow us to draw any firm conclu-
sions, it may be observed that both cases of intentional disruption between EU legal
instruments contained in our sample followed the causal mechanism of Behavioral
Interaction. A preliminary explanation for their occurrence is offered in the section
on Behavioral Interaction in this chapter.
Policy Responses: Frequent in Case of Disruption, Rare in Case of Synergy Collec-

tive policy responses were identified in less than 40 percent of the cases of interac-
tion contained in our sample (table 13.2). While this share may appear low, it needs
to be taken into account that more than 60 percent of our cases had a synergistic
effect. In line with our expectations, more than 80 percent of these synergistic cases
were not responded to at all but were ‘‘consumed’’ without further action (table 13.5).
At all three levels of interaction, response action focused on cases of disruptive
Table 13.5
322
Policy responses, quality of effect, and potential for further improvement

Horizontal international Horizontal EU Vertical Total
Disrup- Neutral/ Disrup- Neutral/ Disrup- Neutral/ Disrup- Neutral/

Synergy tion unclear Synergy tion unclear Synergy tion unclear Synergy tion unclear
Policy
response No. % No. % No. % No. % No. % No. % No. % No. % No. % No. % No. % No. %
Response 9 33.3 16 84.2 8 66.7 3 8.8 9 100 2 33.3 5 12.2 9 69.2 1 50.0 17 16.7 34 82.9 11 55.0
No response 18 66.7 3 15.8 4 33.3 31 91.2 0 0.0 4 66.7 36 87.8 4 30.8 1 50.0 85 83.3 7 17.1 9 45.0
Total 27 100 19 100 12 100 34 100 9 100 6 100 41 100 13 100 2 100 102 100 41 100 20 100
Cramer’s V 0.462 0.770 0.550 0.593
No No No No
Further further Further further Further further Further further
potential potential Unclear potential potential Unclear potential potential Unclear potential potential Unclear
No. % No. % No. % No. % No. % No. % No. % No. % No. % No. % No. % No. %
Response 16 76.2 5 35.7 12 52.2 8 61.5 0 0.0 6 25.0 7 36.8 1 5.0 7 41.2 31 58.5 6 13.0 25 39.1
No response 5 23.8 9 64.3 11 47.8 5 38.5 12 100 18 75.0 12 63.2 19 95.0 10 58.8 22 41.5 40 87.0 39 60.9
Total 21 100 14 100 23 100 13 100 12 100 24 100 19 100 20 100 17 100 53 100 46 100 64 100
Cramer’s V 0.321 0.492 0.369 0.364
Source: Own compilation from database reflected in the appendix; calculation of Cramer’s V for further potential does not take into account
unclear cases.
effects that accounted for a majority of the identified responses (34 of 62). At the EU
level, all disruptive cases were responded to in one way or the other. At all levels,
there is a strong correlation between the variables of quality of effect and policy
response (Cramer’s V for total: 0.593).
Response action occurred in about 57 percent of the cases located at the interna-
tional level and thus far more frequently than at the EU level and with respect to the
vertical cases (both between 25 and 30 percent; table 13.2). This imbalance can be
traced back to two factors. First, disruptive cases that are generally responded to
more frequently are more abundant at the international level. Second, synergistic
cases appear more likely to be responded to internationally (more than 30 percent)
than at the other levels of interaction (table 13.5). We may speculate that this latter
difference may be accounted for by the greater prominence of a formal exchange of
information between institutions at the international level. This type of response
may be less likely to occur vertically as well as at the EU level, where an information
exchange may not be formally established between different EU legal instruments
but occurs through the European Commission.
Potential for Further Improvement Is Widespread In a significant number of cases,

a potential for further improvement by means of additional policy responses was
identified. Experts saw a potential for further improvement in about 30 percent of
the cases of interaction, while they found no such potential in roughly another 30
percent of the cases. In about 40 percent of the cases, the situation remained unclear
(table 13.2). Overall, there is a significant and widespread potential for enhancing
EU and international environmental governance to be exploited by a more system-
atic ‘‘interplay management’’ (Stokke 2001).
Somewhat counterintuitive, potential for further improvement appears to exist in
particular where response action has already been taken. The Cramer’s V of 0.364
points to a comparatively weak but still relevant correlation. Hence, in a majority of
cases in which a potential for improvement exists, actors have already attempted to
improve the situation by means of collective action, but have not realized the full
potential in most of these cases. Only in the case of vertical interaction had a major-
ity of cases in which experts identified a significant potential not drawn a collective
political response. This may be due to the hierarchical relationship between inter-
national and EU law, so that it may be particularly difficult to pursue collective
action to enhance existing synergy or mitigate conflict. However, there are cases
at all levels of interaction in which an existing potential for improvement has
not yet been attempted to be realized by means of collective action (overall twenty-
two cases, accounting for nearly 15 percent of the cases of our sample) (table
13.5).
From our data, it appears that significant benefits from enhancing synergy have
been neglected so far and remain to be reaped. Whereas synergistic cases of interac-
tion more rarely display a potential for further improvement than disruptive cases,
existing opportunities for enhancing synergy have received far less attention than
the threat of disruption. A potential for further improvement was found in all except
one of the twenty-seven disruptive cases that allowed for a clear assessment, as com-
pared with twenty-four of the sixty-four synergistic cases (table 13.6).1 Also, two-
thirds (twenty-one of thirty-one) of the cases that have been responded to and still
have further potential were disruptive. However, about three-fourths of the cases
(sixteen of twenty-two cases) that have not been responded to but possess a poten-
tial for further improvement were synergistic. Accordingly, Cramer’s V indicates a
less relevant correlation between further potential and quality of effect for cases
that have not been responded to than for those that have (table 13.6).
A number of interesting questions may be explored in the future. For example, is
existing potential for improvement through targeted policy responses more easily
exploited in cases of interaction within the same policy field than in interaction
across different policy fields, as one might expect? There are only six cases in which
policy responses appear to have led to a full exploitation of the potential for im-
Table 13.6
Quality of effect, policy responses, and potential for further improvement
Response No response Total
No No No
Further further Further further Further further
potential potential potential potential potential potential
Quality of effect No. % No. % No. % No. % No. % No. %
Synergy 8 25.8 2 33.3 16 72.7 38 95.0 24 45.3 40 87.0
Disruption 21 67.7 0 0.0 5 22.7 1 2.5 26 49.1 1 2.2
Neutral/unclear 2 6.5 4 66.7 1 4.5 1 2.5 3 5.7 5 10.9
Total 31 100 6 100 22 100 40 100 53 100 46 100
Cramer’s V 0.645 0.336 0.525
Source: Own compilation from database reflected in the appendix, excluding cases with
unclear potential for further improvement.
provement (table 13.6). Of these six cases, two involved institutional interaction
within the same policy field (not reflecting the distribution within our overall sam-
ple). This is not in line with our expectation that interaction within the same policy
field should be more conducive to creating synergy and mitigating disruption. How-
ever, the limited number of cases does not allow us to draw any firm conclusions.
Moreover, our analysis did not allow us to assess to what extent a potential for im-
provement has been realized through policy responses.
Weberian Ideal Types of Institutional Interaction
In chapter 2, we introduced three causal mechanisms that provide the analytic

framework of this study. Cognitive Interaction reflects a learning process and relies
on the transfer of information. Interaction through Commitment is based on the
commitment of some members of the target institution to the obligations of the
source institution and an ensuing change of their preferences. Behavioral Inter-
action occurs within the issue areas but outside the decision-making processes of
either of the institutions involved. The basic causal mechanisms reflect the causal
pathways through which influence is channeled from the source institution to the
target institution.
In the following, we move beyond these three basic causal mechanisms and de-
velop more specific ideal types of institutional interaction that help us understand
how interaction operates and why some cases generate disruptive effects, while
others create synergy. We do not attempt to modify, change, or replace the basic
causal mechanisms. Instead, we endeavor to identify patterns of interaction that elu-
cidate systematically the different forms and effects of varying cases of interaction
following the same causal mechanism. Chapters 3–12 of this volume have illus-
trated the variance of forms and effects of different cases of interaction. To take
account of this variance, we further specify and differentiate the general causal
mechanisms by developing ideal types of institutional interaction.
The construction of models or ideal types is a well-known method of social-
science inquiry. However, types and classes have explanatory power only if their
existence is based on a distinct logic (Weber 1976, 1–11). Mere grouping or classi-
fication of empirical cases according to certain individual properties does not
explain anything. Instead, an abstract, theoretical reconstruction of their inherent
logics is needed. Weberian ideal types are abstract models of social interaction phe-
nomena, to which real-world cases can be compared.
While Weberian ideal types are by definition mutually exclusive, real-world cases
may reflect aspects of more than one model at the same time. Ideal types are neces-
sarily mutually exclusive because they follow their own distinct rationales. They are
constructed to elucidate the important components of these rationales. As abstract
models, ideal types reflect the basic pattern of a real-world situation appropriately,
or they do not, but they cannot be empirically right or wrong (Snidal 1985). How-
ever, they grossly reduce the complexity of real-world cases, and are not intended to
provide precise descriptions of all properties of these cases. ‘‘Mixed cases’’ that are
concurrently driven by two or more of the basic rationales and therefore difficult to
classify may exist.
We constructed Weberian ideal types of institutional interaction in an iterative
three-step process oscillating between inductive and deductive reasoning. First, we
systematically assessed our sample of 163 cases of institutional interaction. On the
basis of the coding of these cases as reflected in the database in the appendix, we
grouped the cases of our sample following the same causal mechanism in numerous
different ways in accordance with the variables spelled out above in order to iden-
tify systematic patterns. Second, we constructed distinct rationales of the resulting
groups of cases. In essence, this step was a matter of deductive reasoning directed
at reconstructing the basic logic driving a number of cases. It aimed to identify
groupings that are based on systematic differences in their underlying logics. As a
result, we identified two ideal types of Cognitive Interaction and three types of Inter-
action through Commitment. In contrast, it proved difficult to identify ideal types of
Behavioral Interaction. Third, we sought to assign the cases of Cognitive Interaction
and Interaction through Commitment within our sample to the corresponding ideal
types. This exercise put our families of Weberian ideal types to a first empirical test.
That all our cases of Cognitive Interaction and Interaction through Commitment
could be assigned to the two families of ideal types indicates that our ideal types
cover a broad range of existing interaction phenomena.
This process of deriving ideal types of institutional interaction revealed that few
possible groupings of cases lend themselves to constructing distinct rationales. While
cases could be grouped in various ways, most groups did not allow for the deductive
development of consistent underlying logics. For example, one could group cases of
Interaction through Commitment according to their effects (synergistic versus dis-
ruptive) or the intentionality of action on the part of the source institution, but it
proved impossible to identify an abstract model of the underlying rationale fitting
the relevant cases. Likewise, cases of horizontal interaction at the EU level are not
systematically different from those at the international level, or from the vertical
cases between international and EU institutions. Ideal types of Cognitive Interaction
eventually identified vary with respect to the variable intentionality, while ideal
types of Interaction through Commitment possess distinct key differences.
Although our ideal types of interaction are based on our specific sample of cases,
they are relevant for other samples both from international and EU environmental
affairs and beyond. The types are generally applicable, because they do not originate
from mere classification of the cases within our sample of cases. Instead, they are the
result of deductive reasoning that addresses the question of what drives the cases of
a particular class. There is no reason to assume that cases beyond our sample cannot
reasonably be analyzed by comparing them with our ideal types of interaction.
Hypotheses derived from these abstract models can well be tested against other sam-
ples of cases. However, our ideal types are linked to our specific sample in two
ways. First, we could only identify types of interaction that occur within our sample.
Consequently, there might be cases that do not fit these types and require the devel-
opment of additional ones. Second, the empirical cases and their properties as well
as the distribution of cases between types of interaction can be assumed to change
between samples. We submit that our ideal types also provide a useful framework
of analysis of institutional interaction beyond our sample of cases.
Ideal Types of Cognitive Interaction
Cognitive Interaction can be conceived of as a form of learning across the bound-

aries of institutions. Cases driven by this causal mechanism are based on the transfer
of information alone. Members of the target institution must voluntarily change
their perceptions in response to new information provided by the source institution.
Cognitive interaction is the least intrusive causal mechanism because it depends
completely on the voluntary acceptance of the members of the target institution.
The source institution does not exert significant pressure on the target to adapt. It
does not establish obligations or otherwise change preferences of the members of
the target institution against their conviction.
Intentionality is the crucial distinction between types of Cognitive Interaction.
While ‘‘learning’’ cannot be imposed, it may or may not be intentionally triggered
by the source institution. If the source institution intentionally requests the target
to change in a certain respect, the attention of the members of the target institution
is deliberately drawn to the effects of their decisions on the source institution. If
Cognitive Interaction is not intended by the source institution, members of the target
institution use an institutional arrangement or policy idea of the source institution
as a model. Hence, intentionality determines whether the center of activity rests
with the source institution or the target institution.
Other possible distinctions did not provide a solid foundation for the develop-
ment of distinct rationales. For example, there is no reason to believe that learning
with synergistic effects for the target institution is founded on a rationale that is
fundamentally different from that of learning with a disruptive effect on the target.
Likewise, learning across the boundaries of policy fields or among institutions with
different objectives does not seem to be fundamentally different from learning within
the same policy field or between institutions with similar objectives, although the
latter might be somewhat more frequent.
Policy Model
If Cognitive Interaction is unintentionally triggered, members of the target institu-
tion voluntarily use some aspect of the source institution as a Policy Model. For ex-
ample, the compliance system under the Montreal Protocol for the protection of the
ozone layer influenced the negotiations on the compliance system under the Kyoto
Protocol on climate change because it provided a model of how to supervise imple-
mentation and deal with cases of possible noncompliance (Oberthür and Ott 1999,
215–222; Werksman 2005; chapter 3). The members of the Montreal Protocol did
not establish the model in order to influence the Kyoto Protocol. They did also not
have the ability to impose their model on the target. Instead, the Montreal Protocol
presented an institutional arrangement that the members of the Kyoto Protocol con-
ceived of as a useful precedent in developing solutions for their problem.
In the Policy-Model type of Cognitive Interaction, action seems to concentrate in
the target institution. This impression is due to two characteristics of these cases. On
the one hand, members of the source institution will find it difficult to foresee which
kind of information or decision originating from their collective decision-making
process might prompt interaction of this kind. Any institutional arrangement, deci-
sion, or scientific or technological information from the source institution might
serve as a Policy Model. And numerous types of actors might pick up the informa-
tion or idea and feed it into the decision-making process of the target institution.
This may be done, for example, by one or more member states, by an interested
nongovernmental organization, by the secretariat of the target institution, or even
by relevant individuals. These actors may, but do not have to be active within the
source institution, because information and ideas can also be obtained by surveying
the field, reading reports, or examining institutional arrangements from outside. On
the other hand, it is the actors making the institutional decisions within the target
institution (usually the member states) whom the new information must convince
so that they change their preferences. Only then can we expect a new constellation
of interests to emerge, which might lead to modified decisions within the target
institution.
Learning from a Policy Model is not limited to the unchanged acceptance of a
policy model developed elsewhere. Frequently, a model is not simply copied into
the target institution (‘‘simple learning’’), but it is modified and adapted so as to
fit the particular needs of the target. ‘‘Learning’’ is thus regularly a complex process
of the collective examination and appraisal of the model followed by its modifica-
tion and development (‘‘complex learning’’; see Haas 1990).
Interaction of the Policy-Model type can be expected to strengthen the effective-
ness of the target institution, while being indifferent to the effectiveness of the source
institution. Effects will be largely supportive of the target’s objectives, because the
interaction can materialize only if the target members collectively consider the policy
model to be useful. Actors cannot be forced to learn and they will not pick up
alleged models that they are not convinced of. However, it cannot be entirely ex-
cluded that actors learn ‘‘wrong’’ lessons that prove to hamper rather than enhance
effectiveness. If learning from a policy model leads to better decisions in the target
institution, this will (at least potentially) strengthen the target’s influence on the be-
havior of relevant actors within its issue area (outcome level), which will (possibly)
also enhance the effects at the impact level. In contrast, the source institution will
usually not be significantly affected, because learning within the target institution
and its effects at the outcome and impact levels could only accidentally feed back
onto the original source institution.
The Policy-Model type of interaction highlights that members of an institution
can enhance the effectiveness of their governance efforts by learning from occur-
rences taking place within other institutions. Any institution—be it international
or European, environmental or other—may learn in this way from any other
institution that displays appropriate precedents and deals with relevant problems.
Frequently, negotiators operating in the framework of international and EU institu-
tions already look for precedents. However, this process is so far frequently ad hoc
and incidental. Especially in international institutions, learning processes may be
further promoted by mandating secretariats and conferences of the parties to
systematically look for and investigate models existing in other institutions. Elabo-
rating and institutionalizing routines and mechanisms for doing so may enhance the
potential of interinstitutional learning so as to foster effective governance.
Request for Assistance

Cognitive Interaction may also take the form of an intentional request by the source
institution for assistance from the target institution. The attention of actors within
the target institution is thereby drawn to a particular aspect of the outside world
that they had so far not taken due account of—at least in the eyes of the source in-
stitution. The source institution still has little ability to force the target to adapt
without the latter’s consent and remains completely dependent on the goodwill of
the target. For example, several wildlife-protection treaties have asked the Conven-
tion on International Trade in Endangered Species of Wild Fauna and Flora (CITES)
for assistance. And CITES itself requested assistance from specialized international
institutions such as the World Customs Organization (WCO) and Interpol to rein-
force the effectiveness of its trade restrictions (chapter 7).
In contrast to learning from a Policy Model, the source institution largely frames
the learning process in cases of Request for Assistance. It intentionally triggers the
interaction by a decision without which interaction of this type would not occur.
Moreover, a Request for Assistance will usually be formally transferred by the sec-
retariat of the source institution to the secretariat of the target institution, and it will
be officially fed into the decision-making process of the latter. Evidently, the source
institution cannot force the target institution to adapt. Some learning in the form of
an adaptation of preferences to new information is necessary to produce a reaction
within the target institution. Actors learn either that their institution produces side
effects that the source institution considers undesirable or that an adaptation of their
institution could further strengthen the effectiveness of the source institution.
Its rationale suggests that a Request for Assistance will generate synergistic or at
least neutral effects for the target institution. We cannot expect to find cases with
disruptive effects on the target institution, because a successful Request for Assis-
tance requires that actors operating within the target institution voluntarily take
action in response. It is hardly conceivable that these actors would adopt measures
in response to a Request for Assistance that are detrimental for the target’s policy
objectives. Even if ‘‘negative’’ learning is unlikely to occur, Requests for Assistance
will not necessarily reinforce the effectiveness of the target institution. Considering
frequent overlaps in membership between the institutions involved, the target insti-
tution might find it difficult to reject a request even if it does not benefit from
responding. Hence, we may well expect to find cases in which the effects on the
target institution are indeterminate.
Whereas the Request for Assistance is directed at the target institution, it is ulti-
mately intended to further the effectiveness of the source institution by triggering a
feedback case of Behavioral Interaction. If CITES asks the World Customs Organi-
zation to adapt the international customs codes to its needs, it first of all asks for a
corresponding decision of the WCO. However, CITES members issue the request
because they expect the application of adapted customs codes to facilitate the effec-
tive implementation of CITES obligations. This second case of interaction occurs
within the issue areas governed by the institutions involved outside their decision-
making apparatuses, when customs officers in the port of Hamburg or at the John
F. Kennedy Airport in New York check imported goods.
Requests for Assistance provide an instrument for furthering effective interna-
tional and EU governance. Although the source institution cannot force the target
institution to adapt its rules, a Request for Assistance enables an institution to
draw on other institutions in order to enhance its own effectiveness. To this end,
members of international and possibly also of EU institutions might actively search
for opportunities to instigate learning processes within other institutions. Since
the success of a Request for Assistance largely depends on the positive reaction by
the target institution, the endeavor will be particularly promising if the requested
adaptation is easily compatible with the objectives of the target institution, or, even
better, if it supports them.
Cases of Cognitive Interaction in the Sample

Significant Cognitive Interaction seems to be a comparatively rare phenomenon.
Only 16 of the 163 cases of our sample (9.8 percent) fall into this category. They
are evenly divided between the two types of Policy Model and Request for Assis-
tance (see table 13.7). The reasons may be assumed to differ between the two types.
Clear-cut Policy-Model cases with significant effects are difficult to detect. Learning
from a Policy Model is a ubiquitous, gradual, and frequently tacit phenomenon that
does not always produce traceable evidence. Since we focused on cases of clear and
significant influence, some Policy-Model cases may have evaded identification. In
contrast, Requests for Assistance are readily observable because they are usually
reflected in a formal decision of the source institution. While opportunities for
requesting assistance may be limited, interested actors may also fail to seize further
Table 13.7
Cases of Cognitive Interaction in our sample
Request for
Policy model assistance Total
No. % No. % No. %
Totals 8 100 8 100 16 100
Horizontal international 5 62.5 8 100 13 81.3
Horizontal EU 1 12.5 0 0.0 1 6.3
Vertical 2 25.0 0 0.0 2 12.5
Synergy 6 75.0 4 50.0 10 62.5
Disruption 1 12.5 0 0.0 1 6.3
Neutral/unclear 1 12.5 4 50.0 5 31.3
Same policy field 7 87.5 4 50.0 11 68.8
Different policy field 1 12.5 4 50.0 5 31.3
Response 1 12.5 3 37.5 4 25.0
No response 7 87.5 5 62.5 12 75.0
opportunities that exist in environmental governance. Our sample comprises eight

cases of Request for Assistance, seven of which involve CITES either as the source
or as the target institution. The CITES Secretariat has actively pursued this type of
interaction (see chapter 7). Similar requests have been reported in the literature, for
example with respect to the Montreal Protocol asking the World Customs Organi-
zation (WCO) to adapt (Oberthür 2001).
Our sample suggests that Request for Assistance is particularly relevant to the
horizontally structured realm of international institutions. All our cases of this type
involve horizontal interaction between international institutions, while none of the
horizontal interactions between EU institutions and vertical interactions between in-
ternational and EU institutions fit this model (table 13.7). There are reasons for this
concentration of cases. Horizontal requests for assistance are less likely in the more
integrative institutional framework of EU decision making. Having the exclusive
competence within the EU to initiate legislative proposals, the European Commis-
sion may be expected to address emerging issues of coordination between institu-
tions in the process of preparing such proposals. ‘‘Interservice consultations’’ are
designed to involve, and get agreement of, all relevant Commission services. As a re-
sult, formal requests for assistance possess less relevance within the EU. They are
also unlikely to occur vertically between international institutions and EU legal

instruments because of the status of the EU as a subject of international law that
concludes international agreements alongside nation-states (Schweitzer 2003; see
also Oberthür 1999). The EU is a party to most international treaty systems dealt
with in this study and pursues its interests in changes of an international regime in
the same way as any other party, namely in the relevant international negotiations.
Where international regimes and organizations, in turn, ask the EU to adapt, the EU
is not totally free but is usually required to implement internationally agreed-on
rules in much the same way as any other party. Such a request would thus initiate
a different causal mechanism, namely ‘‘Interaction through Commitment.’’
Our sample demonstrates that learning from a Policy Model is possible across all
formal boundaries and levels, so that it occurs within and across policy fields as well
as at all levels of interaction (table 13.7). We found cases between international
institutions, between EU institutions, and, vertically, between international and EU
institutions. Moreover, learning from a Policy Model is possible across policy fields,
although the fact that this occurred in only one of our eight cases suggests that it
might be easier and more probable between institutions operating within the same
policy field. The fact that response action was rare may be attributed to the fact
that source institutions remain totally unaffected by the interaction, while target
institutions already exploited the learning opportunities inherent in the case.
As expected, cases of the Policy-Model type are primarily about creating synergy.
We identified two deviant cases (table 13.7). One case had disruptive effects. In the
mid-1990s, several parties to the climate change regime wanted to employ the model
of the technology and economic assessment panels of the Montreal Protocol (on the
latter see Parson 1993, 2003) for feeding scientific advice into the political decision-
making process. Other parties blocked the adoption of a similar arrangement pre-
cisely because of its effectiveness. From the ‘‘model’’ of the Montreal Protocol, they
had learned which features of the science-policy interface should be avoided in order
to limit the effectiveness of scientific advice. On the basis of the consensus principle
that prevails in many international institutions, minorities may thus ‘‘learn’’ to
block certain decisions. The quality of the effect of this type of interaction therefore
depends on whether the lessons learned from the Policy Model benefit the leaders or
the laggards within the target institution. The other deviant case is simply difficult to
judge. The Montreal Protocol provided to the WTO a model for admissible trade
restrictions for environmental protection in a multilateral framework, which takes
interests of nonparties duly into account. It is difficult to determine whether this is
conducive to the objectives of WTO (because it mitigates conflict over environmen-

tally motivated trade restrictions) or not (because it leads to acceptance of such
restrictions under certain circumstances).
None of the Requests for Assistance in our sample resulted in disruption. Four of
our eight cases created synergistic effects for the target institution. These cases re-
lated to interaction between CITES and other wildlife treaties. The other four cases
had neutral or insignificant effects on the policy objectives of the target institutions.
Three of them related to CITES’ Requests for Assistance to institutions located in
other policy fields (WCO and Interpol). It may not be surprising that Requests for
Assistance occur both within the same policy field and across different policy fields.
That response action appears to occur more frequently in cases of Request for
Assistance than in the Policy-Model cases may be due to the fact that Requests for
Assistance per se establish some kind of contact between the source and the target
institution that may lead to further coordination or joint activities. However, the
number of cases is too small to draw any firm conclusions.
Ideal Types of Interaction through Commitment
It is characteristic of cases of Interaction through Commitment that commitments

entered into by the members of the source institution change the constellation of
interests within the target institution tangibly. Actors have an interest in being
consistent and developing a reputation of keeping their promises. They may also be
interested in subjecting others to costly obligations that they have already accepted.
Options that are in conformity with the obligation of the source institution are
therefore adopted with a higher probability, while agreement on options that are in
conflict with obligations of the source institution is less likely. Being aware of the
binding force of obligations, actors may even gain an interest in adopting commit-
ments in one institution in order to frame the policy choices available in another
institution. As a result, commitments of the source institution may be exported to
the target institution or may serve to prevent adoption of contradicting decisions in
the target. In contrast to Cognitive Interaction, Interaction through Commitment
does not rely on a transfer of information and learning, but on the binding force of
obligations. While only the source institution can initiate Interaction through Com-
mitment, it is still not capable of determining the action of the target institution.
The causal mechanism of Interaction through Commitment covers a broad range
of phenomena. Some of them create disruption, like the pending conflict between
the WTO and several multilateral environmental agreements (chapter 8), while
others generate synergy. Some of the synergistic cases of interaction reach across
the boundaries of policy fields, while some disruptive ones remain within a single
policy field. Some are intentionally created, while others occur unintentionally. All
these criteria might lend themselves to creating a typology, but they do not support
abstract model types that rely on a clear rationale and cover all relevant cases of our
sample.
While cases of Interaction through Commitment require a significant overlap of
the memberships and the issue areas of the institutions involved, institutions must
differ in some important dimension to create momentum for interaction. In contrast
to Cognitive Interaction, Interaction through Commitment presupposes that some
countries are members of both institutions. Otherwise, no actors within the target
institution would be committed to obligations originating from the source institu-
tion. Likewise, issue areas must overlap somewhat. Otherwise, no commitment of
the source institution could matter directly for the decision-making process of the
target institution. However, a significant difference between the institutions is also
a prerequisite for Interaction through Commitment. In the hypothetical situation of
two identical institutions, we could not expect significant interaction between them.
The three ideal types of Interaction through Commitment developed in the fol-
lowing paragraphs vary with respect to their values of the variable ‘‘key differences’’
introduced in the first part of this chapter. Cases of the ideal type of Jurisdictional
Delimitation are driven by different objectives of the interacting institutions. Cases
of the type of Nested Institutions are driven by significantly different memberships
of the two institutions involved. Cases of the type of Additional Means gain their
relevance from significantly different means of the institutions involved. In each of
these ideal types, institutions are assumed to be congruent as to the other two poten-
tial key differences. As a result, the underlying rationales of these three types of in-
teraction can be expected to differ profoundly because interinstitutional influence
relies on exactly one distinct key difference for each ideal type. To avoid confusion,
it is repeated here that real-world cases of institutional interaction may well be
driven by more than a single rationale. These ‘‘mixed’’ cases do not require separate
ideal types.
Jurisdictional Delimitation
Cases of Interaction through Commitment driven by differences in objectives create
a demand for the delimitation of jurisdictions. We call this ideal type of interaction
‘‘Jurisdictional Delimitation.’’ In the ideal type, the same group of member states
addresses the same issue by identical means within two institutions with different
objectives. As a result, actors are confronted with the prospect of conflicting obliga-
tions concerning the same subject. Consider that international trade is regulated
within the WTO for the purpose of liberalizing trade and thus removing obstacles
to international trade. At the same time, the Cartagena Biosafety Protocol to the
Convention on Biological Diversity (CBD) governs international trade in genetically
modified organisms (GMOs) predominantly for the purpose of protecting the envi-
ronment of the importing countries. In practice, the WTO protects the interests of
exporting countries, whereas the Biosafety Protocol guards the interests of import-
ing countries (chapter 8). While it is perfectly possible that a group of states simul-
taneously aims at liberalizing trade and at protecting their environment, a country
cannot at the same time restrict trade to protect its environment and remove these
trade obstacles. Conflicting commitments may lead to ambiguity because actors
cannot sincerely implement both of them simultaneously. To resolve the conflict,
jurisdictions of the institutions involved must be separated from each other, so that
the substantive overlap of the two issue areas diminishes. Hence, cases of Jurisdic-
tional Delimitation create a demand for the clear allocation of exclusive regulatory
authority.
In some cases, it may be possible to delimit the jurisdictions involved so clearly
that interdependence of governance disappears completely, whereas it continues
to exist on a less contentious basis in other cases. Diverse regional seas regimes
minimize interdependence of their governance activities through territorial separa-
tion of their issue areas. In contrast, delimitation of the jurisdictions of the regime
for the protection of the ozone layer and the climate change regime is founded
on a subject-by-subject approach. Chlorofluorocarbons (CFCs) and other ozone-
depleting substances regulated by the Montreal Protocol are excluded from regula-
tion by the climate change regime, even though most of these substances are also
greenhouse gases. In spite of a gradual delimitation of their jurisdictions, the WTO
and environmental regimes with trade restrictions remain more interdependent.
While international trade is generally governed by the WTO, it is increasingly
recognized that the details of certain trade restrictions established for purposes of
environmental protection fall within the authority of international environmental
institutions. In these latter cases, some contentious issues are assigned to one in-
stitution, although they remain of interest to the other institution involved. Conse-
quently, the regulating institution must take the competing regulatory objectives
duly into account.
Jurisdictional Delimitation cases demonstrate that the issue areas governed by in-
ternational or EU institutions are not ‘‘given,’’ but socially constructed. In coopera-
tion theory, it is usually assumed that the boundaries of an institution’s issue area
are largely preexisting, or externally ‘‘given.’’ While negotiation analysis emphasizes
that constellations of interests can be modified by adding or subtracting actors and
issues (Sebenius 1983, 1992; also Haas 1975), Jurisdictional Delimitation cases
point to the need for occasional manipulation of the boundaries of relevant issue
areas to avoid conflict or fruitless competition between institutions.
Due to their underlying rationale, Jurisdictional Delimitation cases will usually
restrain the effectiveness of both institutions involved. The probability of conflicts
and trade-offs will be high, if two institutions with significantly different objectives
regulate the same issue, because regulatory considerations differ. The target institu-
tion’s effectiveness will almost inevitably be undermined, if the source institution
encroaches on its jurisdiction. By creating conflicting obligations the source institu-
tion enlarges the room for national and subnational actors to interpret the obliga-
tions to their liking and thus to disregard the target’s commitments at least partially.
At the same time, conflicting obligations make it also easier for national and sub-
national actors to disregard the commitments of the source institution so that
they regularly also jeopardize the effective implementation of obligations under
the source institution. Under these circumstances, synergy between the institutions
involved will be rare or totally absent. It can only be expected if one of the institu-
tions dominates the situation and a delimitation of issue areas occurs amicably.
The delimitation of issue areas tends to be a matter of conflict and of the distribu-
tion of power between the institutions involved, rather than of amicable problem
solving. The model type of Jurisdictional Delimitation elucidates when and why
overlapping jurisdictions create a demand for their separation, but not where the
balance will or should lie (nor how and when it will be achieved). The regulatory
competition between the institutions can usually be solved in different ways. Au-
thority for regulating imports of GMOs could be exclusively allocated to the WTO
that would presumably give priority to free-trade considerations, or to the Carta-
gena Protocol, which can be expected to prioritize environmental protection. In
reality, the balance in this case is struck somewhere in between these extremes
(chapter 8). How it is struck is eventually a matter of the political process based on
interests of varying strength, with all uncertainties and attractions inherent in this
process.
The members of both the source institution and the target institution are in a
‘‘mixed-motive’’ situation that resembles the game-theoretic constellation of the
Battle of the Sexes (Stein 1982; Keohane 1984). On the one hand, they possess a
common interest in some sort of separation of jurisdictions in order to avoid fruit-
less regulatory competition and a reduced effectiveness of their respective institu-
tions. Neither side will be served if the institutions involved mutually undermine
their effectiveness. On the other hand, the constituencies of the institutions have
conflicting preferences that make it notoriously difficult to find a mutually accept-
able solution. In the aforementioned conflict between the WTO and the Cartagena
Protocol on the regulation of GMOs, neither side can be content with a conflict that
disturbs both international trade and environmental protection. However, actors
favoring free trade will advocate regulation by the WTO, while countries struggling
for far-reaching domestic environmental regulation will prefer enlarged jurisdiction
of the Cartagena Protocol. The conflict might even extend to different constituent
groups (ministries, advocacy groups) within member states.
The governance challenge consists in arriving at a delimitation of jurisdictions
that balances the diverging interests and realizes the common interests. Conflict
management will tend to identify measures that honor the basic objectives of each
institution, while being least intrusive into the objectives and operation of the other
institution. It does not necessarily require an overarching institutional structure.
Only in the EU are supranational actors such as the European Commission and the
European Court of Justice available to assist in striking the right balance in ac-
cordance with general principles. Balancing of competing political objectives
such as free trade and environmental protection or sustainable development can
occur within either of the institutions pursuing them. Actors that are members of
both institutions are likely to play a major role, because they will usually have the
strongest interest in making contradictory demands compatible.
In Battle-of-the-Sexes situations, an equilibrium found is assumed to be fairly
stable, because neither side can expect to gain from resumption of conflict. This
standard conclusion from game theory suggests that, if the conflicting institutions
are established consecutively, the earlier institution will posses a ‘‘first-mover advan-
tage’’ (Héritier 1996; Mattli 2003). Commitments of the earlier institution will al-
most automatically limit the room for maneuver within negotiations of the later
institution by strengthening the actors preferring the earlier institution’s objectives.
Hence, it will be difficult to change the balance by establishing a new institution that
encroaches on the issue area of the old one.
Occasionally, it may nevertheless prove useful as a political strategy to deliber-
ately raise (potential) jurisdictional conflict—even by creating new institutions. For
example, this may be the case if new regulatory objectives such as environmental
protection are to be promoted in a field already governed by an existing institution.
Consider that the world trade regime (GATT/ WTO), having developed since the
late 1940s, principally covers all international trade issues. Whenever international
environmental regimes established since the 1970s have restricted trade for environ-
mental purposes, they have almost automatically encroached on the established
jurisdiction of the trade regime. However, environmental trade restrictions agreed
on within multilateral environmental agreements are today widely acknowledged
as legitimate, if sufficiently adapted to the needs of the world trade regime (chapter
8). Hence, raising jurisdictional conflict is not per se futile or negative for the effec-
tiveness of international governance. To be successful, it requires a sufficiently high
capacity to seriously endanger the jurisdictional authority of the target. The force of
such threats will depend on their credibility and thus on the ability and political will
of source-institution members to follow up on their initial action, if the target proves
unwilling to adapt.
Nested Institutions
Interaction through Commitment may take place between two institutions that dif-
fer exclusively with respect to their membership, while pursuing identical objectives
and employing the same means. These cases follow the ideal type of Nested Insti-
tutions. Even if addressing identical problems, institutions with divergent mem-
berships may arrive at differing obligations because the relevant constellation of
interests depends on the set of actors involved (Sebenius 1983, 1992). However,
the causal mechanism of Interaction through Commitment requires that member-
ships overlap somewhat. In the absence of any overlap, such as between two re-
gional seas conventions in different parts of the world, no state would be motivated
to adapt its preferences within the target institution to the obligations of the source
institution. The two conditions of difference in and overlap of memberships of two
institutions will be fulfilled in their clearest form if the membership of one institution
forms part of the membership of another institution. In this case, two formally inde-
pendent institutions with similar objectives and regulatory means are ‘‘nested’’ into
each other (Aggarwal 1983; Young 1996). Nested-Institution cases gain their
momentum in particular from the tension between EU legislation or other regional

arrangements on the one hand and larger, in particular pan-European or global,
agreements on the other hand. For example, the decision of the regional regime for
the protection of the Northeast Atlantic (Oslo and Paris Conventions, OSPAR) to
ban dumping and incineration of waste at sea led to the adoption of similar mea-
sures within the global London Dumping Convention (chapter 5; Meinke 2002).
Interaction between Nested Institutions rests on three related mechanisms. First, it
is typically easier to reach agreement within a smaller (e.g., regional) than in a larger
(e.g., global) institution because a higher number of participants usually implies a
greater heterogeneity of interests (Snidal 1994). In international legal systems, the
prevailing consensus principle complicates decision making, and even in the Euro-
pean Union a comparatively small minority can block a decision by the Council of
Ministers. As a consequence, the rejection of a decision by an interested actor or
subgroup of actors is the more probable, the larger a group is. It may be noted that
a greater homogeneity of interests within an institution may also follow from other
factors such as a greater homogeneity of economic or political conditions of the
members. Consequently, a similar momentum could in principle develop between
institutions of a similar size but with partially overlapping memberships.
Second, commitments agreed within the smaller institution will streamline the
preferences of its members in two dimensions. Internally, accepted obligations will
bind members of the institution, where heterogeneous preferences might have pre-
vailed before. Hence, accepted commitments remove options that existed before.
Externally, the members of the smaller institution may develop a common inter-
est in extending the agreed-on commitments to third parties that benefit from the
smaller agreement like free riders. Thus, even members of the smaller institution
who accepted the commitment only grudgingly can be expected to gain an interest
in its wider application, because they have to implement it. This interest will be par-
ticularly strong if implementation of the agreed-on commitments is costly and less-
ens the competitiveness of sincere cooperators (Gehring 1997).
Third, streamlined preferences reinforce the position of a smaller group as com-
pared to the necessarily more heterogeneous individual negotiation positions. Based
on their common commitment, members of the group form a coalition. It will be
more difficult to convince a whole group of negotiators than a single country to
change its preagreed negotiation position. Likewise, it will be easier to offer a single
country a side payment that makes it change its position than a whole group of
countries. Hence, the probability increases that the coordinated position of the co-
alition constitutes some ‘‘focal point’’ (Schelling 1995, 100) around which expecta-
tions for an agreement converge. This does not ensure that diffusion of the policy
measure from the smaller to the larger institution is successful. Although they are
also members of the larger institution, the members of the smaller institution cannot
impose their measures on the broader membership of the target institution. If resis-
tance against the measure is too strong, the diffusion process will stop and interac-
tion will fail.
Influence from Interaction between Nested Institutions can only originate from
the smaller institution and affect the larger institution. If the smaller institution
took over an obligation from the larger one, no additional actor would become sub-
ject to the obligation, so that this transfer would not be relevant for the effectiveness
of governance. The transfer of an obligation to a smaller institution could affect the
effectiveness of governance only if it activated an additional means of implementa-
tion. As we will see in the next subsection, such a transfer belongs to a further type
of Interaction through Commitment that is characterized by the difference in means
available for governance within the two institutions rather than the difference of
their memberships.
The rationale of interaction between Nested Institutions suggests that effects will
largely support the effectiveness of the target institution, and occasionally also of the
source institution, while disruption is unlikely to occur. The assumption of identical
objectives renders disruptive effects on target institutions highly improbable if not
impossible. It is difficult to imagine how the transfer of a policy measure from a
smaller to a larger institution with identical objectives might lead to disruption. It
will most probably contribute to increasing the effectiveness of the target. Interac-
tion will only exceptionally produce disruptive effects, if the policy measure adopted
has unintended negative consequences or interferes with other policy measures exist-
ing within the target institution. However, this result would not any more be an im-
mediate consequence of the rationale underlying cases of Nested Institutions. In the
latter case, it presupposes a certain difference of means between the two institutions
involved. The effectiveness of the source institution will not be affected by interac-
tion between Nested Institutions, even though the resulting transfer of commitments
promotes other interests of the members of the source institution (otherwise this
type of interaction would not come about). If the objectives and governance means
of the two institutions do not differ, positive feedback is difficult to conceive of, but
it is not entirely excluded. Consider that, if the European Union decided to commit
member states to taxing air transport in order to reduce pollution, it might not be
possible to subject planes from third states to similar rules. Under the circumstances,
diffusion of the obligation to a global institution with larger membership would sup-
port the effectiveness of the regional institution.
Interaction between Nested Institutions constitutes a mechanism for policy diffu-
sion within the same policy field and provides opportunities for forum shopping.
Where possible, actors striving for regulation of a particular issue may choose
whether to promote their proposals predominantly in a smaller or in a larger insti-
tution. If they choose the smaller one, they may actively employ it as a ‘‘pilot insti-
tution’’ to exercise ‘‘leadership by example’’ (Grubb and Gupta 2000, 21; also
Young 1991) to subsequently expand an obligation to a larger membership. To do
so, they can create provisions in the smaller institution as a catalyst for change in the
broader institution operating in the same issue area. Whereas regional agreement
may not suffice to bring about broader international change, it can help tip the
balance. Thus, policymakers within the EU and regional agreements should be
conscious of the rationale of interaction between Nested Institutions because passing
of ambitious standards may drive broader international and global agreement.
Additional Means
Interaction through Commitment may also take place between two institutions with
an identical membership and the same objective, if the governance instruments
(means) available within these institutions differ significantly. After adoption of an
obligation within one institution, a given group of actors will comparatively easily
agree on the same obligation within a second institution governing the same issue
area with the same objectives because of their previous agreement. However, such
simple diffusion of an obligation cannot be assumed to change the situation signifi-
cantly for actors that are members of both institutions unless incorporation of the
obligation into the target institution mobilizes an additional instrument (means) of
implementation. In this case, the diffusion will matter for the effectiveness of envi-
ronmental governance, because it will make it more difficult for addressees to side-
step or ignore the obligation.
Frequently, international and EU institutions do not control the full spectrum of
possible governance instruments but differ in the means available to them. For ex-
ample, an institution combating freshwater pollution may set standards that limit
the emission of particular pollutants from particular sources. It may also employ
water-quality standards that provide addressees with considerable freedom of how
achieve them. Moreover, an environmental institution may not define specific limit
values but set procedural obligations such as the requirement to conduct an environ-
mental impact assessment prior to the implementation of certain projects or to open
the permission procedure to the public (chapter 9). Likewise, one institution may
rely almost exclusively on ‘‘soft’’ international law, while another institution resorts
to ‘‘hard,’’ legally binding international law (Abbott and Snidal 2000). The Euro-
pean Union even controls supranational law that is, under certain conditions, di-
rectly applicable within the legal systems of the member states and supported by
the particularly stringent supervisory and judiciary mechanism of the EU (Burley
and Mattli 1993; Alter 2000; Craig and de Búrca 2002). Finally, some institutions
are capable of linking sincere implementation of environmental obligations to the
granting of financial or other benefits, while others are not.
Interaction of the Additional-Means type will regularly raise the effectiveness of
both institutions involved. If the diffusion of an obligation from one institution
to another one with identical objectives and memberships activates an additional
means of implementation, it will support the effectiveness of the target institution at
the behavioral and impact levels. Without the influence of the source institution, the
target would neither have adopted nor implemented the obligation. Since the new
obligation is, by definition, in line with the target institution’s own objective, imple-
menting the obligation will support this objective by activating the target’s means of
implementation in addition to the means of the source institution. At the same time,
activating an Additional Means automatically contributes to a more effective imple-
mentation of the source institution. An increased effectiveness of one institution will
simultaneously reinforce the effectiveness of the other institution, because, according
to the ideal type, objectives, and memberships of the two interacting institutions are
similar or identical. The adoption of the obligation in the target institution can thus
be expected to trigger a case of Behavioral Interaction that feeds back on the source
institution. The interaction between the regime for the protection of the Northeast
Atlantic (OSPAR) and the International North Sea Conferences established in the
1980s may serve as an illustration. While relying on declarations that were formally
nonbinding, the Conferences took place at a high political level and generated polit-
ical pressure. Having politically agreed on the phaseout of certain substances and
activities, it became difficult for the same countries to resist the adoption of substan-
tively identical hard-law obligations within the framework of OSPAR. These hard-
law obligations provided a more effective basis for the protection of the Northeast
Atlantic. Their implementation, in turn, automatically helped achieve the goals of
the North Sea Conferences declarations (chapter 5).
Because of its synergistic effects on the source and the target institution, interac-
tion activating Additional Means allows actors operating within the source institu-
tion to enhance the effectiveness of international and EU governance. This type of
interaction may have its highest potential within the same policy field where objec-
tives of source and target institution are most likely to coincide, but its occurrence
across the boundaries of policy fields is not excluded. Actors and groups of actors
in international and EU institutions can employ the mechanism intentionally to fur-
ther and reinforce their objectives. As in the case of Nested Institutions, interaction
activating Additional Means provides opportunities for forum shopping because
actors can choose in which of the institutions available to launch a particular regu-
latory initiative. As the interaction between the North Sea Conferences and OSPAR
illustrates, actors may even establish an institution in an area already governed by
another institution for the sole purpose of triggering interaction activating Addi-
tional Means (chapter 5). In principle, synergistic effects may also be enhanced
through direct coordination between the interacting institutions.
Cases of Interaction through Commitment in the Sample

Our sample comprises sixty-six cases of Interaction through Commitment. Most of
them fall into the category of Additional Means (forty-two), while roughly one-third
are Jurisdictional Delimitation cases (twenty-one) and only three are of the Nested
Institutions type. Partly, this distribution is due to the fact that the many vertical
cases of implementation of international obligations into EU law were counted as
interaction activating Additional Means because they mobilize particularly hard
supranational law.
In line with the logic of the ideal types, disruption prevails in cases of Jurisdic-
tional Delimitation, while the other two types predominantly produced synergistic
effects on the target institution. The relevance of the correlation between quality of
effect and type of Interaction through Commitment is confirmed by a Cramer’s V of
0.483. Cases of Jurisdictional Delimitation did primarily involve institutions from
different policy fields, while the types of Additional Means and Nested Institutions
appear to have their highest potential within the same policy field where objectives
of source and target institution are most likely to be synergistic. This finding is in
line with the logics of the ideal types, but it also reflects the overall correlation be-
tween the variables quality of effect and policy field. In light of the overall quantita-
tive results presented in the first part of this chapter, it is also hardly surprising that
policy responses occurred far more frequently in the case of Jurisdictional Delimita-
Table 13.8
Cases of Interaction through Commitment in our sample
Jurisdictional Nested Additional
delimitation institutions means Total
No. % No. % No. % No. %
Totals 21 100 3 100 42 100 66 100
Horizontal international 16 76.2 1 33.3 7 16.7 24 36.4
Horizontal EU 4 19.0 0 0.0 14 33.3 18 27.3
Vertical 1 4.8 2 66.7 21 50.0 24 36.4
Synergy 1 4.8 3 100 32 76.2 36 54.5
Disruption 13 61.9 0 0.0 3 7.1 16 24.2
Neutral/unclear 7 33.3 0 0.0 7 16.7 14 21.2
Intentional 4 19.0 0 0.0 25 59.5 29 43.9
Unintentional 16 76.2 3 100 17 40.5 36 54.5
Uncertain 1 4.8 0 0.0 0 0.0 1 1.5
Same policy field 2 9.5 3 100 37 88.1 42 63.6
Different policy field 19 90.5 0 0.0 5 11.9 24 36.4
Response 16 76.2 1 33.3 8 19.0 25 37.9
No response 5 23.8 2 66.7 34 81.0 41 62.1
tion than in cases of the other two types of Interaction through Commitment (see
table 13.8).
Of our twenty-one cases of Jurisdictional Delimitation, thirteen had disruptive
effects, seven produced neutral or unclear effects on the target institution, and only
one generated synergy (table 13.8). Many of the disruptive cases involved interac-
tion between the WTO and international environmental institutions, which is in-
tensely discussed in the literature (see the overview in Brack 2002). They refer to
the pending conflict regarding the admissibility of trade restrictions employed by en-
vironmental agreements either as sanctions to reinforce compliance, as in the case of
the Montreal Protocol for the protection of the ozone layer, or as direct instruments
for the protection of the environment, such as in the case of the Cartagena Biosafety
Protocol on genetically modified organisms (chapter 8). The only synergistic case of
Jurisdictional Delimitation in our sample can be attributed to exceptional circum-
stances, namely a common interest of both institutions paired with grossly asymmet-
ric power relations between them. The climate change regime requested the Global
Environment Facility (GEF) to operate its financial mechanism (Werksman 1996;
Fairman 1996). Because financial assistance was part of the package that developing
and developed countries agreed on under the climate change regime, there was a
need to determine a financial mechanism through which resources would be chan-
neled. Assigning the financial mechanism to the GEF was beneficial for both institu-
tions involved. The GEF acquired a new task within its area of competence, while
the climate change regime spared the task of establishing a separate funding mecha-
nism. The situation was dominated by the climate change regime, where the finan-
cial mechanism was negotiated, while the GEF could merely offer its assistance, with
no leverage at all to support this offer. Under these circumstances, conflict could not
arise.
Jurisdictional Delimitation issues appear to arise more rarely within the EU than
at the international level. While sixteen of the twenty-one Jurisdictional Delimitation
cases are located at the international level, only four cases occurred at the EU level
(table 13.8). As in the case of Requests for Assistance, this may be due to the partic-
ular institutional structure of the EU. The existence of central actors that oversee the
legislative processes as a whole—most importantly the European Commission, but
also the Council of Ministers and the European Parliament—allows the EU to
address these issues in a more centralized and coordinated fashion already in the
process of elaborating legislation. Many Jurisdictional Delimitation issues may thus
be expected to be resolved before they reach the level of open conflict. Nevertheless,
such conflicts occur even within the highly integrated institutional framework of the
EU. A particularly intense political conflict characterized the interaction between the
EU Deliberate Release Directive on genetically modified organisms and EU product
sector legislation as well as the EU Pesticides Directive regarding the authority to
apply specific risk-assessment procedures (chapter 11). Also, the revised Deliberate
Release Directive did not contain any liability regime because this aspect was
expected to be covered in a separate EU directive on environmental liability,
whereas the liability approach of the latter instrument proved to be significantly
less stringent than was originally assumed (chapter 11). In the last case, new legisla-
tion was used to change existing one, when a strict interpretation of the 1979 Birds
Directive by the European Court of Justice led EU member states to soften the rele-
vant provisions through the Habitats Directive (chapter 10).
Vertical cases of Jurisdictional Delimitation occur even more rarely (table 13.8).
This may be explained by the fact that the EU is a subject of international law
alongside nation states. We did not consider tensions between EU legal instruments
and international institutions as a struggle over competences but instead as imple-
mentation difficulties on the side of the EU as a member of the international institu-

tion. Consequently the only case in our sample has very particular characteristics.
EU policymakers attempted to export the precautionary approach towards geneti-
cally modified organisms enshrined in the EU Deliberate Release Directive to other
international institutions that are capable of indirectly influencing the interpretation
of WTO rules (chapter 11).
Whereas most cases of Jurisdictional Delimitation were triggered unintentionally
by the source institution, some were intentionally employed in order to affect deci-
sion making of the target institution (table 13.8). All cases involving the WTO
occurred unintentionally. The WTO does not support free trade in order to interfere
with a number of environmental regimes. Some of the intentional cases of Juris-
dictional Delimitation closely resemble Requests for Assistance. Members of the
climate change regime requested the International Civil Aviation Organization
(ICAO) and the International Maritime Organization (IMO) to regulate greenhouse
gas emissions from international aviation and shipping. While an open conflict over
jurisdictional competence has not yet arisen, this case differs from the model of a
Request for Assistance in that it is not driven by information but by the capacity of
the climate change regime to pass relevant (competing) legislation. This implicit
threat has been a significant motivation for the efforts of ICAO and IMO so far
(chapter 3; Oberthür 2003).
Over time, a delimitation of jurisdictions was achieved in most cases of Jurisdic-
tional Delimitation contained in our sample. The institutions succeeded in balancing
their competing objectives to a large extent. For example, parties to the Montreal
Protocol and other environmental agreements waived trade restrictions for non-
parties that undertook to comply with the international regulations so as to avoid
conflict with WTO rules. How, and in favor of which of the institutions involved,
the balance was struck was a function of the power relation between the institu-
tions, as reflected in the credibility of their threats, the relevance of the ‘‘sticks’’
available to them, and the general importance and political weight of their issue
areas and underlying objectives.
The small number of three cases of interaction between Nested Institutions in our
sample renders the results of a general evaluation highly preliminary. Although our
sample does not contain intentionally triggered cases of this ideal type, their occur-
rence cannot be excluded. In line with the logic of the type we found only cases with
synergistic effects on the target institution. Our three cases also illustrate that inter-
action between Nested Institutions can occur either between a regional and a global
institution or between an EU legal instrument and an international institution. The

nesting of the EU Habitats Directive in the pan-European Bern Convention on the
Conservation of European Wildlife and Natural Habitats is analyzed in detail in
chapter 10. Here, the development of the Natura 2000 program of natural protec-
tion sites under the EU Habitats Directive contributed significantly to the incor-
poration of the concept into the Bern Convention in the form of the pan-European
Emerald Network.
While it is not accidental that our sample does not contain a case of interaction
between Nested Institutions at the EU level, this type might become more relevant
within the European Union in the future. So far, all EU legal instruments apply to
all member states so that their memberships do not differ. However, since the
Amsterdam Treaty of 1998 an opportunity exists for smaller groups of member
states to accelerate and deepen their cooperation within the institutional framework
of the EU in the form of the so-called ‘‘enhanced cooperation’’ (originally ‘‘closer
cooperation’’), which has not been employed to date (Bär, Homeyer, and Klasing
2002). Principally, a subset of EU member states could ‘‘enhance’’ their cooperation
in order to broaden the acceptance of a measure over time, if it is not acceptable to
all member states at once.
Cases of interaction activating Additional Means occurred at all three levels (table
13.8). At the international level, legally binding treaty law provided guidance to the
generation of influential scientific advice. The inclusion of the precautionary princi-
ple in the UN Fish Stocks Agreement (FSA) enabled the biologists involved in the
International Council for the Exploration of the Sea (ICES) to use this principle as
the basis of their advice on sustainable yield of fish stocks, on which domestic
protection measures are based. Furthermore, the fact that the global fisheries regime
had identified fisheries subsidies as a problem leading to undesired fishing over-
capacity helped place the issue on the agenda of the WTO, which has particular
enforcement mechanisms at its disposal (chapter 6). Examples at the EU level pri-
marily consist in an outright referral to obligations enacted elsewhere. For exam-
ple, the IPPC Directive refers to a number of environmental standards established in
other instruments. As a result, the IPPC Directive provides an additional means of
supervision and enforcement of these existing obligations (chapter 9). Likewise, the
EU Environmental Liability Directive defines environmental damage with reference
to existing instruments of EU environmental law such as the Habitats Directive and
the Water Framework Directive (chapters 9–10). While these Directives facilitated
the elaboration of a definition of damage under the Liability Directive, the latter
adds another means (liability) for their enforcement.
Vertical interaction of the Additional-Means type is particularly frequent, ac-

counting for 50 percent of the cases in our sample (table 13.8). This comparatively
large figure includes cases related to the implementation of international agreements
by the EU, which is, together with the EU member states, a party to many interna-
tional (environmental) agreements. For example, the Bern Convention prompted
preparation of the Habitats Directive (chapter 10), the Aarhus Convention on pub-
lic participation led to expanded public participation under the IPPC Directive
(chapter 9), and the EU passed a regulation restricting trade in endangered species
to implement CITES, even though it is, in contrast to its member states, not (yet) a
full member of CITES (chapter 7). Also, air-quality guidelines of the World Health
Organization have been incorporated into the daughter directives of the EU Air
Quality Framework Directive (chapter 12). Such domestic implementation by mem-
bers of an international regime may not appear particularly relevant for the study
of institutional interaction. In all these cases, however, transferal of international
obligations into supranational European law activates the particularly stringent su-
pervisory and enforcement mechanism of the EU.
While interaction activating Additional Means resulted in synergy in three-fourths
of our cases, it may also lead to neutral or even disruptive effects on the target insti-
tution (table 13.8). Especially where international commitments are implemented in
EU legal instruments belonging to other policy fields or pursuing different policy
objectives, tensions may occur. For example, international fisheries rules support a
‘‘greening’’ of EU fish-import rules and a reduction of EU fisheries subsidies that
are hardly compatible with the original objectives of these EU policies (chapter 6).
Furthermore, the Convention on Biological Diversity required restrictions on the
ability to grant patents for living organisms that were at odds with the original
objectives of the EU Patent Directive and resulted in manifest conflict (chapter 4).
While no transfer of an obligation would presumably occur under these circum-
stances horizontally at the international level, the EU is more constrained because,
to the extent that it acts as a subject of international law, it is obliged to implement
international rules.
Behavioral Interaction
Behavioral Interaction is characterized by the fact that one institution directly influ-
ences the effective implementation of another institution it its issue area, rather than
its decision-making process. Whereas the two causal mechanisms of Cognitive Inter-
action and Interaction through Commitment operate at the rule-making (output)
level, Behavioral Interaction is located within the issue areas governed by the institu-
tions involved (outcome level). It will occur, if behavioral changes triggered by
the source institution are also relevant for the performance of the target institution
within its issue area.
Behavioral Interaction is characterized by a high ability of the source institution
to influence the target unilaterally. In contrast to interaction at the output level, it
occurs through the adaptation of actors’ behavior beyond the decision-making pro-
cess of the source institution and does not depend on a decision within the target
institution. The effect on the target might even come about unnoticed by the mem-
bers of either or both institutions, because it emerges as the aggregate result of the
uncoordinated behavior of actors within the two issue areas involved. A collective
decision adopted within the target institution, or the source institution, in response
to the effects of Behavioral Interaction is possible but does not constitute an essential
element of this causal mechanism, because the effect will also occur without a policy
response.
It proved difficult to identify Weberian ideal types of Behavioral Interaction that
are characterized by distinct logics. One could group the cases of Behavioral Inter-
action within our sample into classes distinguished by certain of our core charac-
teristics introduced in the first part of this chapter. Cases vary as to whether or
not they have been triggered intentionally by the source institution, whether they
resulted in synergy or disruption for the target institution, and so on. Unfortunately,
none of these groupings revealed distinct underlying rationales or distinct appear-
ances of Behavioral Interaction. For example, whether a case of Behavioral Interac-
tion is triggered intentionally by the source institution or not, does not significantly
affect the causal pathway through which influence must pass, nor its result.
Our inability to identify meaningful ideal types of Behavioral Interaction is related
to the fact that cases driven by this causal mechanism are almost entirely controlled
by the source institution. Cases of Cognitive Interaction and of Interaction through
Commitment require activity on both sides of the causal pathway, including a deci-
sion of the target institution. Cognitive Interaction requires that the source institu-
tion produces some information that influences decision making in the target
institution, while in cases of Interaction through Commitment an obligation collec-
tively agreed on in the source institution affects decision making in the target. In
both cases, the distinction between ideal types is based on systematic variation in
factors that determine how action on both sides of the causal pathway relates and
comes about. In contrast, Behavioral Interaction depends exclusively on activity at
the source side of the causal pathway, while the effect does not need any particular
activity at the target side to occur. Therefore, all cases are based on the same funda-
mental rationale, so that it is unlikely that different ideal types can be identified.
However, the universe of cases of Behavioral Interaction might reveal different
rationales if considered in connection with other cases of interaction that are regu-
larly located at the output level. Three of the five types identified above, namely
the Request for Assistance, Jurisdictional Delimitation, and Additional Means, are
meant to trigger subsequent, or respond to preceding, cases of Behavioral Interac-
tion. Hence, many, though not all, cases of Behavioral Interaction constitute parts
of longer causal chains comprising at least two separate cases of interaction. We dis-
cuss this broader phenomenon in the next section.
To be sure, we cannot exclude that the core characteristics developed in the begin-
ning of this chapter might simply not be the appropriate ones for identifying mean-
ingful ideal types of Behavioral Interaction. Other core characteristics might have
provided a more fruitful foundation in this respect. For example, we did not elabo-
rate systematically on the different types of actors that may be involved in the causal
pathways. Interaction may involve behavioral changes of member states of the insti-
tutions involved, other states, nonstate actors like nongovernmental organizations,
or relevant economic actors like companies. However, we do not have any positive
indication that these differences might generate distinct rationales.
In contrast to the five ideal types of institutional interaction at the output level
developed above, Behavioral Interaction is indifferent to the quality of its effects.
As the mechanism does not indicate how precisely interaction affects the perfor-
mance of the target institution within its own domain, we cannot derive any mean-
ingful hypotheses about whether we might expect predominantly synergistic or
disruptive effects. We might speculate whether intentionally created Behavioral In-
teraction produces a systematically different quality of effects than unintentionally
triggered interaction, or whether interaction across the boundaries of policy fields
is different from interaction within a single policy field. However, all these assump-
tions are not derived from the causal mechanism. Hence, we may expect that Behav-
ioral Interaction will produce significant synergistic effects on the target institution
in some cases and disruptive effects in others, with no systematic pattern to be
derived from the rationale of the causal mechanism.
While Behavioral Interaction may be employed intentionally to influence the effec-
tiveness of another institution, it rarely needs to be accepted as given and can regu-
larly be influenced by political decision making. In many instances, the members
of the source institution will endeavor to create synergy, if only because they are
simultaneously members of the target institution. This will be generally true for a
policy field like EU and international environmental governance. However, the
causal mechanism may equally well be employed to undermine the effectiveness of
the target institution. This will be predominantly the case if mutually incompatible
objectives are pursued and a conflict over jurisdictional authority prevails. Inten-
tionally created disruption is likely to result in conflict with the members of the tar-
get institution. Options for enhancing synergy or mitigating disruption will in most
cases exist within the source institution that triggered the interaction. Actors operat-
ing within the target institution may also be able to take effective action, because the
institution possesses regulatory authority over its own affected domain. Depending
on the quality of the effect, they can attempt to reduce or increase the susceptibility
of their institution for external influence.
Cases of Behavioral Interaction in the Sample

