Reforming the European Pesticides Regime
Experimentalist Solutions to Regulatory Challenges
Tobias Jung 12290610
Master Thesis Political Science: Public Policy and Governance
Research Project: New Modes of Governance Date: 20.06.2019
Supervisor: Prof. Dr. Jonathan Zeitlin Second Reader: Dr. Maria Weimer
Table of Contents
1. Introduction ... 1
2. Risk Regulation in the European Union ... 5
3. Theoretical Framework: Experimentalist Governance and Co-Production ... 13
3.1 The European Union and the Concept of Governance ... 13
3.2 Experimentalist Governance ... 15
3.3 Co-Production ... 17
4. Methodology ... 19
4.1 Case Selection ... 19
4.2 Data Collection and Analysis ... 21
5. The EU Pesticides Regime ... 22
5.1 A Short History of EU Pesticides Regulation ... 23
5.2 Guiding Principles of the Pesticides Regime ... 24
5.2.1 Hazard Identification ... 25
5.2.2 Precaution ... 26
5.2.3 Sustainability and Substitution ... 27
5.2.4 Harmonisation and Mutual Recognition ... 28
5.3 Pre-Market Assessment of Active Substances and Placing the Product on the Market ... 29
5.3.1 Approval of an Active Substance ... 29
5.3.2 Placing a Plant Protection Product on the Market ... 33
5.3.3 Renewal Procedure ... 35
5.3.4 Post-Market Measures ... 37
6. The Case of Glyphosate: Law-Based Science and the Value of Evidence ... 37
6.1 The Process of the Re-approval of Glyphosate ... 38
6.2 Regulatory Controversies ... 43
6.2.1 Differences between the IARC and EFSA/ECHA Conclusions ... 43
6.2.2 The Issue of Independence and Transparency ... 48
6.2.3 Technocratic Decision Making vs. Democratic Legitimacy ... 49
7. Reforming the Current Pesticides Regime ... 53
7.1 Deficiencies in the Current Pesticides Regime ... 53
7.1.1 Delay of Evaluations ... 53
7.1.2 Ineffective Working of the Zonal System ... 54
7.1.3 Lacking Transparency in the Risk Management Stage ... 56
7.1.4 Excessive Use of Article 53 ... 56
7.1.5 Lack of Stringent Post-Authorisation Monitoring System ... 57
7.1.6 Incompleteness of Assessments... 58
8. Assessing the Reform Proposals ... 59
8.2 Introduction of a Post-Authorisation Monitoring System ... 62
9. Post-Authorisation Monitoring and Experimentalism ... 67
9.1 Scope Conditions ... 67
9.2 Developing a Recursive Meta-Regulation ... 69
10. Conclusion ... 72
11. References ... 77
Annex I – List of Interviews ... 100
List of Abbreviations
BAuA Federal Institute for Occupational Safety and Health BfR Federal Institute for Risk Assessment
cMS Concerned Member State
CLP Classification, Labelling, and Packaging of Substances and Mixtures CLH Harmonised Classification and Labelling
co-RMS co-Rapporteur Member State DAR Draft Assessment Report DG Directorate General DRR Draft Renewal Report CA Competent Authority ECHA European Chemical Agency
ECPA European Crop Protection Association EFSA European Food and Safety Authority EMA European Medicines Agency
EU European Union
GCEU General Court of the European Union GMOs Genetically Modified Organisms GLP Good Laboratory Practice
GTF Glyphosate Task Force
IARC International Agency for Research on Cancer NGO Non-Governmental Organisation
PAFF Standing Committee on Plant, Animals, Food and Feed PAN Pesticides Action Network
PBT Persistent, Bioaccumulative and Toxic for the Environment RAR Renewal Assessment Report
RMS Rapporteur Member State SAM Science Advice Mechanisms
SAPEA Science Advice for Policy by European Academics TFEU Treaty on the Functioning of the European Union
UN-GHS United Nations Globally Harmonised System of Classification and Labelling of Chemicals
WHO World Health Organisation zSC Zonal Steering Committee zRMS Zonal Rapporteur Member State
List of Figures
1 1. Introduction
Over the last centuries, mankind has used its growing intellectual capacity and availability of information to develop new technologies that have secured and expanded society’s prosperity. However, the steady progress in technological advances has not only led to increased control over life, increased safety of life, progress and individual choice (Weimer, 2019: 21) but has also produced unintended negative side effects that might fundamentally threaten the progress and level of prosperity achieved through the very same technologies. In this context, Giddens speaks of ‘manufactured risks’, which are
created by the very progression of human development, especially by the progression of science and technology. Manufactured risks refer to new risk environments for which history provides us with little previous experience (Giddens, cited in Weimer and de Ruijter, 2017: 19)
This paradoxical development of an exposure to risks that are produced by mankind itself was best captured by Beck’s (1992) conceptualisation of a risk society as "a systematic way of dealing with hazards and insecurities induced and introduced by modernisation itself” (Beck 1992:21).
The use of chemical pesticides, according to Hood et al. (2001), comes closest to the risks discussed in the risk society literature, where pesticides refer to “any substance, or mixture of substances of chemicals or biological ingredients intended for repelling, destroying or controlling any pest, or regulating plant growth” (FAO, 2014: 6). Pesticides have a central role in modern intensive farming practices and are products of technological developments in chemistry, but at the same time pose health risks and may threaten the environment (Hood et al. 2001: 32; Cooper and Anderson, 2007; Bozzini, 2017; Cuhra, Bohn and Cuhra, 2016; Schäffer et al. 2018; Storck et al. 2017; Székács and Dervas, 2012). In contrast to these negative effects, it is similarly argued that pesticides are indispensable for guaranteeing a stable food supply (Cooper and Anderson, 2007). By controlling agricultural pests and vectors of plant diseases, by controlling human and livestock diseases, and by preventing or controlling organisms that harm other human activities and structures, pesticides contribute to higher yields and better quality of products (Cooper and Anderson, 2007). Thus, there is a fundamental challenge of, on the one hand, trying to maintain its advantages and, on the other hand, controlling the risks originating from it. To solve this challenge is the task of social regulation which “uses tools aimed at controlling and monitoring actions by agents, which affects the health of human beings and the environment” (Pelaez, da Silva and Borges Aravjo,
2
2013: 643). Pelaez, da Silva and Borges Aravjo (2013: 643) argue that, in times of technological advancement and concomitant risks, social regulation is translated into risk or product regulation, which can be defined as an “activity aiming to minimize and manage risks resulting from the development, production, and use of technological innovation” (Hood et al., 2001: 3).
