The EU Political System Under Stress

Permanent Link:

Material Information

Title: The EU Political System Under Stress Managing Political Uncertainty in Genetically Modified Organisms Policy
Physical Description: Book
Language: English
Creator: Hinova, Diana
Publisher: New College of Florida
Place of Publication: Sarasota, Fla.
Creation Date: 2009
Publication Date: 2009


Subjects / Keywords: European Union
Multi-Level Governance
Precautionary Principle
Genetically Modified Organisms, GMOs
Agency Networks
Genre: bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation


Abstract: How does the EU system deal with competing claims about the appropriate basis for decision-making? Are all claims equal in the EU arena? Do policy outputs match original objectives? This study examines the case of genetically modified organisms (GMOs) policy as a base for considering these questions. It undertakes an assessment of the EU policy machine from beginning to end. There seems to be room for diverse claims in the EU system, but only in times when the system is seeking a broader range of inputs to bolster its legitimacy. Policy outputs match policy objectives only loosely in the cases under consideration, but this does not seem to matter too much, provided the specific regime faces no insurmountable credibility challenges.
Statement of Responsibility: by Diana Hinova
Thesis: Thesis (B.A.) -- New College of Florida, 2009
Bibliography: Includes bibliographical references.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The New College of Florida, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Local: Faculty Sponsor: Hicks, Barbara

Record Information

Source Institution: New College of Florida
Holding Location: New College of Florida
Rights Management: Applicable rights reserved.
Classification: local - S.T. 2009 H7
System ID: NCFE004114:00001

Permanent Link:

Material Information

Title: The EU Political System Under Stress Managing Political Uncertainty in Genetically Modified Organisms Policy
Physical Description: Book
Language: English
Creator: Hinova, Diana
Publisher: New College of Florida
Place of Publication: Sarasota, Fla.
Creation Date: 2009
Publication Date: 2009


Subjects / Keywords: European Union
Multi-Level Governance
Precautionary Principle
Genetically Modified Organisms, GMOs
Agency Networks
Genre: bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation


Abstract: How does the EU system deal with competing claims about the appropriate basis for decision-making? Are all claims equal in the EU arena? Do policy outputs match original objectives? This study examines the case of genetically modified organisms (GMOs) policy as a base for considering these questions. It undertakes an assessment of the EU policy machine from beginning to end. There seems to be room for diverse claims in the EU system, but only in times when the system is seeking a broader range of inputs to bolster its legitimacy. Policy outputs match policy objectives only loosely in the cases under consideration, but this does not seem to matter too much, provided the specific regime faces no insurmountable credibility challenges.
Statement of Responsibility: by Diana Hinova
Thesis: Thesis (B.A.) -- New College of Florida, 2009
Bibliography: Includes bibliographical references.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The New College of Florida, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Local: Faculty Sponsor: Hicks, Barbara

Record Information

Source Institution: New College of Florida
Holding Location: New College of Florida
Rights Management: Applicable rights reserved.
Classification: local - S.T. 2009 H7
System ID: NCFE004114:00001

This item is only available as the following downloads:

Full Text


THE EU POLITICAL SYSTEM UNDER STRESS: MANAGING POLITICAL UNCERTAINTY IN GENETICALLY MODIFIED ORGANISMS POLICY BY DIANA HINOVA A Thesis Submitted to the Division of Social Sciences New College of Florida in partial fulfillment of the requirements for the degree Bachelor of Arts Under the sponsorship of Barbara Hicks Sarasota, Florida April 2009


To my academic sponsors, Barbara Hicks, Frank Alcock and Joseph Mink. Words will never be enough to express my gratitude. To my husband, Tad, and the rest of my family, who have been nothing but encouraging. To my grandfather, Ivan Zlatkov, wh o is the original source of my interest in political science. An finally to my friends and fello w students, who shared whatever New College is with me. i


Table of contents Abstract vii Introduction 1 Chapter 1 The EU political system: procedures and practices 4 Overview of EU institutions: foundations and functions 4 The EU institutions and the balance of powers 6 Typical process of policy making and implementation 8 Policy generation and legislative procedures 9 Different policy instrument s for different objectives 13 Implementation 14 Legitimacy of the EC and the policy process 16 Output legitimacy 17 Input legitimacy 19 Summary: particularities of the EU policy making process 26 Chapter 2 The EU policy making process under stress: the long-suffered question of GMOs 28 The common complication: Uncertainty and its effects on policy making 30 The impasse 36 Setting the stage 38 Policy making, round one: Directive 90/220/EEC 41 Interest entrenchment 44 Deadlock 47 Interest distribution: networks of varied actors 50 External pressure 51 Policy making, round two: Directive 2001/18/EC 53 Trouble on the horizon 58 Conclusions 60 Chapter 3 Implementation of EU policy: agency networks 65 Delegation theories applied to the EU level 65 Constraints and controls 66 Given all these stumbling blocks, why are there agencies at the EU level? 69 Empirical observations on EU delegation 71 Product approval mechanisms: three multi-level agency networks 73 The EMEA regime 75 The ECHA regime 77 ii


The EFSA regime 79 Beyond EU-level product approvals: an implementation network for GMOs 84 Making a point: safeguard clause invocations and general bans 86 Summary and implications 88 Conclusions 90 Chapter 4 Conclusion 93 Appendices 1. Map of PDO, PGI and TSG license holders 97 2. Map of EU member states w ho invoked precautionary measures against GMOs 98 3. Outcomes of DG SANCO implementation inspection missions 99 Bibliography 100 iii


List of tables, boxes, and illustrations Box 1.1: EU institutions and their powers relevant to this study 6 Table 1.1: Lobbying Systems to the European Commission 21 Table 2.1: Interest distribution in GMO policy 50 Table 2.2: Interpretations of th e moratorium on new GMO approvals 51 Table 2.3: Policy making modes 62 Table 3.1: EU-level agencies some key features 74 Box 3.1: Definitions used by the EU institutions in the governance of GMOs to delineate fields of responsibility 80 Table 3.2: Differentiated pathways implementation of the GM regulatory framework 85 Table 3.3: Safeguard clause invocat ions and general bans and their consequences 89 Appendix 1 Map of PDO, PGI and TSG license holders 97 Appendix 2 Map of EU member st ates who invoked precautionary measures against GMOs 98 Appendix 3 Outcomes of DG SANCO im plementation inspection missions 99 iv


List of acronyms Acronym Name of organization BASF BASF Chemical Company BSE Bovine spongiform encephalitis CA Competent authority CAP Common Agricultural Policy CAT Committee for Advanced Therapies CHMP Committee for Medicinal Products for Human Use COMP Committee for Orphan Medicinal Products CPMP Committee for Proprietary Medicinal Products CVMP Committee for Medicinal Products for Veterinary Use Directorate General Directorate General DG AGRI DG Agriculture and Rural Development DG ENT DG Enterprise and Industry DG ENV DG Environment DG RTD DG Research DG SANCO DG Health and Consumer Affairs EC European Commission ECHA European Chemicals Agency ECJ European Court of Justice EFSA European Food Safety Authority EMEA European Agency for the Evaluation of Medicinal Products / European Medicines Agency ENGL European Network of GMO Laboratories EP European Parliament FDA Food and Drug Administration (US) FVO Food and Veterinary Office GM / GMO Genetically modified / genetically modified organism HMPC Committee on Herbal Medicinal Products MEP Member of the European Parliament MON810 GM maize variety, authorized for cultivation in the EU NGO Non-governmental organization OECD/FAO Organization for Economic Cooperation and Development / Food and Agriculture Organization PDCO Pediatric Committee POP Persistent organic pollutant PDO Protected Designation of Origin PGI Protected Geographical Identification QMV Qualified majority voting SRO Self-Regulating Organization SPS Sanitary and Phytosanitary Measures SCF Scientific Committee for Food SCP Scientific Committee for Pesticides SCAN Scientific Committee for Animal Nutrition TBT Technical Barriers to Trade TSG Traditional Specialty Guaranteed WTO World Trade Organization v


vi POLITICAL SYSTEM UNDER STRESS: MANAGING POLITICAL UNCERTAINTY IN GENETICALLY MODIFIED ORGANISMS POLICY Diana Hinova New College of Florida, 2009 ABSTRACT How does the EU system deal with competing claims about the appropriate basis for decision-making? Are all claims equal in the EU arena? Do policy outputs match original objectiv es? This study examines the case of genetically modified organisms (GMOs) policy as a base for considering thes e questions. It undertakes an assessment of the EU policy machine from beginning to end. There seems to be room for diverse claims in the EU system, but only in times when the system is seeking a br oader range of inputs to bolster its legitimacy. Polic y outputs match policy objectiv es only loosely in the cases under consideration, but this does not seem to matter too much, provided the specific regime faces no insurmountable credibility challenges. Keywords: European Union, multi-level governance, uncertainty, precautionary principle, genetically modified organisms, agency networks Barbara Hicks Division of Social Sceinces


Introduction The European Union as a political system draws the attention of scholars from diverse fields because it is innovative in it s balance of multiple levels of governance. Without reinventing the wheel, its constituent states and instituti ons look to political arrangements from varying traditions to resolve particular problems that arise in attempting to coordinate a multitude of policies that are national in their origin. Their relative successes and failures have been examined from f unctionalist, intergovernmental, and transnational perspectives. The literat ure on EU policy outcomes in given policy areas has usually focused on whether these have been more successful than other international actors attempts and whether the EU is keep ing pace with other global actors or forging ahead. But the most revealing cases remain ones where the EUs political system has taken an entirely different course from that which is internationally endorsed. How does the EU system deal with polic y areas where competing claims about the appropriate basis for decision-making are made? Are all claims equal in the EU arena, and what factors make them powerful? Do economic motivations really prevail, and if so, whose those of states or of transnationa l industrial networks? How are competing claims translated through legislation into practicable polic y? Do policy outputs match the original objectives, and is this even considered a para meter for evaluating them in the EU system? These questions lead to an examinati on of the methods that the EU political system uses to deal with stress, and their answers may provide useful lessons for the future, as well as for other supranational organizations that may be faced with policy coordination problems in similar substantive areas. In addition, they lead to an 1


assessment of the EU policy machine from beginning to end. This study examines the case of genetically modified organisms (GMOs) policy as a base for considering these questions. References to EU policy on ch emicals and pharmaceuticals are used to highlight points in the process of special interest. Chapter 1 provides a general overview of the institutional actors in the EU, the balance among them and among different levels of governance, and the decision-making procedures through the process of turning agenda items into legislation and implementing it. The factors and dynamics that enter into the EUs claim to legitimacy are also outlined. Chapter 2 delves into the GMO polic y case to uncover the sources of stress to the EU political system, and why they rose to such a high degree in this case and not others. The aim is to determine what strategi es different actors used in the EU to cope with the contestation of not only policy, but also institutional legitimacy, and to what degree these strategies used were successful. Chapter 3 turns to the post-legislative stage to follow through the complex implementation provisions set out for GMOs, chemicals, and pharmaceuticals. Delegation to semi-independe nt EU agencies is the main theme, as it has been in the EU of late although most impl ementation responsibi lities remain with member states. Chapter 3 concludes with a br ief look at this arena of implementation and an evaluation of the entire frameworks eff ectiveness. Finally, the findings of this assessment and avenues for further research are offered in the concluding chapter. The assessment reveals that the real reasons behind the disconnect between public opinion and policy outputs, rather than lying in deficiencies in one decision-making procedure or another, are complex and mutually compounding. It appears that although not all claims are equal, there is room for di verse claims in the EU system. However, this 2


obtains only in times when th e system is under stress and is seeking broader inputs to bolster its legitimacy. Economic motivations have a definite ro le to play, especially in structuring the interest field, but they are not the sole factor that influences policy making; they are not equally important over time either. Overlap in economic interests between state and higher levels is difficult to describe, because national positions (as indicated by voting within th e Council) also vary over time, so a simplistic relationship must not be assumed. Path dependence is f ound to be a key factor in determining how competing claims enter and transform thr ough the policy making process. Policy outputs match policy objectives only loosely in the cases under considerati on, but this does not seem to matter too much, provided the sp ecific regime faces no insurmountable credibility challenges. 3


Chapter 1 The EU political system: procedures and practices Before tackling the questi ons about how the EU handl es stress from competing interests and epistemic claims, it is necessary to outline how its political system functions in regular conditions, when it is not under stre ss. This Chapter highlights the balance of powers and the typical interac tions among EU institutions, th e different types of outputs they produce, the points of access for societal interests throughout the process of policy making, the legitimacy considerations the EU system faces and how it attempts to handle them. All of these elements can play an importa nt role in shaping a particular policy area as it enters the EU agenda, is transformed into legislation and then implemented. Overview of the foundations and functions of EU institutions The EU has its beginnings in economic cooperation among states with sufficiently convergent policy outlooks. The European Econ omic Community was created in 1957 to take advantage of opportunities of liberali zed trade among its me mbers. Initially it involved six core states1 previously involved in the Eur opean Coal and Steel Community. In 1970 the Treaty of Rome, by which the Eur opean Economic Community had been established in 1957, was revised to allow for the financing of the Common Agricultural Policy (CAP), which continues to consume the largest single share of EU funds (Hix 2005) and has served as a policy-process model for other areas in which EU institutions subsequently became involved in (Wallace, Wallace and Pollack 2005). The CAP has facilitated the establishment of European ag ricultural interests as a strong influence on 1 Belgium, the Federal Republic of Germany, France, Italy, Luxembourg and the Netherlands 4


subsequent policies and has been criticized by external actors as a massive subsidization and protection enterprise. Successive enlargements brought in Denm ark, Ireland and the UK in 1973 and an additional sixteen states from the south and east of Europe into the evolving community. In the 1970s, the smaller states perceived a threat to their individual interests from the potential consolidation of intergovernmental relations as the modus operandi of the Community, since they were not in a position to contend with the main European powers on this basis. They acted to move away fr om intergovernmentalism, but there was not sufficient political will to create a fully supr anational level of governance either, and the union remained based on functionalist ideas (Dehousse and Magnette 2006). By the 1980s, an integrationist thrust ha d become the cornerstone of the agenda set by the European Commission, the main pol icy initiating body. Territorial enlargement was accompanied by expansion into new policy areas. This expansion has been largely uneven, reflecting in part the particular interests of contemporary EU institutions and in part the limitations memb ers states were willing and able to set to it. For instance, much more policy is made at the EU level on competition and the internal market than on labor issues, which remain mostly a national prerogativ e. During the process of expansion, areas seen as congruent with the integrative impetus of the EC were emphasized over other issues. This prioritizati on was based on the goal of ma king integration itself work, and linking issues with the overarching process of integration made the European Commissions involvement in those areas defe nsible in terms of its Treaty mandate, which is to represent the common interests of the Community. In 1991, the Treaty of the European Union codified these expanded comp etences and established the Union more or 5


6 less as it now exists. A decade later the EU adopted the Lisbon Agenda, a set of policy goals aimed at modernizing the economy thr ough innovation and investment in business and education in order to increase its co mpetitiveness with other major world powers. The EU institutions and the balance of powers The current division of powers among EU bodies is summarized in Box 1 below.2 Some areas, like the EU budget and internati onal relations, have been excluded because they are only tangentially relevant to this study. Box 1: EU institutions and thei r powers relevant to this study COUNCIL OF THE EUROPEAN UNION (Council) -made up of member state ministers (can take different forms, i.e. Environmental Council, Economic Council, etc.) -approves legislation through qualified majority voting -can block some legislative and implementation proposals EUROPEAN PARLIAMENT (EP) -composed of 785 representatives who are directly elected in member states -approves the ECs appointment and can dismiss it -shares legislative power with th e Council under co-dec ision procedure EUROPEAN COMMISSION (EC) -appointed by the Council for a 5-year term -made up of 24 policy Directorates gene ral (DGs) and 17 service departments -initiates policy and legislative proposals -drafts implementation measures -has final approval power under some legi slative frameworks, if the Council cannot reach a decision European Council -the meeting of heads of state of th e EU (with EC President present) -sets the general political goals for the EU Court of Justice of the European Communities (ECJ) -composed of 27 judges (can sit in session of 3-5 or of 13 judges) -can interpret ambiguities in EU legislation -ensures compliance wi th the founding treaties 2 For information on EU institutional bodies and procedures, visit the EU web portals glossary:


The European Commission (henceforth EC or Commission) is the chief policymaker. The EC is comprised of multiple Directorates General (DGs) defined by primary policy areas. However, the DGs also have overlapping departments. For instance, biotechnology is a branch of DGs Research Environment, Enterprise and Industry, Health and Consumers, and Agriculture and Rural Development, although each DG approaches the policy area with its own agenda in mind. This type of loose overlap in competences has been a problem in the past leading to infighti ng among DGs about the direction policy shoul d take (Hix 2005, 47). The Council of the European Union (henceforth Council) is comprised of ministers from each member state, and meets in nine different forms according to the substantive area under consideration.3 The Council rules, thr ough voting, to approve or reject legislative and impleme ntation proposals. It is up to the Commission to decide which form of the Council to convene in acco rd with provisions set out in the founding treaties. The Council is join ed by the European Parliame nt (henceforth EP) in its decision-making power under the most predominan t of the several legislative procedures used in the EU. The EP is a body made up of citizens representati ves who are directly elected in their home state. However, critic s often lament poor turnout and voting based on domestic concerns at these elections, which cast doubt on the demo cratic value of the EP as a mechanism of represen tation in EU policy making. 3 The Council of the European Union has been formalized into nine configurations since 2000. These are General Affairs and External Relations (GAERC); Ec onomic and Financial Affairs (Ecofin); Justice and Home Affairs (JHA); Competitiveness; Transport, Te lecommunications and Energy (TTE); Agriculture and Fisheries; Environment; Education, Youth and Culture (EYC); and Employment, Social Policy, Health and Consumer Affairs (EPSCO). (Wallace 2005, 59) 7


Typical process of policy making and implementation At the EU level, there is an even greater separation of execu tive, legislative and judicial powers than within member states (Balme and Chabanet 2008). In tracing the process of making policy and enacting it through legislation, it is perhaps simplest to use a fictional proposal lets call it PROP to discuss the procedures and eventualities of the interplay between EU institutions, and thei r interplay with national institutions in the implementation stage. The present study will focus mainly on th e initiation and to a lesser degree implementation stages, because these stag es are where access points for interest representatives are found. There is most room for influence in the initiation and drafting of legislative proposals by the EC, when the dominant discourse has not yet been settled, and where certain modes of regulation begin to be favored. After policy is drafted and proposals are submitted pursuant to it, the EP has only limited power to introduce revisions, which can furthermore be blocked by the Council. The Council itself presents a key access point throughout the process. Natio nal-level groups who lobby the relevant ministers at home are the main beneficiarie s of this dynamic. Finally, the Commission is also the body that drafts implementation me asures, although there is some room for influence at the national level at this st age because national experts transmit their concerns about EU measures through the focused committee s the EC uses for this purpose (commonly referred to as the comitol ogy system). In addition, strongly opposed national agencies can stall and block particul ar measures, even though the EU institutions try to preempt this outcome by setting out an optimally acceptable implementation program. 8


