The Paradoxes of Transparency: Science and the Ecosystem Approach to Fisheries Management in Europe

Wilson, Douglas Clyde Book — Published Version The Paradoxes of Transparency: Science and the Ecosystem Approach to Fisheries Management in Europe MARE Publication Series, No. 5 Provided in Cooperation with: Amsterdam University Press (AUP) Suggested Citation: Wilson, Douglas Clyde (2010) : The Paradoxes of Transparency: Science and the Ecosystem Approach to Fisheries Management in Europe, MARE Publication Series, No. 5, ISBN 978-90-485-0813-6, Amsterdam University Press, Amsterdam, https://doi.org/10.5117/9789089640604 This Version is available at: https://hdl.handle.net/10419/181372 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc/3.0/ A U P ISBN 978 90 8964 060 4 www.aup.nl The International Council for the Exploration of the Sea (ICES) is the central scientific network within the massive set of bureaucracies that is responsible for Europe’s Common Fisheries Policy (CFP). While spending the past 25 years failing to sustain Europe’s fish stocks, this management system also became adept at making the lives of its scientists miserable. Now it is being confronted by the complex challenge of an ecosystem-based approach to fisheries management. If this combination of a multi-national bureaucracy, hard politics, and scientific uncertainty has made it impossible to maintain many individual fish stocks, how are decisions going to be made that consider everything from sea birds to climate change? The old political saw that “if you can’t solve a problem, make it bigger” has never been put to a test like this! Yet ICES has begun to rise in an impressive way to the scientific challenge of providing advice for an ecosystem approach within the world’s most cumbersome fisheries management system. This book lays out the results of extensive sociological research on ICES and the decision making systems into which it feeds. ICES is finding ways to provide effective advice in the many situations where scientific advice is needed but a clear, simple answer is out of reach. In spite of the difficulties, scientists are beginning to help the various parties concerned with management to deal with facts about nature in ways that are more useful and transparent. Doug Wilson is a Senior Researcher and Research Director at Innovative Fisheries Management – An Aalborg University Research Centre. 5 MARE PUblICAtIon SERIES 5 Douglas Clyde Wilson A U P Douglas Clyde Wilson The Paradoxes of Transparency Science and the Ecosystem Approach to Fisheries Management in Europe 5 The Paradoxes of Transparency A clear analysis of an extremely complex domain—a great achievement, beautifully written. The author leads the reader along with clear, well reasoned arguments, documenting, and explaining everything carefully along the way. James McGoodwin, Professor of Anthropology, University of Colorado This is a very good book. It is original, based on a massive amount of self-generated primary data, embedded in relevant theoretical and methodological debates, and superbly written. A work of real scholarship. tim Gray, Professor at the School of Geography, Politics and Sociology, University of newcastle aup_mare5_paradoxes.indd 1 26-06-2009 12:45:03 The Paradoxes of Transparency MARE PUBLICATION SERIES MARE is an interdisciplinary social-science institute studying the use and management of marine resources. It was established in 2000 by the University of Amsterdam and Wageningen University in the Netherlands. MARE’s mandate is to generate innovative, policy-relevant research on marine and coastal issues that is applicable to both North and South. Its programme is guided by four core themes: fisheries governance, maritime work worlds, integrated coastal zone management (ICZM), and maritime risk. In addition to the publication series, MARE organises conferences and workshops and publishes a social-science journal called Maritime Studies (MAST). Visit the MARE website at http://www.marecentre.nl. series editors Svein Jentoft, University of Tromsø, Norway Maarten Bavinck, University of Amsterdam, the Netherlands previously published Leontine E. Visser (ed.), Challenging Coasts. Transdisciplinary Excursions into Integrated Coastal Zone Development, 2004 (isbn 978 90 5356 682 4) Jeremy Boissevain and Tom Selwyn (eds.), Contesting the Foreshore. Tourism, Society, and Politics on the Coast, 2004 (isbn 978 90 5356 694 7) Jan Kooiman, Maarten Bavinck, Svein Jentoft, Roger Pullin (eds.), Fish for Life. Interactive Governance for Fisheries, 2005 (isbn 978 90 5356 686 2) Rob van Ginkel, Braving Troubled Waters. Sea Change in a Dutch Fishing Community, 2009 (isbn 978 90 8964 087 1) The Paradoxes of Transparency Science and the Ecosystem Approach to Fisheries Management in Europe Douglas Clyde Wilson MARE Publication Series No. 5 Amsterdam University Press Cover illustration: Alyne Delaney and Douglas Clyde Wilson Cover design: Neon, design and communications, Sabine Mannel, Amsterdam Lay-out: japes, Amsterdam isbn 978 90 8964 060 4 e-isbn 978 90 4850 813 6 nur 741 © Doug Wilson/Amsterdam University Press, Amsterdam 2009 All rights reserved. Without limiting the rights under copyright reserved above, no part of this book may be reproduced, stored in or introduced into a retrieval system, or transmitted, in any form or by any means (electronic, mechanical, photocopying, recording or otherwise) without the written permission of both the copyright owner and the author of the book. SERIES FOREWORD As editors of the MARE Publication Series, we are proud to present yet another major work on people and the sea. The topic of Doug Wilson’s important and timely book is the role of natural scientists in fisheries management and environmental governance. Its focus is on the institutions that provide the scientific basis for decision-making with regard to European fisheries policy. A prominent organisation in this context is the International Council for the Exploration of the Sea (ICES), which involves twenty member states and serves as the hub of a network of approximately 1600 scientists. The marine environment is difficult to observe so the scientific uncertainty is very high. Hence the crafting of scientific advice for an ecosystem approach to fisheries management is a complex challenge. How do scientists communicate uncertainty among themselves and with the outside world? And how well does science mix with advice? Wilson lucidly discusses these and other important questions. Drawing on “communicative systems theory”and employing a wide array of data collection methods, Wilson provides deep insights into the challenges and dilemmas involved in providing scientific advice to a demanding political process. Can science really deliver what stakeholders expect and policy makers are asking for, i.e. rapid answers with minimum uncertainty to complex issues, without compromising what science is meant to be? These are all pertinent questions for fisheries management but they also have more general relevance. Indeed, fisheries may well provide a test case for our capacity to deal with a range of environmental issues, in which science is called upon to provide the knowledge base necessary for effective and rational collective action. Thus, this book is also a contribution to the Science and Technology Studies that are now enjoying widespread interest within academic circles and beyond. Finally this book is valuable in helping to cross the divide between natural and social sciences. It demonstrates how sociology of science perspectives and methods can help us understand the contribution of natural sciences, and their representatives, to resolving complex societal issues. Svein Jentoft (Norwegian College of Fishery Science, University of Tromsø, Norway e-mail: svein.je[email protected] Maarten Bavinck (University of Amsterdam, the Netherlands) e-mail: J.M.Bav[email protected] To Prof. Joe Francis Many wonderful people contributed to my formal education, but 20 years is time enough to know which lessons proved the most useful. Others did more to help me learn what to think about, but no one did more to help me learn how to think. NASCO North Atlantic Salmon Conservation Organization NEAFC North East Atlantic Fisheries Commission NFI National Fisheries Institute NGO Non-governmental organization NOAA National Oceanographic and Atmospheric Administration (USA) NSCFP North Sea Commission Fisheries Partnership NUSAP Number, unit, spread, assessment, pedigree OPEC Organization of the Petroleum-Exporting Countries OSPAR The OSPAR Convention for protecting the NE Atlantic Ocean PKFM Policy and Knowledge for Fisheries Management project PNS Post-normal science RAC Regional advisory council REGNS Regional Ecosystem Study Group for the North Sea RSE Royal Society of Edinburgh SAFMAMS Scientific Advice for Fisheries Management at Multiple Scales project SCICOM Science committee SGFI ICES/NSCFP Study Group on the Incorporation of Additional Information from the Fishing Industry into Fish Stock Assessments SGGROMAT Study Group on Growth, Maturity and Condition in Stock Projections SGPRISM Study Group on Incorporation of Process Information into Stock Recruitment Models SSB Spawning stock biomass STECF Scientific, Technical and Economic Committee for Fisheries STS Science and technology studies TAC Total allowable catch ToRs Terms of reference VMS Vessel monitoring system VPA Virtual population analysis, a type of stock assessment model WGECO Working Group on Ecosystem Effects of Fishing Activities WGNSSK Working Group on the Assessment of Demersal Stocks in the North Sea and Skagerrak WGRED Working Group for Regional Ecosystem Description 14 The Paradoxes of Transparency Preface The book consists of three theoretical chapters, four empirical chapters, and a conclusion that seeks to pull the whole thing together. The theoretical chapters explain a number of concepts that I use to organise and relate the empirical material, which is mainly a description of the work of the International Council for the Exploration of the Sea (ICES). Fisheries management is an area of environmental management that contains many potential lessons for how we can relate to our planet in general. The point of social theory is to provide vocabularies and concepts that allow lessons learned in one kind of social endeavour to be translated for use in other kinds. Hence the value of this book, to me, lies as much or even more in the theoretical discussion than in the empirical work. It helps me, and I hope others, when thinking about new problems. However, I realise that the interests of many readers will be simply on fisheries and marine management in Europe, and that what happens in ICES is what they will want to read about. Those readers will likely wish to start with Chapter Four. This should work fine for their purposes. However, I would encourage this group to skim Chapter One for the sake of orientation. I would also suggest that they read Section 2.2.1 about Mode Two Science and Section 2.2.3 about Post-Normal Science and examine Section 3.1 long enough to get a sense of the way I am using the terms saliency, credibility and legitimacy. I believe this effort will lead to a quicker understanding of the presentation of the case material. Douglas Clyde Wilson Hirtshals, March 2009 15 Acknowledgements It is well beyond my ability to list all of the people who contributed to this book. For one thing, the most important group is very large. It includes the many fisheries scientists, fishing industry members, conservation advocates, fishery managers and policymakers across Europe who allowed themselves to be interviewed, questioned and/or observed. They were, almost to a person, helpful and forthcoming. Particular mention should be made of the ICES staff and leadership who directly facilitated this research in different ways: Bodil Chemnitz, Paul Connolly, Gerd Hubold, Hans Lassen, Solveig Lund, Inger Lützhøft, Martin Pastoors, Vivian Piil, and Poul Degnbol, who later also guided me through the labyrinth of DG MARE. Another group deserving special mention are those PKFM colleagues who carried out interviews that I have quoted in the text: Alyne Delany, Petter Holm, Kåre Nielsen, and Jesper Raakjær. Three EU Framework Research Projects have supported this work at various times: the Fifth Framework Programme supported Policy and Knowledge in Fisheries Management (PKFM), Contract no. Q5RS-2002-01782; the Sixth Framework Programme supported Scientific Advice for Fisheries Management at Multiple Scales (SAFMAMS), Project no. 013639; and the Seventh Framework Programme supported Judgement and Knowledge in Fisheries Including Stakeholders (JAKFISH), Grant Agreement no. 212569. This work does not reflect the Commission’s views and in no way anticipates its future policy in this area. Finally, this book would not have happened, and it certainly would not have been readable, without the dedication of Kirsten Klitkou, Innovative Fisheries Management’s secretary, meaning in reality our all-round document coordinator and quality control officer. 17 1. Introduction 1.1 Sensing the need for change The topic of this book covers institutional aspects of providing natural science advice for the fisheries management programmes of the European Union and its immediate neighbours, particularly as they turn towards the Ecosystem Approach to Fisheries Management (EAFM). In many ways it is a straightforward application of the sociology of science. It applies various research methods and theoretical perspectives to the work of fisheries scientists in Europe to try to draw some insights on how we can do a better job structuring science for environmental decision-making in general and for the marine environment in particular. That is the topic but not what the book is really about, to me at least. What drives me to write about the way that science fits into decision-making is an obsession with adaptation. I was trained as a human ecologist; I learned to think of society as a set of social systems embedded in a surrounding set of natural systems. What I know of the sociology of science I learned well after graduate school. I was drawn to it by my reflections on the four simple steps that General Systems Theory outlines for all adapting systems. If a system is going to adapt to its environment it must: 1) sense the need for a change; 2) have its own potential to change in response; 3) have a way to select the change to make; and, 4) have a way to implement and maintain the change (Buckley 1967). The first and third steps rely on a good understanding of nature while the second and fourth steps require the system to be capable of effective collective action. So how does a social system ‘sense the need for a change’when it must adapt to changes in its natural environment? The term ‘social system’can mean a lot of things. In my work, I think of a social system as a patterned set of communicative actions because this definition allows a firm distinction between what is social and what is natural. Society and nature are so intertwined that such a distinction would otherwise be impossible. Is a farmed field social or natural? Defining society as a set of communicative actions means that the substance out of which society is built is shared meaning. This has some strange implications –for example it places our bodies in society’s environment –but it makes for a clear definition of the relevant system boundaries from both the ontological and epistemological angle. If the laws of thermodynamics apply, then the phenomenon is part 19 of the environment of the social system. This also allows, in principle, human ecological reasoning to be applied in lots of areas that are not part of environmental studies, for example to health care. Communicative Systems Theory (CST) (Habermas 1984, 1987; Wilson 2003) provides a set of conceptual tools useful for understanding society as a set of communicative actions. An approach that focuses on the realities between actors, i.e. shared meanings and the systemic requirements for reproducing them, provides a meaningful complement to the atomistic analysis of actors and their incentives that currently dominates social science. If there is interest in the volumes of argumentation for why this ‘communicative turn’makes for good social theory and hence good human ecology, then I suggest going directly to Habermas. Otherwise, as Habermas (1990) himself argues, you will have to judge for yourself whether this approach offers a coherent account of the case at hand. One thing that I hope this book demonstrates is that CST can make a contribution to Science and Technology Studies (STS), a broader field of which the sociology of science is just a part. A great deal of valuable recent work in STS has emphasised the alloyed social and technical nature of the modern world, and particularly the products of science and policy. The idea is that in most aspects of modern social life, scientific and social issues are woven together into complex hybrids (Freudenberg et al. 1995; Holm 2007; Latour 1991). Nature and technology are important drivers of social structures, and social structures determine understandings of nature. The co-production (Jasanoff 2005) of science and society through creative acts equally rooted in humanity and nature is ubiquitous. The zeitgeist of this literature is that an analysis that seeks to maintain the dichotomy between the social and the natural is unhelpful at best and likely to lead to illusion and error at worst. While being unable to deny the empirical importance of the co-production of hybrid phenomena, I have never been comfortable with the conclusion that an analytical distinction between society and nature is unhelpful. One reason, mentioned above, is that the distinction is such a central one to human ecology. Furthermore, a background in critical theory makes one very distrustful of anything that suggests that reasoning about society and nature are similar exercises –I view the naturalisation and technicization of human interactions as a dangerous path. CST allows us to recognise the hybrid nature of a technological society while maintaining a strong analytical distinction between society and nature. This distinction finds practical expression in differences in how people communicate with one another with respect to facts on the one hand, and values and interests on the other. The central CST concept of ‘rational communication’is based on the requirement of establishing mutual understanding between two or more parties. Part of rational communication is the idea that the presuppositions of different kinds of communication are oriented according to different principles. As I will mention frequently and in more detail below, a discussion leading towards mutual under20 The Paradoxes of Transparency standing of facts presupposes the goal of consensus, while a discussion leading towards an understanding about values and interests presupposes a goal of compromise. This distinction does not ignore the emergent properties of socio-technical systems where facts, values and interests are connected in the substance of nearly all the issues we grapple with. It does, however, introduce a type of rationality which reflects our basic intuitions about how people understand each other. It is a basis for recognising the special role of the scientific assertion, and by extension a special role for scientists whose expertise is based on generating such assertions in a transparent way. This recognition, however, does not rely on the notion of science as a white-coated, objective other. For a communications system to be able to adapt, ‘sensing the need for change’has to be more than someone knowing that something important is going on in nature. It has to move to the next step of finding the potential to change in response –most generally, this means it has to become the basis of some sort of collective action, a term I use very broadly to include everything from effective policies and legislation to decisions taken at a community level. The knowledge about nature also has to be turned into something that people talk about, i.e. it has to become part of a discourse, and this talking has to be translated into some kind of collective action that has a chance to bring about change. First we need knowledge, and then we need that knowledge to influence collective action. Knowledge in itself is a complicated social idea. Everybody has some knowledge, but the knowledge that scientists have has a special quality. In environmental studies we tend to think of two groups as having special knowledge about nature: the scientists who have research-based knowledge, and user groups that have experience-based, local, ecological knowledge. The scientists’knowledge is thought to be the best quality. Most people think that when a scientist says that some fact about nature is ‘science’, then that fact is as true as can be. I spend a lot of time with fisheries scientists, and when they tell me things about fish, I almost always believe them without question, and if I do not believe them right away, it is because some other scientist has told me something different. But being somehow truer is not the real, special quality of scientists’ knowledge. After all, they frequently change their minds. What makes the scientists’knowledge special in the imagination of human ecology is the way it links to the next step that is needed if the social system is going to adapt. Science has a uniquely high potential to be the knowledge base for collective action. This uniquely high potential is not a product of science’s truth but of science’s radical commitment to transparency. The scientific method is all about transparency. It is about the clear articulation of knowledge in a way that can be clearly challenged, i.e. held clearly accountable for the truth of its assertions. This idea has an ironic tinge that stems, in fact, from the paradoxes of transparency after which this book is named. Science uses special tools and techniques, especially quantification, to create claims of truth that are 1. Introduction 21 clearly stated and able to be challenged. But these very techniques, meant to guard transparency, require skill and training to understand. They have also developed into cultural images with great rhetorical power. In practice, one might say, science suffers from transparency-induced opacity. Effective adaptation is only possible when there is a link between learning and the decision-making that guides collective action. Our natural environment is large and complex. Keeping track of changes, whether brought about by human action or natural processes, requires large numbers of people and technologies, purely from the information angle. In practice, many of these people will be carrying out direct or indirect environmental monitoring as part of other activities such as extracting goods from the environment, taking pleasure in it, seeking to conserve it, or managing it on behalf of the public. All these activities generate knowledge, some of which will be useful for environmental management. Some of this knowledge will be systematic, but most of it will be anecdotal. This systematic/anecdotal distinction is rooted, as are so many other things important to human ecology, in the question of scale. What makes anecdotal information less useful is not that it is less true; it is that it cannot be linked to information from elsewhere in order to characterise a phenomenon happening on a broader scale level. The point is that the systematic information needed to give a picture of widescale phenomena will rarely be available when environmental changes requiring adaptation are being recognised. Recognition will be based on anecdotal information that must then be supplemented by or formed into systematic information. In short, the knowledge base for adaptation requires a two-way process. It must tap into a lot of diverse information sources, and it must also find ways to shape that knowledge to get a broader picture. Thus, the need for wide and interactive participation in environmental management arises simply through information requirements even before we consider the decision-making aspects of governance. Interactive participation (aka authentic participation, participatory democracy, discursive democracy, etc.) makes effective governance possible within an ecosystem approach. The translation of decisions into action is much easier when more people buy into decisions and the science they are based on. I do not intend to devote space here to showing that broad participation is necessary for good governance. This has been done extensively elsewhere by me and many other people. My general perspective is perhaps not as strong as that of many advocates of participatory governance in fisheries. While participation is one of a set of characteristics of effective environmental management institutions, the most basic of these requirements is that the institutions are rooted in the authority of a democratic government. This is a basic lesson from fisheries co-management systems around the world. Bottom-up effectiveness through participatory democracy happens best in a top-down, formal democracy context. Without this underlying top-down system, co-management and participatory democracy are lost because accountability has to operate both up and down. In any 22 The Paradoxes of Transparency event, here I assume rather than defend the need for broad participation in environmental management and focus on the question of how participation in building a knowledge base by and with scientists is best achieved. Participation and science do not go easily together. We analyze participation by observing the various ‘discoursive themes’that different groups draw on in their discussions and arguments (Klenke et al. 2009). Such themes are formed by linking together facts, values and interests into coherent positions and arguments. Certain facts fit in with certain values and reflect certain interests. By interest here I mean economic and political interests, but there is also an illuminating connection to the other meaning of the term; your work and your values influence what you find personally interesting. The linkages between facts, values and interests have a direct impact on the ways that facts are learned, selected and presented to others. All too often they lead people to stubbornly ignore facts that are not linked to their values and interests. This can be the result of simple recalcitrance, but it happens among people who are committed to finding solutions as well. It is not some sort of moral failing, although it often feels that way when we listen to the people who oppose our positions. It makes arriving at collective action difficult, but not impossible. The way we talk about natural facts 1 is different from how we talk about values and interests. This comes from a principle of ‘communicative rationality’: all speech aimed towards mutual understanding takes place against a cultural ‘preunderstanding’which differentiates between claims about what is true, claims about what is right, and claims about what is sincere (Habermas 1984, p. 100). Claims about natural facts fall in the first category, while claims about values and interests fall in the second and third. The latter are subject to negotiation, and when we discuss them, we are seeking to find a compromise that people can live with. Discussion of facts, however, consists of trying to demonstrate that something is true, and it presumes the goal of agreement about that truth. To arrive at an agreement about what is going on, for example in the environment, participants have to explain how they know what they say they know. The tools and techniques of science are the Rolls Royce of explaining how you know what you know because of the commitment to radical transparency. This is not an ideal that is ever reached; even on a philosophical level, all that can be shown is that something has not yet been disproved, and in the day-to-day practice of doing science to support decision-making, it remains distant. Nevertheless, it is this methodological quality, not the truth of any particular demonstration, that makes science the best basis for developing a knowledge base for collective action. Because broad participation is needed both to mobilise information and for effective governance, striving for this ideal form of demonstration cannot be left only to scientists. Other groups must also participate in creating a knowledge base for adaptive collective action. The elephant standing on the table when applying all this theory to fisheries, and especially to the EAFM, is scientific uncertainty. Because of un1. Introduction 23 managing fisheries, because it solves important political problems around the division of fish resources among EU member states. The CFP is one of the few areas where member states have given EU institutions full decision-making power. The Council of Ministers makes decisions about the management of European fisheries resources beyond 12 nautical miles from each country’s shoreline. Arriving at this agreement required the creation of the rule of ‘relative stability’under which the EU cannot change the historical share that member state fleets have enjoyed for various fish stocks. The main job of the CFP has been to conserve fish stocks while allocating the allowable catch among member states following the relative stability rule. This political requirement fits very nicely with quotas that provide a mechanism for dividing up the fish. The quotas in turn are generated by a group of age-structured stock assessment models. What age-structured models do is follow groups of fish in the same ‘age class’, i.e. fish that are born the same year, through their life spans while trying to keep track of how many fish of a particular age live to become one year older. These models produce, in principle, an estimate of how many fish can be sustainably removed from a fish stock each year. They require a massive amount of scientific effort and data, and therefore a huge infrastructure has grown up around these annual fish stock assessments. The result is a mutually reinforcing system of political and scientific institutions that works well in terms of EU politics, but much less well in terms of sustainability (Holm and Nielsen 2004). The scientific advice system is being asked to change. Several priorities are emanating from ICES clients, who are themselves responding to shifts in European environmental politics. The CFP has failed in terms of sustainability. Indeed, Sparholt et al.’s (2007) analysis of fish stocks in the ICES area found that the implementation of the CFP has had no real impact on the condition of fish stocks. The pelagic stocks that are in relatively good condition currently were already being fished at a moderate level before the CFP was implemented in 1983. More tragically, the high fishing pressure on the demersal stocks that are currently in very poor shape simply continued. Both the fishing industry and conservation groups are deeply dissatisfied with the situation. The industry is pushing for more longterm management plans that allow rational business planning. The conservation groups want to reduce fishing pressure, implement marine protected areas and bring about an ecosystem-based approach to fisheries management (EAFM). The European Union’s developing Marine Strategy strongly reflects these concerns, particularly the EAFM. Both long-term management plans and the EAFM present new and extremely complex problems for scientific advice; the TAC Machine will soon be as politically inadequate as it is environmentally inadequate. How it can be changed, and what it can be changed into, is the challenge that European marine scientists are trying to meet. 30 The Paradoxes of Transparency 1.4 Research methods The qualitative aspects of case study research include in-depth interviews, observation of various public and private meetings, and review of many documents. Notes from observations, informal interviews, and original documents were analysed using NUD*IST textual data analysis software. A detailed description of this process as it was applied to the particular qualitative research goal of identifying the main themes in the discussions around a reorganisation of the ICES Advisory Programme is found in Section 7.3. This research also makes extensive use of the anthropological method of participant observation. Part of the time I was researching and writing this book, I was also serving as the Chair of the ICES Working Group on Fisheries Systems. This not only means that I was a participant in the system I was studying; the assigning of a sociologist to such a role is a reflection of the opening to new perspectives that characterises recent changes in ICES. As any participant observation would, I am sure that this introduces some biases. Indeed, my fisheries science colleagues often tease me about whether I am being an observer or a participant at any given moment. I cannot be aware of the biases that this participation introduces to my analysis –although I do admit to a general admiration of fisheries scientists – so I must be content with cautioning the reader that it is present. This book is based mainly on research that was supported by two European research projects. The Policy and Knowledge in Fisheries Management (PKFM) project, which ran from 2003 through 2005, supported the detailed observation of five scientific deliberations within the ICES system and two meetings overseeing such deliberations. 4 The PKFM project also allowed us to carry out 29 formal interviews with fisheries scientists or close observers of the fisheries science process. The Scientific Advice for Fisheries Management at Multiple Scales (SAFMAMS) project ran from 2005 through 2008. In addition to allowing the observation of ten more ICES meetings 5 and six more in-depth interviews, the SAFMAMS project supported the observation of the March 2005 meeting of the North Sea Commission Fisheries Partnership and the May 2005 meeting of the North Sea RAC Spatial Planning Working Group. In addition, the SAFMAMS project supported nine workshops at various geographical levels on developing scientific advice for fisheries management, the results of which have often contributed to this case study. In addition, a large number of documents were reviewed. They included ICES and STECF reports, the Memorandums of Understanding between ICES and DG MARE, and a number of internal ICES documents, mainly those distributed in conjunction with the meetings being observed. In the end the access to both meetings and documents that ICES allowed for this research was extensive, and I am very grateful for that. It is ICES’s stated policy that quotations from any of their expert group documents must be cleared ahead of time, and this certainly applies to some of the even less 1. Introduction 31 public documents that I was given access to. Therefore, the ICES Director General has reviewed a draft of this manuscript. However, he did not request any changes. The quantitative methodology used in this case was a random sample survey of European marine fisheries scientists employed in the countries around the North Sea, namely Denmark, Norway, Sweden, Belgium, France, Germany, the Netherlands, Ireland, United Kingdom and the Faroe Islands. A total of 465 valid responses were received. The sample size was 900, which indicates a response rate of 51.7% –a relatively high response rate for a non-telephone survey. The survey procedures and methodologies are described in Appendix 1. 