In the context of the European Union, risk regulation is operationalised by relying on three distinctive steps, risk analysis, risk management and risk communication1. Risk analysis is performed by scientific agencies such as the European Food and Safety Authority (EFSA), which evaluates the agent from which risks emanate and assesses whether or not the risk poses a danger. Risk management, on the other hand, is conducted by the Commission and national representatives. In this stage, decision makers deliberate about the normative implications of risk, thus establishing what an acceptable risk is to society. If the risks posed to society are found to be acceptable, the agent under consideration can be approved.
However, disputes over biotechnological technologies such as genetically modified organisms (GMOs) have forcefully proven that the deliberative power of the risk management stage can be undermined. Although scientific assessments provided a positive conclusion with regards to the risks of this technology, an over-politicisation of debates among risk managers resulted in a gridlock and consequently in a failure of the risk regulatory system to produce legitimate and effective policy outputs (Paskalev, 2012; Weimer, 2019; Weimer, 2015; Everson and Vos, 2009). Thus, studying how technological risks are regulated tackles fundamental questions of how science, politics and law interact (Weimer, 2019: 13).
By studying the regulation of pesticides, this thesis aims to unravel the workings of risk regulatory systems for synthetic chemicals. In the European Union, the regulation of pesticides follows a two-tier assessment in which pesticides are further differentiated into active substances and plant protection products.2 An active substance is “any chemical, plant extract, pheromone or micro-organisms that has action against ‘pests’ or on plants, part of plants or plant products” (European Commission, 2019a). A plant protection product, on the
1
According to the General Food Law, risk communication can be defined as ‘the interactive exchange of information and opinions throughout the risk analysis process as regards hazards and risks, risk-related factors and risk perceptions, among risk assessors, risk managers, consumers, feed and food businesses, the academic community and other interested parties, including the explanation of risk assessment findings and the basis of risk management decisions’ (European Communities, 2002: Article 3(13)). Risk communication will, however, not be a focus of this thesis.
2
3
other hand, is the final product containing at least one active substance but that might also contain other ingredients such as co-formulants, safeners or synergists.3
In order to assess the hazardous and risk properties of both the active substance and the plant protection product, the European Union uses an extensive legislative framework comprised of several regulations and directives – at its heard Regulation 1107/2009 concerning the placing on the market of plant protection products. This piece of legislation regulates the assessment and authorisation of active substances and plant protection products. Moreover, it creates a dual approval system that is supported by a multi-level network of national regulatory agencies and European regulatory agencies. The active substance is approved at the European level with the support of an initial scientific assessment of the active substance by a national competent authority (CA), whereas the plant protection product containing that active substance is authorised solely by the Member States. However, the effectiveness of this system has recently been challenged by regulatory controversies surrounding the re-authorisation of glyphosate. Similar to the authorisation procedure of GMOs, risk assessors provided a clear conclusion that glyphosate does not pose risks to human health, yet risk managers failed to deliberate successfully for years (e.g. Bozzini, 2017; Leonelli, 2018; Tosin, Lieleveldt and Wing, 2019).
In addition to the possible difficulties encountered in risk management, risk regulatory regimes in the European Union also face the challenge of accommodating the diversity of its 28 Member States. To meet these challenges and avoid gridlocks in the decision-making process, there is a need to enhance deliberative practices in EU governance arrangements. One way to conceptualise governance arrangements that offer the necessary degree of deliberative practices and ways to accommodate diversity is through experimentalist governance (Sabel and Zeitlin, 2008; 2012). Experimentalist governance arrangements are characterised by the possibility of mutual learning through recursive review of the implementation of framework goals or rules. More specifically, experimentalist governance arrangements comprise four distinct stages: the setting of framework goals, the autonomous achievement of these goals by lower level units, e.g. through national ministries or regulatory authorities, a peer review of the measures employed by lower level units to achieve the framework goals, and the revision of the initial framework goals in the light of knowledge
3 A co-formulant is a substance or preparation used in a plant protection product or biocidal product which is
neither an active substance, a safener, nor a synergist. A safener is a substance or preparation added to the plant protection product that suppresses or reduces its effects on some plants. A synergist is a substance or preparation that enhances the activity of the active substance or substances in a plant protection product (Bourguignon, 2017: 4).
4
gathered through the establishment of autonomous systems. Hence, it is possible for actors to foster deliberation and learn from each other despite their diverse backgrounds. Pesticides policy is characterised through highly diverse agricultural backgrounds in each Member State, i.e. different economical, ecological, environmental and climatic environments; the study of which, along with policy dispute within the risk management stage, offers a rich opportunity to apply the assumptions of the experimentalist governance framework. In order to do so, this thesis will analyse the reform proposals of the current pesticides regime that were proposed following the controversies surrounding the glyphosate re-approval, and it will apply the theoretical insights from experimentalist governance in order to assess the benefits of these proposals (e.g. SAPEA, 2018; SAM, 2018; European Parliament, 2019). In particular, the focus will be on the proposal for a post-market authorisation monitoring system, as literature on pharmaceutical post-authorisation monitoring has shown that such a system can produce experimentalist governance features that lead to more effective policy outcomes (Heim, 2019; Sabel and Simon, 2011). Therefore, the research question of this thesis is as follows: do the
current reforms proposals of the current European pesticides regime, in particular the post-authorisation monitoring system, present a viable alternative to overcome regulatory challenges in light of experimentalist governance theory?
Yet, in order to answer this research question properly, I also ask the following sub-questions:
1. How does the European pesticides regime work?
2. Why is the current European pesticides regime perceived to be ineffective?
3. How much did the re-approval process of glyphosate contribute to the perception that the current pesticides regulatory regime is ineffective?
