Establishing Credibility: Practising Standard-Setting Ideals in a Swedish Seafood-Labelling Case

Journal of Environmental Policy & Planning Vol. 8, No. 2, June 2006, 135 158 Establishing Credibility: Practising Standard-Setting Ideals in a Swedish Seafood-Labelling Case MAGNUS BOSTRÖM Stockholm Centre for Organizational Research, Stockholm University, Sweden ABSTRACT Market-based non-state governance arrangements, many examples of which are seen in the environmental field, require the active approval of a broad group of stakeholders. This paper makes the theoretical argument that credibility is a key issue in the establishment of such arrangements, and examines empirically the effort to develop a trustworthy eco-labelling scheme for seafood in Sweden. Many policy actors view eco-labelling as a particularly credible instrument that consumers and businesspeople can use to demonstrate environmentally friendly behaviour. But establishing credibility is complicated, especially if the issues are controversial and if there is mistrust among the groups. This paper analyses the challenges involved in practising six standard-setting ideals, the fulfilment of which is seen to establish credibility: inclusiveness, independence, auditability, scientific validity, global applicability and the balancing of feasibility versus environmental stringency. The ideals are subjects of framing, debating, power struggles and negotiation; and are dependent upon context, situational and historical factors. The assumed positive relationship between ideals and credibility is complicated because of the challenges involved in practising the ideals. This article draws upon the literature on non-state authority, governance and standardization. KEY WORDS: Eco-labelling, environmental governance, environmental management, private authority, standardization, credibility, fishing politics, trust Introduction Market-based non-state governance arrangements, many examples of which are seen in the environmental field (Jordan et al., 2004), require the active approval of a broad group of stakeholders. This paper makes the theoretical argument that credibility is a key issue in establishing such arrangements; and examines empirically the effort to develop a trustworthy eco-labelling scheme for seafood in Sweden. Eco-labels are markers which are presented to consumers or professional buyers to indicate environmentally beneficial consumer choices. The popularity of such tools for dealing with environmental problems has increased Correspondence Address: Magnus Boström, Score (Stockholm Centre for Organizational Research), Stockholm University, SE-106 91 Stockholm, Sweden. Fax: þ46 8 16 49 08; Tel.: þ46 8 67 47 419/þ46 70 58 46 033; Email: magnus.bostrom@score.su.se 1523-908X Print=1522-7200 Online/06=020135-24 # 2006 Taylor & Francis DOI: 10.1080=15239080600772126

136 M. Boström significantly over the last decade in Northern and Western Europe (e.g. Boström, 2003a; Cashore et al., 2004a; Jordan et al., 2003; Micheletti, 2003). Many policy actors view eco-labelling as a credible instrument that consumers and businesses can use to demonstrate environmentally friendly behaviour. Eco-labels identify the greenest buying options. Such categorical statements must usually be qualified. The self-declarations of corporations are seldom considered to be trustworthy and are often seen as pure green-washing. The eco-labelling strategy is presented as superior, partly because well-recognized environmental organizations (EOs), other experts and business players jointly participate in setting the standard principles, criteria and indicators that comprise the eco-labelling scheme (e.g. Bendell, 2000). Establishing credibility is anything but a straightforward process, however. Debates about eco-labelling in general, about the specifics of particular ecolabelling projects, and about other standard-setting processes reveal a plethora of grand standard-setting ideals such as inclusiveness, transparency, openness, auditability, accountability, independence, global application and scientific validity (cf. Cashore et al., 2004b; Gulbrandsen, 2004; Jordan et al., 2003; 2004; Lipschutz & Fogel, 2002). Eco-labellers hope to establish credibility by referring to such ideals. Practising and fulfilling them may be difficult, however. For example, to involve many actors in the labelling process itself a standardsetting ideal that establishes credibility may lead to many conflicting views about substantive and procedural issues, including conflict over definitions, interpretations and practices of the ideals themselves. This paper analyses how actors strive to establish a credible eco-labelling scheme by trying to negotiate and practice certain standard-setting ideals. A case has been chosen that began with bitter controversies and mutual mistrust a case that demonstrates more clearly the challenges and dilemmas involved in the effort to establish credibility. Empirically it refers to the long, drawn-out process of creating an eco-label for seafood in Sweden designed to promote sustainably managed fisheries. The efforts began in 1996 when the World Wide Fund for Nature (WWF) took the initiative to create an eco-label by proposing the Marine Stewardship Council (MSC) model. But state authorities and other key Nordic actors rejected the MSC model and initiated their own labelling project. Discussions and debates continued within informal groups, but not until 2002 did labelling advocates initiate a more efficient process that assembled the main Swedish stakeholders. The project was run by the Association for Control of Organic Production (KRAV), a well-known Swedish eco-labelling organization and non-governmental organization (NGO). Two years later, KRAV s board decided to accept a final standard proposal. The label is now in operation, although only a few fishing vessels are certified. Many intensive debates took place between 1996 and 2004. This paper analyses the possibilities and difficulties of incorporating the six ideals that were the implicit or explicit subject of debate:. inclusiveness and independence: how to allow for inclusiveness and independence against a background of initial controversies and mutual mistrust among interest groups and the asymmetric power structure of the fish sector;. auditability: how to establish a trustworthy auditing function, given the significant and general mistrust directed at the fishing industry;

