Introduction

The EU’s social dimension was until recently seen by most observers to be in deep crisis. After a blooming period that began in the 1990s and included new directives and the spread of non-binding regulation in the form of Open Method of Coordination to new areas, a somewhat different period began in the 2010s. Some researchers point to changes within the European Semester (the yearly cycle of the EU’s economic policy coordination) in the form of a greater focus on social policy issues from 2012 (Zeitlin and Vanhercke, 2014), whereas others point to the start of the Juncker Commission in 2014 as the turning point (Pochet 2020). Although the ‘social turn’ has been questioned (e.g. Copeland and Daly 2018; Jordan et al. 2021) and others have questioned the extent to which changes in EU-level policies will result in changes ‘on the ground’, few have argued for a total status quo. The Juncker Commission (2014–19) and the van der Leyen Commission (2019-2024) have issued a large number of new proposals and revisions for EU directives and regulations within the EU’s social dimensions and have been able to have a large majority of them adopted, although it has rightly been pointed out that revisions have outnumbered new initiatives and important concessions have been made in many cases (e.g. Vanhercke et al. 2020, p. 156; Copeland 2022). Within the labour market-related parts of Social Europe—which is the focus of the present article—new initiatives that include important new rights and possibilities for workers and citizens are the European Pillar of Social Rights (2017), the Posting Directive (2018), the Working Conditions Directive (2019), the Regulation on the European Labour Market Authority (2019), the Mobility Package on Road Transport (2020), the Work–Life Balance Directive (2021), the Minimum Wages Directive (2022). And more could be added.

The substantial change in priority given to Social Europe issues over a short period is remarkable and, to some extent, surprising. Drivers contributing to this development include, first, the negative side effects of austerity policies that have gradually become more visible and created the desire for alternative policies amongst policymakers as well as European citizens in general (e.g. Schmidt 2020b; Carella and Graziano 2022). Second, with Brexit, the ‘regulation sceptical actors’ (e.g. Mailand and Arnholtz 2015) lost their most outspoken and strongest member. Furthermore, Brexit had already started to impact EU politics following the referendum in June 2016 (Advisory Council on International Affairs 2018, p. 27; Hantrais 2019). Third, a stronger Social Europe potentially could weaken the EU’s legitimacy crisis as it appeared in the first part of the 2010s, including the EU’s lack of legitimacy in the eyes of European citizens (Mailand et al., 2022).

However, such drivers do not automatically lead to changes. Changes are made by actors, and the most important of these are the member states, the European Parliament (‘the Parliament’), the European social partners, and the European Commission (‘the Commission’).

First, with the labour market and social policy being mainly within the competence of the member states, they are obviously important to focus on when searching for drivers of change in Social Europe. Although the number of centre-left governments—supposedly the most pro-Social Europe governments—has increased from its low point at the beginning of the 2010s, they still made up only one-third of EU governments in 2019 (Pochet 2020). Euroscepticism amongst member state populations is another relevant factor to consider in relation to the member states. Euroscepticism was on the rise for a period until 2014, but has been decreasing since then (Kang and Ho 2020). However, the turn came too late to explain the change in policy at the EU-level. Rather, the increasing level of Euroscepticism prior to 2014 was an explanation for the change in policy at the EU-level. Moreover, although Euroscepticism in the population has decreased when measured as trust in the EU, Eurosceptic parties have continued to grow, measured in terms of their turnout in elections (popu-list.org).

Second, regarding the European Parliament, the elections in 2014 and 2019 did not strengthen the political forces supporting Social Europe. The 2014 election strengthened the Eurosceptic parties, and although the 2019 election did not strengthen these parties any further, the centre-left and centre-right parties, which are generally in favour of Social Europe, lost their majority for the first time (Ripoll 2019). Hence, it does not seem to be in differences in the Parliament’s turnouts in elections that one might find an explanation for the change regarding Social Europe.

Third, with regard to the European social partners, although employers' organisations in many member states have succeeded in sustaining their membership, most trade union movements in Europe continue to lose members (Vandaele 2019) and struggle with a lack of organisational resources as well as weak bargaining power and political influence. Therefore, the change in European policies cannot be explained by a strengthening of the ETUC or of its member organisations, and hardly by developments related to BusinessEurope.