4. How do the proposed reforms differ from and/or improve the existing policies?
The remainder of this thesis will proceed in the following way. Chapter two reviews the literature on risk regulation to introduce the broad framework in which the European pesticides policy is embedded. Chapter three introduces the theoretical framework used in this thesis, comprising of experimentalist governance and co-production. Chapter four lays out the methodological choices and a more detailed description of the case study. Chapter five analyses the European pesticides regime, focusing on its history, its guiding principles and its procedures with regards to the approval and renewal of active substances and plant protection products, and subsequent post-market authorisation measures. Chapter six provides an in-depth case study of the renewal of the authorisation of glyphosate. It is argued that the case
5
exemplifies several deficiencies of the current system, including a lack of transparency and the questions which evidence are being and should be used. Moreover, the case exemplifies the inseparable nature of science and law and how they constitute each other, which not only impacts and reinforces the creation of deficiencies of the overall pesticides system but also facilitates the emergence of regulatory controversies. Chapter seven expands on that and presents further deficiencies in the current pesticides regime, including a lack of harmonised guidelines, delayed authorisations, the use of the emergency clause, a lacking post-monitoring marketing system and insufficient consideration of ecotoxicological and epidemiological factors in risk assessments. Chapter eight then sheds light on the reform proposals from academics and policy actors for the current pesticides regime. It has to be noted, however, that due to the limited length of this thesis, not all proposals can be analysed. Thus, the focus is on those reforms that tackle the deficiencies identified in the previous sections, namely proposal to strengthen transparency and the post-authorisation monitoring system. Chapter nine then analyses the post-authorisation monitoring system through the lens of the experimentalist governance theory. It is concluded that when applying the features of experimentalist governance, the post-authorisation monitoring system not only proves to be an effective way to establish recursive mechanisms that inform assessments of risk assessors, but it would also offer a way for Member States to learn from each other. The tenth and final chapter summarises the main findings of the thesis and concludes in two ways. First, EU risk regulatory systems are not static institutional systems but are able to be re-constituted and incrementally changed in policy environments characterised by value and scientific conflict. Second, debates about value conflicts, scientific conflicts and technical policy reforms should be aligned in a united effort to start a debate about a future vision of European food production and the place of pesticides within it.
2. Risk Regulation in the European Union
Literature on risk regulation, especially in the context of the European Union, comprises a vast body of multidisciplinary and interdisciplinary work. Academic literature comes from subjects such as law, politics, policy studies, sociology, and science and technology studies and focuses broadly on three interrelated topics:
1. the institutional and legal set up of those frameworks and their difficulties and challenges;
2. the relationship between science and politics;
6
The objects of scrutiny in this literature range from the regulation of chemicals (e.g. Nordlander, Simon and Pearsson, 2010; Fox, Versluis and van Asselt, 2013) and nanotechnology (e.g. Ehnert, 2017) to biotechnology (e.g. Weimer, 2019; van Asselt and Vos, 2006, 2008; Everson and Vos, 2009; Jasanoff, 2005). The literature on the biotechnology sector, with its focus on the regulation GMOs, has especially contributed to the overall conceptualisation and theorising on risk regulation in the European Union and will be mainly used to illustrate the latter’s strengths and weaknesses (Weimer, 2015; Weimer, 2014; Weimer and de Ruijter 2017; Weimer, 2019; van Asselt and Vos, 2006, 2008; Everson and Vos, 2009; Kim, Klika and Versluis, 2013; Paskalev, 2012).
In his seminal articles, Majone (1994, 1997) argues that the European Union has become a regulatory state. Advances in the harmonising efforts towards the completion and functionality of the Single Market, as well as the objective of protecting citizens from health and environmental risks, led to a gradual increase in the regulatory power of the European Union, most notably exercised through the European Commission and European regulatory agencies (Lodge, 2008; Weimer and de Ruijter. 2017: 3; Weimer, 2019). Hence, the very notion of risk has become an important organising principle in markets and society (Black, 2010).
Besides work on the origin of modern risks, the literature has also tried to define how risk can be understood in a regulatory framework. Hood et al. (2001: 3) gives a bold definition by arguing that risk is the “probability of adverse consequences”. Risk, being the probability of an adverse consequence, distinguishes itself from hazard, defined as the “potential for a substance, activity or process to cause harm or adverse effect” (Lofstedt, 2011: 149). Thus, risk should be understood as the probability of a hazard to materialize. By arguing that risk is connected to a probability, the literature offers the possibility for regulators to manage these risks by reducing the probability to a point where it is unlikely that the risk itself will ever materialize into a hazard for the population. In addition, defining risk as a matter of probability allows risks to be understood as inherently objective and detectable by scientists. To that end, Jasanoff (2017: 19) explains that risk can be treated as “a scientifically demonstrated fact, independent of the persons social and cultural affiliation, let alone the political preferences of the authorities who manage people’s affiliation”. However, a growing body of literature, primarily from science and technology studies, including Jasanoff, argue that the description of risk as an objective fact is flawed. Because, in order to detect and regulate a risk, it is important to raise awareness of what a risk is and whether it can be harmful to society. This process can be described as risk selection (Weimer, 2019: 26). Yet,
7
the process of risk selection is value-laden and touches upon political and normative questions, e.g. what kinds of risks are worth paying attention to, who should be protected, and at what cost.
A different focus of the literature is the strict separation between the different stages of risk regulation. Legal literature stresses that the clear separation of the three steps, risk analysis, risk management and risk communication, is rooted in what is called a rational-instrumental model of regulation. The rational rational-instrumentalist idea is underpinned by the perception that risks “can be ordered, managed, and controlled, and so regulators can implement their legislative mandates in an objective and efficient manner with exercising excessive discretion” (Black, 2010: 337). In that sense, the objectivity upon which the final decision is based is ascribed to a scientific judgment that is conducted during the risk assessment stage. In such a setting, science is treated as culturally neutral, universalistic and objective (Black, 2010: 326). The idea that science should offer an objective analysis of risks is rooted in various food scandals of the 1990s, especially the BSE and Dioxin scandals. These scandals eroded the trust in public authorities at both the EU and national levels and led to questions around their legitimacy. It appeared that experts who were supposed to give objective advice were in fact not objective, and this gravely undermined the image of policy-neutral expertise as both levels (Levidow and Carr, 2007; Stapleton, 2016; European Commission, 2007).
As a result, two trends have been observed. One the one hand, the EU discourse on risk regulation began to be - and still is - dominated by reference to ‘evidence-based policy’ (e.g. European Commission, 2019b; 2019c, 2015a, 2015b), while on the other hand, the entanglement of expert advisors who are directly connected to the decision makers has been broken up. Most notably, the latter idea was implemented by a process of ‘agencification’ within the EU. As it was deemed important to be able to draw on neutral advice, scientific tasks of analysing risks and threats and concluding whether products are safe, have been delegated to semi-independent agencies that are comprised of experts, such as the European Food and Safety Authority. Hence, these agencies have become the institutional embodiment of the separation of politics and science.