Establishing Credibility: Swedish Seafood-Labelling 137. scientific validity: how to link the eco-labelling system to sound scientific advice, despite disagreements among claim-makers about the situation at sea;. global applicability: how to create a local market-based regulatory instrument in a global pro-free-trade context;. feasibility versus environmental stringency: how to maintain a balance between these often-conflicting demands. It must be borne in mind that more is said in this paper about efforts to establish credibility than can be said about the extent to which credibility is achieved. It is too early to assess, for instance, how consumers are responding to the recently introduced eco-labelling scheme. The next part of the paper presents the method, followed by the literature on non-state authority, governance and standardization, in order to theorize about establishing credibility through certain standard-setting ideals in eco-labelling and similar arrangements. A historical overview is then provided of the efforts to initiate seafood labelling in Sweden and the arrangements involved in the incorporation of standard-setting ideals in the eco-labelling project are analysed. The paper ends with a concluding section. Method The case study is based on various documents and on interviews with 15 key persons. 1 The selection of interviewees and text material encompassed a broad array of diverging experiences, attitudes, viewpoints and arguments. Some interviewees worked at KRAV and others represented key organizations in the project and its network. 2 All interviewees played a role in the labelling process, but they were not equally committed to the project. As the interviewees had extensive experience dealing with seafood issues, they were able to provide valuable expert knowledge about various details. More importantly, they were asked to give an adequate and balanced picture of the general views, attitudes, understandings and conflicts surrounding fishery and labelling issues in the organization they represented and in the professional network to which they belonged. Accordingly, the interviewees had to be well aware of their own organizational setting and network. A specific interview guide was designed for each interview because the participants had various backgrounds and sets of knowledge, and played different roles in different phases of the process. There was, however, a standard set of questions that were asked in all interviews (e.g. questions around conflicts and disagreements and their resolution, the topics to be emphasized or treated as secondary in the discussions, questions about the interviewees views of the roles and actions of other parties). The interviews lasted between 60 and 150 minutes and were conducted between November 2002 and June 2004. Because of the length of time between the first and last interview, one person was interviewed twice, and additional questions, developed later in the interview process, were sent via e-mail to some interviewees. Documents have been an equally important data source in this study. Particularly critical have been various KRAV documents, 3 such as memory notes from meetings, two documents sent to stakeholders for consideration (so-called remiss), and stakeholders replies to these documents (remissvar). In December 2002 the first remiss with principles and standard criteria was published and sent to stakeholders, followed by a number of hearings and meetings and a second remiss. These documents have provided a valuable resource for sorting

138 M. Boström through the main topics in the debate and for analysing the actors arguments. Some interview questions were designed to clarify certain positions expressed in the documents. These documents allowed the author to add to his analysis the arguments and viewpoints from stakeholders other than those interviewed. 4 The selection of six standard-setting ideals partly follows an inductive logic. Ideals reflecting arguments, meanings and concerns that appeared central to the debates have been focused upon. However, the ideals are also conceptual constructs that have been developed by reference to other governance, non-state authority and standardization literature, and which in turn have helped aid understanding of the debates. Credibility and Standard-Setting Ideals The concept of legitimacy has been frequently used in the recent literature on governance and non-state authority (Cashore et al., 2004a; Cutler et al., 1999; Hall & Biersteker, 2002; Mörth, 2004; Rosenau, 2003; Tamm Hallström, 2004). New ways of governing and setting rules for societies, organizations and individuals challenge conventional methods of governance and regulation. Regulatory innovators struggle to legitimize their own standards. Any type of governance relies on legitimacy (cf. Weber, 1919/1977), but voluntary non-state governance and market-based arrangements rely on approval from targeted audiences; because they, by definition, lack traditional enforcement capacities associated with the sovereign state (Bernstein & Cashore, 2004; Cashore et al., 2004a). Although the concept of legitimacy can include passive and apathetic consent, such market-based and consumer-orientated instruments as labelling require more active approval. The concept of credibility is especially relevant when considering non-state horizontal relationships and organized collaboration among different types of organizations (Boström, 2006). Credibility can be defined as the perception and assumption that the operations of an actor or agent are trustworthy, responsible, desirable and appropriate (Boström, 2006; cf. Cashore et al., 2004a) and, similar to power, authority and legitimacy, it is a relational rather than a possessional concept (cf. Rosenau, 2003, p. 274). The credibility of an actor s operations must be understood in relation to the perceptions of all relevant stakeholders or audiences (cf. Cashore et al., 2004a; Huckel, 2005). Relevant stakeholders can be actors with a direct interest in the policies and procedures of the standard-setting bodies or actors that are less directly influenced by the policies, but have an important role in approving the standard-setting body. Both groups include state, business and civil society actors, all of whom may develop distinctly different yardsticks from which they assess the credibility of a standard-setting body. Several factors can be seen as essential for establishing credibility or legitimacy. One common distinction in the academic literature on democracy and legitimacy is the difference between input legitimacy (e.g. representation, election processes, transparency, accountability, decision-making procedures, deliberation) and output legitimacy (e.g. problem-solving capacity, effectiveness, outcome and output) (Beisheim & Dingwerth, 2005; Huckel, 2005; Lindvert, 2006; Rothstein, 1992). Similarly, there is a distinction between procedural and substantial sources of legitimacy. Relating to all these aspects are such standard-setting ideals as inclusiveness, representation, participation, openness, transparency,

