II. Highlights of the glyphosate authorisation renewal decision-making process

Initially, glyphosate was admitted to the European market in July 2002, a decision which was based on based on Directive 91/414/EEC. This directive has been replaced by Regulation 1107/2009/EU (on the marketing of Plant Protection Products). In the period 2012–2017 glyphosate was subject to scientific review in light of the expiring authorisation (July 2016). In this process of (re-)authorisation, the role of the so-called rapporteur Member State is key. According to the 2009 Regulation, the producer must submit the application to the Rapporteur Member State.Footnote ⁹ The latter checks the admissibility of the applicationFootnote ¹⁰ but most importantly compiles a draft assessment report that assesses, based on the state of play in scientific research, inter alia mutagenicity and carcinogenity risks.Footnote ¹¹ Germany, the rapporteur Member State for glyphosate, issued – by way of its Bundesinstitut für Risikobewertung (BfR) – a positive report, indicating in particular that glyphosate would not entail significant carcinogenicity risks. Although assessing active substances involves many more public health and environmental aspects, the controversy about glyphosate revolved indeed around the carcinogenicity risks. The responsible EU agency, the European Food and Safety Agency (EFSA), followed with an equally positive assessment, indicating that glyphosate was “unlikely to pose a carcinogenic hazard to humans”. ^{Footnote 12} The International Agency for Research on Cancer (IARC) had, however, concluded that glyphosate would be “probably carcinogenic to humans”.Footnote ¹³ This sparked a controversy about the reliability of EFSA’s findings. ^{Footnote 14}

Nevertheless, the Commission pushed for a renewal of the authorisation of glyphosate, but the adoption of the implementing regulation to this end is subject to the committee procedure system (or “comitology”) based on the Comitology Regulation.Footnote ¹⁵ Comitology concerns a unique feature of EU law, which allows the Member States to control decision-making over executive rule-making. The system was born in the 1960s when the Member States did not wish to overburden the legislature with the regulation of all aspects of the Common Agricultural Policy (CAP) but at the same time sought to avoid a situation in which the Commission would have considerable freedom to implement the CAP according to its own preferences. The solution was found in what is officially labelled as the committee procedure system.Footnote ¹⁶ It obliges the Commission to cooperate with committees of national experts in the adoption of implementing acts. The powers of these committees differ according to the procedure that is prescribed.

The applicable procedure in the case of the glyphosate reauthorisation is the examination procedure, which was introduced in 2011. If the examination committee does not deliver an opinion (which is the outcome if no qualified majority can be established either in favour of or against the draft implementing act), the European Commission shall not adopt the draft act if it concerns (inter alia) the protection of the health or safety of humans, animals or plants. Decisions on the (re-)authorisation of active substances are of a particular kind, as it is not possible to not take a decision: these substances must either be accepted or banned from the internal market. The options in a “no opinion” scenario for the Commission are to amend the draft implementing act and submit the new version to the examination committee, or to submit the original draft act to the appeal committee, which is a higher level committee.

In the glyphosate case, the Committee on Plants, Animals, Food and Feed (PAFF – the responsible committee in the field) could not reach a qualified majority either in favour of or against the proposed implementing act. The German abstention in the voting procedure was an important reason for this. Eventually, when the German representative in the appeal committee voted in favour, the committee could come to support the reauthorisation decision, albeit for a limited period of five years only.Footnote ¹⁷ This allowed the Commission to adopt the implementing regulation, an outcome it probably had not expected in light of the arduous decision-making on sensitive products and substances in the past. In the period 2014–2017 such decision-making persistently resulted in deadlock.Footnote ¹⁸ This was the reason for the Commission to propose amendments to the system of comitology, which will be discussed in the next section. In the meantime, the reauthorisation was disapproved of by Members of the European Parliament, which had just one week before held a public hearing on the Citizens’ Initiative (ECI) “Ban glyphosate and protect people and the environment from toxic pesticides”.Footnote ¹⁹ Inevitably, this gave the impression of the EP being bypassed. Comitology, however, is a system of executive rule-making and for that reason does not imply co-deciding powers for the Parliament. Such powers belong to the Parliament under the ordinary legislative procedure. Thus, the issue of the EP’s involvement essentially revolves around the question whether decisions on the authorisation of active substances should be for the EU legislature to make.

