Governing through consensus? The European Semester, soft power and education governance in the EU

The European Semester and the framing of education policy

Education has been somewhat absent in the existing literature on the ES. The instrument’s educational content has been broadly identified as part of the ‘social dimension’ of EU governance, perhaps reflecting the importance of streamlining key policy areas in Europe today to help in times of crisis. Thus far recent scholarship has approached the instrument mainly from a legal perspective (Bekker, 2015), whereas other scholarly work centres on the CSRs and their implementation in national contexts (e.g. Zuleeg, 2015), or on the role national parliaments play under the ES (e.g. Hallerberg et al., 2018). Several authors have considered the impact of the ES on health policy (e.g. Azzopardi-Muscat et al., 2015) and social policy (Copeland and Daly, 2018; Verdun and Zeitlin, 2018). Yet, some observations have focussed on scrutinising education policy in the ES (e.g. Antunes, 2016; Stevenson et al., 2017), and its connectedness to other policy instruments, including the European Social Funds (e.g. Sorensen et al., 2021). This article provides complementary views on education policy to supplement these existing reviews on the instrument.

The first European Semester cycle took place in 2011 and has completed in total 10 rounds of country-specific recommendation during EU 2020 (see Figure 1). The ES has changed its shape and form over the course of its existence and during the period of this research: its implementation cycle has evolved and the policy framework expanded. Although the ES is presented as independent, it also draws on and possibly shapes the lives of the other interconnected but separate instruments as illustrated in the EC’s official ES timeline (see Figure 1).

Figure 1. Timeline for the European Semester.

Source: EC (2020c). The European Semester and the United Nations Sustainable Development Goals.

Figure 1 depicts how the instrument is driven by the complex set of standardisation and comparison of performance (i.e. the benchmarking for EU 2020) that assume convergence. The ES functions not merely as policy text but it also provides a space to cultivate a bond through governing activity that keeps the instrument alive (Freeman, 2017a). Following Latour (1990), the negotiation process sets the ES ‘in motion’: if policy actors can cohere around common policy terms, the ES will ‘move’. Therefore, the process of co-constructing country-specific recommendations provides legitimacy which builds convergence – allowing soft power to do the governing work. I use the country-specific recommendations and the governing work surrounding their construction as an example to investigate the interactions which enable consensus-building ‘on paper’. The CSRs as policy text possess problems, purposes and agencies held by actors in different social worlds, which may become converted and incorporated into the instrument (Carvalho, 2014). That is, policy instruments are socially constructed and connected to policy targets and goals that draw on data and indicators which are also based on negotiated consensus between various policy actors. For example, the EU countries have translated the EU-level (EU 2020) targets into national targets, each year reporting on their actions to meet these targets via the ES policy cycle. This process illustrates a practice of translation (Freeman, 2009) that takes place within and through various ES activities, shaping knowledge production and use.

Translation as a term is assigned different meanings in the scholarly literature. I draw on sociological approaches that consider translation in connection with policy and practice (e.g. Freeman, 2017b). Therefore, ‘translation’ is here understood as meaning-making of social interaction in policy practice. It allows policy actors to adhere with different worldviews to common objects of interest because ‘translation takes place within discourses in which some kinds of association or translation are legitimated and authorised just as others are excluded or denied’ (Freeman, 2009: 434). I focus on the work of translation; that is, the tools and practices it cultivates and explore them in the meeting-making and drafting of ES policy. I attempt to ‘study up’ how the construction of CSRs generates governing practises in which translation operates.

I draw on scholarly work that place emphasis on processes that replace traditional governing forms, laws and regulation (e.g. Lawn and Grek, 2012; Ozga et al., 2011). Grounded on this perspective, I apply a socio-constructivist enquiry that allows me to look beyond the institutional hierarchies and the rules of representation, focussing on social relationships through discourse (Fairclough and Wodak, 2008). I examine the ES as open to meditation (Jacobsson, 2006) and translation rather than through the ‘top down’ or ‘bottom-up’ actions of institutions and structures. In lieu of scrutinising evidence of policy capacity or policy implications, I focus on the operation of soft mechanisms in constructing, mobilising and monitoring ES policy content.