1.5 Plan of the book This book has two basic parts. The first third, Chapters 1 through 3, is mainly theoretical and is offered to provide background on the sociological concepts used in the case study analysis. The present chapter provides a general theoretical and methodological orientation. The second and third chapters cover most of the relevant theory. They are basically selective reviews of studies and theoretical insights drawn from a wide number of areas in which science and policy-making are joined. Chapter 2 begins by reviewing three general challenges we face in the relationship between science and the rest of society. The first challenge is the ‘inflation’of the science boundary that results from constant institutional pressure to define more and more issues as being capable of resolution using scientific methods. The second challenge is the uneasy relationship between the scientific culture and that of decision-makers, which becomes particularly sharp when scientific ideas about what constitutes evidence clashes with legal ideas about evidence. The last challenge is the uncertainty that not only pervades our knowledge of the environment but is also a general condition of modern society with far-reaching consequences. The chapter concludes with a review of three perspectives on the changing relationship between science and the wider society that have proven useful in research on fisheries science. Chapter 3 focuses on the narrower question of general experience with the production of science for policy advice. A large number of studies of how and when science is effective in aiding policy development have been carried out in many different arenas over the last two decades, especially in respect to environmental protection. Out of these studies have come three qualities of science that tend to improve its effectiveness for policy. These are credibility, legitimacy and saliency. Most of Chapter 3 is spent exploring these three qualities. The chapter then turns to a discussion of the boundary between science and non-science and ways that have been found to work across that boundary. This is followed by an explanation of the idea 32 The Paradoxes of Transparency of the paradoxes of transparency. Chapter 3 concludes with a summary of the main theoretical ideas. The remainder of the book, Chapters 4 through 7, presents the case study. Chapter 4 orients the reader to the institutions and issues of scientific advice for fisheries management in Europe, particularly the role played by the International Council for the Exploration of the Sea (ICES), which is the central institution in the case study. The demands for scientific advice that various clients make on ICES, these clients being the European Commission and various other multilateral marine management bodies, are one of two major forces determining the content of the knowledge base for European marine management. The other major force is the interests and plans of the European marine scientists themselves, and these interests and plans are also played out within ICES to an important degree. These two statements, however, are true for European-level activities. The same two forces operate at national levels through marine science laboratories owned by national governments, which I will refer to collectively as National Fisheries Institutes (NFIs). The two levels are deeply intertwined. Most funding for ongoing, day-to-day ‘turning the crank’on fish stock assessment comes from the NFIs and their respective member state ministries. Much of this money is actually expended through the ICES system. The NFIs relate to each other through ICES, and the directors of these NFIs are the single most influential group in the governance of ICES. They control the bulk of the funds used for ICES activities. Most forward-looking research funding for new approaches comes either directly from the EU or from matching grants made by member state governments to European-level research projects. Chapter 5 focuses on the experiences of the individual scientists who work within the advice production process. It is Chapter 5 that draws the most heavily on a formal attitude survey of fisheries scientists in northern Europe. 6 Much of the information reported in Chapter 5 is based on comparisons of the survey responses of fisheries scientists involved in the advisory process with those who are less involved. A central finding is that these scientists find the advisory production process frustrating and demoralising because it both places tremendous demands on their professional lives and uses their work in ways that do not meet their expectations of what science should be about. Chapter 6 returns in a very direct way to the question of adaptation that is so close to my heart. One of the central challenges that ICES is facing is a demand stemming from multiple clients to produce advice for an ecosystem approach to fisheries management (EAFM). It would be difficult to pull together good scientific information for an EAFM even if it was clear what such information should consist of, but the practical application of the EAFM is unclear from the perspective of both the knowledge base and implementation. ICES has moved ahead as best it can under these circumstances, and the institutional issues they are confronting in the process 1. Introduction 33 provide valuable lessons about how you organise an adaptive learning process. Chapter 7 completes the case study with the story of how ICES has reorganised itself to meet the complex demands of providing scientific advice for the EAFM as well as a set of other demands related to the practical implementation of fisheries management. A central rubric of this reorganisation has been an integration of diverse kinds of knowledge that has presented both technical and institutional challenges. Tracing this reorganisation process allows the case study to examine in depth the political dimensions of creating an effective knowledge base for policy. Chapter 8 draws together a set of conclusions about what lesson this case might provide, in light of the theoretical issues outlined in the first three chapters, for how we organise ourselves to develop a knowledge base for environmental management. This book was a long time coming and seeks to bring together a great deal of sometimes disparate research. The inadequacies of the attempt will no doubt become clear to the reader. If the book is able to make a contribution, most of the credit by far will belong to the fisheries scientists whose activities and insights I am recording and presenting. 34 The Paradoxes of Transparency 2. Some general theoretical guides for understanding the role of science in society 2.1 Three challenges in science and society 2.1.1 Science and culture: Pressures to inflate the science boundary Culture and science are bound together in ways we are often barely conscious of. While the more radical reflections on science and society may not have much practical use, they remind us of the weight of the cultural baggage on the science we are trying to harness to practical ends. To me, the most illuminating of these radical reflections comes from the Frankfurt School. This group of thinkers began to blend sociology and philosophy in the 1930s and 1940s to explain the rise of fascism. Communicative Systems Theory (CST) arose from this tradition a generation later. For them, science is the ultimate expression of ‘instrumental rationality’, i.e. rationality focussed on achieving ends. Technical considerations about how to do things are pushing out moral and practical considerations about why we do things. The why question has become a private matter; our collective responsibility is to provide each other with the services and tools needed to reach our individual goals. Ecological degradation is one place where we are being forced to bring back the collective why, so it is perhaps not a co-incidence that questions about science and participation arise in response to environmental problems. Science is powerful rhetoric. Technical arguments are more convincing than soft appeals to values because we have so fully internalised the difference between negotiation and demonstration, while often missing the practical and rhetorical links between them. Pelletier et al. (2000) did a study using before and after attitude measurements of agriculture stakeholders involved in a participatory action conference. Shifts in attitude as a result of the discussion were much stronger in response to technical arguments than value-based arguments. The presentation of something as ‘science’can be and is used to silence people’s concerns (Beck 1992; Irwin 1995). Kaminstein (1996), for example, analyzed responses to a public 35 meeting in which an agency was presenting information about a toxic waste site in the United States. He found that the technical and bland language calmed the audience. When people did express emotions and worries, they seemed very out of place. The community people reported feelings of frustration afterwards. They experienced the officials trying to be friendly and positive as mockery, and compliments such as ‘that is a good question’made them feel less able to complain, dispute and disagree over what was in the end a political rather than a technical question. The concept of the science boundary, the boundary between what is and is not scientific knowledge and who is and is not a scientist, is an important concept in STS (Gieryn 1983). People make use of this rhetorical power of science. They try to present their values and interests as technical requirements, undermining the credibility of science in general when they do so. This phenomenon, then, might be called ‘inflating’the science boundary. Habermas (1984) warns against a tendency, rooted in a desire for control, to try to redefine culture phenomena into technical ones. When this happens, social relationships are made to appear as natural and inevitable, rather than as the concrete results of real decisions made by real people who could have chosen another route. Herbert Marcuse (1964) points out that instrumental rationality is built into the very heart of natural science; because the methods of hypothesis and experiment consist of prediction and manipulation, and hence the domination of nature, science is inescapably linked to control. Suggesting that something is the appropriate object of a scientific analysis casts that object into a particular and subordinate cultural role. This casting can be and is done inappropriately to the detriment of both science and policy. One point CST makes is that bureaucracies and markets have particularly strong institutional imperatives to make the subjective appear objective, often by getting things stamped as ‘scientific’facts. The argument is that some kinds of institutional coordination require ‘empirically motivated ties’that are based on objective facts, rather than ‘rationally motivated trust’based on social relationships (Habermas 1987). The empirical ties are needed by institutions that rely on pressure to create compliant behaviour, rather than on convincing people to behave a certain way (this distinction is discussed in more detail in Section 3.3.4). Coercive coordination is needed for institutions operating on large scales because it makes behaviour predictable. Markets use market pressures that take final form as take-it-or-leave-it offers, while bureaucracies use legal authority backed up by sanctions. Neither requires the rich and nuanced communication about choices that is found in institutions that coordinate action by convincing people that something is the case. Indeed, that would be impossible because rich communication on large scales would be too costly. To make the pressure work, such institutions have to use decision rules that make reference to something ‘objective’so such decision rules require material facts. Social psychologists have long recognised the different behavioural implications of ‘facts’versus ‘social attitudes’, the difference 36 The Paradoxes of Transparency being whether or not it is possible in principle to check some objective reality (Festinger et al. 1950). This need is clear in respect to bureaucracies that have to meet objective legal standards in decision-making, though –as discussed in the next section –‘objective’when it comes to legal standards means something subtly but importantly different from the scientific idea of ‘objective’. The functional differences between facts and attitudes are also important for how markets function, although this is much less studied because it is so far outside what can be accommodated in standard economic theory. It was shown in one study of an African labour market that natural risks affected labour prices exactly as a market model would predict, i.e. people paid to reduce risk. Risk arising from social relationships, however, such as increased danger of being cheated when working with people who are not kin or not from the same ethnic group, did the opposite of what the market model would predict. People paid more money while taking on greater risks. This was found in data for kinship and, more strongly, ethnicity (Wilson 1998). The bottom line is that institutions trying to coordinate behaviour across large scales have to be able to make operational references to ‘indisputable facts’, while having very limited communicative resources for convincing people that these facts are true. The more questions that can be defined as issues of fact with objectively true answers, the easier it is for such institutions to function because they are able to bring more contingencies under their direct control. Bureaucrats are continually trying to expand the arena of questions that can be answered by ‘scientific facts’into areas governed by moral and practical rather than technical rationality. Scientists, however, have their own reasons to resist these trends. Science depends on ‘rationally motivated trust’to coordinate behaviour (Habermas 1987, pp. 182-184). Scientists must convince other scientists that something is true and make use of rich and complex communications to do so. This requires focussing on questions where a rationally motivated factual truth (Habermas 1987, p. 184) can be reached through processes of consensus, characterised by disinterest and scepticism and oriented around universal criteria such as precise definitions, the falsification of hypotheses and replicability (Merton 1968b). Only this narrowness of focus makes possible the internal trust of both people and results that enables scientific inference (Barnes et al. 1996) because it defines the criteria and extent of such trust. Therefore, scientists resist external pressures to change the subject matter and the operating modes of science. An extended example of both bureaucratic pressure to inflate the science boundary and the fisheries scientists’attempts to accommodate this where they could and resist it when it went too far is the discussion of mixed-fishery management in Section 6.1. The temptation to try to change political, social or cultural phenomena into technical ones is always present. Environmental management requires us to address social behaviour, and so we search for techniques to 37 do so. Management involves manipulation, and it is out of this tension that the question of governance arises when democratic societies seek to address social and environmental problems. This problem is exacerbated when actors seek to obscure rather than clarify the distinction between technical and cultural phenomena. The usual motivation for this is to make a policy choice appear as a technical necessity. As one scientist involved in environmental management put it ‘If ... a manager suggests that a decision is based solely on scientifically-derived biological considerations, the manager either misunderstands the nature of science ... or is deliberately trying to disguise ... a value judgement’(Decker et al. 1991, quoted in Minnia and McPeake 2001). 2.1.2 Clashing cultures and notions of evidence Science for legal and policy uses places demands on science and scientists that challenge their modes of operation. Policy processes use different standards of evidence and burdens of proof than science. Scientists want to see that there is a small probability of a null hypothesis, while policymakers are more concerned with the costs of being wrong (Kinzig et al. 2003). As an exaggerated but illustrative example, a policymaker would hardly want to reject a potential cure for cancer because there was only an 89% chance that it would work. Managers employ scientists, but they see things through different cultural glasses. Cullen (1990) observed interactions between water use planners and limnologists and found many differences in expectations and mutual perception. The managers viewed limnology as a curative practice while scientists saw it as preventative. Planners saw scientists as unable to agree on a conclusion and driven by a need to publish rather than getting a planning process done. They also thought the scientists were poor communicators. The scientists felt isolated and underutilised, they saw the planners as unable to interpret data or even find information, indeed as so ignorant that they did not know what they did not know. They also thought the planners were poor communicators. Cohen et al. (2001) did 55 semistructured interviews with scientists working in eight public-sector research science institutions in the UK. A majority of their respondents objected to the idea of being accountable to politicians and managers whose purposes they saw as being at odds with science. Salter (1988) explores in some detail what happens to scientific practice when it is mandated as part of a policy process. The burden is placed on the scientist to produce work that is sufficiently credible, salient and legitimate to support the policy. Policymakers want science that is intelligible to non-scientific audiences, and in doing so represents a clear body of evidence and appears to be rational. The ideal it is meant to project is that of something free of value judgements, using clear methods that produce credible results. At best, it is characterised by open debate, anonymous 38 The Paradoxes of Transparency peer review, and academic publication. In addition to projecting this sterling public image, it must present the policymakers with clear policy choices. Of course, as Salter points out, real science in support of policy conforms to none of these ideals. It makes moral dilemmas explicit, produces conflicting results that cannot be resolved by further studies, and is often seen as corrupt. The policymakers desire transparent knowledge to facilitate and justify their choices, but the actual transparency just reveals more uncertainty. Where these differences really become apparent is when science is drawn into legal proceedings. As Smith and Wynne (1989) argue, legal institutions have their own ways of defining what counts as fact, and they are not the ways found in science. In court it is the law that decides what the factual question is that the scientist must answer. They further argue that ‘adequate evidence’is fundamentally problematic in courts because of the unremitting scepticism. In Latour’s (1987) terminology, opposing lawyers always push scientific facts back towards the conditions of their production and expose the role that scientists’tacit assumptions, experimental skills, and professional judgements played in the production of knowledge. Within the scientific community these perfectly normal aspects of scientific work are handled by expectations of trust in intellectual integrity and the resulting importance of scientific reputations (Barnes et al. 1996). In court they can be portrayed by opposing counsel as simply unprofessional. 2.1.3 Uncertainty In the 1980s, prominent social theorists began to argue that the West has become a ‘risk society’in which anxiety over uncertainty is the driving force in the development of both the self (Giddens 1991) and institutions (Beck 1992), and that science is the key institution that we look to in order to relieve these anxieties. We now live in a ‘post-modern’society characterised by inescapable uncertainty due to both information overload and the loss of the ability to trust traditional sources of valid knowledge. Science has become an arena where disagreements over how to respond to risk and uncertainty are played out (Irwin 1995). People simultaneously look to science as an institution for answers, while viewing individual scientists and their results with scepticism. The desire for scientific answers is great, while the automatic authority of science is a thing of the past. Funtowicz and Ravetz (1992, p. 251) tell us that in any policy arena where the stakes are high, ‘the political manipulation of uncertainty is now the focus of any relevant epistemology’. As research results presented in this volume strongly illustrate, scientists cast in this central role as arbiters of uncertainty have their own identities as scientists, challenged in ways that have a direct impact on their morale. 39 that the quality control is shifted to agents and mechanisms outside of scientific organisations (Guggenheim 2006) and even outside the scientific community. Recent scholarship has challenged some of the empirical arguments underpinning the Mode Two approach. One thread suggests that the frequency and importance of trans-disciplinary research have been exaggerated. Universities, industry and the government remain highly differentiated, and there are strong structural reasons why the current basic division of labour between the three will continue (Shinn 2002). Indeed, some of the ground that Mode Two is argued to have gained in respect of disciplinary science may even be lost. One piece of evidence for this is provided by a study of the recent history of science funding in Sweden. A number of political challenges to the value of funding independent basic science arose in the early 1990s. A special fund was set up to support work addressing specific problems of concern to industry and environmental regulation. These funds grew to the point where they were seen as a challenge to the university establishment. The universities were able to reassert a commitment to basic research activity within national funding (Elam and Glimell 2004). Nevertheless, as will be seen in later chapters, many of the trends described by the Mode Two approach are present in fisheries science as it is carried out in Europe. 2.2.2 Epistemic communities Peter Haas (1989) developed the concept of the ‘epistemic community’ within the discipline of international relations as a way to try to explain the successful emergence of the Mediterranean Action Plan, a pollution control regime around the Mediterranean Sea. He took the term from the philosophy of science where it was used to mean scientists who share certain assumptions about what questions are worth asking. He argued that the Mediterranean Agreement came about because of the existence of an ecologically oriented epistemic community, made up mainly of people working in the environmental ministries of the various countries. They shaped their governments’policies, hired people who thought like them and gained international support (Haas 1989). The idea of the epistemic community has proven useful to many observers of science-based international management regimes. Haas (1992a, p. 3) defines an epistemic community as: Q 2.1 A network of professionals with recognised expertise and competence in a particular domain and with authoritative claim to policy-relevant knowledge within that domain or issue-area. Although an epistemic community may consist of professionals from a variety of disciplines and backgrounds they have: 46 The Paradoxes of Transparency (1) a shared set of normative and principled beliefs, which provide a valuebased rationale for the social action of community members; (2) shared causal beliefs, which are derived from their analysis of practices leading or contributing to a central set of problems in their domain and which then serve as the basis for elucidating the multiple linkages between possible policy actions and desired outcomes; (3) shared notions of validity –that is, inter-subjective, internally defined criteria for weighing and validating knowledge in the domain of their expertise; and (4) a common policy enterprise –that is, a set of common practices associated with a set of problems to which their professional competence is directed, presumably out of the conviction that human welfare will be enhanced as a consequence. Epistemic communities are a particular type of policy network characterised by this general agreement. Members of an epistemic community share a strong normative orientation (Haas 1992a). Epistemic communities are an example of the ‘international civil society’where people divide their allegiance between domestic constituencies and international peer groups (Engles et al. 2006). Haas (1992b) argues that the success of the Montreal Protocol on Substances That Deplete the Ozone Layer can be attributed to an epistemic community. The first countries to actively encourage control were those in which both the epistemic community and a tradition of pro-environmental sentiment were strong. Once channels between other countries’national administrations were established and the epistemic community broadened, then the other countries began to support action. His analysis found that it was these contacts that made the real difference, rather than public opinion or the actions of environmental NGOs. These things only became important later in the process after government regulations had been introduced. Interestingly, for Haas (1992b) the key actor was the DuPont Corporation, which broke ranks with other chemical companies in an act that was critically important to the momentum for international action. Haas (1992b) argues that the difference was that key decision-makers on this issue at DuPont were all chemists who modified their positions in reaction to advances in scientific understanding. The epistemic community approach takes a more deferential attitude towards science than is usually found in STS. The effectiveness of the scientific community in international regime formation is rooted in the ability of the scientists to reach a consensus and to overcome their natural inclination to extreme caution. These characteristics can be seen in the formation of international regimes that were science-driven, including the Convention on International Trade in Endangered Species and the International Union for Conservation of Nature and Natural Resources (Young 1989). The key to effectiveness is that knowledge on which the epistemic community is based is ‘accurate, accessible, and contributes to the achieve47 ment of collective goals’. It must ‘represent consensus and be provided through a medium that is politically palatable’(Haas 2004, p. 575). The emphasis on consensus has been the characteristic of the epistemic community approach that has brought the most criticism, especially from scholars trained in STS. Lidskog and Sundqvist (2002) argue that Haas was right about the importance in international relations of a common picture of reality to challenge the idea that national interests alone drive outcomes. However, they accuse him of a naive view of science, almost to the point where he introduces a consensus notion of truth as part of the epistemic community argument. Others question the strong programmatic emphasis on consensus because disagreement is not only going to be found within any scientific community; it is a positive force driving new thinking (Young 2004). Both the strength and the weakness of the epistemic community approach are that it posits an ideal situation: a strong consensus among scientists reflecting truth about nature that has clear implications, and policy alternatives that all of the scientists can gather around. It describes what has actually happened in several successfully negotiated environmental protection regimes. These empirical examples, while limited in number, do show that success is possible and provide a set of experiences of success from which lessons can be drawn. The weakness of the epistemic community approach is also rooted in its focus on an ideal situation. What about the majority of environmental problems where the degree of agreement among scientists is a mixed bag of agreement about some facts (and some values), but not others? Haas’s (2004) response is that we must wait for the consensus, and until that consensus emerges, scientific and policy developments must be kept insulated from one another. This response is inadequate. It is grounded in a naive view of both the way the science boundary works and the degree of urgency we face in tackling many of these issues. Uncertainty is a reality we have to learn to deal with. The epistemic community idea sees cooperation as an either/or proposition (Sebenius 1992) that fails to consider the multiple bottom lines that participants in negotiations face. Ways that policy development can move forward do not emerge all at once. What is most critical is that the people involved in ongoing negotiations are able to move forward when the opportunity arises, in a way that their ability to work together in the future is enhanced rather than impaired. Such a community may become a true epistemic community when uncertainty is reduced and the way forward becomes clear, but it must continue to learn and adapt in small steps the rest of the time. 2.2.3 Post-normal science: New forms of scientific practice Silvio Funtowicz and Jerome Ravetz developed the idea of Post-Normal Science (PNS). PNS happens in policy areas characterised by both high 48 The Paradoxes of Transparency uncertainty and high stakes. Under these conditions it is very difficult to keep facts separate from the interests and value positions, especially as they are usually expressed as probabilities. Where the stakes are high, interests and values determine how participants perceive the associated risks. Their basic argument is that such conditions require a more participatory decision-making process. A central concept in PNS is the ‘extended peer community’. To deal with new problems in a high uncertainty/high stakes area, an open dialogue is required because the quality of the science depends on an ‘extended peer review’. The important thing to keep in mind to understand PNS is this link to quality. The idea of the extended peer community is close to, but not synonymous with, stakeholder involvement in science. The extended peer community is about the science itself, it is a new kind of quality control. It is made up of experts, even if some of these people base their expertise on experience-based knowledge, say of a policy process or a fishery, rather than research-based knowledge. The ideas of PNS begin with a set of philosophical developments about the nature of quantification. Their arguments are mathematical ones broadly understood and do not begin with empirical observations the way that both the epistemic community approach and the Mode Two science approach do. It is from this philosophical perspective that they developed the NUSAP notation for quantitative statements. In the NUSAP notation, in addition to the familiar categories of number, unit and spread, Funtowicz and Ravetz (1990) introduce the new categories of assessment and pedigree. Assessment and pedigree are both about the characterisation of uncertainty. Assessment relates to uncertainty rising from problems in the reliability of a quantity. It could be expressed, for example, in arguments about what level of confidence should be placed around a statement. Should it be 95%, 99% or some other level? Assessment is about the different kinds of judgements that one would see expressed by statements such as ‘conservative by a factor of 10’. As Funtowicz and Ravetz (1990, p. 28) describe it, ‘our knowledge of the behaviour of the data gives us a spread, and our knowledge of the process gives us an assessment’. Pedigree relates to uncertainty rising from the border with ignorance, discussed above (Section 2.1.3), in the context of their categorisation of types of uncertainty. Here Funtowicz and Ravetz (1990) step entirely into the qualitative aspects surrounding the quantity. Pedigree addresses the relevant epistemological, historical, sociological and institutional contexts needed if one is to understand the implications of a quantity for policy. They develop ‘pedigree matrices’based on a hierarchy of modes of knowledge: a deductive argument is stronger than an inductive inference which, in turn, is stronger than an analogical argument. All three of them are considered stronger than conventional definitions. They offer a number of examples of pedigree matrices. A very basic example is this pedigree table for research information (Table 2.1). 49 Table 2.1 Research pedigree matrix Theoretical structure Data input Peer acceptance Colleague consensus Established theory Experimental data Total All but cranks Theoretically based model Historic/field data High All but rebels Computational model Calculated data Medium Competing schools Statistical processing Educated guesses Low Embryonic field Definitions Uneducated guesses None No opinion Source: Funtowicz and Ravetz (1990, p. 140) The table establishes a hierarchy which allows the receiver of the quantity communicated with the NUSAP notation to evaluate basic levels of certainty. The NUSAP notation and the pedigree table help clarify the idea that it is in the extended peer community that the relevant forms of quality control for the quantity arise, in contexts of high stakes and high uncertainty. The extended peer community is made up of the various groups that can contribute their perspectives on the policy and their own knowledge. The extended peer community consists of the people who have the knowledge to fill out the pedigree table, knowledge that is found in both the scientific disciplines themselves and in the sociology of knowledge of the policy arena. Ravetz (1999) argues that effective science-based policies in arenas of high stakes and high uncertainty require an open dialogue with all those affected. The extended peer community improves quality by mobilising ‘extended facts’to help develop a shared understanding of the uncertainty in areas of conflicting values and agendas (Healy 1999). This extended peer community is not meant to reduce the authority of science or to make it explicitly political. It addresses the problem that in these high-stakes, high-uncertainty areas, the traditional mechanisms for assuring quality are not adequate. Accredited experts need the assistance of the extended community to get the necessary job done. Establishing the pedigree of a quantity requires a broad understanding of its saliency, credibility and legitimacy that only a broader group can provide. Ravetz (1999) expresses concern that seen out of context, these ideas may appear to reduce the authority of science; he stresses, however, that PNS is not about traditional areas of research, but new areas with high social and economic importance where traditional mechanisms for assuring quality are not adequate. PNS is not meant to be an attack on accredited experts, but a way of describing the kind of assistance needed (Ravetz 1999). The way the relationship between scientific experts and other participants should be structured is one of the most helpful insights of PNS. Funtowicz and Ravetz (1990) often make use of the contract between ‘knowing-that’vs ‘knowing-how’when discussing scientific quality control in contexts of high uncertainty. Traditional science has seen itself basically as the first, but PNS requires a new emphasis on the second. There are 50 The Paradoxes of Transparency several dangers in trying to maintain traditional styles of practice in uncertain conditions. Scientists over-sell their science, feeling that they have to show more confidence and authority than the situation warrants. This leaves them, in turn, facing exactly what they were trying to avoid: a continuous decline in respect for expert claims (Irwin 1995). Science policy settings in Europe and Canada often involve ongoing discussions with interest groups, and these discussions help guard against attempts to justify value-based choices with post-hoc scientific arguments (Jasanoff 1986). Knowing-that is about the ultimate attainment of truth, while knowinghow is about practice. Knowing-how is about using skill, it is rooted in tacit knowledge and not part of the traditional philosophy of science. In resource management scientists playing this kind of role can be seen in certification programmes where management programmes are evaluated according to a complex set of criteria. Issues of uncertainty become areas of ongoing negotiations between the scientists and those who desire certification. Within a high-stakes, high-uncertainty context, scientific skills provide ‘rubrics, guidelines and elicitation procedures, for the expression of uncertainty, for the assessment of quality, and also for the training in both skills’ (Funtowicz and Ravetz 1999, p. 68). Scientists working interactively with others can act as facilitators of transparency, while simultaneously taking advantage of different kinds of knowledge to try to describe and deal with the implications of uncertainty and ignorance. Scientists are currently not trained to be consultants, but it is the skills of the consultant that are required here. These are the skills to work with policymakers and other stakeholders in a process linking the uncertainty and quality of the information with the needs of the policy. They point out that where experts in consulting professions normally have very long, practical, apprentice-type training after their formal educations (e.g. doctors), scientists generally do one major research project under supervision and then are certified as able to operate as an independent scientist. They argue that the ideas of skill and craftsmanship can be the basis of a way to reformulate the science boundary in areas of high uncertainty (Funtowicz and Ravetz 1990). The shift is from trusting science as a ‘truth machine’to trusting the scientific institutions and procedures to make science that works in practical situations (Healy 1999). Again, the strengths and weaknesses of the approach are closely related. Because it begins from an essentially philosophical perspective, the NUSAP notations provide a clear rationale, grounding the science boundary in epistemological rather than practical considerations. At the same time, the emphasis on scientific skills and the consultancy model suggests an appealing programme for implementing science in areas where boundary work can become very difficult. The idea of the extended peer community, however, needs sociological examination. The extended communities involved in science-based policies include groups with conflicting objectives who bring their own sets of relevant facts, values and interests into the 51 discussion. Legitimacy can be lost if scientists are seen as closer to one interest group than another. Given the level of social conflict involved, what are the conditions under which an extended peer community adds to the quality of the science, rather than just adding some chaos to the complexity and uncertainty? All three of these approaches to understanding the changing institutional context of science are helpful in understanding the case study presented here. While each of them can effectively supplement a human ecological approach, none of them addresses the question of adaptation directly. Nor do any of them even begin to address the complexity of understanding scientific institutions that seek to take an ecosystem approach to fisheries management. In seeking to support European fisheries management policy, the scientists in the ICES system both reflect and speak back to these three tendencies. Because they are doing so within a formal policy context, they share a set of more specific demands on their science with other scientists who work with policymakers. It is this exchange between science and policy that Chapter 3 addresses. 