Introduction

Drawing on the view of energy efficiency as the ‘first fuel’ (International Energy Agency, IEA, 2013, 2019), the ‘energy efficiency first’ (EE1) principle was introduced in the vocabulary of European Union (EU) energy and climate policy with the European Commission’s (EC) communication on the Energy Union (EC, 2015) and the ‘Clean Energy for All Europeans’ legislative package (EC, 2016). It was manifested in Article 2(18) of the Governance Regulation (2018/1999), which defines EE1 as (EU, 2018, p. 15):

taking utmost account in energy planning, and in policy and investment decisions, of alternative cost-efficient energy efficiency measures to make energy demand and energy supply more efficient, in particular by means of cost-effective end-use energy savings, demand response initiatives and more efficient conversion, transmission and distribution of energy, whilst still achieving the objectives of those decisions.

In July 2021, the EC, (2021a) proposed to revise the Energy Efficiency Directive (EED, 2012/27/EU), setting the EU’s headline energy efficiency targets by 2030, and including provisions on national energy saving requirements, energy efficiency in the public sector and companies, etc. Among the most important provisions, the proposal entailed strengthened regulation of the EE1 principle, with provisions in a legally binding article mandating member states (MSs) to use the principle in planning, policymaking and major investment decisions in all sectors of the economy. The European Parliament (EP) and the Council of Ministers (Council) reached a political agreement on a recastFootnote 1 EED in March 2023Footnote 2 and the recast was formally adopted by the EP on 11 July 2023Footnote 3 and by the Council on 25 July 2023Footnote 4, making the EE1 principle legally binding (EU, 2023). This means that energy efficiency now should be considered in policy, planning, decision-making on major investments (i.e. euro 100 million each and euro 175 million for transport infrastructure projects) regarding the energy system and in sectors that affect energy supply and energy demand.

EU Commissioner for Energy, Kadri Simson, welcomed the adoption, which significantly raises the EU’s ambition and places a strong emphasis on energy efficiency—establishing EE1 as a fundamental principle of EU energy and climate policy, recognising its vital role in practical policy applications and investment decision-making:Footnote 5

Another milestone has been achieved today towards completing the ‘Fit For 55’ objectives. Our increased ambition and stronger measures on energy efficiency will accelerate the energy transition. The EU’s security of supply will be boosted, and our dependency on Russian fossil fuels will further decrease, in line with the REPowerEU Plan. The strengthened energy efficiency directive will help us achieve these goals collectively across the EU.

But how did this policy change happen in the complex institutional framing of the legislative procedures in the EU and how was the meaning of the EE1 principle formed at the interface of agenda-setting and decision-making under influence of different interest groups? Analysing policy does not only include their effects and effectiveness, but the values and beliefs attached to problem framing, different policy options and the ambiguities of decision-making on policies (Cairney, 2023). To understand the EE1 principle as a policy instrument, one must understand the policy processes and the politics of policy. Applying the advocacy coalition framework (ACF) (Jenkins-Smith et al., 2018; Nohrstedt et al., 2023; Sabatier & Weible, 2007) as a theoretical lens, this article aims at exploring and explaining the policy process and the politics, a tale of creation, of the institutionalisation of the EE1 principle as a legally binding principle in EU energy and climate policy, focusing on the following research questions:

  • Which actors and advocacy coalitions were involved in the policy process?

  • What values and beliefs did they hold on problems and policy solutions?

  • Which were the political conflicts?

  • Which were the paths to policy change?

The paper provides knowledge on how the concept of energy efficiency as the ‘first fuel’ was contextualised into a binding principle to be applied by different actors in the public and private sectors in EU MSs. The paper provides a unique insight into the processes of policymaking and policy change related to EU energy and climate policy in general and energy efficiency policy in particular. It gives researchers and stakeholders a glimpse into the secluded means of communication which usually characterise negotiations in the Council (Naurin, 2007) and between the Council and the EP (Brandsma, 2015; Roederer-Rynning & Greenwood, 2015, 2021). Thus, it can inform scholars of energy and public policy, policymakers and stakeholders on the policy processes, values and conflicts related to policymaking on the clean energy transition. The article will add to the growing literature on the processes of EU policymaking on energy efficiency (Dunlop & Völker, 2023; Dupont, 2020; von Malmborg, 2021, 2023a; von Malmborg & Strachan, 2023). The article will also add to the ACF literature, particularly the understanding of policy change through a combination of paths in mature policy subsystems.

The remainder of this paper is structured as follows. The ‘Notes on theory and method’ section presents the theory framing the analysis, summarises previous research on the EE1 principle and EU policy on energy efficiency, and discusses the method and material used. ‘Views on the EE1 principle’ section presents the results in terms of views on the EE1 principle held by policymakers and other stakeholders. ‘Discussion’ section analyses and discusses the findings, with a focus on beliefs and advocacy coalitions, as well as paths to policy change. Finally, ‘Conclusions and policy implications’ section draws conclusions and presents policy implications and topics for further research.

Notes on theory and method

Advocacy coalition framework

This study uses the advocacy coalition framework (ACF) as a theoretical framework for analysing the policy process on the EE1 principle in the EU. ACF is a network theory developed by Sabatier, (1986, 1988, 1998) and Jenkins-Smith, (1990) to explain how people and organisations work together to make public policy. It is considered one of the most prominent and influential theories for analysing and explaining the policy process and policy change (Heikkila & Cairney, 2018; Pierce et al., 2017).

ACF assumes that policymaking in modern societies is so complex that actors must specialise to be able to exercise any influence. This specialisation takes place within policy subsystems where actors regularly try to influence policies (Sabatier & Weible, 2007). A policy subsystem is defined by a policy topic (e.g. energy efficiency), territorial scope (e.g. the EU) and the actors influencing policy subsystem affairs (Sabatier & Jenkins-Smith, 1999). Importantly, policy subsystems are semi-independent and overlap with other subsystems and are nested within yet other subsystems (Nohrstedt et al., 2023).

ACF considers public policy to be ‘not just the actions or inactions of government, but also the translations of belief systems as manifested by goals, rules, incentives, sanctions, subsidies, taxes and other instruments regulating any given issue’ (Jenkins-Smith et al., 2018, p. 142). Thus, policies can be analysed in terms of belief systems and policy change corresponds to changes in belief systems (Sabatier & Weible, 2007). Jenkins-Smith et al., (2018) argue that policy actors have a belief system structure in three strata: (i) deep core beliefs, (ii) policy core beliefs and (iii) secondary beliefs. Deep core beliefs are fundamental normative values and ontological axioms. They are not policy-specific and can be attributed to several policy subsystems. In contrast to deep core beliefs, policy core beliefs are bound by scope and topic of the political subsystem and thus have topical and theoretical components. They can be normative and empirical, representing the translations of deep core beliefs to one specific policy subsystem (Sabatier & Weible, 2007). Empirically, they include overall assessments of the seriousness of the problem, basic causes of the problem and preferred solutions, such as the purpose of energy efficiency policy and EU targets for energy efficiency. Policy core beliefs are fairly stable over time and more resistant to change than the lowest level of the belief system, the secondary beliefs. These deal with a subset of the policy subsystem or the specific instrumental means for achieving the desired outcomes outlined in the policy core beliefs. They represent what can be described as the actors’ policy preferences with respect to concrete policy options, e.g. specific policy design, policy instruments, budgetary allocations and others. These preferences are more prone to change based on new knowledge and experience (Sabatier, 1998).