The formal EU framework does not acknowledge the ways in which policy actors use technical knowledge, such as the objectives of the EU 2020, to build, negotiate and translate policy. I suggest that this development can be comprehended not only as a method but also as a political solution achieved by coordinating and transforming national governing practices to correspond to European and international indicators (Nóvoa and Yariv-Mashal, 2003). Translation becomes the means to develop governing practices cultivated in the ES as it provides capacity to foster peer learning and co-construction of policy, appearing to dissolve or naturalise political disposition. The translation work is then an interaction in which actors are also defined by the translation process itself and by other actors (Freeman, 2009). In this frame, translation is a means but also an end product of policymaking (Freeman, 2017a).

Research context and design

I understand discourses as a medium of soft power. They co-opt actors into shared meanings and values held by specific groups or institutions, projecting an imaginary of a world that is attractive to those actors. Thus, my empirical approach follows the emphasis on understanding language as a form of social practice. I examine the interaction between texts and talk in the production of country-specific recommendations, investigating how they, as policy texts, create common meanings and rhetoric ‘containing the problems, ideas, and purposes which move between actors and locations and are replicated at different levels of the governing practices’ (Freeman, 2009: 431). Indeed, CSR policy texts connect different forms of agency in different settings (Stritzel, 2017). I approach policy construction as a process of learning and socialisation (Grek, 2017) in which a common discourse on a certain topic or field is created. The interview and documentary data presented here illuminate how the meaning-making of policy purpose is processed and synthesised into country-specific indicators and benchmarks.

In framing my methodological enquiry, I followed Grek (2015: 38) in conceptualising policy as a sharing and cooperative activity, focussing on the social dimension of policy. I identified discursive key concepts and terms emerging from policy text: They focussed on ‘economising’ themes – efficiency, competitiveness, performance and skills. These key concepts informed the data analysis and coding procedures. I used a semi-structured interview format in my sampling design. The interviewees were selected based on gained access but also on the networks they belonged to. The primary interviewees also provided me with contacts for other policy actors within their networks, therefore having a ‘snowball’ effect on the original sampling during my fieldwork.

The content and discourse analysis² focussed primarily on country-specific recommendations in 2011–2016 coupled with selected interviews with key actors involved in the construction of the instrument. The analysis was conducted on the final versions of the CSRs which were approved by the European Council (hereafter ‘Council’). These ‘final’ versions of the recommendations were preceded by several drafting and re-drafting stages that involved various actors in the EC, Council, EP and member states. In total 165³ education-related CSRs were analysed alongside other policy documents related to the ES, such as National Reform Programmes⁴ (NRP) and country reports,⁵ which supplied complementary knowledge about benchmarking in the ES. The thematic interviews were conducted during 2015–2017 with 15 policy actors consisting of MEPs, high-ranking officials, academic experts and national representatives possessing specialised knowledge of benchmarking in the ES. Themes for interview discussions derived from the discourse analysis of aforementioned ES policy documents selected CSRs. The data also encompassed publicly available meeting documents, email correspondence with the interviewees and my personal fieldwork notes – all of which contextualised and supported my interview data.

Grounded on the research described above, the documentary analysis was based on the CSR policy texts that addressed education(al) elements. The mapping of education-policy related CSRs would itself be a topic of a research paper as there are various ways of defining the particular country-specific recommendations that are regarded as ‘education’. The selection of CSRs was scrutinised by examining all Council’s recommendations between 2011 and 2016. Instead of identifying educational content as we traditionally understand ‘education’ within the EU official framework, I refined the discourses that constructed education-related recommendations and referred to them as ‘borderline recommendations’. I have reported this discourse analysis in more detail elsewhere (see Eeva, 2019). These borderline recommendations are identified in the next section but, in short, they seem to spill over into other policy fields – namely economic, employment and social.

The content analysis of the education-related country-specific recommendations between 2011 and 2016 contributes to the existing literature on the ES by (a) illustrating the changing landscape of education governance and (b) identifying the implications for education policymaking in the European Union. Inevitably, CSRs can be read and data presented in alternative ways. The common practice, especially in the literature of political sociology, is to view education-related recommendations as part of the social and employment objectives (e.g. Zeitlin and Vanhercke, 2018) or justified by the ‘EU convention’ (Copeland and Daly, 2018: 1004). With a focus on the relationship between the EU institutions and member states, I address the interactions and processes in the production of ES policy rather than specific policy outcomes. That is, how ‘education’ becomes encapsulated in the CSRs and is conceptualised in the policymaking of the ES.

Co-constructing the European Semester

I present here how education is conceptualised in the country-specific recommendations and then discuss the interaction that defines the policy text and constitutes representation. On this basis, I look at (1) CSRs policy formulation as it relates to education; (2) the interplay between various actors in the construction of the instrument; and, (3) practices cultivated in the governing work of the ES. First, the analysis suggests that education policy has increasingly become associated with employment and economic policies. I argue that the ES operates discursively, through elements of soft power expressed in the country-specific recommendations, suggesting an overarching policy discourse of socio-economising education.

Thus, this paper contributes to wider scholarly debate on the narrative of economisation of education. It connects to the prominent discourse of education performance that is often indicated in data, numbers and standards reflecting the economisation of education (see e.g. Carney, 2012; Ozga, 2008; Steiner-Khamsi, 2004). The meta-narratives emerging from the overarching policy discourse are illustrated in Table 1: the Right skills for the right jobs; Investment in human capital; Quality assurance; and Social cohesion. They indicate how education policymaking is now in a close relationship with economic policy. The knowledge production in the ES policy framework is realised through demands of the knowledge economy effectuated by neoliberal approaches to the economy and governance that put the emphasis on investing in human capital in an ‘audit state’ (Ranson, 2003: 466).

Table 1. Discourse analysis of CSRs: Socio-economising discourse and its meta-discourses (‘borderline recommendations’).

The analysis of the interview data reveals that a consensus on the role of education as a tool for economic recovery enables policy actors to create a governing discourse based on the ES indicators and country-specific recommendations which provide accountability for policymakers. In this context, I argue that CSRs provide a platform where policy can take shape: it allows conflicting ideas and interests to be encapsulated in a consensual way into policy text. In the Commission language, the CSRs are considered as the main tool for evaluating the implementation of the ES and for collecting data to help monitor economic developments in EU member states through common indicators and benchmarks (EC, 2020d). This also seems to reflect and draw comparison to the global economy.

The analysis of the country-specific recommendations shows that the discourse emerging from the policy documents relates to the European-wide, common benchmarks underpinning the EU 2020 strategy. Discourses, such as the socio-economisation of education, provide a way to see how knowledge is operationalised into shared meanings and common understandings as is exemplified by the ‘translation’ or ‘naturalisation’ of the EU 2020 benchmarks into national contexts. These indicators, then, become governing resources enabling policy actors to steer policy through consensus about best practices, a form of soft power. Discourses operate as soft power when they are grounded in information that is presented as evidence-based and reliable (Nye, 2008: 99). Specific knowledge forms such as comparative performance data, often expressed in benchmarks and indicators, provide such legitimacy when they are encapsulated in discourses (Stone, 2013). Moreover, discourses produced by influential sources or actors are attractive (Nye, 2008: 100); their credibility is often acquired in social interaction and through social practices.

Second, soft power takes place through the brokering work in the EU-national interaction resulting in policy reforms which derive from meditative activities (Jacobsson, 2006). This quotation from the European Commission illustrates the capacity assigned for the ES to manage the national-EU interplay and brokering work:

The European Semester is presented here as ‘treatment’ for the crisis in the economy in Europe and beyond, following global prescriptions in pursuit of growth. Before the economic crisis, fiscal coordination at EU level was limited: economic and budgetary policy planning occurred at the national level (EC, 2010) despite the attempts to coordinate fiscal policies through, for instance, the creation of the Economic and Monetary Union (EMU). After the economic and financial crisis there was a need to extend the areas of surveillance and coordination to broader macroeconomic policies (EC, 2015). The ES was established as part of this wider reform of the EU’s economic governance and the EU 2020 strategy.

