Comparing Strategies and Success of Policy Entrepreneurs in EU Energy and Climate Policy Processes

1. Introduction

Given the current climate crisis with rapidly increasing temperatures and more frequent and intense weather extremes leading to extensive damage and losses in nature and societies (IPCC, 2023), climate change mitigation through a clean energy transition (Bressand & Ekins, 2021) is one of the most important public policy areas today. The clean energy transition towards climate neutrality is a cross-boundary problem that cuts through society and governance levels. This second-generation energy transition requires transformation of social organisation involved at an order of magnitude greater than it was for the first-generation transformation focusing on renewable electricity (Bressand & Ekins, 2021).

More pervasive and cross-boundary public policy issues such as climate change and the related clean energy transition, but also sustainable development, fair trading and civil rights have made public executives and legislators targets for broad business and non-business advocacy to ensure one’s often conflicting interests are protected or taken into consideration in public policies (Harris, 2001; Bitonti & Harris, 2017). Thus, liberal democracies have seen an increase of interest groups (IGs) and other stakeholders trying to influence public policy processes and policy outcomes as well as governance arrangements (Baumgartner & Leech, 1998; Klüver, 2013; Bitonti, 2017a, 2017b).

Political science offers a variety of theories, frameworks and concepts to explain policy processes and the roles of stakeholders influencing agenda-setting and decision-making on public policy. One such concept is the policy entrepreneur (PE) defined as “advocates who are willing to invest their resources – time, energy, reputation, money – to promote a position in return for anticipated future gain in the form of material, purposive, or solitary benefits” (Kingdon, 1984, p. 179). They are central actors in political agenda-setting and for policy change as they ‘soften’ the political system for certain ideas and make sure there are packages of problems and policies ready when there is an opportunity to put the problem on the agenda (Mintrom, 2019a). The concept is attractive since it highlights the role of agency in understanding policy change (Mintrom & Norman, 2009; Mintrom, 2019a). In view of the ‘Multiple Streams Framework’ (Herweg et al., 2023), policy change occurs when streams of problem, policy options and politics are connected by PEs who, as individuals or collectives, attempt “to transform policy ideas into policy innovations and, hence, disrupt status quo policy arrangements” (Petridou & Mintrom, 2021, p. 945). A policy option must be promoted by PEs with resources, strategies, and access to the politics stream when there is an open policy window of opportunity where the streams can meet and form the policy community in the policy stream. PEs are not only involved in agenda-setting- on public policy, but also in policy formulation (Gunn, 2017).

PEs from different sectors are and will continue helping formulation of new, innovative climate policies at local, national, regional and global governance levels. In fact, leading scholars on PEs stress the need for more actors to become PEs and “step forward and catalyze change processes” (Mintrom, 2019b, p. 307) in relation climate governance (Mintrom & Leutjens, 2017). Regarding climate governance, Boasson and Huitema (2017) suggests that more research is needed (i) whether PEs that seek change are more active and/or more successful as PEs compared to those that defend the status quo, and (ii) whether there is more successful entrepreneurship in public or in private arenas of governance. They also call for the need to study failure of PEs. Such analysis is crucial for establishing that PEs influence outcomes (Arnold et al., 2023). In addition, Green (2017) and Mintrom and Luitjens (2017) argue for more comparative research on PEs in climate governance. Such research should distinguish PEs from the strategies they use (Green, 2017). Strategies are the causal mechanisms that link the independent variable (PEs) to the dependent variable (some form of policy change). Few scholars have responded to these calls, and the research agenda is still relevant. New theoretical and empirical contributions along these lines could do much to advance our understanding of agency and structure in contemporary politics in climate policy and governance and governance in general (Petridou & Mintrom, 2021).,

Given these research gaps, this paper aims to further the understanding of strategies used by PEs from different sectors and social spheres in energy and climate policy processes and which strategies that are successful or unsuccessful. To meet the aim, I undertake a comparative analysis of motifs, strategies, success and failure of PEs over time, related to new and revised EU policy and legislation to stimulate a clean energy transition for meeting the new EU target on climate neutrality by 2050. I compare four cases focusing on EU supranational policies for: (i) enhanced energy efficiency of buildings, (ii) enhanced energy efficiency in companies, (iii) strengthening the role of energy efficiency in climate policy with the ‘energy efficiency first’ (EE1) principle, and (iv) decarbonizing maritime shipping. The different policy changes were introduced with the adoptions and subsequent changes to the Energy Efficiency Directive (EED) 2012–2023 and the new FuelEU Maritime Regulation (FEUM) 2013–2023. Besides contributing to the literature on PEs, the paper provides valuable knowledge on key actors in policy processes on the clean energy transition, furthering our understanding of the politics of the clean energy and just transition more broadly (cf. Kuzemko et al., 2019; Bressand & Ekins, 2021; Pearse, 2021).

The four cases saw PEs from different spheres of society, i.e. the public sector, the private (non-profit) sector, hybrid sector (both public and private), and civil society, making it possible to compare PE motifs, strategies, success and failure of different kinds of PEs. This adds civil society PEs as a category to the public and private sector PEs stressed by Boasson and Huitema (2017) as well as Brouwer and Huitema (2018). Civil society actors have rendered an increasingly important role in climate governance since the UN climate summit in Copenhagen 2009 (Pandey, 2015; Nasiritousi et al., 2016; Reimer & Saerbeck, 2017; Carter & Childs, 2018). As for private sector PEs, I differentiate profit from non-profit organizations, the latter category including e.g. privately funded think tanks. In addition, the four cases cover different sectors of EU climate policy, with PEs acting in different policy subsystems of different degrees of maturity, from nascent (shipping) to mature (energy efficiency, buildings), to different degrees of collaboration/conflict. Three of the cases focus on policy processes over a decade where changes were made to the legislation. In the fourth case, policy was developed over a decade until legislation was finally adopted. This makes it possible to analyse PEs’ strategies dynamically over time, which is important since the impacts of PEs can operationalize expansively, spanning different phases of the policy process (Green, 2017). The following research questions are addressed:

R1	What motivated PEs and which strategies were used by different PEs in the four cases? Which were the similarities and differences?
R2	Which PE strategies were (un)successful in influencing the policy process in the four cases?
R2	Did the PE strategies and related success or failure change over time?
R4	Did the strategies and related success or failure differ between public sector, private sector, hybrid sector and civil society PEs?
R5	Were PEs seeking change more active and successful than PEs advocating status quo?

The paper is outlined as follows. Section 2 presents the theoretical framework for analysing strategies and success of PEs, while the methodology is described in section 3. The macro-context on policy entrepreneurship in EU climate governance is described in section 4. The four cases are briefly presented in section 5 and analysed in section 6. Since the European Commission (EC) acted as a PE in all cases, and was the only PE to deploy structural entrepreneurship, its role will be analysed and discussed in detail in section 6.3. Conclusions are drawn in section 7.

2. Policy Entrepreneurs in the Policy Process: A Theoretical Framework

Analysing the success or failure of PEs, one must distinguish PEs from other policy actors, identify their motifs and strategies, and the influence of context. Figure 1 captures the conventional model of the PE’s strategic behaviour and impact in the policy process. These factors are described and reflected upon in the remainder of this section.

Figure 1.

2.1. Who Can Be a Policy Entrepreneur?

The PE concept has been accused of fuzziness, meaning different things to different scholars even within the same discipline (Cairney, 2013; Petridou, 2017; Arnold et al., 2023). In early studies, like Kingdon’s (1984), only individuals were considered PEs, but studies of policy entrepreneurship in Europe have added organizations as agents. The broad scope of PEs not only include individuals such as elected politicians, public officials, academics and experts, but also companies, business associations, non-governmental organizations (NGOs), think tanks, other IGs, political parties and public institutions, e.g. the EC, the Council of the EU (Council), the European Central Bank, the European Investment Bank, or national, regional or local governments and authorities (Herweg, 2016; Petridou, 2017; Mintrom, 2019a; Liebe & Howarth, 2020; Pircher, 2020; Heldt & Müller, 2022; Herweg & Zohlnhöfer, 2022; von Malmborg et al., 2023a). Zito (2000, 2017), even refers to “collective entrepreneurship” in which advocacy coalitions acted entrepreneurially to formulate individual policies.

Second, we must distinguish policy entrepreneurship from other actions. Boasson and Huitema (2017, p. 1351) argue that “privileged actors in powerful positions deploy[ing] the regular tools at their disposal and merely do their job, they are not demonstrating entrepreneurship”. This could be interpreted as privileged persons or organizations in powerful positions, like cabinet ministers, EU commissioners, EU institutions or national governments, not being able to act as PEs. But according to Kingdon (1984), entrepreneurship can be deployed by actors in and out of government – elected officials, lobbyists or civil servants – at different levels, and in different domains. They do so if they are “persistent and skilled actors who launch original ideas, create new alliances, work efficiently or otherwise seek to ‘punch above their weight’” (Boasson & Huitema, 2017, p. 1344; Green, 2017). For instance, the current EC president Ursula von der Leyen and former EC commissioner Frans Timmermans has been described as PEs for launching the European Green Deal in 2019 (Kreienkamp et al., 2022; Becker, 2024), and the EU High Representative for Foreign Affairs and Security Policy, Federica Mogherini, acted as a PE and pushed for policy change in EU’s foreign and security policy (Sus, 2021). Schneider and Teske (1992) found that PEs often come from inside the political system, even from the government cabinet (Saetren, 2016). Similarly, Rabe (2004) found that PEs on US climate change policy came from within the federal administration. In addition, partisan effects of Labour policy entrepreneurship have been identified for UK law and order policy (Staff, 2018) and UK education policy (Olsson Rost & Collinson, 2022). In all, acting as a PE depends on a “set of behaviours in the policy process, rather than a permanent characteristic of a particular individual or role” (Ackrill & Kay, 2011, p. 78).

To further distinguish who can be a PE, they are similar to ‘experts’ (Christensen et al., 2023) and ‘lobbyists’ (Bitonti & Harris, 2017) in the policy process, but differ from traditional lobbyists in that they are more actively involved in problem framing and in the technical and legal aspects of the policy solutions to a problem, in an area in which they have expertise. Lobbyists are often paid by a client. In addition, PEs are usually working as disruptive change agents with new policy ideas, catalyzing innovative policy change (Roberts & King, 1991; Capano & Galanti, 2021), while lobbyists often but not always advocate status quo (Mintrom, 1997; Baumgartner & Leech, 1998). However, some PEs are active in blocking policy change and maintaining status quo (Boasson & Huitema, 2017; Arnold, 2021), although some PE scholars dispute that advocating status quo should be related to policy entrepreneurism, that it is rather more consistent with classic IG politics (Green, 2017). Moreover, PEs can come from within the government, whilst lobbyists come from outside the government. Compared to lobbyists, PEs are considered ”heroic” or “crazies” with a “desire to change the world” and ”push humanity forward” (Mintrom, 2019b, p. 320). In that sense, PEs can be considered ‘thought leaders’, to borrow a concept from business strategy literature (Young, 2013; Harvey et al., 2021).

2.2. Motifs and Strategies of Policy Entrepreneurs

Drawing on Kingdon’s definition of PEs, most scholars have assumed that they are instrumentally rational (Bakir & Jarvis, 2017), motivated by a “desire for power, prestige and popularity, the desire to influence policy, and other factors in addition to any money income derived from their political activities” (Schneider et al., 1995, p. 11) or “satisfaction from participation, or even personal aggrandizement” (Kingdon, 2002, p. 123). However, policy actors are rather boundedly rational and motivated by cognitive rationality, i.e. their beliefs and ideas (Zahariadis, 2007; Bakir, 2009). PEs may engage in policy advocacy to prevent opponents with conflicting beliefs from securing ‘evil’ policies, triggering a ‘devil’s shift’ (Sabatier & Jenkins-Smith, 1999). Arnold (2022, p. 26) argues that “[o]ppositional factors, by triggering a value-laden, devil shift-influenced fear of a threat to a desired policy goal, can catalyze policy entrepreneurship”.

Previous research tells that PEs work in an energetic, sometimes activistic, strategic manner with the intention to innovatively change political alignments (Roberts & King, 1991; Capano & Galanti, 2021). Strategies used by PEs are the lines of action taken to reach their aims, the latter which fall in two categories (Boasson & Huitema, 2017):

• Structural entrepreneurship: acts aimed at overcoming the structural barriers to enhance governance influence by altering the distribution of formal authority and factual and scientific information; and
• Cultural-institutional entrepreneurship: acts aimed at altering or diffusing people’s perceptions, beliefs, norms and cognitive frameworks, worldviews, or institutional logics.

Analysing the scholarly literature, Aviram et al. (2020) identified 20 strategies and three traits of PEs: trust building, persuasion, and social acuity. Theses may vary with respect to target audience, level of government at which the PEs operate, sector, and PEs’ professional roles, timing, number and types of actors involved, relationship to development of international politics etc. Brouwer (2015) and Brouwer and Huitema (2018) proposed four categories of strategies (Table 1), which I will apply in my analysis:

Table 1. xxx.