This leaves us with the Commission. A number of very recent studies have already indicated that the Commission is the main driver in strengthening Social Europe. Mailand (2021) argues that the Commission was clearly the main driver in establishing the European Pillar of Social Rights (‘EPSR’ or ‘the Pillar’), which, since 2017, has operated as a mainly legally non-binding—but nevertheless important—umbrella for Social Europe-related initiatives. The Pillar is not only about workers’ rights but also citizens’ rights, and it has a broad perspective on Social Europe, which includes equal opportunities and access to the labour market, fair working conditions; and social protection and inclusion. Although some actors welcomed the Pillar, none other than the Commission had asked for such an initiative, and although concessions were made during the decision-making process, the Pillar survived without too many important changes. The broad scope and coverage of both hard and soft law, and coverage of the areas of several General Directorates’ responsibility, created an institutional ambiguity. This was amongst the factors that enabled the Commission to effectively steer the decision-making process. Likewise, Vesan et al. (2021) find the Commission to be the main driver of the Pillar in a study that especially emphasises the role of the president of the Commission, Jean-Claude Juncker, and his use of the ‘top-units’—the Cabinet and Secretary General—to push through his ideas for the Pillar and socialise the European Semester. Copeland (2022) also clearly points to the Commission as the leading actor in his analyses of the Pillar and other recent initiatives. This author finds the Juncker Commission to be a ‘politicalising bricoleur’, which has revitalised and revised existing tools with the aim of strengthening the Social Europe.

What these studies focus on is almost exclusively the meta-governance initiative the Pillar, where mainly the Commission and the member states play important formal roles. It is therefore unsurprising that the other main Social Europe actors—European social partners and, to some extent, the Parliament—did not appear to be very influential in these processes.

From the perspectives of the research traditions of industrial relations as well as of European integration, it would be interesting to inquire if the Commission also stands out as the most influential actor if the focus is moved to more direct wage and working conditions-related EU-level decision-making processes during the same period. The European social partners and their social dialogue have at least a possibility of having a higher impact because their formal treaty-based roles are greater here than in the meta-governance processes mentioned above.

If the Commission also stands out as the most important actor in direct wage and working conditions-related initiatives, it confirms the Commission’s importance as the main driver in Social Europe, and furthermore implies that the European social partners risk being mere policy-takers rather than policymakers at the EU-level on their core areas of interest. This is problematic from a self-governing perspective, as it implies that decisions are moved away from the actors most affected by them and regulations can be introduced or blocked because of the interests of other actors. Another implication of the possible Commission supremacy and social partners’ policy-taker status is greater volatility in Social Europe-related regulation. Whereas the social partners’ preferences only change gradually over time and only to a limited extent, the change from the Barroso Commission to the Juncker Commission shows an enormous difference in the priority of Social Europe with huge consequences for the content and extent of the regulation. This might result in a stop–go policy, which makes it difficult for member states and other national level actors to prescribe the regulatory context they are to operate within.

So far, only a few studies have analysed the decision-making processes of the directive after the social turn. Studies of the revision of the Posting Directive 2016–18 illustrate that, just like with the EPSR, strong priorities and entrepreneurship from the Commission’s President and the Commissionaire for Employment were key in getting the unlikely revision adopted (Seikel 2019, 2022) in a situation where the member states were very divided on the issue (Steuenberg, 2021). However, supporters of the revision in the Parliament were also important in making the approval of the president of the Commission conditional on the revision and by side-tracking opponents in the Parliament decision-making processes. Moreover, the combination of European employers’ opposition and ETUC support gave the latter stronger access and influence than usual through bilateral contacts with the Commission and Parliament (Seikel 2019, 2022).

To date, no other in-depth peer-reviewed studies have been published on the decision-making processes of new or revised directive after the social turn, although the content of the recent directives—especially the work-life balance directive under EPSR—have attacked some attention from researchers (e.g. Pircher et al. 2024). The closest we get is Copeland’s above-mentioned study of the EPSR as well as a number of initiatives under the umbrella of EPSR, including the revision of the Written Statement Directive and the Work-Life Balance Directive. Apart from pointing to the Commission as the dominant actor, the study also finds that the European Parliament is influential in the cases of both directives, whereas no observations are made about the role of the European Social Partners. The Work– