Introduction

Political interest in Research Infrastructures (RIs) on a European scale has been a new phenomenon, marked in the early 2000s with the launch of the Lisbon Strategy and the European Research Area (ERA) (Hallonsten 2020; Cramer et al. 2020; Ulnicane 2020). Political rhetoric promotes RIs as a blanket approach to research, technology, innovation and Grand Challenges, including the United Nations Sustainable Development Goals and acute crisis scenarios (Czech Presidency of the Council of the EU 2022). Yet, the European Commission’s (hereafter: Commission) definition of RIs is also criticized for its lack of clarity. It includes a diverse set of scientific tools, such as single-sited (telescopes or accelerators) or distributed facilities (networks of laboratories or arrays of instruments), collections and repositories of data and materials (archives, models or bioresources) as well as research vessels (icebreakers or research planes) or computing grids (Cramer and Hallonsten 2020).

Since the early 2000s, European RI policy developed along three policy initiatives: first, the creation of the European Strategy Forum on Research Infrastructures (ESFRI) in 2002 and its publication of a roadmap since 2006; second, the partial rededication of the Framework Programme for Research (FP) that earmarked a budget for the preparatory phases of RIs since the seventh FP (2007–2013, hereafter: FP7); and third, the enactment of the European Research Infrastructure Consortium (ERIC) in 2009, a specific framework governed under EU law that grants legal personality to RIs (Ástvaldsson 2022; Bolliger and Griffiths 2020; Moskovko et al. 2019; Stahlecker and Kroll 2013; Ulnicane 2020). European RI policy developed through, first, the strategic incorporation of incumbents, i.e. “privileged ‘insiders’” (Edler and James 2015: 1259) and “organized interest groups” (Hensell 2022: 153) like, for instance, existing research funding agencies or research organizations through new modes of coordination; second, the Commission’s emphasis of joint responsibility at the supranational level, claiming its own accountability and mobilizing the subsidiarity principle to its advantage; third, the incentivization of conformity to the Commission’s policy agenda through generous financial schemes and fourth, the implementation of tailor-made legislation.

European RI policy added to the complex organisation of Big Science projects in Europe that date back to the end of World War Two. Big Science mainly refers to large and collaborative projects such as the particle physics organization CERNFootnote 1, the synchrotron facility ESRF or the neutron research reactor ILL (Cramer 2020; Hallonsten 2016). They can be characterized as institutionally independent from formal and supranational policy-making processes within the European Union (EU) and its predecessors because they have been based on ad-hoc high-level political agreements among national governments (Krige 2003).

The emergence of a European RI policy constitutes a topical case in research, science policymaking and European integration. Yet, the relative novelty of RI policy on the agenda of the EU in conjunction with its complexity and analytical vagueness may explain why this phenomenon has so far been a blind spot in scholarly research. However, extensive literature already exists in the fields of European research and innovation policy as well as European integration, addressing the Commission’s involvement in the policy processes (Banchoff 2002; Edler and James 2015; Hensell 2022; Kaiser and Prange 2002; Kreppel and Oztas 2017). This scholarly research provides a rich foundation for this article to draw upon highlighting, on the one hand, that the evolution of RI policy testifies to a surprisingly powerful role of the Commission within the domain of research policy that struggled for a long time to lift competences from fragmented national approaches to the supranational scale at all (Banchoff 2002; Flink 2016). These dynamics resemble in some aspects developments in related policy fields such as security research (Edler and James 2015; Hensell 2022) or energy (Maltby 2013) as well as studies that highlighted how the Commission turned into an active shaper of research policy since the early 2000s (Edler and James 2015; Tamtik 2016; Wedlin and Nedeva 2015). On the other hand, RI policy also resonates with a novel understanding of European integration, as evidenced by studies on concepts like “new intergovernmentalism” or “intergovernmental union” (Fabbrini and Puetter 2016). Within these concepts, European integration lacks the comprehensive transfer of nation-state competences to supranational bodies but relies on new forms of coordination outside the legislative treaty frameworks that—at the same time—limit the Commission’s power to act (Fabbrini and Puetter 2016; Kaiser and Prange 2002).

Hence, while this article speaks to current debates in EU studies and research policy, it also amends previous analyses of Big Science as an empirical puzzle within European integration and launches an effort to come to terms with the new phenomenon of RIs (Cramer and Hallonsten 2020). Its added value is a closer examination of the interaction of individual strategies and policy instruments which collectively shed new light on the Commission's efforts to shape research policy in the context of the complex governance arrangements of the EU. The present article thus is a first attempt to provide a picture of the structures and mechanisms behind the emerging European RI policy since the 2000s and its current effects on the landscape of RIs. Our analysis includes the qualitative analysis of policy documents as well as the quantitative analysis of data from the Horizon Dashboard and the CORDIS database. We employ a top-down perspective on policymaking and implementation and hypothesize the institutionalization of a new layer of collaborative RIs beyond historically longstanding and institutionally well-established Big Science projects.