Establishing Credibility: Swedish Seafood-Labelling 139 accountability, auditability, precaution, human rights, sustainability, expert orientation, voluntary participation, deliberation, scientific validity, global application, feasibility and substantial stringency (e.g. Bernstein & Cashore, 2004; Boli & Thomas, 1999; Braithwaite & Drahos, 2000; Garsten & Lindh de Montoya, 2006; Rosenau, 2003; Tamm Hallström, 2004). Such values, principles, or ideals are believed to establish authority, legitimacy, or credibility, first because they are seen as intrinsically positive. Secondly, scholars and policy actors believe that practising such values, principles, or ideals will lead to better outcomes in other words, there is an assumed positive relationship between input legitimacy and output legitimacy or between procedure and substance. Democracy, for example, can be seen as instrumental for effectively dealing with complex environmental problems (Lafferty & Meadowcroft, 1996). Thirdly, a closely related function of ideals is as a tool to guide action (e.g. the precautionary principle) or to enable dialogue among disagreeing interest groups. Sustainable development, for instance, is often thought of as having these mobilizing qualities (e.g. Hajer, 1995). The practise of ideals and principles may not be without problems, however. They may conflict with each other, as seen when the principle of expertise clashes with the principle of representativeness (Tamm Hallström, 2004). And ideals such as sustainability, precaution, accountability and transparency may not be sufficiently concrete to guide action because they are essentially multi-dimensional, ambiguous, and flexible; and may provide weak tools for overcoming power asymmetry. Resourceful actors may deliberately refer to ideals such as transparency in an attempt to create an impression of transparency rather than real transparency, because it involves decisions, selections and, inevitably, judgements about what should be revealed to the public and what should remain hidden (Klintman & Boström, 2006; Thedvall, 2006). An additional problem, less discussed in the literature but the focus here, is the assumed positive relationship between ideal and credibility, which can be far from clear-cut, precisely because of the challenges involved in practising the ideals. This paper focuses on six standard-setting ideals: inclusiveness, independence, auditability, scientific validity, global applicability and the balance between feasibility and environmental stringency. Although they certainly do not constitute an exhaustive list, they were found to be recurring themes in the eco-labelling debates. And implicitly or explicitly they have played a considerable role in this particular case, reflecting, as they do, many of the chief procedural and substantial concerns in the debates. Taken together, they should provide a representative picture of the credibility issues confronting eco-labellers. The purpose here is not to present all possible standard-setting ideals, but to analyse and discuss the problematic nature of the relationships among ideal, practice and credibility in eco-labelling. Inclusiveness Inclusiveness may be seen as an intrinsically good quality, connected to democratic values and representativeness. Inclusiveness can also be seen as instrumental for the mobilization of dispersed resources that are believed to be essential for the operations of the standard-setting body. Those who are included may have power resources such as finances, specific expertise, or symbolic resources that bring added value to the eco-labelling project (see Boström, 2006, for an extended

140 M. Boström analysis). In a market-based governance system, for instance, it is essential for the standard setter to include business actors with central network positions that can direct powerful demands along the entire chain of production and distribution (Cashore et al., 2004a, 2004b; Green et al., 2000; Gulbrandsen, 2004; Jordan et al., 2004). The participation of social movement organizations (SMOs) such as EOs or animal rights groups can bring credibility to the project with their moral authority (Hall & Biersteker, 2002) or symbolic power resources (reflecting collectively shared values, voluntarism, consumer power, for instance) (cf. Boli & Thomas, 1999), and they can be sources of new framings and alternative expertise (Boström, 2003a, 2004a). State actors can assist a non-state-driven project by making it compatible with a wider national or transnational regulatory framework. They can provide practical help in the form of funding, expertise, the adjustment of rules, or auditing; and symbolic help such as approval or authorization. Although a non-state, market-driven governance scheme does not rely on states, and may even explicitly exclude them, states remain powerful actors that can potentially challenge the legitimacy of the scheme (Bernstein & Cashore, 2004). The positive value of inclusiveness can also be related to the specific nature of environmental policy making. Many scholars of environmental policy speak warmly in favour of inclusive policies because of the complex character of environmental problems (e.g. Glasbergen, 1996; Mol et al., 2000; van Tatenhove et al., 2000). For example, Lafferty & Meadowcroft (1996) have argued for co-operative management regimes : through constructive dialogue, reflection, negotiation and compromise, groups with different concerns, knowledge and experiences may be able to shed light on different aspects on the complex problem. But one of several preconditions for such engagement is for the parties to recognize each other as legitimate interlocutors. Developing such mutual recognition is likely to be difficult, especially if there are pre-existing antagonist relationships and deep mutual mistrust among the relevant stakeholders and if they exist under asymmetrical power relationships, as do SMOs and business