Behavioral Interaction is a widespread phenomenon at all levels of interaction. It
drives about half of our cases. Virtually all of the institutions explored in our project
are both the source of behavioral effects on other institutions and the target of such
effects originating from other institutions. Moreover, the cases of Behavioral Inter-
action in our sample are roughly evenly divided between horizontal interaction at
the international level, horizontal interaction at the EU level, and vertical interaction
between EU instruments and international institutions (table 13.9).
Intentionally triggered Behavioral Interaction is particularly frequently used as a
governance instrument at the EU level, resulting in a network of mutually reinforc-
ing sectoral legal systems (table 13.9). This may be due to the important strides to-
ward policy integration that EU environmental policymaking has made in recent
years (Lenschow 2002). As a result, more encompassing instruments such as the
Water Framework Directive as well as crosscutting instruments, such as the Envi-
ronmental Liability Directive and the IPPC Directive, integrate environmental
legislation in a specific sector or add supervision and enforcement mechanisms ap-
plicable across several instruments. Consequently, many cases of intentional Behav-
ioral Interaction at the EU level in our sample concern these more encompassing and
crosscutting instruments. For example, the Water Framework Directive is partially
intended to assist implementation of the Nitrates and Bathing Water Directives.
The IPPC Directive is partially intended to assist implementation of the Water
Framework Directive, Air Quality Framework Directive, Habitats Directive, and
Table 13.9
Cases of Behavioral Interaction in our sample
Unclear
Uninten- intention-
Intentional tional ality Total
No. % No. % No. % No. %
Totals 24 100 53 100 4 100 81 100
Horizontal international 4 16.7 16 30.2 1 25.0 21 25.9
Horizontal EU 15 62.5 12 22.6 3 75.0 30 37.0
Vertical 5 20.8 25 47.2 0 0.0 30 37.0
Same policy field 19 79.2 35 66.0 3 75.0 57 70.4
Different policy field 5 20.8 18 34.0 1 25.0 24 29.6
Synergy 22 91.7 31 58.5 3 75.0 56 69.1
Disruption 2 8.3 21 39.6 1 25.0 24 29.6
Neutral/unclear 0 0.0 1 1.9 0 0.0 1 1.2
Response 6 25.0 26 49.1 1 25.0 33 40.7
No response 18 75.0 27 50.9 3 75.0 48 59.3
Birds Directive (chapter 9). Also, the Environmental Liability Directive is intended
to support implementation, inter alia, of the Deliberate Release Directive, the Habi-
tats Directive, and the Water Framework Directive (chapters 9–11). Relevant efforts
at the international level are less advanced and may meet with more difficulties than
in the institutionally relatively coherent EU framework.
Intentional Behavioral Interaction is less common at the international level and
in vertical interaction in our sample (table 13.9). All vertical cases concern EU legis-
lation implementing international obligations; generally, they follow from previous
interaction of the Additional Means type. For example, the Habitats Directive
supports performance of the Bern Convention (chapter 10), the Water Framework
Directive deliberately supports the performance of both the Convention on Trans-
boundary Watercourses and Lakes and OSPAR (chapter 9), and several EU legal
instruments are designed to support implementation of the climate change regime
(chapter 3). The cases at the international level also follow from earlier interaction
at the output level. For instance, activities of Interpol and the World Customs Orga-
nization in relation to trade in endangered species were entirely designed to assist
implementation of CITES—and as such had themselves been the result of a Request
for Assistance (chapter 7).
The quality of effect varies significantly with intentionality. In the case of both in-
tentional and unintentional Behavioral Interaction, synergistic cases in our sample
outnumber cases with disruptive effects on the target. However, about 40 percent
of the cases of unintentional Behavioral Interaction resulted in disruption, while the
ratio was less than 10 percent for the intentional cases (table 13.9). This may reflect
the deep-rooted interdependence of international and EU environmental institutions,
which makes it difficult to avoid side effects. At the same time, deliberately under-
mining other institutions might be exceptional as far as environmental protection
is concerned. The only two cases of intentionally disruptive Behavioral Interaction
in our sample occurred at the EU level. While this might be surprising at first glance,
disruptive behavioral effects may be created deliberately as a substitute for the
amendment of the target institution in particular within the institutionally dense
and integrated framework of the EU. EU member states used the EU Habitats Direc-
tive of 1992 to weaken the implementation of the 1979 Birds Directive following a
ruling of the ECJ that established a rather strict interpretation of the Birds Directive
not intended by the member states (chapter 10). The EU Deliberate Release Direc-
tive intentionally requested application of a more stringent risk-assessment proce-
dure under EU product-sector legislation that disrupted the latter’s objectives in a
politically charged situation concerning the marketing of genetically modified prod-
ucts (chapter 11). In both cases, it appeared easier to integrate the desired changes
into the new directives that were under discussion anyway than to initiate amend-
ments of the target institutions. Members of source and target were identical and
influence occurred hardly against the will, or at least with the knowledge, of the
members of the target institution. Intentional Behavioral Interaction with disruptive
effects may thus reflect the integrative character of the legal framework of the EU
rather than a higher potential for conflict.
As expected, Behavioral Interaction does not have a clear-cut pattern as to the
quality of effects produced. Roughly reflecting the distribution of cases in our over-
all sample, more than two-thirds of the cases of Behavioral Interaction display syn-
ergistic effects on the target institution, while about 30 percent are disruptive. The
differences between the three levels of horizontal and vertical interaction largely cor-
respond to the differences with respect to intentionality. Thus, both unintentional
and disruptive Behavioral Interaction is more frequent vertically and at the interna-
tional level than at the EU level (compare tables 13.9 and 13.11). Variation between
Behavioral Interaction occurring within the same policy field and across different
policy fields also largely reflects the situation with respect to our overall sample.
Table 13.10
Cases of Behavioral Interaction in the same and in different policy fields
Same policy Different policy
field fields Total
No. % No. % No. %
Synergy 49 86.0 7 29.2 56 69.1
Disruption 7 12.3 17 70.8 24 29.6
Neutral/unclear 1 1.8 0 0.0 1 1.2
Total 57 100 24 100 81 100
Cramer’s V 0.586
Table 13.11
Quality of effect of the cases of Behavioral Interaction in our sample
Neutral/
Synergy Disruption unclear Total
No. % No. % No. % No. %
Totals 56 100 24 100 1 100 81 100
Horizontal international 13 23.2 7 29.2 1 100 21 29.5
Horizontal EU 23 41.1 7 29.2 0 0.0 30 37.0
Vertical 20 35.7 10 41.7 0 0.0 30 37.0
Response 11 19.6 21 87.5 1 100 33 40.7
No response 45 80.4 3 12.5 0 0.0 48 59.3
More than 80 percent of the cases of Behavioral Interaction within a single policy
field resulted in synergy, while about 70 percent of the cases involving different pol-
icy fields had disruptive effects. The Cramer’s V of 0.586 indicates a relevant corre-
lation between both variables (table 13.10).
The frequency of responses to cases of Behavioral Interaction varies with the qual-
ity of effects. In accordance with our general empirical findings, the target institution
tends to reap the benefits without taking further action, especially in the case of
synergy (table 13.11). This helps us understand the higher level of policy responses
to unintentional Behavioral Interaction, which results in disruption more frequently
than intentional Behavioral Interaction (table 13.9). Response action might improve
the situation and may even transform originally disruptive effects into (at least
partially) synergistic ones. For example, the EU Structural Funds originally under-
mined the effectiveness of the Habitats Directive because they funded development
projects that encroached on protected natural habitats. Therefore, Structural Fund
rules were adapted so as to make funding dependent on the proper designation of
habitat-protection sites. The threat to withhold funding now serves as an implicit
sanction to enforce the Habitats Directive (chapter 10).
Where to Go from Here?
Apparently, this volume has not resolved all the issues of institutional interaction; it
may even have raised more questions than it was able to answer. Given the state of
development of research in the field so far, our project had to be in large measure
exploratory. It has in particular made three contributions to the study of institu-
tional interaction. First, the conceptual approach can be employed to analyze and
compare interaction involving both international institutions and EU sectoral legal
instruments at large. Second, we have developed causal mechanisms and ideal types
that drive institutional interaction in general. Third, the project has produced empir-
ical findings on the nature of institutional interaction that may be checked by future
studies. Overall, institutional interaction has turned out to be a more multifaceted
phenomenon than the existing literature suggested. But many interesting research
questions are still waiting to be addressed. In this section, we identify two strands
of promising future research on institutional interaction. The first relates to the fur-
ther development of the analytic approach presented in this volume. The second im-
portant area for future research is the systematic analysis of more complex settings
of institutional interaction.
Toward Further Refinement of the Analytic Approach of This Volume

Future research may be directed at deepening our understanding of particular ele-
ments of institutional interaction, of which we highlight four here. First, future
studies may investigate in more detail the conditions under which the causal mecha-
nisms of institutional interaction are triggered and become operational. Our own
research focused on actual cases of interaction, because we were interested in
enhancing our knowledge of how interaction operates and how it is dealt with po-
litically. We did not delve into the issue of under which conditions interaction
occurs and does not occur. This question may be particularly relevant for interac-
tion at the output level, because the target institution must be motivated to adopt a
decision in order to establish a case of interaction. We can expect to identify situa-

tions in which a potential for interaction exists but in which the target does not re-
spond, so that there are no effects to be observed. Comparing noncases with actual
cases of interaction may help specify under what conditions a potential for interac-
tion is realized. These conditions may vary between the different causal mechanisms
and ideal types.
Second, more detailed studies might investigate the factors that influence the
success or failure of managing cases of institutional interaction politically. In this re-
spect, a first task will consist in developing a clear distinction between, and indica-
tors for, the more or less successful political management of different kinds of
institutional interaction. It might be explored under which conditions an issue of
Jurisdictional Delimitation or a case of unintentional Behavioral Interaction can be
said to have been dealt with successfully. A related subject of relevance both for ana-
lysts and policymakers is the exploration of the role of different kinds of actors, such
as governments, nongovernmental organizations, intergovernmental organizations,
entities like the European Commission, and secretariats of international institutions,
in triggering, and responding to, issues of institutional interaction. Studying how
institutional interaction is managed at the EU level may provide useful lessons for
the management of interaction at the international level. We observed that the EU
has developed particular mechanisms for addressing and managing potential inter-
action issues. We have also been able to link these particularities to the specificities
of the supranational governance system of the EU, in particular to the existence of
the European Commission. A closer exploration of Commission activities might
reveal the specific functions that are relevant for the successful management of par-
ticular types of interaction issues. In a second step it might be possible to identify
functional equivalents that may be employed at the international level—given that
the creation of supranational actors appears to be out of reach for most, if not all,
international institutions.
Inquiry into the conditions under which interaction occurs and its political
management is more or less successful can be based either on a bottom-up or a
top-down approach. Research may start from empirical cases and noncases of
interaction, trying to generalize empirical conditions under which particular causal
mechanisms become operational and political management is effective (bottom up).
Research may also be based on deductive reasoning to identify the relevant con-
ditions and restrictions in theory, which may subsequently be tested by empirical
analysis (top down).
Third, future research might attempt to assess the precise impact of institutional
interaction on the ultimate target of governance, in our case the environment. This
volume focused on cases of interaction with a significant impact on the environ-
ment, or at least the potential for creating such an impact. We refrained from deter-
mining this impact in more detail due to the immense additional research effort
involved in such an assessment as well as the methodological challenges to be
addressed and the uncertain outcome. As in the case of research on the effectiveness
of international and EU institutions (Underdal 2004), however, assessing the impact
on the environment (or other ultimate targets of governance), possibly from a com-
parative perspective, is of obvious relevance for the selection of political priorities in
addressing issues of institutional interaction. It would also enable us to weigh the
impact of varying cases of institutional interaction and to address the question of
whether the aggregate effect of all or a specific subset of interaction cases is positive
or negative (see Hovi, Sprinz, and Underdal 2003).
Finally, empirically robust knowledge about institutional interaction will require
the systematic expansion of the empirical basis on which results rest. The empirical
basis of our project has remained limited. Future research may corroborate or mod-
ify our findings on the basis of more interaction cases involving other environmental
institutions. Reliability of results may also be supported by exploring cases of inter-
action relevant for the efficiency of governance or having longer, more indirect
causal chains—taking into account the concomitant methodological challenges in
demonstrating causality. Expanding the empirical research base to other policy
fields might prove particularly fruitful. While interaction phenomena occur in all
areas of an increasingly densely populated realm of international and EU institu-
tions, institutional frameworks, actors, objectives, and governance approaches dif-
fer. In security relations or in the development-aid sector, for example, governance
relies less on regulation and rule making than on collective action and financial as-
sistance. In other fields such as international trade, the institutional landscape is far
less fragmented than in environmental policy. Future research may explore the rele-
vance of these varying conditions for institutional interaction.
Toward Systematic Analysis of More Complex Interaction Phenomena

Complexity raises a totally new set of issues. Single cases of interaction do not occur
in isolation from each other. Frequently, interaction situations will be more complex
than the analysis of single cases suggests. Eventually, we will not be content with
knowing how and why a particular case of institutional interaction matters. Com-
plex situations raise the problem of ‘‘emergent’’ properties. They may be affected by
so many cases of interaction in such unexpected ways that new properties emerge
that are not inherent in the single cases. There may be a forest to examine in addi-
tion to the many trees it is made of. Therefore, emergent properties will be difficult
to grasp by exclusive analysis of the units.
The analytic concept of this volume can be employed to systematically explore
more complex settings and their emergent properties by recombining individual
cases. While the emergent properties of the forest are related to the coexistence of
the many trees and to their mutual influence on each other’s existence, possible
emergent properties of complex interaction situations will originate from the co-
existence of several, or many, cases of interaction. In essence, the investigation of
institutional interaction in this volume is based on the analytic disaggregation of
complex situations into separate cases with a limited scope and a clear-cut direction
of influence (chapter 2). This approach assumes that complex situations will usually
not only be composed of several cases, but that these cases also differ in some re-
spect (e.g., regarding source and target institution, causal mechanism, quality of
effect, and so on). The systematic recombination of cases would allow us to retain
the single case of interaction as our principal unit of analysis, while exploring more
complex interaction situations and their typical patterns.
In the following, we explore two particular ways individual cases may be related
so as to form more complex interaction situations, namely causal chains and clus-
ters. Cases of interaction may form sequential chains so that an individual case gives
rise to a subsequent case that feeds back on the original source institution or influ-
ences a third institution. Cases of interaction may also cluster around certain issues
and institutions. In this case, several institutions jointly address a particular problem
and contribute to the effectiveness of governance of a certain area.
Causal Chains Two or more cases of interaction will form a causal chain, if one
case triggers one or several subsequent cases. For this effect to occur, it is not suffi-
cient that two or more institutions influence each other in more than one case. The
original case must have changed the situation within the original target institution
sufficiently, so that the latter becomes the source of a further case of interaction
directed at the original source institution or at a third institution. Causal chains
draw attention to the fact that interaction processes in more complex settings may
acquire a momentum of their own, so that the actors and institutions involved are
drawn into an autonomous process that they do not fully control any more.
As a first step, we can identify several ideal types of causal chains of interaction
between two institutions, which have their own logics and restrictions. First, cases
of interaction at the output level may give rise to further cases of output-level in-
teraction in the reverse direction and thus start sequential coevolution processes.
Whereas we might think of numerous combinations, in particular two typical pat-
terns involving the ideal types of Interaction through Commitment can be expected
to develop.
Coevolution processes may lead to a mutually reinforcing and synergistic develop-
ment of the normative structure of the institutions involved. Such chains will most
probably be composed of cases of Nested Institutions and Additional Means. These
types are in significant respects complementary and mostly create synergistic effects
between the institutions involved. They cannot be expected to prompt a second case
of interaction in the reverse direction, however, as long as the target merely adapts
to the normative change originating from the source institution. Only if they give
rise to more far-reaching innovations within the target institution can a feedback
case influencing the original source institution occur. For example, the Bern Conven-
tion drove the preparation of the EU Habitats Directive. As the Directive was elabo-
rated within the EU, the concept of the Natura 2000 network of natural protection
sites was developed, which subsequently triggered the adoption of a similar concept
(Emerald Network) within the Bern Convention (chapter 10). Whereas the first case
of this chain is one of Additional Means, the second is one of Nested Institutions.
An Additional-Means case may also lead to a second Additional-Means case, or it
may follow from a Nested Institutions case.
Causal chains composed of two or more cases of Interaction through Commit-
ment may also lead to the gradual separation of the jurisdictions of two or more
institutions. These chains will typically be composed of Jurisdictional Delimitation
cases, which regularly produce disruptive effects on the target institution. Think of
a decision within the source institution, which interferes with the governance activ-
ities of the target institution. The case will be complete once the target institution
reacts to the changed situation. Once again, this reaction will not set off a feedback
case as long as it merely consists of an adaptation to the jurisdictional claim of the
original source institution. However, if the reaction is sufficiently far-reaching, it
may change the constellation of interests within the original source institution so
that the latter makes a further decision. For example, the WTO, especially through
its Agreement on Sanitary and Phytosanitary Measures (SPS Agreement), con-
strained the use of trade measures related to genetically modified organisms under
the Cartagena Protocol on Biosafety. Negotiations on the Cartagena Protocol, in

turn, precluded that detailed rules on the appropriate risk assessment for genetically
modified organisms were developed within the WTO (chapter 8; see also Oberthür
and Gehring, forthcoming). Whereas this chain will ideally develop over time, its
component cases might also occur simultaneously, if anticipated changes within
one institution are immediately responded to by action within another institution,
which creates a feedback.
It is quite improbable that Cognitive Interaction would give rise to causal chains,
although this cannot be excluded. In cases of learning from a Policy Model, influ-
ence is very diffuse and nontargeted. Since the target institution may learn from
otherwise totally unrelated institutions, it would be completely incidental if this
learning were to cause immediate effects on the original source institution or on a
third institution. Requests for Assistance will usually be all too targeted to trigger
coevolution at the output level. Requests tend to focus on very specific and compar-
atively minor adaptations by the target institution, short of triggering broad changes
capable of causing feedback effects on the original source institution.
Second, in another family of causal chains, interaction at the output level causes
subsequent Behavioral Interaction. In these chains, the source institution causes
normative change within the target institution, which induces behavioral changes
of relevant actors within the target’s issue area, and these changes affect the perfor-
mance of the original source institution within its own domain. Three of our types
of output-level interaction—namely a Request for Assistance, interaction activating
Additional Means, and Jurisdictional Delimitation cases—are particularly suited to
triggering this kind of causal chain. An institution issues a Request for Assistance
not merely to induce a decision of the target institution, but ultimately to generate
a subsequent case of Behavioral Interaction to support performance within its own
domain. Whereas the first of these cases might have a neutral or indeterminate effect
on the original target institution, the second case will usually be synergistic for the
original source institution. For example, CITES asked Interpol and the World Cus-
toms Organization for adaptations that subsequently helped improve the effective-
ness of its own implementation (chapter 7).
Interaction activating Additional Means is also likely to create subsequent Behav-
ioral Interaction. The major difference vis-à-vis Requests for Assistance is that adap-
tation is not entirely based on cognition and conviction, but on the commitment
of actors to obligations originating from the source institution. Like a Request for
Assistance, interaction of the Additional-Means type will be meaningful only if the
additional means controlled by the original target institution helps improve the per-
formance of the original source institution. For example, OSPAR had the effect of
supporting the implementation of the declarations of the International North Sea
Conferences because the soft-law-based North Sea Conferences had previously
accelerated the development of hard obligations under OSPAR (chapter 5). The In-
ternational Council for the Exploration of the Sea (ICES) was able to support the
performance of the UN Fish Stocks Agreement (FSA), because the FSA had previ-
ously motivated ICES to base its estimates of sustainable fish catch on the precau-
tionary principle (chapter 6). In a vertical direction, CITES induced the adoption of
the EU Regulation on trade in endangered species, which subsequently supported
the performance of CITES (chapter 7). In all these instances, a synergistic (or at least
neutral) interaction at the output level causes a synergistic interaction at the out-
come level.
Although less obvious, Jurisdictional Delimitation cases may also create feedback
effects at the outcome level. Ideally, we may think of two rather different conse-
quences of a Jurisdictional Delimitation case on the target institution. Interaction
may lead to normative adaptation, which accommodates the two institutions and,
accordingly, mitigates existing or anticipated disruptive Behavioral Interaction. In
this case, subsequent Behavioral Interaction is not to be expected. However, Juris-
dictional Delimitation may also lead to normative adaptation within the target in-
stitution, which does not completely separate the jurisdictions of the institutions
involved. As a result, it is likely to cause Behavioral Interaction, which undermines
the performance of the original source institution within its own domain. For ex-
ample, the Cartagena Biosafety Protocol was adapted to the needs of the WTO,
especially the SPS Agreement, in a way that exploits the remaining room for
interpretation to a large extent. Thus, the implementation of the Protocol might
undermine the effectiveness of the WTO in the future (Oberthür and Gehring, forth-
coming; see also chapter 8).
The remaining two types of interaction at the output level are unlikely to trigger
subsequent Behavioral Interaction. The Policy Model is once again too diffuse to
trigger feedback cases of Behavioral Interaction in any systematic way. While a
causal chain starting with a Policy Model case cannot be totally excluded, it would
be merely incidental. Cases of the Nested Institutions type will also rarely be capable
of triggering Behavioral Interaction. If members of a regional institution transfer
obligations to a global institution, this cannot be expected to affect the behavior of
relevant actors at the regional level significantly, because these actors are already
bound by the same obligations. Feedback through Behavioral Interaction could arise
only if the transfer led to innovation in the target institution (e.g., the adoption of
stricter rules than in the regional institution) or if the behavior of nonregional actors
within the region could not be controlled by the regional institution itself.
Third, Behavioral Interaction may trigger subsequent interaction at the output
level. In this causal chain, the second case originates from a collective decision
made within the original target institution in response to a preceding Behavioral In-
teraction. For the feedback to occur, this decision has to influence decisions of the
original source institution. It can in particular take the form of a Request for Assis-
tance or raise a Jurisdictional Delimitation issue. While this causal chain is absent
from our sample, its occurrence is possible. Consider that the Kyoto Protocol under-
mines the Convention on Biological Diversity (CBD) by providing incentives for
establishing fast-growing tree plantations. If the CBD responded by adapting its
own rules to mitigate this effect, this would not initiate a second case of interaction.
However, if the CBD asked the climate change regime to change its rules or if it set
strict new rules in order to gain comprehensive jurisdictional competence over the
management of forests, it would create a subsequent case.
These types of causal chains certainly do not exhaust the possibilities of recom-
bining cases to form more complex settings. Thus, causal chains may involve more
than two cases. For example, the causal chains operating entirely at the output level
are rarely independent from Behavioral Interaction. For example, the Bern Conven-
tion not only drove the preparation of the EU Habitats Directive and was later af-
fected by the Directive’s concept of the Natura 2000 network of natural protection
sites. Because the Habitats Directive mobilized the additional means of suprana-
tional law with its particularly strong enforcement arrangements, the Bern Conven-
tion’s effective implementation was also supported (chapter 10). Likewise, mutual
adaptation of the norms enshrined in the Cartagena Protocol and the WTO (includ-
ing relevant interpretations by the WTO’s dispute settlement bodies) are accompa-
nied by mutual influence of the two institutions on each other’s performance within
their issue areas.
Causal chains that relate three or more institutions in systematic ways will also be
of interest. For instance, regulatory competition between WTO-TRIPS and the bio-
diversity regime has influenced a third institution, the Food and Agriculture Organi-
zation (FAO) (see chapter 4; Andersen 2002). Over time, the ‘‘Regime Complex for
Plant Genetic Resources’’ involves even more institutions (Raustiala and Victor
2004). Also, the conflicts between the WTO SPS Agreement and the EU Deliberate
Release Directive on genetically modified organisms regarding application of the

precautionary principle have affected the decision-making process within other rele-
vant international institutions that set standards that are relevant for the interpreta-
tion of the SPS Agreement, including the Codex Alimentarius Commission (chapter
11). The systematic recombination of appropriate causal chains of limited scope
may help develop a more rigorous analysis of such complex settings.
The analysis of causal chains of cases of institutional interaction points to the fact
that interinstitutional influence is not merely driven by the interests and action of the
actors involved. Institutional interaction can itself become an important driver. If
cases of interaction influence subsequent or parallel cases, institutional interaction
gains momentum. Thus, our approach of disaggregating complex settings into a
suitable number of cases and then recombining cases so as to ‘‘reconstruct’’ the
more complex setting is not confined to the mere aggregation of properties of the
units. It may also enable us to reveal typical patterns and ideal types of more com-
plex interaction situations that possess their own logics and rationales. These logics
would not be located within the units, but in the particular forms of their coexis-
tence, and would therefore allow us to grasp the emergent properties of the larger
situation. Admittedly, this issue has not been central to the empirical investigation
of the present volume and needs further elaboration.
Clusters Single cases of interaction can also cluster around certain issues and insti-
tutions. In this case, several institutions commonly address a particular problem and
contribute to the effectiveness of governance within a certain area. While causal
chains address causation between cases of interaction, clusters address settings of
parallel interaction without causation between cases. Problem areas may comprise
a considerable number of institutions, which separately address parts of the larger
problem. For example, the marine environment of the North Sea and the North At-
lantic is subject to the regulation of numerous institutions, including OSPAR, the
North Sea Conferences, the global London Dumping Convention, the MARPOL
Convention of the International Maritime Organization on marine pollution from
ships, institutions governing pollution of particular rivers such as the Rhine, as well
as a number of pertinent EU directives on water and air pollution. Likewise, the
problem area of the protection of nature and wildlife is populated by numerous dif-
ferent institutions, including CITES, the CBD, the Convention on Migratory Species,
the Bern Convention, and several EU instruments. We may identify similar clusters
in the area of air pollution (UNECE Convention on Long-Range Transboundary Air
Pollution, EU directives, IMO regulations, and so on) and beyond. Each single in-
strument focuses on a limited fraction of the overall problem, obliges varying groups
of states, and employs its own means. While this is beyond our effort here, in a
broader perspective the analysis may even extend to interaction with national-level
institutions.
While such clusters will usually be related to underlying problems as well as rele-
vant actors and their interests, they may follow typical patterns that reflect core
characteristics related to the cluster, not the single cases they are composed of. We
found deliberately created interaction clusters supported by crosscutting institutions
that were established in order to influence other specific institutions. For example,
the EU IPPC Directive, introducing an integrated scheme for the authorization of
large-scale industrial plants and comparable installations, cuts across numerous is-
sue areas governed by other EU directives, including the Water Framework Direc-
tive, the Air Quality Framework Directive, the Solvent Emissions Directive, and the
Habitats Directive. The IPPC Directive expressly relates to several of these instru-
ments by its very wording, while it directly affects others or is affected by them
(chapter 9). At the international level, the International North Sea Conferences
have been established not least to provide for a coordinating mechanism for various
institutions relevant for the protection of the North Sea and to set priorities for fu-
ture activities (chapter 5). The Global Environment Facility (GEF) also constitutes
the core of an intentionally created cluster of interacting international institutions.
It is designed to provide for a coordinated approach to the financial mechanisms of
global environmental agreements (Fairman 1996; Werksman 1996), including the
climate change regime, the CBD, and the 2001 Stockholm Convention on Persistent
Organic Pollutants.
Other clusters develop over time so that one of the institutions involved gradually
acquires a lead function. Such clustering around an emergent lead institution may
provide the opportunity for more centralized management of the system of interna-
tional institutions. However, it also creates an informal hierarchy of institutions that
is voluntarily accepted by those involved. For example, CITES has gradually devel-
oped to become the center of a cluster of institutions dedicated to the protection
of wildlife. In this process, CITES even reached beyond its original mandate, for
example by calling on members to join the subagreement on the protection of the
houbara bustard of the Convention on Migratory Species, even though this sub-
agreement is unrelated to CITES and does not address international trade in the spe-
cies. In other cases, CITES responded to requests of other international institutions
by addressing the protection of whales or the vicuña, a species of Andean llama.

CITES itself requested assistance from nonenvironmental institutions in implement-
ing its rules, in particular Interpol and the World Customs Organization (chapter 7).
In the 1990s, the CBD developed a similar approach by coordinating programs and
activities with pertinent regimes such as the Ramsar Convention for the protection
of wetlands (chapter 4). This cluster is even larger and includes CITES.
Our approach provides a basis for capturing emergent properties of such interac-
tion clusters. Recombining individual cases of interaction into clusters enables us to
discover typical patterns and derive their emergent properties. For this purpose, typ-
ical patterns and ideal types of the clusters would have to be developed from the
examples identified above, or from other suitable constellations of cases. Because
this is beyond the scope of our effort here, we have to leave this task to future re-
search efforts.
Toward the Study of Interlocking Structures of Governance Institutions

Eventually, we may want to study the broader consequences of institutional interac-
tion for international society at large, especially within the decentralized interna-
tional system. If governance within the international system, as well as within the
European Union, is not merely the product of more-or-less rationally designed
sector-specific institutions (Koremenos, Lipson, and Snidal 2001), but emerges
from interaction among numerous international, European, and even domestic insti-
tutions, established issues of international and European governance arise anew.
Over time, we may gain a clearer picture of the interlocking structure of interna-
tional governance institutions and EU legal instruments beyond causal chains,
clusters, and limited ‘‘regime complexes.’’ Analysts of European integration have
been more aware of the fact that the many EU legal instruments form parts of larger
policies. The study of institutional interaction reveals that international institutions
also constitute parts of institutional networks that operate in broad problem areas.
Underdal and Young (2004) relate the concern for interlocking structures of gov-
ernance institutions to the new institutionalism in the study of political economy.
Douglass North has argued that a network of institutional arrangements, including
property rights, contractual procedures, liability rules, and credit systems, promoted
the growth of the dynamic markets of the West in the past several hundred years
(North and Thomas 1973). Mancur Olsen (2000) explains the superior economic
performance of democratic over authoritarian systems in a similar way.
The recognition of interlocking structures of governance institutions has impor-

tant ramifications especially for the study of international governance. Although it
may be assumed that every single international institution is established and main-
tained purposively by its respective members, the broader institutional structures
within the international system emerge spontaneously without explicit coordination
between the discrete institutions involved. Purposive and skillful efforts to enhance
the effectiveness of international governance in one sector may be jeopardized by
parallel efforts in other sectors. Eventually, institutional interaction may turn out to
be as important for the overall effectiveness of governance within the decentralized
international order as the skillful design of single institutions. The comparative per-
spective demonstrates that even in the relatively well-organized institutional setting
of the European Union coordination failures and disruption between instruments
occur, while the deliberate use of interaction to promote desired policies is more
advanced than in international relations.
This volume on institutional interaction constitutes a first step toward the com-
prehensive analysis of the broader consequences of international and EU institu-
tions and their interlocking structure. It demonstrates that international institutions
have effects beyond their domains. These effects might generate even broader con-
sequences for international and domestic society, the legitimacy of international
governance, and the actors entitled to participate in decision-making processes or
capable of doing so. Whereas our study has hardly been able to delve into pertinent
aspects of these broader consequences, it has made a start that demonstrates the po-
tential of the effort. The study of interaction between international and EU institu-
tions thus provides a rich and fascinating research agenda for those interested in
international and European governance.
Note
1. This may help explain why a potential for further improvement was particularly frequently
identified at the international level, for disruption also figured particularly prominently at that
level (table 13.2).
References
Abbott, Kenneth W., and Duncan Snidal. 2000. Hard and Soft Law in International Gover-
nance. International Organization 54 (3): 421–456.
Aggarwal, Vinod K. 1983. The Unraveling of the Multi-Fiber Arrangement, 1981: An Exam-
ination of International Regime Change. International Organization 37 (4): 617–646.
Alter, Karen J. 2000. The European Legal System and Domestic Policy: Spillover or Backlash?
Andersen, Regine. 2002. The Time Dimension in International Regime Interplay. Global En-
vironmental Politics 2 (3): 98–117.
Bär, Stefani, Ingmar von Homeyer, and Anneke Klasing. 2002. Overcoming Deadlock?
Enhanced Cooperation and European Environmental Policy after Nice. Yearbook of Euro-
pean Environmental Law 2: 241–270.
Brack, Duncan. 2002. Environmental Treaties and Trade: Multilateral Environmental Agree-
ments and the Multilateral Trading System. In Gary P. Sampson and W. Bradney Chambers,
eds., Trade, Environment, and the Millennium, 2nd ed., 321–352. Tokyo: United Nations
University Press.
Brown Weiss, Edith. 1993. International Environmental Issues and the Emergence of a New
World Order. Georgetown Law Journal 81 (3): 675–710.
Burley, Anne-Marie, and Walter Mattli. 1993. Europe before the Court: A Political Theory of
Legal Integration. International Organization 47 (1): 41–76.
Craig, Paul, and Gráinne de Búrca. 2002. EU Law: Text, Cases, and Materials. 3rd ed. Ox-
ford: Oxford University Press.
Fairman, David. 1996. The Global Environment Facility: Haunted by the Shadow of the
Future. In Robert O. Keohane and Marc A. Levy, eds., Institutions for Environmental Aid:
Pitfalls and Promise, 55–87. Cambridge, MA: MIT Press.
Gehring, Thomas. 1997. Governing in Nested Institutions: Environmental Policy in the
European Union and the Case of Packaging Waste. Journal of European Public Policy 4 (3):
337–354.
Grubb, Michael, and Joyeeta Gupta. 2000. Leadership: Theory and Methodology. In Joyeeta
Gupta and Michael Grubb, eds., Climate Change and European Leadership: A Sustainable
Role for Europe?, 15–24. Dordrecht: Kluwer.
Haas, Ernst. 1975. Is There a Hole in the Whole? Knowledge, Technology, Interdependence,
and the Construction of International Regimes. International Organization 29 (3): 827–
876.
Haas, Ernst. 1990. When Knowledge is Power: Three Models of Change in International
Organizations. Berkeley: University of California Press.
Hasenclever, Andreas, Peter Mayer, and Volker Rittberger. 1997. Theories of International
Regimes. Cambridge: Cambridge University Press.
Héritier, Adrienne. 1996. The Accommodation of Diversity in the European Policy-Making
Process and its Outcomes: Regulatory Policy as a Patchwork. Journal of European Public
Policy 3 (2): 149–167.
Hovi, Jon, Detlef Sprinz, and Arild Underdal. 2003. The Oslo-Potsdam Solution to Measur-
ing Regime Effectiveness: Critique, Response, and the Road Ahead. Global Environmental
Politics 3 (3): 74–96.
King, Leslie A. 1997. Institutional Interplay—Research Questions. Paper Commissioned by
Institutional Dimensions of Global Change and International Human Dimensions Programme
on Global Environmental Change. Environmental Studies Programme, Burlington VT. School
of Natural Resources, University of Vermont.
Koremenos, Barbara, Charles Lipson, and Duncan Snidal. 2001. The Rational Design of
International Institutions. International Organization 55 (4): 761–799.
Lenschow, Andrea, ed. 2002. Environmental Policy Integration: Greening Sectoral Policies in
Europe. London: Earthscan.
Martin, Lisa L., and Beth A. Simmons. 1998. Theories and Empirical Studies of International
Institutions. International Organization 52 (4): 729–757.
Mattli, Walter. 2003. Setting International Standards: Technological Rationality or Primacy
of Power. World Politics 56 (1): 1–42.
regionalen Umweltregimen. Darmstadt, Germany: Deutscher Universitäts-Verlag.
North, Douglass, and Robert P. Thomas. 1973. The Rise of the Western World. Cambridge:
Oberthür, Sebastian. 1999. The EU as an International Actor: The Case of the Protection of
the Ozone Layer. Journal of Common Market Studies 37 (4): 641–659.
Oberthür, Sebastian. 2003. Institutional Interaction to Address Greenhouse Gas Emissions
from International Transport: ICAO, IMO and the Kyoto Protocol. Climate Policy 3 (3):
191–205.
Oberthür, Sebastian, and Thomas Gehring. Forthcoming. Disentangling the Interaction be-
tween the Cartagena Protocol and the World Trade Organisation. In W. Bradnee Chambers
and Joy A. Kim, eds., Institutional Interplay: The Case of Biosafety. Tokyo: United Nations
University.
Oberthür, Sebastian, and Hermann E. Ott (in collaboration with Richard G. Tarasofsky).
1999. The Kyoto Protocol: International Climate Policy for the 21st Century. Berlin:
Springer.
Olsen, Mancur Jr. 2000. Power and Prosperity: Outgrowing Communist and Capitalist Dic-
tatorships. New York: Basic Books.
Parson, Edward A. 1993. Protecting the Ozone Layer. In Peter M. Haas, Robert O. Keohane,
and Marc A. Levy, eds., Institutions for the Earth: Sources of Effective International Environ-
mental Protection, 27–73. Cambridge, MA: MIT Press.
Parson, Edward A. 2003. Protecting the Ozone Layer: Science, Strategy, and Negotiation in
the Shaping of a Global Environmental Regime. Oxford: Oxford University Press.
Raustalia, Kal, and David G. Victor. 2004. The Regime Complex for Plant Genetic Re-
Schelling, Thomas C. 1995. The Strategy of Conflict. 15th ed. Cambridge, MA: Harvard
University Press.
Schweitzer, Michael. 2003. Überblick zum Verhältnis: Völkerrecht–Europäisches
Gemeinschaftsrecht–nationales Recht. In Hans-Werner Rengeling, ed., Handbuch zum euro-
päischen und deutschen Umweltrecht, 679–704. Cologne: Carl Heymanns.
Sebenius, James K. 1983. Negotiation Arithmetics: Adding and Subtracting Issues and Parties.
Sebenius, James K. 1992. Challenging Conventional Explanations of International Coopera-
tion: Negotiation Analysis and the Case of Epistemic Communities. International Organiza-
tion 46 (1): 323–365.
Snidal, Duncan. 1985. The Game Theory of International Politics. World Politics 38 (1):
25–57.
Snidal, Duncan. 1994. The Politics of Scope: Endogenous Actors, Heterogeneity, and Institu-
tions. Journal of Theoretical Politics 6 (4): 449–472.
Stein, Arthur. 1982. Coordination and Collaboration: Regimes in an Anarchic World. Inter-
national Organization 36 (2): 299–324.
Stokke, Olav Schram. 2000. Managing Straddling Stocks: The Interplay of Global and Re-
gional Regimes. Ocean and Coastal Management 43 (2–3): 205–234.
Stokke, Olav Schram. 2001. The Interplay of International Regimes: Putting Effectiveness
Tversky, Amos, and Daniel Kahnemann. 1981. The Framing of Decision and Rational
Choice. Science 211: 453–458.
Tversky, Amos, and Daniel Kahnemann. 1984. Choices, Values, and Frames. American Psy-
chologist 39 (4): 341–350.
Underdal, Arild, and Oran R. Young. 2004. Research Strategies for the Future. In Arild
Underdal and Oran R. Young, eds., Regime Consequences: Methodological Challenges and
Research Strategies, 361–380. Dordrecht: Kluwer.
Victor, David G. 1999. The Market for International Environmental Protecting Services
and the Perils of Coordination. Paper prepared for the International Conference on Synergies
and Coordination between Multilateral Environmental Agreements, United Nations Univer-
sity, July 14–16, 1999, Tokyo. Available at http://www.geic.or.jp/interlinkages/docs/online-
docs.html.
Weber, Max. 1976. Wirtschaft und Gesellschaft: Grundriss der verstehenden Soziologie. 5th
ed. Tübingen, Germany: J.C.B. Mohr.
Werksman, Jacob. 1996. Consolidating Governance of the Global Commons: Insights from
the Global Environment Facility. Yearbook of International Environmental Law 6: 27–63.
Werksman, Jacob. 2005. The Negotiation of a Kyoto Compliance System. In Olav Schram
Stokke, Jon Hovi, and Geir Ulfstein, eds., Implementing the Climate Regime: International
Compliance, 17–38. London: Earthscan.
Young, Oran R. 1991. Political Leadership and Regime Formation: On the Development of
Institutions in International Society. International Organization 45 (3): 281–308.
Appendix: Overview of Identified Cases of
Column ‘‘No.’’ gives the number of the case of interaction in the table.
Column ‘‘Brief provides a very brief characterization of the case of interaction,
description of case’’ including indication of other institution.
Column indicates whether the core institution was the source of the
‘‘Core: T or S’’ interaction (S) or the target (T).
Column provides information on whether the interaction is horizontal (H)
‘‘Hor. or vert.’’ or vertical (V ¼ between an international and an EU institution).
Column denotes whether the effect on the target institution was synergistic,
‘‘Quality of effect’’ disruptive, or neutral/uncertain.
Column denotes whether the interacting institutions belong to the same or
‘‘Policy fields’’ different policy fields.
Column indicates whether the source institution was assessed as having
‘‘Intentional’’ triggered the interaction intentionally or unintentionally (Yes/No);
in some cases, the entry is ‘‘uncertain.’’
Column provides, for cases of Interaction through Commitment, the key
‘‘Key difference’’ difference of the interaction (O ¼ Objectives; M ¼ Membership;
Ma ¼ means).
Column indicates whether or not a collective policy response occurred (in the
‘‘Policy response’’ source institution, in the target institution, or in both) (Yes/No).
Column indicates whether case-study authors identified a further potential for
‘‘Further potential’’ improving the situation by targeted policy action (Yes) or not (No),
or whether such a further potential was deemed to be uncertain.
Column denotes the ideal type of interaction to which the case belongs as
‘‘Interaction type’’ derived from the other variables: Model ¼ Policy Model; Request ¼
Request for Assistance; JD ¼ Jurisdictional Delimitation; Nested ¼
Nested Institutions; Means ¼ Additional Means; Behavioral ¼
Behavioral Interaction (see chapter 13 for the ideal types).
Notes: Cases of implementation of international rules within the EU have always been coded
as ‘‘intentional.’’ Cases that could have shown up twice in the list (with the source or the
target institution as the core) have only been included once. Some cases constitute a chain
(i.e., need to be seen together). The assessment of policy responses and future potential may
change over time.
Hor. Key Policy Further Inter-
374
Core: or Quality Policy Inten- differ- re- poten- action