Table 1. xxx.

For structural entrepreneurship, three strategies are particularly important: (i) creating and working in networks and advocacy coalitions, (ii) strategic use of decision-making processes, and (iii) strategic use of information (Boasson & Huitema, 2017). Through networking, a PE learns the worldviews “of various members of the policymaking community” (Mintrom, 1997, p. 739), which enables the PE to persuade policy actors with high levels of legitimacy or authority to join in (Wihlborg, 2018). As for strategic and smart use of decision-making procedures and venues, this relates to timing and thus launching the policy idea when there is an open policy window (Kingdon, 1984). Finally, PEs can reach their aims by assembling new evidence and make novel arguments (Dewulf & Bouwen, 2012), carrying ideas (Swinkels, 2020) that serve as ‘coalition magnets’ (Béland & Cox, 2016) to convince an appropriately powerful coalition of supporters to back the proposed changes (Boasson & Huitema, 2017; Mintrom & Luetjens, 2017), either by manipulating who gets what information, if information is distributed asymmetrically and information is scarce (Moravcsik, 1999; Zahariadis, 2003), or by strategic manoeuvring, such as providing as little information as possible to one’s likely opponents (Mackenzie, 2010).

As for cultural-institutional entrepreneurship, making people positive to the ideas coming from the entrepreneur, and negative to existing and competing policy or governance arrangements, framing of problems and policy options is the most important strategy (Boasson & Huitema, 2017). As Copeland and James (2014, p. 3) put it, framing is about “strategic construction of narratives that mobilize political action around a perceived policy problem in order to legitimize a particular solution”. How a condition is framed as a problem influences how we think about the problem (Knaggård, 2015), which enables coupling to certain policies, but not to others (Weiss, 1989). To persuade others, a PE must consider the perspectives of various actors and create meanings and frames that appeal to them (Fligstein, 2001). PEs deploy ‘outsider tactics’ by shaping public discourse on problems and policy solutions, or ‘insider tactics’ by working with policymakers to design regulations (Gabehart et al., 2022; Tosun et al., 2023).

Becker et al. (2024) and Petridou et al. (2024) recently suggested that PEs can be considered proactive, performing ‘entrepreneurship by opportunity’, or reactive, performing ‘entrepreneurship by necessity’ as a response to vertical pressure. A sense of urgency and necessity, professional norms and technical skills make reactive PEs utilise actions not underpinned by the proactive quest for opportunities. Compared to proactive PEs, reactive PEs “seek to affect change under conditions of urgency while they do not exhibit intentionality and a priori adherence to a specific policy solution” but look for and advocate “the design and/or implementation of a public policy solution /…/ as the best, but not always the preferred, choice (Petridou et al., 2024, p. 84). Such reactive entrepreneurship is deployed by PEs in management and governance of crises and emergencies, e.g. the COVID-19 pandemic and floodings (Taylor et al., 2023; Becker et al., 2024).

Importantly, Braun et al. (2024) found that PEs employ different strategies over time when interacting with their immediate and wider contexts in attempting to foster policy change – they co-create with policymakers to shape their ecosystems and society at large. This echoes the critique of mainstream research on PEs, claiming that “in contrast to widely held view that PEs promote ideas and operate outside decision-making processes (Mintrom, 2000), PEs are endogenous to policymaking processes” (Bakir & Jarvis, 2017, p. 469).

A related critique is that mainstream research on PEs focus on the meso-level context of policy change, i.e. the immediate context of a particular policy, in policy formulation but dismisses the macro and micro contexts within which a PE is embedded (Bakir & Jarvis, 2017). They demonstrate that context impacts policy entrepreneurship and institutional entrepreneurship, at least in the public sector. Agency of PEs is ”most likely to generate policy and institutional changes when they are reinforced by complementarities arising from context-dependent, dynamic interactions among interdependent structures, institutions and agency-level enabling conditions” (Bakir & Jarvis, 2017, p. 465). Acknowledging Giddens’s (1979) concept of ‘duality’ between agency and structure, PEs and context “should not be viewed in isolation but as linked through strategy” (Zahariadis & Exadaktylos, 2016, p. 62). The macro-level context consists of formal and informal institutions, i.e. a relatively stable collection of rules and practices, embedded in structures of resources and meaning that enable and constrain PE agency (cf. March & Olsen, 2008). The micro-level context includes the social position of the PE, who, be they individuals or organizations, can occupy multiple social positions with different identities in different parts of the policy process, e.g. as decision-maker, academic, framer and broker. This enables PEs to work in different ideational realms, programmes, discourses, building coalitions and generating consensus (Bakir, 2009). In addition, addressing the ‘ahistorical’ nature of mainstream PE research, disregarding that “previous policies affect current debates and institutional innovation, or institutional contexts that shape how, why, and where policy reforms take place” (Bakir, 2009, p. 573), and that the outcomes of PE agency today will affect future debates, there needs to be a feedback loop from policy outcome and other expansive effects (cf. Green, 2017) to context.

2.3. Success of Policy Entrepreneurs

Previous studies indicate that the question whether PE advocacy systematically succeed in promoting policy change is unresolved (Arnold, 2021). Not all PEs are successful (Boasson & Huitema, 2017; Arnold, 2021). Success is defined in different ways. For some, a PE is successful if the advocacy leads to changes in other policy actors’ policy preferences (Teske & Schneider, 1994). This can be compared to policy-oriented learning (Sabatier, 1988). For others, a PE is successful if influencing agenda-setting in such a way that the PE’s pet issue is considered by policymakers (Mintrom, 1997; Mintrom & Vergari, 1998). Yet another view is that a PE is successful if she has actual influence on policy and governance (Boasson & Huitema, 2017), e.g. adoption of specific policy measures the PE sought (Arnold et al, 2017; Crow, 2010; Mintrom, 1997, 2000). What is deemed success depends on the aim of the PE agency.

Mintrom and Norman (2009) assume that success is more likely for a PE who has more characteristics that define her as such, or who deploys entrepreneurial strategies with greater frequency or intensity (cf, Binderkrantz & Krøjer, 2012). Most PEs strive for policy innovation, which consists of initiation, diffusion, and the evaluation of effects that such innovations create, the latter requiring analytical capacities (Jordan & Huitema, 2014). These challenges are central to the work of PEs. It is their willingness to use their positions for leverage and for aligning problems and solutions that increase the likelihood of policy change (Mintrom & Norman, 2009). The ability of PEs to successfully promote policy innovation also depends on their skill at identifying relevant competencies, developing and effectively deploying them (Considine et al., 2009; Meijerink & Huitema, 2010). In addition, a successful PE must understand the concerns of actors they seek to persuade, use social acuity to build teams, networks, and coalitions, be knowledgeable to strategically disseminate information, and be organising, corresponding to political activation and involving civic engagement (Mintrom & Norman, 2009; Aviram et al., 2020; Arnold, 2021). Anderson et al. (2020) adds that the influence of PEs lies not only in their ability to define problems and build coalitions, but also in their ability to provide new and reliable information to elected officials.

Categorizing characteristics, goals and strategies of successful PEs related to anti-fracking policies in New York State, Arnold (2021) suggests three PE archetypes. Activists are highly active and display many PE characteristics and deploy a wide range of strategies to reach several moderate policy goals. Advocates have similar characteristics as activists, but are less active, using fewer strategies and focus primarily on one ambitious policy goal. Concerned citizens manifest rather few PE characteristics and tend to lobby for restrictive/ambitious measures with relatively little facility or effort. Activists and advocates were found to be successful, while concerned citizens had no or negative impact. An interesting implication of her study is that “it may not be necessary for an actor interested in securing a policy goal to go ‘all in’ as an activist.” “Policy change can potentially be achieved through more modest, accessible advocacy” (Arnold, 2021, p. 985).

3. Method and Material

Regarding choice of method to study PEs, Ohemeng and Grant (2023) raise a fundamental question. Research in the last decades has largely focused on providing generalizable knowledge on PE strategies, using quantitative methods. However, this research overlooks why and how PEs utilize certain strategies. To understand in a meaningful way the nuances of PE behaviour, qualitative research methods must be used (Mintrom, 2020). To pursue the analysis for this paper, a comparative multiple case studies approach based on phenomenology has been used. This methodology is suitable because the research questions are qualitative in nature. I adopted a cross-case comparison (Yin, 1994; Eisenhardt & Graebner, 2007) to find and explain similarities and differences. The phenomenological approach includes obtaining rich data, delineating meaning units, transforming meaning into sensitive statements and synthesizing a general structure (Giorgi, 1997; Ohemeng & Grant, 2023).

3.1. Case Selection

Four cases were selected for their theoretical relevance, i.e. their ability to generate as many properties as possible of the phenomenon under study (Glaser & Strauss, 1967):

• Provisions in EED on individual metering and billing (IMB) of heating, cooling and domestic hot water to enhance energy efficiency in multi-family and multipurpose buildings. The provisions were first adopted in 2012 (EU, 2012a) and amended in 2018 (EU, 2018),
• Provisions in EED to enhance energy efficiency in companies across all sectors. The provisions were first adopted in EED in 2012 (EU, 2012a) and amended in 2023 (EU, 2023a). A large majority of MSs and IGs unsuccessfully advocated for amendment already in 2018,
• Provisions in EED to strengthen the role of energy efficiency in EU climate policy by introducing the energy efficiency first (EE1) principle. The provisions were gradually introduced in EU policy documents from 2015 and onwards until EE1 was made legally binding with the 2023 recast of EED (EU, 2023a), and
• Provisions in the FuelEU Maritime regulation (FEUM) to enhance uptake of low and zero carbon (LoZeC) fuels and decarbonise international maritime shipping. The policy was gradually introduced in EU policy documents from 2013 until provisions were adopted in 2023 (EU, 2023b).

The first case deals with innovative policy on energy efficiency of buildings complementing the previous focus on technical measures such as renovation with a focus on changing human behaviour of occupants (von Malmborg, 2021, 2022). But the initial provisions were unconditional and hampered renovation, why the amended provisions put technical and behavioural measures in context and aligned two different EU directives. The second case deals with a complete change of focus on which companies should implement energy audits and energy management systems, from focusing on company size to focusing on annual energy use (von Malmborg & Strachan, 2023; von Malmborg, 2024a). The third case provides a paradigm shift in EU energy and climate policy, requiring that energy efficiency should be treated on par with power production and investments in power grids (von Malmborg, 2023a, 2023b). The innovative idea and impact of the EE1 principle is further manifested through the launch in 2024 of the journal Energy Efficiency First by scientific publisher Elsevier. The final case is about introducing, for the first time ever, EU legislation to decarbonize maritime shipping, implying that EU overtakes the global leadership from the International Maritime Organization (IMO) (von Malmborg, 2023c, 2024b).

The different cases saw PEs from different sectors (social spheres, i.e. public, private, civic), and PEs proactively advocating policy change as well as reactively advocating status quo. Some PEs acted as ‘activists’, while others acted as ‘advocates’. In addition, the four cases cover different sectors of EU climate policy, with PEs working in different policy subsystems of different degrees of maturity, from nascent (shipping) to mature (energy efficiency, buildings, industry), to different degrees of collaboration (energy efficiency, buildings, industry) or conflict (shipping).

3.2. Identifying Policy Entrepreneurs, Their Motifs and Strategies

Arnold et al. (2023) have recently criticized PE scholars of paying too little attention to the methodological question of how to empirically identify PEs. Generally, scholars identify PEs by event-based or process-based approaches “asking elites and experts to point them out, querying secondary sources, surveying possible entrepreneurs, and focusing on high-profile advocates” or “using secondary sources” (p. 659), using a “I know it when I see it” standard to distinguish PEs from other policy actors (p. 660). Using elite or expert input to identify PEs may overrepresent the prevalence of PEs while overlooking less connected or less expert policy actors trying to influence policy. It is a problem that case studies of successful policy entrepreneurship, typically involving an event-based PE identification approach, dominate the literature. Fewer scholars examine failure cases to assess whether policy entrepreneurship was absent or using bad strategies, yet such analysis is crucial for establishing that PEs influence outcomes. According to Arnold et al. (2023, p. 662), the literature needs more careful analyses of cases with varied outcomes to understand whether PE presence or absence correlates with these outcomes in the theoretically anticipated way. Even fewer studies examine cases where PEs were present but unsuccessful and try to explain why they failed. If a jurisdiction fails to adopt a path-breaking policy, the failure may not be solely attributable to lack of policy entrepreneurship. PEs might lack facilitative characteristics or be using outdated or bad strategies to try and advance policy goals.