52 The Paradoxes of Transparency 3. Developing scientific advice for policy 3.1 Saliency, credibility and legitimacy Over the last 25 years or so, many social and natural scientists have begun to ask, some more and some less systematically, under what circumstances scientific findings are able to influence policy. Most of these investigations have been motivated by frustrations over the seemingly slow response of policymakers to environmental issues. One of the most useful efforts to examine the use of science in policy is the Global Environmental Assessment Project, a large comparative research project carried out by the Kennedy School on the policy uptake of results from global scientific assessments (Cash and Clark 2001; Cash et al. 2002; Clark et al. 2002, 2006). As mentioned in Chapter 1, they conclude that the majority of the assessments are not effective in influencing policy, but they do affect the long-term development of an issue through such mechanisms as influencing the issue’s visibility, the stakeholders who will take an interest, the way questions and objectives are framed, and the selection of management alternatives (see also van der Hove 2007). Hence, an approach to understanding international policy that examines the creation of common understanding within a political process is more instructive than simply examining issues of power and interests (Haas 2004). This influence of science on policy is reflected back on science through the influences of the policy process on scientists and their work. Questions of policy and the needs of policymakers have a strong influence on science. As Young (2004) argues, international institutions influence the topics being studied and the kinds of models being created. Policy influences science directly by determining what scientific questions will receive funding, and less directly through changes in standards of quality control. When policy addresses areas of high uncertainty, and therefore competing interpretations among scientists, changes in the standards that determine the ‘best available knowledge’become extremely important (van der Hove 2007). Debates about the credibility and uncertainty of results forming an important part of the negotiations are a very common feature in diplomacy related to environmental issues (Backstrand 2004). Scientists deciding which directions to take in the creation of new knowledge are strongly influenced by their experiences with what has proved useful 53 and interesting to policymakers in the past. This desire for policy influence can also lead to downplaying disagreements in the search for consensus, in spite of the fact that ‘within reason, disagreement is a positive force in the scientific world’(Young 2004, p. 223). The Global Environmental Assessment Project made use of three analytical concepts that proved very useful in understanding the uptake of science to policy. Clark et al. (2002, p. 7) define these three as follows: Q 3.1 ‘Saliency’reflects whether an actor perceives the assessment to be addressing questions relevant to their policy or behavioural choices; ‘Credibility’reflects whether an actor perceives the assessment’s arguments to meet standards of scientific plausibility and technical adequacy; and ‘Legitimacy’reflects whether an actor perceives the assessment as unbiased and meeting standards of political fairness. A scientific result needs all three of these attributes to some degree if it is to influence policy. An important difficulty, Clark et al. (2002) argue, is that there are often trade-offs between the three. Indeed, they suggest that efforts to bolster one usually only succeed at the expense of another. For example, efforts to increase saliency by narrowing the questions to be investigated can decrease legitimacy by making the process appear to have a political bias. Conversely, adding stakeholders to increase legitimacy can reduce saliency as issues are raised that are outside of the policymakers’ remit. While these relationships may usually involve trade-offs, they can also lead to mutual reinforcement and complementarity. For example, an effort to increase credibility by including new knowledge may also increase saliency and legitimacy. These observations form the basis of their empirical investigations of science policy institutions, suggesting that the main differences lie in the ways that they shape and balance the trade-offs among saliency, credibility and legitimacy. It is important to note that these or similar categories also emerge in various ways in studies of the perception of scientific information in local contexts. It is these local contexts, rather than abstract ideas about science, that provide the tools that people use to make sense of the information (Irwin 1995). Michael (1996), for example, conducted a series of interviews with the general public about radon. They were interested in how people understood the categories of ‘ignorance’and ‘expertise’. They found that three kinds of responses were the most common and these three corresponded roughly to credibility, legitimacy and saliency. One was the basic, unsurprising idea that some people know more than others and the experts should be listened to. The second was the notion that understanding radon is ‘not my job’, suggesting an understanding of ignorance and expertise that was based on a legitimate division of labour. The third main response was in terms of saliency, the information about radon was simply not interesting to them. Yearley (1999) held a series of focussed group interviews with different stakeholder groups around air pollution models 54 The Paradoxes of Transparency being used by a local government agency. They found that three factors influenced the public understanding of the models’output. The first two corresponded to legitimacy and credibility. The first was the respondents’ assessment of the trustworthiness and agenda of the agency, and the second was their confidence in their own technical knowledge and ability to judge the technical merits of the model. The third was also related to credibility and reflected the importance that people attach to direct experience: they judged the models in terms of their evaluation of the assumptions about social behaviour that underlay them. All three of these factors were more important than the face credibility of the models based on realisticlooking simulations and projections. 3.1.1 What is saliency? Saliency is very similar to ‘relevance’, in fact ‘relevance’appears in the definition, and my fisheries science colleagues have from time to time wanted to know why social scientists insist on using this weird word instead of just saying ‘relevance’. The difference is that many scientific findings are relevant to policy in the sense that they could logically be considered in making the policy without actually being salient. By using the term ‘saliency’we are emphasising that particular facts become prominent because of their usefulness in responding to the needs of policy development. In practice, policy-making must select from among a broad range of relevant facts to steer activities. The competition among stakeholders in environmental management in developing the knowledge base for management consists of each group trying to move a set of facts from being merely relevant to being salient. This competition happens within management systems that already presuppose a great deal about what facts are salient. Considerable aggravation is experienced by conservation interests, user groups and scientists alike when the facts they see as most relevant are not salient in the decision-making process. Many feel that there ought to be an objective, ‘scientific’way to determine which facts should be salient. We expend a lot of energy in identifying the right ‘indicators’and ‘drivers’and experience a great deal of frustration when these lists quickly become very long. While science can perhaps eliminate some candidates for saliency in an objective manner, the movement of facts from relevance to saliency is unavoidably a political process. 3.1.2 What is credibility? Credibility is about making sure that the scientific result reflects nature as closely as possible. Credibility comes from applying the scientific method, i.e. the testing of a falsifiable hypothesis, along with that method’s guar55 are the only source of sovereignty. Policies are legitimate when policymakers are representative, accountable and placed under public scrutiny… Outcome legitimacy: This vision of legitimacy focuses on the policy eventually made, and not on the process through which it was made. In this case, what makes a policy legitimate is its capacity to solve problems requiring collective solutions, and solve them in a way benefiting the ‘public interest’(2003, p. 76). The distinction between process and outcome legitimacy means that there are two basic ways that people rationally evaluate whether a decision-making process is legitimate. The first is based on the process by which the decision gets made; the second is based on the characteristics of the decision itself. Even when science is evaluated in terms of process legitimacy, the answer to the question in the title of this section is yes, but the scientific legitimacy is very closely tied to the question of credibility. The beholder is in fact evaluating aspects of the process that are directly linked to the question of credibility, and the response to challenges to procedural legitimacy is made by defending the credibility of the science. For process legitimacy to say that a scientific process is illegitimate is tantamount to saying that the methodology followed was unscientific. This is not quite as straightforward as it might seem, though. What I will try to convince you of in this section is that the rational source of process legitimacy for any decision is based on the same underlying principles as the scientific method. The answer to the question is yes because the questions of process legitimacy include but are wider than simply the scientific method –they also extend to surrounding social processes. Is it possible to have a concept of the legitimacy of a process that is not just a set of opinions about what a process should be like? Can there be process legitimacy that is not in the eye of the beholder? Meunier (2003) says that a legitimate process should be ‘representative, accountable and placed under public scrutiny’. That sounds fine. But does it have any basis beyond ‘this is the way we like things done in Europe’? I think there is such a concept, which can be arrived at by an argument I find logical and compelling and, more importantly, which links up directly to our general understanding of what makes science credible. Again I am referring to the concept of communicative rationality that is at the centre of Communicative Systems Theory. To repeat the basic idea, communicative rationality is the logic that we use when we talk to each other to make sense of what we are saying. It can be thought of in a pure sense that would be useful for looking at philosophical arguments, but we also use it in day-to-day speech to understand normal situations. At the social level, it depends on two general rules: that there is no manipulation involved in the communication –meaning that no one is prevented in principle from participating or is forced or paid or tricked into making 62 The Paradoxes of Transparency statements that do not reflect their true beliefs –and that everything communicated is open to question about its validity (White 1988, p. 56). Now this ideal is not meant to describe a real situation. Habermas himself uses the term ‘partly counterfactual’(Habermas and Nielsen 1990, p. 105) meaning that while these ideals exist only in people’s heads, they are still a very real, concrete aspect of a situation. This is because people are always using them to evaluate the communications they are engaged in, and they have to believe that the standards are met, at least to some degree, if they are going to trust the communications enough to be able to build a shared reality. For example, your boss can tell you to be quiet because he says so, but that does not help you and your boss come to a mutual understanding about what your situation is. In fact, it is irrational because it keeps you from reaching a better understanding. We see this ideal operating every day. No one takes statements made in marketing or political contexts very seriously because we know that in these arenas communications are highly manipulated and rarely have to respond to serious questioning. Planning meetings with colleagues –at my workplace and I hope at yours –are taken very seriously, people really have to explain why they see things the way they do, and serious violations of the ideal standards would be seen as major betrayal. Of course, these standards play a role in broader discourses as well. The rules against manipulation are the basis of accusations that scientists have been corrupted by their funding sources. As discussed in the next section, these are perhaps the most common challenges to scientific legitimacy. To a degree, this is all a fancy and detailed way of saying that people have to be fair and listen carefully to one another if they are going to understand each other. It is pretty much common sense. It has to be very common sense because we all have to live together and coordinate our actions all the time, so we all have to make use of these standards in day-to-day speech. But underneath this common simplicity Habermas (1990) has found an elegant result, one that yields a concept of process legitimacy that is not just an opinion. These standards of ideal speech are not the basis of process legitimacy because of their functional usefulness. We cannot assume that a process is legitimate just because it works well to achieve some end. This would then raise the question of the outcome and goals of the process, and these outcomes and goals cannot turn around and determine the legitimacy of the process that created them. Legitimacy is a normative concept, an evaluation of legitimacy requires showing objectively that the process is as it ought to be. You cannot settle an argument about what ought to be by pointing at evidence. However, you can settle it if you show an internal contradiction in the opposing argument. You can prove X by demonstrating that any argument that objects to X already assumes X (Habermas 1990). The ideal speech standards achieve this. In fact, they are very likely the only thing that does. They are inescapable presuppositions of any argument – just by raising an argument you are assuming that they are in force be63 cause otherwise there is no point in making the argument. In other words, you cannot logically raise an argument against the basic rules that make raising arguments meaningful in the first place. Therefore, these commonsense rules of how we make sense to one another also point to context-independent criteria that can always be used to evaluate the rational legitimacy of communicative processes, including decision-making processes. The concepts that we commonly use when talking about process legitimacy, where they are valid, are all ways of trying to uphold the standards of ideal speech. The first three principles of good governance articulated in the European White Paper (CEC 2001a), participation, openness, and accountability, are all rooted in the demands of ideal communication. Meunier’s (2003) ‘representative, accountable and placed under public scrutiny’criteria are as well. ‘Participation’is a restatement of the requirement that no person or argument be excluded, while ‘representativeness’is about trying to find a fair way to do this when the scale of the discussion does not allow everyone to speak. ‘Accountability’is the demand that people account for themselves, that they explain what they think and why they think it. ‘Public scrutiny’,‘openness’or ‘transparency’is in many ways the critical technique for guarding the standards of ideal speech. Transparency makes accountability possible and is the main safeguard against manipulation. What science does that is special, and this is I think the essence of the scientific method, is that it takes these general rules for rationally legitimate decision-making and turns them into a rigorous procedure. The ideal of the scientific method can only be approached, not achieved in practice, in the same way that the standards of ideal speech are never completely met. The scientific method rejects dogma, in principle, no claim is excluded on a priori grounds, and all claims must pass the same universal set of procedures to be verified. In practice, of course, the majority of claims are ‘black boxed’in the sense that they are assumed to have been already established, and notions of plausibility in the tradition of the scientific community play the role of gatekeeper in deciding what claims will be examined. This process is carried out, however, in a way that recognises the application of the same set of procedures to all claims. The scientific method takes accountability and transparency to their logical extreme. The statement of a falsifiable hypothesis sets up the clearest accountability one can imagine. Quantification and experimental replication require the scientists to explain how they know something with such precision that someone else can go out and repeat the experience. This makes quantification and replication the ultimate ideal of procedural transparency. The precision of language that mathematics brings to descriptions of natural phenomena provides the most transparent accounting possible. Ultimately, this radical commitment to the standards of ideal speech is why we turn to science when we want to resolve questions about nature. Science is cast in the role of the arbiter of truth because of the 64 The Paradoxes of Transparency transparency of its philosophically clear, if practically obscure, ability to say how it knows what it knows. The irony is that these procedures for radical accountability and transparency, and the tools of precision they have engendered, are what has made science so opaque to the non-specialist. The question behind the eyes of the beholder under the rational category of legitimation of science, when the beholder focuses his or her judgement on procedures, is the question of scientific credibility. This does not mean, however, that scientific process legitimacy is the same thing as scientific credibility. Scientific process legitimacy places the question of scientific credibility in a wider context; it examines the entire gamut of standards of ideal speech when asking the question of scientific credibility. Process legitimacy demands freedom from manipulation. When a drug company pays for research on the effectiveness of its own drugs, a question of procedural legitimacy is raised –were the results manipulated in order to maintain funding? The response to these questions of manipulation, however, is to seek to demonstrate in a transparent way that established scientific methodologies were adhered to. Process legitimacy demands freedom from the suppression of relevant claims. Science that is selective about its evidence can be very damaging, especially when the results have an influence on policy. A very famous and far-reaching example in fisheries management is the ‘tragedy of the commons’. This is a theory made popular by Hardin (1968) that has had a tremendous influence on natural resource policies. Hardin took a basic result in resource economics showing that when a natural resource is managed under open access –i.e. when anyone can use it –it will be overused to the point where the economic benefits will be wasted and the resource may even be destroyed. He gave this result the catchy title ‘tragedy of the commons’. What Hardin and other adherents to this theory did, however, was to conflate ‘open access’with ‘commons’and conclude that only private property or government control can avoid the tragedy. This ‘tragedy of the commons’parable became common sense in natural resource management for decades, ignoring mountains of evidence that property owned in common by small groups was very often not open access, but rather managed in ways that avoided resource degradation. The real tragedy that emerged from all of this was decades of ineffective topdown management that assumed that when it is not feasible to establish private property rights, only government agencies could manage resources. Finally, outcome legitimacy also points to a kind of scientific legitimacy, although it has little to do with issues of the common good and everything to do with the potential policy impact of the new information. Policy debates are not made up of discussions naming long lists of independent facts. Instead, they are made up of discursive themes. A discursive theme is a way of linking facts, values and interests together into a story line that makes sense. It is these linkages that compete in the discussion. The story lines do change, but only slowly. When a new fact is presented that fits into 65 an existing story line without problem, it is quickly taken up –i.e. it is treated as legitimate. When a new fact does not fit easily into a story line, it is resisted. Such resistance does not necessarily include an attack on the credibility of the process that created it. It can also mean simply ignoring the fact or treating it as unimportant. This kind of rational legitimacy is similar to the saliency of the result, but it is not about how well it fits into a policy process as much as how well it fits into the constellations of facts, values and interests that different stakeholders are bringing to the discussion. 3.1.4 Saliency, legitimacy and research on risk perception A number of studies have been made of the impact of the source of scientific information on its saliency, credibility and legitimacy. Risk studies have shown that people care more about how decisions are made and their fairness than they do about the magnitude of a risk (Chess and Lynn 1996). The perceived saliency of an issue, and related scientific communications, also emerge as important in research on perceptions of environmental risk. Furthermore, credibility and legitimacy are both closely tied to the source of the information. People perceive both themselves and others as having or not having a right to talk about something (Michael 1996). This is not surprising when we consider what might be the most basic observation from the sociology of knowledge: the social location of facts is what determines their effective validity. For example, as Collins and Pinch (1998) point out, you and I base our personal knowledge about nature, e.g. that it is not possible to travel faster than light, entirely on our personal knowledge about society that such information resides with physicists rather than Star Trek scriptwriters. One of the corollaries of this observation, which has been well documented in research, is that in the event people do have personal experience of a phenomenon, they weight that experience much more highly in comparison to information that is merely communicated. Perceptions of risk in general are higher when they are more salient to the perceiver, for example if they know somebody who personally suffered the adverse event (Kolker and Burke 1993). The inverse, however, is not true; higher risks are not necessarily the more salient ones. In fact, people see risks with middlelevel probabilities as the most salient, and high-probability dangers are overlooked because they are more accepted, for example traffic accidents (Douglas 1985). In another example of exaggerated trust in one’s own experience, people overestimate their ability to detect risk themselves. Many non-specialists believe that they should be able to detect risks with their own senses, for example one study of risk perception found that non-scientists thought they could taste pollution (Johnson and Griffith 1996). Judgements about 66 The Paradoxes of Transparency whether scientists and others are aware of risks are not correlated with concern about the seriousness of the risks (Fife-Schaw and Rowe 1996). Whether or not the risk is seen as under one’s own control or as forced on one by another party is also important, not only in perceptions of fairness and legitimacy but in respect to saliency as well. A risk that becomes well-known is generally connected to some issue of legitimacy; therefore, moral concerns are involved not just in the response to risk but in its perception (Douglas 1985). I do not know of any research that directly links these findings about saliency and risk to the question of how stakeholders will perceive the saliency of scientific advice for policy. It is a reasonable hypothesis, however, that advice based on findings linked to experience and previous knowledge of an issue will not only be seen as more credible, they will also be seen as more salient. Legitimacy and source are linked in ways that are filtered through local experience. People dealing directly with the public have completely different ideas than scientists about what makes a communication credible. Chess et al. (1995) interviewed 145 risk communication researchers and practitioners about what scientists and people dealing with policy thought was important in the presentation of risk information. The technical people saw the question of how to express probabilities and communicate evidence to support their assertions as the heart of the problem. The practitioners were much more concerned with how to integrate political and value-based concerns into the policy responses. Legitimacy is tied to the source of the information most often by questions about the interests of the researcher. Scientific findings are often seen as purchased. In fact, the notion that science is an expression of political interests is a form of ‘common sense’among much of the general population; claims of factuality and objectivity, rather than increasing credibility, are often seen as just one more marketing tool (Irwin 1995). McKechnie (1996) in field work in a rural area found that sources are seen as credible only if their self-presentation is consistent with local values and that assertions about having esoteric knowledge actually undercut credibility as the source is seen to lack ‘common sense’. This common sense has a factual basis. It is becoming increasingly common for scientists to have direct financial stakes in outcomes. Comparative, statistical studies of the outcomes of drug trials have found that drug-company funding has a strong influence on results: 89% of company-funded studies showed new drugs to be more effective than older ones compared with 61% of those not company-funded. Moreover, the authorship of articles in major medical journals is routinely hidden; one study found that 29% of articles used such devises as guest authors or ghost authors (Guston 2001b). Frewer et al. (1996) did a series of semi-structured interviews related to the source of information about environmental risks. They found that trusted sources are characterised by multiple positive attributes. Interest67 ingly, while observers of policy process often assign a high weight to ‘independent’sources of information, the people interviewed in this study found that sources which operate under moderate accountability, rather than complete freedom, are trusted more. Television current affairs programmes and quality newspapers are among the most trusted, with the tabloids and government the least trusted. These potential linkages between risk perception, saliency and legitimacy take us close to the central issue of uncertainty. Jasanoff (1986) points out that officially sanctioned risk assessments focus mainly on technical questions, and their guidelines emphasise using multiple sources of uncertainty to make numerical assessments. This in itself has a political dimension as environmentalists feel that risk assessment creates a false impression of certainty. Shackley and Wynne (1996) are concerned that scientists are motivated to demonstrate control over uncertainty, for example by its quantification as risk, because they see uncertainty as a challenge to the authority and use of science in policy-making. The question of legitimacy and source is an important one in the science underlying the Common Fisheries Policy. One question in our survey asked the extent to which scientists in expert groups suspected other scientists of arguing in a way that was consciously biased by that scientist’s national interest (Table 3.1). The overall mean is just below the middle point of 4. Mean answers from those scientists whose last group was directly involved in stock assessment were slightly higher than from the others. Table 3.1 When participating in working or study groups how often have you suspected that a scientist was arguing in a way that was consciously biased by his or her national interests? 1 = never, 7 = very frequently Last group was not a stock assessment group Mean 3.51 N 161 Last group was a stock assessment group Mean 3.89 N 123 Total Mean 3.68 N 284 Relationship is significant at .06. Excluded to reach N of 284: 148 respondents who indicated not having participated in an expert group in the last five years; 20 who did indicate participating but did not make clear which kind of group and therefore did not fit in any of the two categories; six who failed to answer whether they had been in an expert group or not; and seven who did not answer the question on bias. Scientists have also expressed feeling on several occasions that particular countries were withholding contributions of information that they should 68 The Paradoxes of Transparency be making. Perceptions of national perspectives that reflect local fisheries but are not seen as full-blown biases are also found. One Scottish scientist told us that ‘you could never get a Danish scientist to say that industrial fishing is a bad thing’. If other scientists have suspicions of national biases, it is reasonable to suppose that non-scientists have the same suspicions at an even higher level. 3.2 Boundary organisations and objects The science boundary is the line between which claims are or are not science, between which procedures do or do not produce such claims, and between who is and who is not qualified to carry out such procedures. Science in support of policy needs to be careful about where it places the science boundary when trying to increase its credibility, saliency and legitimacy. As we have seen, in situations of high uncertainty, there is a greater need for expanding participation in scientific activities. In addition to the needs of addressing uncertainty, pressures to inflate the science boundary also emerge through policy processes seeking ‘scientific’answers to more and more kinds of questions. In this section I briefly review the research that has been done on placing and maintaining the science boundary. This constant focus on maintaining boundaries around science is critical to the functioning of both science and policy. It allows scientists and policymakers to bring a clear accounting of knowledge, including an accounting of uncertainty, into policy decisions, and this accounting, and the status distinctions that support it, should not be discounted because they cannot be perfect (Collins and Evans 2002; van Zwaneberg and Millstone 2000). The demand for objective standards for decisions, and the conferral of a special status on particular forms of knowledge to protect these standards, is not a denial of democracy; it is a product of it (Porter 1995; Ozawa 1991). One of the most basic insights from science and policy studies is that boundary work must be taken seriously. It cannot be treated naively; neither by assuming that the distinction between what is science and what is policy, advocacy or values is easily made in concrete situations, nor by assuming that it does not really exist. Boundary work as it emerges within policy debates is not an all or nothing affair (Guston 2001b). In more participatory decision-making processes, it is called upon to support specific judgements in circumstances where there are checks and balances on its use. In fisheries management, what might be called the mainstream view of the role of science within the policy process is articulated as follows by DG MARE: Q 3.7 Visibly free of political influence: Scientists in most national administrations are generally placed at a distance from administrative and political 69 pressures by the national fisheries laboratories. Those who are not so distanced quickly lose credibility and influence (CEC 2003a, p. 15). Science must stand off to the side so that the other players in the policy debate have a common set of facts to debate about. Otherwise, following this logic, there is no possibility of a rational conclusion. The most concrete expression of this role is when DG MARE is able to point to scientific conclusions to justify particular management measures. This need is only intensified by the move towards the use of harvest control rules which, in the final analysis, uses the scientific finding as a mechanism to shortcut the policy debate, hence adding predictability and continuity to the polices to everyone’s benefit. On the surface, this mainstream view is an expression of communicative rationality in the way it seeks to separate facts from values and interests. However, as the research described in this section demonstrates, the view laid out in Q 3.7 is an utopian construct with a number of problems. The focus is on the objective assertions made by people assigned the role of being the objective people. In communicative rationality, assertions of fact are unique because they are carried out in discussions that presuppose consensus as a goal, a presupposition that makes transparency about the reasons for the assertion a central concern. Here the scientists’expertise in the methods of transparency is what is critical, not some privileged access to objectivity. The mainstream view expressed in Q 3.7 is not, in fact, communicatively rational at all because it prejudges assertions based on their source. Boundary work is difficult. Scientists, of course, have the primary stake in boundary work. Within STS the early work on ‘boundary maintenance’, particularly Gieryn (1983), has now achieved a seminal and classic status. The basic argument was that within policy deliberations, scientists work to enhance their authority by defining what knowledge is really science, who is really a scientist (Jasanoff 1990; Gieryn 1983), and which questions are to be considered scientific (Dietz et al. 1989). When something is labelled as science, those who are not scientists are de facto barred from having anything to say about its substance (Jasanoff 1990). When this labelling is successful, knowledge becomes a ‘black box’for the non-scientist until another expert challenges it. Constant pressure is seen in policy contexts to make the political appear objective (Ozawa 1991). The constant pressures to inflate the science boundary to treat more and more things as appropriate for a scientific approach was discussed in Section 2.1.1. Seeking to make political decisions seem objective is not necessarily a product of fraud or opportunism –it can arise simply from the nature of the problem to be solved. In science in support of fisheries management, for example, one critical issue is the need for the operational units reflected in the advice to match those of the actual fishing activities that are being managed. Fisheries scientists traditionally worry about units defined by nature. These are mainly single fish 70 The Paradoxes of Transparency stocks, or increasingly biological and ecological interactions between stocks and between the fish and their environment. These are all things that exist independently in nature. They are not, in principle, created by people, and they operate under natural laws. I say ‘in principle’here because attempts to describe and measure these things are created by people, and actions by people have a powerful impact on their condition. Managers, however, manage fisheries which are complexes of fishing ports, fishing boats, and fishing gears that are hybrids of natural, technical and social phenomena, some of which are entirely human inventions (Wilson and Delaney 2005). Fisheries scientists are being pulled more and more into hybrid domains that they share with social scientists and stakeholders. Fishing fleets can, and