The divide in the EU green taxonomy: how conflict impacts the quality of policy advisory systems

Abstract

The second wave of research on Policy Advisory Systems (PAS) dynamics has induced scholars to rethink the received knowledge about the insider–outsider and technical–political divides in policy advice, leading to more nuanced descriptions of PAS configurations. The long-term structural change of PAS has been the focus of this research agenda. In contrast, a research gap exists in analyzing the role of disagreement and conflict in shaping the configuration of advisory systems and their interactions with policymakers. We analyze a case of conflict in the policy advisory system supporting the European Commission in defining the EU Green Taxonomy, i.e., the classification of economic activities that underpins the new EU regulation of public and private sustainable investments. The dynamic effect of conflict is particularly evident in the case of the EU Green Taxonomy, where opposing values and interests have shaped the advisory process, raising questions about how policymakers navigate and utilize contested advice. Our findings contribute to understanding the relationship between pluralistic and conflictual advisory systems and government management of PAS.

Nearly 30 years ago, P. Weingart highlighted a paradox in the growing reliance on scientific knowledge in policymaking: “scientification […] produces its opposite, the politicization of science” (Weingart, 1999, p. 158). This caution is especially relevant today, as the increasing use of evidence to address complex, high-stakes problems often reveals underlying value conflicts, challenging the ideal of purely evidence-based policymaking (Bandola-Gill, 2021; Brown, 2015; Hoppe, 2005; Maasen & Weingart, 2005; Wesselink et al., 2013).

One widely advocated strategy for enhancing the production of policy-relevant knowledge under conditions of uncertainty and high decision stakes is to broaden the community of knowledge producers to encompass all individuals affected by the issue (Fischer, 2017; Kowarsch & Edenhofer, 2015; Scott & Baehler, 2010). The early proposal of post-normal science (Funtowicz & Ravetz, 1993; Ravetz, 2006) and the more recent research agenda on knowledge co-production (Bandola-Gill et al., 2023; Lövbrand, 2011; Wyborn et al., 2019) emphasize the importance of promoting inclusion, public deliberation, and the democratization of expertise.

These approaches assume that policy-relevant knowledge should be produced in a pluralist context, where conflicting interests and values are openly acknowledged and power inequalities addressed (Norström et al., 2020). Recent normative models have re-evaluated the role of disagreement in pluralistic settings, viewing it not as a problem but as a valuable resource. This perspective suggests that disagreement can help uncover partisanship, promote flexible and innovative solutions, and expose misinformation (Machin, 2020; Moore & MacKenzie, 2020). However, there is the risk that pluralism and conflicting values help fuel “a vicious cycle, [where] the […] undermined legitimacy of technical advice and the perception of value-driven advice [lead to] the development of a much more politically driven […] policy advice” (Capano et al., 2023, p. 280).

Despite the growing awareness of the importance of disagreement and contestation in policy advice (Head, 2024), it is still unclear how it influences the use of advice in policymaking. How do policymakers incorporate diverse and potentially conflicting expert advice into their policy decisions? Are they driven to proactively manage a conflictual advisory process to align its outputs with their policy preferences and goals?

We aim to empirically investigate these theoretical problems by analyzing a case where conflicting views impact the supply of policy advice and the relationship between advisors and policymakers. We will do so by exploring the case of the advisory bodies involved in defining the EU Green Taxonomy, i.e., the classification of economic activities that underpins the new regulation of sustainable investments of the European Union (EU).

For this purpose, we will use the conceptual tools made available by Policy Advisory Systems (PAS) research. This research agenda has elucidated several crucial aspects of the evolution of policy advice over time and offers a vantage point to tackle the interplay between increasingly pluralistic sources of advice and the political demand for policy-relevant knowledge.

The paper contributes to advancing the research agenda on PAS about conflict-related dynamics of advisory systems. At the same time, it helps better contextualize a controversial policy choice adopted by the EU in its green taxonomy.

The article is organized as follows. The remainder of the introductory section will briefly discuss PAS research and the role of policy advice in EU sustainability policies. Section “Advising the Commission on finance: the Platform on Sustainable Finance” will present our case, namely the Platform on Sustainable Finance (PSF). The “PSF and the controversy over nuclear energy and fossil gas” section will discuss the controversy between the PSF and the European Commission about nuclear energy and fossil gas. The final section will discuss the results and their implication for the research agenda on PAS.