Alongside the idea of a separation of politics and science, it is also noted that agencies are believed to produce better and more efficient decisions, as well as possess increased accountability, particularly in areas of high technical complexity (Kim, Klinka and Versluis, 2013: 175; Gornitzka and Holst, 2015; Majone, 1994; Rimkuté, 2015: 117). The increase in
8
efficiency is rooted in the setup of those agencies. Several agencies - and the EFSA in this case - have been designed in a networked way that should accommodate the diverse nature of the Member States. In such a networked framework, scientific pluralism is ideally cherished and integrated into a forum in which network members can deliberate and debate in a way that would ultimately result in more efficient and legitimate outcomes (Skogstadt, 2003). For decision-making bodies, the delegation of these tasks and the subsequent reliance on conclusions produced by these agencies also have the side-effect of strengthening the legitimacy of their decisions. Thus, Gorniztka and Holst (2015) call the interactions between risk assessors and risk managers an ‘executive-expert nexus’. They argue that
drawing on scientists as the main information providers would thus legitimize the executive bodies’ autonomous basis for action, independent of national, societal, and parties’ interest, and would potentially buffer it from the political and intergovernmental logic of policy making (Gornitzka and Holst, 2015: 7)
In other words, the final decision should be de-politicized and guided by rational arguments provided through objective scientific advice.
Nevertheless, the idea that the process of risk assessment is neutral is increasingly challenged, especially in the literature of science and technology studies. In a report for the European Commission, a variety of scientists argued that it is often overlooked that the ways in which risk assessments are conducted inevitably rest on normative commitments, i.e. they are influenced by subjective judgments, social values, contestable assumptions and administrative practices (European Commission, 2007: 33; Jasanoff, 2005). Moreover, it is argued that risk assessment is inherently political, as the decisions concerning what constitutes a risk and how risk is measured are influenced by institutional arrangements and assessment practices that are themselves influenced by legal requirements (Motta, 2013). Jasanaoff already observed in 1995 that regulatory science, i.e. the science used for regulatory purposes, “is susceptible to divergent socially conditioned interpretations […] since quality standards tend to be more fluid, controversial, subject to political considerations” (Jasanoff, 1995a: 282)4. Besides the normative critique of the concept of neutral risk assessment, the idea of efficient and objective, i.e. independent, science has become heavily contested. As Satelli and Funtowitz note:
4 Besides the fluidity of standards and their subjection to political considerations, regulatory science can also be
differentiated from research science in several other ways, including that research is conducted under time pressure, ‘truths’ need to be relevant for policy makers and stakes in research outcomes are high (Jasanoff, 1995a: 283)
9
science finds itself under pressure from two major concurrent and intersection crises: one concerns public trust in the evidence produced by science and its institutions; the other, the governance of science and the reproducibility of its results. (Satelli and Funtowitz, cited in Weimer, 2019: 32)
The rationale for creating a risk management stage in the rational instrumentalist logic is rooted in the initial idea of separating science from politics, where risk management enshrines the political stage. It is acknowledged that science might be able to resolve analytical and methodological questions, yet it is neither able nor required to have an answer to the normative dimension inherent in policy making (Jorges, 2001). In the famous Pfizer Animal Health SA v Council case, the EU Court of the First Instance concluded that “scientific legitimacy is not a sufficient basis for the exercise of public authority” (Case T-13/99 Pfizer Animal Health SA v Council [2002]: para 201, II - 3392). It follows that the two stages are underpinned by two different sources of legitimacy. According to Weimer (2019: 43), scientists gain their legitimacy from scientific methodology, whereas policy makers derive their legitimacy from political and legal accountability. It follows that political processes go beyond science and, in the case of the EU, have to accommodate the diverse views of the Member States, necessitating the creation of a place for deliberation. This place for deliberation is embodied in the risk management stage by the comitology committees (Joerges, 2001; Joerges and Neyer, 1997) in which national representatives should ideally follow a problem-solving approach towards acts proposed by the Commission before they can be implemented. The technical nature of risk regulative decision, accompanied by uncertainty, is especially regarded as an ideal subject in which deliberation can place as policy makers can engage in a “collective search for truth and the best policy” (Weimer, 2015: 623).
However, the strict separation between these stages is heavily criticised throughout the literature. One point of criticism has been raised in the context of situations in which the scientific advice provided by the agencies is used as a political vehicle to make unilateral decisions. A case in point is the authorisation process of GMOs in the EU. As several commentators have analysed (see Weimer, 2015; Weimer, 2014; Weimer et al. 2017; Weimer, 2019; van Asselt and Vos, 2006, 2008; Everson and Vos, 2009; Kim, Klika and Versluis, 2013; Klika, Kim and Versluis, 2013; Paskalev, 2012), the European Commission repeatedly followed scientific opinions and authorised GMO products in situations where the national representatives in the risk management stage were not able to vote either for or against the authorisation. The scientific opinion became the decisive tool for justifying a decision, although it is only supposed to inform a political decision. Therefore, the risk
10
assessor was effectively assigned the role of the risk manager (Weimer and Pisani, 2017; Kim, Klika and Versluis, 2013; Majone, 2010), which ultimately led to the perception that the Commission is not a legitimate decisionmaker in the field of GMOs (Weimer, 2015: 640). This behaviour has been ascribed to a dilemma of the Commission, which does not perceive itself to have an electorate mandate to judge policy, but rather perceives itself as gaining legitimacy through superior knowledge and expertise (Weimer, 2019). By deliberately removing the political component from the political decision making, the Commission not only displaced political and institutional differences into technical problems, but politics also underwent a process of scientisation (Weimer, 2014; Vos and Asselt, 2009; European Commission, 2007).
Besides an empirical critique, the separation is also criticised from a normative perspective. From this perspective, a strict separation is perceived as creating mismatches of expectations between risk managers and risk assessors, which inhibits effective policy making. This is, for instance, the case when risk managers do not provide a specific protection objective. Risk assessment then produces a
variety of technical expressions of risk, based on considerations of endpoints, biological responses or other technical parameters that are sometimes only very indirectly related to the protection objectives pursued by the risk managers and policy makers. (SCCS, SCHER, SCENHIR, 2011: 9)
Thus, a situation might be created in which risk assessors do not understand what risk managers need for effective decision making, while risk managers and the public, on the other hand, do not understand what risk assessment can deliver (SCCS, SCHER, SCENHIR, 2011: 10).
With regards to the challenges of risk regulation in Europe, there are two broad issues, namely the challenge of diversity and the challenge of scientific uncertainty. While the former is a particular feature of the EU risk regulatory system, the latter is a general challenge encountered in all risk regulatory system. Concerning diversity, Weimer (2019: 49) notes that
diversity presents an enormous challenge in terms of finding common approaches to risk given both the politics of risk, that is, risk selection, including how to address divergent public risk perceptions in different Member States, and in terms of considering all relevant evidence including context specific and local knowledge.
In the same vein, Versluis (2016) argues that there is no single actor in the EU who can solve complex policy problems on its own. It is against this background that the EU has to walk a
11
fine line between respecting the subsidiarity principle and centralising efforts in order to harmonise the internal market. The latter focus on harmonising the internal market has limited, according to Weimer (2019: 228), the EU’s capacity to respond to political, cultural and value dimensions of risk and technology. Yet the legal framework of the European Union prescribes that its policies also must be reconciled with health and environment concerns. This difficulty in striking a balance between both market integration and health and safety, as well as between centralisation and diversity, has been widely recognised, most notably in the GMO decision-making of the past decade (e.g. Groenleer, 2016).
The second challenge is the handling of uncertainty. Uncertainty is a situation in which risk assessments cannot be based on probabilities, i.e. it is not possible to understand causal relationships or it is altogether impossible to establish a causal relationship in the first place (Weimer, 2019: 27; van Asselt and Vos, 2006, 208; Levidow and Carr, 2007). Thus, uncertainties are difficult to control as they transcend human cognitive capacities (Everson and Vos, 2009: 3; van Asselt, Vos and Rooijackers, 2009: 359). However, as uncertainty is an inherent property of science, policymakers must find ways to deal with it, especially in politically sensitive fields. Paradoxically, one of the ways policymakers deal with uncertainty is with an overreliance on scientific opinions (Versluis 2016). While this might seem illogical, the reason for this is simple: when policymakers are faced with uncertain situations, they have difficulty defining their preferences. Therefore, they require the production of more knowledge and thus become more dependent on scientific advice. Van Asselt and Vos (2006; 2008) call this the uncertainty paradox. Policymakers resort to science and frame their role as providing certainty, although uncertainty might preclude such conclusiveness. This kind of framing can be facilitated by policy makers when requesting a so-called plausibility proof that does not leave room for uncertainties (van Asselt and Vos, 2008: 282). The existence of uncertainty paradoxes can, for instance, be found within the Commission (e.g. van Asselt and Vos, 2006; Weimer, 2019). Weimer (2019) provides an important reason that can be found in the way that the European Commission uses Article 114 of the Treaty of the Functioning of the EU (TFEU), with its two objectives of economic integration while also ensuring high levels of protection of human health and of the environment. In reconciling those two objectives, science, with its ascribed objective meaning, provides a basis for intervention. The reliance on science in this regard produces narrow safety considerations, i.e. case by case decisions. Yet, since the Commission legitimises its actions with superior knowledge, it will want to rely on conclusions that are as certain as possible, thus making it prone to disregard uncertainties or focus only on specific dimensions of the assessments in order to frame
12
scientific outcomes as certain. Concomitant to the intolerance of risk managers to acknowledge and subsequently deal with uncertainties, uncertainty intolerance can be apparent in risk assessors as well. In the case of GMOs, van Asselt and Vos (2008) and van Asselt, Vos and Rooijackers (2009) concluded that EFSA was also intolerant to uncertainties, which affected the decision by the Commission to authorise GMOs based on apparent sound evidence.
However, risk managers do have the opportunity to act upon uncertainties without being trapped in the uncertainty paradox. The precautionary principle is a tool for this, as foreseen in the TFEU and widely acknowledged in literature. This principle allows “the use of provisional measures in cases of scientific uncertainty where a possibility of harm to health exists, even if it is not yet conclusively proven scientifically” (Hristova, 2013: 110). As Fisher (2009: 22) explains, the principle nearly always applies to standard-setting and risk appraisal, that is the process of deciding whether to authorise a product or activity. Nevertheless, it is necessary to establish that there exist scientific uncertainties which might become a problem if risk managers require a plausibility proof, thus leaving out the option that uncertainties exist in the first place. It is also for that reason that the precautionary principle is contested in literature. Bourguignon (2015: 12, 13) summarises the divergent positions. While there are those who see the precautionary principle a step in the right direction due to the possibility of reducing serious and irreversible environmental public health hazards, there are others who see it as flawed because the acknowledgement that uncertainties always exist would mean that the principle can be applied to every decision. Moreover, the application of the precautionary principle could undermine progress by depriving society of useful products and of a source of knowledge, as well as by isolating the EU globally.
While this review mainly presented findings from the GMO case in the EU, it is necessary to shed light on the literature of pesticides as well. While the overall literature on pesticides is vast (Bernhard et al. 2017), it mainly comprises papers in natural science fields such as ecotoxicology, ecology, chemistry and biology. Literature on pesticides with a regulatory focus is, however, scarce. Pelaez, da Silva and Bourges Arajvo (2013), for instance, offers a comprehensive comparative accounts of regulatory regimes of pesticides in the US, Sweden and the EU, with a focus on their institutional organisation as well as with regards to risk and hazard. However, they do not engage in a detailed analysis of the legal provisions. Literature that deals exclusively with the current EU legislative and policy framework focuses on the impact of the EU regulation in the field of pesticides in specific national contexts (Jess et al. 2014) or it comments on the challenges that accompany the
13
current pesticides legislation, focusing on legal challenges (Garcon, 2011; Mereu, 2011; de Vries-Stotijn, 2016) and methodological challenges in the risk assessment (Stehle and Schulz, 2015; Panizzi, Sucis, Trevisan, 2017; Kienzler, Bopp, van der Linden et al. 2016; Hunka et al. 2015). With regards to the questions of how science and law interact in risk regulatory systems, literature on pesticides is likewise less informative than literature on GMOs. Rimkuté (2015) and Alemanno (2013), however, conduct case studies of the ban of neonicotionids focusing on how and why agencies use different forms of scientific expertise, especially a substantiating expertise, i.e. when an organisation promotes and justifies its predetermined preferences, and the interconnection between law and science with regards to the application of the precautionary principle.
Although these accounts provide valuable insights into the working of the pesticides policy, a systematic focus on the political processes underpinning pesticides policy in the European Union and a clear integration of pesticides policy within risk regulation literature is missing (a notable exception on the political processes is Bozzini, 2017). Nevertheless, the recent debate about glyphosate has led to a renewed interest in the politics of pesticides authorisation and the identification of flaws in the current pesticides regime (Tosun, Lelieveldt and Wing, 2019; Paskalev, 2019; Arcuri 2017; Leonelli, 2018; Storck et al. 2017). Moreover, the rapid political development that followed the authorisation of glyphosate has offered new opportunities of research which are not yet covered by the literature. It is these new opportunities that are being looked upon in this thesis.
3. Theoretical Framework: Experimentalist Governance and Co-Production
In order to analyse the current pesticides regime and its reforms, this thesis employs a dual theoretical approach consisting of the experimentalist governance framework developed by Sabel and Zeitlin (2008; 2012) and the co-production framework developed by Jasanoff (2004). The reason for using a dual framework is the following: while experimentalist governance offers a lens to normatively analyse the possible advantages of reform proposals in an institutional and political environment that is characterised by multi-level decision-making and scientific uncertainty with regards to regulatory procedures, the co-production framework helps to understand more clearly the interrelationship between science and meaning-making in the social, legal and political world (Weimer and de Ruijter, 2017: 3).