I view PEs as individual or organizational change agents strategically and entrepreneurially involved in framing and developing innovative policies, presenting them well-packaged to policymakers or political executives. To avoid the pitfalls mentioned, I applied a stepwise process suggested by Arnold et al. (2023) for identifying and distinguishing PEs once potential PEs are identified:

• One marker of significant resource investment (time or money),
• AND: at least one entrepreneurial goal,
• AND: at least one entrepreneurial strategy or characteristics,
• AND: at least one network partner.

To screen potential PEs in the cases, I analysed the responses to the Commission’s public consultations, reports from negotiations in the Council, and reports in pan-European online newspaper Euractive, writing about developments in EU policy. The screening was also facilitated by previous studies of advocacy coalitions in the cases (von Malmborg, 2021, 2023b, 2023c; von Malmborg & Strachan, 2023). The screening identified 50–400 policy actors in each case but singled out 4–40 policy actors in each case being more active and vocal, providing more reports and policy briefs, and co-leading the work in different policy networks and advocacy coalitions, why I did not survey all potential PEs with questions about resources, goals, strategies and network partners. Most policy actors were involved in the case of decarbonizing shipping, fewest in the case of energy efficiency of buildings. After such analysis, seven PEs from the public, private and civil society sectors, meeting the stepwise criteria, were identified in the four cases, the EC through different DGs being a PE in each case (table 2). In all cases, key individuals with positions such as deputy director, head of unit, policy officer or associate, were particularly important for driving policy

3.3. Data

As for data collection on motifs, strategies, success and failure, I combined semi-structured interviews and qualitative text analysis (Table 3). Ten interviews were held with key individuals in the organizations acting as PEs. In addition, three interviews were held with the Swedish energy and transport attachés in the cases of energy efficiency in companies, EE1 and FEUM, since they acted as chairs of the Council energy and transport working groups during final trilogue negotiations between the Council, EC and EP in spring 2023, leading to political agreements and adoptions of the two pieces of EU legislation. Interviewees were asked and told about their motifs and entrepreneurial strategies. They were also asked about their views on the opponents’ strategies in advocacy. Interviews were not recorded but notes were taken. As for texts, I have analysed policy documents from the EC as PE in all cases, policy papers from the Government Offices of Sweden, Ministry of the Environment and Energy (GOS/MEE) as PE on energy efficiency of buildings and companies, and reports and policy papers from the PEs on EE1 and FEUM, looking for perceptions of problems and positions on policy proposals.

In this material, I identified perceptions of problems and positions on climate change, as well as on policies and policy proposals, entrepreneurial strategies and responses to these. Data was coded in relation to problem framing, policy proposal, beliefs and motifs for policy change, entrepreneurial approaches (structural or cultural-institutional (cf. Boasson & Huitema, 2017)), entrepreneurial strategies (see table 1 for a categorisation of different strategies), and kinds of impacts/output.

4. Contexts of Policy Entrepreneurship in EU Climate Governance

Context at macro-, meso- and micro-levels frame who choose to act as a PE, their aims, motifs and strategies. Climate governance in EU is contextually influenced by macro-level factors such as multilevel governance and the formal and informal procedures of policymaking. Each of the cases analysed and compared are further influenced by the meso-level context of EU climate policy. Micro-level contexts differ between the cases and are analysed in section 5.

4.1. Multilevel Governance

The EU is a complex polity (Hix, 2007). It is a semi-democracy – a demoi-cracy (Müller, 2010).¹ Policymaking in the EU truly takes place in a multilevel governance (MLG) setting. The MLG concept was originally used to analyse and describe the governance structure of EU structural policy, emphasising the parallel processes of decentralisation and European integration (Marks, 1993; Hooghe & Marks, 2001). As for MLG in the EU, more than five supranational institutions, 27 national, 1 200 regional, and almost 87 000 local governments (OECD, 2018) are linked in territorially overarching policy networks. To these, a fifth tier can be added, the international, as the EU is also a member of international agreements. These layers interact with each other in two ways: vertically across different levels of government and horizontally with other relevant actors within the same level.

Of the EU institutions, the EC² promotes the general interest of the EU by proposing and enforcing legislation as well as by implementing policies and the EU budget. The EC is the EU’s politically independent executive arm. It is alone responsible for drawing up proposals for new EU legislation and implements the decisions of the Council of the EU and the European Parliament. The European Council and the Council of the EU consist of representatives of national governments in the member states of the EU.³ The European Council is the policy making venue for overall policy guidelines, where negotiations are made by heads of governments and states. The Council of the EU is a single legal entity, but it meets in ten different configurations, depending on the subject being discussed. The Council is a venue for policymaking on specific legislation. Negotiations are held between ministers from the specific policy areas, but most negotiations are done in Council working groups where member states are represented by civil servants from member state governments. The European Parliament is a directly elected EU body with legislative, supervisory, and budgetary responsibilities.⁴ It also performs democratic scrutiny of all EU institutions. It consists of 705 Members of the European Parliament (MEPs), belonging to seven party groups.⁵ MEPs are elected every five years.

Governance and policymaking in the EU can be described as a mix of supranational negotiations and intergovernmental cooperation between sovereign member states in the Council and the European Parliament, codified through a succession of major treaties (Tosun & Graziano, 2022). Treaties are the primary legal basis, complemented by legislative instruments such as directives, regulations, decisions, recommendations and opinions, constituting the means through which the principles of EU integration stipulated in the treaties are applied, disputed and enforced.

According to the treaties, MLG in the EU shall respect competences, share responsibilities and cooperate between the various governance levels. Reference is made to the ‘subsidiarity principle’, enshrined in Article 5(3) of the Treaty of European Union (TEU) (EU, 2012b), aiming to guarantee a degree of independence for a lower authority in relation to a higher body or for a local authority in relation to central government. It therefore places decisions as close as possible to the citizens and ensures that that action at Union level is justified considering the possibilities available at national, regional or local level.

However, there is a constant debate on subsidiarity and the need for collective action on the EU level furthering the European integration. In the FF55 cases, as in EU policy in general, MSs often contest EU policy, based on either sovereignty (subsidiarity claims), or substance (Wettestad et al., 2012; Herranz-Surralés, 2019; Herranz-Surralés & Solorio, 2022). MSs usually want room for manoeuvring, flexibility, related to national circumstances. Since the entry into force of the Lisbon Treaty in 2009, energy, climate and transport policy are shared competence between MSs and the EU, with energy policy being based on Article 194 of the Treaty on the Functioning of the European Union (TFEU) (EU, 2012b). EU climate policy is based on Articles 191–193 of TFEU. EU transport policy is based on Articles 90–100 of TFEU. The question of subsidiarity is particularly challenging in EU policy on energy efficiency of buildings, in focus of the first case in this study, since there is shared competence between EU and MSs on energy policy, but full competence for MSs and no competence for EU on buildings policy (Björklund et al., 2023).

4.2. The Ordinary Legislative Procedure

Decision-making on EU legislation, aiming at European integration, i.e. ‘harmonising’ legislation in the EU, follows what is called the ‘ordinary legislative procedure’ (OLP) (Roedinger-Rynning, 2019). In the system, the directly elected EP must approve EU legislation together with the Council (i.e. the governments of the 27 EU member states) in an act of co-decision. The procedure starts with a legislative proposal from the EC, who holds the right of initiative to put forward legislative proposals in the EU. A call for new or amended legislation can also come from the Council (through Council conclusions) and the Parliament (through resolutions), requesting the EC to put forward legislation. Based on input from a public consultation, lobbyists and PEs and the findings within a regulatory impact assessment (RIA), EC formulates a proposal and addresses it to the two co-legislators simultaneously, which can adopt, reject or amend the proposition. The proposal is also sent to the European Economic and Social Committee and the Committee of the Regions, who can respond to the proposal but do not have a formal role in the decision-making process.

The Council and the EP respectively negotiate and decide on a negotiation mandate for tripartite inter-institutional negotiations (so-called trilogues) between the Council, the EP and the EC which aim at reaching a political agreement between the co-legislators. Trilogues have become a standard operating procedure in EU decision-making (Brandsma, 2015; Roederer-Rynning, 2019). Under co-decision – introduced amid institutional competition between EP and the Council – negotiation in trilogues can facilitate cooperation between the two co-legislators (Roederer-Rynning & Greenwood, 2021), by creating “mutual confidence, and positive trust spirals” and by augmenting “skills at political compromise” (March & Olsen, 1998, p. 960).

Decisions on EU legislation can be taken by unanimity or majority voting, where the former “requires that all parties give their consent, or at least do not actively block an agreement”, while the latter “entails that only a large subset of all parties needs to be on board for an agreement to be reached” (Tallberg, 2008, p. 694). Qualified majority requires that >65 % of the votes support the decisions. A decision can be blocked by a minority coalition gathering >35 % of the votes. Decisions on the negotiation mandate for trilogues are adopted by qualified majority voting in the Council and absolute majority in the EP. The agreement between the Council and the EP following trilogue negotiations requires consensus (Brandsma, 2015; Puetter, 2016; Roederer-Rynning & Greenwood, 2021). Once the text is politically agreed upon, the two co-legislators formally adopt legislation jointly, having equal rights and obligations.

Throughout the MLG policy process, there are frequent and complex interactions between governmental actors and non-state actors such as business associations, companies and civil society organisations and other IGs, trying to influence the EC, the EP, the Council and governments of the MSs. Governments of EU MSs are part of the process within the EU and their respective country, being influenced by lobbyists and PEs but themselves also influencing the EC, the EP and other MSs. In addition, they must adapt their national governance strategies to the decisions made on the European level.

The three institutions are not monoliths. Mentioned, MSs may group in different coalitions within the Council during negotiations on a general agreement (Huhe et al., 2018). Likewise, different party groups in the EP form coalitions during negotiations on the negotiation mandate for trilogues. In the EC, different DGs may dispute on a legislative proposal prior to its adoption before and during the so-called inter-service consultation (ISG). PEs does not address the institutions as such, but selected MEPs and their assistants, policy officers and chiefs in the EC as well as commissioners, and cabinet ministers and civil servants in MS governments.

4.3. Meso-Level Context of Policy Entrepreneurs in EU Climate Governance

The EU began developing climate policy in the 1990s and has since built a broad portfolio of mitigation policy measures and governance tools, including legally binding targets to reduce GHG emissions and energy use, and increase the share of renewable energy sources, and policy measures addressing emissions trading, renewable energy, energy efficiency, land-use and land-use change, transport, taxation and more (Dupont et al., 2023).

Through its privilege on initiating and monopoly to present legislative proposals, the EC is considered a central and ‘natural’ PE in EU politics and governance in many different policy areas, e.g. energy, climate, water, information and communications technology, research, defence, security, social rights and finance (Maltby, 2013; Edler & James, 2015; Becker et al., 2016; Herweg, 2016; Bocquillon, 2018; Simpson, 2020; Vesan et al., 2021; Zeilinger, 2021; Herweg & Zohlnhöfer, 2022; Kreienkamp et al., 2022; Bürgin, 2023; Silander, 2023; Valin & Huitema, 2023; Becker, 2024; Sus, 2024). In the perspective of MSF, the EC frames the problem and links it to policy options, which are then coupled to the politics stream as a package. The EC also participates in Council negotiations and in trilogues. But the EC does not act as a PE every time it proposes new legislation, only when it proposes innovative policies and legislation (cf. Roberts & King, 1991).

Since the early 1990s, the EU, particularly the EC, aspires to exercise global leadership in climate governance and other environmental policy areas (Tobin et al., 2023). To be able to show ‘exemplary leadership’ internationally, the EC has taken several steps to develop transformative climate policy (Dupont et al., 2023). But the EC is not unaffected in its role as a PE. As mentioned, the European Council, the Council and the EP, can request the EC to put forward legislation. The EC usually responds to such calls by proposing legislation. It was the EP that in the early 1990s took the initiative to develop EU’s climate policy and made the EC propose a carbon dioxide tax, which, however, was voted down by the member states (MSs) in the Council (Wettestad et al., 2012). As outlined below, it was the European Council that initially proposed the 20-20-20 targets. PEs and traditional lobbyists are also trying to influence the EC, if PEs a long time before the EC announces a legislation, if lobbyists when the EC has announced that a new policy/legislation is underway.

However, the EC is not monolithic. Although climate action has a dedicated DG (DG CLIMA), there are significant overlaps with the work of other DGs covering policy areas such as energy and transport (DG ENER, DG MOVE), important for EU climate policy according to EGD and responsible for the legislation analysed in this paper. EC policy proposals are often criticized internally in inter-institutional rivalry (Simpson, 2020), due to competing interests, beliefs and policy priorities, can hamper climate policy ambition (Rietig, 2019). It is often a specific directorate general (DG) and its commissioner, or a unit within and its head and specific (group of) policy officers that act as PEs. Mid-ranking officials in the EC often play a crucial role in identifying and utilising a political window of opportunity (Edler and James, 2015).