III. The Member States individually

Member States have a crucial role in the authorisation of active substances in the EU. They impact decision-making through comitology, which is – as elaborated above – a common feature of EU executive rule-making. In the case of the authorisation of specific products and substances an extra element of Member State involvement is added. Specific Member States are designated to act as rapporteurs, which means they compile the file and carry out the first assessment before the EFSA does. In other words, the Member States (at least the Member State rapporteur) are involved from the very beginning of the decision-making to the very end of it when the implementing act is adopted in comitology. The responsibility of Member States as rapporteurs may seem mostly, or even entirely, technical and procedural, but the glyphosate decision-making demonstrated that it may still put the Member State rapporteur in a difficult political situation. Apart from credibility issues of the competent authority, it has questioned the freedom of Member States to abstain or vote against decisions in comitology when their initial position has been positive. Formally, Member States indeed retain this freedom: no legal obligation exists under EU law that would bind the Member States in comitology to assessments made earlier as Member State rapporteur. Undoubtedly, this is at least partly due to the fact that the national representative in comitology (representatives of the ministry) is not the same as the actor which assessed glyphosate in the first instance (in this case, the German Federal Agency for Occupational Protection).

The response prompted by Germany’s role may nevertheless be viewed as a simple sanctioning measure. In May 2019 a Regulation was adopted, taking the rapporteurship away from Germany.Footnote ²⁰ The Regulation opened the possibility for a group of Member States – rather than an individual Member State – to assume the role of rapporteurs. The reasons put forward in the regulation are the expected workload and the complexity of the task related to the evaluation of a specific active substance; a desired repartition of the workload and a pooling of expertise. Nothing in the official considerations suggests that the change of the governance system would have anything to do with the German performance as rapporteur Member State. The composition of the current Assessment Group on Glyphosate (AGG) – France, Hungary, the Netherlands and Sweden – suggests, however, that the Commission and the other Member States were not even willing to trust Germany to be part of this system of collective responsibility.

Broader ramifications of the appointment of the AGG become visible when we zoom out from Germany’s special role. The appointment of a group of Member States reduces first of all the vulnerability of having just one responsible Member State. A common assessment by a group of Member States strengthens the legitimacy thereof, and is less likely to be set aside or overturned. At the same time, however, it may suggest that the assessments are considered to be (at least in part) of a political nature. Indeed, the involvement of a greater number of Member States suggests a greater reliance on input legitimacy arguments. This line of reasoning suggests that scientific assessments are not just purely objective, science-based and procedural elements of the authorisation procedure, but require support from a significant number of Member States. The counterargument, as put forward by the Commission, runs that the complexity of the issue and the need to pool the necessary expertise (rather than the alleged political nature of the issue) require the collective responsibility of a group of Member States. However, no evidence was put forward to demonstrate the inability of the biggest of Member States to adequately deal with the issue. Moreover, if expertise is indeed a crucial factor, it makes little sense to exclude the Member State that has built extensive expertise as a Member State rapporteur from the Assessment group, even if it has indeed lost credibility for the way in which it dealt with the initial assessment.Footnote ²¹ Thus, the suggestion that a political rationale underlies this change in the glyphosate decision-making system cannot be simply dismissed.