The article is structured as follows. We start by reviewing the historical context of RI policy and introducing the Commission’s three major RI policy instruments. Subsequently, we present our methods and materials, before we describe our results. We then systematically discuss our findings and provide an outlook for further research.

The Emergence of Research Infrastructure Policy

Until the 1980s, European research policy was characterized by a fragmented division of responsibilities between the European Economic Community (EEC) and its member states (Guzzetti 1995; Tindemans 2009). This was mainly due to a lack of political will to escalate research policy from a national priority to a European competence (Delanghe et al. 2009; Guzzetti 1995; Papon 2004). The Single European Act (SEA, see Art. 24) in 1986 and the Treaty of Maastricht (Treaty of European Union/TEU, see Art. 130) in 1992 clarified some competences within European research policy (Chou 2012, 2014). But supranational initiatives remained limited by the principle of subsidiarity (TEU Art. 5(3)) regulating the exercise of EU's non-exclusive powers: 

“It rules out Union intervention when an issue can be dealt with effectively by Member States themselves at central, regional or local level. The Union is justified in exercising its powers only when Member States are unable to achieve the objectives of a proposed action satisfactorily and added value can be provided if the action is carried out at Union level” (European Parliament 2024).

These constraints of necessity and added value did not allow for a research policy of a supranational character. The birth of the FPs in 1984 pooled research funding for the first time together into a common effort (Flink 2016; Guzzetti 1995; Ulnicane 2020). Chou (2012) describes the historical trajectories of European research policy throughout the 1980s and 1990s as marked by incremental change by which treaty changes in combination with the growth of the FPs extended the competences of the EEC/EU. However, the EU found itself in a severe crisis in the 1990s that limited the further development of a common research policy because it lacked commitment and involvement from the member states (de Elera 2006; Tamtik 2016).

This difficult decade culminated in the transformative approach of the Lisbon Strategy in 2000 that set out the ambitious goal to turn the EU into the world’s leading knowledge-driven economy and a major player in technology and innovation (European Commission 2000; European Council 2000). At around the same time, the Commission proposed ERA, framed as a kind of internal market for research (European Commission 2000). It called for the creation of a comprehensive strategy to foster science, technology and knowledge production that uses the full portfolio of financial, practical, and legal policy instruments available to the Commission (Chou 2012). Eventually, research policy became a shared competence in 2009 with the Lisbon Treaty (Treaty on the Functioning of the European Union/TFEU Art. 2 and 4), enabling the Commission to initiate new policies and legislation. This could be done under the precondition that they continued to align with the subsidiarity principle and do not interfere with the competences and responsibilities of the member states (Chou 2012; Flink 2016).

The realization of Big Science projects has long been and remains a concern of ad-hoc intergovernmental politics across member states and international partners (Hallonsten 2014, 2016; Krige 2003). However, the launch of ERA in 2000 can be considered as a decisive step for the emergence of a distinct European RI policy, paving the way for a strategy-led and coordinated “European approach” (European Commission 2002: 10). The Commission (2002: 10) considered RIs among those “areas where a European approach is called for, given the levels of funding and the need for them to be given the means to ensure they are able to provide services on a European scale”. It began to claim a complementary role in the governance of European RIs, whereby it aims to coordinate its activities with national governments that remain responsible for the operation and funding of RIs (ESFRI 2018; European Parliament and Council of the European Union 1999, 2002). Considering these recent developments, the tangible relationships between longstanding intergovernmental Big Science projects and the growing engagement of the Commission in RI policy are not yet systematically understood (Cramer et al. 2020).

The European Strategy Forum for Research Infrastructures (ESFRI) and Roadmaps

In 2002, the Commission established ESFRI based on a mandate from the Council (ESFRI 2017). ESFRI evolved from an informal group of delegates from national ministries and was promoted as an advisory body to the Commission (Cramer 2020). It is a self-regulated body with delegates from member states, associated countries and representatives of the Commission, regularly joined by delegates from the European Intergovernmental Research Organisations Forum (EIROforum) (ESFRI 2017). EIROforum unites Europe’s major Big Science projects—CERN, EMBL, ILL, ESO, ESA, ESRF, EUROfusion, European XFEL and ILLFootnote 2—that carry decades-long experiences of operating large-scale research projects. Officially founded in 2002, EIROforum’s establishment has been a direct response to the Commission’s emerging role in research policy and its vision of ERA (Praet and Mazurelle 2001). ESFRI further consults with other stakeholders of the European RI landscape, such as the European Science Foundation and the Global Science Forum (ESFRI 2005). ESFRI has no decision-making power or institutionalized political authority and operates in the modus of the Open Method of Coordination (OMC) (ESFRI 2017). The Commission had introduced OMC in 2000, bringing together European policymakers as well as experts and delegates from the member states to improve mutual learning and policy coordination. Initiated as a general policy tool to realize the goals of the Lisbon Strategy, it became particularly used in research and innovation policy from 2003 onwards (Prange and Kaiser 2005; Tamtik 2012; De Ruiter 2010).