Pluralistic and conflictual policy advisory systems

One of the defining features of the PAS research agenda is to focus on the relational and systemic aspects of policy advice (Halligan, 1995), prioritizing the “examination of how various advisory units and practices interact with one another, and the dynamics of system configuration and operation over time” (Craft & Halligan, 2017, p. 48).

The second wave of research on PAS (Craft & Howlett, 2012, 2013) has focused on the increasing diversity of sources providing policy advice. This shift has resulted in a more significant presence and importance of actors outside traditional public advisory units and publicly funded policy research. Evidence of the growing “externalization” and pluralization of policy advice is the significant role that private organizations, including think tanks and consulting firms, play in shaping policy advice (Diamond, 2020; Marciano, 2024). The growing complexity of modern policy challenges is one reason for the externalization of PAS (Craft & Howlett, 2013). This trend is further shaped by the growing emphasis on inclusive policymaking (Bochel & Evans, 2007) and stakeholder consultation within contemporary policy processes (Bäckstrand, 2006).

A second key development in PAS research was recognizing a trend toward the politicization of policy advice, i.e., “the extent to which partisan or nontechnical aspects of policy form the content of policy advice and thereby favors actors who deal in this kind of information and knowledge” (Howlett, 2019, p. 248). The notion of politicization problematizes the distinction between technical and political advice. It recognizes that technical policy recommendations can be politically utilized to promote specific political agendas and preferences (Daviter, 2015) and supply policymakers with valuable information for political competition.

Consequently, second-wave PAS research has challenged the notion that effective policy advice primarily comes from advisory bodies located in the proximity of government and subjected to a greater degree of hierarchical or contractual control (Howlett, 2014). The assumption that policy advisors are exclusively “speaking truth to power” by supplying technical and non-partisan knowledge on possible policy solutions has also been called into question. A more insightful depiction of advisors’ activity is that of a “fluid, pluralised and polycentric advice-giving […] characterised as ‘sharing truth with many actors of influence’ or ‘weaving’ policy knowledge” (Craft & Howlett, 2012, p. 85).

Craft and Wilder (2017) looked into the dynamic interactions that contribute to shaping the configuration of an advisory system. Their model is particularly insightful for addressing pluralist and politicized policy advice and its effect on knowledge utilization. Craft and Wilder introduced two explanatory variables: a. the “accessibility” of the policy subsystem to which the advice is directed (i.e., degree of openness of the policy community operating in that policy domain) and b. the “ideational compatibility” between the policy advice given and the ideational and material goals of the policy subsystems.

A “collaborative advisory network” operates in an accessible policy subsystem, where policymakers and the relevant stakeholders in that policy domain are open to considering new solutions and different perspectives on issues. If access to the policy subsystems is restricted to actors and ideas aligned with the prevalent goals of policymakers, we have a “hegemonic advisory network”. If, on the other hand, the access is unrestricted and multiple beliefs and prescriptions clash among themselves, and with the goals of the policymakers, the advisory network is “contested”. According to the model, pluralist advisory networks will demonstrate high accessibility, while a low degree of accessibility is linked to corporatist networks. A high level of ideational compatibility fosters consensus-based policy advice, whereas a low level leads to adversarial advisory networks (see Figure 1).

Craft and Wilder discuss the factors that limit compatibility, hindering the “process of sense-making and consensus formation during which interpretations of problems and potential solutions are discursively constructed and justified” (2017, p. 227). They propose four “drivers of policy ideational compatibility”: the background and foreground, normative and cognitive content of the advice (Campbell, 2002); the purpose (learning, political, or instrumental) that motivates policymakers to utilize the advice (Weible, 2008); the nature of the problem, its complexity, salience, and urgency; the relations of trust or mistrust among the participants in the advisory system.

Regarding the dynamics configuration of PAS, an increase in accessibility is anticipated to shift the advisory system from corporatist to pluralist. Moreover, this shift impacts the ideational compatibility of the advisory system:

1. If access involves advisors who do not share the same beliefs and policy preferences as the actors previously engaged in the policy subsystem, we should anticipate a reconfiguration of the PAS into a contested policy advisory network.

Krick (2015) introduced the concept of “hybridity” to capture the evolving nature of policy advice in light of a heightened demand for accessibility to policy advisory bodies (Veit et al., 2017). The principles of participatory governance and inclusive policymaking have driven the expansion and diversification of advisory sources (Krick et al., 2019). As a result, modern advisory bodies integrate a broader array of expertise, combining scientific and professional knowledge with inputs from socioeconomic interest groups and advocacy organizations (Collins & Evans, 2002).