actors. Furthermore, the power resources of participants in inclusively organized projects play multiple and ambiguous roles (Boström, 2006). Participants compete over standard matters by employing their respective power resources, even as their power resources enhance the credibility and authority of the entire labelling organization and project. Hence, there is a delicate mixture of competition and mutual dependency. The inclusion of powerful actors may benefit the whole project; yet there is a risk that powerful actors will utilize their strength merely to advance there own positions, which would risk an undermining of credibility for the entire project. Nevertheless, there is a potential quality that an inclusive labelling organization can add the ability to mediate and overcome particularistic interests. Scholars have observed that repeated interaction in organized networks comprising a wide range of actors can result in mutual learning, mutual trust and common expectations about proper behaviour (Boström, 2003b, 2006; Cutler et al., 1999; Elliot & Schlaepfer, 2001; Rhodes, 2000; Sabatier & Jenkins-Smith, 1999; Wälti et al., 2004). The question is whether and how the Swedish seafood labelling project could employ the standard-setting ideal of inclusiveness despite pre-existing asymmetric power structures, mutual mistrust and controversies among stakeholders (see the Standard-setting ideals in practice section below for a discussion of these background circumstances).

Establishing Credibility: Swedish Seafood-Labelling 141 Independence Such background circumstances may also stand in the way of a second ideal independence which is related closely to inclusiveness. Leading as it does to a combination of roles and resources, inclusiveness can create an image of independence (Boström, 2006). Whereas most individual members reflect a certain interest relative to the aim of the project they are not independent or neutral it is the combination and mutual adjustment of interests that create this image. In the eco-labelling vocabulary, independence is understood in a slightly different way. Third-party labelling schemes (labels established by a body other than the industry or its members) are distinguished from second-party labelling schemes (labels established by industry associations for the products of their producer members) and first-party labelling schemes (self-declarations by the producer, distributor, or seller). A common view among eco-labellers is that only the third party embodies objectivity because the rule setter and the accredited certification body work independently of producers, distributors and sellers of the labelled products. However, because these rule followers actually participate in rule-setting activities, the independence of the labelling organization is somewhat questionable. Accordingly, the inclusion of experts and other stakeholders such as EOs is key to establishing an impression of relative independence if this is possible given the disturbing contextual factors. Auditability A third ideal is auditability, which is understood here in a broad sense. 5 A basic requirement for a credible scheme is the inclusion of independent certification bodies that regularly conduct auditing and on-the-ground inspection in order to verify compliance with standards. Auditing problems can be due to certification dynamics (certification bodies have economic and reputation incentives to approve rather than reject), arbitrary interpretation of certification criteria, information asymmetry (to the advantage of the certified party) expectation gaps and excessive trust in the auditing (Humphrey & Owen, 2000; Power, 1997, 2000). Although such factors mean that certification bodies cannot be as independent as they pretend to be, clear and stringent standards can help facilitate credible certification (Gulbrandsen, 2004). Performance-based standards with measurable standard criteria, in contrast to standards of procedures that regulate only the routines of an organization, strongly enhance the possibilities of undertaking substantive auditing (cf. Humphrey & Owen, 2000). It is therefore a vital task for labellers to consider the formulation of all standard criteria and the design of the auditing function with great care. However, contextual factors may impinge on the possibility of developing a trustworthy auditing function. In this case, a general mistrust towards the entire fishing industry, including suspicions of illicit fishing, would create a challenge for the fulfilment of this standard-setting ideal. Scientific Validity A fourth ideal is scientific validity. Eco-labelling provides a signal to consumers that labelled products are better for the environment than are their conventional equivalents. Such categorical statements must normally be qualified with

142 M. Boström reference to authoritative knowledge claims. However, it is far from straightforward to develop convincing claims about the effect of various production methods on the environment. Reference to natural science and the inclusion of scientists in one way or the other are often seen to be essential in order to establish trust and create an impression of independence. There cannot, however, be a perfect connection between all the specific standard criteria and scientific advice (i) because of great uncertainty and complexity in substantial issues; (ii) because scientists normally disagree among themselves about problem diagnosis; (iii) because practitioners such as fishermen mistrust scientists, who tend not to take the knowledge and practical experiences of the fishermen into account (Johansson, 2003); and (iv) because standard criteria do reflect (and from a normative point of view, should reflect) certain political and ideological standpoints (Klintman & Boström, 2004). Yet, increasingly, scientific validity and scientific consensus are collectively shared ideals in all types of policy making, but particularly in environmental policy making (cf. Lidskog & Sundqvist, 2004). Thus, eco-labellers often try to create an atmosphere of scientific credibility by linking eco-labelling schemes and their related claims to the use of science (frequently using the terminology of life-cycle analysis ; Jordan et al., 2004) (cf. Klintman, 2002). Global Applicability In addressing the fifth ideal, global applicability, Bernstein & Cashore (2004, p. 41) drew attention to the ways in which the legitimacy of non-state governance depends on its fitness or competition with