14
The European Union is conventionally portrayed as a special case in Political Science since it is neither a state nor is it comparable to an international organisation. As Jachtenfuchs (2010: 210) observes, the EU is a “special multi-level system, in which governments, sub-national actors, EU institutions, private interests and other actors interact”. Moreover, the EU, due to its diverse nature, finds it increasingly difficult to find and revise a balance between unity and diversity (Jachtenfuchs, 2010: 211). The classification of the EU as a special multi-level system was partly inspired by the governance turn in political science that was able to capture the nature of this multi-level construct (Kohler-Koch and Rittberger, 2006; Börzel, 2012: 382).
However, before this section proceeds further in explaining the EU’s multi-level character as a governance arrangement, it is worth elaborating on the conceptualisation of the term governance itself. This will help in understanding its impact on studies of the EU as well as on the development and importance of experimentalist governance. The term governance can be summarised as a shift from hierarchical governing towards a more flexible and horizontal form of policy co-production between institutional actors and stakeholders that goes beyond territorial and jurisdictional boundaries of the nation state (see for example Bartolini, 2011). Therefore, governance can be perceived as an umbrella concept for authoritative decision-making (Bartolini, 2011). This change from hierarchical governing was largely a response to a crisis of the state due to, inter alia, its weakened position in an increasingly globalised world and the resulting limits in the state’s problem-solving capacities (e.g. Bevir, 2011). Yet, this does not mean that the term governance only applies to an international setting in which nation states must work together. Likewise, Kohler-Koch and Rittberger (2006: 31) note that the departure from hierarchical governing, or the role of the state as the main ‘steerer’ for realising public goods was abandoned in favour of concepts that were more open to cooperative forms of governing. In this vein, Rhodes (2007: 1246) defines governance as “a new process of governing: the methods by which society is governed”. Since these new methods are not only more decentralised but also include a broader range of actors in the decision-making process, Rhodes argues that governance can furthermore be referred to as “governing with and through networks” (Rhodes, 2007: 1246).
Applying this concept of governance to the study of the EU bears the advantage of being able to account for the ways in which various actors and entities work together in a horizontal and more decentralised manner. From a theoretical perspective, governance thus
15
theorizes EU governance as non-hierarchical mobilizing networks of private as well as public actors, who engage in deliberation and problem-solving efforts guided as much by informal as by formal institutions. (Pollack, 2005: 36)
Moreover, and as a result of this theorisation, the concept of a multi-level system or governance system itself was useful in trying to understand the functional complexity of the EU and an attempt to focus on how the EU works (Versluis, van Keulen and Stephenson, 2011: 28). On a practical level, it was additionally able to offer solutions to accommodate the diverse nature of the EU, and it became clear that ‘one size fits all’ solutions were hardly applicable in the diverse context of the EU (Craig & de Burca,2015). As, Hooghe and Marks (2003: 235f) argued, decision-makers in a multi-level governance system are able to adjust the scale of governance to reflect heterogeneity since the multi-level nature offers flexibility as a result of the dispersion of governance across multiple jurisdictions. Therefore, Pollack (2005: 36) stresses that a collective problem-solving approach offers a normatively superior alternative form of policy making in a multinational Union.
3.2 Experimentalist Governance
Although the literature on governance in the EU offers a variety of approaches to conceptualise the interaction of actors such as ‘deliberative supranationalism’, ‘network governance’, ‘informal governance’, ‘new and soft modes of governance’, they have not been able to fully capture the changes over the years in the architecture of policy-making within the EU (Börzel, 2012). What was missing was a way to categorize the “fundamental design for law making, and the way this design transforms the distinct elements of EU governance by connecting them to a novel whole” (Sabel and Zeitlin, 2008: 272). As a result of an analysis of several policy fields in which they identified these new fundamental designs and distinct elements of EU governance, Sabel and Zeitlin called this new architecture ‘experimentalist governance’ (Sabel and Zeitlin, 2008; 2010; 2012). Broadly speaking, experimentalist governance can be defined as a framework for rulemaking and revision through a recursive review of implementation experience in different local contexts (Sabel and Zeitlin, 2012: 169), whereby experimentalism can be understood as
experimental in the sense that they will be entertained subjects to constant and well-equipped observation of the consequences they entail when acted upon and subject to ready and flexible reason in light of observed consequences (Dewey, cited in Sabel and Simon, 2011: 78)
16
The main procedural characteristics of experimentalist governance are four independent elements that are organised in a cycle. First, framework goals, alongside measures for gauging their achievement, are established in joint action of Member States and EU institutions, which is typically undertaken in consultation with relevant stakeholders from the civil society and businesses. Framework goals can include examples such as good water or good education. Second, lower-level units, e.g. national ministries, regulatory authorities and the actors with whom they collaborate are given the freedom to advance these goals as they see fit. By granting lower levels a high level of autonomy, the principle of subsidiarity is respected. Third, in return for the autonomy granted to them, these lower levels must report on their performance, especially in relation to the agreed indicators, and participate in a peer review process in which the results are compared. Fourth, and last, in light of the results of the former three steps, the framework goals and procedures are periodically revised by the participating actors who established them, augmented by new participants whose views come to be seen as indispensable to deliberation. The cycle then repeats (Sabel and Zeitlin, 2008: 272-274). The advantage of experimentalist governance is thereby that it
systematically provokes doubt about their [actors’] own assumptions and practices, treat all solutions as incomplete and corrigible, and produce an ongoing, reciprocal readjustment of ends and means through comparison of different approaches to advancing common general aims (Sabel and Zeitlin, 2012: 170)
Alongside identifying the procedural workings of this governance system, Sabel and Zeitlin define the form of governance itself as a directly deliberative polyarchy. It is deliberate because, when confronted by a requirement to justify its own positions as well when confronted with various other practices by other Member States, arguments can be used to disentrench settled practices and open up actors to reconsider the definition of group, institutional and national interests. It is directly deliberative as it applies concrete and diverse experiences of actors to current problems, facilitating the emergence of new possibilities for considerations among the actors. Lastly, it is polyarchic because it is a system in which local units learn, discipline and set goals for each other.