No. Brief description of case T or S vert. of effect fields tional ence sponse tial type
Climate Change Regime (UNFCCC and Kyoto Protocol: Chapter 3)
Appendix
1. Compliance procedure of Montreal T H Synergy Same No no No No Model

Protocol on Substances that entry
Deplete the Ozone Layer has (n.e.)
served as a model for the elaboration
of the compliance system of the Kyoto
Protocol.
2. Following the Montreal Protocol, T H Disrup- Same No n.e. No Yes Model
establishment of technology and tion
economic assessment panels under
UNFCCC was blocked by a minority.
3. Montreal Protocol has helped T H Synergy Same No n.e. No No Behav-
phase out ozone-depleting substances ioral
that are also potent greenhouse gases
(GHGs).
4. Montreal Protocol has supported T H Disrup- Same No n.e. Yes Yes Behav-
use of fluorinated GHGs regulated tion ioral
under Kyoto Protocol (hampering
climate protection).
5. Convention on Biological Diver- S H Disrup- Same No n.e. Yes Yes Behav-
sity may suffer from establishment of tion ioral
monocultural tree plantations induced
by climate change regime.
6. Ramsar Convention on Wetlands S H Neutral/ Same No n.e. Yes Yes Behav-
may benefit from additional resources uncertain ioral
for wetland management or suffer
from conversion of wetlands for
carbon sequestration induced by
climate change regime.
7. Convention to Combat Desertifi- S H Synergy Same No n.e. Yes Yes Behav-
cation may benefit from forestry ioral
activities promoted under the climate
change regime that help combat
desertification.
8. Kyoto Protocol has asked Interna- S H Uncertain Dif- Yes O Yes No JD
tional Civil Aviation Organiza- (disrup- ferent
tion to address GHG emissions from tion)
international aviation.
9. Kyoto Protocol has asked Interna- S H Uncertain Dif- Yes O Yes No JD
tional Maritime Organization to (disrup- ferent
address GHG emissions from inter- tion)
national shipping.
10. World Trade Organization is T H Disrup- Dif- No O Yes Yes JD
used as a major argument against tion ferent
elaboration of trade-relevant climate
protection measures (‘‘chill effect’’).
11. World Bank has greened its opera- S H Synergy Dif- No Ma No Yes Means
tions to some extent in response to ferent
climate change regime.
12. Climate change regime has asked the S H Synergy Dif- Yes O Yes Yes JD
Global Environment Facility ferent
Identified Cases of Institutional Interaction
to operate its financial mechanism.

375
376

13. EU Landfill Directive results in T V Synergy Same No n.e. No Uncer- Behav-
reductions of methane emissions and tain ioral
Appendix
thus helps implement the Kyoto (no)

Protocol.
14. EU Renewable Energy Directive is T V Synergy Dif- Yes n.e. No Yes Behav-
to lead to increasing use of non-GHG ferent ioral
emitting energy sources and thus helps
implement the Kyoto Protocol.
15. EU Directive on the Internal T V Disrup- Dif- No n.e. Yes Yes Behav-
Market for Electricity is expected tion ferent ioral
to result, inter alia, in lower energy
prices, counteracting efforts to save
energy and reduce GHG emissions.
16. EU Directives on car emission T V Disrup- Same No n.e. Yes Yes Behav-
standards require cars to be tion ioral
equipped with catalytic converters,
leading to increases of GHG
emissions.
17. EU GHG monitoring mechanism S V Synergy Same Yes Ma No No Means
responds to international monitoring
and reporting requirements under the
Kyoto Protocol.
18. EU Burden-Sharing Agreement T V Synergy Same No M No No Nested
of 1997 facilitated agreement on
strengthened targets under the Kyoto
Protocol.
19. Kyoto Protocol triggered the codifica- S V Synergy Same Yes Ma No No Means
tion of EU Burden Sharing in supra-
national EU law.
20. EU burden sharing helps implement T V Synergy Same No n.e. No No Behav-
the Kyoto Protocol by strengthening ioral
enforcement in the EU.
21. Kyoto Protocol contributed to the S V Synergy Same Yes Ma No Yes Means
emergence of an EU regulatory
framework on fluorinated GHGS
(Regulation and Directive).
22. EU Regulatory framework on T V Synergy Same No n.e. No Yes Behav-
fluorinated GHGS is expected to ioral
lead to reduction of GHG emissions.
23. Climate change regime triggered S V Synergy Same Yes Ma No No Means
adoption of the EU Emissions
Trading Directive.
24. EU Emissions Trading Directive is T V Synergy Same No n.e. No Yes Behav-
expected to lead to reduction of GHG ioral
emissions.
Convention on Biological Diversity (CBD: Chapter 4)
25. Implementation of Ramsar Conven- T H Synergy Same No n.e. Yes Uncer- Behav-
tion on Wetlands helps CBD tain ioral
achieve its targets. (yes)
26. Implementation of Convention on T H Synergy Same No n.e. Yes Uncer- Behav-
Migratory Species (CMS) helps tain ioral
CBD achieve its targets. (yes)
27. CBD has facilitated talks in the UN S H Synergy Same No Ma No No Means

Forum on Forests (and its predeces-
sors), which could build on agree-
377
ments reached under it.

378

28. WTO TRIPS Agreement affected T H Disrup- Dif- No O Yes Yes JD
negotiations of the CBD on equitable tion ferent
Appendix
sharing/access to genetic resources.

29. Rules of WTO TRIPS Agreement on T H Disrup- Dif- No n.e. Yes Yes Behav-
patenting engender behavior at odds tion ferent ioral
with CBD objectives on equitable
sharing/access to genetic resources.
30. CBD objectives on equitable sharing/ S H Disrup- Dif- No n.e. Yes Yes Behav-
access to genetic resources lead to tion ferent ioral
cumbersome patent protection at odds
with WTO TRIPS Agreement.
31. FAO Undertaking on Plant T H Disrup- Dif- No O Yes Uncer- JD
Genetic Resources for Food and tion ferent tain
Agriculture influenced negotiations (no)
on CBD regarding property rights and
access to genetic resources.
32. CBD influenced FAO Treaty on S H Neutral/ Dif- No O No Uncer- JD
Plant Genetic Resources for uncertain ferent tain
Food and Agriculture regarding (no)
access to genetic resources.
33. CBD objectives on equitable sharing S V Disrup- Dif- Yes Ma Yes Yes Means
and access to genetic resources influ- tion ferent
enced preparation of the EU Patent
Directive of 1998.
34. EU Patent Directive may under- T V Disrup- Dif- No n.e. No Yes Behav-
mine the implementation of the CBD tion ferent ioral
objectives on equitable sharing/access
to genetic resources.
International Regime for the Protection of the Northeast Atlantic (OSPAR/INSCs: Chapter 5)
35. International North Sea Confer- T H Synergy Same Yes Ma(M) No No Means
ences (INSCs) have facilitated and
sped up development of the OSPAR
Convention for the protection of the
Northeast Atlantic.
36. OSPAR has expanded the scope of S H Synergy Same Uncer- n.e. No No Behav-
INSCs and helped implement INSC tain ioral
Declarations on marine pollution. (yes)
37. Montreal Protocol on Sub- T H Synergy Same No n.e. No No Behav-
stances that Deplete the Ozone ioral
Layer helped implement INSC com-
mitments on carbon tetrachloride and
methyl chloroform.
38. The climate change regime may T H Disrup- Same No n.e. No Yes Behav-
provide an incentive for CO2 seques- tion ioral
tration in North Sea oil fields, which
may violate OSPAR prohibition of
dumping at sea.
39. London Dumping Convention S H Synergy Same No M No No Nested
benefited from OSPAR that facilitated
agreement on global ban on dumping
and incineration at sea within London
Convention.
40. Implementation of Rhine Conven- T H Synergy Same No n.e. Yes Uncer- Behav-
tion helps achieve objectives of tain ioral
OSPAR. (yes)
41. UNECE Convention on Long- T H Synergy Same No n.e. Yes No Behav-
range Transboundary Air Pollu- ioral

tion contributes to achieving the
objective of OSPAR on nutrients and
379
eutrophication (NOX ).
380

42. INSC Declaration facilitated/triggered S V Synergy Same Yes Ma(M) No No Means
EU Nitrates Directive.
Appendix
43. EU Nitrates Directive helps imple- T V Synergy Same No n.e. No No Behav-

mentation of INSC objectives on ioral
nutrients.
44. INSC Declaration facilitated/triggered S V Synergy Same Yes Ma(M) No No Means
EU Urban Waste-Water Directive.
45. EU Urban Waste-Water Directive T V Synergy Same No n.e. No No Behav-
helps implementation of INSC objec- ioral
tives on nutrients.
46. OSPAR triggered/facilitated inclusion S V Synergy Same Yes Ma(M) No No Means
of an obligation to phase out sewage-
sludge dumping in EU Urban Waste-
Water Directive.
Global Fisheries Regime (UN Fish Stocks Agreement/FAO: Chapter 6)
47. The regional International Coun- S H Uncertain Same Yes Ma Yes Uncer- Means
cil for the Exploration of the (synergy) tain
Sea (ICES) has implemented the (yes)
precautionary approach following its
formal adoption in the global Fish
Stocks Agreement (FSA).
48. ICES contributed to improved conser- T H Synergy Same Yes n.e. No Uncer- Behav-
vation in line with FSA by introducing tain ioral
greater safety margins in its catch (yes)
recommendations and requesting
rapid recovery of troubled stocks.
49. ICES modified its communication and S H Uncertain Same Yes n.e. No Uncer- Request
terminology in response to criticism (synergy) tain
by FAO and regional regimes. (yes)
50. The Northeast Atlantic Fisheries S H Synergy Same Yes M/Ma No Uncer- Means
Commission (NEAFC) responded to tain
the transparency rules of the FSA by (no)
adopting new provisions on access to
meetings and reports.
51. FSA supports implementation of the S V Synergy Same Yes Ma(O) No Yes Means
precautionary approach in the EU
Common Fisheries Policy (CFP)
(through ICES), which requires pre-
agreed decision rules within recovery/
management plans.
52. FSA rules support inclusion of man- S V Disrup- Same Yes Ma(O) No Uncer- Means
agement considerations in EU fish tion tain
import regulations. (yes)
53. FAO International Program of Action S V Disrup- Same Yes Ma(O) No Yes Means
on fishing capacity (IPOA-Capacity) tion
contributes to reform of related EU
rules aiming at a reduction of
subsidies.
54. FAO IPOA Capacity contributed to S V Disrup- Same No n.e. Yes Yes Behav-
reduction of fisheries subsidies in tion ioral
the EU and its member states.
55. EU fisheries subsidies still under- T V Disrup- Same No n.e. No Yes Behav-
mine conservation and management tion ioral
of fish stocks required by global
fisheries regime.
381
382

56. WTO subsidies rules improve infor- T H Uncertain Dif- No Ma/O Yes Yes Means
mation flows and bindingness and (synergy) ferent
Appendix
may therefore help reduce or limit

fisheries subsidies and capacity.
57. The global fisheries regime helped S H Synergy Dif- No Ma/O Yes Yes Means
place fisheries subsidies on the formal ferent
WTO agenda (Doha Round).
58. Convention on International T H Synergy Dif- Yes n.e. Yes Yes Request
Trade in Endangered Species of ferent
Wild Fauna and Flora (CITES) has
contributed to the emergence of FAO
International Program of Action on
the Conservation and Management of
Sharks (IPOA-Sharks).
59. CITES benefited in its implementation S H Synergy Dif- Yes n.e. No Uncer- Behav-
from improved national reporting ferent tain ioral
procedures and records of shark (yes)
catches resulting from IPOA-Sharks.
60. Northwest Atlantic Fisheries T H Synergy Same No n.e. No Yes Model
Organization facilitated adoption
of FSA provisions on inspection,
detention, and arrest of fishing ships.
61. Central Bering Sea Doughnut T H Synergy Same No n.e. No Yes Model
Hole Agreement facilitated adop-
tion of FSA provisions on inspection,
detention, and arrest of fishing ships.
Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES: Chapter 7)
62. The Convention for the Conser- T H Synergy Same Yes n.e. No Uncer- Request
vation and Management of the tain
Vicuña (CCMV) has asked CITES (no)
for help in limiting trade in Vicuña
products and CITES has responded
positively with a COP Resolution.
63. The International Convention on T H Synergy Same Yes n.e. No Uncer- Request
the Regulation of Whaling tain
(ICRW) asked CITES for help in (yes)
limiting trade in whale products and
CITES has responded positively with
a COP Resolution.
64. In response to a call for action by the T H Synergy Same Yes n.e. No Uncer- Request
Convention on Migratory Species tain
(CMS), CITES passed a COP Resolu- (no)
tion in support of CMS concerning
the houbara bustard.
65. Implementation of CITES helps Con- S H Synergy Same No n.e. Yes Uncer- Behav-
vention on Biological Diversity tain ioral
achieve its objectives (several decisions (no)
and resolutions in both institutions).
66. CITES restricts free international S H Disrup- Dif- No n.e. Yes Yes Behav-
trade and thus is in potential conflict tion ferent ioral
with the World Trade Organi-
zation (WTO); it allows for trade
with nonparties complying with its
obligations.
67. CITES asked the World Customs S H Neutral/ Dif- Yes n.e. Yes Uncer- Request
Organization (WCO) for help in uncertain ferent tain
implementation. (no)
383
384

68. WCO supports implementation of T H Synergy Dif- Yes n.e. No Uncer- Behav-
CITES by helping coordinate CITES ferent tain ioral
enforcement and training via national (no)
Appendix
customs organizations.
69. CITES asked WCO to change customs S H Neutral/ Dif- Yes n.e. No Uncer- Request
codes (e.g., for shark products). uncertain ferent tain
(no)
70. CITES asked Interpol for help in S H Neutral/ Dif- Yes n.e. Yes No Request
implementation and enforcement. uncertain ferent
71. Interpol supports implementation of T H Synergy Dif- Yes n.e. No Uncer- Behav-
CITES (e.g., memorandum of under- ferent tain ioral
standing on cooperation). (no)
72. The EU adopted a regulation to S V Synergy Same Yes Ma Yes Yes Means
control wildlife trade in 1982 in
response to CITES, although it was
not a party to CITES.
73. CITES concerns about abolition of S V Synergy Same Yes Ma Yes No Means
internal border controls caused the
EU to strengthen its CITES Regula-
tion in 1997 (revised 2001).
74. CITES Regulation of the EU T V Synergy Same Yes n.e. Yes Uncer- Behav-
supports the implementation of tain ioral
CITES. (no)
75. Abolition of intra-EU border controls T V Disrup- Dif- No n.e. Yes Uncer- Behav-
for goods and persons in the wake of tion ferent tain ioral
the EU Single Market Program (no)
endangers effective implementation of
CITES trade restrictions in the EU.
World Trade Organization (WTO: Chapter 8)
76. Montreal Protocol on Substan- S H Disrup- Dif- No O Yes Uncer- JD
ces that Deplete the Ozone Layer tion ferent tain
granted exemptions from trade restric- (no)
tions to nonparties complying with it
in response to WTO rules.
77. Montreal Protocol did not apply S H Disrup- Dif- No O Yes Yes JD
planned restrictions on trade in prod- tion ferent
ucts produced with ozone-depleting
substances, partly in response to
WTO rules.
78. Montreal Protocol has provided a T H Uncertain Dif- No n.e. Yes No Model
model for WTO of admissible trade (synergy) ferent
restrictions for environmental pur-
poses in a multilateral framework.
79. Montreal Protocol did not restrict S H Disrup- Dif- No O Yes Yes JD
trade in used products relying on tion ferent
ozone-depleting substances, to a large
extent due to WTO disciplines.
80. WTO disciplines affected implementa- S H Disrup- Dif- No n.e. Yes Yes Behav-
tion of the Montreal Protocol by tion ferent ioral
limiting parties’ use of trade measures
with respect to used products relying
on ozone-depleting substances.
81. Cartagena Protocol on Bio- T H Disrup- Dif- No O No No JD
safety was used to block attempts to tion ferent
regulate biosafety under the WTO.
82. Cartagena Protocol includes S H Neutral/ Dif- No O No No JD

preambular language attempting to uncertain ferent
reconcile tensions with WTO.
385
386

83. Cartagena Protocol has adapted S H Disrup- Dif- No O Yes Uncer- JD
its risk-assessment procedure from tion ferent tain
Appendix
WTO’s SPS Agreement, which poten- (yes)

tially constrains its use of trade
measures.
84. In response to WTO rules, Basel S H Disrup- Dif- No O Yes Uncer- JD
Convention on the Control of tion ferent tain
Transboundary Movements of (no)
Hazardous Wastes and Their
Disposal granted exemptions from
trade restrictions to nonparties com-
plying with it.
85. In response to WTO rules, Inter- S H Disrup- Dif- No O Yes Uncer- JD
national Commission for the tion ferent tain
Conservation of Atlantic Tunas (no)
granted exemptions from trade restric-
tions to nonparties complying with it.
86. In response to WTO rules, Conven- S H Disrup- Dif- No O Yes Uncer- JD
tion on the Conservation of Ant- tion ferent tain
arctic Marine Living Resources (no)
granted exemptions from trade restric-
tions to nonparties complying with it.
EU Water Framework Directive (Chapter 9)
87. EU Nitrates Directive supports im- T H Synergy Same No n.e. No Yes Behav-
plementation of Water Framework ioral
Directive.
88. EU Water Framework Directive is S H Synergy Same Yes n.e. No Uncer- Behav-
expected to support implementation tain ioral
of Nitrates Directive.
89. EU Water Framework Directive influ- S H Uncertain Same No Ma No Uncer- Means
ences the amendment of the Bathing (synergy) tain
Waters Directive.
90. EU Bathing Waters Directive T H Synergy Same No n.e. No Yes Behav-
supports implementation of Water ioral
91. EU Water Framework Directive is S H Synergy Same Yes n.e. No Uncer- Behav-
expected to support implementation tain ioral
of Bathing Waters Directive.
92. EU IPPC Directive allows the Water T H Synergy Same No Ma No No Means
Framework Directive to establish
explicit link by defining environ-
mental quality standards.
93. Later EU Water Framework Directive S H Synergy Same Yes Ma No No Means
changes meaning of environmental
quality standards protected under
earlier IPPC Directive.
94. EU IPPC Directive supports imple- T H Synergy Same Uncer- n.e. No Uncer- Behav-
mentation of Water Framework tain tain ioral
Directive by requiring that permits (yes)
do not lead to breaches of established
environmental quality standards.
95. EU Water Framework Directive was S H Synergy Same No Ma No No Means
used for defining damage under
Directive on Environmental
Liability.
96. EU Liability Directive is to support T H Synergy Same Yes n.e. No Uncer- Behav-
implementation of the Water Frame- tain ioral
387
work Directive. (no)

388

97. Convention on Transboundary T V Synergy Same Yes Ma No Uncer- Means
Watercourses and Lakes triggers tain
Appendix
certain implementing provisions in (no)

98. Water Framework Directive contrib- S V Synergy Same Yes n.e. No Uncer- Behav-
utes to achieving the objectives of the tain ioral
Convention on Transboundary
Watercourses and Lakes.
99. OSPAR Convention for the protec- S V Synergy Same Yes Ma No Uncer- Means
tion of the Northeast Atlantic triggers tain
certain implementing provisions in (no)
100. Water Framework Directive contrib- S V Synergy Same Yes n.e. No Uncer- Behav-
utes to achieving the objectives of tain ioral
OSPAR Convention.
101. International North Sea Confer- T V Synergy Same Yes Ma No No Means
ences facilitated/triggered Water
102. Aarhus Convention on public par- T V Synergy Same Yes Ma No Uncer- Means
ticipation and access to information/ tain
justice led to enhanced public partici- (yes)
pation requirements in Water Frame-
work Directive.
103. Water Framework Directive supports S V Synergy Same No n.e. No Uncer- Behav-
achieving objectives of/helps imple- tain ioral
ment Ramsar Convention on
wetlands.
EU Directive on Integrated Pollution Prevention and Control (IPPC Directive: Chapter 9)
104. EU IPPC Directive adopts principle of T H Synergy Same No Ma No No Means
waste avoidance from Waste Frame-
work Directive.
105. EU Waste Framework Directive S H Synergy Same Yes n.e. No Uncer- Behav-
benefits from implementation of IPPC tain ioral
Directive that requires waste avoid-
ance in line with the former.
106. EU IPPC Directive establishes link to T H Synergy Same No Ma No No Means
Air Quality Framework Directive
by requiring that permits do not lead
to breaches of established environ-
mental quality standards.
107. Implementation of IPPC Directive S H Synergy Same Yes n.e. No Uncer- Behav-
supports implementation of Air tain ioral
Quality Framework Directive by
requiring that permits do not lead to
breaches of established environmental
quality standards.
Habitats Directive by requiring
that certified installations do not
adversely affect ‘‘favorable conserva-
tion status.’’
109. Implementation of EU IPPC Directive S H Synergy Same Yes n.e. No Uncer- Behav-
supports implementation of Habitats tain ioral
Directive by requiring that certified
installations do not adversely affect
‘‘favorable conservation status.’’