Policy generation and legislative procedures Policy is initiated almost exclusively by the EC. That said, it is common for member states to urge a particular DG to l ook into the need for EU-level policy, and possible for the EP and private interests to attempt to persuade the Commission that the need exists (Peterson and Shackleton, 2006). The EC is also limited in its ambition by the mandate set out in the intergovernmental treat ies, although it has us ed integration as a catch-all justification for extending its reach. In any case, it must necessarily be with the EC that PROP would originate and enter the process. There are three possible routes or procedures for a le gislative proposal to take, and these correspond to different balances of power between the Council and Parliament. In the consultation procedure, the Council is only required to consult the EP before taking a vote on the proposal and is in fact not bou nd to consider the latters opinion. The Council has a complex system of qualified majority voting (QMV), designed to give consideration to population size of member states. Voting in the Council is extremely important, and that is why a ttempts have been made to weight it in line with considerations of both stat e representation and popular representation. There are 345 votes in total, and each member has between 3 and 29 votes depending on its population. The threshold requirements for passing a proposal will vary by the type of policy concerned; some issues require approval by a majority of member states, while others require 255 votes in favor for the passing of a proposal. In case of doubt, members have the right to insist that a check be conducte d of whether 62% of the EUs population is actually represented by the approving votes. However, this carefully ba lanced setup often leads to deadlock, with Council members unable to reach a decision to adopt or to change 9


a policy. The repercussions of th is outcome also vary by ca se, since there are different stipulations under each legisl ative procedure. In our fict ional case, a QMV approval (225 votes) is all that is required fo r PROP to become EU legislati on; alternatively, it would be dropped from consideration w ithout an approving vote. Because the role of the EP under this procedure is almost purely symbolic, the Single European Act treaty, in effect since 1987, created the coopera tion procedure. This procedure is similar to its successor the co-dec ision procedure, in fact it is almost the same procedure except that under cooperation the Council can block proposals by refusing to issue an opinion. The use of cooperation has been limited to monetary union policy, so we will focus on the more frequen tly used co-decision procedure to illuminate the details of the process. co-decision, whic h involves the three EU bodies most equitably (if that term can be applied at all), covers many contentious issues such as the free movement of workers, the internal market, education, research, the environment, transEuropean networks, health, cult ure [and] consumer protection.4 The co-decsion procedure is what allows the EP to play the role of joint legislator with the Council in these policy areas. Our PROP would go thr ough a first reading by both institutions, at which time they can propose amendments to it A vote of absolute majority is required for the EP to disapprove the proposed legisl ation. If the EP and C ouncil cannot agree on the proposal after a second reading including any proposed amendments, a Conciliation Committee is convened. The Conciliation Committ ee is made up of equal parts Council and EP representatives and is co-chaired by the two bodies Presid ents. The Conciliation Committee is a last resort for reaching agreement and is limited to eight weeks of deliberation (Peterson and Shack leton 2006, 115). Even if a join t text is agreed upon in 4 EUROPA Glossary, 10


the Committee, each institution must then approv e it by absolute major ity. If this does not happen, the proposal fails. Many scholars have found it necessary to comment on the results that increasing the scope of co-decision has had. Some have focused on the enhanced ab ility of the EP to influence legislation, while others hold up the primacy of the Commission in policy generation and its role in facilitating the legi slative process, as well as the wider role of the Council in approving legisl ation, as proof that little has changed. Those who argue that the EUs democratic deficit is lesseni ng often take the former interpretation as a beacon of hope that actual decision-mak ing is becoming more democratic and meaningful. Others are not so optimistic about the effects of expanded co-decision. Giandomenico Majone (2000) claims that th e chief causes of the credibility crisis faced by European institutions are not so much rooted in a disconnect with democratic practices, but lie, on the one hand, in an ill fit between po licy tasks and administrative mechanisms, and, on the other, in the inabilit y of EU institutions to exhibit commitment credibility. The commitment credibility proble m arises when regulation of a particular policy area is susceptible to politicization, in which case el ected incumbents may come under pressure to continually revise it. In Majones view, closer involvement of the European Parliament has served to exacer bate the credibility problem (2000, 285). Many scholars concede that parliamentarization ha s not resulted in the legitimacy benefits which some scholars predicted would be the na tural outcome of increased participation of democratically elected agents (Pollack 2005, 42). As a mean s of bolstering the EUs commitment credibility, Pollack offers two alternatives: constitutionalism, enacted through minimum requirements for transparen cy and participation, which has been 11


embraced to some extent in the EU, or deliberative decision-making (2005, 42). In the deliberative mode of decision-making, polit ical actors are e ngaged in a sincere collective search for truth; it requires that political actors be ready to empathize with others views, that have a shar ed epistemology, and that all ac tors have equal access to the forum of deliberation (Pollack 2005, 43). However, as will be emphasized later, none of these conditions are necessarily met in the EU policy making process. In fact deliberative decision-making may be a non-starter for the EU because while the potential for it was seen to be greatest in exercises of policy coordination (Pollack 2005, 44), it is precisely when these efforts falter that the EU politic al system encounters its greatest problems. The role of national governm ents in the policy process is manifold. They can be active as leaders, insisting on prioritizatio n and stringency of regul ation in a particular policy area. Through ministers in the Council, they must ultimately approve legislative proposals (with varying degrees of involveme nt from the European Parliament), and national agencies are respons ible for implementation. Cl early, these responsibilities provide differentiated access opportuniti es to interest representatives. The latest development in EU intergove rnmental agreements is the Treaty of Lisbon, finalized in 2007. It in cludes important extensions a nd clarifications on certain core principles of EU governance. The Lis bon Treaty Protocol on the Application of Subsidiarity and Proportionality in partic ular could serve as a constraint to the Commissions advance into new policy areas, because it requires that EU-level draft legislative acts justify themselves with re spect to subsidiarity and outline attendant financial and regulatory implications for memb er states (Art.5). The term subsidiarity refers to the commitment by EU member states and instituti ons to resolve legislative and 12


implementation matters at the lowest level of governance that is ad equate and coherent with the single market. The Lisbon Protocol also provides that national parliaments can object to such drafts and indu ce the proposal to be revised or withdrawn through majority (Art.7 and 8) (Treaty of Lisbon 2007, 151-3). However, the Treaty of Lisbon still awaits ratification by the Czech Republic, while Ireland has outright rejected it through a referendum in June 2008. Another underlying driver of many developments in the EU has been the Lisbon agenda, established in 2000 as a blueprint fo r future policy developm ent. Initially, it was rather ambitious, including particular econom ic goals for each state that focused on increasing employment and competitiveness to peak levels by 2010. At a review of accomplishments in 2005, though, it became clear that these goals could not be met, and that generalized targets for all member states were not necessarily practicable or useful. In consequence, the agenda endure a shift in focus and a drastic reduction in specific expectations, with the current buzzword and overarching goal becoming a knowledgebased economy (European Council 2005). The Li sbon agenda came under criticism from environmentally minded observers for its emphasis on research and investment, because it was seen to promote unsustainab le technological applications. Different legislative instruments for different policy objectives Two main types of legislative instrument s are enacted in the EU. The first, the horizontal type of instrument is issued by the Commission and the Council jointly in areas where a need for harmonization of natio nal policies is perceived. This type of instrument takes the form of a directive, usua lly carries within it a general principle that the Community has agreed on to guide policy in that area, and must be transposed, or 13


translated and integrated, into national le gislation through the member states own procedures. Transposition can take a very long time in so me cases, for procedural or political reasons. The second type of instrument is referred to as vertical, or sectoral legislation. It takes the form of regulations by the Commission and norma lly complements particular directives with specific requirements for im plementation. Regulations are not subject to transposition, and must take eff ect in member states at once when they are adopted at the EU level. Finally there are also non-legislative m easures, such as guidance documents and papers issued by the Commission either in prep aration for drafting legislation, or in areas where there is not the necessa ry convergence of preferen ces among member states to achieve it. These documents can indicate broad or specific policy goals for the community and best practices, but they are not binding on member states. Implementation Policy is implemented by national organs directed by the EC, with the process scrutinized under the system of comitology (o r committee procedure). Comitology brings together national and EU agents, or people involved in a substantive policy area through expertise or membership in institutions, in to committees that attempt to coordinate implementation. Temporary committees composed of interest representatives are created when consultation with civil soci ety is seen as desirable, as in cases where the policy area has traditionally drawn a lot of public atte ntion and where there is disagreement among institutional actors as to the pr oper underlying basis for policy. 14


This system was established in 1987, and revised in 1999 on the EPs urgings to enhance transparency and limit the ability of the Commi ssion to choose what comitology procedures to use for each instance of implementation. An additional procedure was created to increase the EPs ro le in implementation. There ar e three distinct procedures under Comitology, with varying degrees of power for the committee. Advisory committees issue opinions that the EC must ta ke into consideration, but it is not bound to follow their recommendations. Management committees can send the ECs plan for implementation to the Council for review if they do not find it appropriate. The Council can then adopt its own measure for im plementation through QMV. Regulatory committees can in a similar fashion send the pr oposed measures for review to the Council and also to the EP. If the Council fails to act within th ree months, the EC can draft measures for implementation, resubmit the original proposal, or submit an amended one. The newly added (in 1999) procedure involves a regulatory committee with scrutiny and stipulates that the Council and EP must review whether implementation measures alter significantly instruments adopted by co-decision. If either institution objects on these grounds, the EC must submit a new or amended proposal. The EC was formerly criticized for making its choice of procedure strategically in line with its own objectives, us ing advisory committees when it foresaw objections to its intended implementation measures (Hix 2005). In 1999, the types of policy associated with each procedure were spec ified in order to avoid th is possibility. Management committees were designated fo r use in areas concerning th e CAP, cohesion, research, and expenditure policies. Regulatory committees, wh ich give the most influence to the EP, are employed in the areas of animal, pl ant and food safety, environment and 15


transportation. Hix (2005) notes that committees rarely issue unfavorable opinions or move to overturn the ECs intended implementation measures but insists that this is a function of the predictability of the comitology system. The EC knows the limits of what is possible to pass through the national experts committees and insures itself against prolonged conflicts by drafting measures that ar e likely to meet with approval or at least assent by national governments. In the event that national governments, w ho are in the final analysis responsible for the implementation of EU policy, find le gislative requirements ambiguous and wish to contest them on this basis they have recour se to the European Court of Justice (ECJ). The ECJ examines the legislation, its comp liance with the founding treaties, and the limitations it is subject to and pronounces on the balance of responsib ilities (and power) between the EC and states in that particular area. The ECJ is also the arbiter in cases where the Commission finds states deliberatel y lagging in the implem entation of specific legislative acts, and it has the pow er to impose financial sanctions. Legitimacy of the EC and policy process The EU has come under criticism for operating in a manner that is not democratically legitimate, especially when policies with impacts outside its borders are concerned. These critiques have focused mainly on th e operation of the EC due to its role as initiator and broke r of agreement in matters of policy and legislation, although a number of analysts have also focused on the predominance of QMV voting and its repercussions for the types of deci sions that are favored (Hix 2005, 86). 16


There are various aspects to the con cept of policy process legitimacy. The legitimacy of policy, the end result of the pr ocess, can be assessed on the basis of the procedures that govern its creation, the degree to which it satisfies interested parties in the aggregate, and the degree to which it addresses the original goal for which it was initiated. Policy making as a pr ocess can be assessed on th e basis of whose input and preferences it aggregates, what types of information it uses, and in what manner decisions are made on sticky issues of standards fo r policy and implementation measures. Output legitimacy Output legitimacy refers to the trust and acceptance by constituents of policy and legislation. It can have differi ng bases, for instance procedural legitimacy or the belief that policy and legislation are arrived at through procedures congruent with the organizing principles of the st ate, or representative legitim acy the idea that output reflects the wants and needs of those interested in the substantive policy area. In the EU context, we can look at several components to evaluate output legitimacy. First, is the output. Does it re flect expertise? Are its mechanisms of enforcement, economic costs, and timeliness appropriate to the objective for which it was initiated and for the substantive policy area? The second basis for evaluation is the procedures that led to this output. Which DGs a nd states took the lead and why? What legislative and implementing procedures were chosen, and how well did these procedures function? These questions lead us directly to the possible strategies for enhancing output legitimacy. The strategies include (1) measur es to increase procedural and substantive transparency (by making the rules that govern institutions and the documents under their consideration available for pub lic review), (2) attempts to increase the breadth of 17


consultation with civil society, and (3) enhancing the role of the EP as a co-legislator. In addition, (4) delegation of stages of the process to t echnocrats can be use d, but this tactic is seen as a subterfuge by the EU institutions who want to avoid politicized aspects of policy and topics where member states cannot reach compromise. Finally, (5) the subsidiarity principle can be used to rest rict the reach of EU institutions over thorny policy areas and contain these at the national level. The convenient method of creating independent agencies with bounded authority to handle politicized aspects of policy making is arguably at tension with the subsidiarity principle, to which EU member states remain committed as evidenced by its highly emphasized role in the Treaty of Lisbon 2007. The fear here is that when standards or trends in policy diverge among member states, actors seeking one or the other extreme will push the debate to the EU level. However, assuming that stances on these issu es are linked to othe r national objectives, the EU institutions cannot effectively reach a coherent policy. As we shall see later on, the result may be an independent agency co mposed of national experts who are tasked with establishing the st andards that political actors coul d not agree on a nd applying them to cases. The question is, is it really necessary to go through this process of forcing policy upward to the EU level, when in the end it is still national experts who have the evaluative role? In prac tice, the answer to this question is rather complex, and we will return to it when examining the GMO case. All five of the strategies for enhancing output legitimacy discussed above have been employed in the EU, sometimes simulta neously, to improve reception of certain 18


policies. We will look at them in more detail when we examine the record of biotechnology policy. Input legitimacy Input legitimacy is essen tially the scrutiny of the pe ople and organizations who influence the policy process in order to dete rmine whether there are procedural or other biases in the process. Input legitimacy has several dimensions in the EU: stakeholder participation and interest repr esentation have to be conducted transparently and balanced with input from established national expert centers. The issue is further complicated by the fact that there is not uniform demand or uniform supply of input across policy areas. The demand and supply of input are interr elated but can be disaggregated for the purposes of analysis, a task that can be ve ry fruitful since both dimensions reveal structural constraints to input legitimacy. Demand side There is broad consensus, at least among those who go beyond the intergovernmental approach to analyzing EU governance, that the European Commission actively structures how input of both stakeholders and experts is supplied to it (Broscheid and Coen 2003, Mahoney 2004, Greer et al. 200 8, Wallace, Wallace and Pollack 2005). Expertise in various substantive issues is ab solutely essential to the Commissions work and specifically for enhancing output legitimacy. The Commission establishes Expert Groups in policy areas that require continuous, technica l, substantive regulation in order to ensure that it is not making policy in the dark; the experts in these bodies serve as a sounding board for the appropriateness of policy, but they also serve as a defense 19


against claims that EU policy is made by out-of-touch Eurocrats. The distribution of these Expert Groups across policy ar eas and their composition reveal a lot about the European Commissions agenda and its prioritization of policy areas and knowledge resources. An analysis of Expert Groups will be undertaken later in this study for the relevant policy area. Hix (2005) outlines three broadly used models of intermediation between competing societal interests by the state: corporatism, where the division is between capital and labor, consociationalism, where th e division is between cultural groups, and neo-pluralism, where divisions are artificially enhanced by state institutions that promote what they think are groups representing the su bstantively relevant societal interest. The EU cannot be defined as either consociati onal or corporatist, because a multitude of groups hold sway in different policy areas, and these divisions fall along cl ass lines in some areas and not in others. There are also elements of neo-pluralism in the EU, since the Commission directly subsidizes certain groups that represent societal interests. Because there are so many extant divisions wi thin policy areas at the EU level, diverse strategies are employed fo r mediating participation and no single model is able to capture this phenomenon in the EU system. The Commission sometimes uses direct subsidies to enhance stakeholder participation, disbursing funds to support organizations that fit its idea of good interest representation vehicles (Mahoney 2004). It also exerts an influence on the extent of stakeholder mobilization indirectly through its agenda set ting and allocation of resources, favoring certain policy areas over others. It is on ly logical, after all, that items prioritized on the EU agenda will be more likely to dr aw the attention and time of civil society 20


organizers. Finally, the process of consultation with civil society representatives that the European Commission uses is similar to that of creating Expert Gr oups. In this sphere, the relatively permanent bodies are C onsultative Committees, while interest representatives not already integrated in the system can self-report on their constituency and the policy areas where their input is relevant. Through their nuanced analysis of lobbyi ng behavior, Broscheid and Coen (2003) establish that the threshold for inclusion of a given organization in the consultation process is lowered as the importance of input to the Commission in that policy area increases. That is to say, the Commission seek s a greater quantity of input in politicized areas, where legitimacy is a greater concern. The EC is also not exempt from the tendency of policy making bodies to select some representation organizations as insider interests and leave others on the outside of the default pool of consultation. These terms are us ed in the lite rature with varying meaning, but for the purposes of this study Broscheid and Coens definition will be taken, whereby insider intere sts are simply those organizations likely to be involved in multiple representation events or venues across policy domains (2003, 168). How they come to be involved in multiple arenas will be discussed further below. Many scholars have attempted to undert ake relatively comprehensive studies of lobbying behavior and success at the EU level. This daunting task was facilitated by the creation of CONNECCS, a regi stry maintained by the Commission that catalogued information about the constituency, policy ar eas of involvement, size, and location of groups that sought involvement in the polic y making process. All the information was self-reported by the in terest groups, which may have motivated them to over-report on 21


questions about their policy fields or constitu ency. Based on studies that drew data from the CONNECCS database, Greer et al (2003) found systematic and predictable variation by state of origin, whereby intere sts from the core of the EU were favored over those from Mediterranean and post-communist states; population si ze of the state of origin and former experience lobbying at the EU level were also factors for inclusion. The preference for seasoned interest s was also found by Mahoney (2004), who discovered additional biases towards organizatio ns representing multi-state constituencies and towards those who concur with the Co mmissions integrationist agenda. It is interesting to note that despit e the strong evidence for inclusion bias, becoming an insider interest was related solely to participation in multiple Consultative Committees and formal discussion venues, suggesting a self-p erpetuating pattern of involvement in the policy process. Since these dynamics are spec ific to each policy making instance, a more detailed analysis of the dist ribution of interest representa tives in areas relevant to biotechnology will be undertaken further on. Based on their analysis of the Commissions relationshi p with the sources of input which stipulates that participants provi de expert and aggregate (across nations) information to it, Broscheid and Coen (2003) model the behavior of lobbyists. They elaborate on three patterns of lobbying at th e EU level that vary by the nature of the policy issue these are represented in Ta ble 1.1. Of the three lobbying systems they identify, the authors conclude that those ch aracterized by small numbers of lobbyists and low demands for output legitimacy (Types I and III in Table 1.1) are most commonly encountered (Broscheid and Coen 2003, 178-9). 22