existing international sets of institutionalised norms already accepted as legitimate in the relevant issue areas, whether in the form of specific rules of multilateral hard law, soft law, or uncodified norms nonetheless understood to reflect shared beliefs about appropriate behaviour. They used the term liberal environmentalism to describe a broad hegemonic shift in ideas that has occurred over the past 30 years a shift that is embodied particularly in the global sustainability discourse. New environmental management tools such as the precautionary principle and the polluter-pays principle have been promoted. They exist in parallel sometimes in harmony and sometimes in conflict with the market pragmatist ideal around global free trade. In fact, among the broad issues in eco-labelling, one of the trickiest has been its effect on free trade. Some member countries in the World Trade Organization (WTO) fear that eco-labelling schemes constitute disguised protectionism, for instance an issue which has been debated in relation to eco-labelling of forest products and seafood (Bernstein & Cashore, 2004; Deere, 1999; Melser & Robertson, 2005; Oosterver, 2005). As long as the physical characteristics of products do not differ, the WTO does not allow countries to distinguish between products produced with special care for the social and environmental consequences and products produced without those concerns. 6 Although these WTO rules can be interpreted as strongly delimiting subjects for eco-labelling, they exert stronger control over mandatory labels than over voluntary eco-labels. Voluntary eco-labelling schemes are largely beyond the direct jurisdiction of the WTO. It is precisely the voluntary and market-based character of particular labelling programmes that potentially offer some room for compromise between the demands of the

Establishing Credibility: Swedish Seafood-Labelling 143 environmental lobby and WTO principles (Melser & Robertson, 2005, p. 53). Yet the WTO trade discourse may exert an impact on the EU and national regulation, which in turn may affect eco-labelling. Therefore, eco-labellers may perceive a pressure to consider how trade issues are affected by a local scheme, such as a Swedish seafood labelling scheme, and how to formulate standards and develop mechanisms that enable free trade. The Balance Between Feasibility and Environmental Stringency Yet another standard-setting ideal could be described as the attempt to find a balance between the opposite poles or ideals of environmental stringency versus feasibility. Cashore et al. (2004a) distinguished among moral, cognitive and pragmatic types of legitimacy, all of which can be sources of eco-labelling authority. Cognitive legitimacy refers to the comprehensibility or taken for grantedness of the actions of an organization, moral legitimacy to the right thing to do and pragmatic legitimacy to the self-interest of relevant audiences. Whereas cognitive legitimacy partially overlaps with scientific validity, the other two variants relate to substantive matters that must be considered in order for the labeller to gain legitimacy. The dimensions are often in conflict, reflecting different interests (scientists, movements and business); and the demanding task is to find a balance among them. Klintman & Boström (2004) argued that the eco-labelling in KRAV the organization described below in the seafood-labelling case operates under an eco-pragmatic meta-framing. Ongoing labelling activities are patterned by a general compromise between two opposite poles: the market pragmatic pole (cf. pragmatic legitimacy) that seeks to frame eco-label products as normal and marketable and the orthodox organic pole that seeks to frame eco-labels as natural and environmentally sound (cf. moral legitimacy). Was such a balance and compromise possible in this intriguing case? Eco-labelling of Seafood in Sweden Fuelled by positive experiences from a similar project about forest labelling in Sweden, 7 in 1996 WWF optimistically suggested the introduction of the MSC label in Sweden. MSC had been initiated by WWF and Unilever (a transnational corporation [TNC] in the food sector), which formed a partnership to promote sustainable and responsible fisheries world-wide (see also Fowler & Heap, 2000; Oosterver, 2005; Weir, 2000). The initiators believed that earlier efforts at creating regulatory frameworks to stop over-fishing were complete failures: Unilever, one of the world s leading buyers of frozen fish, recognised that, unless major fisheries took stronger steps to become sustainable, the company would no longer have access to its valued raw materials (Weir, 2000, p. 119). MSC and WWF started campaigns and organized seminars in order to mobilize support. They were able to obtain sufficient financial and symbolic support, especially among food retailers and the fish processing industry, to start their labelling activities. However, MSC was criticized strongly by fishermen and government authorities from Scandinavian and other countries (Constance & Bonanno, 2000). Interviewees from the Swedish National Board of Fisheries say that they were concerned primarily because a private TNC was a main sponsor and played a central role. They were also concerned about the closed nature of

144 M. Boström MSC, because they were being told that specific information was classified and that the officials were not welcome as participants. Formally, MSC now operates independently of WWF and Unilever, but it was their feelings of dependence, as well as the exclusiveness and opacity of MSC, that made Scandinavian officials averse to the organization. Constance & Bonanno (2000, pp. 134 135) reported that MSC was generally criticized by government authorities as a tool of neo-liberal restructuring over and above the jurisdiction of nation states. In its corporativist political setting, the Swedish fishery administration has, by tradition, shouldered great responsibility for policy making (Hultkrantz et al., 1997). One interviewee from WWF believes that it was difficult for government authorities to accept that private actors stepped in and claimed responsibility for what has traditionally been seen as the sole responsibility and concern of the state. Interviewees from WWF said that there was no point in trying to convince state authorities and others to set up a Swedish working group following the MSC framework because of the strong criticism that would ensue. MSC had an impact in Scandinavia, none the less, and the officials who were initially critical towards MSC admitted that it had initiated a process. 