do, read scientific advice based on their behaviour and then change their behaviour, often in ways to avoid the implications of the scientific findings. Fish stocks do not. When scientists are dealing with such slippery units of analysis as these, it is very hard to separate the ‘political’and the ‘objective’. Tensions on the science boundary have many other sources. Gibbons (1999) argues that there is nothing really new about the boundary problem; reliable knowledge has always been reliable within certain boundaries. What is happening in policy contexts with the risk society is an increase in contestation. Knowledge that is incomplete now is not just knowledge that is waiting for better science; it is knowledge that will be contested. This contestation then puts further strain on the science boundary. Evans (2005) observes one case of scientific controversy where the more one side tried to show themselves as more scientific than the other, the more they broke down the appearance of science, and each began to appear as ‘scientific’as the other. A number of studies have shown that boundaries that are too rigidly maintained can lead to scientific elites being shielded from accountability until a crisis is required to expose the problems (Waller 1994). Increasing competition is also exposing scientific boundaries to increasing strain. Waterton (2005) carried out a large set of interviews in which she asked scientists to reflect on science. She asked about the impacts of the rise of contract-driven science, and many of her examples were from scientists involved in policy support. Scientists invoked scientific norms as something that has been lost fairly recently as science has become more and more contract-driven. Boundary work is a response to the repeated failure of methodological definitions to determine what is and is not science in practical contexts (Evans 2005). Science in practice can no longer be easily reduced to a single methodology or ascribed to a privileged subculture (Gibbons 1999). Halffman (2003) argues that neither the ‘cage model’nor the ‘seamless network’model seem to accurately describe the policy science interface. The ‘cage model’is based on the idea that it is possible a priori to establish criteria about what is ‘science’and what is ‘politics’. It is the model supporting the mainstream view displayed in Q 3.7. In the ‘seamless network’ model, for which Halffman (2003) cites Actor Network Theory as the most 71 Ostrom (2001), who is trying to explain the sustainable management of forests, defines polycentric systems as Q 3.12 the organisation of small-, medium-, and large-scale democratic units that each may exercise considerable independence to make and enforce rules within a circumscribed scope of authority for a specified geographical area (Ostrom 2001, p. 2). For Ostrom the strength of a polycentric governance mechanism is its ability to experiment with diverse approaches and provide for a range of responses to external shocks. While her focus is on sustainable governance, there is a clear analogy to science-based policy-making in conditions of uncertainty. Information processing, including the access of the smaller units to various knowledge sources, their ability to process feedback from policy changes and communicate these experiences to the parallel units, is a key part of the strength of polycentric polities. Polycentric approaches are also complex, redundant and hence expensive, and resistant to top-down steering (Ostrom 2001). The polycentric networks Cash and Clark (2001) see in science for policy contribute for similar reasons. They are able to provide methodological coherence across scale levels while still allowing local specialisation; the redundancy of the system provides for multiple pathways to encourage innovation and flexibility. A polycentric network also facilitates stakeholder capacity building and involvement (Cash and Clark 2001). It is an institutional design that gives form to the continuously operating process of communicative rationality. It is also the appropriate institutional design for an adaptive system. A polycentric network has multiple sources of knowledge and capacities for responding to problems as they arise on the scale level they arise at. They are, as Cash and Clark (2001) tell us, structures that allow more complex interactions to respond to needs. Polycentric networks also reflect current thinking about effective fisheries management in general that I would argue is broadly applicable to the EAFM. This is the idea of placing the burden of proof on those who wish to engage in economic activities and within a ‘results-based’management structure. A belief that this is the appropriate framework for marine management strongly influences my assumptions about the kinds of institutions that can benefit from the reflections I am offering on the nature of transparency and the production of scientific advice. The basic idea is that someone who wants to become involved in some activity in the marine ecosystem, such as fishing, is licensed to do so through a results-based agreement. This means that, in exchange for an otherwise secure and high-quality right to a stream of benefits, they must demonstrate that they honour a series of limits on their ecosystem impacts, for example limits on the catch, bycatch, impacts on habitat, etc. This requires scientific processes both at the level of limit-setting and at the level of particular activity staying within those limits. The economic actors will demand transpar78 The Paradoxes of Transparency ency in the limit-setting process, and managers and other interested parties will demand transparency in the demonstration that limits have been honoured. This would require the development of science in an interactive way for a wide range of possible activities at different geographical scales. If this is the institutional path towards EAFM that is chosen, and I believe it reflects our best lessons from fisheries management, a polycentric network of boundary organisations is bound to emerge. When I am discussing science and negotiation processes in these pages, this is the kind of framework I have in mind. The institutional form for effective adaptation seems to be polycentric networks of boundary organisations. These networks facilitate complex interactions that address emergent boundary issues at appropriate scales in a way that allows us to hold them accountable for both good science and good policy. Guston (2001c) tells us that successful boundary organisations have lines of accountability into both the science and policy communities, while current thinking on effective management institutions tells us that accountability to line managers and interested stakeholders is also critical. To be effective, such accountability requires transparency. So what we are after is a flexible, polycentric network of science-policy boundary organisations that are transparent both among themselves and to society at large. The critical institutional challenge that must be managed to achieve this goal I call the paradoxes of transparency. 3.3 The paradoxes of transparency Chapter 1 argued that the basic social function of science is to achieve the maximum possible transparency of knowledge so that knowledge can be used to guide collective action. Transparency may seem like a fairly simple idea. It is certainly often treated like a simple idea, easily achieved through passing ‘sunshine laws’or inviting someone to observe something. But achieving what we want to achieve with ‘increased transparency’is rarely straightforward. Transparency is not just about observing something going on, it is about understanding what is going on. Understanding a complex situation requires that it be articulated or described clearly and honestly, yet the very techniques used for clarity and openness often seem to lead to transparency-induced opacity. As I went about analyzing the case presented here, I found myself coming back frequently to this idea of the ‘paradoxes of transparency’. Transparency presents itself as a paradox in many ways and situations because the techniques used to achieve it often undermine themselves. This selfnegation is a common pattern. Because of this I thought of calling the idea the singular ‘paradox of transparency’. I chose ‘paradoxes’because I kept coming across so many different ways the self-negation happens, and they did not seem to share a general characteristic, save that they all made transparency more difficult. As this idea began to gel, I tried to give names 79 to the various paradoxes of transparency. I offer four of them in this section. Many of the ideas behind them appear fairly frequently in the literature although, with the exception of the ‘paradox of surveillance’, they are not used as concepts for understanding transparency as such. In this section I describe each of these paradoxes in turn. It is quite possible that, if others find this idea of the paradoxes of transparency useful, they will be able to sketch out others and perhaps develop a more rational scheme than mine. 3.3.1 The paradox of precision and expertise For the problems of creating a knowledge base for the EAFM, quantification is perhaps the main culprit behind transparency-induced opacity. The importance of quantification for science has long been recognised. Kant (1950/1783) argued that judgements of experience are always synthesised out of individual observations, while mathematical judgements are a priori and not based on experience. Quantitative methods are the paradigmatic form of science because they tap into the power of these a priori judgements. They begin with the accurate and consistent measurement of comparable units. When meaningful comparability and accurate measurement can be achieved, mathematical laws are mapped onto the phenomenon under study. Disputes must focus on either the comparability or the measurement of the units, and these two questions must be outlined with the greatest possible clarity if a result is to be acceptable as science. When this is achieved, this mapping of mathematical laws attains the greatest possible transparency of argument because it makes precise replication of results possible, in principle, if often not in practice (Collins and Pinch 1998). Beyond science, the clarity of quantification is also critical for democratic governance. Porter (1995) links the rising importance of quantification to democracy. Decisions made by bureaucrats and politicians, who are open to public criticism, need to be objectively justified. Decisions made by aristocrats do not. The rule of law can no longer rely on aristocratic judgement and is forced to rely on accountancy, statistical and often scientific analysis to provide that justification, usually in the form of some numbers (Porter 1995). The paradox itself arises from the fact that mathematical expertise is a difficult skill to acquire, as well as the fact that even those with the skills do not have the access and time required to assess the comparability or measurement of the relevant units. Finding mechanisms for collaboration and communication that are able to facilitate interaction across levels of mathematical skill is becoming very important in scientific activities and is taking up more and more scientists’time. This is certainly true among European fisheries scientists and is a problem that will be returned to frequently in this book. 80 The Paradoxes of Transparency 3.3.2 The paradox of quantification and reification The idea that quantification is about transparency is not paradoxical simply because mathematics require so much skill. To reify is to treat something abstract with an inappropriate level of concreteness. Reification is a constant danger when using numbers because the aura of objectivity that quantification lends to the measured often obscures aspects of its nature. Arithmetical language actually has a great deal in common with ordinary, informal language (Funtowicz and Ravetz 1990). The formality of the mathematical language is lost when expressed in normal speech, and numbers are used in ordinary speech to mean very different things. Funtowicz and Ravetz (1990) offer an analysis of numbers as a language based on the observation that a discussion about a formal language cannot take place in the language itself. They tell us a joke. A museum guide was showing some schoolchildren a collection of dinosaur skeletons. A little boy asked how old one particular dinosaur was and the guide said, ‘50 million and 12 years old’. The boy frowned ‘That’s a funny age?’‘Well yeah, but they told me it was 50 million years old when I started to work here 12 years ago, so it is just simple arithmetic!’Numbers are used in real speech to mean quite different things, and we laugh at the guard because in this situation that is obvious. This informality inherent in discussing mathematics, however, is often obscured, and the aura of formality remains. Measuring something can change the way it is seen so much that it becomes a different thing. An obvious example is ‘teaching to the test’in educational measurement, but this phenomenon does not require awareness of the measurement process. Porter (1995) argues that the concept of ‘society’itself was in part a statistical construct and that people didn’t really talk about ‘society’before there were statistical handles to do so. Crime rates and unemployment rates made it possible to talk about a society involving collective responsibility instead of just the condition of individuals. The measured items often evolve into standardised categories where individual variation is obscured. Porter (1995, p. 28) quotes an official of the US Bureau of Standards: ‘We have now reached the stage where there is a federally mandated method for measuring almost every physical, chemical or biological phenomenon.’Accommodating variation in measurement practices is impossible for a bureaucracy; even improvements are unhelpful unless made universal. But measurement is hard, and such standards also obscure this truth when reports contain spurious levels of precision. The transformation of landing figures into catch figures and thence into stock assessments is an example anyone in fisheries is familiar with. Measurement problems are easy to forget and ignore. In the American television crime drama ‘Numbers’, applied mathematics is transformed into an initially plausible form of magic by pretending that measurement is never a problem. 81 This combination of the hidden informality of numbers in speech and the ways that measurement transforms and invents phenomena has problematic consequences. A large part of the literature on local knowledge points to communication breakdowns that emerge when the day-to-day experiences of user groups are presented to them in the form of unfamiliar numerical categories. Roepstorff (2000) in his study of fisheries science in Greenland suggests that fishers ‘focus on fish as a living being’and think of them as ‘mass nouns’, while the scientist sees the fish as a ‘count noun’, meaning that the individual fish is a representative of the stock in the sense that the stock is the arithmetic sum of the single fish. Just as in daily speech numbers lose their precision while maintaining their aura, the quantitative presentation of material may also obscure underlying uncertainty. Mathematical models have been called a ‘technology of hubris’when they imply that managed use is feasible where precaution might be the wiser choice. They overstate the known and downplay ignorance, uncertainty and conflict (Harremoes et al. 2001; Jasanoff 2002). In many policy contexts some stakeholders will be distrustful of the models at the same time that others are praising them (Jasanoff 1986). The greatest danger in the application of quantification to policy is when a goal is given special importance, simply because its attainment is easily measured. Quantification creates the possibility for great transparency in demonstrating that something is true, while simultaneously undercutting that transparency with an aura of formality and objectivity, and a precision that is often not justified. 3.3.3 The paradox of surveillance The surveillance paradox is well recognised in respect to negotiations and other processes seeking compromise and understanding among diverse parties. The world-famous Chatham House Rule, ‘When a meeting, or part thereof, is held under the Chatham House Rule, participants are free to use the information received, but neither the identity nor the affiliation of the speaker(s), nor that of any other participant, may be revealed’,is based on this understanding. When it is invoked, participants are much more open to information-sharing and search for compromise than when they have to account to their home constituency for every word they say. Especially the very formal and comprehensive transparency rules such as those found in policy-making in the United States hide as well as disclose important information; openness rules alone do not provide missing information, expose all the hidden assumptions and may even dampen novel interpretations of facts (Jasanoff 2002). Degnbol and Wilson (2008) suggest that it is helpful to distinguish internal and external transparency. Internal transparency addresses the need for stakeholders to be transparent to each other so that areas for potential compromise and common ground can be found. This is not always easy 82 The Paradoxes of Transparency when people involved in negotiations are having everything they say scrutinised by their constituencies. External transparency allows decision-making processes to be accountable to outsiders and finally to the public. Because of this, internal transparency must be limited, and attention must be paid to balancing the two kinds. The paradox is simply that in negotiations, transparency to others –i.e. to the public –reduces the internal transparency that allows negotiators to understand each other and search for possible compromise. The paradox of surveillance leads to complex dances in structuring complexity. One example is the discussion in Section 7.3.5 by the ICES delegate who was willing to have stakeholder observers at physical meetings but not video meetings, because it would be impossible to observe which stakeholders were present and how much they were influencing the scientists. This is not something, however, that applies only to negotiations looking for compromise. It also applies to discussions of nature that are seeking consensus. Pictures of nature can be self-reinforcing. The facts that create the picture may not be wrong, but especially when they fit a particular political narrative, they can make it difficult for contradictory facts to be raised. This becomes a particular problem when facts take on a symbolic importance to particular groups. Groups have unifying symbols, and such symbols take many forms, including particular assumptions about natural facts. In fisheries, as elsewhere, belief in certain facts can act as an ideology that maintains and tests loyalty. This symbolic status shields these facts from criticism. The North Sea experience with illegal or ‘black’landings for fish in the 1990s and early 2000s fits this pattern. Individual fishers were quite open about the extent of the problem privately –in confidential interviews, for example –but it was a long time before it could be honestly addressed in public fora. In fact, this did not really happen until stronger enforcement mechanisms were put in place, making it possible for fishers to comply without simply hurting their own finances while helping no one. 3.3.4 The paradox of scale Often when one hears discussions of transparency, the speaker seems to be referring to making a process transparent to a very wide number of people, or even to the public as a whole. Transparency involving greater numbers of observers leads to greater accountability with respect to a wider range of values. The problem is that the larger the number of people, the greater the distortion of information introduced by the systems meant to make transparency possible. A good example would be the way the activities of one member of a polycentric network of boundary organisations are transparent to the rest of the network. The paradox creates substantial difficulties for both the larger group wishing to observe and the smaller group being observed. In a perhaps confusing way, this large and 83 small distinction applies both up and down a hierarchy of institutions; it can mean a group of high-level decision-makers being held accountable to the entire interested public, as well as one particular fishing fleet being held accountable for their activities. Discussing the paradox of scale with the detail I think it deserves means taking my last dive into explaining the basic concepts of the Communicative Systems Theory (CST) that has guided my research. As I discussed in Section 1.2, institutions are not only arenas of competition and contention, they also require the coordination of behaviour if they are to function. CST focuses on the communicative mechanisms that make such coordination possible. Up until now I have mainly discussed only one of these mechanisms, rational communication. Table 3.2 lists all the communication mechanisms used in institutions, and any institution will use a mixture of some of these mechanisms to coordinate action. The reader should be warned that while the basic ideas behind these mechanisms were inspired by Habermas’s (1984, 1987) concepts, especially that of ‘steering media’,I have taken his ideas in a direction that he would likely not recognise. This means that any incoherence is my fault rather than his. 7 The main difference between the mechanisms is the degree to which they are embedded in what Habermas (1984) calls the ‘lifeworld’, which is the rich mixture of shared background meanings that make communication possible. The mechanisms towards the bottom of Table 3.2 are based on simple mutual understandings that require little or no discussion. A can buy a candy from B, or drive through his country, without even sharing a language. If A has authority over B, A can get compliance from B with very little discussion. The mechanisms closer to the top of the table draw much more heavily on the shared lifeworld. The more embedded mechanisms motivate behaviour by convincing people that something is true or right, and these behaviours are much less predictable than the less embedded mechanisms which rely more on coercion. This is a fairly broad use of the word coercion, in that I apply it to choices in which people are faced with a take-it-or-leave-it decision, as well as situations in which direct sanctions are invoked to win compliance. The model in Table 3.2 is very similar to the state –market –civil society model, the first use of which is often credited to Polanyi (1957), and that is still widely used today (Jentoft and McCay 2003). The main difference is that it focuses on the mechanisms that handle the communication needed to coordinate action, not on the institutions themselves. On lower scale levels, i.e. among fewer people, coordination is handled mainly by coming to a mutual understanding through rational communication. On broader levels, the communications that allow coordination must be streamlined and made more predictable. This means that the orientation of the discussion towards convincing participants that something is true or right begins to break down. As the level on which the institution operates grows larger, coordination requires a presupposition in communications 84 The Paradoxes of Transparency that the opinions and values of individual participants do not matter; hence the use of the word ‘coercion’. This is the reality of what must be in place if behaviour is to be coordinated across broad-scale extents. The informationprocessing problem it presents, however, is a severe loss of richness and nuance in the information that the institution can identify and respond to and that rational communication makes most abundantly available. The result is a high potential for a systematic distortion of communication. Table 3.2 Four continua characterising the CST communications mechanism Communication mechanism Embeddedness in lifeworld of communicative resources How the mechanism elicits behaviour Predictability of outcomes when mechanism is used Scale Examples of institutions that rely heavily on this mechanism for coordination Rational communications More embedded in the lifeworld Relies more on convincing Less predictable outcomes More effective on smaller scales Science Social movements Prestige Influence Local communities Social networks Authority Money Right-of-way Less embedded in the lifeworld Relies more on constraining More predictable outcomes More effective on higher scales Governments Markets Traffic A similar table was originally published in Wilson (2003) Adaptive ecosystem-based management requires above all a capacity for institutional learning, and the paradox of scale has direct implications for how institutions learn. Institutional learning across scales requires that knowledge flows both up and down, and this requires the translation of information into forms useful at another scale level. Such translation often distorts the information in the eyes of its producers, as is often seen when trying to use fishers’local ecological knowledge in management (Holm 2003). The information required can range from data about natural or socio-economic processes, to scientific findings of expert groups, to information about the basis and results of decisions made by lower-level management institutions. A process of diffusion from higher levels is necessary to establish and maintain a system for transparency and accountability. Then a process of concentration of information is necessary to gather and condense the information. Concentration involves packaging information into a form that will be useable on the higher scale level. The information that 85 is the product of these packaging processes is not merely raw data and particular indicators; it includes findings, behaviours, scientific findings and group decisions. One of the interesting lessons for me from this case study is that this whole process requires many kinds of consistency (Section 7.3.1). The packaging process involves four processes of the transformation of knowledge: –Simplification means the filtering of the knowledge to leave only the knowledge that will be useful on the scale level receiving the knowledge. This might be thought of as increasing the ratio of signal to noise in the information; –Abstraction means formatting the knowledge for applicability to wider situations than the one that produced the knowledge; –Codification means categorising the knowledge within a symbol system that will be used to generate further knowledge through comparisons; –Standardisation means simplifying, abstracting and codifying knowledge according to a common system. The heart of the problem of institutional learning across scales is the distortion of communication (Habermas 1987). By communicative distortions I do not mean factual inaccuracies or poor interpretations of information. Nor do I mean the information-packaging processes as such, though it is in these packaging processes that the communicative distortions do their damage. Communicative distortions are severe deviations from the requirements of rational communication: that 1) people can effectively raise any claim without manipulation and 2) discussions of fact are separated from discussions of values and interests. In the diffusion process, claims are often blocked by heavy reliance on coercion-based communication mechanisms. This means, for example, that instructions about how to package information are formalised and even made into legal requirements, which in turn must be very strictly defined in order to be enforced. Enforcement institutions are not geared to take into account questions about how well these packaging instructions handle the information they are meant to communicate. The packaging can become routinised to the point of being ritualised, resulting in everyone being forced to take a cynical distance from the ‘knowledge’being created. This formalisation and specificity of communicative content reduce the richness of the information being gathered, while at the same time allowing the people on the level doing the packaging to be selective about the information they are sending, because all that is really required is formal packaging. They cannot raise claims, and they cannot be effectively questioned. For example, the impacts of this blockage of questions and explanations accompanying required findings on scientists working in ICES expert groups is a central topic in Chapter 5. 86 The Paradoxes of Transparency Wilson and Degnbol (2002) outline an example of this kind of communicative distortion. A group of scientists shared a belief about the condition of a fish stock. The key assumption of the model they chose to define formally as the ‘best available science’for this fish stock, however, was the precise opposite of what they believed to be true. The issue here from the perspective of institutional analysis is not how well the model they selected reflected the real situation in the ocean, as their original belief could also have been wrong. The issue is that a group of scientists was forced by a set of institutional imperatives to say the opposite of what they believed; they were effectively blocked from raising their claim about what they actually thought was going on. What blocked them was a combination of legal requirements for particular model parameters and the characteristics of the institutional peer review process they were operating under. These requirements were put in place precisely to help a large-scale science-policy community ensure the transparency of scientific decisions. These distortions of communication within marine management institutions can be hypothesised to operate in particular directions. I would suggest that they would tend to make management problems appear more tractable than they are. This is because during the simplification and standardisation processes, information about local nuances will be excluded, and this will hide problems. They would also tend to make their management seem more effective than it actually is, because the same information that is formulated through institutional imperatives is used to evaluate the effectiveness of the institution. Furthermore, they may tend to hide linkages between scale levels because the packaging of knowledge from the lower level is given form by questions that are driven by the level and resolution at which the questions are posed. They may also be subject to a positive feedback loop. The responses by higher-level institutions to the results of communicative distortions are to create even more stringent packaging requirements that then increase the distortions. The paradox of scale is a particularly dangerous challenge when trying to figure out how to structure transparent institutions. These four paradoxes mean that transparency can never be taken for granted. It must always be fine-tuned, and as a result transparency can almost never be brought about simply by setting up rules. Too often a simplistic approach to transparency will simply end up making matters worse. 3.4 Summary of the theoretical discussion These first three chapters trace a theoretical path towards understanding science for an ecosystem approach to environmental management. My broadest starting point has been an approach to human ecology that sees the social system as a communicative system governed by shared meanings functioning within a set of environmental systems governed by natural laws. Given this orientation, CST is an informative way to think about 87 official ICES advice emerges. As the NFIs employ these people and pay for their participation in the expert groups, they exercise an important indirect control over these central ICES activities. They also hold considerable formal control over ICES because of their membership on the ICES Council. What exactly they are controlling, and how much they are really controlling it, are open questions. ICES is a lot of things. It is an intergovernmental scientific organisation which is charged with producing scientific advice to be used in official decision-making. It is a primary professional organisation for marine and fisheries scientists in Europe and an important secondary organisation for North American scientists. It is a building in Copenhagen where a professional staff processes data, edits kilograms of scientific advice, and constantly hosts meetings on many subjects. It is also a loose network where approximately 1600 scientists work together to address scientific questions, some of which they are eager to examine and some of which they wish had never been asked. Formally, ICES is an intergovernmental organisation with 20 member countries. Each member country has two delegates on the Council, and the Council is the highest decision-making level in ICES. It is chaired by the ICES President. The two delegates from each country are usually the Director of the country’s NFI and someone from the relevant central ministry. The Directors are the more active of these two groups, and all current members of the Bureau, an executive committee elected by and from the delegates, are NFI Directors. Figure 4.1 Overview of the advice system for marine and fisheries management in Europe ICES must also respond to its clients. The biggest of them is the European Commission, mainly in respect to the Directorate General of Maritime Affairs and Fisheries (DG MARE), but ICES also receives some advice re94 The Paradoxes of Transparency quests from DG Environment. The regular clients include the government of Norway, the OSPAR and HELCOM Conventions, the North East Atlantic Fisheries Commission (NEAFC), and the North Atlantic Salmon Conservation Organization (NASCO). ICES is very dependent on the willingness of scientists to make their expertise available. It is true that many scientists, even while sometimes referring to themselves as volunteers, participate in ICES expert groups because they are assigned to do so by their laboratories. However, the degree of enthusiasm with which they do so is very much under their own control. Furthermore, many other scientists participate in expert groups solely because of their own interest in the issue or the scientific subject matter. As explored in some detail in the coming chapters, mobilising these people’s expertise is a real challenge, and the attitudes of the constituent scientists make up an important part of the political environment of ICES. The next section of Figure 4.1 is the box marked European Union that contains a number of sub-sections. The central actor, although not the final decision-maker, is DG MARE. Like other clients, DG MARE directs funds and requests to ICES. The amount of money involved is not huge; the current Memorandum of Understanding between ICES and DG MARE puts the figure at approximately 750,000 euros for the recurring advice. The group at DG MARE that actually uses the advice, however, does not formally receive the advice directly from ICES. Through a fairly recent change in procedures, official advice must pass through the Scientific, Technical and Economic Committee for Fisheries (STECF). Their job is to receive the ICES advice, review it, add economic information as appropriate, and pass it on to DG MARE. An important difference between STECF and ICES is that STECF is DG MARE’s own body, and its budget is under the direct control of DG MARE. An important similarity between STECF and ICES is that these are very often the same scientists, who are getting their per diems paid by a different public body. DG MARE also has the option to send requests directly to STECF, an option that it exercises frequently for certain types of short-term and very specific requests. In addition to the scientific advice from STECF, DG MARE also considers political and social advice from two stakeholder structures. The Regional Advisory Council (RAC) system was set up by the European Council in 2002 (CEC 2002) and given practical form in 2004 (CEC 2004). They are meant to be stakeholder fora bringing together divergent groups to try to produce a consensus approach to fisheries management problems to help meet management goals. There are seven such RACs; five correspond to the geographical level of a shared sea, the other two relate to particularly far-ranging fisheries. A RAC is required by the Council Decision (CEC 2004) to allocate two-thirds of the seats on its executive council to the harvesting sector and one-third to ‘other interest groups affected by the Common Fisheries Policy’(256/17). The other group, the Advisory Committee on Fisheries and Aquaculture (ACFA), is similar in purpose to the RACs, 95 but operates at a European level representing mainly cross-European groups. ACFA is even more strongly weighted towards economic interests than the RACs are. DG MARE’s output is proposals to the Council of Ministers for decisionmaking. The decision-making power has recently been expanded by the Lisbon Treaty to include a ‘co-decision’by the European Parliament, but how this will influence the process is unclear, and the Commissioner’s current position is that it will not influence the settings of TACs. Up until now their role in fisheries has been advisory. The Council of Ministers, in hectic end-of-the-year sessions, decides on the fisheries legislation for the coming year. While they do issue a number of ‘technical measures’, i.e. rules about how to catch fish and some restrictions on how much fishing will be allowed, the most important decisions are related to setting the TACs for each species and their associated member state quotas. The quotas are governed by the relative-stability rule, under which the quota allocation is based on the proportion of the catch which the member states enjoyed before joining the CFP. Under the CFP the member states are to receive an allocation, and it is up to them how they distribute that allocation to their fishers. Member states are also responsible for passing the relevant laws and regulations, monitoring and enforcing compliance. The EU role in this area is growing rapidly. This includes not just the fisheries management measures but the data-gathering activities as well. Monitoring and enforcement at the member-state level has been the weakest part of the CFP. But data-gathering, behaviour-monitoring and enforcement have improved significantly in the past few years, due to the tightening of standards by the European Union as well as a growing influence of conservation groups at the member-state level in response to the CFP’s poor conservation record. 