Table 1.1: Lobbying Systems to the European Commissio n 5 Type I highly technical and not politicized issues Type II not technical, highly politicized issues Type III not technical, not politicized issues Organizational costs of lobbying High Low Low Number of lobbyists, breadth of interests Small Large Small Societal demand for output legitimacy Low High Low Environmental policies, and even more so policies that straddle the environment to human health interface, frequently attract broad mobilization among interests and are highly charged politically (Tiberghien and Papic 2006, Eckley and Selin 2004). In this sense, the policy areas to be examined here (GMOs and chemicals) represent relatively rare constellations of participants in the polic y process and for this reason it is even more essential to describe their dynamics and draw conclusions about their outcomes. The EP is another venue where interests ma y seek to exert influence. It is often considered more friendly than other EU in stitutions to environmental and consumer protection organizations, although there is segmentation by DG in the degree of receptivity in the Commission and by state in the Council. The EP can serve to alter the policy discourse, in accordance with its f unctions (which are often limited to issuing resolutions) as a forum for the articulation of positions that are merging as crossEuropean norms. Within the EP, legisl ative proposals from the Commission are examined by topically organized committees th at are chiefly responsible for amending these proposals and presenting them to the fu ll plenary session for approval or rejection (Hix 2005, 93). Committee membership is normally correlated with the previous 5 Table constructed on the basis of findings by Broscheid and Coen 2003, 178-9. 23


experience of MEPs but allocation is also ba sed on the loosely formed party groups at the EU level (Hix 2005, 93-6). As outlined above, under the co-decision procedure the EP has a greater role in approving legislativ e proposals, which makes it worthwhile for interests to approach individua l Members of the European Pa rliament (MEPs) either at the EU level or in their home state. The dependence on national agencies for implementation and the importance of voting in the Council increase the incentive for lobbying at multiple levels as well as across borders, when officials outside an intere sts home state prove to be more receptive to change. In the policy initiation and legisl ative stages, actors can attempt to persuade state governments that it is in their interest to promote a certain policy or regulatory instrument. In the implementation stage, they may alert the Commission of noncompliance by a given state, or work with implementing agencies to provide models and information for the successful realization of policy. Supply side The supply side of input legitimacy is best dealt with in the specific context of a case, since what is being suppl ied and who is in a position to supply vary greatly with the substantive issue. We will return to exam ining the supply of input dynamic later on, but there are nonetheless some general observati ons that obtain acro ss cases in the EU. The supply of input is affected by vari ation in the resources and discourses available to stakeholders, both of which can have constraining effects on their ability to organize. Aside from the Commissions pref erence for certain type s of organizations (long-standing, aggregative ones), we have to take into account the availability of financial and human resources to non-govern mental organizations (NGOs) and other 24


groups that might have a stake at the EU-level policy process. It is also widely accepted among analysts of EU policy that industry is in a better default position to exert influence on the process (Rosendal, 2005) than many of these groups, due to the market-oriented, integrationist focus of most EU policy a nd the dominance of centralized information structures in industry. An additional disparity among those seek ing to influence EU policy can be discerned in the area of expertise. It was only in the 1990s that ethical considerations began to be linked to objectiv e scientific information, a nd these considerations have still not attained an important position among the types of information that should inform policy in practice (Abels, 2002). Jasanoff (2005) has described this bias extensively, at both EU and national levels, toward natura l science as purportedly purer and less politicized than interdisciplin ary analyses that take into account the social and cultural effects of technology. These perceptions lead to the marginalizat ion of alternative scientific interpretations and the inputs that bring them forth, from the sub-national level up through the chain of institutions. All these elements feed into a general idea that the quality of information being supplied matters not only for policy-makers and policy outcomes, but also for the suppliers. Presumably, policy is made more le gitimate by being based on information of a good quality; but this statement is tautological if the parameters of good quality are not specified. In reality, those parameters are in cessantly renegotiated as attention to the policy process and its underlying principles shif t, and yet understandi ngs of what good information is and whether it has been obtained have a continual impact on legitimacy. Access to policy-makers can be assumed to increase in parallel with the quality of 25


information supplied. This principle highlig hts the learning proc esses inherent in consultation and suggests that policy maki ng may become better over time as a more comprehensive and effective consultation netw ork is formed. This relationship offers insight into a factor for the selection of in sider interests: policy-makers will prefer to make insiders those interest representative s who have reliably prov ided them relevant information in the past. Clearly, the capacity of interest representatives to generate or aggregate relevant and accurate information is a key factor that distinguishes them from one another. Summary: Particularities of the EU policy making process The momentum of EU policy expansion has come from the desire to make economic integration more complete and comp rehensive, and it has led to efforts to harmonize practices and standards in various policy areas, although not always without resistance from states. Member states have ensured that measures to constrain this expansion have been built in to the founding treaties and ar e reinforced through decisionmaking procedures. In recent years, the EU ha s become concerned with its international position and is making concerted efforts to pr omote its knowledge-b ased capacity. This drive to expand research and development and make them a cornerstone of the EU economy has resulted in a privileged status fo r novel technologies, but this status is not uncontroversial. Societal actors wishing to contest the EU agenda have a variety of access points throughout the process of policy making to assert their claims on normative and procedural questions. These claims can be made during the ECs consultation periods, when comments from indivi dual and organization actors are solicited on specific 26


proposals, or directly to MEPs and minister s at the national level. Intra-institutional divisions can also shape the final policy out come, as delegative choices made at every step of the way determine the forum and cast of personalities responsible for moving the process along. Choices have to be made as to which DG must take th e lead on a particular legislative proposal (despite frequently ove rlapping competences), on which form of the Council will vote on the proposal (a clear signal to internal a nd external actors as to how the issue has been framed and what di scourse is becoming dominant), and which Committees in the EP have the opportun ity to propose amendments to it. On the receiving end, the cr iteria for inclusion as insider interests will tend to privilege some actors more than others. The preferences that emerge from a survey of interests involved with the Commission and Pa rliament are linked to the experience of the organization, its constituency, and the cong ruence of its goals with the ECs policy outlook (generally, integration). On reflecti on, it becomes clear that these criteria, if applied in the same way to all actors, would privilege industrial gr oups active at the EU level because of their pro-integration focus as well as their role as employers. In addition, industrial actors in most cases have a stronger resource base than non-profit or voluntary associations, and this factor can be importa nt for policy-makers who are seeking input primarily as a source of information. N onetheless, non-industrial interests are not completely marginalized. The Commission even goes so far as to lend support to some civil society organizations that are, by the criteria discu ssed above, good representatives of multi-national constituencies. This neopluralist balancing is necessary, because competing agendas within the EU instituti ons seek to draw on diverse sources of information and public support to enhance their legitimacy. 27


Chapter 2 The EU policy making process under stress: The long-suffered question of GMOs The attempt to build working routines of governance around biotechnology has called into question some of the founding assumptions of liberal democracy Sheila Jasanoff (2005, 274) Food safety regulation in the EU is embe dded in broader goals and controversies of European integration Grace Skogstad (2006, 235) The differences in European and American policies toward GMOs have less to do with either economics or culture than with the emergence of a new European approach to risk regulation David Vogel (2001, 18) What lies at the root of complications in the policy making process as regards biotechnology? Some analysts have argue d that outward-looking economic concerns have been an obstacle in this area, while others look at much less frequently cited culprits, namely the divergent approaches to formulating and testing knowledge among European countries as well as across the Atlantic. Sheila Jasanoff (2005), one of the most outspoken and influential among the latter camp, contrasts several approaches employe d by nations and societies to framing biotechnology within the contex t of their polity. She argues th at these differences have implications for the type of regulatory m echanisms applied. Initially, biotechnology was predominantly looked at as a process, a fram e consolidated in the United States during the 1975 Asilomar Conference. Th is conference was itself c onvened in response to the National Research Councils Berg Letter, wh ich called for a temporary moratorium on recombinant DNA experimentation due to potential risks (Cantley 2004, 261). At the conference, an emphasis on degrees of hazard was laid down as the chief evaluative concern with regard to applications of genetic modification. As biotechnology increasingly entered the market, a new fram e emerged that saw it as an industry of 28


products, and the regularization of bringing products to the mark et resulted in a tendency against precaution (Jasanoff 2005, 52). Finally, Jasanoff contrasts these approaches with one that treats biotechnology as a program of activity: neither simply a novel scientific process nor a cornucopia of new products, but a program of state sponsored activity that threatened important individual and collectiv e values (2005, 54). Interestingly, both the product and program frames afford the possibility of convergence between expert and bureaucratic opinions in favor of the expa nsion of applications into the market. Some who focus on the international trade implications of th e EUs regulations offer a very different view on the reasons behind the complex institutional dynamics and their seesawing outcomes. This alternative hinges on the fa ct that, at the outset of the controversy, most producers of GM crops were located in the United States, and a large share of their exports was intended for the European market. Proponents of the regulatory protectionism argument insist th at the delays in resolving the internal conflicts in the EU, and the resultant stricter standards for GM products, were a concerted effort to shield, variably, European farmers, European biot echnology firms, or both. There are some who refute the argument that the EUs comparativ ely stricter regulation of GMOs serves the goal of protecting its own companies agains t US products on the basis that European biotechnology firms have to be equally as competitive as foreign firms in both scientific and marketing terms, meaning that regulati ons would hurt them equally (Rosendal 2005, 90). Others have pointed out that regulat ory barriers that ri sked slowing down the development of GM products were not seen in Europe as particularly costly in terms of foregone economic opportunities, because these opportunities were in any case likely to be seized by American rather than domes tic firms (Cadot, Suma-Eisenman, Traca 2001, 29


19). Since there have been no outright declara tions from any EU institution to indicate explicitly that trade advantage consideratio ns have tainted the debate, this argument remains in question, and is best dealt with in the context of specific events. As we track the evolution of EU regulatory measures on GM Os below, we will pay attention to points that are relevant in assessing the va lidity of the protectionism argument. The common complication: Unce rtainty and its effects on policy making Scientific uncertainty is present in these cases due to the limited knowledge we have of projected and unforeseen adverse effects on organisms and ecosystems both in direct and indirect contact with the products This scientific uncertainty leads to a heightened attention to the decision-making process and th e quality of information on which decisions are based, thus raising the i nput and output legitimacy requirements. The approach commonly taken by the EU institutions in the face of politicization of scientific uncertainty is a deficit model, whereby the publ ic is thought to be distrustful of GMOs, or whatever the product in question may be, becau se it is ignorant about them (Abels 2002, 2). This approach is shared by industry. Th at is to say, the predominant position among policy-makers and industrial representatives is not that the public is making a choice based on information and concerns alternative to mainstream ones, but rather simply that the public does not know enough. This position can be found in the litera ture as well as in direct conversation with European official s (Jasanoff 2005; Abels 2002; EC Official A 2009). The case of GMO regulation has many sim ilarities with that of chemicals regulation. In both cases, there was significant attention from the public and an unevenly distributed industry. Most inte restingly, in both cases the a ttempt to harmonize regulation 30


at the EU level led to a deba te on the applicability and us efulness of precaution and the particular standards and measures to be us ed in implementing policy, the latter being essentially a debate on methods to decreas e or delimit uncertainty. All the parties involved acknowledged the existence and importa nce of these debates. The third case, pharmaceuticals regulation, provides a somewhat varied story for comparison. In this case, the preexisting national re gulation was the densest, By c ontrast the degree to which the public was mobilized was not significant. In short, the uncertainties inherent in the process and the standards for authorizati on of new products were shielded from politicization by the consent of both industr y and national agencies to the evolving regulatory regime. Because the condition of uncertainty prompts regulators to seek more and better information and this quest is reinforced by an overarching reliance on the deficit model, the proliferation of risk-oriented resear ch findings can (and did in the GMO and chemicals cases) lead to more and different uncertainties about a wh ole range of issues. More importantly, whether risk-based research proliferates uncertainty or simply delivers results received in a negative way, it can lead to distrust towa rd the regulatory system and the industry proposing the novel technology. This process is described by Levidow, Carr and Wield (2000, 199) when they discuss ho w, in the 1997-9 period, expanding the definitions and the scope of research on adverse effects of GMOs6 simply led to a shift in focus toward finding acceptable thresholds of GM content, and not to a broadening of 6 Council Directive 90/220/EEC of 23 April 1990 on the deliberate release into the environment of genetically modified organisms, Art. 4 (1) Member States shall ensure that all appropriate measures are taken to avoid adverse effects on human health and the environment which might arise from the deliberate release or placing on the market of GMOs. 31


the body of knowledge or consensus on which of these effects should be regulated and how. The problem of transformative uncertainty plagued even the case of chemicals policy, in which initially there was consensu s on the shortfalls of the system, but not necessarily on the way ahead (Pesendorfer 2006, 96). The essential problem in this case was that there simply was not data on the eff ects chemicals currently in use could have on humans and the environment. This, of cour se, made it impossible to make comparisons with newly developed altern ative products or to harmoni ze national standards for the control of chemicals in any meaningful way. Prior to the establishment of the bioaccumulation of these chemicals as a scientific fact, the mere uncertainty about whether they persisted in biomass long beyond their ingestion had not generated sufficient concern to catalyze regulation. The 2001 Stockholm Convention on Persistent Organic Pollutants (POPs), which entered into force in 2004, garnered international acknowledgement of the adverse effects of bioaccumulation.7 Because of chemical policys close interrelation with agricultural and environmental policy, both usually hammered out at the national level, the question of establishing toxicity standards, testing and evaluation methods became rather politicized. These discussions in the chemicals case even provoked a scholarly debate on whether uncertainty should be communicated to the public, in which Grill and Hansson argue persuasively against what they term epis temic paternalism, or the withholding of information about uncertainty, on the grounds that a) information is absolutely required 7 Based on a Commission proposal (COM (2003) 333 final), the European Parliament and the Council adopted Regulation (EC) No 850/2004 [which]entered into force on 20 May 2004. The new Regulation goes further than the international agreements empha sizing the aim to elimin ate the production and use of the internationally recognized POPs. 32


for citizens to participate meaningfully in de mocracy and b) defining the parameters of uncertainty provides a better basis for private and public choices than unbounded uncertainty, which results in anxiety and distrust (2005, 648-51). Perhaps if these insights were brought to bear more closely on those espousing the defi cit model of public opposition their strategy for conciliation coul d be advanced. Recently, EC officials are conceding a more narrow conception of the de ficit, namely that while the public may have its own valid opinions and beliefs a bout such products, it is not well enough informed about the body of la w that governs their authoriza tion and the safeguards it puts in place (EC official C 2009). But acceptance of the procedure and safegua rds crafted by the EC is contingent on more than mere understanding of them. A concept created by Jasanoff captures well what else is needed, and what else is lacking in the case of GMOs. As originally defined, Civic epistemology refers to the institutionalized practices by which members of a given society test and de ploy knowledge claims used as a basis for making collective choices[M]odern technoscientific cultures have developed tacit knowledge-ways thro ugh which they assess the rationality and robustness of claims that seek to order their lives; demonstrations or arguments that fail to meet these tests may be dismissed as illegitimate or irrational. (Jasanoff 2005, 255) In other words, all of society is active to some degree in determining the standards for how to evaluate scientific claims and thei r sources, and how to properly value them in relation to claims based on ethical and social considerations. Herrick elaborates a similar but more specified conception when she discusses cultures of GM, or systems of questions and opinions raised by biotechnology that are internalized by the public, whose members then socially reinterpret legislativel y defined risk in orde r to legitimize certain discourses (2005, 287). The key merit of these conceptions is that they allow for the 33


differing knowledge-legitimation practices across states to be considered as factors in explaining contrasting (and c onflicted) policy outcomes. The differences highlighted by a civic epistemologies approach are incorporated in the institutional-legal contexts of states Thus, in the United States, where the burden of responsibility is commonly placed on th e producer and litigation is employed to enforce obligations, adversarial legalism prevai ls, while in the EU authority is dispersed among states and institutions with supermajority thresholds required for policy change, resulting in a meditative styl e (Skogstad 2006, 219-20). The c ontrast in contexts makes each system more amenable to predetermine d mechanisms for holding third-party agents accountable to the public for ex ternalities they produce, which is arguably the main issue behind the GMO controversy. In the United States, the principle of substantial equivalen ce is applied with respect to GMOs, and it is also used by Organization for Economic Cooperation and Development / Food and Agriculture Organizat ion (OECD/FAO). This principle is based on the premise that the use of different pr oduction practices to yield products that are essentially the same should lead to processbased discrimination. But the principle still leaves ambiguous the exact degree of equiva lence that is cons idered substantial (Lieberman and Gray 2006, 596), and results in virtually any conceivable GMO being exempt from potential labeling by the FDA (Herrick 2005, 290). A similar principle was employed in the EU under Directive 90/220/ EC, where mutual recognition by member states of each others assessment was the main procedure for approval of new products. However, the meditative style of the European Union gave room for conflicts as to the appropriate standards for appr oval to persist, and this method had to be reformed. 34


Directive 2001/18/EC supplanted it with the principle of precaution, supplemented by regulations for preand post-market control, so under the current regulatory regime the process and product frames are bot h in operation (Herrick 2005, 289). But how did precaution come to be a favored principle in a collection of states first and foremost bound together by econo mic objectives? The precautionary principle first entered the realm of international law, although not as a fully binding principle, at the 1992 United Nations Conference where it was incorporated in several documents (Victor 2001, 316). Although it is interpreted diffe rently in various contexts and agendas, the precautionary principle essentially en tails the commitment to take all actions necessary to prevent harm to the environment a nd human health before that harm occurs.. Since the early 1990s, environmentally minded NGOs and politicians have attempted to extend its application to a multitude of policy topics at all levels of governance with varied degrees of success. There has b een for some time a tendency for advocacy coalitions, or broad networks with similar pa radigmatic views, to emerge in the EU along a familiar fault line, where interests in favor of precaution are pitched against ones more narrowly focused on industrial progress (Pes endorfer 2006, 97). In Vogels assessment, while the precautionary principle cannot be divorced from science its growing popularity in Europe reflects the perception th at scientific knowledge is an inadequate guide to regulatory policy because it priv ileges responsiveness to public opinion as a legitimating factor for policy making bodies (2001, 29). The discourse concentrated on precaution became explicitly relevant to GM Os with the drafting of the Cartagena Protocol in 2000, which required that exporters have the exp licit permission of receiving states for shipments of GMOs (Victor 2001, 315). 35


To summarize, some analysts believe the lack of faith in the EUs regulatory regime is simply a result of the incons istency of enforcemen t of GMO regulation between the national governments and the s upranational level or an implementation deficit (Victor 2001, 304). However, even if this is in fact the case, the inconsistencies in enforcement are not the mere outcome of time-lag, but have been undergirded by disagreements on the appropriate sta ndards and measures. Those who see an epistemological basis for the divergent regul atory approaches on the two sides of the Atlantic argue more persuasively that the biggest problem faced by EU institutions in terms of meeting their constituents demands re mains that they are confusing a lack of knowledge (uncertainty) with trust (credibi lity) (Jasanoff 2005, 253-5). In confusing the two and insisting that measures to decreas e uncertainty are sufficient to ensure the legitimacy of the regulatory regime, the institutions and industrial interests are contributing to the simultaneous politicization of science and scientificization of politics (Abels 2002, 2) which is arguably at th e root of public discontent, and which has as yet unknown long-term political consequences. The impasse Abels (2002, 9) insists that opposition to GMOs and the trajectory of the debate was linked closely to questions about the na ture of the EU polity. The same idea is echoed time and again by other scholars: Thr ough the 1990s, the high level of scientific uncertainty about the effects of GMOs on th e natural environment and on human health has been accompanied by a similarly high level of political dispute and public distrust of regulatory authorities in Europe (Rosenda l 2005, 96). Even those who approach the issue from a trade perspective and deny the ac tual importance of uncertainty tend to agree 36


with this general assessment, claiming that what drives consumers concerns is a distrust of the motives of their regulatory authoriti es, the honesty of their politicians, and the objectivity of their scientists, rather than fear of any ge nuine danger (Victor 2001, 320). Given that so many different aspects of the collective civic epistemology of the EU are distrusted and challenged, how can analysts and policy-makers approach this impasse? The GMO case is emblematic of a situation described by Vogel and Ansell as contested governance, or a perva sive sense of distrust that challenges the legitimacy of existing institutional arrangements [accompan ied by] widespread public debate (2006, 10-11). Contested governance in this conception is more than just the outcome of politicization. It is instead a precondition for a deepening of contestation, because it shifts the conditions for successful mobilization by interests and privileges certain types of protest (Ansell, Maxwell a nd Sicurelli 2006, 119) because the public becomes sensitized to certain appeals In this situation it appears indeed that distrust (and not uncertainty per se) is what allows for political entrepreneurship, by raising the stakes for the quality and variety of information employed in policy making and in some cases redefining the policy problem altogether. However, even the more nuanced approach afforded by an analysis using the lens of contested governance fails to fully capture the EU context because of the importance of interplay among multiple levels in that system. In Tiberghien and Papics version contested multi-level governance consists of a govern ing body that is simultaneously undergoing a transition of gove rning paradigms and fragment ation of its institutional structure (2006). They posit that because uncertainty is pervasive and highly visible in this situation and the legitimacy of the governing body is questioned, contested multi37


level governance allows political entrepreneurs at different levels to assume roles that are disproportionate to them under normal conditions. Few would dismiss the claim that the EU is experiencing a legitimacy crisis that has forced it to adapt to its own and the publics changing criteria for pa rticipation and policy effectiven ess, at the same time as it is reconfiguring its relationship to environmental policy (in terms of both the EUs role vis-a-vis national governments in creating it, and the policys impact in legitimating EU institutions). These simultaneous processes are an apt illustration of contested multi-level governance as discussed thus far. Setting the stage At the beginning of the 1980s, the biotechnology sector was seen by EU institutions to hold much promise, and it became an area targeted for expansion under the Fifth Research Framework. It was intended to regain a leading competitive role for Europe in the field of products heavily reliant on research for their development. This attitude towards the field clearly falls with in the program definiti on proposed by Jasanoff (2005, 54), but it would soon meet challenges from civil society. By contrast, the European chemical industry did no t need explicit support, as it was already a leader in the world market. The importance of the chemicals i ndustry to the EU, and its role as part of the single market agenda, allowed policy makers to insulate this area by holding back substance-specific regulation to protec t economic growth (Pesendorfer 2006, 95). In the chemicals regulation case, there was not a simultaneous need for new marketing and retail strategies to stake out a place in th e European market. This difference, or the biotechnology industrys immaturity, is what opened up the GMO issue to alternative framing by the opposition. 38