8 On behalf of the Nordic Council of Ministers, the Nordic Committee of Senior Officials for Fishery Affairs took the initiative to draft criteria for the labelling of fish products. No Nordic scheme has been introduced as yet, but the Nordic work did lead to the publication of a report entitled An Arrangement for the Voluntary Certification of Products of Sustainable Fishing (Nordic Council of Ministers, 2000). If a labelling scheme were to be established, the report concluded, it was critical that it be traceable and transparent and that it include all concerned parties (cf. auditability and inclusiveness). Other key points were that (i) a labelling arrangement should focus only on over-fishing, and should not include other environmental issues (making it feasible); (ii) that it must relate to the FAO s Code of Conduct for Responsible Fisheries, in order to underline the fact that it builds on international consensus (global application); and (iii) that state authorities should provide scientific material and auditing, among other tasks. Thus, the report touched on the standard-setting ideals and, implicitly, albeit consciously, criticized the MSC model for not fulfilling them. However, one interviewee from WWF said that she was disappointed because the report provided no solutions to problems identified by WWF and MSC. In addition, she thought that the writing of the report was a closed process: WWF was not invited to take part, which is inconsistent with the Nordic group s emphasis on inclusiveness, transparency and openness. One might argue that the principal achievement of the report may have been to disseminate a view of the labelling instrument as feasible, especially to more hesitant stakeholders, such as fishermen. In parallel with the establishment of a Nordic working group, WWF and an association for the Swedish fish industry (Fiskbranschens Riksförbund) initiated their own informal and recurrent meetings among Swedish stakeholders. Fiskbranschens Riksförbund heard demands by the Swedish and foreign food retailers for verification that fish products come from sustainable fisheries. Several meetings were held during the late 1990s, and included various organized actors in the field: food retailers; processing industries; fishermen s associations; EOs and other SMOs; and state authorities in the fishery, agriculture and food sectors. Despite many seminars, there was little accomplished outside of the

Establishing Credibility: Swedish Seafood-Labelling 145 actors gradually strengthened agreement about the relevance of eco-labelling. According to one interviewee representing the fish industry, it was mainly the difficulties in finding financial resources that slowed down the process; but an interviewee from WWF thought that the process was badly focused and disorganized, and that the discussions were merely scraping the surface. Several participants believed that KRAV was an eco-labelling organization capable of assuming primary responsibility for conducting a more efficient process. KRAV s normal activity is to set and issue standards for organic production. 9 Because wild captured seafood falls outside the definition of organic food, KRAV had no expert knowledge invested in this subject. Another problem was that KRAV is a Swedish eco-labelling organization, which may cause problems in trade issues, thus threatening the ideal of global application. But KRAV s advantages outweighed its disadvantages. It was the general opinion of the informal working group participants that the project desperately needed an organization with established and efficient routines. KRAV represented an inclusive and hybrid organizational form to be analysed below and was well known and appreciated by the public. In June 2000, KRAV offered to co-ordinate the development of standards. In 2002 it managed to obtain the financial support of the EU; the Swedish Ministry of Agriculture; the National Board of Fisheries; a number of County Councils; and, through Svensk Fisk (aswedishfishassociation),theswedishfishindustry.indecember2002, a first preliminary framework with principles and standard criteria a so-called remiss was published and sent to various stakeholders for consideration. During the 2003 hearings, seminars and other specific meetings were held between KRAV s Project Group and stakeholders, and a second remiss with standard principles and criteria was published and sent to all stakeholders. In November 2003, a final standard proposal was presented; and it was approved in March 2004 by KRAV s Board of Directors. By spring of 2006, a few vessels have been certified for the eco-labelling of shrimp, and an application for labelling herrings had been approved as well. Some basic facts about the standards are summarized in Table 1. Table 1. The KRAV standard for eco-labelling of wild captured fish and wild captured shellfish summarized briefly The standard has five groups of rules, which apply to every part of the fishing sector from fishing to retailing. The first group concerns competency requirements for the environmental certification body, which focus on such issues as quality assurance (approval according to relevant accreditation organ such as IFOAM or ISO), decision-making structures and the conducting of fisheries assessments. The second group of rules concerns the assessment of stocks. Eco-labelled fishing can only be allowed if the stock is deemed to be within safe biological limits, following a precautionary approach defined by the FAO in the Code of Conduct for Responsible Fisheries and by the ICES. The third group of rules concerns the fishing vessel, as it is the individual fishing vessel that obtains the license to catch eco-labelled fish or shellfish. The rules concern such required routines as documentation and plans for regular, adequate education, as well as specific requirements for the motor, waste and chemicals. A fourth group of rules concerns fishing methods that should minimize by-catches and damage to the surroundings. The last group concerns landing and processing, including restrictions on the use of additives, waste, energy consumption and traceable storing. ICES, International Council for the Exploration of the Sea; IFOAM, International Federation of Organic Agriculture Movements; ISO, International Organization of Standardization. For more information, see http://www.krav.se/vildfisk (February 2006).