In March 2007, the European Council adopted the so-called 20-20-20 EU headline targets on climate mitigation and energy (Council, 2007):

• To reduce GHG emissions by 20 % by 2020 taking 1990 emissions as the reference,
• To increase energy efficiency to save 20 % of EU energy use by 2020,
• To reach 20 % of renewable energy in the total energy use in the EU by 2020, and
• To reach 10 % of biofuels in the total consumption of vehicles by 2020.

In 2008, then president of the EC, José Manuel Barroso, presented a climate and energy policy package (EC, 2008), including these targets and legislation to meet these targets, by (i) limiting carbon dioxide (CO₂) emissions from vehicles, (ii) developing capture and storage of CO₂, (iii) inviting each MS to reduce their GHG emissions, (iv) reforming the EU emission trading system (ETS), and (v) introducing a renewable energy directive. Shortly after, the EC presented a proposal to introduce an energy efficiency directive (EED) (EC, 2011), adopted in 2012 (EU, 2012a). In 2015, the president of the EC, Jean-Claude Juncker, presented the Energy Union Strategy (EC, 2015), followed by the ‘Clean Energy for all Europeans’ legislative package (EC, 2016b), amending or recasting several EU directives, including EED, and introducing the regulation for governance of the Energy Union and climate action (EU, 2018).

Shortly after, the new EC President Ursula von der Leyen, acting as a key authoritative PE in the politics stream, capable of connecting agenda and decision-making (Kreienkamp et al., 2022), presented the European Green Deal (EGD) (EC, 2019) in response to the climate and environmental challenges facing the world and manifested in the 2015 Paris Agreement. EGD was also a response to bottom-up pressure for more ambitious climate policies, with unprecedented levels of public engagement on climate change, the success of the Fridays for Future strikes, and the EC targeted by NGOs such as the European Climate Foundation (Kreienkamp et al., 2022). Von der Leyen coined EGD as “Europe’s man on the moon moment” (Hutchinson, 2019), and it is one of six priorities for the 2019–2024 period of the von der Leyen EC. It is a roadmap to make the EU’s economy sustainable by turning climate and environmental challenges into opportunities across all policy areas and making the transition fair for all. It is also “a new growth strategy aimed at transforming the EU into a fair and prosperous society, with a modern, resource-efficient and competitive economy where there will be no net GHG emissions by 2050 and where economic growth is decoupled from resource use” (EC, 2019, p. 1). In spring 2021, the EU legislators approved, as part of implementing EGD, a new European Climate Law (ECL) (EU, 2021), which states that the EU GHG emissions will be reduced by 55 percent by 2030 compared to 1990 levels, and that the EU, by 2050, will be the world’s first climate-neutral continent. Reaching the targets require contributions from all sectors of the economy. To meet these targets, the EC, acting as a PE, presented in July 2021 a complementary and interconnected set of proposals as part of the so-called Fit for 55 (FF55) legislative package (EC, 2021g), to make the EU climate, energy, land use, transport and taxation legislation fit for reducing net GHG emissions by at least 55 % by 2030.

The legislation and policies included in the three policy packages all aimed at transforming EU climate policy and governance to increase its legitimacy as a leader in global climate governance while also furthering European integration.

5. Strategies of Policy Entrepreneurship in EU Climate Governance

5.1. Enhanced Energy Efficiency of Buildings

5.1.1. The Problem and Policy Issue

The EC (2021a, 2021b) estimates that 40 % of final energy in the EU is used in buildings and the sector accounts for 36 % of total GHG emissions in the EU. Improving energy and climate performance of buildings is important for decarbonizing our societies by 2050. Energy efficiency of buildings can be improved by different measures, spanning from technical-structural measures to organizational and behavioural measures (von Malmborg et al., 2023b).Energy efficiency of buildings has been an issue on the EU policy agenda for more than 50 years (Economidou et al., 2020; von Malmborg et al., 2023c), and EU policies have expanded in scope and size over time as the EU has become more ambitious. But legislation such as the EU Energy Performance of Buildings Directive (EPBD), first adopted in 2002, has mainly focused on technical-structural measures.

In 2011, the EC Directorate-General for Energy (DG ENER) proposed complementary legislation on individual metering and billing (IMB) in the new EED to stimulate behavioural changes, e.g. lowering the in-door temperature and closing windows. EC (2011) proposed that individual meters should be installed unconditionally in each apartment of multi-family and multipurpose buildings and that billing should be based on actual consumption (EC, 2011). The reasons given by the EC for provisions on IMB stressed the need of final customers (and final users) to be better informed of their own energy use to reduce it.

However, the governments of Sweden and Finland and IGs in the two MSs, saw the EC proposal as a threat, negatively impacting housing policy which is national competence from a subsidiarity perspective as well as counteracting more effective energy efficiency renovations, and acted as PEs to maintain status quo to avoid the ‘evil’ of IMB. IMB could save energy by 0–20 %, while deep renovations could save 60–100 %. In Finland and Sweden, so-called gross rent is the dominating pricing model in multi-family housing, where costs of heating, cooling and domestic hot water are included in the rent. In continental Europe, net rent is the dominating price model, where costs of heating, cooling and domestic hot water are paid separately, on top of the rent. The governments of Finland and particularly Sweden, the latter represented by a deputy director of GOS/MEE, argued that that gross rent and no IMB would provide incentives for energy efficiency renovations and improve the energy efficiency of buildings in the EU, while IMB requires a shift to net rent that reduces incentives for renovations. Taking a systems perspective on policy to enhance energy efficiency of buildings, they saw a conflict between EED and EPBD. Worth noting, the deputy director of GOS/MEE holds a PhD in environmental systems analysis and was trained in systems thinking. Other MSs in the Council did not understand the issue at stake and did not support the governments of Finland and Sweden in the Council negotiations. This was because IMB is mainly an issue where buildings are heated with district heating or cooled with district cooling, which is the case in Sweden and Finland but not the case in most MSs. In the eastern and central European MSs, where district heating is common, tenants do not have the opportunity to influence their own indoor temperature with thermostatic valves, why IMB is a non-issue.

In late 2015, the EC announced a revision of the EED. At an early stage, contributing to agenda-setting, GOS/MEE proposed the introduction of proportionality as an important condition for the installation of IMBs in existing buildings – that it shall be installed in existing buildings “where it is technically feasible and cost-effective to be proportionate to the potential energy savings”. After negotiations, were the Council and the EP turned out to have identical positions (Council, 2017; EP, 2017), provisions of the original EED on unconditional requirements for IMB of heating and cooling in new buildings or buildings that underwent major renovations were deleted. The same applied to the unconditional requirement for IMB of domestic hot water in buildings that are undergoing major renovations. In the revised EED, there was only one unconditional requirement for IMB for domestic hot water in new buildings, which was a fallback position for GOS/MEE. In addition, provisions on IMB in new buildings were clarified, referring to metering at unit level in multi-apartment buildings or multi-purpose buildings, and mentioned final users in addition to final customers.

5.1.2. Motifs and Strategies of Policy Entrepreneurs

When the IMB provisions were initially negotiated in 2011–2012, the EC acted proactively for policy change, while the governments of Sweden and Finland reactively advocated status quo. When EED was amended in 2016–2018, EC wanted to maintain status quo, while GOS/MEE, with support from the Finnish government, proactively advocated policy change. The different strategies used by EC and GOS/MEE as PEs in the two phases are summarised in table 4.

In both phases, EC and GOS/MEE used problem framing and strategic use of information to gain attention and seek support for their respective positions. There was no difference if the PE advocated policy change or status quo. However, it is evident that PEs advocating status quo focused only on these strategies, trying to persuade others by the good argument. Using the archetypes of Arnold (2021), they can be considered ‘concerned citizens’, representing housing owners and tenants. Their motif was to avoid new policies to counteract the effectiveness of existing policy, thus a ‘threat-centred’ policy entrepreneurship to avoid the EC secure its ‘evil’ policy – a ‘devils shift’. To their defeat, the governments of Finland and Sweden did not understand the value of interacting with other policy actors in 2011–2012. As mentioned by the deputy director at GOS/MEE, he thought that good arguments underpinned by statistics would the job. Nor did national IGs of building owners in Finland and Sweden opposing IMB understand the value of interacting with MSs and IGs at the EU level. Social acuity was low. Opponents of IMB in Sweden had for long focused on influencing the Swedish government in the phase of implementing the EU provisions into national legislation rather than influencing actors at EU level during negotiations of the EED (von Malmborg, 2021).

Aiming for policy change in different phases, both EC and GOS/MEE used a broader palette of strategies. EC acted as and ‘advocate’, while GOS/MEE acted more ‘activistically’ (cf. Arnold, 2021). Compared to 2011–2012, the strategy of GOS/MEE, led by the deputy director of the MEE in charge of the EED dossier, acting as a PE to change the IMB provisions in 2016–2018, changed considerably. Motivated by negative experiences from negotiating the original EED and heavy critique from Swedish IGs when EED should be transposed and implemented, claiming the IMB provisions being illegitimate, the deputy director of GOS/MEE and Sweden’s energy counsellor in Brussels decided it was necessary to take a new approach to advocacy through proactive policy entrepreneurship (von Malmborg, 2022). In addition, EC had opened an infringement case on Sweden for noncompliant transposition of IMB provisions, and changing the provisions could take away the legal ground for the infringement case. Among other things, GOS/MEE acted early to frame the problem, build coalitions of likeminded and to network with policymakers in all EU institutions. Three months after the announcement of the revision and nine months prior to the EC proposal was presented to the Council and the EP in November 2016, GOS/MEE presented a position paper with proposals for revised provisions on IMB. The purpose was to influence the EC proposal but also to influence other MSs. Three months before the EC proposal, GOS/MEE invited and teamed up with national IGs critical towards IMB, including the tenants’ union, and the national associations of public and private housing owners, energy companies and local authorities. Once the EC proposal was presented, GOS/MEE continued to present position papers on IMB and met with rapporteurs on EED in the EP as well as influential people among MSs in the Council, including the Presidency, and this had an impact. Even more, so did the teaming up with EU-level IGs like Housing Europe (HE) and European Property Federation (EPF), representing public and private building owners, and International Union of Tenants (IUT, representing the energy end-users and thus consumers), and three Swedish MEPs in the EP committees for industry, research and energy (ITRE) as well as environment, food safety and public health (ENVI), the latter being rapporteur on the EED file for ENVI. An important effort and activity of GOS/MEE and the coalition opposing IMB, besides presenting co-ordinated position papers, was the organizing of a seminar in the EP. In April 2017, GOS/MEE together with HE, EPF and IUT organised a lunch seminar in the EP on improving energy efficiency in buildings (von Malmborg, 2022). The seminar was hosted by two Swedish MEPs representing Renew Europe and the Greens. Since GOS/MEE and other actors in the advocacy coalition opposing IMB were negative to the division of IMB and renovation of buildings, the seminar focused, by issue linking, on how EED and EPBD could promote a coherent and holistic approach to reducing energy use in buildings in a cost-effective way. What are the respective roles of building owners and residents to improve the energy performance of buildings? What is the potential of IMB and behavioural changes compared to technical-structural energy efficiency renovations? In what ways can consumers be empowered? By putting IMB in relation to energy efficiency renovations of buildings a better narrative was told than if the message would only have been that IMB is bad. At the seminar, the chair of HE referred to experiences of IMB in Swedish member companies, showing that energy use increases, not decreases, because of IMB. In addition, she claimed that IMB would dramatically change housing and tenancy policy, which in turn have negative impacts on energy efficiency and thus counteracts the aim of the EED, and that IMB could increase energy poverty and thus influence social policy (von Platten et al., 2020).

The extensive work of GOS/MEE as a PE and its allies opposing IMB impressed other MSs, which supported the proposals from the governments of Sweden and Finland to add proportionality as a precondition for IMB in the Council’s general agreement and mandate for trilogue negotiations. At the same time, the governments of Denmark and Germany were less vocal as proponents of IMB and they did not oppose the position of the Swedish and Finnish governments put forward to the Council. Sweden’s position regarding new buildings and buildings undergoing major renovations was further developed and demands for IMB in buildings undergoing major renovations were deleted in the Council negotiations, as were requirements for IMB for heating and cooling in new buildings. The entrepreneurial strategies of GOS/MEE also influenced the EP. The positions of the Council and the EP were identical and similar to the position put forward by GOS/MEE and its allies why the legislation was amended accordingly.