4.1.2 Gathering the data for fisheries advice The various scientific survey vessels owned and operated by the NFIs produce the valuable ‘fisheries independent data’that constitute the most important input, at times even the only empirical input, to the stock assessment models. The cost of scientific surveys added up to 34 million euros, making surveys the largest single expenditure in the entire scientific advice system. ‘Fisheries dependent data’that comes mainly from sampling landings constitutes the second largest expenditure of 19 million euros. As discussed below, the quality of these data has been the single most prominent issue in European fisheries management. These data-gathering expenditures can be compared to the 7 million euros that pays for all the analysis, stock assessments and advisory work (EASE 2007). These resources have been invested somewhat unevenly among the different fish stocks, with the northern area, including the North Sea and Baltic demersal stocks, attracting the most resources. Some 40% of the resources was 96 The Paradoxes of Transparency spent on these stocks, while the rest of the demersal stocks received 20%, pelagic stocks 30%, and deep sea stocks 5% (EASE 2007). In 2000 the European Union established a common regulation for the collection and management of fisheries data for all member states. This regulation is known as the Data Collection Regulation (DCR). The DCR is the answer to a need for harmonised procedures for data collection that remain the same from one year to the next. It is meant to provide an overview of the activities of the fishing fleets, total catches, price trends and the economic situation in the sector. The DCR has undergone some developments since its first adoption in 2000. The reason for this is especially the 2002 reform of the Common Fisheries Policy through which the EU management of fisheries has been moved from a concern with how to manage individual stocks to a concern with how to manage the social and economic units of European fisheries and fleets. The main trend within the development of the DCR has been towards meeting the need for multi-species fleetand area-based advice. It is anticipated that from 2009, data will be collected by fleet and fishery, rather than by stock. In March 2008 a new version of the DCR was adopted that is intended to prepare the system for the EAFM. The new DCR introduces satellite monitoring of vessels. It expands the EU-funded data collection programmes from one to three years. It places more emphasis on the gathering of social and economic data so as to provide a basis for impact assessment of new legislation and to allow monitoring of the performance of the European fleet. Paragraph 8 in the proposal for the new version states as follows: Q 4.1 Data collected for the purposes of scientific evaluation should include information on fleets and their activities, biological data covering catches, including discards, survey information on fish stocks and the environmental impact that may be caused by fisheries on the marine ecosystem. It should also include data explaining price formation and other data which may facilitate an assessment of the economic situation of fishing enterprises, aquaculture and the processing industry, and of employment trends in these sectors (CEC 2007, p. 9). To many fisheries scientists, the data directive guidelines of 2002 actually implied a decrease in international coordination. The data directive had established sampling schemes for all of the EU member states to adhere to, but by doing that, it had also de-emphasised former cooperative sampling programmes. Since that time a new system has been created under more direct EU auspices. New internal mechanisms were needed to ensure the international coordination of sampling. The sampling of discards was one of the areas which proved to be a challenge. While all member states should engage in a national sampling programme, it was left to the individual member states to decide how to estimate the discards. There was no international coordination in place to ensure that sampling oc97 curred from all the fleets fishing any particular stock, and this in turn prevented the full utilisation of discard estimates into the stock assessments. The international coordination of landing samplings was complicated further by the very heterogeneity of the authorities responsible for enforcement and inspection (e.g. the navy, the police, customs, etc.). The scientific methodology relating to the length and ageing of landings was also in need of international coordination as some laboratories had no established expertise. Tools were required for analyzing and storing international sampling results, and the division of responsibilities for the management and development of the international sampling system needed to be clarified. These issues are being addressed to some degree. The concerns in the previous paragraph were themes in interviews done in 2003 and 2004. International coordination is increasing, especially through programmes designed at the regional (e.g. North Sea) level. Here is a respondent from the data programme at DG MARE speaking in 2007: Q 4.2 Interviewer: How much of your time do you spend talking to people in member state ministries? Respondent: Rarely, I would say. With the data collection system we have regular meetings with member states, normally when we are evaluating their national programmes. We are not in the process of evaluating their national programmes in 2008. We have bilateral meetings, and my unit is participating in all this regional coordination work. This is a good occasion where you meet not the higher level decision-makers but the people involved in implementation. These are observers and people coordinating observers, people dealing with data collection issues. Data for fisheries advice are not available in either the quality or quantity DG MARE would like to see. According to one of our respondents from DG MARE, they had the sense that processing data for advice is not attractive to the employees of the national fisheries institutes. They thought this true even to the extent that those who do it are not seen as scientists, and there is little money available for these activities in comparison with research projects. One point of the new initiatives was to make some investments to change this situation. The impact of stock assessment work on scientists’careers and working conditions is the central theme of Chapter 5. A respondent at DG MARE: Q 4.3 The data collection framework has been running for five years now. If you listen to an end user like ICES, they say that the programme is functioning well. It has stabilised the flow of data to the assessment work. That does not always apply to the quality of the data. That is a problem, and most of the problems are with missing landings, misreporting, black landings or whatever you want to call it. 98 The Paradoxes of Transparency Fisheries data also raise questions of confidentiality and control. Because of the concerns of member state governments, reflecting in turn the concerns of their fishers, the current legislation does not allow DG MARE to have its own databases, and they are only allowed to use data for 20 days. Getting data is actually easier for scientists when they request it wearing an ICES hat than if they request it wearing an STECF hat. If they want information for STECF, the request has to go from DG MARE to the Ministry, but if they want it for ICES, they can just make a request directly to the NFIs. DG MARE can also make the data from fisheries inspections available to ICES, although with a number of restrictions. An agreement between DG MARE and ICES about the use of community fisheries inspections data for scientific purposes was described by an ICES official. It is a long and complex document. It specifies exactly what DG MARE is willing to provide, requires that ICES not publish the data and restricts access to members of relevant expert groups. Furthermore, ICES must only analyse the data for purposes of assessing catch statistics for assessment and advisory purposes, and in ways that are restricted in terms of geographical resolution. ICES is also restricted from making comments about individual member states, let alone individual fishing vessels. Some member states also specifically bar scientists from access to data from the vessel monitoring system (VMS) satellite tracking. While efforts at standardisation have been critical, problems with fisheries data are extensive and often politically charged. The most basic issue is the cooperation of fishers with the data programmes and the related problem that the fish catch is not the same thing as the recorded fish landings. This happens because of discarding of fish at sea or ‘black’landing of the fish. There are many possible reasons for both problems (Wilson 2000). Very briefly, discarding may be because of ‘high-grading’, throwing out fish caught earlier to accommodate more valuable fish caught later, or ‘regulatory discards’where fish are caught that are illegal to land for quota or other reasons. ‘Black’landings may be a way to avoid either fishing regulations or taxes. At the ACFM meeting in September 2004, the issue of Eastern Baltic cod was discussed as a particularly difficult example. Landings from this stock were estimated to be underreported by as much as 35%. Yet the scientists have been concerned that if they pushed this issue, they would get even less cooperation from the industry. These concerns are reflected on the member-state level as well. Some countries in the area have a rule that scientists can only use official data, so they created a category called ‘unallocated’data to detach the estimated misreporting from national data. When this decision was reviewed, these scientists received comments back that this category was not acceptable. This kind of restriction means that the data cannot be evaluated. It creates ‘black boxes’ in the system and undermines scientific credibility of the scientists involved, even up to the ICES level. One scientist complained that if one names a specific country in trying to address misreporting, this will just make the situation worse. Indeed, this person suggested that this worsen99 ing may even extend to relationships with the scientists from that country. Another scientist from the region said that for these reasons, they never mention the countries where misreporting is undermining the value of the data. Very similar problems have been found in the North Sea and elsewhere. Data on discards and bycatch are both particularly sensitive. One scientist we interviewed was charged with putting together discard information so that it would be commonly available to scientists throughout ICES. This proved to be a problem because various countries did not submit their information while other countries submitted all that they had. The database they created was very patchy, although he argues that the new EU regulations have improved this situation. Three countries had already collected the discard data as part of a research project, but when these data were desired by an ICES study group, one country would not provide it because of a disagreement with their fishing industry about whether the results were representative. A good example of the problems around data and fishers’cooperation is Scotland’s experience. Scotland has been a particular place of contention with respect to discard information. Scotland was the earliest place to begin a regular programme of hosting observers on board fishing vessels to gather such data. Once the data were made available, they drew attention from both the EU and local conservation NGOs that increased pressure on the industry. This led in turn to a feeling among the industry that the data that they were helping to provide were being used against them in a way that penalised them more than other fishers who had not provided the data. In at least one case this led one of the fishers’organisations to withdraw from providing more data. Interestingly, Scottish scientists felt some kinship with the industry on this issue. On two separate occasions Scottish scientists in interviews indicated that Scotland had made more than their fair contribution to data-gathering, though their reaction was that it was high time other countries started picking up their slack rather than that Scotland should pull back. A scientist we interviewed in 2004 who works directly with the Scottish observer programme reported that it is becoming more difficult to solicit cooperation. This is something that has built up gradually, but the quota setting for 2003 seemed to have been a watershed event. At that point it became clear that fishers’organisations no longer wanted to provide cooperation at the previous levels. It was not the discards issue per se, but a general attitude towards government establishments that contribute towards quotas and management. The data went through the labs, and the fishers felt that they were only providing information that would be used against them. The observer programme still gets cooperation; our respondent argued that the personal approach was still the most important. But this withdrawal of cooperation has had an impact on the observers’morale, and these are not easy positions to fill or train people for. It has also, on rare occasions, meant a hole in the sampling scheme; he estimated 100 The Paradoxes of Transparency roughly that five trips failed in 2003 out of a total of 70. The most extreme expression of industry refusal to cooperate with data collection took place in Northern Ireland where fisheries officers were kept from sampling landings through threats of violence. A scientist involved in the expert group says that this activity has had an impact on the Irish Sea stock assessments. The increased emphasis on data-gathering and especially investments in general fisheries enforcement should begin to make a difference. For one thing, video surveillance on fishing boats is beginning to appear rapidly. Enforcement is a critical support in bringing about any kind of cooperation. Even fishers with the best of intentions are forced to cheat in a system where there is no enforcement; otherwise they hurt their own families while benefiting no one. Earlier in this decade, when the enforcement and data problems were finally beginning to get considerable attention, a pattern in interviews with fishers was for them to begin by talking about how much they had been misreporting their landings, and how much they wanted to see a system where they did not have to. Hopefully, such a system is beginning to be put in place. 4.1.3 The National Fisheries Institutes In this chapter and throughout this book I refer to what may be the most important institutions in the whole advisory process using the single term ‘National Fisheries Institutes’or more often NFIs. I thought it was important here to offer some very short descriptions of a few of these organisations. What follows is not an in-depth history or analysis, indeed the information comes almost entirely from the institutes’own websites. Such an analysis is not necessary for adequately describing the case at hand. What I hope this section does do is give the reader an idea of both the similarities and diversity of these entities that I for the most part treat almost as a unit. They have diverse cultures, quite proud scientific histories, and different organisational structures, while also having both similar missions and a need to respond to similar stakeholder groups. They are the employers of most of the scientists interviewed for the book, and hence their primary institutional identity. Some are government entities fully encompassed by the ministry in charge of fisheries, others are parts of universities. Both DTU Aqua and IMARES recently changed from independent research institutes to university research centres. After the short descriptions some general remarks about the NFIs complete the section. The Institute of Marine Research in Bergen, Norway The Institute of Marine Research (IMR) is the largest centre for marine science in Norway. The scientific focus of IMR is on aquaculture and eco101 systems. The main task is to advise Norwegian authorities on ecosystems and aquaculture, and half of IMR’s revenue comes from the Norwegian Ministry of Fisheries and Coastal Affairs through earmarked advisory contracts. IMR organises its marine and ecological research into 19 research groups, and together they cover a wide range of research on e.g. different maritime species, health, animal welfare, genetics, oceanography and data maintenance. The subjects of fisheries dynamics and fish capture constitute two fields of their own within this structure of 19 research groups. The group of fisheries dynamics studies capture rates, effort and fleet development, collects fishery-dependent data and performs monitoring services. The fish capture group, for its part, seeks to develop capture methods that are at once environmentand resource-friendly and energyefficient. IMR has recently started to work more closely with the industry in order to improve the data collection and information flow between fishermen and scientists. This is being done through their ‘Reference Fleet’ programme where IMR pays a small group of fishermen to provide detailed information on their fishing activity and catches. IMR’s history as an institution goes back to 1860 with a parliamentary proposal to fund practical-scientific fishery research in 1859 (Nordstrand 2000; Schwach 2000). This led to the first investigations of herring and cod fisheries in the early 1860s. In 1900 the Norwegian parliament established its first fishing directorate and divided it into a practical-administrative and a scientific branch. The scientific branch was reorganised in 1947 and developed into the institute named IMR. IMR remained part of the Norwegian fishery directory until 1989, when the former scientific branch, now an independent institution, split from the administrative one. Today, the two old branches continue to cooperate, and they share the same physical facilities. Due to its national advisory function, the main geographical areas of research interest are the Barents Sea, the Norwegian Sea, the North Sea and the Norwegian coastline. In addition to the many activities near home, IMR also supports fisheries research and management in developing countries. Through their ‘Centre for Development Cooperation in Fisheries’IMR has worked with Asian, African and Latin-American countries and describes itself as heavily engaged in development aid activities. IMR also engages in international research activities which are not attached to development aid. It advises international organisations and commissions and works with international sister institutions including 50 years of cooperation with the Russian PINRO Institute. The Centre for Environment, Fisheries and Aquaculture Science in Lowestoft, England For England and Wales, CEFAS acts as the main aquatic scientific research and consultancy centre. It provides the UK government and other clients 102 The Paradoxes of Transparency with research, advisory, monitoring, consultancy and training services. The goal of CEFAS’s scientific endeavour is to help preserve the aquatic environment, develop sustainable management and protect the public from aquatic contaminants. CEFAS is engaged in a broad range of research and advisory activities within both fisheries and the broader field of marine science. CEFAS’s research area is divided into the three categories of ecosystem interactions, organism health and resource management. Fisheries research is a subfield of its own under the latter and refers primarily to the exploration and assessment of the impacts of human activity and management on aquatic resources. Back in 2003 CEFAS began to work directly together with British fishermen due to a funded initiative by the UK government. Working together with fishermen organisations, CEFAS is building up relationships between fishermen and scientists in scientific cooperation. What is today known as the ‘Centre for Environment, Fisheries and Aquaculture Science’started out as a fisheries laboratory in the town of Lowestoft in 1902. It was established as the UK’s contribution to the then newly created ICES. Being an executive agency of the UK government’s Department for Environment, Food and Rural Advice, CEFAS is strongly connected to a UK national political agenda and has to provide the UK government with a range of services. CEFAS is mandated by the central and local government to licence deposits at sea, undertake visits and recommend enforcement actions. Their history is an extremely proud one. It was at the Lowestoft laboratory that Raymond Beverton and Sidney Holt developed the basic concepts and equations on which modern fisheries management is based. From a research and funding perspective, CEFAS’s profile is national, European and global. Most of CEFAS’s current projects are ordered and funded by the European Union and have either the North Sea or Europe as their area of research. The UK government remains CEFAS’s other large customer. Having international aid agencies and national government bodies from outside Europe on its list of customers, CEFAS also engages in global-scale research and consultancy. Fisheries Research Services in Aberdeen, Scotland In Scotland, the Fisheries Research Services (FRS) acts as an advisor for the Scottish and UK governments. The primary task of FRS is to advise these governments on marine fisheries and aquaculture and the protection of the aquatic environment and its wildlife. FRS became a government agency in 1997. It is an agency under the Scottish Government Marine Directorate and is a purely government-owned institution. Its research activities are to a large extent confined to the Scottish context and defined according to government policy interests. 103 Thus, a practical solution is reached that will likely last until someone decides to make an issue out of someone else’s participation in an expert group. It is interesting to note here the reference to the EAFM. This is the first mention of the related issues of the science/advice boundaries and the mobilisation of the many different kinds of expertise that the EAFM will require. These issues will be dealt with much more in Chapter 6. Figure 4.2 The current structure of ICES (www.ices.dk) Figure 4.3 outlines the ICES structure before the changes that began in January 2008. The major difference is the creation of ACOM in place of what had been the Management Committee for the Advisory Programme (MCAP) and the three previous advisory committees: the Advisory Committee for Fisheries Management (ACFM), the Advisory Committee for Ecosystems (ACE) and the Advisory Committee for the Marine Environment (ACME). According to its current Memorandum of Understanding with DG MARE, which is available on www.ices.dk, ICES provides recurring advice for 38 species or species groups. Recurring usually means annual, but particularly poor data or special characteristics in the biology of the animal may require a different time interval. The ‘38 species’also implies a much higher number of fish stocks, and the annual fisheries advice is of the order of 1600 pages long. In addition, ICES provides non-recurring advice through special requests that are negotiated on a case by case basis. Table 4.2 is provided to give some examples of the kinds of advice requested. Some of these requests may be so-called ‘fast track’requests (examples in Table 4.3) that ICES must organise and respond to under exceptional time pressure. 110 The Paradoxes of Transparency Figure 4.3 The structure of ICES before the reorganisation (ICES 2007c) Nevertheless from a DG MARE perspective, in spite of the fast track system, the advisory system is not designed to respond swiftly enough to urgent requests (CEC 2003a). DG MARE is considering making greater use of short-term contracts and more stringent prioritisation procedures in order to address issues which arise in the EU political arena or in international negotiations, for example, that need quick scientific assessments. Because ICES is geared towards addressing the ongoing annual rhythm of stock assessments and generating advice for TACs, it is difficult for them to organise resources to address such requests from DG MARE. 111 Table 4.2 ICES special requests pending in May 2008 Client Topic Date received ICES response ICES groups Client deadline EC-DG MARE First 3 as examples of a total of 10 special requests Long-term management of the NEA mackerel stock and fishery 25/01/2007 Progress report ACFM Oct. 2007 Final response 25 April 2008 Ad hoc group As ICES response Management plans for NS saithe, NS herring and herring IIIa 27/03/2007 Saithe 25 April 2008 Herring stocks 27 June 2008 WGHMP, WGNSSK Ad hoc group As ICES response Revision of Salmon Action Plan –Baltic 09/08/2007 26/06/2008 Ad hoc group and WGBAST (13-16 May) As ICES response NASCO Advice for 2008 12/07/2007 09/05/2008 WGNAS 15/05/2008 NEAFC First 3 of 10 Blue ling spawning aggregations 30/11/2007 23/05/2008 WGDEEP As ICES response Stock structure of Sebastes mentella 30/11/2007 Early 2009 SGRS + attached workshop Not settled yet Advice along the lines of that given recently regarding deepsea species on the appropriateness of the introduction of potential management units for redfish in the Irminger Sea and adjacent waters 30/11/2007 23/05/2008 WGDEEP As ICES response Norway IUCN criteria for red-listing marine fish species 20/09/2007 To be decided To be decided Not settled yet Norway Management goals for the harp seal stock in the northeast Atlantic 27/02/2008 To be decided To be decided ;Not settled yet 112 The Paradoxes of Transparency Norway, EC and the Faroes Mackerel, unreported landings 25/02/2008 To be decided To be decided Not settled yet OSPAR First 3 of 14 An assessment of the changes in the distribution and abundance of marine species in the OSPAR maritime area in relation to changes in hydrodynamics and sea temperature 01/07/2006 ACE June 2007 9 June 2008: final response Many SGs and WGECO involved 09/06/2008 Peer review of further nominations for threatened and/or declining species and habitats 01/07/2007 01/02/2008 Ad hoc group 01/02/2008 Development of proposals for Environmental Assessment Criteria 01/07/2007 23/05/2008 WGBEC 02/06/2008 Thanks to ICES for providing this information. 113 Table 4.3 Fast track requests to ICES in 2005 Client Fast track request DG MARE DG MARE and Norway DG ENV NEAFC I BSFC OSPAR HELCOM Norway 1. Compile status list of EU fish stocks 2. Sole in IIIa B new information to be included in re-assessment of stock 3. Bycatch of common dolphin 4. Advice on deep-sea stocks 5. Long-term management of Baltic cod 6. Request on restocking of glass eel 7. DNA analysis of Baltic salmon 1. Long-term management advice 1. Influence of sonar on marine mammals and fish 1. Information on stock identity of Sebastes mentella and quantitative information to allow spatial and temporal limitations in catches 2. Advice regarding the proposal for the protection of vulnerable deepwater habitats 3. Stock assessment methods for Atlanto-Scandian herring and blue whiting stocks 4. NEA mackerel stock assessment methodology 1. Advise on areas with the Gotland Deep and Gdansk Deep where the hydrological condition allow for a successful cod spawning in 2005 1. The design of one-off surveys to provide new information for a number of OSPAR Chemicals for Priority Action 2. Quality Assurance of Biological Measurements in the northeast Atlantic 1. To coordinate quality assurance activities on biological and chemical measurements in the Baltic marine area and report routinely on planned and ongoing ICES inter-comparison exercises, and to provide a full report on the results 1. Catch of NEA cod and haddock for 2006 Thanks to ICES for providing this information. 4.1.5 Scientific, Technical and Economic Committee for Fisheries The STECF has been producing reports since at least 1995, but the current STECF was legally created by the same European Council legislation that reauthorised the CFP in 2003. Article 33 of that document reads as follows: 114 The Paradoxes of Transparency Q 4.6 –1. A Scientific, Technical and Economic Committee for Fisheries (STECF) shall be established. The STECF shall be consulted at regular intervals on matters pertaining to the conservation and management of living aquatic resources, including biological, economic, environmental, social and technical considerations. 2. DG MARE shall take into account the advice from the STECF when presenting proposals on fisheries management under this regulation. Members of STECF are appointed for three-year terms by DG MARE. There are currently 32 members and 39 reserve members. The STECF holds a plenary meeting twice a year. The fisheries advice is reviewed, and an Annual Economic Report is produced. Part of this is an estimate of the economic impact of the current ACFM advice. When scientists attend STECF meetings, DG MARE pays for travel expenses, but they remain employed by their home organisations even though they are working on assignments put to them directly by DG MARE. Member states can be more willing to send people to STECF than ICES because they see it as closer to the decision, and work at ICES is paid for by the NFI. As mentioned above, STECF and ICES expert groups draw on the same pool of people. In recent years STECF has begun to have similar recruitment problems to those found in ICES. The Joint Research Centre, a Commission entity which is acting as the STECF Secretariat, is having increasing problems recruiting scientists for expert groups. According to a respondent at DG MARE, the problem is much more finding people than it is having the money to put them to work. Some in DG MARE do not find this continued dependence on the employees of other institutions adequate because it prevents the expansion of STECF tasks beyond what the goodwill of the NFIs and other employers allows. The need for a true in-house science capacity at DG MARE is an ongoing debate. Some scientists working at DG MARE strongly believe that DG MARE needs its own in-house fisheries science capacity. While the ICES system with STECF review may provide independence, and hence increased legitimacy, it is inefficient. A Commission employee describes their position this way: Q 4.7 This is quite a strong opinion, actually, from those who are writing these regulation proposals. They have good reasons to have that opinion; I understand perfectly where it is coming from. All these concerns about whether it is legitimate outside the house and whether it is scientifically credible in terms of being properly peer-reviewed and based on models that have been scrutinised becomes less of a concern. It is not really that the in-house advice may have less standing in civil society generally, in the public debate that is not an issue. If you are under pressure and you have to say so many days for a hundred different fleets, you have to come up with a number for the next regulation, you just need that number to come from somewhere, and as long as it is on the best possible technical basis you 115 could just consider it to be engineering rather than science, and it may be perfectly valid without having all these features that you would need to have legitimate and credible science. The relationship between ICES and STECF can be tense. One issue is that, in spite of the fact that they are competing for the same people, it has been difficult to avoid overlap and duplication. Another is feedback when changes in the advice are recommended. A respondent who works in the Advisory Programme at ICES: Q 4.8 Interviewer: Do you find DG MARE helpful partners in terms of eliminating ICES STECF overlap? Advisory Programme Leader: Yes and no. Yes, when we have face-to-face talks and we resolve issues, but not in terms of planning. I find it very difficult to understand the logic of DG MARE about when they use groups in ICES and when they set up different groups in STECF. And in general there is very little communication from STECF to ICES. So if there’s a group evaluating North Sea flatfish management plan, we don’t see it, and there’s no formal exchange of reports. One of the most fascinating things in this entire system is how essentially the same fisheries scientists seem to adapt to quite a different scientific culture when working in STECF or working in ICES. There is a story that has been repeated many times now about STECF and ACFM. It is outlined in detail in Section 6.1 because it illuminates a great deal about the workings of the science boundary in the advisory system. The reason for mentioning it here is that the way it is so often repeated, and even has become part of the cultural mythology of the advisory community, illuminates the tension that exists between STECF and ICES, even though they are usually the same people. The story is that the ACFM had refused to do mixed fishery assessments because of a lack of data on discards that they felt made the analysis impossible. After this happened, DG MARE brought several of the same scientists who had been at that ACFM meeting to a meeting of the STECF and asked them to do the analysis that ACFM had refused to do. Those scientists did provide the requested results to DG MARE. When the story was retold at ICES, it was also mentioned that some DG MARE employees had said that this outcome demonstrated that ICES should have been able to do the analysis in the first place. The ACFM members who related this anecdote were quite offended by this. One scientist got a laugh at an ACFM meeting by joking that ‘DG MARE is better than ICES because they are able to do more work with less data’. The following is from an interview with a scientist active in STECF. He describes the relationship between STECF and ICES advice in terms of the actual practice of how they receive the ICES advice and then pass it on to DG MARE. He also mentions this same story about the mixed fishery assessments: 116 The Paradoxes of Transparency Q 4.9 Interviewer: So you think, that’s how STECF is working, as a sort of an extra quality control? Respondent: Well, not so much quality control. It just adds, you know, if you look at our annual stock review that we do in STECF, in 99%, or at least 95%, we just say ‘No comment’,or‘We agree this, that or the other, but we note that’. And the only reason that there is an STECF comment is that somebody has thought something was questionable with the ACFM report and that always comes back to the member state, to the scientists on the ACFM from the member state that put in the request. I would suggest. I mean we don’t sit around and read the whole of the ACFM report in that STECF subgroup and start asking questions. We say ‘All right, do we have any questions about this advice? And if we have, let’s look at it. Interviewer: But that’s only one part of the STECF role, and the other, it has a wider mandate than that? Respondent: Yeah, but it’s actually an important part because it’s written in the STECF regulations that you have to provide an annual review of stocks. Um, it has a wider role, and it’s a consultation role, it’s a consultation committee for DG MARE. So DG MARE can ask it any issue that it wants, really, and sometimes there’s actually a bit of naughtiness on DG MARE’s part, if you like. They try to ask questions that support the answer they first thought of, and if that answer isn’t supported then they often come down and say, well, why can’t you do this? You know, a couple of years ago there was an example ... all the heavyweights came into the room and said, ‘look we need a table in your report of what comes out of this mixed fishery forecast’, and of course we all smelled a rat. And reluctantly we agreed that we would put a table in as an example, and of course as soon as it was in the report, it was used as the proposal to start negotiations. So that was a bit err, we weren’t very happy about that. The importance of these references to the table being ‘an example’that they were not ‘happy’about should not be exaggerated. STECF made some very specific distinctions about the questions they were and were not willing to answer (Section 6.1). Previously in the interview, he also said, however, that sometimes ICES is simply not pleased to have another committee second-guess its advice. He believes that STECF should be seen as another opportunity for DG MARE to get needed work done, even if this closer association sometimes provides ‘scope for people to be rather naughty’as he later puts it. Members of ACFM have been known to go back to the members of their home institutions and report that they do not agree with the result in the ACFM report, and that they intend to try to get the advice changed at STECF. Another stock assessment scientist told us that DG MARE organises STECF and related meetings on short notice ‘because they need something that is called ‘science’to be used as the basis of a decision’. He sees the less structured STECF as almost a way ‘DG MARE exploits scientists’. These are classic tensions between gaining some legitimacy while losing some saliency through increasing the distance between managers and scientists. 