146 M. Boström The result can be seen as a general compromise between a feasibility pole and an environmental stringency pole (see below). An important organization SSNC has been reluctant to accept this compromise. SSNC has neither officially approved nor officially criticized the eco-labelling scheme, preferring instead to wait and see how the labelling scheme works in practice. According to a key interviewee from the Project Group, this decision was seen as problematic because SSNC is the most trusted EO in Sweden. With 170 000 members, it reflects consumer power. SSNC does support the project in principle, but, because of some critical issues, it cannot declare official support. 10 Some of these issues will be revisited those which touch upon the main theme of this paper: how to create a credible labelling scheme. Standard-setting Ideals in Practice Inclusiveness and Independence Coping with an Asymmetric Power Structure and Mutual Mistrust and Controversies Among Stakeholders The inclusion of many different, antagonist interest groups with diverse knowledge claims was a huge challenge for the organization of the process. In Sweden, as in many other countries, debates about fishing methods have been impregnated with controversies, and interviewees maintained that representatives from fishermen s associations and EOs could not even sit in the same room initially. Against that background, it is difficult to imagine how a labelling dialogue could be achieved. However, KRAV s hybrid organizational form, which fit the ideals of inclusiveness and independence from the viewpoint of several stakeholders, ensured that the labelling project was not totally captured by any single interest group (cf. Boström, 2006). KRAV has been able to include both SMOs and the big players in the field of agriculture as member organizations, and has developed collaborative relationships with government authorities. 11 Because many actors in the fish industry, especially fishermen and their associations, have been highly suspicious of EOs, they would not have accepted an ordinary EO to do the principal labelling in the same way that the SSNC is eco-labelling various products and services according to its label Good Environmental Choice. The EO, SSNC, owns Good Environmental Choice and it alone can decide on standard principles and criteria (although, in practice, SSNC must obviously communicate with relevant industries in its labelling activities in order to establish credibility) (see Micheletti, 2003 on SSNC s Good Environmental Choice). Many business actors see KRAV as an organization representing environmental values without being completely occupied by EOs, and as an organization which understands economic prerequisites. Interviewees from the National Board of Fisheries, who dismissed MSC for being too private and closed, perceive KRAV as being a semi-official organization into which government authorities have access despite the fact that they are not members. SMOs such as WWF also affirm KRAV (despite the fact that WWF was disappointed after the dismissal of the MSC option). Thus, KRAV could be seen as a good enough option for all participants an option that would give them all some checks on their counterparts. No party objected to KRAV running the project, but the certification of a sustainable fishery was a new arena for KRAV. It could not use its normal procedures

Establishing Credibility: Swedish Seafood-Labelling 147 to develop the standard criteria, which meant that the project had to develop specific organizational forms for this specific project with great care. Questions of how to define inclusion and representation soon emerged. The work was eventually structured into three groups: the Project Steering Group, the Project Group and the Reference Groups. The Project Steering Group, which made decisions about funding and major priorities, consisted of representatives of the fishermen, fish processors, retailers, professional buyers and marine research. The Project Group, which consisted of KRAV personnel, was the unit that co-ordinated and conducted the bulk of the work. Its members wrote standard proposals, and its primary task was to secure acceptance for the project in the Reference Groups, which comprised six categories of actors: scientists, professional fishermen, retailers, environmental and consumer organizations, the fish processing industry, and government authorities. The business side fishermen, the fish processing industry and retailing represented the majority in the group with the decision-making power: the Project Steering Group. If the ideal of inclusiveness was to be wholly fulfilled, one might ask if the EOs, for instance, should not have a more central position in the formal structure, as is the case in many other eco-labelling arrangements (Boström, 2003a). However, this choice of organizational form must be understood in relation to an asymmetric power structure in the fish sector, which, by tradition, has favoured large-scale fisheries. Fishing policies have, by tradition, been exercised primarily within an iron triangle of co-operating decision makers from the Swedish Ministry of Agriculture, the National Board of Fisheries, and the Swedish Association of Fishermen (SFR); thereby excluding opponents such as EOs and the Swedish Anglers Association (Hultkrantz et al., 1997; Sjöstrand, 2003). The groups within the iron triangle had few