5.2. Enhanced Energy Efficiency in Companies

5.2.1. The Problem and Policy Issue

Business in industry, transport, energy and the tertiary sector account for 75 % of the EU’s final energy use (Eurostat, 2023), which makes companies an important sector to develop policies for promoting energy efficiency. Research on corporate energy management in different sectors suggests that ‘energy audits’ (EAs) and ‘energy management systems’ (EnMSs) are necessary but not sufficient for improving energy efficiency in companies (e.g., Thollander & Palm, 2012; Lee & Cheng, 2016; Schulze et al., 2016; Kluczek & Olzsewski, 2017; McKane et al., 2017). They could help overcome lack of and asymmetric information about the most important measures to take to improve the energy efficiency of a process, system or organisation. According to EED, an EA is ‘a systematic procedure with the purpose of obtaining adequate knowledge of the energy consumption profile of a building or group of buildings, an industrial or commercial operation or installation or a private or public service, identifying and quantifying opportunities for cost-effective energy savings, identifying the potential for cost-effective use or production of renewable energy and reporting the findings’ (EU, 2023a, p. 30). EED defines an EnMS as ‘a set of interrelated or interacting elements of a plan which sets an energy efficiency objective and a strategy to achieve that objective, including monitoring of actual energy consumption, actions taken to increase energy efficiency and measurement of progress’ (EU, 2023a, p. 29). Like environmental management systems (EMSs) and quality management systems, EnMSs are based on the Deming plan-do-study-act cycle for continual improvement (Cleary, 1995), and can be certified according to standards of the International Organization for Standardization (ISO), ISO 50001.

EAs have been regulated in the EU law since the entry into force of the EU Energy Services Directive (ESD) in 2006 (EU, 2006). According to the ESD, MSs should ensure the availability of efficient, high-quality EA schemes which are designed to identify potential energy efficiency improvement measures, and which are carried out in an independent manner, to all final consumers, including smaller domestic, commercial, and small and medium-sized industrial customers. After a proposal by the EC (DG ENER), provisions on EAs were made mandatory for large companies with the adoption in 2012 of the original EED (EU, 2012a), which also repealed the ESD. EED provided that MSs shall ensure that enterprises that are not SMEs are subject to an EA carried out in an independent and cost-effective manner by qualified and/or accredited experts or implemented and supervised by independent authorities under national legislation. The original EED also introduced voluntary provisions on EnMS.

When transposing the provisions of the original EED, MSs encountered challenges as for how to relate to the concept of ‘large companies’, which are defined as ‘not being small or medium-sized entreprises (SMEs)’, and thus for identifying the target group of national legislation. The challenge for MSs arose when the headcount, economic turnover and balance sheet analyses were made to identify if a company is large or not. This is since the analysis should consider ownership and thus linked and related enterprises. For a company to be an SME, it must be autonomous. This means that a large company can consist of several small or medium-sized but linked or related entities, located wherever in the world. This made it difficult for MS law enforcement agencies to identify which companies should undertake mandatory EAs. Availability of data on entities in other countries was limited unless law enforcement agencies in different MSs and jurisdictions collaborated. Consequently, MSs addressed the situation and treated companies, particularly multi-national companies, in different ways, hampering the level playing field for companies on the EU internal market. This also led to frustration among MS civil servants and companies.

As part of FF55, EED provisions on energy efficiency in companies were changed. Provisions were changed from focusing on large companies to focusing on companies with high energy use. Companies will be subject to an EA if their annual energy use exceeds 10 TJ. The recast EED also includes requirements for companies with the highest energy use to implement certified EnMSs. All enterprises, including SMEs that exceed 85 TJ of annual energy use, will have to implement a certified EnMS. The new provisions are deemed one of the three most important policies and measures of the EED in terms of contributions to energy savings and increased energy efficiency in the EU (EC, 2021b).

5.2.2. Motifs and Strategies of Policy Entrepreneurs

The problems with the 2012 EED provisions on EAs were identified by MSs already some years after the entry into force of the directive. There was a clear dissatisfaction among MSs with the performance of the EA provisions of the EED. The issue was raised by MSs in an EU-wide concerted action project on implementing EED, which sparked several studies by consultants, academics and the EC (Serrenho et al., 2015; EC, 2016c; Eichhammer & Rohde, 2016; Hirzel et al., 2016; Nabitz & Hirzel, 2019). These studies confirmed the critique by MSs and catered for consensus among MSs and the EC on the problem description and the need for policy change.

A policy window to re-examine and change the provisions was opened in 2016–2018, when the EED was undergoing a limited amended. Motivated by negative experiences of the original provisions, impossible to implement in a way that safeguarded a level playing field for EU companies, and with frustration among MS civil servants and companies, a group of 13 MSs⁶ led by GOS/MEE 7 proactively presented a joint position paper in early 2017 with a proposal for how to change the legislation. This was supported by all other MSs but one. The Maltese Presidency of the Council, who led the negotiations in the Council in the initial deliberations, as well as the EC as an “entrepreneurial gatekeeper” selecting, rejecting or reshaping the ideas that float around in the “policy primeval soup” (Kingdon, 1984, p. 140), were hesitant to include changes to the provisions on EAs in the amended EED since the article was not included in the EC proposal. They feared that opening of the article in the Council would send a signal to the EP, who could then put forward proposals for other amendments, e.g., that companies must take measures on the outcomes of the EAs and implement energy efficiency measures. The EC, Malta and the successive Council Presidencies wanted to keep the amendment of the EED limited, avoiding distortions in the negotiations with the EP. This non-action in terms of not including the issues of EAs in the 2018 amendment of the EED was a way by the EC and the Council Presidency to deal with the policy risk that the EP would come up with new controversial policy ideas (cf. Leong & Howlett, 2022). This was frustrating to the deputy director of GOS/MEE, who claimed that the EC “manipulated the Council Presidency to intrude the possibility of sovereign MSs to bring new policy ideas to the table”, and “distrusted the Council of being able to deal with potential policy proposals of the EP in trilogues”.

The issue was buried but re-opened by the EC with the recast of EED in 2021-2023. The proactive policy entrepreneurship of GOS/MEE in 2017 supported by IGs such as the Coalition for Energy Savings (CfES) and European Union Alliance for Saving Energy (EU-ASE), researchers and think tanks such as Centre for European Policy Studies (CEPS), the European Council for an Energy Efficient Economy (ECEEE), and ISI Fraunhofer Institute had born fruit. There was consensus among policy actors already when the EED was amended in 2018, that thresholds should be based on annual energy use in the future, instead of company size as in the original EED, but the policy window was not open. Thus, this was not an issue in the negotiations on the recast EED in the Council, nor in the EP or in the trilogues.

In summary, the different strategies used by EC (DG ENER) and GOS/MEE as PEs in the policy process on EED provisions for enhanced energy efficiency in companies are presented in table 5. Both EC and GOS/MEE acted as ‘advocates’ according to Arnold’s (2021) archetypes.

5.3. Strengthening the Role of Energy Efficiency in Climate Policy

5.3.1. The Problem and Policy Issue

The clean energy transition needed to limit global warming to 1.5 °C above preindustrial levels and meet the UN sustainable development goal on affordable and clean energy requires massive deployment of energy efficiency and renewable energy sources (IPCC, 2023). According to the International Energy Agency (IEA, 2021), energy efficiency can contribute with more than 40 % of the global GHG emission reductions needed to meet the Paris Agreement targets. IEA (2019) refers to energy efficiency as the first fuel in the clean energy transition.

With reference to the ‘first fuel’ narrative, and influenced by entrepreneurial advocacy of RAP and ECF, the EC under president Jean-Claude Juncker introduced the energy efficiency first (EE1) principle in the EU energy and climate policy narrative with the Energy Union communication (EC, 2015) and the 2016 Clean Energy for All Europeans legislative package (EC, 2016b), presenting it as the overarching theme of EU energy and climate policy. EE1 was introduced as a non-binding principle in EU legislation with the regulation on the governance of the Energy Union and climate action (Governance Regulation), defining it as: 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 (EU, 2018, p. 15).

The EE1 narrative was also included in the EU Strategy for Energy System Integration (EC, 2020). The 2021 EC proposal to recast the EED (EC, 2021a, 2021b) as part of EGD and FF55 included stronger provisions on EE1. The EP and the Council formally adopted the recast EED in July 2023 (EU, 2023). This made EE1 a legally binding principle for EU MSs to apply in policy, planning and decision-making on investments regarding the energy system and in sectors that affect energy supply and energy demand, exceeding euro 100 million each and euro 175 million for transport infrastructure projects. EE1 as an overarching, binding principle complements two other guiding principles of EU energy and climate policy, i.e., cost-effectiveness and consumer protection. EE1 indicates a paradigm shift contrasting the earlier path-dependent agenda for EU energy policy (Lafferty & Ruud, 2009; Kettner & Kletzan-Slamanig, 2020; Herranz-Surralés & Solorio, 2022), disconnecting the energy supply-side from the energy demand-side.

5.3.2. Motifs and Strategies of Policy Entrepreneurs

The introduction of the EE1 narrative in EU energy and climate policy, from the 2015 Energy Union strategy to a binding principle in the EGD/FF55 recast EED was much influenced by proactive policy entrepreneurship by the Brussels based, global private sector think tank Regulatory Assistance Project (RAP),⁸ the international environmental NGO (ENGO) European Climate Foundation (ECF),⁹ and the hybrid sector Energy Efficiency Financial Institutions Group (EEFIG)¹⁰, an expert group comprising over 200 organisations working with energy efficiency investments in EU. RAP, leading the work in the consortium, was motivated by a dedication to accelerate the clean energy transition through thought leadership, taking a systems perspective on energy policies traditionally addressed as silos, while ECF was motivated by a will to save the world from climate catastrophes and provided funding. RAP and EEFIG had strong analytical resources. Since 2015, they provided guidance in reports, policy papers and through meetings with EC on how EE1 has been and can be implemented as well as its benefits. EEFIG, working towards the financial sector to increase investments in energy efficiency, stressed that energy efficiency investments are strategically important for the EU and financial institutions, and that a binding EE1 with ‘multiple benefits’ framing can help the sector.

However, there was a lack of a politically understandable problem framing in 2015 and no policy window opened for making EE1 binding. The politics stream was not receptive to the techno-economic needs for systems approach and merits of consumer flexibility proposed by RAP and ECF. It ended with EE1 being defined in the Governance Regulation, without requirements for application in practice. It took until 2020-ish for the policy window to start open by a push from the problem stream, with clear problem narratives. Firstly, the Paris Agreement target to limit the temperature increase to 1.5 °C above pre-industrial levels, and consequential demands for extensive GHG emission reductions. This led the EC to present the EGD and EU to adopt the ECL. Second, Russia’s second war on Ukraine led the EC to present the REPowerEU Plan and the associated EU Save Energy Strategy, which stressed the need for stronger energy efficiency policy. These two external ‘events’ or ‘shocks’ are examples of the importance of “seizing synchronistic moments when such windows open from political situations ‘above’ or ‘below’” (Kreienkamp et al., 2022, p. 735). Third, indicators showed a large untapped economic potential for further energy savings. Fourth, the increased integration of renewable energy sources in the power system requires a systems approach to the green energy transition, with consumer flexibility and demand-side measures (Pató & Mandel, 2022). Fifth, the emergent discourse about ‘multiple benefits’ of energy efficiency (Fawcett & Killip, 2019), where energy efficiency, like a Swiss army knife, helps solving all kinds of problems (von Malmborg et al., 2023c). The coupling mainly followed the mechanisms of ‘consequential coupling’ (Zahariadis, 2003) and ‘problem surfing’ (Boscarino, 2009). The task for PEs was to find a solution to the problem. But there was also a tendency towards ‘doctrinal advocacy’ (Zahariadis, 2003) or ‘problem-focused advocacy’ (Boscarino, 2009), were the EC, led by DG ENER, as a PE in making EE1 binding looked for a problem that suited a given solution. EE1 was established as a non-binding principle with the Governance Regulation, but for making EE1 binding, the EC had to look for better problem narratives than in 2016–2018 when the EED was amended, and the Governance Regulation negotiated.

Active entrepreneurship by RAP, ECF and EEFIG as ‘advocates’ (cf. Arnold, 2021) inspired the EC to ‘advocate’ and sequentially introduce the narrative of EE1 in EU policy, first in the Energy Union communication (EC, 2015), second in the Clean Energy for All Europeans package (EC, 2016), manifested in the Governance Regulation (EU, 2018b), in the Strategy for Energy System Integration (EC, 2020), and finally as a self-standing, binding article in the recast EED (EU, 2023). The views of RAP, ECF and EEFIG were supported by other IGs, particularly CfES, an ENGO gathering more than 500 business and civil society associations, 200 companies, 1500 cooperatives and 2500 cities in favour of energy efficiency, and EU-ASE. They all stressed that the “EE1 principle should be systematically and consistently applied in EU law” (CfES, 2021, p. 27). RAP, EEFIG and ECF, acting as PEs primarily coupling the problem and policy streams, provided guidance in reports, policy papers and through meetings with the EC on how EE1 have been and can be implemented as well as its benefits for more than five years (Bayer, 2015; Rosenow et al., 2016; EEFIG, 2015; ECF, 2016; RAP, 2016). To sequentially influence the EC, they used so-called ‘salami tactics’ (Zahariadis, 2003), breaking favoured policies into modest segments and presenting them sequentially, steering policymakers toward the desired outcome without raising alarm at any one stage. These organisations were also successful in linking the problem and policy solution to part of the politics stream, influencing the EP rapporteur, calling for a very strong application of EE1.