117 The importance of the formal distinction between in-house and independent science is debated. In the private sector in-house science has actually declined because of costs and because high uncertainty means that firms are not sure what knowledge they need (Gibbons et al. 1994). However, some have suggested that in policy matters, this is a very important variable. Clark et al. (2002) report that how much a scientific assessment is carried out by or under the control of the subsequent users is one of the top three features in determining the future use of that scientific assessment. This is a function of increased saliency. Alcock (2004) makes a similar claim in a study of the organisation of fisheries management, while simultaneously pointing out that fisheries science under control of the management agencies also raises suspicion among stakeholders. His analysis, however, conflates in-house (in his terms ‘embedded’) science with top-down management. He characterises the pre-cod collapse Canadian system as embedded, while seeing the US system as disembedded because of the existence of the Regional Management Councils. This characterisation underestimates both the relative degree to which scientists have influence over NOAA Fisheries and the degree to which NOAA Fisheries has influence over the Regional Management Councils. His suggestion that the more independent science has a higher legitimacy among stakeholders would also fail in a comparison between North America and Europe, where ICES is formally entirely independent of the Commission and where the legitimacy crisis in fisheries is particularly focussed on science (Schwach et al. 2007). The bottom line seems to be that what is really important with respect to saliency is the ease of communication between policymakers and scientists and, with respect to legitimacy, the perceptions of independence, rather than in either case the legal relationship between the two groups. 4.1.6 DG MARE DG MARE uses two main types of biological advice from ICES. These are the annual advice on TACs and the more specific special requests described above. Many of the special requests derive from derogation requests, meaning requests by member states for the special application of regulations, but many also derive from ideas about technical measures for specific fisheries. DG MARE will also sometimes request ICES advice with respect to strategic fisheries management directions. Needs for advice are identified in DG MARE, usually from someone charged with writing the regulations or with handling the annual negotiations both among member states and between the EU and its fishing neighbours. The annual advice is delineated in the Memorandum of Understanding between the DG MARE and ICES. Until recently, ICES produced advice in June and October on management of a range of different 118 The Paradoxes of Transparency stocks. This timeline is now being moved up in a process called ‘frontloading’in order to accommodate more input from stakeholders. When the advice arrives from ICES, DG MARE immediately asks STECF for an opinion, and they create a study group of approximately ten people and write a report commenting on the advice. A respondent at DG MARE described this process as ‘quite extensive, they do a thorough job, not a rubber stamp, but basically they do end up agreeing. They comment on all issues relevant to DG MARE’. Simultaneously, DG MARE is already beginning to draft measures because of time pressure. If STECF does recommend something different than what ICES did, they then make the required modifications. The final proposal is sent to the Council of Ministers. Once this proposal is made, the negotiations start. The derogation requests can become quite extensive and are usually dealt with by STECF. DG MARE is, in fact, required to respond to any citizen, so of course any communication from a member state ministry asking for a derogation must be responded to. A Commission respondent said that at any such request, ‘the machinery starts and a group in STECF is formed’. However, this is not quite as ad hoc as it sounds. The requests from each member state are bundled, and most come in the beginning of the year right after the Council of Ministers has made its final decisions. When this system was set up in 2002, the member states sent many requests that got no response from STECF beyond ‘we don’t have the data needed to evaluate this’. Without an analysis a derogation will not be granted, so more and more often the member states now provide some sort of backing for the request including a scientific rationale, which gives STECF something to review. The final decision is made by the Council of Ministers, and the decision about what will be sent on to them remains with the full-time DG MARE staff. The final political decision-making process for the CFP sits in this relationship between the Council of Ministers and DG MARE. The Council of Ministers reflects the desires of the member states, while DG MARE represents the European perspective. The Council makes its decisions by ‘majority vote’, meaning majority in the EU’s weighting system in which larger countries have more votes. The Council has the final decision, but DG MARE is not at all powerless because only DG MARE has the power to propose. The Council can only approve their proposal. So when DG MARE makes a proposal to the Council, its role has not ended. Once the proposal is made, a negotiation starts in the form of ‘if you propose this, we will agree to it’. With uncertain advice, everything becomes even more negotiable, and this creates a dilemma. The scientists at DG MARE say that they struggle with the extent to which they should try to foresee the reaction of the Council and be more precautionary. In the end, a respondent said, they usually try to give the advice based on what is seen as the most appropriate reaction to the kind of uncertainty involved. The Council generally does not consider itself bound by the precautionary approach in the sense that 119 in the judgement itself. These are the requirements of an effective ‘extended peer review’. This also implies that institutional forms are needed that are able to limit the size of the group where these trust-based mechanisms for addressing uncertainty have to take place –i.e. nested systems and the separation of different levels and types of transparency mechanisms. This is a central question in the case to which I return in the final chapter. The following exchange took place at the ACFM meeting after the discussion related above between DG MARE and ICES. It is a long excerpt, but very illustrative of the problems that they are facing and of how both quantitative advice and its qualitative context come together in trying to formulate the advice that is really relevant for how they see the actual condition of the fisheries. The discussion is about how to formulate the advice for North Sea cod in a mixed-fisheries context. The text under discussion says that the cod catch should be ‘as close to 0 as possible’, recognising that there will be discards of cod from fishing boats targeting other species. Q 4.14 Scientist One: Last year we gave strong advice because the stock was in desperate and dire state, and we wanted to prevent its commercial extinction, this kind of wording takes us back to wording that got us criticism in the past. ‘Close to 0 as possible’what is that? Now we give unequivocal advice, and we get attacked. I don’t want to be unhelpful to managers but ... I am unhappy with this kind of phrasing. We had a long discussion from last year and Scientist Two’s text is a reaction to manager feedback, but my view is that the state of the stock has not changed. Do we just bend? We need to be helpful, but do we bend to every wind? Scientist Two: No and this should not be seen as a retraction from last year or the seriousness of the situation, I agree to anything that says it is as bad as it was, but in mixed fisheries, if we say the catch should be zero, then we are saying ‘Close all demersal fisheries’. There is no way of getting around saying that mixed fisheries should prioritise clean [i.e. little bycatch of species of concern] fisheries, but there will not be 0 catches, unless you want to say close all demersal fisheries. Scientist Three: Yeah, how much do we read into changing ‘reduced catch of cod’to ‘no catch of cod’. What do we want to advise? Scientist One: If we are really serious then we say this is the advice and the caveat comes with it, if there are other reasons they decide to go ahead, fine, but we should make sure they recognise what they are doing. Scientist Two: We could have an opening statement saying the catch should be zero and all fisheries closed, then continue with this text. Scientist Four: I agree to a large extent, but it should be made conditional on the implementation of the recovery plan that would take account of the mixed fisheries.(Observer’s notes at the Advisory Committee for Fisheries Management meeting, October 2003) 126 The Paradoxes of Transparency Not only must the advice for cod within the mixed-fisheries context be communicated using text, 10 how the advice will be phrased is directly affected by both the ongoing discussions with the managers and their speculations about how the advice will be taken up and used. Scientist Four’s contribution is perhaps the most interesting as it indicates a desire to make the advice conditional on a specific set of management measures. In this particular comment, such reflexivity is perhaps especially ironic because this scientist is referring to taking ‘account of the mixed fisheries’ when DG MARE is asking for tools to do just that. Tools that, in a few days, the ACFM is going to decline to provide because of discomfort with the units of analysis required and the levels of uncertainty attached to them (Section 6.1). While this is a particularly dramatic example, comments indicating that scientists need feedback from managers are very common at both the ACFM (now ACOM) and assessment expert group levels. The following exchange addresses this question even more directly, also in terms of how to handle mixed-fishery advice and this time in the context of how to apply the precautionary approach to which ICES is publically committed. It took place at the committee in charge of the Advisory Programme: Q 4.15 Scientist One: This is how ACFM must deal with this; when they give multispecies advice, it must be consistent with the precautionary approach. I thought the interpretation is that below Blim they had to go to 0 catch ... Scientist Two: No, they have to have a recovery plan. Scientist One: But as long as there is no recovery plan in place, they have to advise 0 catch. The recovery plan is not in force now. Scientist Three: It is not ACFM that has to recommend a certain recovery plan. Scientist One: No they have to recommend the catch level and that must in some way be consistent with the precautionary approach. Scientist Four: No. Scientist One: I’m surprised you don’t think so. Scientist Three: We can ask Scientist Five about that. Scientist Four: You are right in that the reference points are from ACFM, they take the precautionary approach into account, but they don’t check against it afterwards. Scientist One: We should check how close we are to these goals. (Observer‘s notes at the Management Committee for the Advisory Programme meeting, September 2004) It is in these kinds of comments that we can see how the line between the science and management has become truly reflexive. MCAP actually wants to follow up and see how the advice is actually being used in terms of ICES’s precautionary goals, presumably to inform how future advice will be set. A lack of follow-up on what happens to ICES’s advice is a systemic problem, as discussed in Chapter 7. At one point in this October 2003 ACFM meeting, a document was presented suggesting guidelines on the use of language in the report text. The word sustainable and its relationship to the fisheries management targets, for example, were discussed at length. There was some fear that they 127 should not be tied too closely as ICES may be changing the way it defines these reference points. Another concern was that after they had gone through all this trouble to define what they mean, readers may not even bother to turn to these definitions. It was suggested, albeit factiously, that instead of using a word like ‘sustainable’in the text that a code such as ‘Concept One’be inserted instead so that people would be required to look up exactly what they meant. It is not just the ACFM that takes the text of the advice very seriously in terms of seeking to limit how the advice will be used. This is also true of scientists in the expert groups that feed into the ACFM. Consider the following exchange: Q 4.16 Scientist One: Yes, you are adding another rinkydink. We should stop pretending we know how many fish there are. Scientist Two: That is where we are going. The trend is there, but the scale is wrong. Scientist One: The system will use it at the Council of Ministers. Scientist Two: That is why I want all these caveats. (Observer’s notes at the Working Group on the Assessment of Demersal Stocks in the North Sea and Skagerrak meeting, September 2003) The text here is the mechanism that allows the scientists to continue to see themselves as doing science while having to articulate results in the midst of great uncertainty. The scientific norm of organised scepticism (Merton 1968b) retains a good deal of force. Scientists want to be either sure or silent, with the text they can at least show where they are sure about what they do not know. Not all scientists are comfortable with this use of text, and this depends on the kinds of advice being developed. Consider the following point made by a scientist at a meeting in reference to the development of advice for EAFM. Q 4.17 If you give loose, non-quantitative advice to managers, then that is what they want because it gives them permission to do what they want, so you shouldn’t do it because they get used to a casually written narrative essay. (Observer’s notes at the Consultative Committee meeting, September 2004) Text is important to others as well. One of our respondents, who is a negotiator for the fishing industry and who was one of the observers at the ACFM, put it this way: Q 4.18 Respondent: [In] the consultations between Norway and the EU for example where they use the ACFM advice, from that I know how much they look into the wording of the advice, so it was quite interesting to see how it was done. Interviewer: The wording? Do you mean the text? Respondent: Yes, the text exactly. They look into the text and say why they use this 128 The Paradoxes of Transparency word, why do they use ‘may’instead of ‘must’and is there a hidden meaning in this, they really look into it during consultations, and it was interesting to see how aware they are of the importance of how they put the words, how they make the text. So while the primary interest of DG MARE may be the numbers found in tables, when these numbers are thrown into a political process, the text sometimes does have an important impact on how the numbers will be used –at least according to one person who is both deeply involved and financially interested in the outcomes. The textual aspects of the advice confront the limits of its role as a scientific recommendation. This is nicely illustrated by the desire of the scientist in Q 4.4 to make the interpretation of the advice contingent on a particular management approach. In the traditional view of science and policy, science is supposed to be the objective other that provides the parties involved in the policy negotiations with an agreed basis for discussion. Yet here it is seen to be very difficult to produce the science without being already involved, at least to some degree, in that discussion. This provides an illustrative grounding for Jasanoff’s (2002) argument that to ‘politicise’science by making it available for public scrutiny and input can promote the interests of both science and democracy. Struggles over the science boundary apply as much to people as they do to results. When someone has been stamped a ‘scientist’, the power to designate a fact is to be used at the behest of the bureaucracy, not the individual scientist. The designation is based on employment rather than on education. In the mainstream view of the role of science in fisheries policy, introduced in Chapter 3, the role of scientists is something you are hired to do rather than something you are trained to be, though obviously the training is a prerequisite to the hiring. Indeed, one scientist told us that not all members of expert groups have university degrees. One that we talked with had been hired as a technician and received a degree in statistics after he was already a regular participant in the assessment group. Biologists working for the industry are lobbyists and negotiators, even if scientists find them easier to work with than they do fishers. Administrative bureaucracies often seek to define scientists (and themselves) as non-stakeholders. They resist seeing scientists as stakeholders because if they are stakeholders, then they bring their own values and interests to the debate, and not merely facts to be used at the discretion of others. Again in the words of DG MARE (Q 3.7, CEC 2003a, p. 15), if the scientists want ‘credibility and influence’, they must keep their ‘distance’. From the scientists’perspective, the distance may make the promise of influence look somewhat empty when, as in Q 4.4, they are trying to guess the managers’intentions. The scientist who had his ‘wrist slapped’in Chapter 3 reflects this way on the science boundary: 129 Q 4.19 Interviewer: But how do you make that distinction when sort of, the boundary between what is science and when it turns into management? It’s not always that clear cut. Respondent: It’s not always clear and …yeah, one can end up in a deep hole [laughs]. I’m always aware of it, but it isn’t always successfully avoided. But that’s also quite an exciting part of the job as well, I think. Interviewer: You like it? Respondent: Yes, I like, I like the confrontational aspects of –of science, industry, officials, and the tension that’s associated with that. I find that quite interesting. In fact, it’s the only interesting part of the job at the moment. There’s no science left. That’smy impression. Not in the role I have. As we will see in Chapter 5 the feelings and experiences that the respondent is pointing to with his exaggerated phrase ‘there’s no science left’is both widespread among fisheries scientists and more problematic for many of them than this quote would indicate. Here is another example of a Commission scientist describing the need for a feedback process in the face of the fact that a simple ‘the advice is X’ ignores the interdependence between advice, management measures, and fishing activities. The topic under discussion is how to develop a management strategy approach to management: Q 4.20 Commission Scientist: As a manager I can illustrate some points here. Baltic cod is illustrative; it is a single-species fishery. The BSFC [Baltic Sea Fisheries Commission] developed a management plan to maintain the two stocks above the Bpa, they set a harvest control rule, and an increase in biomass of 30% per year. The advice came out that this was OK, but the unreported landings were too high to make it real. At the same time we got new advice that because of this uncertainty, they refused to give us shortterm assessment. So the advice is ‘no fishing’. These are the two problems. We don’t get the science to implement it, so the plan on paper does not work. It is very clear that in addition to the SSB objective, we have to address the problem of scientific input and the availability of knowledge, a feedback process.(Observer’s notes at the Advisory Committee for Fisheries Management meeting, September 2004) In the following quote, from the same Commission Scientist in response to the October 2003 ACFM decision about multi-fishery advice, the last sentence borders on lamentation. Q 4.21 What we have now is in some ways weak advice. We say you do this, you do that. There is no data besides one table offered about interactions. It leaves managers to decide how they can define fisheries and take bycatch considerations. We have the same situation as last year. We are on our own in Brussels. (Observer’s notes at the Advisory Committee for Fisheries Management meeting, October 2003) 130 The Paradoxes of Transparency No one involved in the production of scientific advice in support of fisheries management would suggest that the drawing of the science boundary is not problematic. DG MARE sees the problem in the context of the mainstream view of the role of science in policy (Q 3.7), and this leads them to cast the problem in terms of the clarity of statements of management objectives. ‘Unclear statement of objectives’is a phrase very often heard in critiques of policy by natural scientists. DG MARE describes the problem in detail as follows: Q 4.22 One of the difficulties with much current scientific advice is that the division of labour between the scientist and the manager is sometimes confused. Some scientific advice may be based on assumptions about policy objectives that are the responsibility of the manager, with the result that the advice becomes open to question because of its policy assumptions. It is therefore important that requests for scientific advice be formulated in a way that leaves no doubt as to what assumptions the scientists are being asked to make. At least two approaches are possible. The first is for the management authority to state clearly what its management objectives are and to ‘impose’those constraints on the scientists. This approach might be followed by the Community in the case of agreement on multi annual management plans, where targets in terms as, for example, biomass, fishing mortality rate, yields or catch stability could be fixed. The second is for the management authority to request advice on different management options before deciding on which one to choose. In this case, those giving advice would be required to identify the assumptions underlying such options and to indicate the alternative strategies to be followed. Greater clarity concerning the assumptions about policy objectives will be needed (CEC 2003a, p. 13). They want to make the process tighter and the goals clearer. The science process itself remains a linear one with requests for sets of facts being responded to with the provision of facts. This makes an interesting contrast with the scientists from both ICES (Q 4.6, Q 4.8, Q 6.1, Q 6.4) and DG MARE (Q 6.4) calling for greater reflexivity, conditional advice and ongoing interactions across the science boundary. The second choice they mention, however, is very close to the idea of scenario-based participatory modelling that is an important emerging model for new kinds of scientific practice. Much like the mainstream view of the role of science, in the face of these actual experiences, the notion of clear statements of management objectives contains utopian elements. Management objectives are set by political negotiations in the face of changing environmental conditions. Science is done in a highly variable climate of uncertainty from both physical and social sources. Hence, in fisheries, objectives that are clearly stated are almost always very abstract and become unclear as soon as they begin to be operationalised. While it is important to keep trying to make 131 them as long-term as possible (e.g. greater use of HCRs), they will never stay clear for very long, and their clarity cannot be relied upon as the basis of a smoothly running linear advice system. 4.3 Conclusion Within the advisory system of the CFP, there is a marked desire for increased centralised control that is having some real impact. This can be seen in the desire for an in-house advisory system, the increased reliance on STECF, and the ongoing tightening of the European data-gathering and fisheries-monitoring systems. These changes are in response to very real needs if the CFP is going to become a more effective fisheries management system. However, the many costs involved, both financial costs and costs in legitimacy and democratic governance, simply reflect the fact that the CFP is an attempt to manage fisheries on a continental scale. To some extent this reflects the biological requirements of managing shared stocks, but this has been intensified for reasons of European politics. For very good reasons rooted in the complexity of the required information and knowledge, the global trend in natural resource management in the past 30 years has been towards making decisions at the lowest possible level. This begins with trying to leave as much discretion in the hands of individual fishers as is feasible and then moving upwards. The creation of the RACs reflects this in Europe, but the way this creation was such a babystep in the direction of the cooperative management systems used in the rest of the developed world is another reflection of how slow Europe has been in modernising its fisheries management. The tendencies towards increased centralisation, while necessary to make the CFP work, move in the opposite direction. Larger-scale systems have to rely much more on mechanisms that shortcut communications. They squeeze out the communicative rationality that makes systems sensitive to the need for change. This squeezing out of communicative rationality takes various forms in this case study, but most of the important ones have to do with giving space for reflection and review of where the science boundary needs to be set. When observing ICES, I rarely heard the ‘engineering versus science’ comparison that I encountered in DG MARE (Q 4.7). In ICES science versus advice is the much more common distinction. This indicates a large jump in understanding of both science and advice between the two institutions. The engineering rhetoric is a way to side step the question of scientific legitimacy from peer review. Engineering does not require peer review; it demonstrates its credibility when the engineered product works. DG MARE is looking at this in a very similar way, but ‘works’does not mean here, at least not in the first instance, ‘accurately characterised the state of a fish stock’. The history of the CFP indicates that ‘works’has not meant maintaining stocks of fish. ‘Works’means sufficient to allow the regulatory process to move forward. This is a very short-term perspective. 132 The Paradoxes of Transparency It is not driven by the desires that individual scientists at DG MARE have for the way the system should work, it is driven by the pressures they are under to keep the TAC Machine running. For the Advisory Programme to operate as an effective boundary organisation, even in the relatively simple problem of single-species fisheries management, it must make time and space for reflection under strong pressure from the broader system for clear and quick results. As will be discussed in Chapters 7 and 8, this is an important point where the question of social power forces its way back into what is on the whole a functionalist analysis of what is needed for an EAFM. As we will see in later chapters, the debate within ICES between science and advice is very much a parallel of this discussion; advice there is what allows the science to work for the client. Both groups are looking for Guston’s (2001b) serviceable truths. What counts as serviceable within the management bureaucracy seems to be much less dependent on the procedures of scientific legitimacy. From DG MARE’s point of view, however, this is not because they do not care how credible the number is, it is because the system that produced the advice is so large and expensive that they have no realistic way of assembling a better one. Scientific boundary work is more than simply determining what sort of facts and results will be stamped ‘science’and which will not. Also included are questions about who will play the role of science, what data the scientists will have available to them, how they will present their work; and even the ways that they will behave as they play their scientific roles are in dispute. The boundary is very vulnerable to pressures, and the scientists doing the boundary work find themselves badly squeezed. This becomes acute when uncertainty rises high because then issues of transparency of argument and methodology become entangled in issues of trust and participation. The next chapter gives a glimpse of just how squeezed these scientists feel. 133 5. Attitudes and working conditions of ICES advisory scientists With the co-authorship of Troels Jacob Hegland The focus of this chapter is the experience and attitudes of individual scientists within the fisheries advisory system. Most of the information is taken from the survey of fisheries scientists, but we have added a number of quotes from meetings and in-depth interviews where this helps give a fuller picture. One important task is to compare the experience of fisheries scientists who are more involved in the advice generation system with that of their colleagues who are less involved. Most of the tables draw comparisons between scientists who work for different kinds of employers or based on the type of the last expert group they participated in, which we consider the best single measure in the survey of participation in the advisory system. We consider this measure to be a good one, but far from perfect, as explained in Appendix 1, where the details of the survey methodology are outlined. Basic information about expert group participation is given in Table 5.1. The types of attitude scales used here are the most useful for the two tasks of comparing differences in attitude among groups and looking for correlations among the attitudes themselves. It is also reasonable to make rough statements about where on a scale a group of respondents scores, e.g. ‘well above the neutral point’ or ‘around the neutral point’, but these scales are not meant to invent and then measure precise differences among peoples’attitudes. This chapter has three parts. The first focuses on the impact of the advisory system on scientists’careers and working conditions. The second focuses on scientist’s attitudes towards the precautionary approach that frame much of how fisheries scientists see the meaning of their advisory task. The third section focuses on scientists’attitudes towards the advisory task itself. 5.1 Advice provision, career and working conditions 5.1.1 Expert group participation The simplest way to get at working conditions through a survey question is to ask: ‘Please rate your overall job satisfaction.’The results related to place 135 5.1.3 Research and publications Participation in assessment expert groups negatively influences the scientists’opportunities to publish in peer-reviewed publications. One question was: ‘How much does your job encourage or hinder you from producing the number of peer-reviewed publications you feel you would like to be producing?’The mean for all scientists is almost exactly the neutral four (Table 5.3). However, for the group of scientists who last attended an expert group that dealt with assessment, the mean is 3.25. Scientists who did not attend an expert group meeting in the last five years constitute the other extreme with a score of 4.54. The difference is even more pronounced if we look at NFI scientists alone (Table 5.4). It is not simply publications as such; publications represent the ability to remain focussed on valuable research. One scientist currently chairing an assessment expert group suggested in an interview that the amount of travelling involved in ICES and EU work meant that he could seldom ‘sit and pursue a line of research’. As many as one-quarter of the scientists indicated that they would absolutely be willing to forego career advancement in order to spend more time doing research. We say ‘absolutely’because this figure is based only on those who checked 7 on the scale of 1 to 7; if 6 is also included the figure is over 50%. Ironically, in most institutions employing scientists, it is quite normal to be promoted out of research to research administration or other administrative duties. The two first questions reported in Table 5.6 relate to how important it is to the respondent to do research rather than other tasks. NFI scientists and scientists employed in academic institutions score relatively high on their desire to do research. Several scientists added as a comment to the question that they have already passed over an advancement possibility because of the administrative duties involved. NGO scientists score lower on these questions, which is not surprising as these positions usually have no research component. In all employment categories, advancement is perceived as leading to fewer chances to do research. However, NFI scientists are more pessimistic than scientists from academic institutions and NGO scientists. This corresponds to the over-representation of NFI scientists among those mentioning administration in the open-ended question on changes (Table 5.7). The most interesting information in Table 5.6 is, however, the distance between the scientists’emphasis on research and the possibilities of doing that later in their career –the distance between wishes and expectations. NFI scientists show the biggest difference. NGO scientists end up with a very small difference, based mainly on the lesser desire to do research. The relative match between wishes and expectations is, no doubt, part of the reason for the high job satisfaction that NGO scientists reportedly enjoy (Table 5.2). 142 The Paradoxes of Transparency Table 5.4 Means of scales of NFI scientists’perceptions of working conditions by type of expert group Not attended expert group meeting in last five years Last expert group did not deal with assessments Last expert group did deal with assessments Total Mean N P Please rate your level of overall job satisfaction. 1 = I am not at all satisfied, 7 = I am very well satisfied Mean 5.08 5.46*** 4.78*** 5.11 208 0.01 N378388 In the past three years how has the pressure you experience on the job changed? Only those with same job title for last three years included. 1 = decreased a great deal, 7 = increased a great deal Mean 5.14 5.18 5.44 5.3 169 0.31 N286180 Approximately how many days did you spend on job-related travel in 2004? Mean 41 41 50** 45 207 0.1 N368388 How do you feel about having to travel this much? 1 = I would enjoy travelling more, 7 = this is far too much travel Mean 3.64*** 4.24 4.42* 4.21 206 0.01 N368288 How much does your job encourage you or hinder you from producing the number of peer-reviewed publications you feel you would like to be producing? 1 = severely hinders, 7 = strongly encourages Mean 4.69*** 4.02 3.23*** 3.81 205 0 N488386 * indicates significance at 0.10, ** indicates significance at 0.05, *** indicates significance at 0.01. Asterisk indicating significance refers to differences between the category and the other two categories combined. Excluded in all questions are 20 who did indicate participating in an expert group but did not make clear which kind of group and therefore did not fit in any of the three categories, six who failed to answer whether they had been in expert group or not, 21 who did not identify their employer, and 209 who did not work in NFIs. Furthermore, one did not rate job satisfaction; two did not rate the changing pressure, and 38 did not confirm holding the same job title for last three years; two did not indicate how much travelling they did in 2004; three did not indicate what they felt about the amount of travelling; and four did not rate the employer’s support for peer-reviewed publications. 