The biotechnology industry is characteri zed by a high degree of competitiveness that is a crucial hurdle to companies with less capital both for research and development and for securing a share of the market (Schurman 2004, 258). Each product is undergirded by large capital investments in the development stages, and so any potential increase in marketing costs can be a threat to its viability. Because complex and lengthy authorization procedures increase the bur den of pre-market investment, industry generally opposes limits on the duration of products market authorization, regulations requiring labeling and traceability, and the potential development regulation at the EU level governing the coexiste nce of conventional and GM crops (Rosendal 2005, 88). Civil society opposition dates back to th e earliest days of the expansion of biotechnology applications, but the explicit focus against GMOs was slow to develop. Schurman asserts that most of what these grou ps did in the first fi fteen years of struggle involved the mobilization of counter-expertise (2004, 252; see also Vogel and Ansell 2006, 99). Within the field of agricultural produc tion itself, initially there were no strong lobbies in the case of GMOs, although some an alysts have observed in hindsight that conventional farmers were in a good position to act as incumbents (Schurman 2004, 248). The decisive factor in the absence of an organized pro-GMO l obby at ground level was simply that when the issue first arose, most varieties of GMOs were not crops widely grown in the EU (Kruzer and Cooper 2007, 1039). Before embarking on an analysis of the s ubsequent actions by civil society groups to organize the opposition, it is useful to consider one key question with regard to the expert base they were developing in the earl y stages: what was this expertise counter to? First and foremost, it countered an unque stioning reliance on industry-generated 39


information, whic h had disproportionate sway in terms of providing the research used in policy making (Rosendal 2005, 99). In this sense, organization on the part of civil society represented a challenge to th e input legitimacy of the EU institutions. In addition, it countered an exclusivity of focus on criter ia based in and calib rated through natural science. This second challenge speaks to bot h input and output legitimacy and is indicative of a push toward a more widely en compassing type of policy to come from the EU level. The theme of appropriate breadth and which disciplines are to be involved through expert representation continues to be brought up in current debate, because the meaningful integration of socio-economic a nd ethical concerns in the authorization procedure has proven difficult (EC official E 2009). One factor in the precautionary principl es popularity with the public and some institutional actors in Europe is undoubtedly the series of regulatory failures associated with food safety that plagued the continent in the late 1980s and the 1990s. The bovine spongiform encephalopathy (BSE) crisis in part icular is invariably linked to the framing of the GMO debate. Its consequences we re very damaging to the Commissions credibility as a guarantor of the safety of food products placed on the single market, because initially the EC adopted the British governments unconcerned position toward societal concerns over BSE (Vogel 2001, 24). Eventually, as the disease migrated to humans, the Commission was forced to ban the export of beef altogether. The BSE crisis discredited completely, in the publics view the regular system of expert-led risk assessment. In 1999 another food scare, this time originating in Belgium and concerning dioxins, reiterated these concerns (Vogel 2001, 23). The food scares, and the failures in the food safety regulation sector that they belied, had to be addressed through a major 40


reorganization of the Commissions capacities in this area and corresponding changes at the national level. EU-level changes included the creation of the Health and Consumer Affairs DG (SANCO) and the European Food Safety Authority (EFSA). Policy making, round one: Directive 90/220/EEC Perhaps the first question to ask about the policy process fo r GMOs is, how did this particular policy issue come to be h eaded by DG Environment when several other services could lay legitimate clai ms to it? As each DG brings to the table the flavor of its own priorities and particular expertise, it is highly significant from a path dependency point of view which of them will be charged with the main responsibility for controversial policy proposals. So much is this the case, that recently the EC established the possibility for shared responsibility in policy where more than one DG saw itself as directly involved. The track record estab lished through the chemicals policy case had its bearing in the delegation of this new t horny area: DG-Envir onment was given the leading role ( chef de file ) based on their experience in regulating dangerous substances and hazardous industrial activities (Ros endal 2005, 85). Jasanoff considers this the battleground where the dominant frame for th e topic was first established, in a power struggle among the directorates, in which DG X II (Research) ultimately lost to DG XI for Environment (2005, 79). Similarly, Lieberman and Gray see the dive rgent approaches as a precautionary one (DG ENV Environmen t) and a developmental one (DG RTD Research) (2006, 596). In Jasanoffs view, the comprehensive approach endorsed by DG Research through several specially designated bodies in the 1980s simply could not work with the vertically segmented structure of the Commission, so a narrower view of the applications of biotechnology was adopted (2005, 82). This comprehensive approach, 41


which attempts to treat applications in pha rmacy, agriculture, and i ndustry in a coherent manner and with equal degrees of regulatory control, is the current goal toward which biotechnology advocates are attemp ting to steer the EU (Moll 2009). Over the same period, while the various DGs attempted to establish their conception of the role of biotechnology in th e European marketplace, the industry began to organize itself and foster linkages with the institutions. During the 1980s, industrial interests became more unified, and by 1989 th eir organization, the Senior Advisory Group on Biotechnology (SAGB) and the Co mmissions Biotechnology Coordination Committee (BCC, whose main goal was to increa se EU competiveness in the area) were closely linked (Rosendal 2005, 91-4). On April 23, 1990 Directive 90/220/EEC on the deliberate release into the environment of genetically modified organisms was presented to and approved by the Council of Environmental Minister s. In this sense, the issue continued to be treated as falling under an environmental agenda, but no t for long. In fact, as a horizontal measure and thus a law that applies immeadeately to all actors in the EU without the need for transposition to national contexts, this Direct ve transcended the vertical structure of the Commission, but it also led to many further disagreements among member states. The first step in seeking authorization for the marketing or cultivation of a GMO in a particular state is the submission of an application to that effect to the national competent authority (CA) of that state. If the CA assesses the appl ication positively, it is forwarded to the EC and other member states who may object to it within 60 days. These objections launch a conciliati on phase, during which EC, member state, and expert representatives attempt to reach agreement. If agreement is reached, the application 42


moves on to a QMV vote in the Council, which should be the final authorization. In case a QMV decision cannot be reached in the Coun cil, the Commission is free to make the final determination. In essence, this dire ctive enshrined the principle of mutual recognition of assessment standards among member states in law. Eighteen GM products have been approved for marketing, and one for cultivation, under 90/220/EC (Skogstad 2003, 327). The maize vairety approved for cultivation, known as MON810, remains the only approval of this kind and is primarily grown by animal feed producers in Spain. This concentration of GM feed gr owers has led to the consolidation of Spains positi on as solidly pro-GMO in term s of voting behavior at te EU level. Abels cites mutual distrust of national agencies risk assessment procedures as the main cause of the insurmountable fr equency of comitology conflict under the procedures set out for approval of GMOs by this directive (2002, 6). The directive also provides, in keeping w ith the precautionary principle, an article that allows member states to suspend the au thorization of particular GMOs within their own territory if there are new indications th at it poses risks to human health and the environment.8 The justification for such a suspen sion is left ambiguous within the directive, but the procedures laid out in it require that scientif ic data underpin all decisions about risk. Interest entrenchment 8 Article 16: 1. Where a Member St ate has justifiable reasons to consid er that a product which has been properly notified and has received wr itten consent under this Directive constitutes a risk to human health or the environment, it may provisionally restrict or prohibit the use and/or sale of that product on its territory. It shall immediately inform the Commission and the othe r Member States of such action and give reasons for its decision. 2. A decision shall be taken on the matter within three months Directive 90/220/EC, UriServ/ i=CELEX:31990L0220:EN:HTML 43


Meanwhile concerns over the degradation of farmland led to the creation of novel legislative instruments in 1992. According to Kruzer and Cooper, Brussels thus pushed a sociocultural redefinition of the CAP that tied it to environmentalism, landscape preservation, and animal welfare (2007, 1040). The redefinition entailed national-level policies to extensify rather than inte nsify farming, whereby less taxing farming techniques were to be given preference. Correspondingly, several new legal instruments established special status pr oducts. Northern Europe benef ited from the protection of organically produced foods under Regulations 2092/91 and 2078/92, while the Mediterranean gained special recognition fo r traditional products designated under the protected designation of origin (PDO), protected geographical iden tification (PGI) and traditional specialty guaranteed (TSG) system9 (Kruzer and Cooper 2007, 1040-1). These legislative changes brought into being a contin gent of farmers with a stake in protecting their particular agricultural practices from the threats of intensification and possible contamination by varieties of lesser or sign ificantly different quality as a matter of economic interest. In the years to come, the interest group constructed around the protected status of these agricultural produc ts would be of key importance for swaying their respective states nationa l positions on GMO authorizati on and for reinforcing civil societys claims directed at the EU level of governance. The climate in the public sphere change d dramatically in 1996, when US firms first began to send shipments of GM crops to Europe. Greenpeace, who would prove key in assisting smaller organizations to stra tegize because of its strength in balancing long 9 These regulations provide for products produced in certain regions, or according to certain regionallybased processes, to be trademarked and protected from unfair competitive practices. Refer to Appendix 1 for map of European states holding such certifications. 44


and short-term goals, symbolic and institutional politics, levels of policy, and diverse frames (Rosendal 2005, 104), took immedi ate action by meeting and protesting the shipments in rowboats. The concerns articulated by Greenpeace include not only those based on the general principle of precauti on when it comes to the environment and consumer safety, but also larger, philosoph ically based concerns with the de facto ownership of life forms that results from the patenting of GMOs. Although other opponents to GMOs did not pick up these obj ections, the physical entry of GM products in the EU provided a mobilizing opportunity that allowed them to emphasize the food safety aspects (Schurman 2004, 252; Rosendal 2005, 97). In 1997, Regulation 258/97, usually referred to as the Novel Food Regulation furnished implementation measures for the Di rective 90/220/EEC but more importantnly, it introduced a harmonised, two-step proce dure involving national food agencies and the Commission in lieu of the procedurally simple but controversial mutual recognition principle for approval of GMOs (Abels 2002, 8) However, problems continued to plague the GMO regime, as member states invoked sa feguard clauses against particular GMOs throughout 1996-1998. Faced with mounting criticism, the Commission came out with the White Paper on Food Safety, which recognized the need for input by experts from beyond natural science, but remained focused on organized interests (Jasanoff 2005, 76). This criticism seems well-f ounded because in the same year, Commission President Jacques Santer successfully insisted that an umbre lla organization be established to represent the biotechnology indus trys interests (Moll, January 2009). The creation of EuropaBio resulted in overall improvements in the internal unity of the sector (Rosendal 2005, 91). This development is a perfect illust ration of the discussion in Chapter 1, which 45


points to the Commissions activist role in structuring interest representation at the EU level. Attempts at coordinated activity by industr ial interests continued, as in 1999 seven prominent companies established Council fo r Biotechnology Information to promote knowledge about the indus try, but at the same time some companies began to withdraw, with three of the largest companies seeking to sell their agricultural divisions (Schurman 2004, 243-4). Spurred on possibly by the consolidation of in dustrial interests, and in reaction to their informational campaigns, civil societ y undertook its own initiatives. Kruzer and Cooper claim that the anti-nuclear discou rse, which relied on a combination of incalculable risks and minimal consumer bene fits, served as the model frame for the oppositions portrayal of the GMO issue ( 2007, 1043). More crucially, NGOs expanded efforts to persuade retailers that stocking GM products would be harmful to their sales. These efforts were rewarded when, in the summer of 1998, Iceland Foods, a large supermarket chain in Britain, initiated a domino effect of voluntary reta iler rejections of GM products (Schurman 2004, 254). This was not only a huge success for the opposition in terms of blocking GMO producers access to the market but an indication of its strategic cunning. The opposition understood the st ructural constraints of the food market and exploited them to its advantage. The food market is characterized by asymmetric dependence, meaning that retailers have a wi de variety of suppliers of food to choose from, while food producers and processors are much more constrained in the array of actors to whom they can supply their products In addition, food retailers are in intense competition with one another and for this reason they face enormous reputational costs. In other words, most food retailers are gi ant conglomerates, they have long-standing 46


relationships with food producers and can dictate what large commercial farmers grow (Kruzer and Cooper 2007, 1054). Although retailers are not inhere ntly averse to GMOs in themselves, their reluctance to stock GM products funnels the negative public opinion and results in a domino effect on the choices of food growers (Rosendal 2005, 99). Deadlock Starting in 1997, Austria, Luxembourg, Greece, France, Germany and Italy instituted unilateral bans ag ainst particular GM products, invoking the safeguard clause of 90/220/EC. These measures could have been justified within its provisions if the countries had found new scientific evidence of risks to the environment or human health posed by these products or th e technology in general. However, the Scientific Committees for Food, for Pesticides, and for Animal Nutrition (SCF, SCP and SCAN) investigated the matter and concluded that th is condition was not met by the unilaterally acting states (Victor 2001, 303). Even so, the Eu ropean Council, the fi nal arbiter in this type of situation, could not come to agreem ent on how the unilateral bans should be dealt with, and they remained in place (Victor 2001, 304). Adding to the gridlock problem, in 1999 the Environmental Council, as a consequence of two declarations by cohorts of resistant states10 to the effect that approvals should not continue in the absence of labeling and traceability mechanisms, decided to revise 90/220/EC to make it mo re stringent (Lieberman and Gray, 598-9). Even this minority of member states was su fficient to prevent QMV from being reached by the Commission on any approvals because many states routinely abstained. These 10 The French position was signed by Denmark, Greece, France, Italy and Luxem bourg; and the other was signed by Austria, Finland, Belgium, Germany, Netherlands, Spain, and Sweden. The declarations thus represented, respectively, 92 and 91 of 168 votes required for QMV (of a total of 237), indicating that at least 77 additional votes were rejections or abstentions in each case under consideration. 47


abstentions were partly motivated by the same concerns that prompted the moratorium: a lack of confidence in the technology or the regulatory regime, political discord at the national level, and a number of other potential reasons. Analysts who have attempted to link voting records to national pos itions have had a difficult time in part because of the large number of abstentions at each vote. Anothe r factor that frustrates such attempts is the inconsistency vis--vis stated national po sitions with which some ministers will vote on particular applications (EC Official D 2009). Despite these difficulties and the limited conclusions to be drawn from voting records in the Council, it remains clear that the dissent of several states and the ambiguous position of others was sufficient to prevent approvals from being grante d under 90/220/EEC after 1999. The scenario described above is quite si milar to the events that allowed for revision of chemicals policy. There also some states, Sweden in particular, took the lead in advocating a precautionary frame. In 1998, the Swedish initia tive was supported by Austria, Denmark, Finland and the Netherlands, driven at first by an advocacy coalition composed mainly of actors from their na tional bureaucracies [which] presented a joint paper for a broad strategic approach in an informal meeting of the Environment Council (Pesendorfer 2006, 104). In this case also the proposal had to be tempered with other interests because of the privileged position of the chemicals industry within the EU. The main struggle, in contrast to the GMOs case, played out in parliament where the precaution agenda and the competitiveness agenda were in conflict (Smith 2008, 76). Eventually the competitiveness discourse came to dominate, and the resulting regime still favored market principles. In the chemicals policy case, there was c onsistent input from 48


industrial actors through the policy making pr ocess and especially in the EP committee battle (Smith 2008, 79).The moratorium on GMO approvals was, by many accounts, what catalyzed the biotechnology industrys efforts to get actively involved in the policy process (Rosendal 2005, 93), similarly to the way in which the p hysical entry of GM crops from the US catalyzed civil society to coordinate its e fforts and expand its repertoire of action. Formerly, the industry had seen any regulation whatsoever as detrimental to its goals and perhaps believed that the single market agenda would preclude stricter approval procedures. Faced with a de facto moratori um on new approvals, firms began to engage in earnest not only consumers but also the EU institutions. Seve ral industrial actors conducted enormous informational campaigns that, in light of th e negative attention formerly attracted to the case by anti-GMO activists, served only to consolidate the publics opinion that the biotech industry was pushing products that were somehow undesirable (Schurman 2004, 253). The industrys informational strategy was premised on the deficit model, assuming that more information to consumers would make them feel safer. In reality consumers lacked not simply information about GMOs, but confidence in them (Rosendal 2005, 99) and the increased ex posure to the issue did not leave them with a more favorable opinion. Since framing and the utilization of salie nt discourses become more influential when uncertainty is articulated as an overarching concern, the legitimacy challenge can affect any of the actors seen to promote the source of that uncertainty Levidow, Carr and Wield affirm that the agricultural biotechno logy industry was itself facing a legitimacy crisis, and when informational tactics did not work, it became dependent on standards 49


established by regulatory bodies for legitimati on of its interests a nd products (2000, 204). The changes precipitated by the moratorium we re emblematic of an important turning point in the policy process: suddenly, former ly resistant industry found itself cornered, and understood its interests to be shared with those of (at least some factions of) the EU regulatory bodies. Interest distribution: networks of varied actors It is useful to describe, in a coherent manner, the distribution of interests in the GMO case. The approach employed in this info rmal exercise is similar to the one taken by Pesendorfer (2006, 99-103) in mapping out the generalized associations of actors of varied types around specific policy objectives The created categor ies do not refer to positions that are precise or constant over time, especially not pro-GM or anti-GM positions. The focus is instead on tendencies for these actors to act in favor of employing precaution with regard to GMOs or in opposition to the principle as a basis for the EUs policy on these products. For this purpose, only states with a significan t interest in GM cultivation are included in the anti-precaution network, and only states that have enacted bans on GM products are included in the pro-precaution network. Table 2.1: Interest distribution in GMO policy Anti-precautionary network Pro-precautionary network States Spain, Portugal, Romania Austria, France, Germany, Luxembourg, Italy, Greece, UK, Poland Umbrella NGOs, Industrial actors or local actors Europabio Monsanto, Bayer, BASF, Syngenta Greenpeace, Friends of the Earth National and regional level NGOs Institutional actors DG RTD, DG ENT DG ENV 50


External pressure The United States had, since the first prot ests against its pr oducts, attempted to exert its influence on national governments and the Commission so as to facilitate the entry of GM products into the European market. Understandably, as the de facto moratorium was officially recognized instituted for an unspecified period of time, the US intensified this pressure by appealing to an international forum. After the required consultations with the EU, the United States Canada, and Argentina each requested the establishment of a World Trade Organiza tion (WTO) Panel in August 2003 (Lieberman and Gray 2006, 593). Lieberman and Gray (2006, 603-5) offer thr ee interpretations of the moratorium (juxtaposed in Table 2.2 below) that are all justifiable by the observable events, and suggest that this ambiguity has been us eful for the Commission and national-level institutions in shirking direct responsibility. A fourth interpretation, with international law as its referent, emerged fr om the WTO Panels deliberation. Table 2.2: Interpretations of the moratorium on new GMO approvals Proponents 1) Media, NGOs, politicians 2) US, some EU states 3) European Commission 4) WTO Panel Scope Crops and food Crops and potentially food none Crops and food (not derivatives) Period 1999 to 2004 1998 to present never 1999 to 2003 Basis of conclusions Informal blocking by some states through QMV in Regulatory Committee (EC) and Environmental Council Sees Environmental Council decision to allow national bans as endorsement of moratorium Simply a hiatus; Previously approved products continued to be imported and experimental release approvals continued Failure to comply with obligation to carry out timely approval procedures resulted in undue delay 51