The ACF asserts that the actors’ behaviours and beliefs are embedded in informal networks, and that policymakers look for allies among people and organisations who share beliefs, particularly policy core beliefs, among, e.g. parliamentarians, government officials, leaders of interest groups, judges, researchers and intellectuals from various levels within the jurisdiction of policy. In addition, if actors have a significant degree of coordination, they form a coalition of actors—an advocacy coalition (Sabatier & Weible, 2007; Weible et al., 2011). Coordination means the formal or informal organised behaviours of actors in a coalition towards shared goals (Nohrstedt et al., 2023). Coordination can be strong, when actors deliberately plan their activities with allies, or weak, when actors in a coalition may unconsciously act in relation to their allies’ activities (Weible & Ingold, 2018). In most subsystems, the number of politically significant advocacy coalitions will be quite small, i.e. 2–4 important coalitions; the number being limited by all the factors which push actors to coalesce if they are to form effective coalitions (Sabatier, 1988).

Previous research

There are few academic papers on the EE1 principle. Rosenow et al., (2017) assessed the extent to which the ‘Clean Energy for all Europeans’ package keeps the promise of putting energy efficiency first, concluding that the package falls short of comprehensively reflecting the EE1 principle. Mandel et al., (2022) and Yu et al., (2022) analysed its potential role as a decision-making tool that prioritises demand-side resources over supply-side alternatives whenever these provide greater value to society in meeting decision objectives. Chlechowitz et al., (2022) analysed the application of the EE1 principle in EU member states. Based on limited data, they found that most countries fail to ensure an equal treatment between supply and demand-side resources and neglect the multiple benefits associated with energy efficiency improvements. EE1 is an important principle for reducing the costs of the energy transition, but it is not easily implemented in different sectors. Pató and Mandel et al., (2022) argue that the electricity sector is particularly relevant for using the EE1 principle. They show that an ‘EE1-compliant regulation guarantees that consumers can offer their flexibility and get compensated at market value and requires that distribution systems operators use them whenever they provide more net benefit that network investment’ (Pató & Mandel, 2022, p. 11). Von Malmborg, (2023a) analysed the political meaning of the EE1 principle using discourse analysis. He revealed a dispute among legislators and other stakeholders whether energy efficiency policy and the EE1 principle should aim at exploiting multiple benefits or climate change mitigation only. Interdiscursive communication settled the dispute and actors agreed in line with the multiple benefits discourse.

The EE1 principle shall be applied to policy, planning and investment decisions. Economidou et al., (2022) investigated the lessons learned from EU MSs national energy efficiency policy planning during 2007–2020. They found that policy planning has improved, particularly with the reporting requirements and two-step submission procedure under the EU Governance Regulation. Embedding energy efficiency in the larger climate policy framework, considering how it can support socio-economic challenges and multiple benefits, is of great importance.

Brown and Wang, (2017) reviewed the arguments of sceptics and advocates of energy efficiency. They found that a growing number of sceptics are voicing their concerns about the existence of ‘low-hanging fruits’, arguing that the energy efficiency opportunity is systematically exaggerated. They underscored many legitimate and important issues for analysts who take short cuts, fail to consider counterfactual evidence and jump too quickly to conclusions about positive benefits of energy efficiency. Advocates, in turn, counter by noting that methods are available to address these concerns and that the field has matured significantly. Also on policy framing, Fawcett and Killip, (2018) analysed the potential arguments of proponents of energy efficiency of buildings, referring to the multiple benefits of energy efficiency. Their experience is that these multiple benefits arguments are most persuasive when linked to the values and priorities of decision-makers and politicians, most of whom do not value energy efficiency as a benefit in itself. Kerr et al., (2017) found that recognition of multiple benefits may not equate with multiplied policy support, and instead it is more likely that different rationales will have relevance at different times, for different audiences.

MSs often contest EU policy on energy efficiency. Such contestation has been both sovereignty-based (subsidiarity claims), and substance-based (Wettestad et al., 2012; Herranz-Surralés, 2019). Dupont, (2020) analysed the strategies of the EC to manage the MSs’ contestation of EU policy on energy efficiency. It was concluded that the EC employ three main strategies to manage these contestations: (i) framing and reframing energy efficiency to enhance and consolidate authority at EU-level and to mitigate contestations over this authority; (ii) developing the legal framework; and (iii) applying flexibility in policy instruments and employing mixed soft and hard governance tools.

Dunlop, (2022) used discourse analysis to trace the evolution of EU energy efficiency policy over seven decades. Her analysis illustrates significant variations in the concept of energy efficiency over time, including how it is defined and measured. The meaning of energy efficiency gradually becomes broader, more detailed, and complex, as new concepts are attached to the term. She finds that a constant across the decades is that energy efficiency policy generally favours economic interests over social or environmental ones.

Deters, (2018) analysed how policymakers use various alternative decision arenas to avoid inconsistency between goals and measures on energy efficiency in the EU in a policy resulting from conflicting interests pulling in different directions within a heterarchical institutional setting. To avoid deadlock, negotiators rely on consensus-building techniques such as watering down, issue redefinition, and the setting of targets without actions. These techniques facilitate moving away from the status quo, but they come at the expense of coherence. Deters shows that alternative decision arenas may bypass conflict, and thereby make the use of consensus-building techniques unnecessary, resulting in more coherence.

There are few studies focussing on the entire policy process related to energy efficiency in the EU. Both von Malmborg, (2021) and Dunlop and Völker, (2023) analysed policy change related to the amendment of the EED in 2016–2018. Von Malmborg, (2021) analysed the advocacy coalitions and their beliefs as well as the paths to policy change in EU related to individual metering and billing (IMB) in multi-family and multi-purpose buildings, one of the policy instruments provided by the EED to improve energy performance of buildings. He found that the previously minor coalition, opposing IMB, grew and gathered enough support to outweigh the dominant coalition who was in favour of IMB when the EED was revised in 2018. This led to an internal shock and policy-oriented learning that changed the provisions on IMB. Dunlop and Völker, (2023) explored how the meaning of energy efficiency was negotiated through the way it is measured. These political negotiations are used as a case to empirically explore politics of measurement, i.e. practices and procedures through which the conceptualization and measurement of energy efficiency get entwined, and in which a seemingly ‘technical’ discussion weaves together scientific and methodological aspects with political, societal and environmental issues. In this way, their work explored how processes of politicisation and depoliticization in the formulation of energy efficiency indicators contribute to bringing about the very governance object that is energy efficiency. Von Malmborg and Strachan, (2023) analysed the policy process of the recast EED in 2021–2023, with focus on policy for energy efficiency in industry. They found large heterogeneity among MSs’ views, and between the Council and the EP, which were settled through deliberative negotiations and policy-oriented learning. Policy change was also a result of external and internal shocks.

Method and material

The exceptional and idiosyncratic nature of policy-making in the EU, with co-decision by the Council and the EP, calls for qualitative theorising and thick descriptions of empirical case studies, which invite readers to evaluate the interpretation of the data (Heisenberg, 2008). Thus, the research underlying this paper was undertaken as a qualitative case study using process tracing (Beach & Pedersen, 2013; George & Bennett, 2005), with the ACF as a theoretical lens for analysing the policy process leading to policy change.

The ACF has been applied in more than 150 studies of policymaking in the EU, of which most focus on policymaking on MS level. Only 15% of the studies are analysing supra-national policy making, i.e. EU policy, including analysis of how interest groups (IGs), MSs and the EU institutions are acting (Ingold, 2022). Applying ACF to studies of EU policymaking on a supranational level poses three methodological challenges (Ingold, 2022): (i) delineating the policy subsystem and choosing the appropriate venue(s), (ii) identifying actors and coalitions differentiating between formal and informal perspectives and (iii) challenges of data accessibility and availability. In addition, Henry et al., (2022) call for transparency and efforts to standardise elements of approaches in ACF research.