143 5.1.4 Funding and administration One of the things that the Mode Two approach emphasises is the relationship between the way science is funded and how it is carried out, especially in relationship to quality control. One of the most important changes in fisheries science in recent years has been the change towards reliance on ‘soft’money that scientists or institutes have to apply for. This has affected working conditions in several ways: Q 5.5 When we first came here ... in those days the government used to give us a pot of money and say, ‘here’s your pot of money. Do what you like with it, as long as you keep us happy.’Um, internally we would then compete for shares of that money to do interesting biological studies that would actually help. The scientist indicates that the current funding arrangements are not leading to better research, rather the contrary. This particular scientist suggests that scientists had previously been able to decide for themselves and produce better –or at least more interesting –research results instead of being caught in fulfilling contractual demands. Table 5.5 Type of last expert group by seniority Type of last expert group Seniority Definition Non-assessment group Assessment group Total N Row % Col % Row % Col % Very senior PhD before 1986 or MSc before 1984 68 32 69 33 20 Senior PhD 1986-2001 or MSc 1984-1999 54 46 135 51 58 Junior PhD after 2001 or MSc after 1999 50 50 46 16 22 Total N 143 107 250 Relationship is significant at .09. Excluded in all questions are 148 who did not attend any expert group, 20 who did indicate participating in an expert group but did not make clear which kind of group, six who failed to answer if they had been in an expert group or not, and 41 who did not report the year of their terminal degree. 144 The Paradoxes of Transparency Table 5.6 Means of scales of perceptions of research and career by type of employer NFIs Acad. NGO EU Other private Other gov. Total Mean P How important is it to you that much of your time is spent doing research rather than advocacy, administration or management? 1 = not very important, 7 = very important 5.62* 5.96*** 3.30*** 5.42 4.83*** 4.41*** 5.49 0 Would you be willing to forego career advancement in order to spend more time doing research? 1 = no never, 7 = yes, definitely 5.09 5.37** 3.78** 4.97 4.65* 5 5.06 0.05 In the job you have now, how does advancement affect your chances to do research that you think is valuable? 1 = far fewer chances, 7 = many more chances 2.74*** 3.52*** 3.2 2.91 3.26 2.56 3.01 0 Difference between a scale based on first two questions and the third question -2.61*** -2.14 -0.22*** -2.24 -1.45*** -2.22 -2.27 0 Total N of difference = 418 211 103 9 31 48 16 * indicates significance at 0.10, ** indicates significance at 0.05, *** indicates significance at 0.01. Asterisk indicating significance refers to differences between the category and all other categories combined. Excluded to reach N of 418: 22 respondents did not identify their employer; five did not rate importance of research vis-àvis other tasks; ten did not rate willingness to forego career advancement; and ten did not rate effect of advancement in relation to research. A reliability analysis between the two first scales yielded a Cronbach’s alpha of .634. 145 The scientists in our survey were asked to give a short description of the most important changes they have experienced in their work activities since assuming their current position. This was an open-ended question, meaning that no answers were suggested. As it turned out, answers to this question fell into two categories. One group of scientists discussed concrete changes in fisheries research issues, for instance the increasing focus on ecosystems. The other group discussed changes in their working conditions. Two issues recur in many of the answers: funding sources and administration. The answers of several respondents included both, for example: Q 5.6 A lot of pressure (and time spent) to secure funding for research and less time to actually do research. More bureaucracy, meetings and management responsibilities. To get an idea about any interesting differences between those mentioning funding and/or administration and those who did not, the answers to this question were coded so that they could be compared with other questions in the survey database. Answers were placed in two categories with respect to both funding and administration. For funding, answers that referred to changes in funding sources were placed in one category, while answers that did not mention funding or only mentioned general changes in funding levels were placed in the other. For administration, answers mentioning management, administration, bureaucracy, or meeting activity were placed in one category, and all other answers were placed in the other. Two coders were used as a check against bias (Table 5.7). Sixty scientists mentioned changes in funding sources. This is a considerable number given that the responses were the result of a completely open-ended question about general changes. Several mentioned that the need to write applications for funding puts additional pressure on them. Some also argued, as in Q 5.5, that the changes in funding sources have had implications for the type of research carried out. One respondent argued as an example that the ‘decrease in central funding and the need to seek outside funding [is] limiting the opportunity to undertake basic research’. Differences in the perspective on funding relate also to the type of employer. Table 5.7 shows that 24% of the scientists in academic organisations mentioned this change, but only 14% of the scientists in the other groups combined. Another question asked how much the scientists’employers were encouraging them to participate in externally funded research. The answers revealed significant differences across the categories of employers. The overall mean was 5.7 on a scale from 1 to 7, which suggests fairly strong encouragement. NFI scientists were close to the overall mean, whereas scientists from academic institutions gave a significantly higher rating (Table 5.2). 146 The Paradoxes of Transparency Scientists employed in academic institutions are also the second highest employment group in experiencing increased pressure on the job (Table 5.2). It is likely that this feeling of increased pressure is related to the stronger emphasis on external funding. One academic scientist said with respect of funding that he felt a ‘pressure to do it, but no other tasks are taken away to permit this’. The survey data also show that those who mention sources of funding as an important change also feel a greater increase in work pressure (5.82 v 5.38/7 P = 0). NGO scientists feel the least pressure from their employer to obtain external research funding. This may mean that the pressure on them is to find funding for advocacy, or that NGOs tend not to use their scientific staff for fundraising because it is carried out by others. Table 5.7 Percentages mentioning issues of funding and administration by type of employer Subjects mentioned without prompting in replies to the following open-ended question: ‘Please give a short description of the most important changes you have experienced in your work activities since assuming your current position’ NFIs Not NFIs Acad. Not acad. N Administration as important change Mentions issue 39 26 117 Does not mention issue 61 74 242 Funding sources as important change Mentions issue 24 14 60 Does not mention issue 76 86 299 N 183 176 82 277 Total N 359 Rounded percentages. Excluded: 22 did not identify their employer and 84 did not describe the most important changes. NFIs against all others combined gave a Chi-Square of .01 for administration. Academia against all others combined gave a Chi-Square of .03 for funding. The open-ended question on most important changes in work activities also generated many answers, almost one-third of the total, mentioning administration along the lines of Q 5.6. Scientists employed in NFIs mention these issues significantly more often than others do (Table 5.7). Some 39% of NFI scientists mention administration as opposed to 26% of all other groups combined. One scientist offered the observation that there is a‘much greater emphasis on administration and monitoring of work targets with no improvement in work output’. Administrative pressures are to some extent related to pressures from the industry, the EU and other elements. In particular, assessment scientists have to be increasingly aware of how they formulate their advice, how they present uncertainties, that their assessments are consistent from spe147 cies to species, etc. As discussed in Chapter 7, these issues are having a tremendous impact on how ICES is organising itself, certainly with respect to the traditional advisory structures, but increasingly in terms of changes in funding structures and administrative loads. These pressures are having a number of unintended, negative consequences on the lives of these scientists. Table 5.8 Patterns in expert group assignments by gender All respondents Type of last expert group Total P No WG Non-assessment WG Assessment WG Men 76% 84% 80% 80% 0.15 Women 24% 18% 20% 20% Total N 148 167 123 438 Employees of National Fisheries Institutes Men 68% 86% 82% 81% 0.08 Women 32% 14% 18% 19% Total N 38 83 87 208 Excluded in all questions are 20 who did indicate participating in an expert group but did not make clear which kind of group, six who failed to answer if they had been in an expert group or not, and one who did not indicate gender. The second question also excludes 21 who did not identify their employer, and 209 who did not work in NFIs. 5.1.5 Gender Fisheries science in support of fisheries management has noticeable gender patterns, and some female fisheries scientists have related experiences of discomfort in professional situations because of their gender. Compared to many other sciences, the proportion of female scientists is low. Participants in all of the meetings we observed were largely middle-aged white men. Only 20% of our survey respondents are women, and this distribution holds fairly evenly across all employment categories. Women’s numbers within fisheries science are increasing. Among the most junior third, the respondents who received their last degree within the past seven years, 32% are women. Among the most senior third, respondents who received their last degree 17 or more years ago, only 5% are women. Women rate their job satisfaction slightly lower than men do at 5.09/7 compared with 5.36 (p = .09). A weak pattern is visible in the assignment of men and women to types of expert groups (Table 5.8). Women are overrepresented among those who do not attend expert groups and under-represented among those attending the more desirable non-assessment expert groups. Assignments to assessment expert groups show no gender 148 The Paradoxes of Transparency pattern. This pattern is the same, but less pronounced, among the more junior third. It is statistically significant at 0.10 only among NFI employees (Table 5.8). The overall picture seems to be a slowly changing working environment where full gender integration is being approached, but still not reached. 5.2 The precautionary approach Most fisheries scientists understand their advisory task as using the best available science to provide advice for implementing a precautionary approach to fisheries management. The precautionary approach is not itself a scientific concept, but rather a framework for the delivery of science. It is not merely an official doctrine. In expert group and other meetings there are constant references to the precautionary approach as the guiding principle for making judgements in uncertain situations. As discussed in Chapter 3, in a way similar to the EAFM, it is a very important ‘scientific ideology’in fisheries science in Europe. Table 5.9 Responses to attitude scales about the precautionary approach by type of employer Type of employer NFIs Academia NGO EU Other private Other gov. All P It is critical that fisheries management be risk-averse and chooses lower fishing pressure when stock condition is uncertain. Strongly disagree = 1 ... 7 = strongly agree Mean 5.71 5.83 6.3 5.3 5.44 5.00* 5.66 0.12 N 220 105 10 33 54 17 439 To what degree should judgements made in preparing scientific advice be influenced by the precautionary approach? never = 1 ... 7 = always Mean 5.94 6.21*** 6.60* 5.39*** 5.49*** 6 5.92 0 N 219 104 10 33 55 17 438 * indicates significance at 0.1, *** indicates significance at 0.01. Asterisk indicating significance refers to category compared to all other categories combined. Excluded in both questions are 22 who did not identify their employer. Furthermore, four did not answer the question on risk-adverse fisheries management, and five did not answer the question on judgements. The precautionary principle as a general norm in environmental management came to the fore through the Rio Declaration (UN 1992). The United 149 Nations Conference on Straddling Fish Stocks and Highly Migratory Fish Stocks (UN 1995) first articulated the principle for fisheries with the following definition: Q 5.7 States shall be more cautious when information is uncertain, unreliable or inadequate. The absence of adequate scientific information shall not be used as a reason for postponing or failing to take conservation and management measures (UN 1995, p. 6). The precautionary principle is the official doctrine of the Common Fisheries Policy, and indeed of the EU. The following is from the Council Regulation dealing with the 2002 reform of the Common Fisheries Policy: Q 5.8 (3) Given that many fish stocks continue to decline, the Common Fisheries Policy should be improved to ensure the long-term viability of the fisheries sector through sustainable exploitation of living aquatic resources based on sound scientific advice and on the precautionary approach, which is based on the same considerations as the precautionary principle referred to in Article 174 of the Treaty (CEC 2002, p. 59). The treaty referred to here is the treaty establishing the European Community. The precautionary approach is how the precautionary principle becomes operational. ICES mentions greater emphasis on the precautionary approach in the discussion of its basic mission in its strategic plan, which was formally adopted by the contracting parties in 2002: Q 5.9 The Mission statement is noteworthy in terms of the evolution of ICES. Marine ecosystems are inclusive of fisheries, but much broader and more complex. The emphasis on marine ecosystems does not diminish the priority that ICES will give to fisheries. Advice on fisheries will continue to be a prominent part of the ICES programme, with an increased application of the precautionary approach and within a wider ecosystem context (ICES 2002). Beyond simply doctrine, it is widely supported among fisheries scientists. In the US research, 80% of the marine scientists we surveyed agreed with the statement, ‘It is critical that fisheries management be risk-averse and choose lower fishing pressure when stock condition is uncertain’, while 44% agreed strongly (Wilson et al. 2002). Table 5.9 lays out the relationship between two statements about the precautionary approach and the type of employer. No relationship was found between these scales and type of expert group. The first statement is a restatement of the precautionary principle, and it gets high agreement across all categories, the lowest being scientists working in the ‘other government’category. The other statement related directly to the question of the degree to which precaution should influence judgements in preparing 150 The Paradoxes of Transparency advice. Scientists working for the European Commission still scored above the centre of the scale on this question, but below all other categories. The highest scorers were the NGO scientists, six of whom chose 7, with the other four choosing 6. Even with this kind of general support for the precautionary approach, real disagreements and confusions exist about how it should be implemented. The following quote from ACFM is typical of many such discussions about advice and is a good illustration of both the basics that the scientists agree on and the pressures and temptations that they struggle with in the implementation of precaution: Q 5.10 Scientist One: Would it be better to highlight that the estimate is close to Blim, and whether it is above is to a large extent how you do the model, and we can’t really tell. We can shortcut the discussion and say that apparently we are very close and it is not an either or. Scientist Two: So we just say that Blim is very sensitive to the assumptions made in the forecast and a 40% reduction would rebuild to Blim. It would be very difficult to say we think this would do the job in accordance with the precautionary approach. We have tried to do the best we could with the management plan and now we are sliding on to saying yes as if these numbers are certain. The precautionary approach is exactly that, we are not in a position to say the stock will recover [General agreement around the table]. (Observer’s notes at the Advisory Committee for Fisheries Management meeting, October 2004) A conceptual framework has been developed and used for the inclusion of the precautionary approach in Europe that identifies ‘precautionary reference points’. The basic idea is that stock assessment models are used to identify Blim, which is the lowest level a spawning stock biomass should ever be allowed to get, and Bpa, which is the larger spawning stock biomass used as a target in order to reduce the chance that a stock is ever fished down to Blim. This conceptual framework is what Degnbol (2003) describes as stochastic predictability: Q 5.11 ... because [in the implementation of the precautionary approach] the basic concept of predictability was maintained, but the predictions of the effects of management measures were expanded to include an estimate of the associated uncertainty (2003, p. 40). He argues that while the precautionary approach is about accepting the fact of uncertainty, this approach treats it as a supplementary consideration within the standard and traditional framework of stock assessment models and their associated management techniques. The fundamental question of the conditions of predictability themselves were not examined. This is an excellent example of Shackley and Wynne’s (1996) transformation of uncertainty. 151 Q 7.62 Interviewer: So what kind of more change would you like to see? Delegate: I think that we have now changed the advisory process, we have decided that we should adopt the ecosystem-based approach, but still, we have the thematic expert groups, and then we have clients or customers, use whatever word you like, who ask for advice on a single-stock basis. So, what I think we should do is that we should have in a way regionalised the ICES advisory process, not the science process. I think the science process needs to be thematic, but the advisory process should be organised in a kind of regionalised way, that in a way causes the process itself to use a broader, more ecosystem-based approach. Interviewer: What kind of regions would you be looking at? Delegate: I would prefer the LME [Large Marine Ecosystem] level. Some scientists also saw the combination of the three advisory committees into the one ACOM as a way of forcing an EAFM agenda: Q 7.63 Scientist One: As long as ACE and ACME existed, if we got a nonfisheries advice request, they had another place to send it, this was because they could not get them into the fish groups. From many years I know that when fish stock assessment working groups were assigned things that did not include SSB and F, these things did not get done. Now they have no other place to send them. Now it will be sent to people who do fisheries advice, and we will put people in the room who know about the bycatch species –like birds for example –when you have a client that is mandated to protect these species. If we give catch option advice ignoring other requests to ICES, that catch option advice is irresponsible advice, not merely not integrated. (Observer’s notes at the Annual Meeting of Assessment Working Group Chairs, February 2008) It is an interesting twist in this statement that the speaker is thinking of advice to clients based on what those clients are ‘mandated’to do, rather than more directly in terms of the client’s specific requests. The question of the EAFM and client requests also appears in the next quote. One of the interviewed delegates, the same one who considers regionalisation so important in quote Q 7.62, is not so optimistic about how effective a regional approach can be because client requests arise from the TAC Machine: Q 7.64 Interviewer: I’m interested in a point you made a little while ago, where you said that you should have worked more on ecosystem approach, but …the rationale for ACOM is to bring the three advisory committees together so that it becomes more integrated. So, how was that a failure? Delegate: Unless you change the process, the underlying process, which means [the way] you compose the actual working groups that work out the data to do the advice …Interviewer: I was hoping they would be doing that on a regional level, like you’ve described. Delegate: No, but they won’t…it has never been really discussed. The advice groups will meet to compose 254 The Paradoxes of Transparency advice about some kind of request …they will produce advice for the North Sea herring, the Norwegian pout …etc. …because that’s what they are being asked for. It is important to note here that the reorganisation adopted by the Council allows, but does not mandate, the organisation of review and advice groups by regions. It is ACOM that will determine how regionalised ICES’s work will become, and there are real disagreements over the desirability of this because of consistency issues, as discussed above in the part of Section 7.3.1 that discusses ‘ensuring consistency’. A basic question this delegate raises, however, concerns the group level where integration takes place. This question is linked to regionalisation because ACOM will decide the degree to which expert groups, review groups and advice drafting groups are tied to regions. This question was the most common way in which integrated advice and the EAFM arose within the reorganisation debate. It is also a question that some scientists feel has been left hanging: Q 7.65 Scientist One: It seems like the integration stuff is always so vague, and the responsibility seems to fall on the assessment working groups. (Observer’s notes at the Annual Meeting of Assessment Working Group Chairs, February 2008) The group level at which integration will take place is critical. Many ICES scientists believe that real integration will never happen unless it takes place before the advice drafting group level and certainly before ACOM. Given the breadth of ACOM’s mandate, there is little additional expertise available there for advice drafting, their job is to make sure that the advice is fit for its purpose. Even at the level below ACOM, if an advice drafting group finds itself looking at nothing but a set of distinct, independent assessments of different issues within some region, then they will be badly handicapped in trying to produce integrated advice, especially in the few days they have available to them. In the terms of the next quote, they will not have time to integrate advice; they will only have time to concatenate it. For many ICES scientists, integrating information for EAFM advice has to begin at as low a level as possible. One ICES leader explained it this way: Q 7.66 What the EAFM does is make one discuss a lot of different issues, almost all of which would be discussed anyway somewhere in ICES, but discuss them in one place and relative to each other. It is the people who can talk to all the different specialists and see how the individual parts interrelate to each other that are in short supply. That breadth of perspective has to be present in the review process, because if advice is to be integrated, then expert group reports need to be reviewed in two ways. One is relative to the specialised disciplinary content, which is the traditional and easier part of the review. The other is when this expert product is to be linked 255 with others, the reviewers have to ask questions about how correct it would be to interpret and apply the results in ways that might not have been on the minds of the specialists who created the work being reviewed. And of course, the drafting groups have to have those people with the breadth of experience to understand several different sets of specialist results well enough to frame decent integrative views of them, and they tie the pieces together in a whole. Otherwise you don’t get ‘integrated advice’, you get ‘concatenated advice’. Indeed, this ICES leader gave this as the most important reason why he supported keeping the advice and review groups together. Given the limited number of people who are able to handle the breadth of the specific issues that arise with respect to the EAFM, ICES cannot expect to be able to move advice along four separate levels and have it be truly integrated. Integrated advice should begin at the expert group level, and then be reviewed and carried forward to advice by the limited expertise that is actually available, and which is unlikely to be able to staff both a review and an advice drafting group. 7.4 Conclusion Two more quotes provide illustrations of what I think are the main conclusions that can be drawn from this discussion of the restructuring of ICES. The first asks and answers a very interesting and basic question: Q 7.67 Scientist One: Why do we come to ICES? If we had a choice, would we choose ICES? Scientist Two: Here we sit in a big community and we are happy, then we are home and have another reality? How important is it? Scientist Three: It is important that the programmes come in and are more integrated and cross-disciplinary, ecosystem-based, this is why we want ICES. Scientist Two: It can do things we can’t do alone ... Scientist Four: One argument is that it is more than just Europe. Scientist Five: We do need ICES, but we need more people, and we are being pulled into a number of organisations, ICES has to work more efficiently with other organisations. (Observer’s notes at the Resource Management Committee meeting, September 2007) These scientists did not reflect at all on ICES as an intergovernmental provider of official advice. They know that is why it is there –it is the reason why they do not ‘have a choice’–but it is not their motivation. ICES provides a network that makes it possible for these scientists to achieve their goals, particularly with respect to the EAFM. Yet, their desires for ICES reflect some of its contradictions. They want the network, but they want it to work more efficiently, and it is under-resourced. They see in the ICES system a source of some of the frustrations described in Chapter 5. The 256 The Paradoxes of Transparency next quote also addresses the question of the network, but it is the voice of someone who is directly responsible for the intergovernmental aspects of ICES: Q 7.68 Delegate: In my opinion, it’s the distributed networks that gives ICES the authority to speak. Interviewer: OK. Delegate: It is as simple as that, because what you get is that you get every scientist to promote their views, their results, their findings, and then they have a kind of scientific consensus process that automatically feeds in the authority for ICES to speak. Because everybody has been participating, everybody has the opportunity to say their ideas, and, if there is agreement it is allowed to show it ... And the strength is they are very distributed ... In my opinion, ICES authority is based on the scientific activity in its distributed groups. If it had not been for the number of distributed groups, you wouldn’t have the authority, because then there would have been the possibility that many would rest on the outside saying: Hey, you didn’t hear me! This delegate is arguing that the network is the source of ICES’s authority because it makes criticism of results possible. No one is silenced, and this is what makes the authority of ICES’s voice so strong. The network provides a social context that allows scientific process legitimacy in the sense described in Section 3.1.3. It allows both the scientific credibility and the ‘something more’(3.1.3) that gives the credibility of the scientific process legitimacy in the eyes of its beholders. He is describing something that is a bit contradictory: a distributed network that speaks with one voice! Yet, I think there is an important insight here, even if it seems paradoxical. What he is describing is what ICES, which often seems to have two faces, actually does do. This paradoxical structure has some great gifts, as will be discussed in the final chapter when I try to pull this case together. However, it often seems to mean that ICES just continues in a state of constant reform. The final decision made by the Council forced ACOM to separate the advice and review groups, but left other critical decisions in ACOM’s hands. These include the ToRs of the expert groups, which will in turn determine how low in the group hierarchy integration begins to be considered, and the degree to which the review and drafting groups will reflect ecosystem regions. The creation of ACOM is certainly one step towards organising the EAFM, but it cannot be said that any of the main issues are really resolved at this point. They are in the hands of the new structure. 257 8 Conclusion 8.1 A case of adaptive learning In spite of the fact that the system of marine management that ICES feeds into has not been a successful one, I believe that this case study is a substantially positive example, and that ICES has lessons to offer to other institutions seeking to produce knowledge that can support an ecosystem approach. Of course, I cannot claim that ICES has been ‘successful’in a general sense because the EAFM is only really beginning in earnest, and ICES plays one set of roles among many in determining policy outcomes. However, I do think that ICES has approached the EAFM in a way that is an example of ‘double-loop’learning at an impressive magnitude. Doubleloop learning is the idea that institutions not only learn and change in response to that learning, but that they also learn about how they learn. ICES has attempted to mobilise a vast array of expertise to meet a complex problem. For an intergovernmental organisation of this size and complexity, it has made an impressive start. What is it about ICES that has made this adaptive learning possible? That seems to me to be the question around which to fashion a concluding chapter. Asking the question this way is not meant to ignore ICES’s problems and constraints. It is to do quite the opposite. The fact that ICES is part of a bureaucratic management system which has often made the lives of its scientists miserable while failing to sustain fish stocks is precisely what makes the question interesting. My response to this question can be summarised this way: a. the ICES network contains a number of ‘creative tensions’that provide space for ongoing reflection; b. these creative tensions are very costly from the perspective of a largescale bureaucratic system attempting to manage fisheries, and finally an EAFM, so there is a lot of pressure to ‘resolve’them by suppressing or ignoring them; c. the ICES network, however, is made up of so many distributed power centres in constant negotiation that the creative tensions are very difficult to suppress; and d. out of these constant negotiations new mechanisms for transparency are being developed that are subject to the paradoxes of transparency 259 and require tightly drawn social spaces and associated languages to resolve the paradoxes. This concluding chapter describes these four points. After doing so, it turns to a short reflection on the usefulness of the CST theoretical perspective. Then the final discussion focuses on the implications of the lessons from this case study for the institutions that evaluate, select and maintain policy responses within an EAFM. The central recommendation is to structure institutions generating and using knowledge for an EAFM by ordering them according to larger and smaller communicative scales, based on the complexity of the mechanisms needed to ensure transparency. I briefly describe this in terms of a results-based EAFM approach. 8.2 ICES’s creative tensions The major form taken by the creative tensions within ICES is the contrast between the Advisory Programme and Science Programme cultures. For a number of reasons, this contrast should not be overdrawn. For one thing all ICES scientists draw on a general scientific culture that is much more pervasive and stronger than any sub-culture within ICES. The two scientific ideologies that have repeatedly arisen in this case study –the EAFM and the precautionary principle –vary somewhat in levels of commitment (Table 6.1), but they are still very generally accepted. Furthermore, for any of the tensions I describe below, there are far more than two different positions. In fact, given that these are cultural ideas, and we are talking about scientists here, it is likely that ICES contains 1600 positions on each one. Nevertheless, I am satisfied that the distinction between the cultures of the science and advice sides of ICES is a useful one. It does describe fairly well the main poles of these tensions. Most importantly perhaps, the distinction is not my creation; ICES scientists commonly use it themselves. 8.2.1 Placing the science boundary One tension between the science and advice sides is contrasting attitudes about the placement of the science boundary. This distinction became very clear in the discussions of how ICES should approach scientific review. ‘Review’remains the key word in describing mechanisms for deciding what will and will not be declared science; but ideas about what review means vary. On the science side, an ideal model of review, derived from the more general notion of ‘peer-reviewed science’, retains substantial force. Peer review requires trained experts in the subject of the review, and it should be independent, not carried out by anyone who was involved in the work that is being reviewed or otherwise with any sort of stake in the outcome. It 260 The Paradoxes of Transparency should be substantially divorced from the advisory process and focussed only on the factuality of the description of nature on which the advice should be based (Q 7.32). On the advice side, practical experience with the questions of credibility, legitimacy and saliency of advice has made them much more open to an extended peer community (Funtowicz and Ravetz 1990) understanding of review in the development of advice. Illustrations of how practical issues have this effect is found in discussions of how to explain advice to managers (Q 4.11, Q 4.12 and Q 6.1) and even in suggestions that advice should be made conditional on management actions (Q 4.14). A philosophy has emerged among some advisory scientists, reflecting the idea that developing advice, including reviewing the outcomes of expert groups, should be based as much on general scientific literacy and skills as it is on specific technical expertise. Review should be rooted in experience with advice, rather than in very specific forms of substantive expertise, i.e. ‘less technical review and more overall review of quality’(Q 7.31, see also Q 7.30). This is not to say that scientists on the advisory side downgrade the value of technical review or reject the ideal model of review in principle. But there is recognition that it is not enough, and ‘the more’that is needed may not fit the ideal model of review or involve the retention of all scientific decisions in the hands of ICES scientists. The discussion of the skills of review and advice, especially the contrasting of technical expertise with the scientific literacy needed to write advice, provides some insight into the practical meaning of an extended peer community. The emphasis in Q 7.30 is on the ability to distinguish the degree of underlying uncertainty in the results. The extended peer community needs exactly this kind of expertise; judgements about uncertainty require the ability to judge scientific methods and background knowledge about the sources of the data. Judging the degree of uncertainty in a practical sense also requires judgements about the implications of that uncertainty for relating the results to the practices to be regulated. 8.2.2 Scientific activities within ICES A second creative tension between the advisory and science cultures concerns scientific activities within ICES. This is mainly played out in issues over the role and function of expert groups. A large part of this is simply the degree to which scientists are used to putting their particular interests aside and responding to a specific set of advice-driven ToRs. These are differences about what sorts of questions are really worth the time of a group of scientists when they go to an international meeting to produce some product. These difficulties here are nicely visible in Q 6.29 where the chairman, who was used to working on the advice side, was perfectly ready to ignore the surface temperatures that everyone –himself included –agreed were the key driving factor in the relationships being examined. 