In the WTO questions of food safety are mainly governed by the Codex Alimentarius Commission, while the Agreemen t on the Application of Sanitary and Phytosanitary Measures (SPS Agreement) may a llow states to place restrictions on food imports provided there is scientific justifica tion. This justificati on must be established through a risk assessment of biological and ec onomic consequences of the introduction of the product (Victor 2001, 308). The 1995 case brought by the United States and Canada against the EU for restrictions on the import of beef raised with the use of hormones established two precedents of importance for th e GMO case: the ability to use scientific opinions from alternative sources, provided th ey meet the WTO panels expectations for qualification, and the Appellate Bodys d ecision that risk assessment cannot be disaggregated from risk management (Vic tor 2001, 312). The charges in the GMO case were multi-faceted: that the Commission and the states who instituted unilateral bans had violated the SPS Agreement as well as the Agreement on Technical Barriers to Trade (TBT) (Lieberman and Gray 2008, 40). There is no doubt among close followers of the dispute that the transatlanti c regulatory conflict over GM Os has exposed some of the ambiguities of the [international] legal fram ework for product standards, as it required the panel to arbitrate on a variety of ques tions (Cadot, Suma-Eisenman, Traca 2001, 11). However, in its final decision, the WTO panel was careful not to vali date either the US approach or the EU approach to GMO regul ation (by ignoring ques tions of substantial equivalence altogether) a nd confined itself to judging that the EU was only in violation of the SPS Agreements requirement to carry ou t its own approval procedures in a timely manner (Lieberman and Gray 2008, 41, 45-6). Quite predictably civil society activity and public opinion have also been influenced by the complaint to the WTO, but instead 52


of popularizing GMOs the dispute has unified the opposition around the cause of defending EU interests (Rosendal 20 05, 96; Lieberman and Gray 2008, 50). In addition, the WTO panel included a statem ent in its final ruling that ensures that the conclusions of the Biotech Products case do not preclude: a) the future suspension of approval procedures in Europe and b) the instigation of moratoria on approvals of GMOs in other countries (Lieberman and Gray 2008, 47). This specification is especially important give n the precedent-setting nature of WTO panel rulings. It indicates that the complainants have failed to achieve one of their likely objectives for bringing the disput e to the WTO to begin with. The WTO assessment notwithstanding, there is broad agreement among observers and EC officials that pressure from the US, espe cially in the blunt form of a case before a WTO panel, was a major motivator for compromise among the internal factions of the Commission (Lieberman and Gray 2006, 6067; EC Officials A, B and D 2009). Many take this result to indicate that a protec tionist agenda was i ndeed behind the EU institutions dilly-dallying on approvals. However, such a conclusion presents a logical fallacy, if not an over eagerness to find evidence of protectionism. Policy making, round two: Directive 2001/18/EC The initial proposals for the amendment of 90/220/EC were oriented in the directions of deregulation and precaution (Levidow, Carr, Wield 2000, 203). In addition, with the advent of debate on labeling and traceability, the commercialization of GM foods was transformed from an all-or-nothing ou tcome into a series of negotiable hurdles (Levidow, Carr and Wield 2000, 199). Greenpeace in particular capitalized on this 53


development, channeling its efforts toward the establishment of labeling requirements, rather than outright rejection of GMOs (Skogstad 2003, 331). The product approval process under Direct ive 2001/18/EC hinges on a scientific opinion by the European Food and Safety Au thority (EFSA) which addresses risks to human health and the environment (see chap ter 3 for details). The EFSA submits its opinion to the Commission and member stat es. The Commission must, within three months of receiving the EFSA Panels opinion, issue a draft risk-management decision as to whether the product is authorized for mark eting and if any specifi c restrictions should apply to it. The draft decision is submitted to the Council, where it must be approved or rejected by QMV, but this element of the approval process has pr oven elusive. In the customary absence of a QMV decision, the ap plication is returned to the Commission who then has sole authority to make the decision. So far, all GMO applications that reached this stage have been approved. There is a clear bias towards approval in the way this process is structured once an application is submitted for conciliation, there does not seem to be any way to reject it. Since the reformed directive still relies on this method, it does not present an improvement in terms of making decisi on-making more refl ective of community positions. In fact, there has never to date b een an approval through QMV, because it is simply too difficult to attain. There has no t been a rejection based on QMV either. A blocking majority would requir e 255 out of 345 votes, but on this issue abstentions still often account for 100 votes or more. Since these do not count as negative votes a blocking coalition seems im possible (Kruzer and Cooper 2007, 1053). With no qualified majority, the Council is unable to exert its influence on the decision, and the EC is 54


obliged determine whether to a pprove or reject the applicat ion to. The practice of doing so has come under severe criticism from NGOs, who cons ider the Commission to be acting in disregard of popular and national opinions, and perh aps outside the lines of its mandate (Vetier, 2009). The EC itself has informally discussed whether this practice is appropriate several times, and has concluded that it is because it makes no procedural violations in a regime that is functioning w ith the assent of member states and society, and presumably legitimate (EC Official E 2009). In addition to the di rective, two regulations11 were introduced in 2003, one superseding the Novel Foods Regulation (258/ 97) and another that defines the labeling and traceability requirements to be applied to GM products in the EU. Abels characterizes these measures as enacting a one door one key principle (2002, 7) because they are vertical instruments and t hus are effective immedi ately at the national level. The threshold established for adventitious presence of GM content is 0.9 percent, beyond which proportion the labelin g requirements come into effect (Rosendal 2005, 86). However, these regulations were not retroactive, leaving products approved before 2003 outside their reach. In addition, there are post-market monitoring requirements and timelimited approval for the market release of GM products, both of which increase the burden of evidence on the ap plicant (Levidow, Carr, Wield 2000, 203). Some see the labeling and traceability requirements as a co mpromise between the factions within the Commission that accommodates both the mark et and consumer protection agendas (Lieberman and Gray 2006, 596). 11Regulations 1829/2003/EC on Genetically Modified Food and Feed and 1830/2003/EC concerning the traceability and labelling of genetically modified organisms and the traceability of food and feed products produced from genetically modified organisms and amending Directive 2001/18/EC (see Ch.3) 55


A study funded by the European Commission12 canvassed opinions toward the GMO regulation regime prior to Directiv e 2001/18/EEC and conc luded that the technocratic model has led to an impasse [in reaching harmonized standards and the need for] redefining the policy problem (Levi dow, Carr and Wield 2000, 205). However, the directive can be said to have failed in this in tended redefinition. In Garcias analysis the main objectives were to extend and clarify th e scope of the Deliberate Release Directive, to improve the administrative procedures a nd their efficiency a nd to introduce common principles of risk assessment (2006, 4). Thes e objectives clearly fall short of redefining the policy problem, as they build on former de finitions of it (extending the scope of the former regime), introducing administrative improvements (refining the same approach), and merely restate the actual main objectiveto harmonize risk assessment and management standards. Ansell et al.s analysis of the significance of this last directive is similar, in that he argues it is largely a continuation of the mutual recognition principles around which the previous regime was organized and more revealingly, a compromise with preexisting structures (Ansel l, Maxwell and Sicurelli 2006, 340). In essence, the Commission was constrained in its scope of policy options by the need to incorporate the nationa l agencies created under the fo rmer regulatory regime into the new one (EC official C 2009). The conclusi on that the new directive is the result of path dependence and is congruent with the mutual recognition principle is troubling indeed, since at the outset of the debate it was precisely distrust and disagreement among member states on assessment standard s that necessitated harmonization. 12 The study examined public opinion in ten European countries as it was expressed in interviews, media coverage, official documents, and a focus group 56


The final element of the GMO regulatory framework established by Directive 2001/18/EC is a supranational body of scientific experts that is to advise the Commission on the technical aspects of pa rticular product applications. In the chemicals policy case, the EC attempted delegation to national agenci es with expanded stakeholder consultation, which were endorsed at both EU and national le vels, but the content of the information to be consulted was too highly technical, and the costs of this experiment were judged unacceptable (Pesendorfer 2006, 108). Eventually, the chemicals regime moved in the direction of EU-level coordina tion of standards with limited stakeholder participation. In the biotechnology policy area, there was agre ement among institutional and national actors that an agency was needed, and that th e need to harmonize standards persisted, but the argument centered on whether EFSA should have a regul atory or simply informational mandate (Buonanno 2006, 264). The debate and the outcomes of the EFSAs operation will be dealt with in more detail in Chapter 3. Observers and NGO actors have criticized the directive for perpetuating reliance on the deficit model because the improvements it offers over the former procedure are increased transparency of discussions and c onsultation of stakeholders, which have been slow to take root. It seems that critics are right in claimi ng that policy-makers are failing to distinguish between building trust and in creasing knowledge of th e substance of policy (Abels 2002, 12). Following the directive four of the la rgest producers of GMOs worldwide, Syngenta, Monsanto, DuPont, and Ba yer Cropscience, relocated their production facilities from Britain to the US in 2004 (Rosendal 2005, 90). This concerted pullout could potentially be in terpreted in two ways: industry fe lt so burdened by the regulatory 57


regime (whether in its current form, or in anticipation of coexistence regulation) that it relocated; or they thought the marketing challenges posed by a hos tile public, recently reinforced by the labeling and traceability requirements, could not be overcome profitably. However, we must be cautious not to make too much of this apparent retreat, since the same companies conti nue to be leaders in the worl d market, as well as in the share of GM food imports approved for the Eu ropean market. In essence, by relocating they were able to shield themselves to so me degree from locally driven negative press attention while maintaining their position in the market. Trouble on the horizon Although many of the demands made by thos e resistant to GMOs were met by the directive, a crucial field was left out, one that would pur portedly deal directly with possible contamination of non-GM crops. Co existence of GM and conventional crops, the real issue for those motivated by envi ronmental concerns or the protection of vulnerable crops for niche markets, was such a thorny one that the European Commission sought to stay well away from it. Consequently, coexistence is still re gulated (or is not in some cases) at the national level. According to Beckmann and Wesseler (2005), the decision to employ the subsidiarity principle and leave regulation of coexistence up to national governments will lead states to de vise diverse ex-ante regulation mechanisms (already in effect in many states) and ex-pos t liability frameworks (only a few of which have been established), although EU law prevents them from banning GM crops altogether. The justifications for this vari ation are almost univers ally agreed upon by farmers, EU officials, and non-governmental entities: farming conditions, configurations 58


and practices are themselves too multivariate across the territory of Europe to be encompassed by a single horizontal piece of legislation (Vetier 2009; Moll 2009; EC Officials A, B, C and D 2009). Despite the valid ity of this obstacle, disparities in state policies, with coexistence regimes ranging fr om no regulation to a few meters of buffer zone, to several kilometers (Beckmann and Wesseler 2005, 14) create differently favorable regulatory environments, and leav e room for GMO applic ants to pick and choose where they initiate th e process of application appr oval. Taking a more optimistic view, Jasanoff sees this as an accommodation t echnique, with the subsidiarity principles application in GMO coexistence regulation se rving as a metaphor fo r the coexistence of different civic epistemologies (2005, 280). A number of issues have arisen that are even more pressing than coexistence regulation. Although a threshold fo r adventitious content of GM crop was established for food products in order to proceed with labe ling regulation, the same was not done in the area of animal feed. This presents a very urgent problem for the EU, as it depends on foreign feed producers for up to 87 percent of its needs (Moll, 2009) and is currently forced to reject feed with any GM content. In addition, Ro mania had a sizable population of farmers producing GM soy for animal feed w ho were forced to search for alternatives when the state acceded to the EU in 2007. Th ese farmers, and livestock producers more broadly, have begun to engage EU institutions in attempts to arrive at a compromise similar to that in GM food products. Countervailing them are interests clamoring for labeling of animal products from animals r eared with GM feed (Vetier 2009). Similar contestation is underway in the area of importe d plant seeds. These multiple interests and the differentiated approaches they advocate ha ve led to stagnation at the EU level, but 59


have so far not attracted the attention of the public as broadly as the GMOs in food debate. Conclusions How far were the motivations driving th e GMO regulation controversy in the EU centered on economic interests? This questio n must necessarily find its answer in juxtaposing economic and other types of fact ors that motivated particular actors and catalyzed the key developments in the case. The economic dimensions seem most salient in the discourse surrounding GMOs and the goa ls envisioned behind pr oposed legislation at three points: (1) at the outset of the debate when bi otechnologys role in the EU economy was still unchallenged, (2) when the pr ospect of labeling allowed for alternative products to benefit from this special status, and above all, (3) when US trade interests became a factor in catalyzing greater coop eration within the EU. In other words, undoubtedly the economic and distributional co nsequences of GMO regulations entered member states as well as the Commissions co nsiderations, but they were not cause for a concerted strategy in either th e precautionary or the opposite direction. The very lack of concerted action in the EU, and even within the Commission, indica tes clearly that the explanation of the complications in the GMO regulation story is not simply that they were a protectionist ruse. In the end, did reform of the GMO re gulatory regime resolve the original problems that necessitated it? The short an swer may appear to be no because disagreements persist in many areas among member states and among a host of other actors, and there are significant impl ementation deficits (see Chapter 3). However, the new regime has been functioning for over seven years now, and challenges to its 60


legitimacy, whether from within the EU or fr om the WTO, have not been sufficient to derail it. Scholars hold varied and contrasti ng conclusions as to what the reforms to the GMO regulatory regime actually amount to. It appears that democratic principles were sacrificed to the primacy of science although some scholars affirm that the very opposite took place (Skogstad 2006, 234). It may be useful for the purposes of a general conclusion to retrace the reform process st ep-by-step, matching up the turning points identified above with shifts in the mode of policy making being employed at those junctures, in order to clarif y the chief considerations that affected the final outcome. Helen Wallace (2005, 77-89) suggests that, if one takes a holistic view of EU policy making, it is possible to identify seve ral policy making modes that the community has used over time. The modes are proposed to have a relation to the substantive policy area they are usually used in. The main char acteristics of the modes of policy making are summarized in Table 2.3. Even at first glance, the regulation of GMOs entails all but one of the policy making modes, especially when these modes are considered in light of the usual types of regulation they entail. Although the implications of coexistence regulation for agricultural policy are carefully being avoided by the EU institutions, the logic of competitiveness, the research-driven econo my, and the fact that biotechnology represented a novel policy area already situate the GMO is sue in at least three different possible modes. Therefore a more detailed look at the steps taken to arrive at the final regime is necessary in order to situate th e GMO case more accurately in this framework of modes. 61


Table 2.3: Policy making modes (charact eristics relevant to GMO case are highlighted) Mode Policy making Policy execution Involvement of other EU institutions Stakeholder / expert relations National level involveme nt Cases (A) Community method high threshold for revisions EC EC, delegation to agencies Council of Ministers provides issue linkages, ECJ arbitration Coopted by EC Engageme nt of national agencies, not much role for elected officials CAP (B) Regulatory mode moving up to EU level to escape domestic constraints EC (policy objectives) Council of Ministers (standards) Executed at national level Growing role for EP, ECJ active in enforcement Extensive consultation, cooptation by EC, noneconomic interests channeled through EP Mutual recognition principle Economic integration, competitiven ess More effective for products, not processes (C) Distributional mode / multilevel governance EC designs the programs Budgeted in Council of Ministers MEPs begin to represent regional interests Coopted by EC (subnational competitio n for funds) Cohesion, structural funds, Research frameworks (D) Policy Coordination / OECD technique EC sets up epistemic communiti es that bridge institutions Integrated topically Integrated topically Coopted by EC, large role for independent experts Used to extend mandate claim over new policy areas = unclear political ownership (E) Transgovernmentalism European Council and Council of ministers Cooperation by experts across nations EP and EC marginalized excluded Key national policymak ers involved NATO, space program Initially, the GMO question was only regarded by the Commission as an aspect of the EUs research program (in the 1980s). The first legislative act specifically dealing with the issue, Directive 90/220/EEC, was ch aracterized by a lack of clear political ownership by any particular st ates or DGs. Both of these are elements of the policy 62


coordination mode (D), which furthermore corresponds most closely to one of the enduring objectives of the GMO regime, to harmonize the standards for approval of GM products. However, the more typical regulatory mode (B) is also descriptive of events in the GMO case, in particular the growing role of the ECJ as interpreter of EU legislation (see Chapter 3) and the increasing pressure from the EP to adhere to transparency and accountability standards. Other elements of the regulatory mode of policy making are evident in the attempt to maintain mutual recognition as the principle underpinning national-EC relations and in the extensive consultation with stakeholders leading up to the revision of 90/220/EC. Finally Directive 2001/18/EC moved the regime closer to the original community method (A ) by assigning implementation responsibilities to the Commission with the assistance of an agency network at the EU and national levels (see Chapter 3). The policy coordina tion (D) thrust was still present, however, as indicated by the segmentation of implementation responsibil ities and the special role of independent scientific experts in the regime. In other words, the GMO policy making process exhibits elements of a regulatory mode and policy coordination with an outcome that would normally characterize the community method b ecause of the high threshold for revision, which is indicated by the reaffirmation that despite continuous inability to reach QMV decisions, the system is functioning as intended (see Chapter 3). But what does this mean outside of the narrow context? The GMO case can be taken to indicate some likely trends for si milarly thorny policy areas in the future. Clearly, the intersection of many agendas that are typically segmented in the EU governance structure (market integration, agriculture, environmental policy, consumer protection) necessitated experi mentation with different polic y making modes as particular 63


obstacles arose. The salience of both th e economic dimension and the precautionary concern prevented biotechnology applications in food from being compartmentalized. The inability of member states to agree on th e appropriate frame in which to weigh these contradictory approaches led to deadlock that could only be overstepped when it was made very urgent by the US bringing a case be fore the WTO. We can likely expect that in the future novel policy areas, especially ones involving technol ogical applications surrounded by uncertainty, will also encount er coordination problems across member states. It is difficult to predict which policy areas are likely to lead to contestation in the EU system, since biotechnology applications other than foods have gone unopposed. It may be that as DNA manipulation yields market able applications, such as artificially produced human organs, or applications to be used in the production of food or medicines, these topics may become more c ontentious due to the existing spectrum of diverging views based in ethics religion, and profitability. If the hypothesis that it was the regulatory crises in food safety that primed the public for mobilization against GMOs is correct, then policy coordination problem s may be exacerbated as the discourses employed in such cases build on each other, creating lasting strategies and networks of mobilization that respond to contestation of EU governance. Another aspect of GMOs in food considered a precondition for contestation is the immediacy of the concern for EU citizens as consumers, unfiltered through indire ct or secondary relations to the single market agenda. If the EU institutions exte nd their reach into social policy, or even security policy, more issues may arise that have this urgency for citizens, for instance the use and storage of biometric information. 64