The policy subsystem analysed in this paper is delineated by the jurisdiction of the EU and EU policy on energy efficiency. Energy efficiency is one of five dimensions of the (EC, 2015) and a policy area in itself. However, the policy subsystem is linked to the larger policy subsystems of climate policy. But energy efficiency policy and climate policy are regulated by different articles of the Treaty of the functioning of the EU (EU, 2012). Energy efficiency policy is based on Article 194, whilst climate policy is based on Articles 191–193. In addition, they dealt with in different Council constellations and different committees in the EP. Boundaries of a policy subsystem are not fixed and often contested as policy actors redefine the policy issue by expanding or contracting the scope of conflict (Nohrstedt et al., 2023). Accounting for its expansion (Dunlop, 2022; von Malmborg et al., 2023), the EU policy subsystems on energy efficiency have endured for about 50 years and can be considered mature. Sufficient time has passed for policy actors to ‘learn enough about the issue to form and fortify their belief systems, mobilize and maintain coalitions, and support or oppose policies and programs’ (Nohrstedt et al., 2023, p. 132).

As for venues in the policy process, I focus on (i) the public consultation held by the EC prior to presenting its proposal for a recast EED, (ii) deliberations and negotiations in the Council and (iii) deliberations and negotiations in trilogues between the Council, the EP and the EC. These venues make it possible to identify opposing beliefs of different actors. The latter two are highly institutionalised, less accessible to researchers, with fewer documents to code and difficult for ACF scholars to research even though they might still be decisive to explain policy outcomes and change (Ingold, 2022).

When it comes to actors having beliefs and being part of advocacy coalitions, I have chosen to include actors at different levels in EU policymaking, i.e. the EC as agenda-setter, the Council and the EP as co-legislators, the national governments as the ones representing MSs in Council negotiations, companies, industry associations, environmental groups and think tanks. This approach is justified by the fact that they are all part of the EU policymaking process. It is a key assumption by the ACF that actors and coalitions seek to maximise their advantage; that, regarding the EU, could mean coalition activities and coordination at different levels and related to different EU institutions (Sabatier, 1998). Including governments of MSs as actors in studies using ACF is relevant, since they are considered key actors in general EU policy studies (e.g. Lundgren et al., 2022; Naurin & Wallace, 2008; Pollack, 2014). MS governments:

  • Receive EU legislation such as the EED for transposition into national legislation;

  • Try to influence the EC as agenda setter, before it puts its proposal on the table (Björkdahl, 2008);

  • Try to influence the EP rapporteur and other members of the EP (MEPs) (von Malmborg, 2022);

  • Try to influence each other, and find coalitions; and

  • Sometimes ally with IGs to influence the EC and the EP (von Malmborg, 2022).

The approach seeing MSs as actors is not common in ACF studies of EU policymaking, but has been applied previously (e.g. Rietig, 2018; Tosun & Schaub, 2017; von Malmborg, 2021; von Malmborg, 2023b; von Malmborg & Strachan, 2023). However, identifying beliefs of MS governments is a challenging task in studies of EU policy, since their positions are usually the result of negotiations within the country, and they can change with the next election.

To empirically identify advocacy coalitions, scholars search for clusters or groups of actors sharing a certain degree of belief (Ingold, 2022; Sotirov & Winkel, 2016). The collaboration is often overlooked. It is thus a bottom-up approach to identifying coalitions of like-minded actors. Coalition membership is mostly informal to the stage that it might be a construct of the researcher without actors knowing about the mere existence of or about their membership in a coalition (Weible & Ingold, 2018). As for coalitions among IGs in the case of EE1, there are formal ones like the Coalition for Energy Savings (CES) and European Alliance to Save Energy (EU-ASE) with a long-term focus on several pieces of legislation such as EED and the energy performance of buildings directive, and informal ones which band together related to specific issues such as EE1.

Based on ACF and previous research on policy processes and policy change regarding energy efficiency policy in the EU (von Malmborg, 2021; von Malmborg & Strachan, 2023), three propositions are made:

Proposition 1

There exist at least three advocacy coalitions, around the EP, the EC and the Council respectively.

Proposition 2

The EP advocates stricter and more far-reaching policy core beliefs and secondary beliefs in favour of the EE1st principle than do the EC and the Council, whilst the Council advocates a weaker role of the EE1 principle.

Proposition 3

Policy change followed an external shock and a negotiated agreement.

Empirical data to answer the research questions was collected through qualitative analysis of official, semi-official and classified documents related to the EU negotiations on the EE1 principle (Table 1). In addition, non-papers and reports from interest groups advocating the EE1 principle were analysed. The topic has had little media coverage, with negotiations being rather private and somewhat ‘confidential’.

Table 1 Documents analysed and their relation to the policy process
Full size table

These documents were chosen since they include data on the positions (beliefs) of the different actors involved in the policy process. To get a deeper understanding of positions of companies, business associations, environmental groups and think tanks, in addition to those mentioned in the EC public consultation, searches were made using Google. Searches were made for ‘recast+EED’, ‘energy+efficiency+first’, ‘efficiency+first’ and ‘efficiency+first+EU’.

Decision-making in the Council and in the trilogues is secluded (Elgström & Jönsson, 2011; Reh et al., 2013; Roederer-Rynning & Greenwood, 2021) and most research on EU policy draws on voting results since it is hard for scholars to get access to the negotiations. Through cooperation with the Swedish Ministry for Infrastructure (SMI), I was given access to SMI reports from the 15 meetings of the Council’s energy working party, three technical inter-institutional trilogue meetings and two political trilogue meetings. Such sharing of confidential information for research purposes is generally very rare (Lundgren et al., 2022). These provided a unique opportunity to analyse positions and changes in positions of different actors. This data also gave insights into which MSs collaborated with each other. This is a methodological merit. However, Sweden’s reports from negotiations are confidential, why positions of individual MSs cannot be revealed. I judge the likelihood that the findings would be systematically affected by bias to be limited. Swedish officials’ reporting from the meetings should have no incentives to falsely convey the positions of other EU MSs and the EP to the Government Offices of Sweden, since those positions are used to formulate Swedish negotiation strategies.

The qualitative text analysis was done manually, and searched for narratives, views and beliefs of various actors (collectives, not individuals) on energy efficiency in general and the EE1 principle in particular, and the reasons given for these. What policy options did the EC propose and what were the justifications for these proposals? What were the responses and counterproposals presented by the governments of different MSs, the Council as an institution, the EP rapporteur and the EP as an institution and other stakeholders such as companies, business associations, environmental groups and think tanks? Did the positions (beliefs) change during the policy process? First, data was coded as beliefs at different levels, from policy core beliefs to secondary beliefs. Dichotomies of beliefs, particularly policy core beliefs and secondary beliefs, were grouped as lines of dispute. The actors were then plotted in a matrix to find out who shared beliefs. A second coding was done to find out who collaborated with who. Significantly fewer actors collaborated than shared beliefs.

Views on the EE1 principle

The policy process leading to the recast of the EED, including mandatory provisions on the EE1 principle, started with the adoption of the Governance Regulation (2018/1999) in December 2018 (Fig. 1). The EE1 principle was introduced as a non-mandatory principle in the Governance Regulation. The EC proposal for a recast EED followed a public consultation held in winter 2020/2021. Negotiations in the Council and the EP commenced in the autumn 2021 and continued until summer/autumn 2022. Trilogue negotiations between the Council, the EP and the EC started in autumn 2022 and continued until March 2023, when an agreement was reached.