In summary, the different strategies used by EC (DG ENER), RAP, ECF and EEFIG as PEs in the policy process on making EE1 a binding principle with the recast EED are presented in table 6.

5.4. Decarbonisation of Maritime Shipping

5.4.1. The Problem and Policy Issue

Shipping is a key sector for the EU economy (EC, 2013). About three quarters of all international trade in and out of the EU is carried by sea (European Maritime Safety Agency, 2023), and the sector contributes with EUR 149 billion to EU GDP and two million jobs in 2020, of which 685,000 at sea (EC, 2021e). Indicators showed that rising GHG emissions from maritime shipping is problematic. The EU shipping sector emitted 138 million tonnes CO₂ in 2018, corresponding to some 11 % of all EU transport CO₂ emissions and 3–4 % of total EU CO₂ emissions (EC, 2021e). The GHG emissions from EU maritime transport (i.e., emissions related to intra-EU routes and incoming and outgoing routes), increased by +48 % between 1990 and 2008 (EC, 2013). EU-related emissions from shipping are expected to further increase by 51 % by 2050 compared to 2010 levels. According to the EC (2021d, 2021e), the current fuel mix in the maritime sector is made up of >99 % fossil fuels. This is due to insufficient incentives for operators to reduce GHG emissions and a lack of mature, affordable, and globally utilisable alternatives to fossil fuels in the sector. Policy measures to decarbonize maritime shipping would best be regulated at international level, where the IMO adopted an initial climate strategy in 2018, with targets for reducing GHG emissions (IMO, 2018). This strategy was revised in July 2023, setting a net-zero emissions target by mid-century, and interim targets for 2030 and 2040 (IMO, 2023). However, no agreement was met on binding measures to reach the targets, why EU legislation is the only one in the world.

The issue of decarbonizing the EU maritime sector entered the EU policy agenda with the 2013 Strategy for integrating maritime transport emissions in the EU’s GHG reduction policies (EC, 2013). However, prior to the FEUM regulation, proposed by the EC in 2021 as part of EGD/FF55 and adopted by the Council and the EP in July 2023 (EU, 2023b), international shipping was the only mode of transport not included in the EU’s commitment for GHG emissions reductions. The policy window for FEUM started to open by a push from the problem stream, as a response to global warming and the need to reduce GHG emissions from all sectors. But views on the emission reduction targets for the sector and the meaning of ‘decarbonizing maritime shipping’ differed among policy actors. There were mainly two PEs in the FEUM case – the EC (through DG MOVE) and the European confederation of green mobility NGOs, Transport & Environment (T&E), both suggesting innovative policy change, but with different targets and different policy options.

Heavily influenced by the incumbent maritime and fossil fuels industries (InfluenceMap, 2023), such as the European Community Shipowners’ Associations (ECSA), FuelsEurope, the International Chamber of Shipping (ICS), the World Shipping Council (WSC), CLIA Europe (a cruise lines trade association), and SEA Europe (a group representing shipbuilders and maritime equipment manufacturers), the EC, under lead of DG MOVE, proposed a technology neutral and goal-based policy with GHG intensity targets (EC, 2021d). EC (2020) argued that a 90 % reduction in transport sector GHG emissions is needed by 2050 to achieve economy wide climate neutrality by 2050. In its RIA accompanying the proposal for the FEUM regulation, the EC (2021e, p. 1) argued that LoZeC fuels “should represent 6–9 % of the international maritime transport fuel mix in 2030 and between 86–88 % by 2050 to contribute to the EU economy wide GHG emissions reduction targets”. With reference to high capital expenditures and long commercial life of existing vessels and high costs of the most effective but also most expensive LoZeC fuels, i.e. electrofuels¹¹, the incumbent shipping industry and the EC argued that transformative technology is costly with a high risk of technology lock-in and stranded assets. Given the inherently global nature of the shipping sector, an important framing factor was the importance of flag neutrality and a favourable regulatory framework, a level playing field, for ports, ship operators and shipping companies (EC, 2021d). This framing and policy were in large supported by the shipping industry and MSs in the Council.

As an alternative to the EC proposal, and taking a different point of departure, T&E in coalition with Seas at Risk (SaR), an association of ENGOs dedicated to marine protection, and the Getting to Zero Coalition (GtZC)¹² called on EU legislators to set a clear target for zero-emission shipping by 2050, in line with the Paris Agreement, and make sure that new policy measures and interim targets work towards achieving the EU target of climate neutrality by 2050. T&E in collaboration with other IGs and progressive companies from the maritime value chain had a different view than the EC on how to stimulate innovation related to LoZeC fuels, particularly zero-carbon fuels such as electrofuels. T&E and allies claimed that the technology-neutral approach proposed by the EC will see ship operators choose the cheapest fuel options to cut emissions, i.e., liquified natural gas and biofuels, fuels that ENGOs oppose. T&E and allies argued that electrofuels are key to reach full decarbonisation by 2050. The share of green electrofuels must be at least 18 % in 2030 and 85 % in 2040, along strong energy efficiency measures (T&E, 2022a). Thus, T&E together with members of GtZC, the Clean Air Task Force (CATF), a US-based environmental organisation pushing for technology and policy changes needed to achieve a zero-emissions planet at an affordable cost, and many progressive companies and organisations across the maritime value chain (T&E, 2022b), wanted lawmakers to spur the uptake of electrofuels, via the inclusion of technology specific measures such as multipliers and mandatory sub-quotas. T&E proposed EU lawmakers to set a sub-quota of 6 % electrofuels and increasing from 2030 and bridge the cost-competitiveness gap with other alternative fuels via the introduction of a multiplier of 5 for electrofuels. According to T&E and its allies, high(er) sub-quotas will stimulate market development and deployment of electrofuels. A technology specific policy to stimulate innovation (production, provision and deployment) of the most advanced (but also most costly) sustainable maritime fuels, i.e. electrofuels, will drive down costs and help make the ensuing rapid scale-up and uptake of electrofuels commercially viable by 2030, making the 2050 clean energy transition end-date within reach. This could help unlock massive investments and job opportunities in Europe and globally, as well as contribute to making the EU the leading global supplier of zero-emission shipping and fuel production technology. This is also important from a strategic autonomy perspective, giving first mover advantage to the global hydrogen economy, T&E claimed. This framing and policy option was in large supported by the EP and some MSs in the Council.

After a combination of deliberative negotiations and bargaining, the Council and the EP reached a political agreement on FEUM in March 2023 (von Malmborg, 2023c, von Malmborg, 2024). The co-legislators agreed on legislation with increased ambitions for international GHG reductions compared to the proposal of the EC, but not to 100 % by 2050 as proposed by T&E, and they introduced a multiplier and potentially a sub-quota to stimulate innovation of electrofuels. The new regulation (EU, 2023b) was formally adopted by the EP and the Council in July 2023, establishing the world’s most ambitious legislation to stimulate decarbonisation of maritime shipping, beyond the borders of the EU. In short, the FEUM provides that:

• GHG intensities well-to-wake should be reduced by 2 % from 2025, 6 % from 2030, 14.5 % from 2035, 31 % from 2040, 62 % from 2045 and 80 % from 2050, and
• a sub-quota of 2 % for electrofuels will be adopted by 2034, if an analysis of the electrofuel market shows that a multiplicator is not enough to drive market development for electrofuels beyond 1 % in the fuel mix by 2031.

These requirements would apply to ships above 5 000 gross tonnes and encompasses 100 % of their intra-EU voyages and 50 % of their voyages between EU ports of call and ports of call located in third countries.

5.4.2. Motifs and Strategies of Policy Entrepreneurs

As mentioned, incumbents were successful in lobbying the EC as agenda-setter, proposing more moderate emission reductions and a technology-neutral, goal-based approach with no special treatment of the more innovative but also more expensive electrofuels. Incumbents were a barrier to an ambitious policy proposal of the EC, but they did not manage to stop special treatment of electrofuels in the final agreement of the EP and the Council on FEUM. Concerted action of ENGOs and progressive companies in the shipping value chain, led by T&E as a PE, gathered enough support in the EP and among MSs to stand the grounds against heavy lobbying by incumbent companies in the shipping and fossil fuels industry. The focus on leadership in the transition towards a hydrogen economy was important to gain political support among MSs and in the EP.

The motif of T&E to act as a PE was an ambition to counter the ‘evil’ fossil fuel lobby and advance EU climate policy as much as possible for the EU to reach zero GHG emissions and climate neutrality by 2050, saving the world from climate catastrophes. Opposing the beliefs and influence from incumbents, T&E emerged as a ‘threat-centered’ PE presenting a devil’s shift to the ‘evil’ policy proposal of the EC (cf. Arnold, 2022). To succeed, T&E put forward its views on the problem and its solution, both in the consultation process before the EC presented its proposal, and once the proposal went live. In the process of problem framing, T&E joined forces with other IGs, such as SaR, CtZC, CATF and many other companies and organisations across the maritime value chain to leverage the knowledge and skills of others towards a common goal (T&E, 2022b), e.g. knowledge about the technical feasibility and market development plans related to zero-carbon fuels. T&E had strong analytical capacity inhouse. To convey its messages and gather support for the alternative policy option, T&E used reports, open letters and participation in meetings and conferences. T&E, also participated in a targeted stakeholder consultations at the European Sustainable Shipping Forum and the European Ports Forum. T&E did also write op eds in online media such as Euractive. Advocating a different policy option than the EC and wanting to influence the Council and the EP once the EC had presented its proposal, T&E did also put much in building a broad coalition of ENGOs, IGs, elected politicians and companies in the shipping industry and fuels industry. In all, EC actes as an ‘advocate’ while T&E acted as an ‘activist’ PE (cf. Arnold, 2021).

In summary, the different strategies used by EC (DG MOVE) and T&E as PEs in the policy process leading to adoption of the FEUM regulation are presented in table 7.

6. Analysis and Discussion

6.1. Comparing Motifs and Strategies of Policy Entrepreneurs’

In the four cases described, there were in total seven organizations acting as PEs, some of which were the same in several cases (Table 8). In the cases of energy efficiency of buildings (IMB) and companies (EA/EnMS), respectively, the main PEs were the EC (DG ENER) and GOS/MEE (more specifically a deputy director at the Swedish Ministry of the Environment and Energy). In the case of EE1, the main PEs were the EC (DG ENER) and RAP, ECF and EEFIG. As for EC (DG ENER), leading on EED, the organization was similar in 2018 and 2012, but different in 2023, with new directors, heads of units and policy officers. Finally, in the case of FEUM, the main PEs were the EC (DG MOVE) and T&E. As mentioned in section 2.2, most PEs seek policy change. Thus, they can be considered proactive PEs (cf. Becker et al., 2024). However, it happens that PEs act to maintain status quo, reacting to ‘evil’ policy proposals by the EC. Out of the four cases, two included reactive PEs that at different times advocated status quo. The EC (DG ENER) advocated policy change while GOS/MEE advocated status quo regarding IMB in the process leading to the adoption of the original EED in 2012. However, in the process leading to the amendment of IMB provisions in 2018, the roles were switched. GOS/MEE, as an ‘activist’ advocated policy change to amend the ‘evil’ existing IMB provisions, while the EC advocated status quo. As for EED provisions on EAs/EnMS, GOS/MEE advocated policy change and the EC opted for status quo at the time of the revision of the EED in 2018. When EED was recast in 2023, both the EC and GOS/MEE advocated policy change. In the other cases, all PEs sought policy change.

6.1.1. Motifs

As for motifs, the EC aimed for an ambitious enough innovative package of policies and legislation that could make the EU meet its new climate targets set out in EGD and ECL. According to the heads of units in DG ENER and DG MOVE, contributing to European integration by presenting harmonised legislation was also paramount. As for GOS/MEE, the other public sector PE, active in two cases, the main motif when acting as a proactive PE can be interpreted as revenge and rectification, and a will to correct illegitimate thus ‘evil’ EU legislation causing problems for public administration in MSs as well as companies across the EU, all in all counteracting the aim of increasing energy efficiency. Finally, PEs from the private and hybrid sector and civil society, no companies or business associations but think tanks and NGOs, were motivated by their dedication to provide the most innovative climate policies in their respective fields of expertise, truly contributing to 100 % GHG emission reductions by 2050. ENGOs wanted policy to ‘save the world’. In two cases, PEs from the public sector and civil society were ‘threat-centred’ (cf. Arnold, 2022) and motivated by a desire to counteract ‘evil’ policy proposals from the EC in a ’devil shift’ (cf. Sabatier & Jenkins-Smith, 1999): GOS-MEE in the case of energy efficiency of buildings, and T&E in the case of decarbonizing shipping. In two cases, energy efficiency of buildings (GOS-MEE) and EE1 (RAP), PEs advocated policies based on a systems approach, aligning different EU laws that otherwise contradict each other. Notably, the lead PEs in these organizations both hold PhD degrees, which none of the other leads did. This is not significant for the strategies chosen, but it says something about their social positions and expertise needed to propose ground-breaking policy change.