Based on a recommendation by the Competitiveness Council in 2004, ESFRI was tasked to prepare a roadmap that outlines the future needs for RIs in Europe based on a clear methodology (Council of the European Union 2004). The first roadmap in 2006 listed 35 RIs (ESFRI 2006). Updates of the roadmap were published in 2008, 2010, 2016, 2018 and 2021 raising the number of included projects to 63 in 2021 (ESFRI 2021). The preparation of the roadmaps in 2006, 2008 and 2010 seemingly followed an ill-defined ad-hoc methodology and received considerable critique (ESFRI 2006, 2016; Hallonsten 2020). It was only in 2014 that ESFRI began to develop a clear methodological approach. It included evaluation criteria and procedures for the submission of roadmap candidates, clarifying that EU member states, associated countries and members of EIROforum are eligible to submit proposals for the roadmaps (ESFRI 2016). Furthermore, from the 2016 roadmap onwards, ESFRI requires the governmental commitment of at least three member states or associated countries—the threshold has formerly been one member state—as minimum eligibility criterion for new roadmap project proposals (ESFRI 2006, 2016).

The Framework Programmes for Research (FP)

Since the implementation of FP1 in 1984, these programs gained in importance for research on a European scale (Banchoff 2002; Flink 2016; Tamtik 2016). Throughout the 1980s and 1990s, the efforts of the Commission in funding research and technology were frequently described as representing a mere industry-oriented “spending policy” (Chou 2012: 1053), highlighting the centrality of the FPs in European research and development activities, but also the lack of a strategic vision. The budgeted sums of the FPs have evolved constantly since 1984, from €3.75 billion for FP1 (1984–1986) to €50 billion for FP7 (2007–2013) and €80 billion FP8/Horizon 2020 (2014–2020) (Banchoff 2002; Flink 2016). With the introduction of ERA and the Lisbon Strategy in 2000, the FPs remained central in European research policy but became linked to a broader strategic approach to govern European research.

FPs contain a sub-set of elements that define and regulate objectives and eligibility criteria, including specific programmes and annual work programmes according to which calls for projects are issued. The negotiation of FPs is based on a complex co-decision procedure among the Commission, the member states and the European Parliament that may involve multiple revision cycles (Flink 2016). The Commission occupies a particularly strong position due to its right to initiate legislation processes which is the case for the FPs (Cini 1996). Negotiations about the FPs also follow a tight time-schedule and failure to reach a compromise would signify a harsh symbolic and political setback for the Commission, the member states and the European Parliament (Flink 2016).

In the 1980s and 1990s, funding for transnational access to research facilities across Europe, termed “major installations”, “large installations” or “large instruments” had already been included in the FPs (Moskovko 2020: 132). FP5 (1998–2002) seemingly was the first occasion in which the notion of RIs was implicitly introduced (European Parliament and Council of the European Union 1999). FP6 explicitly highlighted the emerging European approach to RIs, arguing that “[t]he development of a European approach with regard to research infrastructures, and the carrying out of activities in this area at Union level, can make a significant contribution to boosting European research potential and its exploitation” (European Parliament and Council of the European Union 2002: 17).

From FP7 (2007–2013) onwards, a funding scheme was put in place for RIs with the goal to optimize the use of existing RIs and to support the creation of new RIs within the specific programme Capacities (European Commission 2014; European Parliament and Council of the European Union 2006; Council of the European Union 2006). The budget targeted the preparatory phase or major upgrades of RIs as well as integrating activities and support actions for existing RIs. It also offered a dedicated budget for e-RIs (electronic RIs) that collect, store and share data and knowledge. The budget for RIs in FP7 amounted to €1715 Mio (European Commission 2010). However, this sum only was the agreed outcome of the co-decision procedure and was reduced to around 43% from the Commission’s originally proposed amount of €3961 Mio (European Commission 2005).

The funding schemes from FP7 onwards were tailored to the work done by ESFRI. For instance, the 2007/2008 work programme of the specific programme Capacities stipulated that “[o]nly research infrastructures projects which are included in the 2006 ESFRI Roadmap (…) will be eligible for support at the first call” (European Commission 2007: 12). Similarly, the 2012 work programme stated that project applications should build “primarily upon the work conducted by the European Strategy Forum on Research Infrastructures (ESFRI)” (European Commission 2011: 6, 16). FP6 had also already featured a program for the construction of new research infrastructures, but it was not bound to ESFRI and its roadmap (Stahlecker and Kroll 2013).