However, a directly deliberative polyarchy requires three broad conditions. First, there needs to be a multipolar distribution of power, i.e. there is no centralised power that is able to impose its will. Second, the underlying political environment in which it is embedded in needs to be diverse, and actors within that environment need to be highly interdependent. In the EU, the political environment is characterised by diversity with its 28 Member States and various regions. Third, subjects that need to be addressed are required to be characterised by
17
strategic uncertainty. That is, although there is a lack of clear and precise goals, and there is no precise knowledge as to how these goals should be achieved, it is realised that only coordinate action can effectively tackle these issues. With regards to risk regulation, uncertainty can be observed in many cases in which it is not possible to establish scientific conclusiveness as to whether or not risks might materialize into hazards. However, by accommodating these factors and by transforming them into a main driver of learning, this governance arrangement offers the possibility to transform an obstacle of integration into a main asset in order to achieve high standard outputs.
3.3 Co-Production
The theoretical lens of co-production helps to uncover the “dynamics of how science begets meaning in the social, political, or legal world and vice versa” (Weimer and de Ruijter, 2017: 9). Co-production is defined as
shorthand for the proposition that the ways in which we know and represent the world (both nature and society) are inseparable from the ways in which we choose to live in it. Knowledge in its material embodiment are not at once products of social work and constitutive forms of social life: society cannot function without knowledge any more than knowledge can exist without appropriate social support (Jasanoff, 2004: 2)
The concept of co-production has had an influential impact on studies of risk regulation as it intellectually challenged the idea that risk can be treated as an objective phenomenon. On the contrary, risk is understood as being framed, or in the words of Jasanoff (2004: 4), “co-production can be seen as a critique of the realist ideology that persistently separates the domain of nature, facts, objectivity, reason and policy from the culture, values, subjectivity, emotions and politics”. To exemplify this line of argument, it is useful to consider that risk is conventionally treated as a scientific fact which can be analysed and controlled. Through the idea that risks can be controlled, policymakers can legitimise their actions by appealing to their capacity to protect people from these risks. However, the initial process of identifying a risk that necessitates a scientific fact finding is inherently political, and thus influenced by normative and social considerations. This is because it is the policymaker who decides that a scientific finding constitutes a risk from which the population needs to be protected. As such, a risk itself can be regarded as a determinant of social order, that is, it does not act exogenously upon society but “it arises from ‘within social arrangements’” (Weimer and de Ruijter, 2017: 9). Utilising this view challenges the perception that treating risk governance is a mere technocratic exercise, but instead re-defines it as highlighting the “often visible role of
18
knowledge, expertise, technical practices and material objects in shaping, sustaining, subverting or transforming relations of authority” (Jasanoff, 2004: 4).
In the context of the EU, co-production plays an important part in the self-understanding of the EU as a legitimate policy actor. In risk regulation, science and politics are perceived to be separate, meaning that subject matter governance is rationalised by appeals to scientific facts and the EU itself as a governance arrangement for political subjects, i.e. its citizens (Jasanoff, 2017: 20). Thus, the EU, by using science as a basis for regulatory action, legitimises itself as a protector of citizens from risks. If the EU manages to protect its citizens by being able to effectively identify risks and by being able to effectively regulate, i.e. control, them, EU institutions are regarded as successful. Therefore, Jasanoff (2017: 24) claims that science has played a constitutional role in European governance. The idea of scientific excellence as a legitimising factor of EU governance can, moreover, explain policies around delegating tasks of scientific fact-finding to experts in agencies. By outsourcing these tasks to independent agencies consisting of the respective experts, the value of scientific excellence is underlined as not only rationalising the separation of risk assessment and risk management, but it is also used to reinforce the EU’s claim to know what is right for the entire European polity and its subsequent claim to legitimacy. However, as the example of EFSA shows, this delegation can also lead to a situation in which experts are deliberating behind closed doors, thus allowing less public influence and scrutiny (Jasanoff, 2017: 28).
A second point in which co-production is identified is the process of standardisation. Standards themselves are products of political consideration. Safety standards are a case in point. While on the one hand, safety standards are driven by the desire to harmonise markets, they are, on the other hand driven by the desire to prevent markets from being threatened by their own products. As such, by establishing those standards, the world is ordered in forms of allowed and allowable transactions while making others undesirable or impossible. In discussion about what a standard should look like, it is likely that various divergent perspectives as to what is deemed satisfactory in standardisation are brought together to determine what hazards are claimed to be unacceptable to the public. It takes political judgments to decide on what is deemed acceptable and what should be controlled through standardising. Results of such judgments and deliberations are then enshrined in laws and regulation. According to Fisher (2014: 165), regulation thus creates demarcations and boundaries that characterise objects as hazardous or harmless, safe or risky, natural or unnatural, therefore shaping the understanding of the world of political subjects. Moreover,
19
regulations establish not only the endpoint of risk assessments, but they also prescribe a way to assess them or standardise the responses to them. The latter is then the practical way of implementing the preconceived understanding of the world. Therefore, Fisher (2014: 165) argues that “co-production occurs through legislative frameworks and the operations of law through administrative practice.”
To conclude, co-production applied to EU risk regulation reveals that the EU itself, through its scientific agencies, claims a superior knowledge for the entire polity and thus constructs an image of a united Europe. Moreover, it relies on the perception of science as being objective and absolute, and which it uses to back up its political power through the reliance on science. Thus, how regulators collect evidence, how they frame their technical arguments and whether they are challenged by the public are relevant to the development and maintenance of political order (Jasanoff, 2017: 34f.). Conflicts over regulatory procedures and outcomes might therefore be defining moments in the constitution of political legitimacy and political order, making it interesting to use these theoretical insights in understanding the European pesticides policy, especially in the case of glyphosate.
4. Methodology
4.1 Case Selection
In order to answer the main research question and the sub-questions, this thesis conducts a qualitative case study of the EU pesticides regime as well as a within and in-depth case study of the authorisation procedure of glyphosate. The choice of analysing the European pesticides regime is based on the assumption that it offers a particularly interesting case in the field of risk regulation, as pesticides are commonly represented with two divergent views, namely the positive view that pesticides are essential for modern agricultural practices and the negative view which stresses the adverse consequences of its use. Moreover, pesticides policy within the EU involves scientifically contested decisions and the prospect of uncertainty. Thus, it is appropriate to speak of pesticides policy as a ‘wicked problem’ in which policy-making processes “are fraught with conflicting interests, radically divergent problem framings, competing alternative solutions, and highly uncertain outcomes” (Hoppe, 2011: 8). The characterisation of pesticides policy as a wicked problem, therefore, has an impact on the potential on the emergence of regulatory conflicts, as regulation must consider both the agronomic and economic advantages alongside the negative consequences for health and for the environment (Pelaez, da Silva and Bourges Aravjo, 2013: 643).