CHAPTER 3 Implementation of EU policy: agency networks In considering the implementation of the GMO regime, it is necessary to follow the dispersion of responsibil ities among EU and national act ors. Broadly speaking, the implementation of the product approval pro cess (established unde r the horizontal measure 2000/18/EC) is carried out by th e Commission, while the implementation of post-authorization requirements (under several vertical instruments) is mainly the responsibility of national governments, with provisions for oversight by the Commission. In the case of product approval, the EC has s een fit to establish a supranational body, the European Food Safety Authority (EFSA) with advisory power in matters scientific and has similarly required the establishment of co mpetent authorities (C As) in each member state as a prerequisite to fulfillment of post-authorization requirements. Because delegation to expert agencies seems to have become the modus operandi of the European regulatory machine, it is necessary to lay dow n a foundation for the analysis of this tool for governance. Delegation theories applied to the EU level There is no single definition of the te rm agency as ap plied to governance structures, and the definitions vary by field of study, because different characteristics and aspects of agencies are emphasized across disc iplines. Here, the term refers to bodies set up by the Commission to carry out narrowly de fined and highly technical implementation tasks. Barbieri and Ongaro examined the cu rrent 21 such EU agencies and identified the following characteristics as typical: the agenci es have independent le gal status, externally determined statutes, are governed by an executive director and confined to 65


implementation tasks, have very limite d autonomy in financial, personnel and organization decisions, and are not subject to a performance contract (2008, 413-4). With regard to the European Union, there are two main trends in theorizing delegation to agencies, which largely mirror essential cleavages in scholarship focusing on the national level. One of these main stra nds is principal-agent theory, which has been used extensively by Kelemen (2002) in his a pproach to emerging el ements of the EU governance system. In his view, these agen cies serve to extend the Commissions influence and reach while remaining within the constraints placed on EC power by national governments through th e founding treaties (2002, 95). Based on this functional premise, Kelemen insists that agencies are only created when there is agreement among the principals (the Commission itself and memb er states) on the substance of policy, and are necessarily set up in a way that leav es legislative power to the EC (2002, 110). Looking more closely at the motivations of member states, Tarrant and Kelemen conclude that agency creation will occur wh en the distributional consequences are not detrimental to a (qualified) ma jority of states (2 007, 5). In other word s, delegation will occur when it results in ne gative outcomes only for a sm all group of states who are unable to block a QMV vote, and simultaneously in positive outcomes (whether direct or indirect, through linkage to othe r policy areas) for a group of states who are ab le to reach a qualified majority in favor It is also likely that the increasingly common use of delegation at the EU level may make more me mber states amenable to such arrangements in the future, increasing the potential pool of candidates for the latter group. The principal-agent model draws criticis m from a camp oriented toward more constructivist explanations. For instance, Borras, Koul atakis and Wendler (2007) 66


consider the principal-agent model too limite d for understanding EU agencies, because in their view top-down functional pr essures are not the only factors that have an effect on agency creation and design (585). Those who want to go beyond the principal-agent model usually argue that agencies become culture-scapes in themselves and a common commitment driven by professionalization give s rise to agency-spe cific interests (Hix 2005, 28). However, like the more narrowly focused proponents of principal-agent theory, analysts of this cons tructivist bent agree that the convergence of preexisting preferences in a particular policy domain is a condition for agency formation (Borras, Koutalakis and Wendler 2007, 586; Groenl eer, Boijn, and Kuipers 2005, 10). Constraints and controls Delegation to agencies gives rise to concerns about democratic control and accountability. These problems are addressed within the prin cipal-agent model under the names political and bureaucratic drift. Politi cal drift refers to the danger that future political office holders will dismantle or tran sform the agency to suit their own agenda, while bureaucratic drift refers to the ability of the agency to develop its own policy agenda and pursue goals that depart from the tasks originally designa ted to it (Kelemen 2002, 96). Some scholars argue that static cont rols, built into the design of the agency through its statute and procedures, can safeguard the principals from these perils (Majone 2000, 290-4). One key mechanism for control is the stipulation that agencies founding statutes are designed by the principa ls (Barbieri and Ongaro 2008, 143). The agencies created by the Commission have their legal basis in secondary legislation, or in legislati on created by the Commission, and not the founding treaties of the EU which have the explicit assent of memb er states. Due to the sovereignty concerns 67


highlighted above, the founding treaties clearly limit the powe rs of the Commission to delegate tasks to external bodies. This measure is also neces sitated by the EUs commitment to balancing not only the traditional tensions between legislative and executive powers on the one hand, and effici ency and democratic accountability on the other, but also betw een the interests of individua l members and the cornerstone integration agenda. An early case adjudicated by the ECJ in 1985, Meroni v High Authority, nonetheless defended the Commissions pr erogative to delegate within the following narrow constraints: The delegated tasks were the ECs responsibility and are limited to implementation only The Commission remains responsible for oversight of the agency Delegation can not result in cha nges in the balance of national and EU-institutional powers (Majone and Everson 2001, 143) This judgment became known as the Meroni doctr ine and is still the main referent for the constraints placed on the Commission and its delegated agencies. Its implications will be discussed below in reference to particular cases. Additional control measures th at can be used to ensure that agencies carry out their functions in accordance with the objectives for which they were created include public transparency requirements, performan ce requirements set by the principals, and accountability to relevant ju dicial and parliamentary bodi es through periodic review (Majone and Everson 2001, 148-50). Verifying the extent to which these controls are applied to EU agencies, on paper and in practice, is beyond the scope of this study. It is important to keep in mind that a broad de legation reduces decisi on-making costs since the principal does not have to invest resour ces in working out the details of regulation, 68


but it increases the cost of controlling th e agencys discretion (Majone and Everson 2001, 154), a relationship which is also influential in the design of agencies and control mechanisms by the Commission. Given all these stumbling blocks, why ar e there agencies at the EU level? Delegation to independent agen cies is frequently regard ed as a means of ensuring commitment credibility at the national level (Majone and Everson 2001, 132-4). The commitment credibility problem arises when regulation of a partic ular policy area is susceptible to politicization, in which cas e democratic governments themselves may come under pressure to continually revise it. Th e possible result is severe inefficiencies in the given sector resulting from regulatory defici ts in periods of adjustment. This scenario is quite akin to the events in the GMO case (described in Ch.2). Majone sees the parliamentarization of EU-lev el regulation as a serious thr eat to the credibility of its commitment to integration (2000, 285). Essentially, commitment credibility problems are the same phenomena as political drift, but re interpreted from the point of view of the electorate and not the principal. Because of their role in insulating regulators from shortterm political turmoil, agencies derive much of their legitimacy from their independence. Nevertheless, they must still be subject to democratic accountability, and this tension is the basis of the European Parliaments char ges that EU agencies are not sufficiently accountable and transparent (Kelemen 2002, 102). In the EU, delegation to agencies has al so been interpreted as a response to a capacity crisis in the EC due to the expansion of the institutions mandate into more and increasingly technical areas. This capacity crisis was exacerbated by the collective resignation of the Commission under Jacques Santers presidency amid a scandal about 69


appointments and qualifications (Peterson 2006, 86-7). The establishment of independent agencies can serve to privilege the role of specialized experts in deliberation on a given topic by insulating them from actors with purely political motiva tions (Kelemen 2002, 96). It could potentially also expand access to civil society depending upon the particular procedural and performance requirements (Majone and Everson 2001, 139). In addition, some scholars have pointed to the integrativ e benefits of specialized agencies in the European context: diverse national practices can be unde rstood as valuable expert resources for a different harmonization model rather than as deviations to be endured or suppressed. For EU regulation in general, local variations in environmental values provide an essential basis for European standards. (Levidow, Carr, Wield 2000, 205) This perspective suggests that divergent approa ches to particular scientific questions can be brought together in a more productive ma nner in the incubated environment of an independent agency than they could in an intergovernmental set ting, where political agendas are likely to prevail. In any case, we can not fully answer th e question why EU decision-makers resort to agencies unless we consider the alternatives. One alterna tive to an agency established by the EC is reliance on a self-regulating or ganization (SRO) which entail much lower transaction costs, but this set up presents its own problems. SROs are associations of enterprises involved in the same industry w ho come together to produce standards of conduct congruent with their ow n as well as governmental and societal views on best, or, occasionally, minimally acceptable, practices Majone and Everson note that SRO standards are non-binding and responsibility for them is diffuse, questions persist about the reliability of self-monitoring, and they face possibility of cap ture by the interests being regulated (2001, 130-40). Look ing at this alternative in the EU context, Tarrant and 70


Kelemen determine that the choice betw een a European agency and [an SRO] is typically, in its pure form, a choice be tween, on the one hand, an institution which combines majority decisions subject to a Commission veto and, on the other hand, an institution which features neither of these mechanisms (2007, 8). Empirical observations on EU delegation Observers have identified two waves of EU agency creation, each with its own set of design principles that became important poi nts of contestation. The first wave occurred in the period leading up to 1994, and saw the creation of six agencies in fields of socioeconomic policy. These bodies removed from the expanding workload of the Commission highly tec hnical, laborand resource intensive activities (Kelemen 2002, 101). The second wave of agency creation occu rred after the entry in to effect of the Treaty of Maastricht, which adjusted the ba lance of power between the Commission and EP in terms of influence ove r agency design and oversight. Accordingly, agencies created in this second wave have been subject to requirements for the inclusion of stakeholder representatives on their management boards, stricter statute defi nitions, and closer oversight by the EC and EP of their tasks and procedures (Kelemen 2002, 104-5). Some already existing agencies, such as the Eur opean Medicines Agency (EMEA), were also reorganized in order to facilitate th e transparency and accountability agenda However, in some cases the inclusion of non-state representatives comes without voting rights, as in the case of the European Chemicals Agency (ECHA), leaving the predominantly intergovernmental mode of decision making in place despite increased access to proceedings for particular designated stakeholder groups. 71


Because of the segmented structure of the Commission, which determines to a large degree the way in which delegated impl ementation by agencies is approached, it is very difficult to make general statements about this phenomenon. Most studies of EU agencies have been, until recently, focused on a particular policy area or a close comparison of several agencies. More and mo re scholars are now at tempting to derive generalizations and models applicable to the increasing number of EU agencies, in pace with the growing recognition that they have b ecome a lasting and characteristic feature of EU governance. In just such an attempt, Barbieri and Ongaro (2008) apply a public management approach to categorizing EU ag encies along several dimensions identified as constitutive of public agencies, in orde r to determine whether they can describe a European type of agency (398-400). They find a high degree of homogeneity in the features of EU agencies as compared to homogeneity found among agencies the national level (Barbieri and Ongaro 2008, 413-4). The implications of their findings appear to confirm the supposition that agencies at the EU level are little more than functionally specialized extensions of th e Commission. Clearly, the mode of control via statute design is employed by the EC, along with attempts to ensure that the executive director shares the principals objectives, as evidenced by appoi ntment procedures for this position (see Table 3.1). In addition, the Meroni doctrin e restrictions serve to protect the Commissions preponderance in policy ma king, including budgeta ry and personnel decisions. The picture painted t hus far is indeed of agencies of a very specialized and restricted nature, but it could perhaps be ma de clearer by consideri ng several particular examples. Product approval mechanisms: three multi-level agency networks 72


The preponderance of member state influe nce in the Council, whose approval is necessary for the creation of an agency at the EU level, use of the hub and spokes structure to ensure that these agencies re main reliant on preexisting national competent authorities (CAs) for particular steps in the implementation of product approval procedures (Tarrant and Kelemen 2007, 31). Ot her scholars have also insisted that avoiding isomorphism of the national CAs, or the mere dupl ication of form and function at the EU level, has been a way for EU agen cies to enhance their credibility (Groenleer, Boijn and Kuipers 2005, 19). The resulting s ituation leaves a lo t of room for the continuation of mutual recognition of nationa l standards in less c ontroversial areas and the performance of aggregative functions by th e EU agencies in areas where scientific judgments have become politicized. The following sections will examine the stru ctures and procedures that make up the three regulatory networks re levant to cases discussed in this study with a view to illuminating the implications of these networks design and practices for effectiveness and democratic accountability. Some key feat ures of the EU agencies concerned are summarized in Table 3.1. 73


Table 3.1: EU-level agenci es some key features 13 EFSA EMEA ECHA Tasks Review environmental and human health risk assessment, monitoring and emergency plans; Issue opinion to EC Coordinate evaluation by national agencies; Verify dossiers and testing proposals; Coordinate evaluation by national agencies; Prioritize substances; Publicize list of priorities Management board 10 member state reps, 4 from either consumer or industry organizations, 1 EC rep 27 member state reps, 2 patients organization reps, 1 doctors and 1 veterinarians organization reps, 1 EC rep, 1 EP rep 27 member state reps, 3 EC reps, 3 independent experts appointed by EP, 3 interest reps without votes Executive Director Proposed by EC, approved by management board Proposed by EC, approved by management board Appointed by management board Committees / procedures 1 Scientific Committee 10 Scientific Panels All appointed by management board CHMP, CVMP, COMP, HMPC, PDCO, CAT Made up of national reps Risk Assessment, Socio-Economic Analysis All appointed by management board Budget EU budget subsidy (debate on whether to become feebased) EU budget subsidy Fee-based (main part of budget) EU budget subsidy Fee-based (intended to be self-financing, but fails to be so) 13 Table based on Groenleer, Boijn and Kuipers 2005, Stanton 2005, and EU agencies webpages: EFSA/AboutEfsa/efsa_locale-1178620753812_WhoWeAre.htm 74


European Medicines Agency (EMEA) The EMEA is rather emblematic in studies of EU delegation, because it is on the freer end of the spectrum of autonomy from the Commission, and so it draws the attention of many scholars who study this m ode of EU governance. Nonetheless, it would be a grave oversight not to claim that the EM EAs history can be ta ken as representative of any general patterns. The context of its creation was exactly that predicted by ideal type models, namely one characterized by a high degree of pref erence convergence among member states and a low degree of public attention. Risk assessment provisions: Within the pharmaceuticals authorization regime, there are two distinct procedures. The decentralized procedure applies to traditional products (essentially all but biotechnology pharmaceuticals), and is still based on mutual recognition, meaning that for a large variet y of products the authority remains with member states. The decentralized procedure requires applican ts to submit their case to a national competent authority (CA) first, a nd once it has been approved, to five more national CAs. At this stage the additional national CAs may submit objections to the Commissions Committee for Proprietary Medi cinal Products (CPMP) within 3 months of receiving the application, and the CPMP must pronounce its opinion within 2 months. However, the CPMPs role in this procedure is only advisory. Objections to particular applications arose very frequently, and ar guably for political reasons, and time limits were frequently violated (Vogel 1998, 4). Formerly all products underwent the decentralized procedure, but with regard to biotechnology applications in the pharmaceuticals industry (referred to as red biotech by the industry) this coordination method failed to facilitate the necessary ha rmonization of national standards (Majone and 75


Everson 2001, 134). The centralized procedur e, instituted since 1987, was designed especially for biotech and high technology products, since Br ussels reasoned that it would be easier to harmonize standards that had not yet been created (Vogel 1998, 5). Under the centralized procedure, applications are made directly to the CPMP, which has 7 months to evaluate them. National CAs can then make objections within a month. Initially, the centralized pro cedure was also not binding. It was only after the Maastricht Treaty in 1992 that it became so, and the CPMP became, along with other relevant committees (see Table 3.1) a part of the EMEA In the end, for biotechnology medicines, the capacity and authority to conduct risk assessment rests with the national competent authorities (CAs), under th e EMEAs coordination, and the final decision on authorization still rests with the EC. Majone considers the EMEA and the pharmaceuticals regulatory regime the greatest success story of EC regulation, because he sees it as a credible EU agency whose establishment required a decrease in member st ate authority, but resulted in efficiency improvements to the authorization process for products under the centralized procedure (2000, 281). On the other hand, the EMEA has also drawn criticism, al beit not as ardent as that directed at some other EU agencies, for its failures in aspects that have become increasingly important in evaluating EU governance. Abraham and Lewis portray the EMEA as the epitome of transparency failure citing its refusal to provide them with assessment studies for a project aimed at di scerning whether a leveling down of national standards was occurring (1999, 1664). They also question the regimes credibility on the grounds that it presents signifi cant room for capture by indust rial actors, because narrow time limits and the motivations of national agencies who collect fees from the applicant 76


place the burden of proof on those contesting th e safety of assessed products rather than on the applicant (Abraham and Lewis 1999, 1665). An additional cause for criticism against the pharmaceuticals regime is the re liance of the EMEA on information provided by producers, which also forms the basis of the risk assessment by CAs. These elements of the frameworks design result in very clos e ties between regulators and producers that begin in the product development stages and carry through the application process. All of these elements contribute to what has been described by a minority opinion as a highly privatized form of corporatist regulation with the EU, in which [consumer] health interests are defined very mu ch within the industrys age nda (Abraham and Lewis 1999, 1665). European Chemicals Agency (ECHA) The former regime divided chemical su bstances into two lists, one of those exempt from regulation, and the New Substances List, but this system still did not encompass many chemicals and hindered i nnovation due to draw n out assessment and approval timeframes (Stanton 2005 232). This sy stem was comprised of three horizontal and one vertical instruments that combine market integration and consumer protection goals,14 but concerns that both of these goals are not well served necessitated reform (Hansen and Blainey2006, 270). The Registration, Eval uation and Authorization of Chemicals (REACH) regime, in effect since June 2007, is still base d on risk-assessment, with precautionary procedures envisioned for the most noxious chemicals. The Environmental Council adopted it, and it is e xpected to encourage the rapid substitution 14 Directive 67/548/EEC and Directive 1999/45/EC (classification, packaging and labeling of dangerous preparations); Directive 76/769/EEC (restrictions of marketing and use of chemical substances); Regulation 793/93/EEC (evaluation and control of existing chemical substances); Hansen and Blainey 2006, 270 77


of dangerous chemicals with alternatives safer for human health and the environment (European Commission 2006). The European Chem icals Agency was created in the same month in fulfillment of the REACH regime. Risk assessment provisions: capacity and authority to conduct risk assessment rests with CAs, under ECHAs coordination as in the pharmaceuticals case. There are provisions for ECHA to collect a pplication fees in certain cases but applicants have been able to manipulate these provisi ons so that they only pay ap plicable CA fees, creating a significant financial deficit for the agen cy (Dancet 2008). The final decision on application approval rests with the EC, who is also responsible for the oversight of downstream uses in the marketplace. Oversi ght is carried out through requirements for producers and downstream users to report quantities and purposes of chemical use or resale directly to the Commission (793/93/EEC). These requirements are clearly distinct from the GMO regimes, where post-market monitoring is not centralized. Hansen and Blainey see the ECHA and th e REACH regime as a continuation of horizontal decision making on the allowable standards for risk posed by chemicals, combined with a market-bas ed approach to information dissemination (2006, 278-9). Before its actual institution, Stanton claimed that the draft REACH regime represents an unstated application of the precautionary principle (2005, 235). There is general agreement that in the chemicals case, harm onization of national st andards is achieved ( Winter 2006, 60). The verdict on the incorporation of the precautionary principle is not so clear, because the regulation15 does not mention the precauti onary principle, at least 15 Regulation (EC) No 1907/2006 of the European Par liament and of the Counc il of 18 December 2006 concerning the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH), establishing a European Chemicals Agency, amending Directive 1999/45/EC and repealing Council 78


not explicitly [but uses] the term risk to describe the situation that may trigger management measures (Winter 2006, 61). Despite this ambiguity, the precautionary principle has become a part of the EU body of treaty law, and does not need to be reiterated in secondary legislati on in order to be applicable. European Food Safety Authority (EFSA) The EFSAs functions are limited to furnishing scientific advice with regard to specific decisions before the EC that fall under food safety and identifying emerging topics that policy-makers may want to consid er with regard to technology. A detail that attests to the very limited role of the EFSA is that it is termed an authority, and not an agency like the other bodies under consider ation, although it still shares the general characteristics of EU-level agencies discussed above. The EFSA works on issues in all areas of food safety, but this study confines itself to its functions in the GMO regulatory regime. The EFSAs creation was dire ctly preceded by and linked to several food safety crises in the EU (see Chapter 2). Although it b ecame clear as a result of these regulatory crises that more capacity to identify and a ssess risks and to manage them was necessary at the EU level, con cerns about delegating such enormous responsibility to a concentrated group of experts persisted. It was thought at the time that to give the EFSA a regulatory mandate was perhaps to concentrate too much power and discretion over this already thorny policy area in a single, expert dominated, entity. The Commission finally determined to limit the EFSAs mandate to informational purposes, citing among the reasons behind this decision the Meroni doctrine and the importance to maintain Regulation (EEC) No 793/93 and Commission Regulation (EC) No 1488/94 as well as Council Directive 76/769/EEC and Commission Directives 91/155/EEC, 93/67/EEC, 93/105/EC and 2000/21/EC 79