The European Research Infrastructure Consortium (ERIC)

ERIC is a specific legal framework, governed under Community law for the operation of European RIs (Ástvaldsson 2022; Council of the European Union 2009, 2013; Moskovko 2020; Moskovko et al. 2019). As of early 2023, 25 RIs have been granted ERIC status (European Commission 2023). ESFRI had initiated the enactment of a novel legal framework for RIs due to the fragmentation of existing frameworks (e.g. consortium agreements, limited liability company) at national and international levels (Moskovko 2020). The creation of a novel framework also seemed important, because ESFRI anticipated future European RIs as being of mainly distributive nature. A legal framework that positioned between an international organization and a national facility was deemed most suitable (Moskovko 2020).

The ERIC framework emerged shortly after research policy became a shared competence. Membership criteria and objectives were laid down in two regulations of the Council in 2009 and 2013 (Council of the European Union 2009, 2013). RIs that are eligible for the ERIC framework need, among other aspects, to be “necessary for the carrying-out of European research programmes and projects (...).” (Council of the European Union 2009: Art. 3). They should furthermore add value to the cohesion of ERA by providing transnational access for researchers or enhancing the mobility of knowledge (Council of the European Union 2009). The 2009 regulation stipulated that “[m]embership of an ERIC should comprise at least three Member States and may include qualified associated countries and third countries other than associated countries as well as specialised intergovernmental organisations” (Council of the European Union 2009: Art. 9), whereby member states have to hold the majority of votes. The 2013 regulation lowered the requirements for membership stipulating that “membership of an ERIC must include a Member State and two other countries that are either Member States or associated countries” (Council of the European Union 2013: Art. 1). The voting procedures were also altered, now stating that “Member States or associated countries shall hold jointly the majority of the voting rights in the assembly of members” (Council of the European Union 2013: Art. 1).

The ERIC framework also carries much symbolic value: The Commission decides to grant ERIC status to RIs based on an extensive application process (Council of the European Union 2009, 2013). ERIC then needs to be added to the full name of the RI and the Commission also provides an identification plate during the initiation ceremony (Moskovko 2020).

Materials and Methods

The European RI policy comprises three different policy instruments: first, the creation of ESFRI in 2002 and its publication of roadmaps since 2006; second, the partial rededication of the FPs; and third, the enactment of ERIC in 2009. Taken together, they constitute the basis of strengthened coordination of RIs in Europe. To obtain a systematic overview over the scope, characteristics and effects of the European RI landscape, we systematically reviewed ESFRI’s roadmaps and the ensemble of RIs that have adopted ERIC status.

We initially identified 74 RIs that were featured on one or more ESFRI roadmaps since 2006 (ESFRI 2006, 2008, 2011, 2016, 2018, 2021) as well as 25 RIs that had adopted the ERIC framework since 2009 (European Commission 2023). For each of these cases, we identified the RI’s self-declared costs of its preparation phase as published on ESFRI’s roadmaps as well as the associated project funding from FP7 (2007–2013), Horizon 2020 (2014–2020) and Horizon Europe (2021-) from the Horizon Dashboard (see Data Availability Statement). Our analysis begins at FP7, because only then a dedicated budget was allocated to support preparatory phases or upgrades of RIs (European Parliament and Council of the European Union 2006 Annex II). Based on this data, we excluded 15 RIs from further analysis, because they were not addressed by all policy instruments under investigation. The excluded cases were either a) never listed on a roadmap, b) removed from roadmap, c) merged with other projects, d) completed or e) they did not receive funding from the infrastructure budget within the FPs but from Euratom (Table 1).

Table 1 RIs Excluded from Analysis (n = 15).
Full size table

We obtained a set of 61 RIs that can be sorted in two categories. The first category (Group 1) consists of 32 RIs that represent those projects that fulfil three criteria. First, they already adopted the ERIC framework or can be characterized as ERIC applicants; second, they appeared on one or several of ESFRI’s roadmaps; and third, they benefited from targeted FP funding for their preparatory phase (Table 2). We consider this group as the triple crown of European RI policy as the projects in this group are intertwined with all three RI policy instruments available to the Commission.

Table 2 Group 1: ERICs and ERIC applicants (n = 32).
Full size table

The second category (Group 2) consists of 29 RIs that were featured on ESFRI’s roadmaps and received FP-funding without adopting the ERIC framework. They operate under diverging legal frameworks and most of them testify of Europe’s historical and organizational experience with intergovernmental ad-hoc integration of Big Science projects (Table 3).

Table 3 Group 2: RIs without ERIC Framework (n = 29).
Full size table

Results

We subsequently analysed both groups for differences and similarities with respect to their membership structure, the extent of their funding from the targeted FP schemes, and their affiliation to scientific fields. Regarding membership structure, both groups exhibit considerable in-group heterogeneity.