However, inclusivity comes with challenges. Interactions within hybrid advisory bodies often involve “a mixture of arguing and bargaining [that] allows for the reconciliation of conflicting interests and respect of differing values” (Krick, 2015, p. 489). Yet, consensus through this process of “negotiated knowledge” production is far from guaranteed, particularly in high-stakes policy discussions.

Hybrid advisory bodies are expected to include a wide range of ideas regarding desirable policy objectives and practical policy solutions. As in Craft’s and Wilder’s model, the diversity of perspectives and interests increases the likelihood of disagreement. Krick suggests that policymakers’ purpose for using the advice affects their interactions with conflictual hybrid PAS. The objective of obtaining diverse policy suggestions, for learning or problem-solving purposes, is compatible with the demand for pluralist and adversarial policy advice. On the contrary,

1. when policymakers are motivated by the political usage of policy advice—i.e., reinforcing the legitimacy of pre-existing political decisions—they selectively adopt expert opinions that align with their preferences and goals.

By selectively engaging with experts, policymakers can prevent conflicting ideas about policy solutions from undermining the effectiveness of the advisory process as a means of legitimizing their policy preferences.

The investigation into the management of PAS offers valuable insights into the strategies that policymakers can deploy to ensure the PAS’s alignment with their policy preferences. PAS management refers to the ability of policymakers to steer the advisory system, influencing its knowledge outputs and impact on policymaking (Craft et al., 2024). Marciano and Craft (2023) outline several management practices that policymakers can use to influence the functioning of advisory bodies. These practices include, among other things, adjusting their operating procedures, resources, membership, visibility, and mandates. Such interventions can help reshape the structure and performance of PAS. From hypotheses (1) and (2), we derive our final research hypothesis:

1. The goal of aligning conflictual PAS with existing policy preferences explains why policymakers engage in PAS management.

Policy advice in the EU green taxonomy

Research on PAS has focused on examining how different geo-institutional regimes influence the functioning of advisory systems The insightful studies on the Westminster tradition of policy advice (Craft & Halligan, 2020) were complemented by works on other jurisdictions (Albert & Manwaring, 2019; Galanti & Lippi, 2023; Hustedt, 2019; Laage-Thomsen, 2022; Pattyn et al., 2022). While the literature has examined several European countries, only limited research has been devoted to studying the nature and behaviors of PAS at the EU institutions’ level. Research on advisory systems at the EU level has confirmed the general trend toward externalization and politicization discussed by second-wave PAS studies (Metz, 2013; Rogacheva, 2019)¹. These studies, however, have not clarified the impact of disagreement and conflict on the production and use of policy advice.

Policy advice has been recognized as a crucial component of the EU policy process since the Maastricht Treaty (Paré & Montpetit, 2009). As the executive branch of the EU government, the European Commission and its Directorates-General receive support for their policy work through various advisory activities, similar to how Ministries operate in national governments (Metz, 2015). The EU institutions have taken a leading role in implementing innovative regulations for stakeholder consultations and expert advice², which became benchmarks for member States’ national regulations (Krick, 2021). Virtually, all the EU agencies engage in stakeholder dialogues and seek the advice of expert groups (Arras & Braun, 2018). The European Commission, however, has a peculiar position in the EU demand for policy advice. It continuously “consumes” policy recommendations and has formal relationships with external advisory bodies (Gornitzka & Sverdrup, 2008). The Commission is tied to this vast advisory system for a crucial institutional reason deriving from the reform of the legislative process initiated by the Treaty of Lisbon, which introduced the delegated acts (2008/C 115/01, Article 290). When the Commission exercises its power to initiate legislative initiatives, drafting proposals that the European Parliament and the Council will amend or reject, it can consult with expert groups as a means, among others, to receive feedback from stakeholders. However, when the Commission intervenes on laws already approved by the Parliament and the Council with non-legislative acts—the delegated acts by which the Commission amends or supplements existing laws without altering their core determinations—it must include in the policy process the expert groups it formally recognizes (Brandsma, 2016)³. After the Treaty of Lisbon reform, the power over the technical specifications that the Commission can exercise through its delegated acts is somehow checked by the expert committees the Commission consults. As shrewdly noted by K. Siderius and G.J. Brandsma, this arrangement causes the expert groups to “simultaneously act as lobbyists and ‘fire alarms’” (Siderius and Brandsma, 2016, p. 1298).