democratic accountability for food sa fety policy (Buonanno 2006, 264). Additional concerns were articulated by the Council, speak ing out in defense of subsidiarity, and the Parliament, insistent on high standard s for transparency (Buonanno 2006, 269). These concerns still resurface occasionally in regard to particular food safety policies. The final structuring and design of EFSA embodied an attempt to di saggregate risk assessment from risk management in order for the Commissi on to retain its legislative role but still benefit from a concentration of scientific expertise. The way in which a distinction between risk assessment and risk management is intended to work can be discerned in Box 3.1. Box 3.1: Definitions used by the EU inst itutions in the gov ernance of GMOs to delineate fields of responsibility 16 Handling of risk(s) should follow a structured approach, which consists of three basic elements Risk assessment can be described as a process of evaluation of the attendant uncertainties, of the likelihood and severity of an adverse effect(s)/event(s) [and] comprises four steps: hazard identification, hazard characterization, exposure assessment and the integrative risk characterization The risk assessment is a scientific exercise. Risk management is the process of weighing policy alternatives in the light of a risk assessment(s) and of other relevant evaluations, and, if required, of selecting and implementing appropriate control options. Risk communication is the interactive exchange of information and opinions throughout the risk analysis process concerni ng risk. It should i nvolve not only risk assessors and risk managers, but also cons umers and a wide rage of other actual and potential stakeholders. The European Food Safety Authority, and the attempted disaggr egation of risk assessment from risk management, have drawn criticism by observers and applicants for failing to present an adequate adjustment from the previous regime. For instance, although risk communication is supposed to be ongoing and all-encompassing, and 16 Updated Guidance Document for the Risk Assessment of GM Plants and Derived Food and Feed. The EFSA Journal (2008) 727, 14-15. 80


responsibility for it presumably shared by all, it is not clear what types of information should be communicated and under what condi tions. More importantly it is not clear whether there is expected to be any type of feedback from risk communication into the structures that manage it, which means th at communication can easily flow one way only as predicted under the deficit model. The ar tificial divorce between risk management and assessment has also been interpreted as an obs tacle to the standardiz ation of assessment measures (Majone and Everson 2001, 135). Risk assessment provisions: Under the provisions of Regulation 1829/2003/EC, companies wishing to apply for authorization for market release of GMOs must approach a national competent authority. The CA has 90 days to compile an assessment report. The CA must inform the EFSA of such applicati ons, and the EFSA transmits this information to all other member states and publicizes it It includes a techni cal dossier containing methods of detection of the GM content and sc ientific studies carried out in the process of assessing risks to consumers and the envi ronment. If the application is found to provide insufficient information, a request for additional data can be made of the applicant by either the EFSA or the CA. Af ter the initial CA report is completed, other member state CAs may submit reasoned obj ections, and unless th ese are withdrawn during a period of conciliati on, their particular causes ar e forwarded to the EFSA. Once the application has been submitted to the EFSAs Scientific Panel on Genetically Modified Organisms, which can task the competent authority with conducting further specific assessments, the Pa nel has six months to issue an opinion. Upon completing its evaluation, th e Panel issues an opinion in official advisory capacity to the Commission and member states. The EFSA Panels opinion must also be 81


publicized and citizens may make comments on it to the Commission within a month. Thus far, of 48 opinions on the safety of GMOs for food and feed us es, all were judged as safe as conventional varieties and un likely to have adverse effects on consumers and the environment; only one opinion indicate d the Panel was divided in its conclusions, and only two opinions recommended furt her environmental risk assessment.17 Regulation 1829/2003/EC also specifies the conditions under which the EFSA can be challenged (Art. 36): Any decision taken under, or failure to exercise, the powers vested in the Authority by this Regulation may be re viewed by the Commission on its own initiative or in response to a request fr om a Member State or from any person directly and individually concerned. To th is effect a request shall be submitted to he Commission within two months from the day on which the party concerned ecame aware of the act or omission in question. t b Assuming full transparency of the procedures and opinions of the EFSA, citizens and states can hold it to account within the legislative framework. However, these ideal conditions rarely obtain within the time lim its for submission of concerns, so the Commission is the chief principal to whom the EFSA must account. For the most part their relations are amicable, because EC offici als are forced to bala nce their expectations of the EFSAs performance with concerns that challenging particular Panel opinions is tantamount to undermining a regime that th ey worked very hard to establish. This situation can make individuals in certain positions quite influential. One example is Commissioner Stavros Dimas, w ho was active in framing the coexistence debate in terms of farmers freedom to choos e which crops to grow and insisting that subnational voluntary associations of GM-free farmers should not be repressed (Dimas 2006, 5). More importantly Dimas decided in 2007, with the support of several member states, 17 Data from the EFSAs Register of Questions, which catalogs all questions posed to it by EU regulatory authorities. Only adopted opinions were included. Withdrawn applications and ones in progress were excluded. 82


to challenge the EFSAs opinions on two varieties of GM maize proposed for marketing and cultivation (Vetier, 2009). After compulsory reconsideration, the GMO Panel reiterated its opinion that the varieties we re as safe as conventional maize; it was subsequently requested by the Commission to c onsider evidence to th e contrary from 11 scientific studies conducted by independent experts, but the Pa nel still insisted that these did not contradict its conclusions (EFSA, 29 October 2008). These developments precipitated internal tension in the EC, and generated the agenda for discussion at its periodic orientation debate wh ere officials from across servi ces come together to discuss common positions on topics related to GMOs (EC officials B and E 2009). As a result of such conciliation exercises, most EC offici als maintain publicly that since the Dimas episode genuine attempts have been made to balance the input of expertise to the EFSA and as a result EFSA is trusted, period (EC Official C 2009). Most of the actors identified in the an ti-GMO network (see Chapter 2) still consider the EFSA biased in favor of bi otechnology because of the composition of its panels, and more generally because it has ne ver delivered a negati ve opinion. Its track record may not necessarily belie a bias if a pplications are only submitted for products that meet the EUs criteria, a condition that is likely to obtain given the biotechnology industrys high capacity to make such strate gic distinctions. However, the challenges by Commissioner Dimas weaken this argument. In addition, Greenpeace has furnished the GMO Panel with alternative scientific studies and challenged the validity of particular studies used in its risk assessments, but to no avail ( Vetier 2009). As for the biotechnology indus try, generally there are no complaints as to the scientific soundness of the EFSAs opinions an d a cooperative relationship seems to have 83


emerged. Industry actors nevertheless frequently target the EC in their complaints over delays caused by its requests for further info rmation, because when these occur the clock is stopped, and time elapsed in furnishing the information does not count toward the deadlines to which the Commission and EFSA are subject. Beyond EU-level product ap provals: an implementation network for GMOs Following the Santer Commissions resignation, Romano Prodi became President of the newly formed Commission, which underwent restructuring to bri ng it closer in line with its expanding responsibilities and the expe ctations of the public One change was the creation of the Health and Consumer Af fairs DG (SANCO). Although the Environment DG (ENV) had formerly headed all matters re lated to GMOs, the aftermath of the BSE crisis and the reactive surge of distrust led to GMO regulation being housed under DG SANCO (Levidow, Carr and Wield 2000, 2001). Despite this change though, implementation of the regulation pertaining to cultivation of GMOs remains the responsibility of DG ENV, while coexistence regulation, if it ever materi alized at the EU level, would fall under DG Agriculture (AGRI) s responsibilities (s ee Table 3.2). This division of labor mimics the compartmenta lization that normally leaves policy areas within the purview of a single DG. However, since the DGs also have an incentive to protect their influence over relevant policy ar eas, optimization of ta sk delegation is not necessarily a priority. Herrick offers a simple explanation for the trifurcation of implementation responsibilities for policy concerning GMOs: since GM crops are simultaneously and inextricably tied to fears over human health and the environment, risk management will always be in contested territory (2005, 288). Since not all of GMO policy falls neatly within a segment of the Commission, it was necessary to distribute 84


implementation responsibilities among severa l DGs. However, the substance of the policy is not the only cause for this compli cated scheme; path dependence is always a factor in EU governance, because each compromise requires considerab le conciliation. In addition to being treated under several DGs, the present GMO regime presents a classic case of member states insisti ng on a continued role for the CA s that the previous regime required them to create and fund (EC offi cial B 2009), leading to the hub-and-spoke structure predicted by Tarrant and Kele men (2007) in post-market measures. Table 3.2: Differentiated pathways im plementation of the GM regulatory framework Legislative act and substantive area 2001/18/EC Research Purposes (field trials) 2001/18/EC Cultivation of crops Regulations 1829/2003/EC and 1830/2003/EC Market release (food imports) Guidelines from EC [no legislative act] Coexistence of GM and conventional crops Executive National CAs DG ENV DG SANCO DG AGRI Principle Subsidiarity principle Precautionary principle Single market principle Best practices, not binding Procedure Manufacturer applies to national CA that can authorize research after consulting the public 1. Manufacturer applies to national CA and EFSA, submitting environmental and human health risk assessment, monitoring and emergency plans. (If rejected, can apply in another state) 2. Conciliation with other states (EC) 3. Co-decision: Council must reach decision through QMV (if not, EC makes the decision) 1. Manufacturer applies to national CA and EFSA, submitting environmental and human health risk assessment, monitoring and emergency plans. (If rejected, can apply in another state) 2. Conciliation with other states (EC) 3. Co-decision: Council must reach decision through QMV (if not, EC makes the decision) N/A 85


In order to make the GMO regulat ory framework enforceable, DG RTD established the European Network of Gene tically Modified Organisms Laboratories (ENGL) in 2002. This network consisted, initiall y, of 45 laboratories whose tasks were to develop and validate methods for tracing GMOs, and thr ough central coordination to furnish best practice examples (European Commission 2002a). The ENGLs purposes were publicly framed in terms of protecting consu.mers right to c hoose certain products by establishing reliable testing facilities to execute labeling re quirements (European Commission 2002b). Despite the central coordination of this network and its expansion with the growth in Union membership, its role is limited to matters concerning scientific methods of the laboratory facilities. The implementation of post-market mechan isms is the responsibility of national governments with oversight from DG SANCO. Since its creation in 1995 DG SANCOs scope of responsibilities and the demands pl aced on it by national regulators have simply exploded. Because of the ECs general imperative to extend the integration agenda into new policy areas, new areas of legislativ e development are prioritized over the implementation of existing requirements. This leaves, in practice, a lot of discretion to the designated CAs, who are requi red by 2001/18/EC to monitor GM content in marketed products through practices out lined in Regulation 1830/2003. Implementation of post-market GMO labeling and traceability DG SANCO conducted several missions to assess the implementation of the GMO regime from 2005 to 2007. Each mission involved a small number of states, and each particular inspection involved one or a few CAs within that state, as well as a review 86


of the general legislative and institutional pr ocedures in place. The results of these missions are available through the Food and Ve terinary Offices database of inspection reports18 (see Appendix 3). In all cases (which include 15 countries in a total of 17 instances France and Italy, significantly, having been inspected twice each) the mi ssions found national provisions mostly adequate. Ne arly all of the member states that were examined had deficiencies in sampling met hods vis--vis the Commission s recommendations in this regard, and several states had allocated inadequate financial and human resources. Interestingly, in 2006 the UK still did not have its own laboratory and sent samples for testing to other member states. These im plementation deficiencies were addressed through recommendations made to the member states, to which they were required to respond with action plans. One naturally wonders how these particular states were chosen for inspection. No rationale is provided within th e mission reports, alt hough it is reasonabl e to assume that former knowledge within DG SANCO about the progress of implemen tation across states informed its choices. A pattern also emerges when considering these choices in the larger context of GMO policy: the overlap is not pe rfect, but 10 of the 15 inspected states were active in articulating a strong position for or against the GMO regime (Table 2.1). By focusing its inspections in stat es that have been outspoken in defending one or the other position, the Commission can point to its equal treatment of both camps and its effort to even the playing field for GMO products in accordance with its regime. Presumably, it was concern about the ade quacy of the action plans submitted in response to the first inspection and their overall lack of w illingness to comply with the 18 87


GMO regime that resulted in Italy and France being inspected repeatedly. Both of these states had also invoked safeguard clauses, and both hold a high number of PDO and PGI licenses (see Chapter 2 and Appendix 1). In addition, France is a special outlier because its failure to transpose Directive 2001/18/EC into national le gislation, from the entry into effect of the directive in 2002 until 2008, eventually resulted in a lump sum sanction by the ECJ of 10 million Euros (European Court of Justice, 2008). Making a point: safeguard clause invocations and general bans There have been nine invocations of the safeguard clauses under Directives 90/220/EEC, then 2001/18/EC, and one unde r Regulation 258/97, then 1829/2003/EC (see Chapter 2 for details). In addition, an Austrian province banned the cultivation of GMOs in 2003, and Poland issued a genera l ban on GM seeds in 2006, which is not aimed at a specific product and thus can not be defended under the safeguard clauses. Since these bans were not premised on a safeguard clause for a part icular product, after initial informal attempts to dissuade the stat e from this course of action the Commission determined to initiate infringement proceedin gs before the European Court of Justice. The Austrian case resulted in an ECJ ruling th at the measure must be terminated, while judgment on the Polish ban is still pending. The Commission and particular stakeholde rs also contested the invocation of safeguard measures against specific GM produc ts as political actions to prevent the single-market aspect of GMO regulation from coming into effect. The first step in contesting a safeguard measure is to submit the decision and its justifications to the EFSA for an evaluation, since the justifiability of the measure turns on the availability of new 88


scientific information that would contradict the original risk assessment for the particular product. In all cases published so far, the GMO Panel has not found this condition to be satisfied by the state enacting the safeguard measure. After this initial step, the Commission will typically attempt to persuade the member state to terminate the measure. In cases where the measure remains in place, the Commi ssion has the option of submitting the case to the ECJ for infringement proceedings. The ECJ can rule on the appropriateness of the safeguard measure as we ll as on which party s hould bear the costs resulting from it. However, the Commission has not resorted to this procedure in cases where a particular product wa s concerned, reserving it for general bans only (Anonymous Commission offici al (d), 2009). Table 3.3: Safeguard clause invocations and general bans (highlighted) and their consequences 19 Date of invocation Invoking member state EFSAs GMO Panel opinion / ECJ proceedings Feb 1997 Austria No new evidence; product withdrawn by EC in 2007 Mar 1997 Luxembourg No new evidence; product withdrawn by EC in 2007 Nov 1998 Greece No new evidence; product withdrawn by EC in 2007 Jun 1999 Austria No new evidence May 2000 Austria No new evidence Jan 2005 Hungary No new evidence Jan 2006 Greece No new evidence Apr 2006 Austria No new evidence Apr 2007 Germany N/A Measure withdrawn May 2007 Jul 2008 Austria Not published Mar 2003 Austria EFSA: No new evidence ECJ: Joined Cases T-366/03 and T-235/04, ECJ ordered the action dismissed; Austria appealed the decision, but the appeal was dismissed (Joined Cases C-439/05 P and C-454/05 P) May 2006 Poland EFSA: N/A ECJ: EC initiated infringement proceedings Apr 2008 (Case C-165/08) 19 Table compiled from data found in searches of EU databases Rapid (press releases) and CURIA (recent case law, and EFSAs database of opinions 89


Summary and implications Implementation of the post-market stages of GMO regulation, which cover all but the single variety approved for cultivation in th e EU, is in the hands of member states and they can determine the resources allocated to it and degree of vigilance that is exercised. Despite generally satisfactory results, the Co mmissions inspections of national facilities revealed that many were lacking in both respects, with inadequate resources and sampling practices being the most common cr iticisms to CAs. The vigilance of the competent authorities does not come under scrutiny in a regularized manner, and the growing demands put on DG SANCO may preven t it from addressing the issue in the absence of a crisis. The Commission is furt hermore conservative in challenging unilateral safeguard clause invocations, probably for f ear of catalyzing anothe r deadlock like the one that occurred in the late 1990s. But in the meantime, uneven implementation leaves room for products that should be subject to labeling and traceability requirements to fly under the radar in some states. Conclusions By most accounts, harmonization of the standards employed in the regulation of pharmaceuticals and chemicals is, if not a fa ct, well underway. In both of these policy areas, industrial actors have been supportive of the new regimes coor dinated at the EU level. The same is not true with regard to GMOs. These differences, and the wide variation in national positions in the area, have prevented the same level of harmonization from being reached beyond the very narrow scope of legislation governing GMOs in food. 90


In the GMO case, implementation inspection reports seem to indicate that policy coordination has been effective under the na rrow scope of Directive 2001/18/EC and the accompanying regulations, especially after Fr ance was forced to fully transpose the legislation into its national code. However, this framework only cove rs GMOs in food on the market and for cultiva tion, and agreement on other contentious issues remains elusive. This inability to extend the re gime indicates that there has not been harmonization in states episte mological or ethical approa ches to the technology, and there is no unified basis on which to build sta ndards in the areas of seeds and coexistence of crops, or even in animal feed. The GMO regime has become enforceable, but not necessarily enforced. National contexts vary greatly in their permissiveness of the presence of GM plants in agriculture and the market. Although violations are sanctionable under EU and in many cases national law, where national CAs are underper forming the burden to identify the shortfalls and challenge them falls either on pr oducers or consumers. Producers clearly have more concentrated motivations and resources to bring the law into effect, leaving a troublesome picture if GM products do indeed have adverse effect s in the long run. What is more, there does not seem to be any meaningful feedback process from the implementation stages back to policy-makers. Beyond inspection of the facilities and legal changes in national context, the Co mmission does not necessarily evaluate the practicability or usefulness of the measures it adopts. In the absence of significant challenges to it, the GMO regime will persist. The Commission has so far judged that NGO challenges to the EFSAs practices can be handled within the regime, and that they do not point to deficiencies that would requir e revision. This stance is typical of the way 91


in which the other agency network regimes are defended against criticism by insisting that increased transparency is a suffici ent remedy. However, transparency does not necessarily address the marginalization of a ll but like-minded natural scientists from the expert committees, which is a main concern for NGOs. Neither did the Commision conclude, despite having examined the question several times, that its practice of approving pr oducts in the absence of QMV decisions by the Council (in every single case!) is a phenomenon that warrants revision of the GMO framework. Although the European Court of Ju stice has had occasion to interpret some aspects of the GMO legislative framework upon queries by member stat es, it has not been compelled by the Commission to consider th e validity of safeguard measures. These observations indicate that the Commission, who must articulate and represent the common European interest, will fight to retain any proced urally sound semblance of agreement among member states. It was only disagreement, and the international consequences of it, that caused the GMO regi me to be revised, and nothing short of such cataclysmic events is likely to uproot the current version. Mere criticism in the public sphere does not outweigh the high costs of r eaching such an arrangement to begin with. 92

PAGE 100

CHAPTER 4 Conclusion The chief source of momentum for EU po licy making has always been integration driven by the single-market objective. Recently the focus for integrating has become an increase, on the aggregate, in the investments in research and development as a vehicle to the EUs desired global role as a leader. For their part, member states traditionally resist the integrative impetus primarily in order to protect their contro l over certain policy domains seen as key to national objectives. Part icipation by interest groups is defined in congruence with this agenda, therefore natu rally privileging indus trial and multi-national actors. Societal actors have access to the policy making pr ocess via specially designed consultation procedures by the Commission, di rect lobbying before MEPs, and access to national-level ministers and peak politicians. In order to bolster its own legitimacy and acquire specialized information, the Commission has made attempts to ensure th at at least some civil society organizations have consis tent access to policy maki ng in areas relevant to their activity. An examination of the interests involved in the GMO policy case and its outcomes quickly reveals that economic motivations we re not uniformly influential over time, or equally important across nati onal, institutional, and civi l society actors. Additional concerns were made salient by the complexity of the policy area and the inability of the Commission to compartmentalize it into a si ngle discourse, a singl e policy making mode, or even a single DG. This phenomenon indi cates that the simultaneous presence of scientific and political uncertainty can pr event the EUs regularized mode of policy 93