experiences dealing with EOs. One key interviewee from the Project Group said that he wanted to have EOs represented in the Steering Group, but that it was impossible because of the hostile relationship between the fishing industry and EOs. The Project Group believed that business must be allowed to own the process. The relationship was so infected that EOs could not be included in the inner circle. Because of the need to buy legitimacy from EOs, the Project Group anyway had to consider the EOs viewpoints on standard criteria with care. Another delicate issue as in all eco-labelling cases was how to include consumers. Consumers are expected to fill important roles by utilizing their power at the end of the production chain, but they can also exercise their power by voicing their concern in the standard-setting process. However, within KRAV, consumers are not participating directly, but are represented by retailers (cf. Marsden et al., 2000), by an EO with many members (e.g. SSNC), or even by the labelling organization (KRAV) itself. A great puzzle for the project, then, was what the consumer might think about a seafood label. In order to improve understanding of consumer interest in and expectations of a seafood label, the Project Group conducted a survey (Lönn, 2003). However, this survey showed merely that consumers greatest expectation was for non-toxic ecolabelled seafood something that the scheme seemed to be unable to take into account in any consequential way. 12 This specific organizational design was inclusive in a biased way, which could be a challenge for the ideal of independence. However, no actor officially complained about the design, and it seemed sufficient to establish credibility, at least among the participating stakeholders. KRAV s familiar name did well in

148 M. Boström this regard. It is important to emphasize indications of mutual learning and increased mutual trust among several of the actors as a consequence of the process. The repeated interaction over time via the inclusive standard-setting process has indeed helped actors, including fishermen s associations and WWF, to become less antagonistic toward each other. This situation was mentioned by most of the interviewees but not by the SSNC respondent. Yet, it is relevant to ask if and how this biased inclusiveness affected the choice and formulation of standard criteria, an issue which will be revisited. A Trusted Certification for a Mistrusted Sector? Auditability A basic requirement for a credible third-party scheme is the existence of independent certification bodies that perform audits and on-the-ground inspections regularly in order to encourage compliance with standards. Credible certification requires both stringency in the standards and trustworthy auditing by the certification body. One of the most intensely debated issues has been how to accomplish credible certification, given the lack of confidence in the fishing industry in Sweden. There is a commonly held suspicion in Sweden that fishermen tend to cheat. In official reports (Hultkrantz et al., 1997) and EO journals (Sveriges Natur, 2001/06 Fusket som tömmer haven ), it is claimed that fishermen make incorrect reports about capture and by-capture, that they practice illicit fishing and that they make use of loopholes in the regulations. Hence, much is demanded by the system for auditing and verification. To be sure, KRAV makes lavish references to grand principles such as transparency and traceability in its standard proposal. But what concrete measures does it suggest in order to verify compliance with standards and generate trust? KRAV s idea is, in part, to rely on various self-made documentations (about by-catches and location of catches, for example), which is complemented by a satellite-based auditing system that would be provided by state authorities. Fishing vessels must be equipped by a Vessel Monitoring System (VMS) for automatic reporting of position and movement, thereby providing information about if and where the fishermen are fishing. Whereas some commentators viewed the criteria for satellite-based auditing as probably the most important regarding consumers confidence for the eco-labelling (Co-op, remiss 2), other parties (for instance SSNC and MSC) were not convinced, but viewed the entire auditing function as the Achilles heel of the framework. One objection was that documentation of by-catches cannot be effective unless there is complementary on-board inspection. Interviewees said that there was discussion over whether or not inspectors should accompany the fishermen on their fishing trips. MSC also commented on that issue: There are number of requirements related to the fishing vessel s operations...that cannot realistically be guaranteed unless there is monitoring at-sea (MSC, remissvar 2). However, KRAV understood that it would be too expensive to conduct on-the-ground, on-dock, or on-board inspections. In addition, it is difficult to conduct surprise sample inspections. (It is probably easier but not easy to do sample on-the-ground inspections in forests and agriculture.) In a sense, the eco-labelling system exists because of great mistrust towards the environmental performance of the conventional fisheries. Yet, if the process is to succeed, there must be some trust placed in the eco-labelled fisheries to comply with the eco-labelling criteria. This is the type of trust paradox described