According to the Commission’s Decision (3301 final 30 May 2016), the expert groups can be formed by five types of experts (Article 7): (a) individual experts (of either academic or professional background) appointed in their personal capacity, (b) individual experts who represent sectoral or public interests, (c) representatives of economic or civil society organizations, (d) representatives of Member States’ national or local administration, and (e) representatives of public agencies from the EU, third countries, or international organizations. The Decision establishes that non-institutional experts should be selected through a public call for participation. Although there is a nominal commitment to a “balanced representation of relevant know-how and areas of interest” (Article 10), no formal procedure or standard has been defined to direct the selection process towards this goal.⁴

The EU green taxonomy

Our case study focuses on the advisory system involved in the EU Taxonomy, i.e., the classification of economic activities that underpins the new EU regulation of public and private sustainable investments. The taxonomy is a central piece in the regulatory effort the EU undertook to move forward with the sustainable transition envisioned by the European Green Deal, reaching the target set by the Fit for 55 policy package (reducing the GHG emissions in the EU by at least 55% compared to 1990 levels by 2030), and the EU carbon neutrality by 2050. These policy initiatives refer to the taxonomy to identify economic activities aligning with the EU’s environmental and social sustainability definition. The taxonomy intends to provide science-based guidelines to policymakers and the business community to set sustainability goals and improve their disclosure policy. It does not directly regulate economic activities since it does not discriminate between investments that can or cannot be carried out in the EU, nor does it set a mandatory level of sustainability performance that enterprises should reach. Nonetheless, it is expected to impact the cost of capital of public or private investments, contributing to moving financial flows away from unsustainable or less sustainable activities (Sørensen, 2009; Tettamanzi, et al., 2024).

The Taxonomy Regulation (2020/852), adopted by the Parliament and the Council in 2020, identifies six environmental objectives to which economic activities labeled “taxonomy-aligned” should contribute: (a). climate change mitigation, (b) climate change adaptation, (c) the sustainable use and protection of water and marine resources, (d) the transition to a circular economy, (e) pollution prevention and control, (f) protecting and restoring biodiversity and ecosystems Each economic activity in the taxonomy contributes significantly to at least one of these objectives while not hampering the others, according to the Do No Significant Harm (DNSH) principle adopted by the EU. Activities may either directly contribute to one of the six goals or be deemed an “enabling” activity, i.e., an activity that improves the environmental performance of another taxonomy-aligned activity. Finally, Article 10(2) of the Regulation establishes another type of economic activity that qualifies as taxonomy-aligned: the “transitional activities”, for which there is no technologically and economically feasible low-carbon alternative. These two labels (enabling and transitional activities) helped broaden the scope of the taxonomy by substantially increasing its definitional flexibility, a precondition for the controversial inclusion of nuclear power and fossil gas among the taxonomy-aligned energy sources, as we will discuss in greater detail below.

Advising the commission on finance: the platform on sustainable finance

The Commission Directorate-General for Financial Stability, Financial Services and Capital Markets Union (DG FISMA) formed multiple advisory bodies during the long gestation of the regulation on sustainable finance. The High-Level Expert Group on sustainable finance (HLEG)⁵ and, subsequently, the Technical Expert Group on sustainable finance (TEG)⁶ supported the definition of the EU strategy for sustainable finance since 2016. The TEG, in particular, contributed significantly to the 2020 legislative proposal on the taxonomy⁷. In both cases, most of the advisory bodies’ members were representatives of financial corporations or interest groups. In the case of the HLEG, financial actors were the absolute majority of the participants (60%), whereas in the TEG they were the relative majority (32.5%).

The Taxonomy Regulation contained the provision for creating a new advisory body (Article 20) replacing the TEG, the Platform on Sustainable Finance (PSF). Figure 2 presents the Composition of the PSF.

It is readily apparent that, in the new advisory body, the presence of the financial sector has been scaled down, while non-financial actors have gained centrality. Experts from business associations, NGOs, and researchers comprised more than half of the participants. The representatives of business associations increased from only 2 in the TEG to 12 in the PSF. The main sectors represented in the platform were the industrial ones (spanning from heavy industries to construction and chemical industries). NGOs’ representatives grew significantly, too, from 3 in the TEG to 13 in the PSF, forming the relative majority group.

The structure of the PSF

The 67 stakeholders participating in PSF (57 members and 10 observers) were distributed into four subgroups.