IV. The Member States collectively: comitology

The collective role of the Member States entered the limelight when the glyphosate renewal decision-making reached comitology. The assessment of the events must be considered in light of the origins of comitology and the assumptions on which it has been based. The first of these assumptions is that the EU legislature cannot simply regulate each and every aspect of all EU policies.Footnote ²² As is the case in other jurisdictions, the EU equally relies strongly on executive rule-making for the implementation of its policies. Particular to the EU system, however, is the principle that executive rule-making authority is decentralised: the prime responsibility for the implementation of EU legislation lies with the Member States.Footnote ²³ The Commission’s authority to adopt implementing acts is secondary and limited to situations in which a need for uniform conditions for the implementation EU acts exists.Footnote ²⁴ Thus, the Commission’s implementing powers are situated in the supranational zone which borders on one side the area of Member States’ individual powers to implement EU policiesFootnote ²⁵ and on the other the EU’s legislative zone, in which the Member States collectively decide on legislation in the Council. It proved difficult for the Member States to accept such an ‘in-between’ zone in which they would be excluded from the decision-making altogether. Moreover, they realised that implementing powers may still entail significant regulatory choices. Thus, they refused to accord the Commission a blank cheque.Footnote ²⁶ Involving national experts in comitology compensates to some extent for the loss of authority of the Member States to adopt implementing measures themselves. The third subparagraph of Article 291 TFEU establishes the competence for the EU legislature to regulate how the Member States control the Commission, which awards the system Treaty (and thereby constitutional) status. The scope of implementing authority of the Commission (and thus of comitology) is not only defined by the enabling provision from the parent legislative act. The CJEU has established that implementing acts must comply with the “essential general aims pursued by the legislative act” and must be “necessary or appropriate for the implementation of that act without supplementing or amending it”.Footnote ²⁷

The Treaty of Lisbon and the subsequent adoption of the new Comitology RegulationFootnote ²⁸ have brought important changes to comitology. Most notably, no longer do any of the current procedures directly involve the Council (or the EP). This was different under the old regime. The Regulatory Procedure with Scrutiny (known under its French acronym PRAC) gave the strongest powers of oversight and allowed the Council and the European Parliament to block the adoption of implementing acts. In the set-up of the Treaty of Lisbon, this procedure has been replaced by the concept of delegated acts (Article 290 TFEU).Footnote ²⁹ This formally excludes issues subject to delegation from the comitology system. Fierce institutional battles have been the result of this, as the European Parliament has consistently favoured delegation over implementation, while the Council has been a strong proponent of implementation instead. The possibility to ex ante impact the decision-making through national experts has, for the Council, been too precious to easily relinquish.Footnote ³⁰ The CJEU has had to decide on the demarcation between delegation and implementation but in doing so it has left considerable discretion to the legislature.Footnote ³¹ In particular, the degree of discretion awarded to the Commission is for the Court not a relevant factor for the choice between implementation and delegation.Footnote ³²

Two other procedures from the old comitology regime – the regulatory procedure and the management procedure – equally involved the Council if there was disagreement on lower levels. The present-day procedures, however, no longer provide any role for political institutions. The current most far-reaching procedure is the examination procedure. This procedure does include an appeal possibility which the Commission has had to accept.Footnote ³³ But the appeal procedure merely involves a committee of higher-ranking (but not political) representatives. The CJEU has made clear that it will strictly review whether the procedures laid down by the Comitology regulation have been observed.Footnote ³⁴ All in all, the comitology system post-Lisbon is arguably more technocratic in nature and is positioned more distinctly from the EU’s political institutions.

Here we arrive at the fundamental question on the foundations of comitology. Where does it derive its legitimacy from? Is comitology based on the need for technical expertise or rather on a need for political control?Footnote ³⁵ The dichotomy has been crucial in the glyphosate dossier as well. If one argues that the adoption of implementing acts essentially involves the translation of technical knowledge and expertise into law, comitology may be viewed as a form of objective and evidence-based decision-making aimed at effective and efficient problem-solving. In such a technocratic perspective, comitology is essentially based on output legitimacy. This brings along an argument to keep politics out of the decision-making altogether. An opposing view is that every elaboration of general legislative norms involves at some level a balancing of interests and values. Indeed, especially in fleshing out the details of an issue, crucial regulatory choices are made. Thus, “the technical” can never be fully separated from “the political”. This means that there is an inherent political dimension to comitology. Consequently, alternative legitimacy sources must be tapped into. This explains the European Parliament’s quest to get a stronger grip on comitology. However, this view not only relies on external democratic accountability mechanisms. Alternative legitimacy sources may equally be explored by viewing comitology as a form of deliberative democracy. In such a perspective, comitology serves as a forum for political processes and as a coordinating mechanism between supranational and national and governmental and social actors.Footnote ³⁶ This co-existence of different and even conflicting views on the nature of comitology and its legitimacy sources is bound to create institutional and political tensions.Footnote ³⁷ The ambiguous position of the Commission on comitology further fuels its undetermined status. On the one hand, the Commission has sought to free itself as much as possible from what it considered as unwarranted constraints of its executive autonomy.Footnote ³⁸ On the other hand, the regulatory practice has demonstrated a close and effective cooperation between the Commission and committees of national experts that led some scholars to the observation that comitology has integrated or fused national and European administrative domains.Footnote ³⁹