Group 1 (Table 2) includes newly formed European consortia that joined forces from already existing yet largely uncoordinated national initiatives. Examples include AnaEE, an experimental platform for ecological and environmental studies, ICOS, a network dedicated to the observation of greenhouse gases or BBMRI, a European approach to integrate national biobank infrastructures (Ciais et al. 2014; Clobert et al. 2018; Vuorio 2017). Group 1 also contains European RIs that existed as coordinated European initiatives long before the emergence of ESFRI and ERIC, such as CESSDA, a coordinative effort of social science data archives founded in the 1970s, or the ESS, a cross-national survey on public opinion and behaviour patterns founded in the 1990s (Lindstrøm and Kropp 2017). Group 1 further contains two newly-built, single-sited large-scale RIs: the European Spallation Source ERIC, a neutron source, and ELI ERIC, a laser infrastructure, hosted by Czech Republic, Hungary and Romania (ESFRI 2021). Both resemble conventional single-sited Big Science projects with massive financial investments and technological groundwork. The members of Group 1 share that they have decided to adopt the ERIC legal framework. The reasons for this seem likely to vary on a case-by-case basis (Ástvaldsson 2022; Gaskell et al. 2013; Lindstrøm and Kropp 2017; Moskovko et al. 2019). Lindstrøm, and Kropp (2017) conclude for the ESS that ERIC escalates beyond a mere legal framework but provides the project with organizational and financial stability by aligning asynchronous national funding cycles.

Group 2 (Table 3) includes projects that are associated with members of EIROforum and other intergovernmental mega-projects as well as projects that have not yet decided on their future legal framework. The first subset includes four upgrade projects of major single-sited Big Science projects that are High-Luminosity Large Hadron Collider (HL-LHC) at CERN, ILL 20/20 at ILL as well as ESRFUP and ESRF Extremely Brilliant Source at ESRF. It also includes the European XFEL, a free-electron laser, and FAIR, an antiproton-ion accelerator, that are two Big Science projects that were recently built or are under construction as well as the preparatory works for SKA, a radio telescope project, and the construction of ELT, another large-scale telescope, under the auspices of the European Southern Observatory. While all these facilities qualify, from a legal point of view, for the ERIC framework, the upgrade projects already rely on sustainable legal frameworks and long-term funding through intergovernmental agreements. FAIR, European XFEL and SKA, for their part, receive strong contributions from non-European partners: Russia is the second biggest shareholder in FAIR and European XFEL, and Australia and South Africa are host countries of SKA (Cramer 2020; Walker and Chinigò, 2018). Here, it can be assumed that a link to a legal framework at the EU level is rather undesirable or a hindrance.

The second subset of Group 2 (Table 3) consists of various facilities that either have chosen national legal frameworks such as the German GmbH (limited liability company) or the Belgian AISBL (international non-profit association), or whose legal framework is still undecided according to the ESFRI roadmap (ESFRI 2021). The undecided cases represent a mixture of RIs, ranging from projects in astronomy (EST or ET), to social science (GGP, GUIDE), data science and digital research (SLICES, SoBigData++), energy production (MARINERG-i) or health and food research (EIRENE RI, EMPHASIS, EU-IBISBA). While it is difficult to find a common denominator beyond the fact that they were represented on the roadmaps and received FP funding, it cannot be ruled out that some of these RIs may apply for ERIC status in the future.

Regarding the amount of FP funding received, the results indicate that Group 1 and Group 2 differ in several respects from each other (Table 4). The median amount of funding received by the projects in Group 1 for their preparatory phase is almost €4 million. This ranged from funding for the European Spallation Source ERIC, a neutron research facility, that obtained only 2% of its preparation costs from the FP to EU-Solaris ERIC, a distributed RI for solar power and thermal energy, that received 809% (sic!) of its declared preparatory budget from FP7. The median amount of funding in Group 2 is slightly higher and amounts to €4.25 million. This ranged from funding for EuPRAXIA, a European consortium that developed a plasma research accelerator, that received only 2% of its declared preparation phase costs through FP funding to the upgrade of the ILL, a neutron research reactor, that received 330% (sic!) of its declared preparation costs from the FP. The total funding amounts of EU-SOLARIS and ILL seem surprisingly high, but so far, there is no reason to believe that it is erroneous data. Differences between Group 1 and Group 2 were not statistically significant (two-tailed Wilcoxon rank sum test; Group 1 (n1 = 31), Group 2 (n2 = 28), p =.72).

Table 4 Median of absolute and relative FP funding for the preparatory phase of RIs in Group 1 and Group 2. Sources: Calculation based on data available in the RI ESFRI raw data set (https://doi.org/10.6084/m9.figshare.22378384)
Full size table

The funding situation changes when analysed through relative FP-funding as a share of full costs. In Group 1, the median of relative costs amounts to 95%. This means that FP funding covered almost all costs (95%) of the preparatory phase for at least half of the RIs in Group 1. In contrast, the median value of the same variable in Group 2 shows that the share of funding was much lower (40%). This means that half of the RIs in Group 2 received less than half (40%) of the costs from FP funding to cover the costs of their preparatory phases. In combination with the information on the absolute amount of preparatory costs in both groups, we can see that members of Group 2 compensate their costs through alternative funding channels. Non-parametric statistical testing indicates that the inter-group differences in the share of FP-funding for the preparatory phase are significant (two-tailed Wilcoxon rank sum test; Group 1 (n1 = 28) and Group 2 (n2 = 23); p = .02; EU-SOLARIS was excluded as an outlier) and indicate a medium effect size (r=.33).

To assess the differences and similarities between the two groups regarding their affiliations to scientific fields, we used ESFRI’s taxonomy that divides RIs into six domains (Figure 1). The majority of RIs in Group 1 are associated with the domain of Health and Food; RIs in the domains of Environment and Social and Cultural Innovation following right behind. The affiliations of Group 2 show a different picture with a comparatively large number of RIs in the domains of Physical Sciences and Engineering and Data, Computing and Digital Research Infrastructures respectively. These differences are statistically significant (Fisher’s exact test, two-tailed, p = .0001).

Fig. 1
figure 1

Sources: ESFRI 2006, 2008, 2011, 2016, 2021 and authors’ own compilation of data of the RI ESFRI Roadmap dataset (https://doi.org/10.6084/m9.figshare.22360552)

Distribution of RIs Categorized by ESFRI Domain and Group.

Full size image

Discussion

The results confirm our hypothesis that a new layer of European RIs has emerged that institutionally extends beyond the existing intergovernmental Big Science projects. We interpret our empirical observations as the result of the interplay between the policy instruments available to the Commission, the interests and capabilities of incumbents in the field of RIs and a novel understanding of European integration in which coordination is key. European RI policy then developed through, first, the strategic incorporation of incumbents through new modes of coordination; second, the Commission’s discursive emphasis of joint responsibility at the EU level, thus claiming its own accountability and mobilizing the rules laid down in the subsidiarity principle to its advantage; third, the incentivization of conformity to the Commission’s own policy agenda through generous financial schemes and fourth, the implementation of tailor-made legislation. These four elements are embedded at various levels of the EU's governance system, namely: the social integration of actors, the discursive level of policy-making, the allocation of funding flows as well as the juridification of research initiatives that is the prevalence of legal frameworks or regulations governing research activities. These levels unfold in parallel and are interwoven.

First, ESFRI was initially promoted as a means of bottom-up policy coordination with the goal to foster stimulation for emerging or already existing RIs. The establishment of ESFRI and its use of OMC seems to testify of a reciprocal coordinative approach of the Commission to build up legitimacy. Here, OMC escalates beyond a mere instrument of bottom-up and non-binding policy learning but becomes a potentially powerful tool through the intervention of the Commission as a supranational body (Kaiser and Prange 2002). The Commission did so by engaging, first and foremost, with a limited number of “privileged ‘insiders’” (Edler and James 2015: 1259) and “organized interest groups” (Hensell 2022: 153). In the case of ESFRI, these included representatives of the member states and associated countries as well as delegates from EIROforum (see also Hensell 2022).

In the early 2000s, the Commission encountered a long-established landscape of Big Science institutions. The Commission’s agency in the field of research policy was further limited by the subsidiarity principle. However, the Commission took a two-track approach. On the one hand, it acted as a policy entrepreneur, keen to advance its vision of a European approach to RIs, attracting the attention of unfolding RIs (Garud et al. 2007). On the other hand, ESFRI can also be interpreted as a coordinative concession to the existing landscape of powerful stakeholders from major national funding agencies or EIROforum. ESFRI’s position as a mediator between existing institutions and new consortia can also be plausibly read from the balanced distribution of the cases between Group 1 (32) and Group 2 (29). Mature RIs—such EIROforum’s members—have secured budgets and stable legal frameworks, provided through governmental commitments. They have never been dependent on the Commission’s financial incentives. An important indication for the powerful role of EIROforum is that nominations for the ESFRI roadmap can be made equally by the Commission, the member states and EIROforum. Co-opting these important and established actors has thus been paramount to legitimize the Commission’s own RI policy agenda (see also Kaiser and Prange 2002). However, several emerging RIs that eventually adopted the ERIC framework highlighted that their establishment has been triggered by the Commission’s emerging political interest. Examples include SHARE, a survey on social, economic and environmental policies, EATRIS, a network of biomedical translation centres across Europe and CLARIN, a distributed infrastructure uniting universities, archives and libraries (Blümel 2018; Börsch-Supan et al. 2013; Fišer and Witt 2022; Kim et al. 2020).