Figure 3 illustrates the composition and relationships of the PSF during its first term (Q4 2020–2022), the reference period for our case study. Data were processed using UCINET software and its extension NetDraw (Borgatti et al., 2002). The data used for the graphical representations were collected from the Commission’s register of experts.⁸

The main criterion determining the ties and arrangement of nodes in the network is the membership of the organizations in the various working groups of the Platform. Entities are represented through specific shapes that indicate each type of actor in the network. A box represents Universities, Research Institutes, and Think Tanks. An upward triangle marks Banks and Financial Institutions, while a downward triangle stands for Companies or Business Groups. A circle inside a box represents EU Institutions and Bodies. A squircle symbolizes International or Intergovernmental Organizations. Three superimposed tringles identify NGOs. A square depicts trade and business associations. A circle represents trade unions. A plus identifies additional entities and organizations, while a diamond denotes subgroups within the platform. Lighter symbols represent PSF’s members, while darker ones represent observers.

Figure 4 shows a similar representation, except that pending nodes (i.e., nodes connected to only one node in the network) have been removed.

This operation made the network more compact and focused, highlighting the organizations participating in multiple working groups and emphasizing their central role in cooperation among the different groups.

PSF’s contribution to the taxonomy

The advisors in the PSF were divided into multiple working groups. The Technical Working Group (Subgroup 1⁹) had to carry on the work started by TEG on technical screening criteria. The Data and Usability group (Subgroup 5) advised the Commission on the availability and quality of the data the taxonomy was mandating companies and investors to disclose, as well as on the issues that the users of the taxonomy may have while abiding by its requirements (such as misalignments between different regulatory frameworks, regulatory overload, etc.). The Negative and low impact activities group (Subgroup 3) contributed to extending the taxonomy beyond green activities. Finally, the Social Taxonomy group (Subgroup 4) produced policy recommendations on extending the taxonomy to social goals (such as labor rights, human rights, equality, non-discrimination, etc.).

All the participants in the PSF signed a confidentiality agreement with the Commission. For this reason, no full recording of their meetings exists, and they declined our interview requests. Data collection relied on the reports published by the subgroups, the Commission, and other advisory bodies (see Section “PSF and the controversy over nuclear energy and fossil gas”), as well as the minutes of the 26 meetings of the PSF between October 2020 and October 2022. Document analysis (Wood et al., 2020) was conducted to improve our understanding of the interactions within the PAS and its outcomes.

The PSF’s advisory role was broader than its predecessors, as it extensively assisted in preparing various delegated acts to the Taxonomy Regulation. Drawing from the work done by the TEG and the PSF, the Climate Delegated Act (Commission Delegated Regulation 2021/2139) introduced the technical screening criteria for mitigation and adaptation into the Taxonomy Regulation. The second delegated act, the Disclosure Delegated Act (2021/2178), detailed the taxonomy-aligned information that non-financial and financial actors must disclose according to Article 8 of the Taxonomy Regulation. The third delegated act, the Complementary Climate Delegated Act (2022/1214), led to a serious controversy between the Commission and the PSF.

PSF and the controversy over nuclear energy and fossil gas

The Complementary Climate Delegated Act, prepared by the Commission in 2022, addressed the most challenging areas for the taxonomy and its advisory bodies: nuclear power and fossil gas. The PSF’s predecessor, the TEG, considered energy production from fossil gas taxonomy-aligned only if it could meet a strict emission threshold through rapid technical innovation, such as extensive carbon capture and storage. TEG recognized the contribution of nuclear power plants to decarbonizing the EU energy mix. However, it could not conclusively assess its potential negative impact on non-climate-related EU environmental goals (i.e., the DNSH principle). The TEG could not reach a univocal recommendation on nuclear energy and suggested an in-depth technical inquiry from a dedicated expert group.

Through the Complementary Climate Delegated Act, the Commission classified these two energy sources as taxonomy-aligned, designating them as environmentally sustainable transitional economic activities. It lowered the emission threshold for fossil gas projects securing construction permits before 2030 and included the construction of new nuclear plants, the operation of existing ones, and research into next-generation nuclear technologies within the taxonomy’s transitional activities.

The activation of an ad hoc advisory process

Rather than relying on the expertise of the PSF to evaluate the alternative courses of action for regulating nuclear energy, the Commission chose to “set up a specific process for an in-depth assessment of the DNSH aspects of nuclear energy” (EC—European Commission, 2022, p. 3). The Commission’s in-house science and knowledge service, the Joint Research Centre (JRC), was asked to comprehensively evaluate nuclear energy production from the point of view of the EU’s non-climate sustainability objectives. The JRC—initially named Joint Nuclear Research Centre—was established by the treaty creating Euratom (11957A/TXT, Article 8). Although the JRC has expanded its research activities beyond nuclear energy, it continues to play a leading role in EU nuclear research¹⁰. In response to the request from the Commission, the JRC produced a report confirming that existing nuclear technologies comply with DNSH assessments and that managing contaminated waste is safe (Abousahl et al., 2021).