Fig. 1
figure 1

Timeline of the policy process leading to the institutionalisation of the binding EE1 principle

Full size image

The commission proposal

As part of the ‘Fit for 55’ package, the EC proposed in July 2021, following the EU ordinary legislative procedure (OLP) (Roederer-Rynning, 2019), a recast of the EED, which among other things included strengthened provisions on the ‘EE1’ principle. According to the EC (2021a), the proposal for a recast EED is an important step towards climate neutrality by 2050, where energy efficiency will be treated as an energy source in itself, corresponding to energy efficiency as ‘the first fuel’ (IEA, 2013, 2019). The central role of energy efficiency is supported by the EE1 principle. As for the role of the EE1 principle in energy policy, planning and management, recital 2 of the EC (2021a) proposal states that:

The energy efficiency first principle is an overarching principle that should be taken into account across all sectors, going beyond the energy system, at all levels, including in the financial sector. Energy efficiency solutions should be considered as the first option in policy, planning and investment decisions, when setting new rules for the supply side and other policy areas. /…/ Energy efficiency improvements need to be made whenever they are more cost-effective than equivalent supply-side solutions. That should help exploit the multiple benefits of energy efficiency for the Union, in particular for citizens and businesses. Implementing energy efficiency improvement measures should also be a priority in alleviating energy poverty.

The EE1 principle has been recognised as a key element in the strategy for the integration of the energy sector (EC, 2020). The proposal for a recast EED regulates for the first time the EE1 principle in a separate article with requirements for implementing measures in the MSs, such as ensuring that energy efficiency solutions are considered in energy systems and non-energy sectors, planning, policy and major investment decisions.

Article 3 of the proposal introduces a new provision on the principle of EE1, which will provide the legal basis for the application of the principle. It includes an obligation to take energy efficiency solutions into account in policy, planning and major investment decisions in energy systems and non-energy sectors, including subsidised housing. The EC did not clarify what was meant by major investments. The provisions shall ensure that the principle is applied where relevant and properly monitored in all areas. The EED does not specify how this is to be done in view of the broad scope of the principle.

In order to facilitate the implementation of the provisions of the EED, the EC, (2021b, 2021c) issued a recommendation to MSs on 28 September 2021, including guidance on how the EE1 principle should be interpreted and applied in different contexts. The EC’s recommendation and guidelines aim to transform the concept from principle to practice, and to encourage the correct implementation of the energy efficiency principle and make it more operational. The recommendation to MSs identifies specific measures to be taken to ensure the correct application of the principle, whilst the accompanying guidelines support the practical implementation of the principle with practical solutions and refer to measures in various sectors, from energy supply and distribution to end-use sectors.

Views of interest groups

The introduction of the EE1 principle in EU energy and climate policy was heavily influenced by the advocacy of the Regulatory Assistance Project (RAP) and the European Climate Foundation (ECF), who provided the EC with reports timely for the Energy Union strategy and the ‘Clean Energy for all Europeans’ legislative package (e.g. Bayer, 2015; ECF, 2016; RAP, 2016; Rosenow et al., 2016).

Efficiency First is the foundation on which a successful Energy Union can be built. Energy is an invaluable resource and we must use it efficiently. Europe has a strong case to promote energy productivity as a key driver of growth and security.

Investments in energy efficiency, whether end-use savings or demand response, can create export potential, lower fuel imports, improve air quality, create jobs, and reduce fuel poverty. They can minimise more expensive investments in generation and the grid. And they can enable Europe to deliver carbon reductions faster and at lower overall cost. (Bayer, 2015, p. 4).

Prior to putting forward its proposal for a recast EED, the EC held a public consultation during the winter 2020/2021.Footnote 6 In all, 344 answers were collected. A vast majority of these welcomed the EE1 principle. Further, 244 IGs added that the EC provide more information to users on energy efficiency and energy consumption of products and infrastructures, considering their life cycle and develop guidelines on implementation in relevant policy, planning and investment decisions. More than 150 IGs required that the EE1 principle is applied to all relevant national energy policies related to the whole energy value chain, on EU and/or national level. In its response to the public consultation on the recast EED, CES, which gathers more than 500 business and civil society associationsFootnote 7, 200 companies, 1500 cooperatives and 2500 cities in favour of energy efficiency, stressed thatFootnote 8

In line with the EE1 principle, the revision of the EED offers an opportunity to strengthen the requirements for energy market actors to improve the consideration of demand resources in the energy market from the planning phase to market operation, including recognising the value of energy efficiency and more efficient conversion, transmission and distribution of energy as well as demand response in grid operations.

CES, with support from EU-ASE, stressed that the EE1 principle should be systematically and consistently applied in EU law, in line with its definition provided by the Governance Regulation. To that end, CES welcomed the EC’s intention to produce guidelines to help its implementation, which the EC did in the autumn 2021. In addition, and similar to the position of the EP, it is important to monitor and enforce its implementation both at EU and national level. CES followed up its answer to the public consultation with feedback to the EC proposal for a recast EED, welcoming the introduction of a new article providing a legal basis for the EE1 principle.Footnote 9 However, it was stressed, the article should be reinforced by deleting the reference to ‘major’ investment decisions, include a stronger obligation to develop cost-benefit methodologies to take into account the wider benefits of energy efficiency and define clear compliance criteria to assess its implementation. This was similar to the EP position.

In January 2022, the EU Energy Efficiency Financial Institutions Group (EEFIG) initiated a working group, comprising, e.g. the European Investment Bank (EIB), the European Bank for Reconstruction and Development (EBRD) and German KfW Development Bank, on applying the EE1 principle in sustainable finance. The working group aims to produce guidance on operationalising EE1 that can be used by financial institutions. EEFIG, (2022) claims that:

Public development banks such as EBRD, EIB and KfW have a clear focus on energy efficiency and have taken steps to operationalise energy efficiency first. There is much for the private sector to learn from their approach. EE1 must fit within wider environmental policies and cannot sit alone.

Not all IGs were in favour of the EE1 principle. Large state-owned energy utilities like ČEZ from the Czech Republic and EDF from France, and the German industry association Verband der Industriellen Energie- und Kraftwirtschaft (VIK) were critical.

In principle, the EE1 principle is misleading as it does not allow for a cost-effective competition with other decarbonisation options. (ČEZ, VIK)

The EE1 principle is not cost-effective and does not lead to prioritising the most important actions for climate. (EDF)

A similar critique was put forward by the Belgian technology industry association Agoria:

EE1 does not mean ‘energy efficiency only’. In some circumstances it is often more cost effective to focus on the end objective of climate neutrality by 2050.

This position was shared by Eurelectric. Yet more negative were the Swedish Forest Industries (SFI) and Jernkontoret (the business association for Swedish iron and steel industries):

Climate mitigation should be prioritised before energy efficiency. In some applications energy efficiency and climate mitigation cannot be combined. (SFI)

EE1 is a rule that leads in the wrong direction. It should be climate measures and emission decrease first. The private sector should be excluded. (Jernkontoret)

In common for the critical voices, climate change mitigation is the prime objective, and they question, like Shove, (2018), that energy efficiency should be given highest priority among carbon mitigation measures. In that sense, they were part of a different discourse than the EC, the EP and most MSs in the Council.

Views of member states and the Council

Negotiations in the Council working group for energy started in September 2021, with the EC presenting its proposal and answering questions from MSs. In the first rounds of discussions and negotiations, led by the Slovenian Presidency, half of the MSs, predominantly small and medium sized from across the EU, declared their concern and the importance that the new provisions do not increase administrative burdens related to reporting and do not decrease cost-efficiency. About the administrative burden, the EC replied that it sought to reduce it by linking rules to existing processes, such as reporting under the Governance Regulation. The EC emphasised that the EE1 principle should be applied throughout the economy. The EC stated that defining ‘non-energy sectors’ could include transport, buildings, information and communication technologies and water as identified in the EC guidelines.