The ES is a site of translation work as well as itself the practice or a technology in which translation can take place. Indeed, public policy instruments are not inactive and purely technical, they may define terms and produce a change through technicalities without necessarily agreeing on the aims or principles of reform (Lascoumes and Le Gales, 2007: 16). These technicalities may allow the ES to disconnect from political goals and enable persuasive interaction between the national and European (Borrás and Jacobsson, 2004). The conditions of translation are constituted on paper in country-specific recommendations as policy texts. Consensus building is integral to the translation work because it enables transformation of the presentation of (national) policy agendas, for example, into shared meanings such as the CSRs. Such policy engagement and exchange can be seen in the shaping of national and European policy into practices that enable comparison (i.e. CSRs) through compatible standard-setting (i.e. national EU 2020 targets). This process, cultivated in translation, is key in making visible the discursive aspects which define the space in which European education policymaking operates.

Third, I considered the ways policy actors responded to the governing work of the country-specific recommendations outside the formal (institutional) ES framework. Exploring the unofficial interaction provides an interesting enquiry because the work of translation often remains hidden in the policy process (e.g. Freeman, 2009). The EC and national governments have developed unofficial practices to maintain dialogue in the production of CSR policy. This interaction is often omitted in literature on international relations and political science, but sociological approaches (e.g. Ozga et al., 2011) recognise its characteristics and the governing practices which are unofficial by nature. Following Wodak (2015), the interplay between frontstage and backstage deliberations can shed light on understanding how policy moves in CSRs and how consensus is sought as a response to the Commission’s official ES cycle (see Figure 1). The official cycle represents, in Wodak’s (2015) words, the frontstage politics whereas governing practices in the ‘backstage politics’ appear to be drawn from the practices of the politics du couloir, ‘conversation in the corridors’ (p. 30).

The analysis suggests that CSR policy is established through stages of translation that occur, in the context of the ‘official’ ES timeline, through informal or backstage policymaking practices:

1. informal consultation of country reports by member states prior to their publication;

2. country team visits; and

3. policy engagement in the Commission’s ES national offices.

This consensus building activity and interplay between the member states and the EC is, in the official ‘Brussels speak’, indicated as ‘dialogue on the proposal for CSRs’ (see Figure 1). The interplay takes place between the frontstage and backstage deliberations, employing various levels of knowledge production that illuminate the ‘life span of the instrument’ (Eeva, 2019). In speaking about the increasing consultation with member states, this interviewee from the EC illustrates the construction process of CSRs:

The work of translation takes form in the unofficial meeting-making and drafting of documents, and it is ‘achieved on paper’ generating consensus (Freeman and Maybin, 2011: 166). The excerpt illustrates the provision for national governments to respond through informal engagement to the drafting of the country reports before they are published in February (see Figure 1). In this process, the construction of the CSRs becomes itself a means of realisation and implementation (Freeman, 2017a: 147): the depiction of the ‘backstage’ elements in the cycle of the ES indicate the rather invisible stages of the work of translation in governing between member states and the Commission.

I suggest that the country-specific recommendations are a platform for actors to organise their preferences, such as political interests, into a manageable form. Country-specific recommendations have become a site where the EC and national contexts are ‘performed’ to seek policy consensus. Following this line of analysis, consensus enables the work of soft power by generating policy solutions in CSRs which are then employed, for example, to frame policy purposes: member states’ response to the ES suggests that the ES is used as an effective policy strategy to support agenda setting at home. This governing work takes form in the practice of social interaction (e.g. consultation) in which policy decisions are defined and encapsulated in CSRs and other ES related policy papers, namely country reports and NRPs. The following excerpts exemplify the process of unofficial deliberation in policy engagement against which the official account becomes redefined:

It is this policy engagement where narratives and different forms of agencies and their social and political effects are connected and reconfigured; that is, the official dialogue is coupled with ‘backstage’ policy practices. The analysis of interviews shows that the consultation on country reports is followed by visits conducted by Commission delegations to members states once the reports are published. Issues covered in the country reports are often established well prior to the publication of these reports as they have been discussed in various arenas at the European level and through unofficial dialogue, as this interviewee suggests:

In this process, various actors – policymakers, experts, officials, civil servants, academic experts and other stakeholders – at European, national, transnational, local and regional levels, are engaged in the meeting-making in which the country-specific recommendations get constructed. The persuasive element in these activities becomes more apparent through the meeting documents, agendas and presentations that are publicly available online. This excerpt discusses how talk becomes constituted on paper and encapsulated both in CSRs and country reports:

This interaction illustrates the governing capacity of soft power: the ability to make policy visible through discourse where policy text is recontextualised and adjusted to the policy needs of regional, national and European levels. The implications of this interplay between Brussels and the national are then shown in policy purpose, demonstrating how meaning-making is expressed and recreated in text. I argue that both the EC and Council employ a range of activities within their structures for policy work arranged between various stages and groups of experts combining the ‘national’ and ‘European’ governing work. This interviewee further illuminates the stages of negotiation within ‘the ES cycle’ in the Council:

This type of interaction is often presented as an embedded procedure within the institutional framework and considered as part of the official and frontstage policymaking. However, the analysis indicates how meeting and negotiation becomes a governing device which operates through persuasion and consensus building in the European-national interaction. As Freeman (2017a: 143) puts it: ‘translation is not merely pronounced but it is a product of cultivation and choreography’. To identify common European meanings, actors address shared (European) problems through these meditative activities by aligned benchmarking. The emerging policy discourse of socio-economising education is a good example of such discourse as it has gained prominence in the governing of education and has also become normalised in the national contexts. The similitude is reinforced by synthesising and presenting online the data on CSRs and national reports from each member state. For example, the Commission-run online platform on the ES not only presents the ES documents for each member state but also indicates nation-specific statistics and indicators in relation to Europe 2020 targets.

To conclude, the analysis suggests that governing happens in the work of translation, in its tools and practices making the ES move and change as it is being constructed and recontextualised. The meeting-making shapes the ES as much it is, itself, the technology that creates and maintains the instrument. The aim then appears to be providing a platform for the policy actors, people, policies and instruments it contains to mobilise ‘common ways of thinking and working’ (Freeman, 2017a: 140).

Governing through consensus: New governing practices

This section discusses the new governing practices that emerged in the education sector as a response to the ES. I examine how these new governing practices make policy move. The repertoire of soft power in the ES allowed for interpretation by various policy actors involved, enabling them to generate meanings, such as the socio-economising discourse, that were useful for specific policy purpose and action. I focus here on two examples of governing work in the Council regarding the ES and country-specific recommendations: (a) Education Committee’s ‘joint letter’ to the Employment Committee and (b) the country-supportive recommendations employed by the Education Committee. Each practice developed into unofficial fora for the exchange of views on the ES and education-related recommendations as my analysis of interviews shows. These practices were generated as part of the ‘peer review’ meetings with Directors General from Ministries of Education in which CSRs were discussed among the EC, Council and member states (policy actor 03 and policy actor 08). These policy arrangements in the domain of education show how discursive practices ‘move’ within the space where European meanings are crafted or translated by policy actors involved in the process.

First, the joint letter of Council’s Education Committee was a response to the heightened role of education in the ES and CSRs, and in particular, a reaction to the part assigned to the Employment Committee in the Council. In fact, the Employment Committee was given the main responsibility for education policy as it related to the ES from the launch of the instrument. This was agreed in the six pack negotiations (in words of policy actor 02). Consequently, the Education Committee, which consists of representatives from national ministries and the EC secretariat, was left without a formal platform to engage with education-related country-specific recommendations. The education sector (in words of policy actor 02) ‘reacted’ to this development by addressing the education recommendations in a letter to the Employment Committee. According to the interviewees’ analysis, the function of the letter was to bring forward the Education Committee’s views on the Employment guidelines⁷ that the Council, together with the EC, communicates to the member states. The joint letter, drafted in consensus among the Education Committee members, illuminates how the socio-economising discourse was being negotiated. It exemplifies how the ‘skills’ discourse was conceptualised and circulated in the relations between texts and interactions (Fairclough, 2010). This policy actor speaks about the new practice that was established during the Danish Presidency in 2012:

This excerpt depicts how the pressure to reach European standards, indicated through the ES and expressed in CSRs, allows flexibility in the formal powers and competence issues. According to the Treaty on the functioning of the European Union (TFEU), the Council is to ‘adopt employment guidelines, specifying that they must be consistent with the broad economic policy guidelines’ (EC, 2017). These guidelines function as a basis for drafting CSRs and they also contain elements of education policy: The sixth guideline concerns skills development to ‘enhance labour supply’ in terms of access to employment, skills and competences. The following excerpt captures the consensus building practice cultivated by the joint letter:

The work of translation defines the relationship between governing and knowledge. ‘Skills’ becomes a reference point against which policy actors assign meanings of concepts such as ‘lifelong learning’, ‘adult education’ and ‘work-based learning’ as they relate to respective national policy frameworks. The attempt to construct a shared understanding of what ‘works’ best to drive the policy is a production of consensus seeking between 28⁸ member states. It illustrates how platforms, such as CSRs, enable divergent views and interests to be expressed in a consensual way in social interaction. Yet, tensions and contention remain as the letter itself developed into a lengthy policy text from what was originally ‘a short piece of text’ (in words of policy actor 02). This shows how translation, although attentively managed, is able to move the action forward. As a result, the product of the translation work, such as the joint letter, becomes a tool to monitor the ES in terms of education policy. Originally an informal practice, the joint letter has now become a formal practice: The Council publishes annually a document called ‘Resolution on education and training in the European Semester: ensuring informed debates on reforms and investments’ (Council of the European Union, 2020), encapsulating the function of the former joint letter.

Second, this novel practice illustrates the broadening development of the policy field of education within the EU framework. The cooperation in the field evolved outside of the legal structure (Novoa and Lawn, 2002), laying the basis for the division of competences within the EU. The Treaty (TFEU) states that the EU has ‘shared competencies’ in the area of social policy (Article 4 of the TFEU) and ‘supporting competence’ in the area of education. The so-called community action in the area of education has formed into open policy coordination without legal status. Against this backdrop, the integration of education into the ES resulted in generating non-politicised, ‘unofficial’ governing practices to overcome the problem of legitimacy (in words of policy actor 03); that is, the issue of national subsidiarity. Therefore, the education sector mobilised an informal practice in the ‘peer review’ meetings with Directors General for Higher Education: the crafting of a tailor-made approach for the ES. Instead of country-specific recommendations, the education sector adopted the country-supportive recommendations. This reflects the source of authority for national governments through their political participation in agenda setting but also the capacity of consensus building in generating instrumentation for governing the European education policy field. A policy instrument needs to be flexible so that it can be easily transformed to ‘fit’ for a policy purpose and agenda, as this policy actors from a national delegation in the Council explains:

The ES can be disconnected from the actual policy goals to facilitate consensus building among EU member states. This interaction illustrates how ‘soft power functions through persuasion and deliberation creating “voluntarily” agreed consensus on performance’ (Eeva, 2019): it depicts soft governance as self-managing, a process in which policy becomes apparently depoliticised through the translation work (Lawn and Grek, 2012: 78). Indeed, I suggest that country-supportive recommendations contain the rhetorical rules for cooperation in the policy sector, structuring actors’ behaviour and ‘the movement of discourses across the national and European levels and in the interaction between actors’ (Grek, 2014: 267). This illuminates how discourse becomes a governing resource organising the relations between systems and institutions beyond hierarchies. Member states prefer these unofficial channels of meeting-making and discussion in the context of ES because they have, according to the analysis of interviewees’, ‘no official impact or consequence in policy terms’. This development encompasses the idea of naturalisation of the European dimension into domestic settings in policy terms allegedly reducing the political levers. Speaking from the EC’s point of view, this policy actor unfolds the social character of knowledge in the instrument:

This interplay between the EC and the national demonstrates how norms, patterns of thoughts, identities and subjectivities are expressed and reproduced in text as vehicles of discourse (Freeman and Maybin, 2011: 158). It enables a dialogue for those countries that voluntarily open up while maintaining control of the policy direction. The analysis of this paper then suggests that the ES is an instrument of soft power through its ability to generate consensus of measures and tools – such as the joint letter and country-supportive recommendations – through discourse.