The Commission chose not to submit the report to the PSF for comments and instead submitted it to two other expert groups. The Group of Experts referred to in Article 31 of the Euratom Treaty is a registered expert group that the Commission consults about radioactive contamination and related health risks. This advisory body released a note supporting entirely the conclusions reached by the JRC (except for the opposing opinion of a single member of the group)¹¹. The second expert group was the Scientific Committee on Health, Environmental and Emerging Risks (SCHEER)¹², which the Commission can consult about health and environmental risks related to pollutants. The opinion given by the SCHEER stood in stark contrast with Article 31 experts and the JRC. SCHEER described the evidence of the compatibility between existing nuclear technologies and the DNSH principle as insufficient, and it stigmatized the acceptance of existing industrial standards and safety protocols as sufficient to prevent environmental and health risks¹³.

Despite the remarks from SCHEER—which the Commission acknowledged in the final draft of the delegated act—the JRC report was the primary source of policy knowledge used in the decision regarding nuclear energy.

The Commission did not proceed along a similar path when considering the issue of fossil gas. No existing or ad hoc advisory body was brought in to produce technical guidance on the topic. The Commission maintained that the decision was based on the climate modeling developed for the Fit for 55 Package, and no further technical advice was needed on the topic¹⁴.

Between December 2021 and January 2022, the Commission requested the PSF to provide feedback on the draft delegated act. The report produced by PSF rejected the proposal¹⁵. The new technical screening criteria for nuclear and gas electricity generation were deemed inconsistent with the Taxonomy Regulation, as existing knowledge of nuclear and gas technologies suggests that neither should qualify as green activities substantially contributing to sustainable transition.

The effect of the PSF policy recommendations on the Commission was almost entirely null. The Commission dismissed the PSF’s criticisms in the explanatory memorandum accompanying the Complementary Climate Delegated Act (EC—European Commission, 2022). It deemed the PSF’s recommendations inadmissible, arguing that they failed to acknowledge that nuclear and gas are activities “for which at present there is no technologically and economically feasible low-carbon alternative” (EC—European Commission, 2022, p. 8)”. Consequently, these activities were considered sustainable investments under Article 10(2) of the Taxonomy Regulation.

The aftermath

The delegated act was approved in February 2022 and passed the EU Parliament vote in July 2022. In the aftermath of the event—which brought some brief attention to the taxonomy in the media—five Civil Society Organizations¹⁶ resigned from the PSF, expressing dissatisfaction with the Commission and doubts about the credibility and the future development of the taxonomy¹⁷.

Furthermore, one EU Member State, Austria, decided to file a case against the Commission at the European General Court (Case T-625/22). Austria is seeking the cancellation of the delegated act on the premise that the Commission failed to comply with the procedural rules (by insufficiently involving the expert groups) and that the new delegated act contradicts the Taxonomy Regulation on several grounds. In April 2023, two additional cases against the Commission were filed, one by Greenpeace (Case T-214/23) and the other by multiple CSOs, including TT&E and WWF (Case T-215/23), seeking to force the Commission to initiate an internal review of the delegated act.

Case discussion: conflict in the EU sustainable finance advisory system

The literature emphasizes that single-case studies, such as ours, can offer valuable insights for future comparative research and serve as powerful tools for theoretical experimentation, particularly when cases are strategically selected for their explanatory potential or critical significance (Bennett & George, 2005). Two key factors drive the decision to examine the development of the EU taxonomy as a theory-generating single case study. First, the case is characterized by a high degree of diversity among advisors and involves a high complexity and decision stakes issue. These attributes provide an ideal context for exploring conflictual dynamics in PAS, offering insights into how disagreement influences policy advice production and its use by policymakers. Second, the taxonomy is particularly significant for the future of EU sustainability regulation, making it a critical subject for in-depth analysis.

The dynamics of the sustainable finance advisory bodies

Hypothesis (1), presented above (see Section “Pluralistic and conflictual policy advisory systems”), is confirmed by the changing configuration of the advisory bodies supporting the Commission on regulating sustainable finance. In Craft’s and Wilder’s terms, the HLEG and the TEG operated within what can be described as a “hegemonic advisory network.” Although access to the advisory bodies was not formally restricted, they comprised mainly a uniform group of experts, including financial professionals and representatives from the banking system, who shared a common understanding of the policy issues and goals they were addressing. The ideational compatibility between the advisory bodies and policymakers was high, as shown by the wide use of the advice produced by HLEG and TEG in the Commission’s legislative acts.