During the French Presidency, in spring 2022, four MSs from North and Central Europe stressed that there should be higher flexibility for MSs to use the EE1 principle. A blocking minorityFootnote 10 consisting of eight large, small and medium-sized MSs from all over EU argued that only public sector policy, planning and investment decisions should be covered. Six MSs from around EU considered that ‘major investments’ should be clarified. Two small MSs from Northern and Central EU proposed that only investments of a certain size, more than euro 50 million each or euro 75 million for transport infrastructure projects, should be included in a cost-benefit analysis. Three MS, including a large MS, suggested that the economic thresholds suggested should be tenfold, making only very large projects be required to apply the EE1 principle. In the final agreement, these thresholds were set at euro 150 million each and euro 250 million for transport infrastructure projects. The EC was hesitant towards limiting the size of projects and limiting the scope to the public sector, since it would reduce the scope and water down the principle. In the end, the suggestion to limit the scope to the public sector did not get enough support to make it into the Council’s general approach.

Views of the European Parliament

Once the EC proposal reached the European Parliament (EP), its role is to form consensus among parliamentarians and present proposals as a plenary resolution for the trilogue negotiations with the Council and the EC. The EP’s Committee on Industry, Research and Energy (ITRE) was responsible for managing the EED, including writing the parliamentary report and managing the interinstitutional negotiations. In late February 2022, the EP rapporteur on the recast EED, Danish Member of the European Parliament (MEP) Niels Fuglsang, representing the Group of the Progressive Alliance of Socialists and Democrats (S&D), presented his draft reportFootnote 11 for voting in ITRE. He stated that the introduction of a legally binding EE1 principle is an important step acknowledging the important contribution that energy efficiency can deliver on its multiple benefits. The rapporteur suggested increasing the scope of the EE1 principle to cover all relevant energy related investment decisions, including in the public and private financial sector. Additionally, the rapporteur also proposed the introduction of a common EU methodology with a minimum set of indicators that take into account the wider benefits of energy efficiency. In mid-September 2022, the EP adopted the proposal for a negotiation mandate of the rapporteur, with minor amendments clarifying the strong role of the EE1 principle. The negotiation mandate was supported by the European People’s Party (EPP), the Socialists and Democrats (S&D), Renew Europe (RE) and the Greens/EFA.

Trilogue negotiations

Informal trilogues have become a standard operating procedure in the EU OLP (Brandsma, 2015; Reh et al., 2013; Roederer-Rynning & Greenwood, 2015, 2021). Informal trilogues provide an alternative to the formal back and forth of readings between EP and Council. In trilogue meetings, the institutions are represented by negotiating delegations tasked with facilitating and finding a legislative compromise within and between institutions. For the EP, this delegation includes, but is not limited to, the rapporteur, the shadow rapporteurs, the committee chair and an EP vice president, whereas the Council is represented by an ambassador of the rotating presidency and experts on working party level (Häge & Naurin, 2013). Trilogues are complemented with interinstitutional technical meetings, with no ambassador and MEPs participating. Most of the negotiations are done in these technical meetings.

Trilogue negotiations between the Council, the EP and the Commission were initiated under the Czech presidency in October 2022 and finalised during the Swedish presidency in March 2023. Negotiations were initially going slowly as a result of deadlocked positions from both sides, i.e. on the scope of projects and investments where application of the EE1 principle is required. Commenting on the EP position, a handful of small, medium and large MSs from Western and Southern EU wanted to keep to the threshold of the Council’s general approach. However, ten MSs from Northern, Central and South-Western EU indicated that they could be flexible to lower the threshold, including more projects, decisions and investments. One Northern MS expressed openness to a lower threshold, if writing is added to avoid duplication of work in relation to new buildings, where the EU energy performance of buildings directive and national legislation already require near-zero energy buildings. The EC emphasised that higher thresholds reduce the scope of the application and therefore the purpose of the EE1 principle but were positive to MSs wanting to find compromises.

Trilogue negotiations did not only focus on the sectors and thresholds for applying the EE1 principle. The institutions did also look into the recitalsFootnote 12. The EE1 principle has a prominent role in the recast EED. It is referred to in 15 of 156 of the recitals of the adopted text, but only 5 of 66 recitals in the EC proposal. In particular, the Council and the EP added several recitals explaining the rationale behind the legal text in articles and providing important input as for how the principle should be applied. As for the aims of the EE1 principle, advocated by the EP, recital 16 of the recast EED (EU, 2023) tells:

The energy efficiency first principle should be applied taking into consideration primarily the system efficiency approach and societal and health perspective, and paying attention to security of supply, energy system integration and the transition to climate neutrality. Consequently, the energy efficiency first principle should help increase the efficiency of individual end-use sectors and of the whole energy system.

Regarding the focus of EE1 in practice, recital 18 of the recast EED (EU, 2023) explains which that the whole value chain of the energy system should be considered when applying EE1 principle:

The energy efficiency first principle implies adopting a holistic approach, which takes into account the overall efficiency of the integrated energy system, security of supply and cost effectiveness and promotes the most efficient solutions for climate neutrality across the whole value chain, from energy production, network transport to final energy consumption, so that efficiencies are achieved in both primary energy consumption and final energy consumption.

After several political trilogue meetings, an agreement was reached between the Council and the EP on 10 March 2023, implying that the EE1 principle is to be applied on policy, planning and investment decisions exceeding euro 100 million each and euro 175 million in the transport sector. The public as well as the private sector are included. The agreement was reached after deliberative negotiations finding a middle ground, with no bargaining on the EE1 principle. The recast EED was formally adopted by the EP on 11 July 2023Footnote 13 and by the Council on 25 July 2023Footnote 14.

Discussion

Beliefs and advocacy coalitions

As described in ‘Views on the EE1 principle’ section, the EC, the EP, MSs in the Council and IGs have different views on the EE1 principle. All endorse it on a general level but presented different views on details and on its application. Lines of dispute relates to (i) the purpose and meaning of energy efficiency policy, (ii) the size of projects covered (all projects or only very large projects) and (iii) which sectors to be covered (the public sector or both the public and private sectors) (Table 2). Policy core beliefs and secondary beliefs are identified along these lines. No deep core beliefs were identified.

Table 2 Beliefs in the political subsystem related to the EE1 principle
Full size table

The EC, as agenda setter, and backed by a majority of IGs as well as the EP, argued in its proposal for a recast EED that the EE1 principle is a necessary decision tool to ensure cost-effective decarbonisation of the economy, including enabling the transition to a future powered by renewable energy, and to reap the multiple benefits of energy efficiency. Across energy systems, it asks the question: Would it be cheaper or more valuable to help customers invest directly in energy-saving actions and demand-side response, rather than paying more for supply-side networks, fuels and infrastructure? The result is a more cost-effective allocation of resources across the energy system, including in the many emerging opportunities for customer engagement. The EE1 principle will provide what has been a missing link in fully implementing two other guiding principles of energy and climate policy: cost-effectiveness and consumer protection.

The EP, with support from CES and EU-ASE, wanted to strengthen the role and expand the scope of the EE1 principle to cover all relevant energy related investment decisions, including in the public and private financial sector. Additionally, the rapporteur also proposed strengthened reporting and the introduction of a common EU methodology with a minimum set of indicators that consider the wider benefits of energy efficiency.