Second, the Commission operated on a discursive level within the limits set by the principle of subsidiarity. To assert a shaping role in RI policy, it was necessary to show, on the one hand, that individual states lack the capacity to implement appropriate policies and, on the other hand, for the Commission to demonstrate tangible added value through its actions. This was primarily accomplished by aligning and re-aligning the purpose of RI policy to emerging strategic discourses of EU’s policymakers and mobilizing them as windows of opportunity. The Commission began to discursively link its vision of a better coordinated European RI landscape to EU’s Grand Challenges. The concept came into fashion in Europe in the mid-2000s and was taken up by the Commission at the same time and spelled out in more detail in Horizon 2020 (Burgelman et al. 2014; European Commission 2008). Here, the Commission exploited the argument that “challenges affect and address everyone, neither scientific-disciplinary nor national-political pettiness can master them alone or deny them—they call for post-disciplinary and post-national STI policies” (Flink and Kaldewey 2018: 18). At around the same time, the Commission began to call for larger investments into science and research to keep up pace with the United States or Japan. It began to consider an increasingly bold and forward-moving role in RI policy, arguing that

“The Community can and should play a catalysing and leveraging role by helping to ensure wider and more efficient access to, and use of, the infrastructures existing in the different Member States (…)” (European Parliament and Council of the European Union 2006: 31).

In the aftermath of the financial crisis in 2008/2009, RI policy discourses shifted “to a more pragmatic recent focus on ‘sustainable research infrastructures’” (Ulnicane 2020: 91) with the “ambition to minimize redundancies and increase efficiency within this system” (Bolliger and Griffiths 2020: 114). A similar example is the continued discursive emphasis on the added value that the Commission’s RI policy would bring for the EU, its member states, societies and economies. ERICs, for instance, need to “represent[s] an added value in the strengthening and structuring of the European Research Area” (European Commission 2015: 6). Similarly, the 2006 ESFRI roadmap emphasized that its collection of RIs of pan-European interest might help policymakers in assessing the “added value of capacity building and the socio-economic returns of RIs” (ESFRI 2006: 86). These dynamics point to a broader tendency in EU policy and funding: For the justification of new policy strategies, the EU frequently resorts to the added value this policy will bring to the member states and the public which attributes competence and responsibility to the EU (Flink and Kaldewey 2018; Rip and Voss 2013; Wanzenböck and Frenken 2020).

The connection of the Commission’s RI policy to the changing discursive agenda of supranational research policy also is evident in the categorization of the various RIs into different domains (see Figure 1). It is no coincidence that ESFRI’s domains mirror the terminology of the Grand Challenges discourse. On the one hand, the distribution of RIs across the different domains shows that the triple crown of Group 1 is more strongly located in the domains associated with these Grand Challenges and thus closer to the discourse that was significantly propelled by the Commission. Members of Group 2, on the other hand, are more likely to be situated in traditional fields of Big Science and basic research. ESFRI's rhetorical development of the classification scheme of domains apparently developed traction for the design and preparation of new RIs on the roadmaps.

Third, the Commission did not only claim competence at the discursive level, but also underpinned this with financial incentives tailored to the approach of ESFRI and its roadmaps. The targeted funding scheme in FP7 was a first step for the Commission to obtain influences on national RI policies beyond ESFRI’s OMC mechanism and to provide incentivization of conformity to its own policy agenda through generous financial means. However, at no point did the Commission exceed its competences. Firstly, the preparation of the FPs in coordination with the member states was already part of its tasks. Secondly, it could be convincingly argued during the FP negotiations that RIs of supra-regional significance warranted funding. The scope of applicants to the specific RI calls in the FPs was severely limited, based on whether projects have been ex-ante included in ESFRI’s roadmaps. This role of ESFRI was further legitimized by its inscription in the joint decision of Council and Parliament on the design of FP7 (European Commission 2006). The Commission hence gradually developed into the central sponsor of the preparatory phases for the triple crown, namely the RIs of Group 1.

Fourth, the European RI policy further manifested through the enactment of ERIC. In contrast to ESFRI, ERIC did not emerge from the OMC, but has been the result of high-level negotiations and consensus among the Commission, the Parliament and the Council (Chou 2012). ERIC is a legally binding instrument that aligns the governance of RIs—that remain to be funded by national governments—to EU law. It further testifies of a tight coupling between the purpose of the RI and the strategic goals of EU research policy, providing the Commission with regulatory agency over the purpose of the RI (Ástvaldsson 2022; Moskovko et al. 2019). Ástvaldsson (2022: 60-61, emphasis in original) argues that “the ERIC is not a legal form where members enjoy full autonomy as to the purpose of their collaboration.” Rather, “an ERIC shall pursue EU objectives in the interest of the EU” (Ástvaldsson 2022: 61). This is a strong assessment, but it illustrates how tightly the ERIC framework is coupled with the ambitions of RI policy.