389
390

Birds Directive by requiring that certi-
Appendix
fied installations do not adversely

affect ‘‘favorable conservation status.’’
111. Implementation of EU IPPC Directive S H Synergy Same Yes n.e. No Uncer- Behav-
supports implementation of Birds tain ioral
Directive by requiring that certified
installations do not adversely affect
‘‘favorable conservation status.’’
112. Emission limits under EU Large T H Uncertain Same No Ma No Uncer- Means
Combustion Plants Directive (synergy) tain
(incl. revision) ensure minimum stan- (no)
dards, thus limiting flexibility in issu-
ing permits under IPPC Directive.
113. Emission limits under EU Solvents T H Uncertain Same No Ma No Uncer- Means
Directive ensure minimum stan- (synergy) tain
dards, thus limiting flexibility in (no)
issuing permits under IPPC Directive.
114. EU Environmental Management T H Synergy Same No n.e. No Uncer- Behav-
and Auditing Scheme (EMAS tain ioral
Regulation) facilitates implementation
of IPPC Directive in that EMAS certif-
icates can be used to simplify permit-
ting procedures.
115. IPPC Directive supports implementa- S H Synergy Same No n.e. No Uncer- Behav-
tion of EMAS Regulation by pro- tain ioral
viding incentive to establish EMAS. (no)
116. EU IPPC Directive establishes link to T H Uncertain Same No Ma Yes Uncer- Means
Environmental Impact Assessment (synergy) tain
(EIA) Directive by allowing use of (no)
impact assessment under EIA Direc-
tive in permitting procedures.
117. EU EIA Directive facilitates imple- T H Synergy Same No n.e. No Uncer- Behav-
mentation of IPPC (in that impact tain ioral
assessment can be used under IPPC). (no)
118. EU Seveso II Directive on industrial S H Synergy Same Uncer- n.e. No No Behav-
accidents benefits from IPPC Directive tain ioral
that promotes an integrated approach
and contributes to a greening of
installations so that the impact of any
accident would be minimized.
119. Aarhus Convention on public par- T V Synergy Same Yes Ma No No Means
ticipation and access to information/
justice has led to amendments of rules
on public participation under IPPC
Directive.
120. OSPAR Convention for the protec- S V Synergy Same No n.e. No No Behav-
tion of the Northeast Atlantic benefits ioral
from implementation of IPPC Direc-
tive (reduction of pollution of North-
east Atlantic).
121. Helsinki Convention for the pro- S V Synergy Same No n.e. No No Behav-
tection of the Baltic Sea benefits from ioral
implementation of IPPC Directive
(reduction of pollution of Baltic Sea).
122. Convention on Long-Range S V Synergy Same No n.e. No No Behav-

Transboundary Air Pollution ioral
benefits from implementation of IPPC
391
Directive (reduction of air pollution).

392

123. UN Framework Convention on S V Synergy Same No n.e. No No Behav-
Climate Change benefits from ioral
Appendix
implementation of IPPC Directive

(reduction of emissions of greenhouse
gases).
EU Habitats Directive (Chapter 10)
124. EU Birds Directive has served as T H Synergy Same No n.e. No No Model
a model for Habitats Directive (in
parts).
125. EU Habitats Directive has supported S H Synergy Same Yes n.e. Yes Yes Behav-
achievement of objectives of Birds ioral
Directive, including by strengthening
certain aspects of protection of bird
sites.
126. Strict interpretation of site protection T H Uncertain Same No O No Uncer- JD
under the EU Birds Directive by (disrup- tain
European Court of Justice results in tion)
Habitats Directive being weaker than
initially proposed.
127. EU Habitats Directive includes provi- S H Disrup- Same Yes n.e. Yes Yes Behav-
sions that directly weaken level of tion ioral
protection afforded to sites designated
under Birds Directive.
128. A reference to Natura 2000 was S H Synergy Same No Ma No Uncer- Means
included in the EU Water Frame- tain
work Directive. (no)
129. EU Water Framework Directive T H Synergy Same Yes n.e. No Uncer- Behav-
supports achievement of objectives of tain ioral
Habitats Directive. (no)
130. Habitats Directive is used in part S H Synergy Same No Ma No No Means
for defining damage covered under
EU Environmental Liability
Directive.
131. EU Environmental Liability T H Synergy Same Yes n.e. No Uncer- Behav-
Directive introduces additional tain ioral
incentive for behavior that is com- (no)
patible with the Habitats Directive.
132. EU Structural Fund projects that T H Disrup- Dif- No n.e. Yes Yes Behav-
conflict with nature conservation tion ferent ioral
objectives undermine implementation
of the Habitats Directive.
133. Cross-compliance rule in the EU T H Synergy Dif- No n.e. No Yes Behav-
Structural Funds is used to encour- ferent ioral
age member states to submit site lists
under the Habitats Directive, and to
avoid damage to sites.
134. OSPAR Convention for the protec- T V Synergy Same No n.e. No Uncer- Behav-
tion of the Northeast Atlantic sup- tain ioral
ports implementation of Habitats (yes)
Directive in the marine environment.
135. Necessity to implement Bern Con- T V Synergy Same Yes Ma No No Means
vention on the Conservation of
European Wildlife and Natural
Habitats in the EU drives prepara-
tion of Habitats Directive.

393
394

136. Habitats Directive supports S V Synergy Same Yes n.e. No Yes Behav-
achievement of objectives of Bern ioral
Appendix
Convention.
137. EU Habitats Directive Natura 2000 S V Synergy Same No M Yes Uncer- Nested
program provides model/reference tain
point for development of Bern (yes)
Convention Emerald program.
138. Habitats Directive has adopted T V Synergy Same No n.e. No Uncer- Model
concepts from the Convention on tain
Biological Diversity, which was (no)
negotiated simultaneously.
139. Habitats Directive supports achieve- S V Synergy Same No n.e. Yes Uncer- Behav-
ment of objectives of Convention tain ioral
on Biological Diversity in the EU. (yes)
140. Agreement on Small Cetaceans T V Synergy Same No n.e. No Yes Behav-
of the Baltic and North Seas ioral
(ASCOBANS) under the Convention
on Migratory Species supports
achievement of objectives of Habitats
Directive at the national level.
141. Pan-European Biodiversity and T V Synergy Same No n.e. No Yes Behav-
Landscape Diversity Strategy sup- ioral
ports implementation of the Habitats
Directive, particularly through its
focus on ecological corridors between
sites.
EU Deliberate Release Directive (DRD: Chapter 11)
142. EU Deliberate Release Directive S H Disrup- Dif- Yes n.e. Yes Yes Behav-
(DRD) established second regulatory tion ferent ioral
hurdle requiring precautionary risk
assessment of genetically modified
organisms (GMOs), thus disrupting
authorization under EU sectoral
legislation.
143. DRD requires EU sectoral legisla- S H Uncertain Dif- Yes O Yes Uncer- JD
tion to provide for ‘‘similar’’ (origi- (disrup- ferent tain
nal DRD) or ‘‘equivalent’’ (revised tion) (yes)
DRD) environmental risk assessment.
144. DRD required labeling only for cer- S H Disrup- Dif- No n.e. Yes Uncer- Behav-
tain GMOs and thus undermined the tion ferent tain ioral
EU Novel Food Regulation that (no)
aimed to introduce more general
labeling of GM food.
145. Amended DRD (1997) allowed for S H Disrup- Dif- No n.e. Yes Uncer- Behav-
labeling irrespective of actual GMO tion ferent tain ioral
content, thereby undermining EU (yes)
sectoral legislation that aims to
introduce unambiguous labeling of
GMOs according to certain thresholds
of GM content.
146. Unclear scope of DRD’s precaution- T H Disrup- Dif- No O Yes Uncer- JD
ary risk assessment led to EU decision tion ferent tain
to evaluate secondary effects under (yes)
Pesticides Directive, rather than
DRD.
395
396

147. As an increasing number of member S H Disrup- Dif- Uncer- n.e. Yes Uncer- Behav-
states pursued evaluation of second- tion ferent tain tain ioral
Appendix
ary effects under DRD, authorization (yes) (yes)

of products conforming to Pesticide
Directive was negatively affected.
148. Liability of GMOs was not included T H Disrup- Same No O Yes Yes JD
in revised DRD because it was to be tion
dealt with in upcoming EU Environ-
mental Liability Directive.
149. EU Environmental Liability S H Synergy Same Yes Ma No Yes Means
Directive refers to revised DRD and
partially covers GMOs.
150. EU Environmental Liability T H Synergy Same Yes n.e. No Yes Behav-
Directive is expected to have a posi- ioral
tive effect on the implementation of
the DRD.
151. DRD’s precautionary approach leads S V Disrup- Dif- No n.e. Yes Yes Behav-
to delays in GM product approval or tion ferent ioral
even bans, thereby restricting inter-
national trade (at odds with WTO
SPS Agreement).
152. Partial ‘‘internationalization’’ of S V Uncertain Dif- Uncer- O/M No Yes JD
DRD’s precautionary approach may (disrup- ferent tain
lead to modification of WTO SPS tion)
Agreement’s concept of precaution.
153. WTO dispute-settlement mechanism T V Disrup- Dif- No n.e. Yes Yes Behav-
provides WTO SPS Agreement with tion ferent ioral
effective instrument to impose rudi-
mentary precaution on DRD.
154. DRD’s labeling and traceability provi- S V Disrup- Dif- No n.e. Yes Uncer- Behav-
sions restrict trade, thereby undermin- tion ferent tain ioral
ing objective of WTO TBT (yes)
Agreement.
155. WTO dispute-settlement mechanism T V Disrup- Dif- No n.e. Yes Uncer- Behav-
provides WTO TBT Agreement with tion ferent tain ioral
effective instrument to enforce compli- (yes)
ance, possibly undermining the DRD.
156. DRD constituted input for formula- S V Synergy Same No n.e. No No Model
tion of UNEP Technical Guide-
lines on Biosafety.
EU Air Quality Framework Directive (Chapter 12)
157. EU Single Market has amplified a T H Disrup- Dif- No n.e. Yes Yes Behav-
trend of increasing transport emis- tion ferent ioral
sions and hence disruptively affected
the process of improving air quality.
158. EU National Emission Ceilings T H Synergy Same Yes n.e. Yes Yes Behav-
Directive sets emissions ceilings, ioral
which will help achieve air-quality
targets.
159. EU Auto-Oil I Directives set T H Synergy Same Yes n.e. Yes Yes Behav-
stricter fuel standards and vehicle- ioral
emission limits, which will help
achieve air-quality targets.

397
398

160. EU SAVE Program (directives on T H Synergy Dif- No n.e. No No Behav-
minimum standards of energy effi- ferent ioral
Appendix
ciency and related labeling) contrib-

utes to increased energy efficiency,
which helps reduce emissions and
improve air quality.
161. EU Renewable Energy Directive T H Synergy Dif- No n.e. No No Behav-
contributes to enhanced use of renew- ferent ioral
able energy sources, which helps re-
duce emissions and improve air
quality.
162. Convention on Long-Range T V Synergy Same No Ma No No Means
Transboundary Air Pollution
developed the critical-loads concept,
which facilitated the EU process of
setting air-quality standards.
163. World Health Organization (WHO) T V Uncertain Dif- Yes Ma Yes Uncer- Means
produces air-quality guidelines that (synergy) ferent tain
have heavily influenced EU air-quality (no)
standards.
Index
Acid rain, 222. See also Convention on Belgium, 69, 91, 109, 120, 254
Long-Range Transboundary Air Pollution; Bern Convention (on the Conservation of
Large Combustion Plants Directive European Wildlife and Natural Habitats),
Actors 15, 30–31, 233, 236–246, 254–255, 348–
knowledge of, 35 (see also Knowledge) 349, 353, 360, 363–364. See also Emerald
perceptions of, 35 Network
preferences of, 31–32, 35–38 effectiveness of, 246, 363
rationality of, 35–36 secretariat of, 242, 246
Additional Means. See Interaction through Best Available Techniques (BAT), 115, 217–
Commitment 219, 224–227
Africa, 92, 159, 239, 275 BAT reference document (BREF), 224–
Agreement on Small Cetaceans of the Baltic 225
and North Seas (ASCOBANS), 166, 236 Biodiversity regime, 33, 44, 363. See also
Air Quality Framework Directive, 29–30, Convention on Biological Diversity
219–221, 285–304, 315, 349, 352, 365 Birds Directive (of the EU), 219–221, 234–
Argentina, 173 241, 346, 353–354
Asia-Pacific Economic Cooperation (APEC), Brazil, 86, 95, 167
131
Australia, 71, 133, 189 Canada, 70, 167, 189, 193–194, 273
Auto-Oil Cartagena Protocol on Biosafety, 14, 81,
Directive, 286, 289, 295, 299–300, 302 181, 184–185, 187–193, 200–201, 275–
Program, 286, 290, 295, 297–299, 301 278, 280, 336–338, 345, 361–363. See
also Genetically modified organisms
Bathing Waters Directive (of the EU), 119, Causal influence, 5–6, 11, 26–28, 30, 32,
208–209, 352 358. See also Causal pathway
Behavioral Interaction, causal mechanism of, Causal mechanisms, 8–13, 31–44. See also
8, 10, 12, 14, 20, 39–42, 44, 58, 71, 82, Cognitive Interaction; Interaction through
84–85, 93, 108, 110, 116, 119, 122, 131, Commitment; Behavioral Interaction;
161, 163, 165, 186, 208, 210–211, 214, Impact-Level Interaction
219–221, 227–229, 236, 238, 249, 253– Causal pathway, 7, 19, 27, 29, 33. See also
255, 263–264, 288, 302, 308, 312, 315, Causal Influence
319, 321, 325–326, 331, 343, 349–357, Cause-effect relationship. See Causal
361–363. See also Causal mechanisms influence
400 Index
Chile, 133, 173, 189 Convention on the Protection and Use of