In contrast, eight MSs wanted to reduce the scope to the public sector, and to introduce a very high threshold for when investment decisions should be required to apply the EE1 principle. The Council, in its general approach prior to the trilogue negotiations with the EP and the EC, approved a rather high threshold to define what is a major investment decision, but disapproved limiting the scope to the public sector. That MSs are part of two advocacy coalitions confirm previous research, that there is higher heterogeneity in EU policy communities than in national ones (Bache, 2013; Herweg, 2016). Differences in culture, economy and politics between MSs make it unlikely that their governments agree on value acceptability, tolerable costs, normative acceptance and receptivity in the Council.

Since no MS in the Council questioned the EE1 principle as such, at least outspoken, it seems that the critical perspectives put forward by representatives of energy utilities and heavy industries have found little ground among the EU lawmakers. However, MSs’ concerns about increased administrative burdens, calls for flexibility and the ambition to scale down the scope of the application of EE1 to the public sector, together with the introduction of economic thresholds for major investments, are backed by the negative voices from the private sector. By introducing flexibilities and limiting the scope of the EE1 principle to policy, planning and decision-making and to major investments in the public sector, some MSs gave voice to the critics of EE1 in the Council negotiations. This indicates that a group of MSs in the Council, in fact a blocking minority, wanted a weak EE1 principle.

According to ACF, policy actors look for allies among other actors sharing the same beliefs and coordinate in advocacy coalitions (Nohrstedt et al., 2023). Coalitions are a hallmark of negotiations in the EU (Häge, 2013). By joining forces in coalitions, MSs but also the EC and the EP may pool bargaining power and achieve outcomes that are more favourable than what could have been achieved by each MS or institution individually. Coalitions in the EU Council mainly come in two forms (Tallberg, 2008). First the traditional country groupings, such as the Franco-German alliance, the Benelux, the Nordic-Baltic Six and the Visegrad states (Naurin & Lindahl, 2008). These groupings are characterised by their long-term nature, high level of institutionalisation and anchoring in historical experiences, cultural affinities and geographical proximities (Lundgren et al., 2022). The second type are the issue-specific coalitions that are formed in respect of particular dossiers, such as the ‘Frugal Four’ on EU fiscal matters (Lundgren et al., 2019). These coalitions bring together likeminded states on a specific dossier and are characterised by their issue specificity, lower degree of institutionalisation and higher level of fluidity (Lundgren et al., 2022). As for coalitions among IGs, there are formal ones like CES and EU-ASE with a long-term focus on several pieces of legislation such as EED and the energy performance of buildings directive, and informal ones which band together related to specific issues such as EE1.

In all, there was four advocacy coalitions regarding the EE1 principle (Fig. 2), including the EC, the Council, EU MSs, the EP and different IGs and grouped according to their views on the strength of the EE1 principle. From those advocating that only very large projects in the public sector should be covered, to those advocating that all projects in the public as well as private sector should be included. The coalition including the EC and the majority of IGs, as well as the coalition including the EP and CES and EU-ASE, are dominant in terms of actors, several hundred or thousands in both cases. The EC, CES and EU-ASE were ‘principal’ actors in their respective coalitions, meaning that ‘they regularly and broadly engage with coalition members’, whilst the EP was an ‘auxiliary’ actor, that ‘engaged intermittently and often serve as coordinator (or bridge) across coalitions or different subsystems’ (Nohrstedt et al., 2023, p. 136). MSs were divided and those who wanted to limit the scope to the public sector were part of the same advocacy coalition as the critical IGs. The coalition wanting the weakest role of the EE1 principle was subordinate and a weaker one, which did not get its positions into the Council general agreement. These four advocacy coalitions could be grouped into two larger discourse coalitions based in their policy core beliefs, related to their views on the meaning of energy efficiency policy (cf. von Malmborg, 2023a). Those who wanted a week EE1 principle also saw climate change as the main purpose of energy efficiency policy, whilst those who wanted a stronger EE1 principle justified energy efficiency policy by its multiple benefits. The coalitions formed around the EE1 principle were not organised around the common North-East-South pattern. Rather, they were formed around the specific dossier, lower degree of institutionalisation and higher level of fluidity, e.g. the blocking minority wanting to focus the EE1 principle to the public sector with very high thresholds.

Fig. 2
figure 2

Advocacy coalitions related to the EE1 principle. The beliefs of Member States (MSs) were divided. (1) European Parliament, Coalition for Energy Savings, European Alliance to Save Energy, Regulatory Assistance Project. (2) European Commission, majority of interest groups. (3) Council, other Member States. (4) Blocking minority of Member States, large energy utilities, heavy industries

Full size image

Unlike political parties, advocacy coalitions are rarely formal entities (Weible & Ingold, 2018). Their networks vary from sharing beliefs and discursive arguments, deliberate coordination of political activities to implicit alliances where allies settle into roles and niches that complement each other. For instance, some coalition members might deploy outsider tactics by organising protests and shaping public discourse, whilst others might deploy insider tactics by working with governments to design regulations (Gabehart et al., 2022). As for the coalitions on the EE1 principle, there was much coordination among members. MSs wanting to limit the scope to the public sector were well coordinated and formed a blocking minority. Häge, (2013) argues that MSs band together to form blocking coalitions, with eventual decisions reflecting a compromise between these coalitions (and therefore a consensus between all MSs). The same held true for IGs critical towards the EE1 principle, but also among stakeholders positive to the EE1 principle. CES, ECF, EU-ASE and RAP encompassed several thousands of IGs, companies, municipalities and NGOs. As regards coalition-formation, König and Junge, (2008) suggest we need to examine more closely the relationship between EC proposals and agenda-setting, on the one hand, and at how the EC exploits potentially favourable coalitions in the Council and the EP, on the other hand. During negotiations in the Council, the EC made clear that limiting the size of projects and limiting the scope to the public sector would reduce the scope and water down the principle. The blocking minority for limiting the scope did not use its veto and the suggestion to limit the scope to the public sector did not make it into the Council’s general approach.

Paths to policy change

The ACF theory associates major policy changes with changes in policy core beliefs, i.e. significant shifts in the direction or goals of a subsystem, and minor policy changes with changes in secondary beliefs, e.g. change in means for achieving the goal (Sabatier, 1988, 1998; Sabatier & Jenkins-Smith, 1999). Advocacy coalitions often disagree on proposals related to these components, and policy debates therefore revolve around diverging preferences regarding initiatives of either change or preserve governmental programmes (Jenkins-Smith et al., 2018; Nohrstedt et al., 2023; Weible et al., 2011).

The ACF model sets out four paths to policy change: (i) external shocks, (ii) internal shocks and other internal events in the political subsystem, (iii) policy-oriented learning and (iv) negotiated agreements (Jenkins-Smith et al., 2018; Pierce et al., 2020). It is asserted that at least one of these, or some combination thereof, is a necessary but not sufficient source of change in the policy core attributes of a governmental programme. Another proposition of ACF regarding policy change is that the policy core attributes of a governmental programme in a specific jurisdiction will not be significantly revised if the advocacy coalition that instated the programme remains in power within that jurisdiction—except when the change is imposed by a hierarchically superior jurisdiction. In the previous section, we saw that the main conflict was between secondary beliefs, focusing on the role of the EE1 principle in EU energy and climate policy. Policy change in the case of the EE1 principle followed a combination of three paths, not two as proposed: (i) an external shock, (ii) a negotiated agreement and (iii) policy-oriented learning. That policy change follows a combination of processes confirms previous ACF research (Nohrstedt et al., 2023).