6.1.2. Types of Entrepreneurship

From the case descriptions (section 4), it is evident that policy entrepreneurship was mainly of a cultural-institutional nature, aiming at altering or diffusing people’s perceptions, beliefs, norms and cognitive frameworks, worldviews, or institutional logics (cf. Boasson & Huitema, 2017). However, the actions of EC (DG MOVE) related to decarbonization of maritime shipping can also be seen as a case of structural entrepreneurship, aiming at overcoming the structural barriers to enhance governance influence by altering the distribution of formal authority (see further discussion on the EC as a PE in section 6.3). The EC was the only PE to use structural entrepreneurship. In most cases, PEs aimed at altering perceptions, beliefs, norms and cognitive frameworks and worldviews of different policy actors and policymakers in the Council and the EP. However, policy entrepreneurship of GOS/MEE on energy efficiency of buildings and companies in 2016–2018 was also addressing the EC (DG ENER), aiming to influence the EC proposal for an amended EED. The EC (DG ENER) was the main addressee of RAP, ECF and EEFIG in the case of EE1. They aimed at influencing what the EC proposed in different policy documents.

6.1.3. Strategies

But what strategies were used by the different PEs? Did strategies differ depending on which sector they came from, if they were advocating policy change or status quo, and over time? As summarised in table 9, it seems that strategies of PEs in the four cases analysed and compared do not differ significantly between the public, private and civic spheres of society. They use all four types of strategies. Though, it can be noted that public and civil society PEs try to raise attention and support also by media attention. In addition, it seems that civil society PEs focus on building broader coalitions than public sector entrepreneurs, covering hundreds or thousands instead of tens of organisations. In addition, civil society PEs acted more like ‘activists’, while private and hybrid sector PEs acted more like ‘advocates’ (cf. Arnold, 2021). Their policy ideas served as ‘coalition magnets’ (cf. Béland & Cox, 2016). The cases of EE1 and FEUM saw RAP, ECF, EEFIG and particularly T&E build coalitions including MS governments, EU politicians, companies, business associations, local authorities and NGOs. In comparison, public sector entrepreneurs focused on building coalitions with MS governments, EU politicians and business associations, leaving individual companies, local governments and NGOs behind.

6.2. Differences When Advocating Policy Change or Status Quo

Table 10 summarises the results on strategies employed by PEs advocating policy change or status quo. It is worth noting that maintaining status quo was the aim in two cases (IMB and EA/EnMS) and advocated only by PEs from the public sector (GOS/MEE and the EC/DG ENER). It is evident that PEs aiming at policy change act like ‘activists’ or ‘advocates’ and use a broader palette of strategies than PEs aiming at maintaining status quo, acting more like ‘concerned citizens’. They use more attention- and support seeking strategies as well as relational management strategies. When GOS/MEE advocated status quo instead of introducing provisions on IMB for enhancing energy efficiency of buildings in 2011–2012, it only used attention- and support seeking strategies. When the EC advocated status quo in relation to provisions on enhancing energy efficiency in companies in 2016–2018, it mainly used linking strategies but also referred to bad timing to stop policy change. This confirms the importance of the policy window of opportunity for PEs transferring experiences and new knowledge to policy learning and policy change. Despite the knowledge of EC in 2016–2018 that provisions on EAs would better be changed, it took until 2021 before a policy window was open by a push from the politics stream and the EC proposed such change.

6.3. Differences over Time

Next research question relates to changes of strategies over time. In the case of energy efficiency of buildings (IMB), both GOS/MEE and the EC changed strategies from 2011–2012 to 2016–2018, as they switched from advocating status quo to policy change and vice versa. Advocating policy change, GOS/MEE’s approach was more elaborated and included many more strategies compared to when GOS/MEE advocated status quo. The change of approach, from the ‘concerned citizen’ to the ‘activist’ was very much dependent on the idea and efforts made by the deputy director in charge of EED at GOS/MEE. He actively approached key persons in different IGs to take them onboard and make a joint action towards the EC, the EP and the Council. Similarly, the EC used fewer strategies to maintain status quo when the IMB provisions were renegotiated in 2016–2018. In the case of enhancing energy efficiency in companies, the EC also switched from advocating policy change to maintain status quo, then back to policy change. The policy debate was most heated in 2016–2018, when GOS/MEE fought for policy change and the EC advocated status quo. When the EC finally proposed policy change in 2021, based on previous proposals from GOS/MEE and allies, all parties where already onboard and the need for active entrepreneurship was limited.

In the case of EE1, RAP, ECF and EEFIG as well as EC used ‘salami tactics’ to sequentially introduce the EE1 narrative in EU climate and energy policy. Initially, they focused on attention- and support seeking strategies, framing the problem, proposing policy and using information in a strategic way. Problem framing developed over time, as the politics stream was not receptive for including EE1 as a binding principle until the policy window opened around 2020. Thus, timing was important. Now, EE1 was framed in relation to five different problems. Advocating EE1 as a binding principle also required coalition building and networking to attract broad support. Finally, the case of policy for decarbonizing maritime shipping. The issue was raised by the EC in early 2010s and introduced in the strategy for integrating maritime transport emissions in the EU’s GHG reduction policies (EC, 2013). This strategy proposed a three staged approach, starting with introducing requirements for shipowners to monitor, report and verify CO₂ emissions, adopted in 2013. Next step was to introduce targets for GHG emission reductions, which were proposed in the Climate Target Plan (EC, 2020). The final step was to develop policy instruments to meet the targets, e.g. FEUM. Similar to the strategies for introducing EE1, this staged approach of the EC can be seen as ‘salami tactics’, although the proposal for policy instruments to reduce GHG emissions was presented only in the third stage.

In all, this confirms the findings of Braun et al. (2024) that PEs employ different strategies over time when interacting with their immediate and wider contexts in attempting to foster policy change. PEs are endogenous to policymaking processes and co-create with policymakers to shape their ecosystems and society at large (Bakir & Jarvis, 2017).

6.4. Coalition Building

As seen in all cases, and used by PEs from all social spheres, coalition building is an important strategy to successfully influence policy change. Through careful construction of coalitions/teams, it is possible for a PE to find effective ways to leverage the knowledge and skills of others towards a common goal. Policy networks, as repositories of knowledge, know-how, war stories, and professional gossip, can be vital sources of information for PEs. Knowing how to effectively tap local and broader networks can do much to help PEs advance their advocacy efforts (Petridou & Mintrom, 2021). The approach of GOS/MEE in 2016–2018, forming a broad normative coalition advocating change of the IMB provisions, shows the importance of coalition building for small MSs to be successful as a PE in the EU (cf. Selck & Kuijpers, 2006; Björkdahl, 2008; Panke, 2012; von Malmborg, 2022). In the case of EE1, RAP, ECF and EEFIG joined with CfES and EU-ASE, encompassing in total 5–6 000 organisations. In the case of FEUM, T&E gathered with more than 200 ENGOs, business associations related to sustainable fuels and sustainable shipping as well as progressive companies across the maritime shipping value chain. Notably, the informal advocacy coalition built by RAP/ECF/EEFIG and T&E also included formal coalitions such as CfES and GtZC respectively.

6.5. Comparing Success and Failure of Policy Entrepreneurs

Most studies of PEs focus on successful cases. Boasson & Huitema (2017) and Arnold et al. (2023) argue for more studies also of unsuccessful cases. “Understanding why entrepreneurs are more successful in some cases than in others requires study of unsuccessful cases as well. /…/ Only by juxtaposing cases with different outcomes can we gain better understanding of how contextual factors may impede or enable entrepreneurs to influence policy and governance” (Boasson & Huitema, 2017, pp. 1351-1352). As mentioned in section 2.3, success or failure can be related to outcomes such as (i) influencing agenda-setting, (ii) changes in other policy actors’ policy preferences, and (iii) actual influence on policy and governance.

From the four cases, we find that entrepreneurship of GOS/MEE was unsuccessful in (i) maintaining status quo regarding IMB in 2011–2012, and (ii) advocating policy change regarding energy efficiency in companies in 2016–2018, but successful in advocating policy change regarding IMB in 2016–2018 and energy efficiency in companies 2021–2023 (von Malmborg, 2021, 2022; von Malmborg & Strachan, 2023). Despite initially being unsuccessful in changing policy on energy efficiency in buildings and companies, GOS/MEE was successful in setting the agenda at an early stage in both cases, and particularly successful in changing the perceptions of other policy actors on energy efficiency in companies. Second, the strategy of the EC maintaining status quo related to IMB in 2016–2018 was rather unsuccessful (von Malmborg, 2021, 2022), but its strategy to maintain status quo on energy efficiency in companies in 2016–2018 was successful in terms of not changing the policy (von Malmborg & Strachan, 2023). In both cases, the agenda was set by the Swedish government, through GOS/MEE.

The strategies of RAP, ACF, EEFIG and EC, acting as ‘advocate’ PEs to sequentially introduce the EE1 narrative in EU climate and energy policy were very successful in terms of agenda-setting, changing perceptions, and changing policy, as EE1 was adopted as a binding principle in 2023 (von Malmborg, 2023a, 2023b). In the FEUM case, strategies of both the EC as an ‘advocate’ and T&E as an ‘activist’ were successful regarding agenda-setting and changing perceptions, and moderately successful in terms of influencing policy change, with the Council and the EP bargaining to find a compromise that included elements of both the EC and the T&E narratives (von Malmborg, 2023c, 2024b). T&E was particularly successful in gathering support for an alternative policy design to influence the EP and enough powerful MSs in the Council to call for higher emission reductions and special treatment of electrofuels. In all, there seems to be no difference in success related to the sector from which a PE comes (Table 11).

However, we can find a clear pattern regarding PEs advocating policy change or status quo and PEs acting as ‘activists’, ‘advocates’ and ‘concerned citizens’. Evidence from this study indicates that PEs advocating policy change are more active and use more elaborated strategies than PEs advocating status quo on the same policy issue (IMB and EA/EnMS respectively), and they are more successful in changing the perceptions of policy actors and influencing actual policy change (Table 12). But how can these differences be explained? Since the issues and macro- and meso-level contexts were similar for both kinds of PEs in the two cases, this difference could be explained by differing ‘actor specific’ factors, i.e. micro-level context (cf. Mahoney, 2007, 2008; Klüver, 2011, 2013; Bakir & Jarvis, 2017; Arnold, 2021). To be successful, a PE must understand the concerns of actors they seek to persuade, use social acuity to build teams, networks, and coalitions, be knowledgeable to strategically disseminate information, and be organizing, corresponding to political activation and involving civic engagement (Mintrom & Norman, 2009; Aviram et al., 2020; Arnold, 2021). When GOS/MEE advocated status quo on IMB in 2011–2012, advocacy and policy entrepreneurship in EU affairs was new to the persons in charge, with little legacy of EU advocacy in the energy division at the ministry. They knew how to disseminate information strategically but held little social acuity in EU circles and did not understand the value of teaming up with relevant competencies and actors with relevant personal positions in networks and coalitions. The transaction costs were high. In summary, GOS/MEE used bad strategies to try and advance policy goals which led to PE failure (cf. Arnold et al., 2023). The EC, maintaining status quo on IMB and EA/EnMS in 2016–2018, were more knowledgeable as a PE, but was taken by surprise by the new entrepreneurial approach taken by GOS/MEE, and particularly that GOS/MEE managed to build a coalition of IGs that for the first time ever had a shared position. This came as a shock to the EC head of unit, who was caught off guard and failed in maintaining status quo. A Swedish consultant in the field of energy, based in Brussels, was also surprised and told that she had never before seen GOS/MEE act in this proactive way in Brussels. (von Malmborg, 2022).

In the case of energy efficiency in companies, GOS/MEE acted proactively as an ‘advocate’ PE in 2016–2018 and gained support from all MSs and the EC, but the policy didn’t change until 2023. The initial lack of success can be explained by a not fully open policy window, and perhaps more important that the EC as PE wanting to maintain status quo acted as an ‘entrepreneurial gatekeeper’. In this case, PE strategies were good, but there was a lack of facilitative characteristics (cf. Arnold et al., 2023).