China, 86, 171–172, 193–194, 273 Transboundary Watercourses and
Clean Air for Europe Program (CAFÉ), 286, International Lakes, 209–216, 228
294, 301 Convention to Combat Desertification
Climate change regime, 3, 21, 28–29, 35, (CCD), 56–57, 59
44–45, 53–74, 84, 108, 261, 333, 336, Council of Europe, 239, 242–243, 245–247
345–347, 353, 363, 365. See also Kyoto Council of Ministers (of the EU), 25, 63, 69,
Protocol; U.N. Framework Convention on 72, 114–115, 145–147, 223, 240, 260,
Climate Change 265, 268–269, 289, 292–293, 297, 340,
Codex Alimentarius Commission (CAC), 346
193, 272, 276, 364. See also World Trade
Organization Deliberate Release Directive (DRD) (of the
Cognitive Interaction, causal mechanism of, EU), 10, 259–280, 346–347, 353–354.
8, 10, 12, 14, 20, 35–37, 42, 44, 56, 131, See also Genetically modified organisms
134, 142, 149, 161, 170, 175, 187, 236, labeling provisions, 263
265, 293, 308, 312, 319, 325–335, 349– risk assessment under, 260–267, 269–273,
350, 361 277, 279, 346, 354
Policy–Model type, 14, 35, 328–334, 361– traceability provisions, 261–262, 264, 278
362 Directive on Integrated Pollution Prevention
Request-for-Assistance type, 14, 330–334, and Control (IPPC) (of the EU), 10, 29–
346–347, 351, 353, 361, 363 31, 46, 58, 205–229, 236, 290, 312, 348–
Compliance 14, 37, 40, 345 349, 352, 365
Cross-compliance, 237–238, 252–253 Directive on Motor Vehicle Air Pollution
Conflict. See Effects of interaction, disruptive Control (of the EU), 297
Convention for the Conservation and Directive on the Conservation of Natural
Management of the Vicuña (CCMV), 158, Habitats and of Wild Flora and Fauna. See
162, 171–177 Habitats Directive
Convention on Biological Diversity (CBD), 3, Directive on the Deliberate Release of
14, 27, 39, 59, 78–98, 162–164, 187– Genetically Modified Organisms into the
190, 192–193, 201, 236–238, 277, 336, Environment. See Deliberate Release
363–366 Directive
Convention on International Trade in Disruption. See Effects of interaction
Endangered Species of Wild Flora and
Fauna (CITES), 3, 10, 14, 46, 82–83, 96, Ecuador, 86, 91, 133, 173
130–131, 157–177, 315, 330–332, 334, Effects of interaction
349, 353, 361–362, 364–366 disruptive, 1, 3, 12–14, 46, 58–59, 87, 89,
Convention on Long-Range Transboundary 91, 96–97, 108, 131, 161, 165, 186, 193,
Air Pollution (LRTAP), 107–108, 221, 200, 218, 233–234, 238–239, 247, 253–
289–290, 364. See also Acid rain; Large 255, 262–263, 270, 279, 289, 309–326,
Combustion Plants Directive 328, 330–335, 341, 344–345, 349–355,
Convention on Migratory Species of Wild 360, 362
Animals (CMS), 82–83, 96, 157–158, neutral/indeterminate, 46, 161, 309–310,
160–172, 177, 238, 364–365 317–320, 322, 324, 330, 332, 334, 345,
Convention on the Conservation of Antarctic 349, 353, 355, 361–362
Marine Living Resources (CCAMLR), synergistic, 1–4, 9, 12–13, 46, 58–59, 68,
185–187 73, 82, 94, 97–98, 103–104, 107–108,
Index 401
121, 131, 148–149, 161–165, 176–177, Environmental Action Programs of, 240,
208, 211, 214–215, 218–219, 227–228, 287, 291, 293
234, 238–239, 245–247, 255, 293–294, product sector legislation of, 259, 261, 279,
299–300, 302, 309–311, 314–326, 328, 346, 354
330, 332–335, 337, 344–345, 347, 349– scientific advisory committees of, 265, 268,
356, 360–362 295
Emerald Network (of protected sites), 15, Single Market, 56, 162, 166, 289
233, 239, 242–246, 348, 360. See also Structural Funds of, 3, 233, 235, 237–238,
Bern Convention 247–255, 356
Emission limits, 54, 73, 206, 219, 222–228,
286, 289–290, 295, 297, 302–303 Feedback effects, 14, 30, 331, 341, 360–363
Emissions Trading Directive (of the EU), 58 Finland, 69, 120
Enforcement, 15, 25 Food and Agriculture Organization of the
Environmental Liability Directive (of the United Nations (FAO), 10, 24, 83–86, 90,
EU), 208–209, 236–238, 262, 264, 346, 96, 98, 129–131, 134–145, 148–150,
348, 352–353 163, 261, 363
European Commission, 2, 6, 23–25, 34, 63, France, 70, 91, 109, 165, 253, 268–270
72, 109, 114–115, 120–121, 145–147, Friends of the Earth, 270
158, 166, 206, 210–211, 213, 216, 219,
222–227, 233–235, 240–243, 245–255, General Agreement on Tariffs and Trade
287–289, 291–301, 303, 315, 323, 332, (GATT), 21, 87, 132, 157, 182, 186–187,
338, 346, 357 189, 196–200, 271, 339. See also World
European Court of Justice (ECJ), 6, 23–25, Trade Organization
72, 74, 119, 121, 165, 213, 237–238, Genetically modified organisms (GMOs), 10,
242, 338, 346 14, 81, 183–184, 187, 192, 259–266,
European Economic Community, 6, 110, 269–270, 276, 278, 336–338, 345–347,
145, 147, 216, 239–240. See also 354, 360, 361, 364. See also Cartagena
European Union Protocol; Deliberate Release Directive
European Environment Agency (EEA), 117, primary effects of, 259
245–247, 255, 288, 299–300 secondary effects of, 259, 262–263, 265–
European Parliament, 6, 23–25, 91, 223, 271, 279–280
232, 240, 253, 289, 297–298, 346 Germany, 69–70, 91, 106, 109, 120–121,
European Union (EU) 168, 223, 253, 269
accession countries of, 218, 243–245 (see Global Environment Facility (GEF), 55–57,
also European Union, enlargement of ) 80, 345–346, 365
Burden-Sharing Agreement on cutting Governance, 1, 5–9, 13–16, 21, 23, 26, 34,
greenhouse gas emissions of, 37, 54, 58, 39, 41–42, 44–45, 53, 104, 207, 229,
68–74 308, 310, 313–315, 318, 329–331, 336,
comitology committees, 6, 25 338–339, 341–342, 344, 352, 357–360,
Common Agricultural Policy (CAP) of, 218, 364, 366–367
238, 247, 271 EU environmental, 1–3, 9, 11–13, 227,
Common Fisheries Policy (CFP) of, 128– 308, 316–318, 321, 323
130, 144–149, 238, 311 global, 20
Common Transport Policy of, 238 international environmental, 1, 3, 9, 11–13,
enlargement of, 243–244, 255 (see also 54, 128, 308, 316–318, 321, 323, 352
European Union, accession countries of ) Greece, 69, 119–120, 160, 250
402 Index
Greenhouse gases (GHGs), 21, 29, 37, 53– 289, 295, 303, 316, 321, 326, 332, 352,
59, 61–69, 71–73, 221, 336, 347. See also 354
Climate change regime; Kyoto Protocol; ideal types of, 11, 13–14, 325–351
United Nations Framework Convention intentionality of, 13–14, 40–41, 97, 139,
on Climate Change 184, 211, 293, 302, 312–313, 317, 321,
Greenpeace, 119, 270 327–331, 347, 350, 355
recombination of cases of, 15, 31, 359,
Habitats Directive, 3, 10, 15, 30–31, 56, 82, 363–364, 366
208–210, 219–221, 233–239, 241–256, vertical, 10–13, 45, 56, 73, 165, 186, 208,
348–349, 352–354, 356, 360, 363, 365. 210–211, 214, 219, 222, 228, 236, 264,
See also Natura 2000 285, 289–290, 302–303, 316–323, 332–
and Sites of Community Importance, 234– 333, 344–346, 352–354
235 Institutions. See International institutions
Special Areas of Conservation (SACs) of, Intellectual property rights (IPR), 27, 79, 81,
234, 241 85–89, 93, 95–97, 182 (see also World
Special Protection Areas (SPAs) of, 234 Trade Organization, Trade Related
Aspects of Intellectual Property Rights
Iceland, 133 Agreement of )
Impact-Level Interaction, causal mechanism Interaction. See Institutional interaction
of, 9, 11, 20, 39, 41–42, 44 Interaction through Commitment, causal
Implementation review, 29 mechanism of, 8, 10, 12, 14, 20, 36–38,
India, 86, 172 42, 44, 56, 58, 60, 68, 71, 73, 84–85,
Institutional Dimensions of Global 108–109, 121, 129, 136, 140, 148–149,
Environmental Change (IDGEC), 4, 21 161,165, 186, 191, 199, 208, 210, 219,
Institutional interaction. See also Behavioral 222, 228, 236, 239, 254, 263–264, 290,
Interaction; Cognitive Interaction; Impact– 293, 308, 312, 314, 319, 325–326, 333–
Level Interaction; Interaction through 350, 360
Commitment Additional-Means type, 15, 335, 341–345,
analysis of, 9, 11, 26, 29, 31, 308–309, 327 348–349, 351, 353, 360–363
cases of, 3, 7–11, 19–20, 27, 29, 31–33, Jurisdictional-Delimitation type, 14, 335–
35, 45, 307, 309–316, 357, 359–360 339, 344–347, 351, 357, 360–363
causal chains of, 10–11, 28, 33, 41, 44, Nested-Institutions type, 29, 335, 339–342,
308, 358–364 344–345, 347–348, 360, 362
clusters of, 16, 31, 82–84, 261, 263, 359, Intergovernmental Panel on Climate Change
364–366 (IPCC), 55, 62, 64
complexity of interaction situations, 7, 15, International Civil Aviation Organization
20, 26, 29–31, 211, 356, 358–359, 363– (ICAO), 53, 57–68, 73, 347
364 International Commission for the
core characteristics of cases of, 308–310, Conservation of Atlantic Tunas (ICCAT),
350–351 181, 184–186, 193–201
effects of (see Effects of interaction) International Convention for the Protection
existing analytical approaches, 5, 20–22 of the Marine Environment from Pollution
disaggregation of complex situations, 7, by Ships (MARPOL), 61, 64, 364
29–31, 359, 364 International Council for the Exploration of
horizontal, 10–13, 21, 45, 56, 73, 161, the Sea (ICES), 128–131, 139–150, 348,
176, 208, 219, 228, 236, 261, 263, 286, 362
Index 403
International Criminal Police Organization Issue linkage, 21–22, 42, 104, 134, 301,
(Interpol), 14, 46, 162–164, 330, 334, 314
353, 361, 366 Italy, 69–70, 120, 160, 214, 268
International institutions
broader consequences of, 4, 16, 366–367 Japan, 69–71, 86, 134, 193–194, 278
coevolution of, 30–31, 38, 360–361 Jurisdictional Delimitation. See Interaction
effectiveness of, 1, 3–8, 13–16, 20–21, 26, through Commitment
32, 34, 42–46, 310, 329–331, 333, 337–
339, 341–343, 351–352, 358–359, 364, Knowledge, 5, 8–9, 11, 34–35, 40–42, 113,
367 132, 135, 143, 167–169, 236
financial mechanisms of, 55–57, 345–346, Kyoto Protocol, 3, 9, 14, 26–27, 29, 33, 37,
365 39, 42, 53–74, 83–84, 107, 328, 363.
impact of, 2, 20–21, 34–37, 39, 42, 44, See also Climate change regime; United
358 Nations Framework Convention on
issue areas of, 4–6, 8–9, 19–20, 23–24, Climate Change
27–29, 33–34, 38–42, 44, 313, 325, 329,
331, 335–337, 339, 342, 347, 349–350, Landfill Directive (of the EU), 57–58
361, 363, 365 Large Combustion Plants (LCP) Directive (of
memberships of, 3, 8, 14–15, 36, 38, 309, the EU), 30, 206, 219–220, 222–228. See
313, 330, 335, 339–343, 348 also Acid rain; Convention on Long-Range
outcome of , 7, 34–35, 39–42, 329, 362 Transboundary Air Pollution
output of, 7, 34–38, 40–44, 349–351, Law, quality of
360–362 hard law, 104, 116, 118, 214, 313, 343
International Maritime Organization (IMO), soft law, 15, 103–104, 108, 112, 114, 118,
24, 53, 57–68, 73, 347, 364–365 121, 205, 212, 215, 285, 291, 302–304,
International North Sea Conference (INSC), 313, 362
3, 103–104, 106–107, 109–110, 112– supranational law of the EU, 15, 72, 205,
122, 142, 209, 343–344, 362, 364–365 213, 228, 343
International organizations, 5, 6, 24. See also Leadership, 68, 73, 104, 106, 301, 342
International institutions Learning, 8, 12, 14, 19, 22, 36, 68, 73, 108,
International Plant Protection Convention 131, 144, 150, 270, 325, 327–331, 333–
(IPPC), 272, 276–278 334, 361
International regime for the protection of the Living modified organisms (LMOs). See
Northeast Atlantic. See Oslo and Paris Genetically modified organisms
Conventions; International North Sea London Dumping Convention, 107–108,
Conference 340, 364
International regimes, 4–6, 19–20, 22–26, Luxembourg, 68–69, 120, 268
28, 36, 41, 45, 103, 333, 349, 366. See
also International institutions Mexico, 86, 91, 167, 194
International Union for the Protection of Montreal Protocol, 3, 14, 21, 29, 35, 39,
New Varieties of Plants (UPOV), 85, 88, 41, 54, 56–58, 61, 107, 184–185, 328,
90–91, 94 332–333, 336, 345, 347. See also Ozone
Ireland, 69, 120, 165, 214, 250, 253 layer
Issue areas. See also International Multilateral environmental agreements
institutions, issue areas of overlap of, 8, (MEAs), 2–4, 96, 164, 182–184, 186–
12, 21, 36, 38, 87, 335–336 187, 205, 278
404 Index
National Emissions Ceilings (NEC) Directive 150, 183, 188, 190, 260–262, 264, 266,
(of the EU), 289–290 271, 273–279, 347
Natura 2000 (network of protected sites), Preference constellation, 38
15, 209–210, 233–239, 242–247, 251–
254, 348, 360, 363. See also Habitats Ramsar Convention on Wetlands, 56–57,
Directive 59, 82–83, 96, 159, 165, 209, 211, 366
Nested Institutions. See Interaction through Renewable Energy Directive (of the EU), 57–
Commitment 58, 289–290
Netherlands, 69, 91, 109, 120, 253, 264 Request for Assistance. See Cognitive
New Zealand, 67, 133 Interaction
Nitrates Directive (of the EU), 3, 107, 115, Russia, 142–143, 167, 171–172
117, 119–122, 208–209
Noncompliance. See Compliance Scientific uncertainty, 105, 276–279, 318
Nongovernmental Organization (NGO), 8, SCM Agreement. See World Trade
20, 27, 34, 36, 38–42, 160, 176, 214, Organization
240, 252, 255, 270, 288, 328, 351, 357 Single European Act, 112, 165, 241
North Atlantic Treaty Organization Solvent Emissions Directive (of the EU),
(NATO), 28, 39 219–220, 365
Northwest Atlantic Fisheries Organization Source institution, 6–11, 14, 16, 19–20, 27–
(NAFO), 130, 143–144 29, 31–36, 38–46, 307–367. See also
Norway, 63, 67, 109, 117, 146 Institutional interaction
South Africa, 86, 91
Organization for Economic Cooperation SPS Agreement. See World Trade
and Development (OECD), 131–132, Organization
134 Sweden, 69, 109, 120, 268
Oslo and Paris Conventions (OSPAR), 15, Synergy. See Effects of interaction, synergistic
59, 104–105, 107–109, 112, 114–122,
209–211, 221, 236–237, 340, 343–344, Target institution, 6–8, 10–11, 14, 19–20,
362, 364 27–29, 31–46, 307–367. See also
Commission of, 104, 107–109, 111–114, Institutional interaction
117–118 TBT Agreement. See World Trade
Ozone layer, 3, 14, 21, 29, 39, 41, 54, 57, Organization
107, 185, 328, 336. See also Montreal Thailand, 160, 195
Protocol Trade Records Analysis of Fauna and
Ozone regime. See Montreal Protocol Flora in Commerce (TRAFFIC), 160–161,
173
Patent Directive (of the EU), 83, 91, 93, TRIPS Agreement. See World Trade
349 Organization
Pesticides Directive (of the EU), 260, 262–
263, 265–271, 279–280, 346 Urban Waste–Water Directive (of the EU),
Persuasion, 36–37 107, 115, 117, 119
Philippines, 133, 196 United Kingdom (UK), 69–70, 108–110,
Policy Model. See Cognitive Interaction 112, 118–121, 223, 226, 254, 264, 268–
Polluter-pays principle, 212 270, 299
Precautionary approach/principle, 106, 110, United Nations Conference on Environment
112, 118, 121, 128–131, 140–146, 148, and Development (UNCED), 80, 84, 86
Index 405
United Nations Conference on Trade and Agreement on Sanitary and Phytosanitary

Development (UNCTAD), 95 Measures (SPS Agreement) of, 182–183,
United Nations Convention on the Law of 185, 189, 191, 262, 264, 271–280, 360,
the Sea (UNCLOS), 21, 128–129 362–364
United Nations Economic and Social Council Agreement on Technical Barriers to Trade
(ECOSOC), 84 (TBT Agreement) of, 182–183, 189, 196,
United Nations Economic Commission for 198, 259, 262, 264
Europe (UNECE), 22, 205–206, 211–212, Appellate Body of, 182, 187, 197
215–216, 221, 290, 364 and Codex Alimentarius Commission, 193,
United Nations Environment Program 272, 276, 364
(UNEP), 2, 55, 97, 135, 137, 158, 160, Committee on Trade and Environment
173, 262, 264–265 (CTE), of 135–137, 164, 183–184, 186,
United Nations Fish Stocks Agreement 198, 201
(FSA), 128–130, 136–142, 144–145, and dispute settlement, 92–93, 133, 181–
147–150, 200, 311, 348, 362 184, 186–187, 197, 262, 264, 271–272,
United Nations Framework Convention on 274–278, 363
Climate Change (UNFCCC), 3, 9, 26, 29, enforcement mechanisms of, 93, 139, 184
53–55, 61–63, 65, 68, 72, 74, 84, 107– judicial decisions of, 187
108, 221. See also Kyoto Protocol; Ministerial Conference of, 137, 183–184,
Climate change regime 193, 278
United States (U.S.), 63, 69–71, 74, 86, 91, objectives (free-trade objectives) of, 46,
94, 133, 167, 189, 192–195, 259, 270, 161–162, 181, 184, 200, 271, 337
273, 275–276, 278, 291 secretariat of, 183, 186–187, 198
Subsidies and Countervailing Measures
Water Framework Directive (of the EU), Agreement (SCM Agreement) of, 132–
116–117, 205–216, 219, 221, 227–228, 133, 137
236–237, 348, 352–353, 365 Trade Related Aspects of Intellectual
World Bank, 57–59, 134 Property Rights (TRIPS) Agreement of, 27,
World Conservation Monitoring Centre 201, 79–81, 83–89, 91–98, 363 (see also
(WCMC), 160–161 Intellectual property rights)
World Conservation Union (IUCN), 157, World Wide Fund for Nature (WWF), 134,
159–161, 166–173, 176 138, 160–161, 173, 249
World Customs Organization (WCO), 14,
46, 157, 162–164, 177, 330–332, 334,
353, 361, 366
World Health Organization (WHO), 56,
285–287, 288–298, 301–303, 315, 349
Air Quality Guidelines of, 290–295
World Intellectual Property Organization
(WIPO), 85, 89, 94, 96–97
World Trade Organization (WTO), 3–4, 10,
14, 21–22, 24, 27, 29, 39–40, 56–59, 79–
98, 129–139, 149–150, 161–164, 176,
181–201, 259–260, 262, 264, 271–280,
311–312, 315, 333–339, 345, 347–348,
360–363
Global Environmental Accord: Strategies for Sustainability and Institutional
Innovation
Nazli Choucri, series editor
Nazli Choucri, editor, Global Accord: Environmental Challenges and International

Responses
Peter M. Haas, Robert O. Keohane, and Marc A. Levy, editors, Institutions for the
Earth: Sources of Effective International Environmental Protection
Ronald B. Mitchell, Intentional Oil Pollution at Sea: Environmental Policy and
Treaty Compliance
Robert O. Keohane and Marc A. Levy, editors, Institutions for Environmental Aid:
Pitfalls and Promise
Oran R. Young, editor, Global Governance: Drawing Insights from the
Environmental Experience
Jonathan A. Fox and L. David Brown, editors, The Struggle for Accountability:
The World Bank, NGOs, and Grassroots Movements
David G. Victor, Kal Raustiala, and Eugene B. Skolnikoff, editors, The
Implementation and Effectiveness of International Environmental Commitments:
Theory and Practice
Mostafa K. Tolba, with Iwona Rummel-Bulska, Global Environmental Diplomacy:
Negotiating Environmental Agreements for the World, 1973–1992
Karen T. Litfin, editor, The Greening of Sovereignty in World Politics
Edith Brown Weiss and Harold K. Jacobson, editors, Engaging Countries:
Strengthening Compliance with International Environmental Accords
Oran R. Young, editor, The Effectiveness of International Environmental Regimes:
Causal Connections and Behavioral Mechanisms
Ronie Garcia-Johnson, Exporting Environmentalism: U.S. Multinational Chemical
Corporations in Brazil and Mexico
Lasse Ringius, Radioactive Waste Disposal at Sea: Public Ideas, Transnational
Policy Entrepreneurs, and Environmental Regimes
Robert G. Darst, Smokestack Diplomacy: Cooperation and Conflict in East-West
Environmental Politics
Urs Luterbacher and Detlef F. Sprinz, editors, International Relations and Global
Climate Change
Edward L. Miles, Arild Underdal, Steinar Andresen, Jørgen Wettestad, Jon Birger
Skjærseth, and Elaine M. Carlin, Environmental Regime Effectiveness: Confronting
Theory with Evidence
Erika Weinthal, State Making and Environmental Cooperation: Linking Domestic
and International Politics in Central Asia
Corey L. Lofdahl, Environmental Impacts of Globalization and Trade: A Systems
Study
Oran R. Young, The Institutional Dimensions of Environmental Change: Fit,
Interplay, and Scale
Tamar L. Gutner, Banking on the Environment: Multilateral Development Banks
and Their Environmental Performance in Central and Eastern Europe
Liliana B. Andonova, Transnational Politics of the Environment: The European
Union and Environmental Policy in Central and Eastern Europe
David L. Levy and Peter J. Newell, editors, The Business of Global Environmental
Governance
Dennis Pirages and Ken Cousins, editors, From Resource Scarcity to Ecological
Security: Exploring New Limits to Growth
Ken Conca, Governing Water: Contentious Transnational Politics and Global
Institution Building
Sebastian Oberthür and Thomas Gehring, editors, Institutional Interaction in
Global Environmental Governance: Synergy and Conflict among International and
EU Policies

Sebastian Oberthur

Uploaded by

Copyright:

Available Formats

Sebastian Oberthur

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Sebastian Oberthur

Uploaded by

Copyright:

Available Formats

Institutional Interaction

foreword by Oran R. Young

edited by Sebastian Oberthür and Thomas Gehring

The MIT Press

Library of Congress Cataloging-in-Publication Data

Institutional interaction in global environmental governance : synergy and conﬂict among

Series Foreword vii

8 Interactions between the World Trade Organization and International

A new recognition of profound interconnections between social and natural sys-

we can proceed to build up an understanding of more complex interactions featur-

Sebastian Oberthür and Thomas Gehring

Clare Coffey Alice Palmer

APEC Asia-Paciﬁc Economic Cooperation

OSPARCOM OSPAR Commissions

International and EU environmental governance is affected by the interaction of

The Relevance of Institutional Interaction

The heavily fragmented institutional structure of international environmental gover-

We ﬁnd both synergistic and disruptive cases of institutional interaction in inter-

ronmental governance by presenting the results of a large comparative exploration

communication processes. EU member states and the European Commission decide

The Meaning of Institutional Interaction Generally, institutional interaction will

Causal Mechanisms of Institutional Interaction Causal mechanisms help us under-

from theories of international institutions, cooperation theory, and negotiation

The Design of the Empirical Investigation

Impact-Level Interaction because it is frequently based on complicated natural

Empirical Findings and Conceptual Development

Empirical Findings The comparative examination of a larger set of cases of inter-

actors may promote governance by establishing a smaller ‘‘pilot’’ institution in order

the original source institution or inﬂuences a third institution. Cases of interaction

This chapter introduces the shared conceptual framework of the contributions to

the target institution (Cognitive Interaction). Or it may commit its members to

Existing Approaches to the Study of Institutional Interaction

Much of the empirically founded research on institutional interaction has so far

nesting of the Multi-Fiber Agreement within the General Agreement on Tariffs

International and EU Institutions as the Interacting Units

In the present volume, we focus exclusively on deliberately established (‘‘negoti-

Both formal international organizations and more loosely institutionalized regimes

Despite the speciﬁc conditions of EU policymaking, such EU legal instruments share

directives and regulations, always including the associated decision-making pro-

The Concept of Institutional Interaction

In this section, we develop a concept of institutional interaction (see also Gehring

Establishing a Cause-Effect Relationship between Institutions

To establish a case of institutional interaction, we must explore the inﬂuence

Dealing with Complexity: Identifying ‘‘Cases’’ of Interaction

The strategy of disaggregating complex situations into a suitable number of clear-

Causal Mechanisms of Institutional Interaction

The Concept of Causal Mechanisms

Interaction Inﬂuencing the Decision-Making Process of the Target Institution

Cognitive Interaction The decision-making process of an international institution

Interaction through Commitment Institutional interaction directed at the decision-

Possible Further Causal Mechanisms It is difﬁcult to imagine that an international

to inﬂuence the decision-making process of the target institution directly, actors

Interaction Inﬂuencing the Target Institution at the Outcome Level: Behavioral

hypothetical causal pathways are hardly relevant in reality. In particular, we might

Interaction Inﬂuencing the Impact Level of the Target Institution

are modiﬁed upon Cognitive Interaction or Interaction through Commitment may

Consequences for the Empirical Investigation of Institutional Interaction

Fearon, James D. 1998. Bargaining, Enforcement, and International Cooperation. Interna-

The International Regime on Climate Change

gases, namely hydroﬂuorocarbons (HFCs), perﬂuorocarbons (PFCs), and sulfur