261 Most of the other scientists wanted to work with their ‘pet data sets’simply because they addressed the most important factor. The chair knew that these temperatures –key factor or not –had no saliency for policy, and the group’s time was better spent on things that actually mattered. When considering an EAFM, the question of how expert group time will be spent is not trivial. Perhaps the main organisational challenge for ICES in contributing to an EAFM is mobilising expertise that has not previously been part of the mainly fisheries-driven advisory process. The expert group culture on the science side is much more used to doing what is interesting to them, rather than what is interesting to a client. ICES’s leverage over the way ToRs will be emphasised, or even addressed, is limited. The scientists who attend these groups tend to be somewhat more senior than those who attend the assessment groups (Table 5.5), and this also strengthens the ‘we are volunteers at ICES’attitude. While the ICES leadership may sometimes feel like they are herding cats, it is not entirely a bad thing that scientists in expert groups follow their own interests. These interests are about understanding ecosystem processes, and the inquiries cannot be limited to results with immediate use to policy setting. 8.2.3 Adequate Science for Advice Another tension exists around ideas of what constitutes adequate science for advice. This is the tension that I am the least comfortable characterising as being between the advisory and science sides. While one side of the debate, those advocating ‘soft predictability’, tends to be from the advisory side, strong advocates of having quantitative forecasts as an important goal are also found on the advisory side. Managers want quantitative advice because solving the political problems they face starts with having a divisible quantity. This is not at all the same thing as the scientific motivation for quantification, which is about precision for the clear statement of hypotheses and potential replicability. Nevertheless, the high value that general scientific culture gives to quantification is drawn upon in support of the need for quantitative advice. This is evident, for example, in the disparaging references to a ‘word game’and ‘intellectual essays’in Q 6.44. This is broadly evident in the overall vision of developing integrated models for EAFM. Scientists also see quantification as a mechanism for increasing their own control over policy, as is illustrated for example in Q 4.17. An internal connection exists between the desire for quantification and ICES’s concern with ‘consistency’. Within ICES the various types of consistency are what allow the different parts of the network –including the broader network of clients and ministries –to negotiate and hold each other accountable for the results. They are mechanisms of transparency forged in the face of the paradoxes. This case study has identified three 262 The Paradoxes of Transparency types of consistency –Types C, D and E –that most directly link the social and technical aspects of advice. Of them, Type E is the most sociological in nature and attempts to keep track of the coherence of a particularly complex set of decisions. Type E consistency is consistency in the way scientific methods and scientific advice are linked across time, space and species. As argued in Section 7.3.1, this kind of consistency has very little to do with science as such, in fact the questions it asks are not particularly coherent from a purely scientific perspective. For Type E consistency, quantification plays still another distinct and critical role. It names the methodologies in their linkages to advice functions. The failure of Type E consistency in the herring incident (Section 7.3.2), for example, was identifiable through the equations used in the spreadsheet. To use a bit of STS jargon, the science/ advice relationship is ‘inscribed’by the methods of quantification, and therefore answers to questions about Type E consistency will consist of the comparison of equations and measurements. Textual information in most cases will prove too imprecise to allow evaluation. Quantification here is a tracking mechanism for ensuring the transparency of managerial rather than scientific decisions. It is a form of accounting in both senses of the term. The other pole of this axis, soft predictability, is a child of uncertainty. Soft predictability describes scientifically plausible scenarios while steering away from precise numerical forecasts about future states of nature. It has emerged from the experience of the advisory side that humility is a virtue when one tries to base advice on predictions about the future. Soft predictability recognises that many ecosystem processes are going to have to be understood qualitatively, based on categorical variables and indicators. It asks that focus be placed on the substance of the issues people care about, in a manner that is not defined by the degree to which the issues are amenable to measurement and modelling. Soft predictability also means that decisions are going to be made using judgements that will be difficult to document objectively in the sense that other scenarios may be equally plausible. This means that the transparency needed for collective action cannot be achieved through methodological rigour, not even for technical experts. Soft predictability interferes with Type E consistency as every issue has to be worked out on its own. It will be very difficult to meet the constant political demand that the links between methods and advice reflect similar judgements. This suggests that we should begin to think of Type E consistency as a requirement for consistency of social practices and forms of interaction related to advice production, rather than as a requirement for consistency in mathematical or technical procedures. Qualitative inference requires a participatory approach to translate categories into advice because there are no other ways to get the needed transparency. Hence, to some of the advisory scientists soft predictability is not a loss of quality, it is a way to improve quality by moving away from reliance on predictions, putting more emphasis on uncertainty, being more 263 tists to be the observers are really in a position to benefit even from this formal transparency. What is really needed, but which is costly from a bureaucratic perspective, is experimentation with ways that scientists can help stakeholders develop transparency with each other, with the ultimate goal of agreement about what is found in nature. Formal transparency does not come anywhere near tapping into the real power of transparency in accounting for how one knows what one knows. This will require other techniques within an extended peer community, based on practices of scientific facilitation towards the creation of serviceable consensuses on facts within a resultsbased EAFM framework (Section 8.6). The need for this is evident in the strong desire that scientists have expressed for more interactive, advice-producing fora. Building interactions into the science for advice mobilises transparency at a level that can make a difference. One indication is the desire in Q 6.1 to move beyond directly using models to produce numbers that serve political requirements to building political options into the models as alternatives. This is an improvement on the less transparent habit of putting conservative assumptions into models in the name of the precautionary approach. It implies acknowledging and moving away from the ‘hiding of political values within models’approach. Scale plays an important role here as well. One pattern that has been noted in effective, cooperative approaches to marine management is the importance of cross-scale institutional linkages (Wilson et al. 2005; Degnbol and Wilson 2008). An example would be when an NGO, concerned with a narrow set of issues across a broad geographical area, makes an alliance with a local government concerned with a broad set of issues in a small area. Cross-scale linkages are critical, because they team groups with agendas that operate on various scale levels. This is going to become even more critical as we move towards an EAFM. Cross-scale linkages often make possible the small-scale interactive fora in which communicative rationality can take place, even when addressing problems at the large-scale level. These activities bring the scale of interaction down to a level where the paradoxes of transparency can be worked through in a way that avoids or reduces gaming. This is beginning to happen more frequently in Europe (Hegland and Wilson 2008). When scientists become involved, activities focus on creating boundary objects that link science and policy. The scientists function as facilitators in science-based decision-making processes. An area where this is taking important form, addressed in other publications (Hegland and Wilson 2008), is participatory modelling using scenario-based approaches that place the uncertainty at the centre of the negotiation. The bottom line is that formal procedures to achieve transparency directly, whether through rigorous application of scientific ideals or through formal observer programmes, often encounter the paradoxes of transparency and end up making things more obscure, while they may be good in 270 The Paradoxes of Transparency and of themselves. The real power of transparency is achieved when scientists address questions of the science boundary directly within an extended peer community. 8.4 The benefits of distributed power This case confirms and perhaps expands our understanding of the importance of the polycentric network for science in support of policy discussed in Section 3.4. Both Ostrom (2001) and Cash and Clark (2001) emphasise the information-processing aspects of polycentrism. Ostrom discusses how polycentrism facilitates accessing various knowledge sources, identifying feedback potentials and experimenting with different approaches. Cash and Clark point to an improved capacity to provide coherence across scale levels while still allowing local specialisation. The ICES case certainly confirms these points, but it also focuses our attention on the importance of the distribution of authority and control of resources. Ostrom (2001) characterises polycentric approaches as hard to govern, and the frustration of the ICES leadership with the simultaneously voluntary and non-voluntary character of participation in ICES activities is a good example. ICES is an intergovernmental organisation, a professional association and a loose network of scientists, and this multiple character is a critical aspect of how it functions. To various degrees, depending on the urgency of the issues to be addressed, the individual’s seniority, and the attitude of his or her employer, the ICES scientist is both constrained to participate in an ICES expert group and has to be convinced to do so. The scientists within the ICES network are related to various centres of resource control, usually cooperating but often contending. Through the delegates, the NFIs and the relevant ministries have formal control of ICES, but the scientists in the ICES network can tap into many sources of funds for their activities. While the Advisory Programme is formally independent of the Commission, it remains to some degree dependent on it for funds. When the scientists are operating through STECF, they are more responsive, but not fully controlled in that STECF expert groups are temporary structures made up of NFI employees who are also part of ICES. Along the advice production chain some activities are funded by the clients who want the advice and some are funded by the NFIs. Again the relationship is neither completely voluntary nor completely constrained. Both the clients and the NFIs, to a variable but growing degree, have to respond to pressures from both user groups and conservationists that are channelled through their ministries. The European Commission is not just a client; along with the national ministries, it is also a source of funding for a multitude of marine-science research projects and focussed tenders. Many of the NFIs very strongly encourage their scientists to compete for these projects so that they can get a substantial part of their salary paid by them. Many of these projects 271 are long-term and exploratory, while others are just a small step away from advice provision. Who comes to ICES expert groups on the science side, and what they do when they get there, are strongly influenced by these projects. One of the strengths of the widely distributed ICES network, with its voluntary yet constrained working relationships, is that it forces negotiations between its various power centres that must make use of communicative rationality, and this creates room for reflection. The gamut of negotiations and discussions required in the reorganisation of the advisory system, and the number and complexity of the issues that were discussed, make this very clear. The reference in the latter part of Q 7.32, that the ICES ‘volunteers’cannot be ordered to make a piece of advice, is a fine illustration of how this system works to resist the systemic pressures described in the previous section. To address an EAFM, ICES needs to be able to tap into an extremely wide range of expertise. The scientists in Q 7.67 see ICES as a network that enables the projects and activities where integrated, cross-disciplinary work takes places and makes it possible to develop an EAFM. ICES is developing a way to mobilise the network to make EAFM advice possible by the shifting-focus strategy that will take individual issues one at a time and push them as far as they can towards preparation for advice. This will mean both anticipating and understanding the implications of ecological events and moving integrated models as far as feasible. The strategy will lead to knowledge about ecological linkages, and sometimes even integration, being developed by the main experts in a form that can be applied by the less expert but still scientifically literate. The shifting-focus strategy is designed to enable a network (and eventually an extended peer community will be required) that can respond to the complexities of adapting to ecosystem changes. It will require resources and a huge commitment from many experts from many disciplines. As it seeks to respond to the EAFM challenge, ICES benefits from its complex power distribution. The requirements for constant negotiation slow the influence of systemic pressures within the large management bureaucracy that threaten to squeeze out communicative rationality. Such communications are costly and difficult to structure formally, hence the squeezing, but they are needed if science for an EAFM is to be possible. This implies that the cooperation needed from across the network to develop an EAFM cannot be forced, making the motivation of ICES scientists critical. These scientists are seeing many former assumptions about what it means to be a scientist and to do science questioned. They are tired of being asked questions they cannot answer, of being asked to create certainty, and of spending long hours in activities they do not see as science. Some are beginning to see and act on new interactive styles of science where they use their transparency skills with clients and stakeholders to help them design realistic management strategies. These scientists are committed to a precautionary approach to marine management, and the 272 The Paradoxes of Transparency EAFM promises vehicles for achieving that in ways that deal in a much more authentic fashion with uncertainty. It is a good bet that their motivation and creativity will allow the development of institutions that can create and maintain the knowledge needed for an adaptive, ecosystem-based approach to marine management. 8.5 Theoretical ruminations: CST and STS A central theoretical goal of this book was to explore how Communicative Systems Theory (CST) could make a contribution to Science and Technology Studies (STS). CST suggests the possibility of a meaningful approach to understanding hybrid natural and social phenomena that has a systematic place for an analytic distinction between nature and society. Trying to understand human society as a system adapting to its environment requires finding a way to define system boundaries, and defining society as a meaning-based communicative system separate from its material surroundings does that. Understanding adaptation, in my opinion, begins with examining science as an institution and how science is linked to possibilities for collective action. The vehicle for the linkage between CST and STS is the use of the Habermasian notion of rational communication. This is a sociological usage of the idea, not the much more common philosophical one. Communicative rationality is what allows people to make sense to one another day-today. It never reaches the ‘ideal communicative situation’of the philosophical system, nor does it need to, rather it describes criteria people actually use to judge if the communication they are engaged in is leading to a mutual understanding. In my interpretation, communicative rationality also includes Habermas’s (1984) arguments about the communicative interpretation of the objective, social and inner worlds. This is the idea that reaching mutual understandings takes different forms and follows different orientations in respect of those worlds, so that discussions of fact must be oriented towards a consensus about what is true, while discussions of values and interests must be oriented towards an agreement about what is fair. Science, in trying to achieve the radical transparency that is the ideal goal of the scientific method, formalises and protects rational communication. Habermasian systems theory is built around this concept, and in this linkage lies the promise that CST can contribute to STS. CST focuses on the quality of communicative systems within a political context. Rational communication’s internal connection to the scientific method makes it possible to directly relate science to a system concept of society. Therefore, one aspect of this study that I hope has demonstrated the utility of CST for STS is the analysis of scale that is pulled together in Section 8.3, based on the theoretical perspective laid out in Section 3.3.4. Scale is a funny concept. On the one hand, we know a great deal about it, at least judging from the amount of literature. On the other hand, a great deal of 273 this writing boils down to an acknowledgement that scale is really important. Science for an EAFM is confronted with problems stemming from both the social and institutional scale. I think CST provides some new tools for understanding scale and institutions that are relevant for science studies. One aspect involves examining how communication mechanisms effective on larger scales (Section 3.3.4) create systemic pressures to inflate the science boundary. These pressures complicate the delicate negotiations around boundary work, boundary objects, and building effective boundary organisations. Section 3.1.3 uses communicative rationality to illuminate the relationship between scientific credibility and process legitimacy. The argument is that process legitimacy results from applying the same set of basic principles that scientific credibility is based on –i.e. communicative rationality expressed as the radical transparency of the scientific method –to the broader social context in which the scientific activity takes place. I would argue that the basic point of the extended peer review in a situation of high stakes and high uncertainty is to provide legitimacy to the ‘serviceable truth’when the uncertainty is too high to establish credibility. If this is reasonable, then this link between credibility and legitimacy based on rational communication sets up criteria that extended peer review should meet to achieve that goal: that there is no manipulation involved in the communication, that anyone involved can raise a question about any claim being made (White 1988), and that discussions of facts are oriented around finding a consensus about truth while discussions of values are oriented around finding a fair compromise. ICES has not formally adopted the idea of an ‘extended peer review’, even while the Advisory Programme has increasingly moved in this direction in practice. This informality has its helpful aspects, but it pays a price in not delivering the increased legitimacy that a more formal extended peer review would. Currently, most extended peer review involves the clients and is more oriented around increasing saliency than legitimacy; although this is shifting with the increasing importance of the RACs. ICES’s main effort to increase legitimacy through extended participation in scientific activities has been the ‘transparency through observers’. This is a good idea, but its effectiveness is severely curtailed by the paradoxes of transparency. Developing a set of rules for extended peer review grounded in rational communication would be a helpful addendum to this strategy. One result of the application of CST that I did not plan was the idea of the ‘creative tension’(Section 8.2). The paradoxes of transparency were not something I had in mind when I began the study of ICES. The central role played by questions of transparency within the case forced me to think about the role that transparency and accountability played from a CST perspective. What I have concluded is that transparency and accountability act as integrating devices that allow institutions to achieve the simultaneous need to structure both competition and coordination in situations where mechanisms for constraining behaviour are insufficient. Transparency 274 The Paradoxes of Transparency and accountability are aspects of communicative rationality, i.e. they are institutional forms that guard the ability to raise claims. They are ways of structuring situations that make a mutual understanding possible. Within ICES the disagreements that emerge between the science and advisory sides, in some respects the product of an ideological competition, are transformed into creative tensions through a combination of mechanisms of accountability and fairly wide distribution of power within the network (Section 8.4). The concept of creative tensions, which I first heard working as a community organiser, strikes me as a useful one as a guide for institutional design. A related idea is the various types of consistency that emerged from the study. Types C, D and E consistency are hybrid concepts that help clarify how co-production is happening in ICES. What is interesting to me is how type E consistency in particular expresses a new set of skill-based tacit understandings of review. Although it is not commonly understood within ICES, as befits an emerging tacit understanding, the review criterion of type E consistency is aimed at legitimacy rather than credibility. It is particularly well suited to, in fact in some sense it demands, an extended peer review. In Section 8.6 below I apply this idea to EAFM institutions. I would expect that the main ideas developed here could prove useful in the analysis of other areas of developing science for policy and even in other aspects of STS. Understanding the development of other extended peer communities may be enhanced through the linkage of the credibility and legitimacy of science in terms of rational communication. The analysis of the types of consistency is very similar to a research agenda on ‘commensuration’that is already underway in other areas of social life (Espeland and Stevens 1998). The CST approach to social systems and adaptation should be able to help clarify the links between the environment, bureaucratic or market-based management, and science in other contexts. 8.6 Living with the paradoxes: A results-based EAFM framework When working to resolve the paradoxes of transparency, the term that ICES scientists use most frequently is consistency. The forms of consistency are mechanisms that they are creating to make transparency and accountability possible among constantly negotiating power centres dealing with very complex issues. These centres are found both within ICES and among stakeholders and clients, so an extended peer community is involved that includes different kinds of expertise. These mechanisms are, I think, a bit more than simply the kinds of language developments one would expect in any sub-culture, for example, the creation of jargon words that allow people to save time when talking to people familiar with a field. All of the different uses of the term ‘consistency’are about ways of making 275 things comparable. They are a language of accountancy even when they are non-numeric. They are all tools that allow the different power centres within the network to account to one another for the decisions, i.e. the science/advice boundary judgements, they are making. They are critical to the function of the science-policy network. To a degree, they reflect tacit knowledge, meaning that knowledge about how to use the forms of consistency is not fully articulated. This can be seen in the way the same word ‘consistency’is used to express several different and very important bases for judgement. They rely on the kind of scientific literacy we were introduced to in Q 7.30 that is required when scientists must make judgements about matters where they are not technical experts. They must get a sense of how certain or uncertain the knowledge is, and in order to do this they look, more or less consciously, at how settled the knowledge is in its application. This implies a judgement about how consistently it is applied. Once again attempts to create transparency lead to a new paradox. Accounting for scientific judgements depends on the development of new kinds of mechanisms for comparison that contain elements of tacit knowledge. Within the scientific network charged with ensuring that our decisions about the ecosystem rely on the most transparent knowledge possible, we are required to trust a sub-culture that is at times difficult to describe, even for its initiates. We have already recognised (Barnes et al. 1996) how dependent the scientific community is on trust. Now, as we seek to expand the transparency of the scientific community to include the translation of science into advice and action, we find that the mechanisms of transparency, while not limited necessarily to scientists or dependent on specific forms of technical expertise, require a good deal of initiation to make them work. Transparency is always limited so knowledge institutions have to rely on a degree of trust. Dealing with uncertainty further demands the development of trust, because procedures of transparency are harder to bring to bear. This is part of what must be understood with ‘extended peer review’. There is a strange relationship between transparency and trust. They can and must substitute for one another when one of them fails, but when both are present, they reinforce and strengthen one another. The experience at ICES is that ensuring transparency in a complex environment creates a demand for new mechanisms of comparison that are themselves complex and must be developed through experience and slowly articulated through processes of reflection. Once more we are brought back to scale. Because of the complexity of the issues, the size of the group of people who have access to the languages required for transparency will necessarily be limited. This is social scale I am referring to here, not necessarily geographical scale, and these groups would often be created through geographical cross-scale linkages. I would argue that an EAFM will require a nested results-based system, organised around both sets of economic activities and geographical areas. 276 The Paradoxes of Transparency The science to support these institutions will require the conscious attainment of Type E consistency, achieved through consistent procedures and styles of practice, rather than attempts to create a purely technical consistency in forms of advice. At a minimum, three levels would be involved, more might be required in practice, although this would not be desirable. We can think of them as communicative “spheres”arranged one inside the other like Russian dolls. Each sphere would have a complex interior characterised by skill-based mechanisms of transparency, such as shared ideas about consistency rooted in a context-specific scientific literacy. These complex interiors would be encased, inside and out, i.e., facing both the sphere further out and the sphere further in, by simple shells. An example might be a group of managers, scientists, fishers, tourism representatives, NGOs and other stakeholders who are concerned with implementing an EAFM on a shared regional sea. The outer sphere would be a public process of limit-setting that would create publicly sanctioned limits on a series of potential impacts to protect ecosystem integrity. The middle sphere would again include many of the same stakeholders, for example recreational and commercial fishers, managers, scientists and NGOs, who are concerned with a bay (it could as easily be a set of mixed fisheries or other sub-set of issues), and who are charged with translating the broader public limits into operational limits on fishing in their bay. This might be limits both on catch and on certain types of habitat impacts such as bottom trawling or numbers of speedboats. This group would form a subculture which does not share specific expertise, but does share the kind of scientific literacy about their bay that allows internal transparency and the reasoned assessment of uncertainty. Once this group has translated the limits, the inner sphere, a smaller group of e.g. scientists and recreational and commercial fishers would develop scenarios for catching and sharing those fish within the constraints imposed by habitat considerations. Again, from an interior perspective this would be a sub-culture that shares considerable scientific literacy about the specific set of problems they need to address. The limit-setting process in the outer sphere would be dealing with both environmental and social complexity, and they would be accountable in limit-setting both to protect ecosystem functions and to allow as much economic activity as possible given the constraints of protecting ecosystem functions. But the product would need to be relatively simple: they would need to set up indicators of compliance for the middle sphere. Two kinds of simple indictors would be required that would form the simple outer shell of the first sphere of decision making. The first kind of simple indicator would be a set of social indicators that would need to be developed that the middle sphere would have to meet to show that it is made up of people who fully balance the relevant interests. This is the absolutely critical requirement that the paradoxes of transparency create for any EAFM. It is just as important as the technical indicators. What may not be obvious here is that it is in the outer sphere that 277 addressing an EAFM as a social dilemma is most required. It is very difficult for local interests, which are focussed so much on their own immediate problems, to be open to raising and dealing with new issues. This openness must be enforced from above. The second kind of simple indicator produced by the outer sphere is, of course, a set of simple ecological indicators that demonstrates that the decisions that the middle sphere is making are sustainable. The indicators will have to be based on some sort of integrated model of the ecosystem; the question of environmental events would be addressed in the middle sphere because such events happen too rapidly for effective response at the higher scales. These models must be very simple to get the job done. This will often mean that they address broad processes. Many of the more aggregate properties ecosystems are more tractable than individual properties. Total production, for example, correlates with many community properties, while trying to predict what happens to individual stocks is much more difficult. The United States has, for example, set a cap on total biomass removals from the Gulf of Alaska. What seem to be needed to set limits on exploitation are processes that are broad and simple enough so that models can provide useful forecasts of the implications of changes. Fisheries scientists are also developing a set of simple indicators with very general applications, based on the size of the fish (Hall et al. 2006; Pope et al. 2006). In the middle sphere, the need for inner complexity and outer simplicity repeats itself. Limits are going to have to be clearly expressed. The processes that set those limits involve a sub-culture that is scientifically literate in dealing with the complexity of for example their bay. These groups are going to have to address environmental events and relate them to implementation of limits on activities. The impossibility of full transparency again requires that interests be balanced and that the limits themselves are clear. Finally, in the inner sphere operational plans are created to meet the limits set by the middle sphere. The critical issue will be the burden of proof. Those who wish to pursue the economic or recreational activity will be responsible for ensuring the transparency that will allow them to be held accountable for staying within those limits. These ideas are not entirely utopian; several existing institutional models approach them. The Marine Stewardship Council, for example, has created a broad set of indictors for sustainable fishing and uses scientist certifiers to work in detail with fishing fleets to decide how the indicators can be fairly measured and met in their particular situation. The complex details of ecosystem interactions are handled by a fairly small group, and provisions are made for any interested party to get familiar with the issues and comment on them. This model for handling information is a very good one, independent of the form that the external accountability takes. A government licensing programme could use this approach just as well as a private ecolabelling scheme. 278 The Paradoxes of Transparency The ICES network is uniquely positioned to contribute to such a multilevel approach to knowledge institutions for an EAFM. The distributed network is able to handle the development of many different kinds of scientific groupings involving cross-scale linkages. They are well ahead, for example, of the EU in moving in this direction. The RAC expert groups are the beginning of a structure for organising interested parties, but they are still quite aggregated. ICES scientists have already shown leadership in moving fisheries management in Europe in a more responsive direction. They are also open to addressing new problems working with groups of stakeholders. To create an adapting social system, we must balance complexity and simplicity to allow as much transparency as possible, recognising that mechanisms for transparency create their own obscurities. The ICES case indicates that to achieve this we need to learn to trust and to pay the costs of a wide distribution of decision-making power in creating and using science in an ecosystem-based approach. The hypothesis it suggests is that negotiations among multiple power centres create the space for open, rational discussion, without which a social system can never sense the need for change. 279 Habermas, J. 1984 The Theory of Communicative Action: Volume I Reason and the Rationalization of Society. Boston: Beacon Press. 1987 The Theory of Communicative Action: Volume II Lifeworld and System: A Critique of Functionalist Reasoning. Boston: Beacon Press. 1990 Moral Consciousness and Communicative Action. Cambridge: MIT Press. Habermas, J. and T.H. Nielsen 1990 Jurgen Habermas: Morality, Society, and Ethics: an Interview with Torben Hviid Nielsen. Acta Sociolgia 33:93-114. Haeuber, R. 1996 Setting the Environmental Policy Agenda: The Case of Ecosystem Management. Natural Resources Journal 36(Winter):1-28. Halffman, W. 2003 Boundaries of Regulatory Science: Eco/texicology and Aquatic Hazards of Chemicals in the US, England and the Netherlands. Dissertation, Science Dynamics, University of Amsterdam. 2005 Science Policy Boundaries: National Styles. Science and Public Policy 32 (6):457-467. Halpern, B.S. and R.R. Warner 2003 Matching marine reserve design to reserve objectives. Proc. R. Soc. Lond. B 270:1871-1878. Hardin, G. 1968 The Tragedy of the Commons. Science, 162:1243-1248. Harremoes, P., D. Gee, M. MacGarvin, A. Stirling, J. Keys, B. Wynne, and S.G. Vaz 2001 Late lessons from early warnings: the precautionary principle 1896-2000. European Environmental Agency: Environmental issue report No 22. Harwood, J. and K. Stokes 2003 Coping with uncertainty in ecological advice: lessons from fisheries. TRENDS in Ecology and Evolution 18(12):617-622. Healy, S. 1999 Extended peer communities and the ascendance of post-normal politics. Futures 31:655-669. Hegland, T.J. and D.C. Wilson 2008 Participatory Modelling in EU Fisheries Management: Western Horse Mackerel and the Pelagic RAC. Deliverable 7B of Scientific Advice for Fisheries Management at Multiple Scales (SAFMAMS), EU 6th Framework Specific Support Action No 013639. Maritime Studies (2009) under review. Hilborn, R. 2006 Faith-based Fisheries. Fisheries 31:554–555. Hilton-Taylor, C. 2000 IUCN Red List of Threatened Species. IUCN, Gland, Switzerland and Cambridge, UK. 286 The Paradoxes of Transparency [Document text truncated for crawler view.]