PAGE 101

making from proceeding in predictable ways. But is their presence sufficient for the advent of contestation of the system itself? It appears that a high degree of public distrust and mobilization on its basis is necessary for this to occur. Evaluations of the outcomes of contested multi-level governance in the GMO case vary greatly, depending on the starting point of the analysis, but the fact remains that implementation is uneven across member stat es and the Commission is not compelled to reevaluate the regime in the absence of a co nflict that would prevent it from functioning altogether. This outcome may indicate that, rather than having reached a qualitatively better outcome by reformulating its original objectives and preferred methodologies through the protracted contestation period, the EU may have been pushed closer to a least common denominator outcome by this situation. There is no question that scientific expert ise occupies a very special place in the EU policy making apparatus as an input resource. This special role is reinforced both by novel technologies periodically creating regul atory deficits and by the EUs stance toward knowledge as an objective and a means to future development. The GMO case demonstrates that the relationship of scientif ic expertise to EU policy is vulnerable, if policy outputs are seen as inadequate or in sufficient, to very di sruptive contestation by other actors with differently based claims. It is possible to anticipate that a dynamic of contestation may develop in policy areas that involve immediate impacts on EU ci tizens and novel technol ogical applications that are difficult to categorize. Contestation is likely to accelerate as discourses reinforce each other. But if the type of policy making paralysis seen in the GMO case is indeed contingent on large-scale public distrust and mobilization, and these fail to materialize in 94

PAGE 102

response to particular novel t echnologies, it may be equa lly likely that a method of handling the attendant un certainty becomes regularized in some way Already it is possible to observe such a dynamic in relation to uses of GMOs in animal rearing and seeds, which do not attract much public atte ntion although they continue to be contentious within the EU political system (s ignificantly, there are no policy outputs yet). It is possible that without priming, such as that provided by the BSE crisis, citizens simply can not maintain their focus on such i ssues and they will be dealt with out of the public eye, despite conflicts among institutiona lized actors as to principles and policy making modes. The multitude of delegative choices in the Commission, Parliament and Council might allow for multiple policy agendas to be balanced while the legislative process moves forward. Avenues for further research One direction for further study would be to examine the efforts at increasing transparency that have been made in various EU institutions and to assess whether they have the intended results. Since these reforms have been suggested as remedies to a multitude of problems, especially with EU-l evel agencies, any negative findings from such a study should be rather disconcerting. A whole other fertile area fo r research is in comparisons of the EU with other political systems. Comparisons with the Un ited States are alrea dy abundant in many policy areas. A sphere that has attracted less attention is that of comparison with international law. At first glance such endea vors may not necessarily be fruitful because of the very different conditions in which th ese systems operate. But the coordination of national positions necessary in both cases, an d the EUs lackadaisi cal attitude about 95

PAGE 103

implementation in some policy areas, may ma ke such comparisons rather interesting, especially concerning environmental policy. In terms of a more philosophical interroga tion of the policy making process, it is worthwhile to examine further the multivaria te role of expertise in the EU political system, as well as the qualifications that defi ne expert status. In addition, a study of how the overarching EU agenda, expressed in such catch-all phrases as the single market and knowledge-based society, functions in times when the system is under stress (does the current slogan become a unifying element for diverging positions, or is it identified as a principle that has run its cour se and is now part of the problem?) may reveal more about the dynamic of legitimation in multi-level governance. 96

PAGE 104

Appendix 1: Map of PDO, PGI, and TSG license holders Original map source: 97

PAGE 105

Appendix 2: Map of EU Member States who invoked precautionary measures against GMOs Original map source: 98

PAGE 106

Appendix 3: Outcomes of DG SANCO im plementation inspection missions Year State DG SANCO recommendations regarding 2005 Spain Unclear designation of responsibilities and poor coordination; Laboratory accreditation 2005 Italy Poor coordination among regions 2005 Netherlands Unclear consequences for infringement of food labeling requirements 2005 Portugal Limited resources allocated; Sampling not in line with EC recommendations 2006 Belgium Sampling not in line with EC recommendations 2006 Czech Republic Sampling not in line with EC recommendations 2006 Denmark Sampling not in line with EC recommendations 2006 France Delayed transposition of Directive 2001/18/EC; Sampling not in line with EC recommendations 2006 Hungary Sampling not in line with EC recommendations; 2006 Poland Sampling not in line with EC recommendations; Enforcement deficit 2006 Slovakia Sampling not in line with EC recommendations 2006 United Kingdom No central plan or laboratory (samples sent to other states); Sampling not in line with EC recommendations 2006 Austria Sampling not in line with EC recommendations; Limited resources allocated; Insufficiently transparent; Enforcement deficit 2007 France Sampling not in line with EC recommendations; Insufficiently transparent 2007 Greece Sampling not in line with EC recommendations; Ensure rights of food op erators to independent expert opinion 2007 Italy Limited resources allocated; Laboratory accreditation 2007 Romania Delayed transposit ion of Directive 2001/18/EC; Limited resources allocated 99

PAGE 107

BIBLIOGRAPHY Abels, Gabriele. 2002. Experts, Citizens, and Eurocrats Towards a Policy Shift in the Governance of Biopolitics in the EU. European Integration online Papers (EIoP) Vol. 6 No. 19. URL: Abraham John and Graham Lewis. 1999. H armonising and competing for medicines regulation: how healthy are the Europ ean Union's systems of drug approval? Social Science & Medicine Vol.48, 1655-1667. Ansell, Christopher K., R. Maxwell and D. Sicu relli. The Asymmetries of Governance, 329-350 in Vogel, David and Ch ristopher K. Ansell editors. 2006.. What's the Beef?: The Contested Governance of European Food Safety MIT Press: Cambridge, MA. Balme, Richard and Didier Chabanet. 2008. European Governance and Democracy: Power and Protest in the EU Rowman & Littlefield Publishers, Inc.: Lanham, MD, USA. Barbieri, Dario and Edoardo Ongaro. 2008. EU Agencies: What is Common and What is Distinctive Compared with National Level Public Agencies? International Review of Administrative Sciences Vol.74, No.3, 395-420. Beckmann, Volker and Justus Wesseler. 2005. Governance of Genetically Modified Crops in the EU Post-Conference Paper, Problems for Polycentric Governance in the EU: Challenges for Cooperative Governance Berlin, Humboldt University, June 15-18. Borrs, Susana, Charalampos Koutalak is, and Frank, Wendler. 2007. European Agencies and Input Legitimacy: EFSA, EMeA and EPO in the Post-Delegation Phase. Journal of European Integration, Vol.29 No.5, 583-600. Broscheid, Andreas and David Coen. 2003. I nsider and Outsider Lobbying of the European Commission: An Informa tional Model of Forum Politics European Union Politics Vol. 4 no. 2 Pages 165. Buonanno, Laurie. The Creation of the Eur opean Food Safety Authority, 259-280 in Vogel, David and Christopher K. Ansell editors. 2006. What's the Beef?: The Contested Governance of European Food Safety MIT Press: Cambridge, MA. Cadot, Suma-Eisenman, Traca. 2001. Trade-related issues in the regulation of genetically modified organisms Paper pr epared for the workshop on European and American Perspectives on Regulating Genetic ally Engineered Food, Insead, 7/8 June 2001. Cantley, Mark. 2004. How Should Public Policy Respond to the Challenges of Modern Biotechnology? Current Opinion in Biotechnology Vol.15, 258-263. 100

PAGE 108

Dancet, Geert. 2008. Speech by Mr. Geert Da ncet, Executive Director of European Chemicals Agency at Hearing at the Envi ronment, Public Health and Food Safety Committee of the European Parliament: Brussels, July 14. Dehousse, Renaud and Paul Magnette. 2006. Institutional Change in the EU in The Institutions of the European Union 2nd ed., Peterson, John a nd Michael Shackleton, eds.,Oxford University Press: New York, NY USA. Dimas, Stavros. 2006. Speech by Stavros Dima s, Member of the European Commission, responsible for Environment Coexistence of Genetically Modified, Conventional and Organic Crops: Freedom of Choice. Conference on GMO Coexistence, Vienna, April 05. Eckley, Noelle and Selin, Henr ik. 2004. All talk, lit tle action: precaution and European chemicals regulation Journal of European Public Vol.11, No.1, pages 78-105. URL: European Commission. 2006. Environment: Commission welcomes Council action on REACH, climate change and marine protection, Press Release IP/06/1826. December 18. European Commission. 2002 a. The European Network of GMO Laboratories (ENGL) Memorandum 02/279. December 4. European Commission. 2002 b. Commission Launches European Network of GMO Laboratories to Improve Traceability in Food Chain Press Release IP/02/1795. December 4. European Commission Official A. January 2009. Interview by author. European Commission Official B. January 2009. Interview by author. European Commission Official C. January 2009. Interview by author. European Commission Official D. January 2009. Interview by author. European Commission Official E. January 2009. Interview by author. European Council. 2005. Working together for growth and jobs A new start for the Lisbon Strategy Communication of the Spring European Council COM (2005) 24; Brussels, Belgium. February 2. URL: European Court of Justice. 2008. Judgment of the Court of Justice in Case C-121/07 Commission v French Republic: The Court Orders France to Pay a Lump Sum for Failing to Comply Swiftly with the 2004 Judgment of the Court Establishing its 101

PAGE 109

Failure to Fulfill Obligations Concerning GMOs Press Release 87/08. December 9. European Food Safety Authority. 2008. Request from the European Commission to review scientific studies related to the impact on the environm ent of the cultivation of maize Bt11 and 15071: Scientific Opinion of the Panel on Genetically Modified Organisms (EFSA-Q-2008-679) The EFSA Journal Vol 851, 1-27. October 29. European Food Safety Authority. 2008. Updated Guidance Document for the Risk Assessment of GM Plants and Derived Food and Feed The EFSA Journal Vol 727. European Union. 2007. Treaty of Lisbon Official Journal of the European Union C 306. December 17. Garca, Paula Rey. 2006. "Directive 2001/18/EC on the Deliberate Release into the Environment of GMOs: an Overview and the Main Provisions for Placing on the Market" Journal for European Environmental Planning & Law Vol.3, No.1, 3-12. Greer Scott L. Elize Massard da Fonseca Ch ristopher Adolph. 2008. Mobilizing Bias in Europe: Lobbies, Democracy and EU Health Policy-Making European Union Politics Vol. 9, No. 3 Pages 403 Grill, K. and S. O. Hansson. 2005. Epistemic Paternalism in Public Health Journal of Medical Ethics, Vol.31 No.11, 648. Groenleer, Martin, A. Boijn and S. Kuip ers. 2005. The autonomous development of European Union agencies: Regulation medici nal products and food safety in Europe. Paper presented at the EGOS Colloquium: Berlin, June 30-July 2. Hansen, Bjorn G. and Mark Blainey. 2006. R EACH: A Step Change in the Management of Chemicals, Review of European Community and International Environmental Law, Vol.15 No.3, 270-280. Herrick, Clare B. 2005. Cultures of GM': discourses of risk and labelling of GMOs in the UK and EU. Area Vol. 37 No. 3, 286 294. Hix, Simon. 2005. The Political System of the European Union 2nd ed. Palgrave Macmillan: New York, NY USA. Jasanoff, Sheila. 2005. Designs on Nature: Science and Democracy in Europe and the United States Princeton University Press: Princeton, NJ. Kelemen, Daniel R. 2002. The Politics of 'E urocratic' Structure and the New European Agencies, West European Politics Vol.25 No.4, 93-118. Kruzer, Paulette and Alice Cooper. 2007. What's for Dinner? European Farming and Food Traditions Confront American Biotechnology Comparative Political Studies, 102

PAGE 110

Vol. 40, No. 9, 1035-1058. Levidow, Carr and Wield. 2000.Genetically Modified Crops in the European Union: Regulatory Conflicts as Pr ecautionary Opportunities. Journal of Risk Research Vol.3 No.3, 189-208. Lieberman, Sarah and Tim Gray. 2006. The so -called 'moratorium' on the licensing of new genetically modified (GM) products by the European Union 1998-2004: a study in ambiguity. Environmental Politics, Vol.15 No.4, 592-609. Lieberman, Sarah and Tim Gray. 2008. The World Trade Organizations Report on the EUs Moratorium on Biotech Products: The Wisdom of the US Challenge to the EU in the WTO. Global Environmental Politics Vol.8 No.1, 33-53. Mahoney, Christine. 2004. The Power of Institutions: St ate and Interest Group Activity in the European Union. European Union Politics Vol. 5 No. 4 pages 441. Majone, Giandomenico. June 2000. The Credib ility Crisis of Co mmunity Regulation. Journal of Common Market Studies Vol. 38 No. 2 pages 273-302. Majone, Giandomenico and Michelle Everson. 2001. Institutional Reform: Independent Agencies, Oversight, Coordinati on and Procedural Control in Governance in the European Union: Cahiers of the Forward Studies Unit Edited by Olivier De Schutter, Notis Lebessis and John Paterson, Luxembourg: Office for Official Publications of the European Communities, 130-169. Moll, Nathalie. 2009. Interview by author. EuropaBio Office, Brussels. Pesendorfer, Dieter. 2006. EU Environmenta l Policy under Pressure: Chemicals Policy Change between Antagonistic Goals? Environmental Politics Vol. 15, No. 1, 95 114. Peterson, John and Michael Shackleton, eds. 2006. The Institutions of the European Union, 2nd ed. Oxford University Press: New York, NY USA. Pollack, Mark A. Theorizing EU PolicyMaking in Wallace, Helen, William Wallace and Mark A. Pollack. 2005. Policy-Making in the European Union 5th ed. Oxford University Press Inc.: New York, NY, USA. Rosendal, Kristin G. 2005. Governing GMOs in the EU: A Deviant Case of Environmental Policy-making? Global Environmental Politics Vol.5, No.1, 82-104. Schurman, Rachel. 2004. Fighting Frankenfoods : Industry Opportunity Structures and the Efficacy of the Anti-Biotech Movement in Western Europe. Social Problems Vol.51, No.2, Pages 243. 103

PAGE 111

Skogstad, Grace. 2003. Legitimacy and/or Policy Effectiveness?: Network Governance and GMO Regulation in the European Union. Journal of European Policy Vol.10, No.3, 321-338. Skogstad, Grace. Regulating Food Safety Risks in the EU: Comparative Perspective, 213-236 in Vogel, David and Christopher K. Ansell editors. 2006. What's the Beef?: The Contested Governance of European Food Safety MIT Press: Cambridge, MA. Smith, Mitchell P. 2007. All Access Points ar e Not Created Equal: Explaining the Fate of Diffuse Interests in the EU British Journal of Politics and International Relations Vol 10, No 1, 64-83. Stanton, Lyndon. (2005) European Chemical s Industry Faces Regulatory Upheaval Environmental Law Review Vol 7, No 4, 231-7. Tarrant, Andy and R. Daniel Kelemen. 2007. B uilding the Eurocracy : The Politics of EU Agencies and Networks Paper Prepared for the Biennial Eur opean Union Studies Association Convention: M ontral, Canada, 16-19 May. Tiberghien, Yves and Marko Papic. 2006. Crisis of Democratic Governance and Centrifugal Forces in the EU: The Struggl e between Periphery and Center over the Governance of GeneticallyM odified Organisms. Paper Pr epared for presentation at The Annual Conference of the Canadian Po litical Science Association (CPSA) at York University in Toronto, June 1-3. Vetier, Marta. 2009. Interview by author. Greenpeace Europe Office, Brussels. Victor, Marc. 2001. Comment: Precaution or Protectionism? The Precautionary Principle, Genetically Modified Orga nisms, and Allowing Unfounded Fear to Undermine Free Trade. The Transnational Lawyer Vol. 14, 295. Vogel, David. 1998. The Globalizati on of Pharmaceutical Regulation Governance: An International Journal of Policy and Administration Vol. 11, No. 1, 1. Vogel, David. 2001. Ships Passing in the Night: The Changing Politics of Risk Regulation in Europe and the United States European University Institute: Robert Schuman Center for Advanced Studies. Working Paper No. RSC2001/16. Vogel, David and Christopher K. Ansell editors. 2006. What's the Beef?: The Contested Governance of European Food Safety MIT Press: Cambridge, MA. Wallace, Helen. (2005) An Institutional Anat omy and Five Policy Modes in Wallace, Helen, William Wallace and Mark A. Pollack. 2005 Policy-Making in the European Union, 5th ed. Oxford University Press Inc.: New York, NY, USA. 104

PAGE 112

Wallace, Helen, William Wallace and Mark A. Pollack. 2005. Policy-Making in the European Union, 5th ed. Oxford University Press Inc.: New York, NY, USA. Winter, Gerd. 2006. Risks, Costs and Alternat ives in EC Environmental Legislation: The Case of REACHReview of European Community and International Environmental Law, Vol. 15, No 1, 56-65. Webpages: ECHA bout/organisation_en.asp EMEA EFSA EFSA Register of Questions http://registerofquestions.efsa.europa.e u/roqFrontend/questionsList.jsf?nocache=1236 619531372 ECJ database sp/cgi-bin/ Implementation reports (cited in Appendix 5): DG(SANCO)/8102/2006 Final Report of a Mission Carried out in Belgium from 13 to 17 February 2006 in Order to Evaluate Offi cial Control Systems for Food and Feed Consisting or Produced from GMOs DG(SANCO)/8104/2006 Final Report of a Mi ssion Carried out in Slovenia from 6 to 10 March 2006 in Order to Evalua te Official Control Systems for Food and Feed Consisting or Produced from GMOs DG(SANCO)/8109/2006 Final Report of a Mission Carried ou t in Hungary from 22 to 26 May 2006 in Order to Evaluate Official Contro l Systems for Food and Feed Consisting or Produced from GMOs DG(SANCO)/8110/2006 Final Report of a Mission Carried out in the Czech Republic from 15 to 19 May 2006 in Order to Evaluate Official Contro l Systems for Food and Feed Consisting or Produced from GMOs DG(SANCO)/8116/2006 Fina l Report of a Mission Carrie d Out in the United Kingdom from 12/06/2006 to16/06/2006 Concerning Controls on Food and Feed Containing, Consisting of or Produced from GMOs DG(SANCO)/7666/2005 Final Report of a Mission Carried Ou t in the Netherlands from 19/09/2005 to 23/09/2005 Concerning Controls on Food and Feed Containing, Consisting of or Produced from GMOs DG(SANCO)/7653/2005 Fina l Report of a Mission Carrie d Out in Italy from 06/06/2005 to 10/06/2005 Concerning Controls on F ood and Feed Containing, Consisting of or Produced from GMOs 105

PAGE 113

106 DG(SANCO)/8105/2006 Final Report of a Mi ssion Carried Out in Germany from 06/03/2006 to 10/03/2006 Concerning Controls on Food and Feed Containing, Consisting of or Produced from GMOs DG(SANCO)/7669/2005 Final Report of a Mi ssion Carried Out in Portugal from 17/10/2005 to 21/10/2005 Concerning Controls on Food and Feed Containing, Consisting of or Produced from GMOs DG(SANCO)/8086/2006 Final Report of a Mi ssion Carried Out in France from 15/05/2006 to 19/05/2006 Concerning Controls on Food and Feed Containing, Consisting of or Produced from GMOs DG(SANCO)/7632/2005 Final Report of a Mission Carried Out in Spain from 07/03/2005 to 11/03/2005 Concerning Controls on Food and Feed Containing, Consisting of or Produced from GMOs DG(SANCO)/8100/2006 Fina l Report of a Mission Carrie d Out in the Slovak Republic from 30/01/2006 to 3/02/2006 Concerning C ontrols on Food and Feed Containing, Consisting of or Produced from GMOs DG(SANCO)/8106/2006 Final Report of a Mi ssion Carried Out in Poland From 20/03/2006 to 24/03/2006 Concerning Controls on Food and Feed Containing, Consisting of or Produced from GMOs DG(SANCO)/2007-7224: Import controls on food and feed of non-animal origin, General Audit Austria Response of the Competent Authorities of Romania to the recommendations of the Mission Report ref. DG (SANCO) 7186/2007 on official control of GMO food, feed and seed. Response of the Competent Authorities of France to the recommendations of Mission report ref. DG (SANCO) 2007/7223 in order to assess the official control systems in place for food hygiene within the meaning of Regulation (EC) no 852/2004 Response of the Competent Authorities of Greece to the recommendations of Mission report ref. DG (SANCO) 2007/7199 on official control of GMO food, feed and seed. Response of the Competent Authorities of Italy to the Recommendations of Mission report ref. DG (SANCO) 7193/2007 in order to assess the of ficial control systems in place for food hygiene within the meaning of Regulation (EC) no 852/2004