The problem framing and policy proposal of T&E on decarbonizing of shipping were transformative. Similarly, RAP, ECF and EEFIG inspired the EC to present a transformative policy change on EE1. As highlighted by Kreienkamp et al. (2022), the ‘hyper-consensual’ environment of EU politics poses a great challenge to PEs seeking to advance transformative policy change. The central role of consensus-building in EU policy change serves to keep a path dependent dynamic, where no single actor can prescribe policy and minority interests can be easily mobilised to block reform (Hix, 2007). T&E and allies had no influence on the EC proposal, some influence on the Council, but was rather successful as an ‘activist’ PE towards the EP. The durable entrepreneurship of RAP, ECF and EEFIG as ‘advocates’ got the EC to propose gradually stronger provisions on EE1. The EC itself was receptive and acted strategically on ‘external shocks’ to reframe energy efficiency policy and the role of EE1 to make it appetising for EU politicians. T&E, RAP, ECF and EEFIG were all successful in coupling all three streams according to their views. In all, this was not only a ‘coincidence of propitious conditions’ (Palmer, 2015, p. 281) in the EU politics stream, but a strategically outlined act of entrepreneurial agency, either as an ‘activist’ or as ‘advocates’. In all, this echoes the findings of Ringel et al. (2021, p. 9) that “especially hybrid stakeholders [such as EEFIG and the coalition of EEFIG, RAP and ECF], combining industry, think tanks and NGO actors, can take a strong role for consensus building” on EU energy efficiency policy. The situation was similar in the case of FEUM, where T&E built a broad coalition of industry, think tanks and NGOs. Similar findings were reached by Carter and Childs (2018), analysing the policy entrepreneurship of ENGO Friends of the Earth in UK climate policy. The strategies used by T&E and ECF (with RAP and EEFIG) confirms research in political sociology that success of ENGOs in advocacy on EU environmental policy depends on “effective coalitions, political opportunities in the form of either powerful allies in the EU institutions or, conversely, the possibility of mobilizing grassroots members to act in particularly influential member states, are found to be central to explaining influence over EU decisions” (Parks et al., 2023, p. 106).

In all, these findings confirm the findings of Arnold (2021) that PEs characterized as ‘activists’ and ‘advocates’ use a broader set of strategies and are more successful in influencing policy change, while PEs being characterized as ‘concerned citizens’ are less active and less successful in policy change, although they can have an impact on agenda-setting and influencing other policy actors’ beliefs. In addition, it can be noted that civil society PEs tend to be more activistic than PEs from the public and private sector, although GOS-MEE deployed activistic agency to change EED provisions on IMB in 2016–2018. The characteristics and type of entrepreneurship employed by PEs also seem to depend on the target group of advocacies. Compared to GOS-MEE and T&E targeting the EC, MSs in the Council in MEPs and party groups in the EP as ‘activists’, RAP, ECF and EEFIG acted as ‘advocates’ to influence the EC. The wider the target group is, the more activistic PEs, at least if they are not from EU institutions. EC also targets MSs, MEPs and party groups, but act as an ‘advocate’. It holds a special position in EU policymaking, with a formal privilege to propose new legislation, thus in less need of activistic entrepreneurship.

6.6. European Commission as a Policy Entrepreneur

Through its monopoly on presenting legislative proposals, the EC is a central and ‘natural’ PE in EU politics and governance (e.g. Herweg & Zohlnhöfer, 2022). But the EC is not unaffected in its role as PE. The EC is heavily dependent on external expertise, and strong lobbying and policy entrepreneurial advocacy are carried out on the EC from various quarters, including private profit and non-profit, public and civil society organisations from the EU and abroad, before EC presents its proposals for legislation (Broscheid & Coen, 2003; Chari & O’Donovan, 2011; Coen et al., 2021). As described in section 4, this happened in all cases of this study. Sometimes to the better for the climate, sometimes to the worse.

Kreienkamp et al. (2022, p. 735), analysed transformative change of EU climate policy from a combined perspective of MSF and multilevel governance, emphasising “the importance of creating windows of opportunity, but also seizing synchronistic moments when such windows open from political situations ‘above’ or ‘below’”. The importance of such situations from ‘above’ was apparent for the EGD/FF55 policy proposals of the EC in 2021.

The framing of climate and energy policies in the EU has been analysed, focusing on individual policy packages at specific points in time (Maltby, 2013; Palmer, 2015; Bocquillon, 2018) as well as over time (von Malmborg et al., 2023c). From around 2005, the EC acted for a shift in political norms, successfully framing energy import dependency as a problem requiring an EU-level solution (Maltby, 2013). Palmer (2015) found that persuasive framing enabled the PE to impinge agenda-setting processes, while boundary work enabled the PE to forfeit an existing policy considering widespread criticism. The Juncker energy and climate policy package from 2009 gathered support among stakeholders from framing it as a way for the EU to gain leadership in the green energy transition (Bocquillon, 2018). The situation was similar with EGD/FF55 (Kreienkamp et al., 2022).

Framing of EU energy efficiency policy has expanded over time, from focusing on energy security of supply, via climate change to energy poverty and other ‘multiple benefits’ (von Malmborg, 2023a; von Malmborg et al., 2023c). The EC have a central role in this change, both as a structural PE (changing power relations) and a cultural-institutional PE (changing worldviews). As found by von Malmborg et al. (2023c), EC acts to expand its reach into areas where the EU does not hold much legal competence according to the Treaty of the Functioning of the EU (TFEU) (EU, 2012b), such as buildings and social affairs policy, through the domain of energy efficiency policy by referring to its ‘multiple benefits’. From the EC perspective, energy efficiency policy is a Swiss knife, being able to solve many different problems, including energy poverty. From the critics’ perspective, it is a Trojan horse, were the EC in a non-transparent way, so-called ‘harmonization by stealth’, tries to sneak in new policy domains through disguising them as energy efficiency policy to increase EU competence in relation to national governments. In that sense, the policy entrepreneurship of the EC is structural, furthering the European integration, and included actions aimed at changing governance influence by altering the prevailing distribution of authority in a way that gives the EC more power (cf. Boasson & Huitema, 2017).

The situation was similar in the FEUM case. Despite shared responsibilities between EU and MSs on energy, climate and transport policy, enshrined in TFEU, FEUM is a regulation, not a directive, with limited flexibility for MSs. The rationale of the EC to propose a regulation was that EU policy on decarbonisation of shipping must consider the fact that maritime transport services in EU and between EU and non-EU countries can be provided by operators of all nationalities. The global nature of the sector underlines the importance of flag neutrality and a favourable regulatory framework, which would protect the competitiveness of EU ports, ship owners and ship operators. A level playing-field for ship operators and shipping companies is critical to a well-functioning EU market for maritime transport. In addition, FEUM encompasses ship owners and ship operators in EU as well as third countries, stretching beyond the borders of the EU. FEUM has generated a multilateral process including IGs from national, EU and international levels. The role of the EU in IMO and its strong support for multilateralism provided motivation to advance on maritime climate policy to lead by example. FEUM was framed by the EC as an opportunity to get global leadership in the clean energy transition and to reinforce the legitimacy of the EU and its institutions.

These findings contradict the claim by Brouwer (2015) and Brouwer and Huitema (2018) that the strategic behaviour of public sector PEs is not about double-dealing or playing nasty tricks. This creativity of the EC to be a successful structural PE, being a ‘purposeful opportunist’ in interpreting rules and procedures in an ever more complicated and contested EU, was highlighted also by Copeland (2022) and Becker (2024). Some may applaud the EC for trying to pressure MSs into engaging with the EGD, especially if they see it as promoting the delivery of ‘public goods’. Others might find that the EC transcends its mandate (Becker, 2024). In that sense, the policy entrepreneurship of the EC was structural and included acts aimed at changing governance influence by altering the prevailing distribution of authority and strategic use of information in a way that gives the EC more power (cf. Boasson & Huitema, 2017). In addition, Dreger (2014, p. 176) argues, that the EC problem framing powers can be strategically employed to ‘technocratise’ subsequent policy debates, ‘pull[ing] political actors toward those grounds where the EC has a home field advantage’. Proposing a regulation instead of a directive on FEUM also gains the EC ‘implementation powers’ (Bauer & Becker, 2014; Crespy & Menz, 2015).

This phenomenon of ‘competence creep’, where the EU acts outside of its powers and slowly expands its competences beyond what is conferred upon it by its members, informally increasing the powers of the EC while reducing the powers and flexibility of MSs, has been acknowledged for centuries (Garben, 2019). With the Lisbon Treaty, attempts were made to combat ‘competence creep’, but several flaws in the Treaty still makes it occur (Fox, 2020). Despite being legal, the ‘competence creep’ comes with democratic deficits related to accountability and legitimacy and should be combated to reinstate and reinforce the position and powers of both the national and European legislator in taking the important decisions that impact, directly or indirectly, the lives of European citizens (Garben, 2022).

Similar to Edler and James (2015), the case of FEUM with two competing problem framings (von Malmborg, 2023c, 2024b), but also the case of EE1 with problem framings evolving over a long period and two competing discourses (von Malmborg, 2023a), indicate that ambiguity about the problem is a key feature in the complex process of framing and mobilisation. Thus, ambiguity can be seen as a multi-dimensional and dynamic phenomenon that changes its nature and function over the different stages of the agenda setting process, and that opens a window of opportunity for the EC to take on a more structural entrepreneurial role. In all, this suggests that impacts of PEs, particularly PEs employing structural entrepreneurship, on democratic norms and principles such as accountability and legitimacy should be analysed (cf. von Malmborg, 2023d, 2023e).

7. Conclusions

This paper aims to further the understanding of strategies used by PEs from the public, private and civic sector in the policy process, if they differ and which strategies that are successful/unsuccessful. Do strategies change over time? It also aims to elucidate whether PEs seeking policy change are more active and successful than PEs advocating status quo. Addressing these questions makes the paper contribute theoretically to research gaps in the literature on PEs (Petridou & Mintrom, 2021; Arnold et al., 2023). To meet these aims, I have made a comparative analysis of motifs and strategies and related success or failure of PEs related to new and revised EU legislation to stimulate a clean energy transition to reach the new EU target on climate neutrality. Four cases focusing on EU supranational policies were compared: (i) enhanced energy efficiency in buildings, (ii) enhanced energy efficiency in companies, (iii) strengthening the role of energy efficiency in climate policy with the EE1 principle, and (iv) decarbonisation of maritime shipping. The comparative analysis further contributes to the theory on policy entrepreneurship in climate governance (cf. Boasson & Huitema, 2017; Green, 2017; Mintrom & Luitjens, 2017).

In the four cases analysed, there were in total seven PEs, some of which were the same in several cases. Three were public sector actors, one was a private sector actor, one a hybrid sector actor and two were civil society actors. The EC was a PE in all cases. Policy entrepreneurship was mainly of a cultural-institutional nature, aiming at altering or diffusing people’s perceptions, beliefs, norms and cognitive frameworks, worldviews, or institutional logics (cf. Boasson & Huitema, 2017). However, the actions of EC (DG MOVE) on policy for decarbonizing maritime shipping and to some extent of EC (DG ENER) strengthen the role of energy efficiency in climate policy, can also be seen as structural entrepreneurship, aiming at overcoming the structural barriers to enhance governance influence by altering the distribution of formal authority. In all cases but making the EE1 principle binding, where PEs aimed at influencing the EC to propose new policy, PEs were mainly targeting different policy actors and policymakers in the Council and the EP. However, policy entrepreneurship of GOS/MEE in 2016–2018 was also addressing the EC (DG ENER), aiming to influence the EC proposal for amending the EED.

The comparative analysis shows that motifs for engaging as PEs differ. Some are acting more instrumentally rational, while others are engaged to present a competing policy to prevent the EC to secure an ‘evil’ policy. Such ‘threat-centred’ entrepreneurship (Arnold, 2022) was deployed by a public sector PE, but not in all cases it participated, and by civil society PEs. Strategies of PEs in the four cases do not differ significantly between the public, private and civic spheres of society. They use all four types of strategies. Though, results indicate that PEs targeting policmakers in the Council and the EP tend to be more activistic than PEs targeting the EC. In addition, civil society PEs focus on building broader coalitions than public and private sector entrepreneurs, including not only business associations and MS governments, but also EU politicians, companies, local authorities and NGOs. Similar results were achieved by Carter and Childs (2018). As for success of PEs, the analysis shows no evidence that PEs from a certain sector would be more successful than others in setting the agenda, changing perceptions of policy actors, or influencing actual policy change. However, we can find a clear pattern regarding PEs advocating policy change or status quo. Evidence indicates that PEs advocating policy change are more active and use more elaborated strategies than PEs advocating status quo on the same policy issue. Both kinds of PEs are successful in agenda-setting, but PEs advocating policy change, be they ‘activists’ or ‘advocates’ are more successful in changing the perceptions of policy actors and influencing actual policy change. Threat-centred PEs, driven by beliefs and ambitions to counteract ‘evil’ policy proposals seem to be as successful as more instrumentally rational PEs if they know how to deploy different resources and strategies.

This paper builds on a comparative study of four cases with in total seven different PEs. Results should be seen as indications (hypotheses) of differences in motifs, strategies, success and failure of different types of PEs with different aims for their entrepreneurship. More comparative research is needed to verify or falsify the ‘hypotheses’.