The turning point for this configuration was the creation of the PSF. The advice produced by the PSF and the policy objectives pursued by the Commission became less compatible, and the PAS moved to a “contested advisory network.” As seen (see Section “Advising the Commission on finance: the Platform on Sustainable Finance”), the public call for participation in the PSF resulted in a large and diverse advisory body compared to the previous two expert groups. Representatives from various sectoral interest groups, environmental advocacy groups, EU institutions, and financial professions joined the PAS at a crucial time since the Commission was working on translating the general objectives of the Taxonomy Regulation into regulatory requirements. Each participant anticipated the possibility of orienting the regulatory process towards what they see as the best possible scenario for sustainable finance. The belief that a significant decision is imminent and that “the time is right” to achieve or block a breakthrough (Goodin, 1998; Howlett & Goetz, 2014) helps clarify why more actors sought access to the PAS.

The shift in the configuration of the advisory network started when the mostly homogenous PAS on sustainable finance operating since 2016 was opened to alternative representations of policy problems and solutions (see Figure 5). The controversy on nuclear and gas revealed the growing disagreement within the PSF and the incompatibility between the Commission’s policy objectives and the advisory body.

The disagreement among PSF experts on nuclear and fossil gas is rooted in diverging ideas on what constitutes an actual transition to a sustainable development model. To illustrate this point, we can briefly consider the case of three PSF members participating in multiple working groups (see Section “The structure of the PSF”).

The WWF has published several critical documents about the inclusion of nuclear energy and gas in the taxonomy,¹⁸ stating that it would effectively mean “the end of the taxonomy.”¹⁹ WWF’s stance on nuclear and gas energy production follows its prioritization of renewable resources, environmental protection, and stricter regulatory interventions for sustainable transition (WWF EU—WWF European Policy Office, 2024).

On the opposite side of the spectrum, BusinessEurope—the EU lobbying group for the national business federations of 36 European countries—defended the inclusion of nuclear and gas. The organization publicly urged Members of the European Parliament to vote in favor of the Complementary Climate Delegated Act²⁰. BusinessEurope emphasizes containing the cost of the energy transition for European enterprises, as it advocates for a competitiveness-enhancing and business-friendly pathway to sustainability²¹.

Finally, the Climate Bonds Initiative (CBI)²² has differing views on nuclear energy and natural gas. It considers fossil gas a significant source of carbon emission, which should be rapidly phased out. In contrast, CBI views nuclear energy as a viable technology for decarbonization despite concerns related to nuclear waste disposal²³. It expressed confidence in the stringent criteria outlined in the Complementary Climate Delegated Act to provide the right signals to financial markets and facilitate a combination of flexible private investment decisions and clear public guidance, which is central to CBI’s transition model²⁴.

Managing PAS: the role of the Commission

As noted above (see Section “PSF and the controversy over nuclear energy and fossil gas”), the Commission chose not to refer to the PSF subgroups (see Section “Advising the Commission on finance: the Platform on Sustainable Finance”) to define the technical screening criteria related to nuclear and gas. The Commission stated that the information provided by the TEG, along with feedback gathered from public consultations and hearings with institutional stakeholders, was adequate for determining the technical aspects of the decision regarding gas (EC—European Commission, 2022). As for nuclear energy, the Commission tasked its public advisory service to clarify a central point of contention—whether nuclear energy production is compatible with the DNSH principle—and to draft the technical screening criteria for its inclusion in the taxonomy. The Commission excluded the PSF from revising the JRC’s report and limited its (mandatory) input on reviewing the delegated act, establishing a tight deadline of three weeks for submission.

The Commission managed the advisory process by operating on the mandates of its advisory bodies—advocating to itself the definition of the extended criteria for fossil gas—and selecting an internal source of advice for reviewing the thorny issue of nuclear energy. Following Marciano and Craft (2023), we can describe the management of the PAS as “authoritative” since the Commission demonstrated both the willingness and the capacity to intervene in the production of policy-relevant knowledge. The pluralization and externalization of the PAS were countered by the Commission procedure, which can be viewed as an issue-specific re-centralization of the advisory process.