The glyphosate renewal process has been rich in comitology. In 2016, the Commission tabled a proposal on two occasions which both resulted in “no opinion”. Thus, a short extension of the authorisation was agreed to give the European Chemicals Agency (ECHA) time to draw up a report and for EFSA to present additional findings.Footnote ⁴⁰ The intention was to create the best circumstances for a well-founded and legitimate decision. When the findings of ECHA and EFSA became available in 2017, however, the decision-making circumstances had deteriorated even further. The contestation of EFSA’s independence and the submission of the Citizens’ Initiative to ban glyphosate complicated the discussions between the Commission and the Member States. As a compromise, a shorter renewal period became the most realistic alternative, but the PAFF in November 2017 again delivered a “no opinion” result. Very soon after, however, (the temporary extension of the authorisation was to expire by the end of 2017) the matter was put to the Appeal committee which could reach a qualified majority in favour of a shorter renewal period of five years.Footnote ⁴¹

In reaction to the arduous glyphosate decision-making, the Commission proposed a fourfold change to the comitology system, with the particular aim of avoiding “no opinion” outcomes by the committee of appeal.Footnote ⁴² The first proposal was to make public the voting behavior of individual Member States.Footnote ⁴³ According to the second proposal the voting rules would be eased, in the sense that abstentions and absences would not be considered as participating Member States for the calculation of the qualified majority. More problematic are the proposals to include the Council in the decision-making. The third proposal would essentially entail that the Council could act as the appeal committee, after at least one initial meeting of the appeal committee composed of representatives of lower rank.Footnote ⁴⁴ In addition, and this is the fourth proposal, the Commission could ask the Council as an institution for an opinion in case of “no opinion” in the appeal committee.Footnote ⁴⁵

These latter proposals are problematic as they blur the line between implementing and delegated acts. Indeed, a key distinguishing feature is that the systems of control are founded on different rationales. Whereas delegation is subject to institutional control (by both the Council and the European Parliament), implementing acts are controlled by national representatives on expert level.Footnote ⁴⁶ For Advocate General Jääskinen this reflects the dichotomy of interests that underlie the distinction between delegation and implementation: the primary concern for delegated acts, as instruments of “quasi-legislation”, is democratic accountability.Footnote ⁴⁷ By contrast, in case of implementing acts the primary responsibility of the Member States with regard to the implementation of EU law is the main consideration.Footnote ⁴⁸ Thus, the dichotomy between delegation and implementation reflects a “politics of separation” that characterises the EU’s legislative system (including the distinction between legislative and non-legislative acts, as will be seen later).

The Commission’s reform proposal is thus contrary to the Treaty of Lisbon’s constitutional arrangements on legislative acts. Both the third and fourth proposal entail a role for the Council in comitology (either formally or factually), a situation which the Treaty of Lisbon explicitly sought to bring to an end.Footnote ⁴⁹ Moreover, the Commission expressly justified involvement of the Council by the need to be able to allow for a political discussion on sensitive issues.Footnote ⁵⁰ This is not in line with the rationale for the system of control for implementing acts and, indeed, must rather be qualified as a consideration that fits the system of control over delegated acts. As an important consequence, it would be difficult to maintain the position that the European Parliament should remain excluded from comitology. Indeed, if control over the Commission is limited to national representatives at expert levels, there is sense in excluding the EP. This reasoning loses its credibility, however, if comitology at the appeal stage is guided by the need for “political discussion on sensitive issues”.