20
On a more general level, the choice of the pesticides regime is due to its overall importance for the chemical sector. Pesticides are the most used man-made chemicals in the world (Schäffer et al. 2018), and Europe is the largest consumer of pesticides (Enserink, Hines, Vignieri, Wigginton, 2013). In terms of its advantages, it is stressed that pesticides are essential for guaranteeing a stable food supply, which is necessary to feed a constantly growing population. Yet despite these positive attributes, the application of pesticides can have undesirable negative consequences for sensitive environmental systems and for human health (Storck et al. 2017: 102; Bourguignon, 2017; Cuhra, Bohn and Cuhra, 2016; Bernhardt, Rosi and Genner, 2017). The increased attention on the negative consequences has therefore led some commentators to argue that the benefits of pesticides have largely been outweighed by their costs (Bourget and Guillemaud, 2016). Moreover, pesticides are also contested for their connection to GMOs. It is stressed that producers of both GMOs and pesticides might only allow the selling of a package in which their GMO products are resistant to the herbicides they sell, thus enabling farmers to cultivate their farmland more efficiently and reduce the use of pesticides5 (Hsaio, 2015). This, however, might create dependencies of farmers on certain firms that use their powerful position to limit both the former’s freedom of choice as well as pricing mechanisms. Therefore, pesticides use in this regard might create a socio-economic justice concern (Weimer, 2019: 6). However, it has to be noted that the use of GMOs is not common in Europe, hence the latter point does not have severe implications for the European market.
Moreover, policy developments in the pesticides field display issues that have also been observed in other parts of the risk regulation literature, for instance in GMO regulation. The strong reliance on science, political gridlock in comitology committees, and regulatory controversies over scientific opinions are present in singular authorisation procedures for pesticides, of which glyphosate is one. That is also the reason why glyphosate has been chosen as a within-case study. This within-case will, on the one hand, exemplify how regulatory procedures of pesticides regulation work in practice, and, on the other hand, illustrate the weaknesses and controversies surrounding the authorisation process (Storck et al. 2017: 1028). Moreover, as this thesis will analyse reform proposals of the current
5
It should, however, be noted that the combined use of GMOs and pesticides does not necessarily lead to an overall reduction of pesticides use. The literature points out that, for instance, the use of pesticides had been reduced for its application on modified Bt (Bacillus truingiensis) crops, whereas it has increased for other uses, e.g. with herbicide weeds (Hsaio, 2015). In case of glyphosate tolerant crops, Perry, Ciliberto, Hennessy and Moschini (2016) found similar results in that adopters of genetically modified glyphosate tolerant soybeans used 28% more herbicides than nonadopters while adopters of genetically modified insect resistant maize used around 11 % less insecticides.
21
authorisation system, the glyphosate case is insofar important to analyse as it triggered a debate about how to reform the system. This debate was most visibly epitomised through the submission of a European Citizens’ Initiative and the accompanying public attention as well as through the establishment of the PEST Committee in the European Parliament. Thus, glyphosate as a within-case study is treated as an extreme case for the current authorisation system as it tests the latter’s limits and offers the potential to reveal deficiencies which might lead to its overhaul.
When analysing this case, a focus is placed on uncovering “the relations between possible causes and observed outcomes” (Venesson, 2008: 231). It is important to understand the actors involved, their decisions, as well as their impacts on political and public dynamics. This should help to craft a thorough picture of a chain of events that is able to make a complex case understandable (Della Porte, 2008). However, besides the focus on the internal validity of this case, studying pesticides policy in the EU as well as the authorisation process of glyphosate can also be situated in the broader in the framework of risk regulation and, more specifically, within the literature on agro-chemical risk regulation and co-production.
4.2 Data Collection and Analysis
To analyse this case, a dual strategy of document analysis and interviews has been employed. Moreover, to cross-check both data sources, the method of triangulation will be employed, i.e. “using more than one method or source of data in the study of social phenomena” (Bryman, 2012: 392). Secondary literature on risk regulation from a variety of academic subjects as well as on pesticides is used to get a basic overview of the policy field, and, a wide range of primary sources is utilised, drawn from EU institutions and agencies, national authorities and case law. Commission and agency documents include regulations, directives, audit reports, staff working document, strategy papers, press releases, conclusions and guidance documents. Parliamentary documents include hearings, written answers to questions, commissioned studies, impact assessments and resolutions. Moreover, grey literature such as stakeholder reports and position papers will also be used to incorporate the views of a wide range of actors.
Besides the documentary analysis, elite interviews were conducted to get a deeper insight into the subject matter of the overall EU pesticides policy and also on the case of glyphosate. The aim of these interviews was to “move beyond written accounts that may often only represent an official version of events” and to “gather information about the underlying
22
context and build up to the actions that took place” (Tansey, 2007: 767). Moreover, interviewing as a research method was chosen because of the sensitivity of the subject, e.g. its high potential for political conflict, which requires trust between the researcher and the interviewee (Bymann and Bell in Kentrup, 2013). The interviewees were selected according to a purposive sampling approach in which “persons are deliberately selected for the important information they can provide that cannot be gotten as well from other choices” (Maxwell, 2008: 235). After an initial screening of the primary and secondary literature, the most important actors and stakeholders on the European and national level were identified and contacted. Actors and stakeholders included Commission officials, parliamentarians, agency staff from the EFSA, the European Chemical Agency (ECHA) and national authorities, NGO representatives as well as industry representatives and policy consultants. The inclusion of a broad variety of stakeholders should balance the flow of information that would inevitably result from the broad opinion base, as especially industry representatives and NGOs hold contrasting opinions in the field of pesticides (Bozzini, 2017: 10f.). Fourteen people were contacted, several of which agreed to be interviewed. Nevertheless, due to issues with finding appointments, no responses after follow-up mails, and issues with data confidentiality, only three people could finally be interviewed. These included a Commission official from DG SANTE, a representative from ECHA and a representative from the German Federal Institute for Risk Assessment, BfR.6 The interviews had a semi-structured set up, lasted between 30 minutes and 65 minutes, and were either conducted by phone or face-to-face. The questionnaires for the interviewees did not follow a standardised model, so that the questions could be tailored according to the institutional background of the interviewees. Using interviews also allowed for triangulation of data. As Tansey (2007: 766) writes: “when documents […] and secondary sources provide an initial overview of the events or issues under examination, interview with key players can be used to corroborate the early findings”. Thus, the validity of the data for the case is additionally increased and buttresses the robustness of the overall findings.
5. The EU Pesticides Regime
As has by now become clear, pesticides regulation is a vital task in both ensuring that the Single Market can function and in ensuring the protection of health of humans, animals, and the environment. Since scientific methods and knowledge, e.g. in the field of ecotoxicology, have steadily increased and improved in the last years (see Bozzini, 2017), regulation in the
6