This conclusion raises the question of why the Commission invested so much effort in establishing an ad hoc procedure for nuclear and gas. In a Q&A prepared after the approval of the Complementary Climate Delegated Act, the Commission justified the fast-paced procedure with the need “to help accelerate the transition to a cleaner energy future.”²⁵ Since the taxonomy is expected to influence investment decisions within the EU, and given that the energy transition is a crucial aspect of decarbonization, the urgency of this issue likely played a significant role in the Commission’s actions. However, the awareness that nuclear and gas entailed high decision stakes and “widely known divided positions” (EC—European Commission, 2022, p. 5) might have had a role in setting a parallel advisory process to the PSF. The Commission closely followed the work of the PSF’s subgroups, as the analysis of the PSF meetings’ minutes helps appreciate. Subgroup 3 was in charge of elaborating advice on extending the taxonomy categorization beyond green activities to include intermediate and significantly harming activities—the area of operation of the PSF where the discussion on nuclear and gas would naturally belong. In the draft report, submitted 5 months before the draft Complementary Climate Delegated Act, the subgroup considered categorizing investments in innovative gas energy production with lower CO₂ emission as “intermediate,” a step below the green taxonomy-aligned activities. Moreover, no explicit mention was made of nuclear energy. The report was not well-received by the Commission, which requested clarification on multiple aspects, effectively inducing a forced revision of the work.

The relationship between the Commission and subgroup 3, which included all three actors analyzed above (see Section “The dynamics of the sustainable finance advisory bodies”), exemplifies the growing ideational incompatibility within the PAS. Mitigating this incompatibility is a crucial reason for actively managing the PAS.

Concluding remarks

The case study explores the interaction between the pluralization and hybridization of PAS and conflict. Including diverse stakeholders—such as interest groups and advocacy organizations—within advisory systems enhances pluralism and increases the likelihood of entrenched disagreements over policy solutions. In the resulting “contested advisory networks,” conflicting ideas about desirable outcomes and recommended courses of action permeate the advisory system, increasing the likelihood of incompatibilities between the advice provided and policymakers’ priorities.

In the case of the EU Green Taxonomy, these dynamics incentivized policymakers to adopt specific PAS management practices, intervening in the mandates and functioning of advisory bodies to align policy advice with their strategic objectives.

This reflects broader trends identified in PAS research, where pluralistic advisory systems often require active management to reconcile diverse inputs with policymaking priorities.

The Commission’s management procedure—relying on a single, in-house advisory service with limited, expedited engagement from other experts—seems inconsistent in generating better, more nuanced, and innovative policy-relevant knowledge regarding nuclear and gas. Achieving such a goal would have benefited from a longer, collaborative process of evidence gathering, analysis, and peer discussion in dialogue with policymakers. Instead of facilitating the production of technical knowledge to improve policy choices, the Commission’s approach seems designed to obtain advice that could symbolically legitimize its pre-existing policy preferences.

Hypothesis (2), presented above (see Section “Pluralistic and conflictual policy advisory systems”), is confirmed by the case analysis. By reducing the role of the PSF in advising on nuclear and gas, the Commission sought to ensure it remained in control of the advisory process while preventing conflicting ideas over policy solutions from undermining its utility as a legitimization tool. While the first objective was fully achieved, the second was only partially successful. The attempt to manage conflict ultimately became entangled in the controversies it sought to resolve. Scientific polarization spilled into the public debate about including nuclear and gas in the taxonomy, with both environmental advocacy groups²⁶ and industrial lobbyists²⁷ invoking science to bolster their positions and accusing opponents of being unscientific. Moreover, the more openly critical participants exited the PSF, publicly criticizing the Commission’s handling of the matter.

Our case study underscores the relationship between pluralistic and conflictual advisory systems and government management of PAS. Our findings reveal that disagreement is a central variable in understanding the dynamic evolution of advisory systems. In the EU institutional context, hybrid (pluralistic and value-laden) advisory bodies are structurally integrated into policymaking as sources of technical knowledge and collective scrutiny of policy decisions. Incompatibilities and conflicts between advisors and policymakers—particularly over high-stakes decisions—will likely drive active PAS management, confirming hypothesis (3), presented above (see Section “Pluralistic and conflictual policy advisory systems”). However, managing the advisory process alone does not guarantee avoiding conflict. In our case, the EU Commission’s management of the PAS maintained control over the advisory process, but this was achieved at the cost of open conflicts among advisors, increased polarization of scientific positions, and the defection of key civil society advisors who publicly disagreed with the Commission.

To build upon the findings of this single case study and enhance their robustness, future research should incorporate a comparative approach at the EU and national levels. By examining multiple cases across geo-institutional contexts and scales, it will be possible to test the generalizability of the insights presented here and refine theoretical implications for PAS research.

Conflict of interest

None declared.

Footnotes

References