Furthermore, from a constitutional perspective it would be more convincing to seek solutions in another direction, namely in increasing (rather than diminishing) the Commission’s executive discretion. In light of the current legitimacy problems of the EU, and especially in light of the efforts of scholars and other actors to address the executive dominance problem,Footnote ⁵¹ this direction may seem counterintuitive. Yet, in this case there are good reasons to consider exactly this. It would entail that in situations in which Member States’ viewpoints (even in second instance) do not result in a clear collective position, the Commission should not need to search further to find legitimacy for its decisions. Indeed, comitology implies a system of control over the Commission, but this does not necessitate a situation in which the Commission (and, indeed, thereby society) should be paralysed where the Member States are unable to effectively exercise such control.Footnote ⁵² It would be more obvious to interpret a “no opinion” outcome as the Member States abandoning their chance to impact the decision-making. Moreover, granting the Commission the power to decide would establish a role for the Commission to act as an arbiter between apparently opposing national views or interests. In other parts of EU policy, this is exactly what the role of the Commission is.Footnote ⁵³

However, this requires consideration of Article 5(4) of the Comitology Regulation. This provision contains a list of situations in which the Commission may not adopt an implementing act in case of “no opinion”. The rationale of this provision is that some issues may be too sensitive to be adopted without sufficient committee support.Footnote ⁵⁴ The provision specifically mentions issues pertaining to the health or safety of humans, animals or plants. Formally, however, this exception does not apply here. The Plant Protection Products Regulation, adopted pre-Lisbon, initially referred to the regulatory procedure under the old Comitology Decision.Footnote ⁵⁵ The Comitology Regulation of 2011 automatically aligned the old regulatory procedure to the examination procedure. By way of transitional measure it provided that the only exception to the Commission’s discretion to adopt an implementing act (for measures previously subject to the regulatory procedure) in case of “no opinion” would be when the basic act would specifically prohibit this. The Plant Protection Products Regulation, however, does not appropriately reflect the transformation from the old Comitology Regulation to the new Regulation and in particular the change from the regulatory procedure to the examination procedure. Now that the new Comitology Regulation displays such a strong sensitivity to the need for sufficient support in case of implementing measures that relate to concerns of public health and safety of humans, animals and plants, the Plant Protection Products Regulation should make it explicit why it would deviate from that principle. The arguments to support this may be easy to find – especially as we are dealing here with reauthorisation procedures rather than initial authorisations – but they need to be specified. Thus, an amendment to the Plant Protection Products Regulation to justify why Commission discretion should be relied upon here would be appropriate and would take away the impression that the reauthorisation processes under the regulation are at odds with the Comitology regulation and Article 5 thereof in particular.

Apart from avoiding deadlock – especially in cases of (re-)authorisation in which “no-decision” after “no-opinion” is not an option – awarding the Commission the power to take the decision in case of no-opinion aligns better with the constitutional arrangements on implementing acts and their position within the wider Treaty framework as introduced by the Treaty of Lisbon. Whilst accepting the difficulties associated with the “politics of separation”, the reality is that the Treaty of Lisbon has sought to differentiate implementation from delegation under Article 290 TFEU and from legislative acts.Footnote ⁵⁶ Unlike delegation, issues subject to implementation do not qualify as “quasi-legislation” and should thus not involve a prominent role for the Council or the EP.Footnote ⁵⁷ Awarding the Commission the competence to decide in case of “no-opinion” would equally confirm the nature of comitology as a system of control over the Commission, rather than as a system of “co-decision” between the Commission and the Member States. The persuasiveness of the latter argument obviously depends on where one stands with regard to the nature of comitology and its legitimacy sources as outlined above.