An external shock

As for paths to policy change regarding the EE1 principle, from a mere aspiration in the Governance Regulation to a legally binding principle in the recast EED with mandatory requirements for MSs to apply it in planning, policymaking and investment decisions in all sectors of the economy, it can first of all be explained as a consequence of an external shock, i.e. as a consequence of decisions in another policy subsystem. The EC proposal for a recast EED, with a proposal on the EE1 principle, was part of a package to make EU legislation ‘Fit for 55’, following the ‘European Green Deal’ and the adoption of an EU climate law with strengthened climate targets for 2030 and 2050. It is also important that it was a new Commission, under von der Leyen, who proposed the change. The introduction of the EE1 principle in the Governance Regulation was a proposal of the previous Commission (under Juncker).

A negotiated agreement

The different policy options put forward by the EC, the Council and the EP created dualities between related but incompatible frames. Different strategies, e.g. elimination, assimilating, polarisation, reframing or connection, can be utilised to deal with the duality (Dewulf & Bouwen, 2012). In the case of EU decision-making on new or changed legislation, splitting the difference through some form of compromise is another strategy to deal with differences, variously referred to as mutual adaptation of the different parts to each other (Thomas Jr., 1995), splitting the difference or bargaining as jointly seeking means to split the differences, set trade-offs or take turns (Kindler, 1988). Both poles are thus somewhat acknowledged but never fully credited.

The new legislation on EE1 is a negotiated agreement—a result of deliberative negotiations in the Council and the EP, and trilogue negotiations between the EP, the Council and the EC. Negotiated agreements mean that the policy change does not correspond to the policy goals of the dominant or minority coalition, but a negotiated middle ground. Negotiated agreements may emerge in a variety of ways but are facilitated by collaborative institutions conducive to negotiation (Nohrstedt et al., 2023). This is the case of EU decision-making, with co-decision of the Council and the EP seeking consensus on new, amended or recast directives and regulations (Brandsma, 2015; Häge & Naurin, 2013; Naurin & Wallace, 2008; Reh et al., 2013).

As for the Council general approach on the EE1 principle (see ‘Views of member states and the Council’ section), a middle ground was found among the different coalitions. No MS opposed the Council general approach on these points, despite there being a blocking minority to focus only on the public sector. As argued by Heisenberg (2008), the Council prefers to negotiate agreements rather than to proceed to voting, and that the negotiations frequently are successful in the sense that agreements are concluded. The Council general approach in this case was met by the MS ambassadors in the Council of Permanent Representatives, some days ahead of the Council of energy ministers’ meeting on 27 June 2022. As for negotiations in the EP, a compromise text was agreed upon by S&D, EPP, Renew Europe and the Greens/EFA. The text was approved by the ITRE committee by voting which was won by the dominant coalition. The EP plenary also approved the compromise by voting.

As for the trilogue negotiations, the starting points of the EP and the Council were rather far from each other. The EP wanted the EE1 principle to be applied to all policies, planning and investment decisions, in all sectors, public as well as private, whilst the Council wanted to limit the scope to very large investment decisions, exceeding euro 150 million each and euro 250 million for transport infrastructure projects. In addition, the EP called for extensive reporting, whilst the Council called for limited reporting. In the negotiations, which were of a deliberate nature, the Council and the EP found a middle ground focusing on investments exceeding euro 100 million each and euro 175 million for transport infrastructure projects.

Contrary to the findings of Reh et al., (2013) and Brandsma, (2015), that an increasing number of co-decisions by the EP and the Council are met informally and secluded by fast-track agreements before the EP and the Council adopted their negotiation mandates, the agreement on the recast EED and the EE1 principle was met after the EP and the Council adopted their negotiating mandates before trilogues commenced, the EP in plenary and the Council at a minister’s meeting. However, it was an early agreement and following an informal and secluded process. Under co-decision, informalisation is in line with—and in application of—the Lisbon Treaty. Informal decision-making thus plays out within the EU’s formal legislative process, from which it is distinct along four dimensions: a restricted, noncodified set of decision-makers operates in a secluded setting, social interaction is structured by informal rather than codified and enforceable rules, and informal compromise must be legitimised through the formal process of rubber stamping (Brandsma, 2015; Reh et al., 2013; Roederer-Rynning & Greenwood, 2015, 2021). In the case of the EE1 principle, there were high expectations for early agreements on all legislative acts in the ‘Fit for 55’ package, with the ‘European Green Deal’ (EC, 2019) being one of six priorities of the von der Leyen Commission. The issue of the EE1 principle was highly politicised, both in the EP and the Council, with lines of dispute related to the meaning of the principle, which sectors and which policies and investment decisions should be covered (see ‘Beliefs and advocacy coalitions’ section). Positions of IGs were varied, with strong advocates both for and against a strong EE1 principle. This confirms the assertion of Reh et al., (2013) that legislative issues that raise public interest and political opposition will not be subject to first reading agreements, because EP and Council constituencies will demand broad and public legislative debate. They expect such demands to be particularly pronounced where issues are salient or redistributive. Where issues are salient, interest and opposition derive from a dossier’s relation to sovereignty or from its strong symbolic relevance.

The mode of negotiation to be found in the EU decision-making processes is contextually determined (Elgström & Jönsson, 2011). Empirically, it is demonstrated that day-to-day negotiations in the EU are to a large extent problem-solving exercises. Under certain circumstances, however, conflictual bargaining occurs. The pattern varies with level of politicisation and type of policy, and according to the stage in the decision-making process. In addition, processes of learning have resulted in changes in the EU’s negotiation style: problem-solving has become increasingly institutionalised within the EU machinery (Elgström & Jönsson, 2011). Negotiations on the EE1 principle in the Council, the EP and between the Council and the EP were of a deliberate, problem-solving nature. Under co-decision—introduced amid institutional competition between EP and the Council—negotiation in trilogues can facilitate cooperation between the two co-legislators, by creating ‘mutual confidence, and positive trust spirals’ and by augmenting ‘skills at political compromise’ (March & Olsen, 1998, p. 960).

Consensual decision-making, thus, is a key characteristic of the Council, although the sources of consensus are debated among researchers (Naurin & Lindahl, 2008; Naurin & Wallace, 2008). But what type of consensus is created in the Justus Lipsius, Lex and the Europa Buildings? A strong trend in democratic theory for some years has been to emphasise the value and importance of deliberation as a way of producing consensus and creating legitimacy for collective decisions (Naurin & Lindahl, 2008; Naurin & Wallace, 2008). Democratic politics, according to this view, should be more about giving good reasons than forcing or striking deals (cf. Cooke, 2000; Habermas, 1984). Deliberation means trying to reach agreement through the force of the better argument—convincing others of the right thing to do—rather than bargaining via threats and promises. This was the case for negotiations on the recast EED. Bargaining, according to normative deliberative democratic theory, is a perfectly legitimate way of reaching agreements in the marketplace. But we should not buy and sell public policy. In politics—at the forum—arguing is the morally superior way of interaction (Naurin & Wallace, 2008).

The Council general approach as well as the EP position and the co-decision were met through deliberation. But it is hard to generalise with respect to the status of the Council as a deliberative body. Deliberation certainly happens, as in this case, but only under specific circumstances. In particular, the level of politicisation is important (Naurin & Wallace, 2008). The issues at stake regarding the EE1 principle was not the main issues at stake, rather the EU energy efficiency target and national contributions and national energy savings obligations, and the issues of the EE1 principle, was, despite its important role put forward by the Commission, not highest on the public agenda. Although deliberation by its normative proponents is designed to be a method for conflict resolution, there is a clear limit to how much conflict this mode of interaction can manage (Naurin & Lindahl, 2008; Naurin & Wallace, 2008).