In summary, the advent of the Commission’s role in shaping the landscape of European RIs is bound to the complex interaction of different policy instruments and the various levels of the EU governance framework. Due to the fundamental treaties (TEU and TFEU), the Commission could not act unilaterally within research policy. Instead, it had to rely on the cooperation of the member states and existing interest groups. The establishment of ESFRI can be understood as a manifestation of this institutional setting. The development of European RI policy also bases on a novel understanding of European integration that relies on new forms of coordination among member states and supranational bodies. European integration hereby lacks supranationalisation of national competences but strengthens the capabilities of intergovernmental actors, such as the Council (Fabbrini and Puetter 2016; Kaiser and Prange 2002). This applies, on the one hand, to ESFRI’s mandates that are based on Council decisions. It also matters, on the other hand, for the dedicated RI budgets within the FPs as well as the enactment of the ERIC framework that both required high-level consensus and co-decision among the Commission and intergovernmental actors, such as the Council or the member states’ governments. All these aspects of RI policy highlight that “[i]n the absence of the possibility or willingness to delegate the enforcement of common policy objectives to supranational bodies, member state governments need to be constantly brought in line in order to enable the EU to act” (Fabbrini and Puetter 2016: 486) which “reinforces policy and political interdependencies between the member states and existing EU institutions” (Fabbrini and Puetter 2016: 484).

The discursive emphasis on the added value of a coordinated RI policy for addressing global challenges, fostering regional cooperation and creating economic benefits can be understood in light of these interdependencies. The Commission needed to justify its mandate for a joint RI policy in order to gain leverage in negotiations concerning the mandate of ESFRI, the configuration of the FPs, and the implementation of the ERIC framework. Even after a dedicated RI-budget was made available in FP7, ESFRI’s role remained central as funded projects almost always were part of the ESFRI roadmap. ESFRI hence took a gate keeping position, illustrating the close connection between the integration of stakeholders and allocation of funding. The distribution of FP funding between the two groups under study suggests a relatively even allocation of financial resources between ERICs and other consortia and international organizations (see Table 4). This allocation could be interpreted as the Commission's effort to strike a balance between various stakeholders of ESFRI to advance its agenda.

Despite the institutional constraints and the necessity to coordinate with member states and interest groups, it is striking that the Commission managed to seize a particularly influential role in RI policy. The Commission emerged as the central interface linking the diverse instruments and governing bodies. This influence is apparent through its adoption of the OMC as ESFRI's operational method, substantial sponsorship of the preparatory phases for a particular group of European RIs, and its role in awarding and overseeing the ERIC framework.

Conclusion

Since the early 2000s, the Commission began to create a strategy-led policy approach to European RIs to optimize their coordination within the emerging ERA. The Commission’s vision and understanding escalated beyond RIs as mere providers of scientific services and research resources. But it also considered the ability of RIs to connect Europe and its research landscape with lasting effects on Europe as a transnational space of science and technology (Schipper and Schot 2011).

Taken together, the vague definition of RIs and the associated political interest can be regarded to have shaped policy directions, because it allowed the Commission to advance and secure its own vision of what RIs are and which of them to prioritize and support. The Commission’s push towards a better coordinated European RI landscape in combination with the incentive-based realization of a targeted FP funding scheme, allowed it to bring the topic to the European level, thus emphasizing its own responsibility. The ensemble of coordinative, incentive-based, and legal policy instruments strengthened the coordination of RIs on a European scale and eventually led to the establishment of a separate layer of more EU-oriented organizations, partly co-opting the traditional European RI landscape of intergovernmental Big Science organizations. These developments have monopolized political attention, turning the Commission’s RI policy into a normative starting point for national policymakers and the scientific communities. While there is no formal and binding connection between the three investigated policy instruments, our findings that RIs receiving extensive FP funding for their preparatory phases are more likely to pursue ERIC status suggest the presence of broader informal mechanisms linking them (see also Stahlecker and Kroll 2013).

Our analysis is limited insofar as we consider the Commission as a uniform actor. Alternative approaches point to the fact that different commissioners and mid-ranking administrators may pursue different goals (Edler and James 2015; Bürgin 2017). Moreover, the group of 15 excluded RIs indicates that there is a considerable residue of RIs that do not directly fit into a blueprint-like policy. For these cases, explanations must be found to understand, for instance, why and how CERIC-ERIC and JIV-ERIC—that were never listed of ESFRI’s roadmap—decided to apply for ERIC status. While recent scholarly research began to unpack such policy complexities and intra-organizational dynamics to reveal interdependencies that affect how emerging policies are strategically pursued and implemented with a multi-governance framework, further research on the case of RI policy needs to be done.