V. Legislation

The glyphosate reauthorisation process is essentially a case of the (re-)politicisation of an issue placed within a technical decision-making context.Footnote ⁵⁸ Despite the controversies and conflicting interests that manifested itself in the decision-making process, the end result was still an implementing regulation and the matter thus did not reach the legislative level. At a deeper level of analysis, this politicisation demonstrated that the “politics of separation”Footnote ⁵⁹ on which the EU legislative system is built has something inherently precarious about it. The possibility of separating technical from political decision-making and distinguishing fundamental legislative choices from the elaboration thereof in technical detail seems in practice close to legal fiction. Thus, the mismatch between the formal status of the decision as an implementing act and its perceived political relevance fuelled discontent in the EP.

The criticism from the side of the EP on the glyphosate decision-making questions the dichotomy between legislation and executive rule-making. Such a dichotomy is not particular to the EU. Indeed, the EU shares with other constitutional orders a system of conditions and limits on the scope of the executive’s powers to adopt acts of general application.Footnote ⁶⁰ A key element thereof is of a formal nature: executive rule-making authority in the EU is derived from legislative acts. Such parent acts contain an authorisation clause to enable the Commission to adopt implementing or delegated acts. In the case of the glyphosate authorisation, the parent act – the Plant Protection Products Regulation – confers, in Article 13, the Commission with the power to adopt implementing regulations on the authorisation of active substances. The formal legality of the reauthorisation decision is therefore not at issue. Only a change to Article 13 would alter this, for example to exclude glyphosate from the reauthorisation process through comitology. Indeed, it is part of the legislature’s discretion to decide that decision-making on glyphosate would need to remain in the legislature’s domain. At the moment this is not a real option, and it is even unlikely whether the European Parliament itself would actually support this.

However, the dichotomy between legislative acts and executive rule-making in the EU has a substantive dimension as well. “Essential elements” of an issue have to be regulated at the legislative level.Footnote ⁶¹ This limits the powers of the Union’s executive institutions (which may not regulate such elements), but it equally limits the discretion of the EU legislature in what it may delegate. Nevertheless, the concept of “essential elements” is obviously elusive, and still allows the EU legislature to define to a great extent the border between legislation and executive rule-making. The CJEU has clarified, however, that political choices, based on a balancing of conflicting interests qualify as such in any case.Footnote ⁶² This still allows for flexibility, but it has at least given more concrete shape to the concept.Footnote ⁶³ It would not be too complicated to make a case that the glyphosate reauthorisation indeed involves essential elements in the sense defined by the CJEU. Tosun et al have demonstrated that glyphosate is widely used in conventional farming in Europe and has been employed for non-agricultural purposes as well. Thus, its authorisation involves “weighing the economic benefits of pesticides against the health and environmental costs associated with their use”.Footnote ⁶⁴ Although they do not make the point, this could easily justify why the renewal of glyphosate would need to be a matter for the legislature.Footnote ⁶⁵ The latter conclusion would impose itself even more if we would consider the approach that the Advocate General proposed in the Schengen Border Code case. He suggested to consider the sensitivity of the issue, the impact and intrusiveness of the measures proposed and the centrality of the measure to the parent act as key criteria. At least on the basis of two of the three criteria, glyphosate reauthorisation could qualify as an essential element.Footnote ⁶⁶

Beyond the “essential elements” doctrine, the EU legislature could in any case decide to keep the decision-making on market access of any plant protection products in its own hands. Whether this would be based on the essential elements doctrine or not, it would in any case imply a rejection of the current system, which is based on a legislative framework that lays down the procedural and substantive parameters of the authorisation process but leaves the actual authorisation decisions to the executive.

There is a compelling logic to this division as well. It creates a meaningful division of responsibility according to which the adoption of substantive and procedural parametersFootnote ⁶⁷ is largely in the hands of the legislature, whereas the actual application thereof is viewed as technical decision-making. In other words, rejecting this division and accepting that concrete (re-)authorisation decisions may qualify as essential elements as well, would imply that distinguishing between technical and political decision-making is rendered obsolete.