Policy-oriented learning

ACF addresses policy-oriented learning, defined as ‘relatively enduring alterations of thought or behavioural intentions which result from experience and which are concerned with the attainment or revision of the percepts of one’s belief system’ (Sabatier, 1988, p. 149). It is found to be one of the most common path of policy-change in studies applying the ACF as a theoretical framework (Jenkins-Smith et al., 2018; Pierce et al., 2017). Learning is associated with changes in belief systems of coalition members that include not only the understanding of a problem and associated solutions, but also the use of political strategies for achieving objectives (Jenkins-Smith et al., 2018). As discussed by Jenkins-Smith et al., (2014), von Malmborg, (2021) and von Malmborg and Strachan (2023), policy-oriented learning can happen across belief systems, which could lead to policy change.

Such policy-oriented learning across belief systems appeared in the Council, where deliberative negotiations on the EE1 principle was key, trying to reach agreement through the force of the better argument—convincing others of the right thing to do—rather than bargaining via threats and promises (Naurin & Lindahl, 2008; Naurin & Wallace, 2008). The rotating Council Presidency had an important role, chairing negotiations, drafting compromise texts, forcing debates, facilitating reflexion, and building coalitions.

The most critical MSs in the Council learned from others and accepted the secondary beliefs of the middle-ground positions. The EC was hesitant towards limiting the size of projects and limiting the scope to the public sector, since ‘it would reduce the scope and water down the principle’. Thus, the blocking majority arguing for limiting the EE1 principle to the public sector did not use its veto power to block an agreement but adopted the same position as the rest of the MSs and the Council reached a general agreement unanimously. There was no voting in the Council and no MSs openly opposed the secondary beliefs expressed in the Council’s general agreement once it was decided. This was despite there being a blocking minority calling for limiting the EE1 principle to the public sector earlier in the negotiations. The situation was similar in the EP ITRE committee, where proposals of the supporting parties were taken onboard without conflict. As for the trilogues, a middle ground was found between the Council and the EP, lowering the threshold of policies, projects and investment decisions to be covered by the EE1 principle compared to the Council general agreement. The EP abandoned its position that all policies, projects and investment decisions should be covered. This policy-oriented learning across belief systems was facilitated by favourable conditions (cf. Jenkins-Smith et al., 2014; Nohrstedt et al., 2023), such as the mature and semi-collaborative policy subsystem, few actors holding extreme beliefs, the moderate level of conflict, a policy venue with focus on consensus and deliberation and an honest policy broker (Table 3).

Table 3 Factors facilitating policy-oriented learning in policymaking on the EE1 principle
Full size table

Conclusions and policy implications

The aim of this paper was to the explore and, as far as possible, explain the political process in the EU policymaking on the EE1 principle, from proposal to political agreement and adoption of the co-legislators. Why did the EC propose that the principle be made binding on the MSs, and how did IGs and the co-legislators—the EP and Council—react to the EC proposal? Which were the political conflicts? What processes lead to the political agreement on the EE1 principle as a legal institute in the EU? The study applied the ACF as a theoretical framework. It was proposed that (i) there exist at least three advocacy coalitions, around the EP, the EC and the Council respectively; (ii) the EP advocates stricter and more far-reaching policy core beliefs and secondary beliefs in favour of the EE1st principle than do the EC and the Council, whilst the Council advocates a weaker role of the EE1 principle; and (iii) policy change followed an external shock and a negotiated agreement.

Ingold, (2022) identifies three methodological challenges when applying ACF to studies of EU policymaking on a supranational level. The policy subsystems analysed is the EU system of energy efficiency policy, which has developed over 50 years and can be considered mature, with a high degree of collaboration between different actors. The policy venues analysed are the public consultation preceding the EC proposal, negotiations in the Council and negotiations in the trilogues. As for actors and coalitions, I included EU institutions, MS governments, NGOs and other IGs. Including MS governments is an unorthodox approach in ACF, which has mainly been applied to analyse policy change at national level but is increasingly used in studies of EU policy. Finally, as for data accessibility and availability, I got access to unique, confidential data from negotiations in the Council and trilogues. As proposed by Henry et al., (2022), being transparent about the methodological choices is key, which I have been.

It is concluded that there were initially four different advocacy coalitions, sharing different policy core and secondary beliefs, i.e. the actors’ views on the purpose of the EE1 principle and their policy preferences with respect to concrete policy options (e.g. specific policy design, policy instruments and others) related to the EE1 principle. One coalition, including eight critical MSs from across EU and IGs from electricity utilities and heavy industries, wanted a weak EE1 principle with the scope limited to the public sector and very large investments. The second coalition, including 19 MSs and the Council, wanted to limit the scope to large investment decisions and wanted limited reporting for MSs. These two shared policy core beliefs that the purpose of energy efficiency is to mitigate climate change. A third coalition, including the EC and the majority of IGs, wanted a strong EE1 principle with focus on decision making in the public as well as private sector. The fourth and final coalition included the EP and the largest IGs in favour of strong energy efficiency policy, such as CES, ECF and RAP. This coalition favoured all decision-making in the public as well as private sector, and strong reporting, calling for the strongest application of the EE1 principle. The advocacy coalitions including the EC and the EP shared policy core belief that energy efficiency policy is justified with its multiple benefits. The coalitions formed on the EE1 principle were not organised around the common North-East-South pattern. Rather, they were formed around the specific dossier, lower degree of institutionalisation and higher level of fluidity, e.g. the blocking minority wanting to focus the EE1 principle to the public sector with very high thresholds. After negotiations in the Council, the four coalitions were reduced to two, one around the Council and the EP, one around the EP.

As for policy change, it is concluded that the adoption of the EE1 principle as a binding principle follows a combination of processes. First, an external shock to the energy efficiency political subsystem, namely the adoption of an EU climate law following the Paris Agreement, which led to the ‘Fit for 55’ legislative package, including, among other things, a proposal to revise the EED. Moreover, it is a negotiated agreement between the Council and the Parliament, which was carried out as a predominantly problem-solving exercise rather than bargaining between the Council and the EP. Given the deliberative nature of the negotiations, it opened for interdiscursive communication (cf. von Malmborg, 2023a) and policy-oriented learning across belief systems in the policy system.

As for policy implications, the paper provides a unique insight into the processes of policymaking and policy change as well as the politics related to the EE1 principle and can inform scholars of energy and climate policy and different stakeholders of energy efficiency policy on the workings of the EU institutions, MSs and IGs. It provides valuable knowledge for policymakers and stakeholders outside EU on how the concept of energy efficiency as the ‘first fuel’ can be contextualised into a binding principle, to be applied by actors in different sectors. It can also help various stakeholders to better shape their advocacy strategies in future EU policymaking. The identification of beliefs and advocacy coalitions may help policymakers and other stakeholders become more aware of their own and others’ beliefs and how these could be changed. As important is the differentiation of policy core beliefs and secondary beliefs: which beliefs can be easily changed, which cannot.

Based on this research, a number of topics for further research are proposed. First, how is the EE1 principle implemented in MSs? What political conflicts are there? In addition, it would be relevant to analyse and compare the policy processes and the politics of other files of the ‘Fit for 55’ package to gain better understanding of EU policy for the clean energy transition. As for ACF related issues, this paper showed that the ACF is a potent theory for the study of EU policymaking. But treatment of MSs in the ACF and the operationalisation of their beliefs represent a conceptual challenge of the ACF on international/supranational levels, which has not yet been adequately addressed in the literature. Finally, more studies can be done to analyse the factors shaping policy-oriented learning and its relation to policy change.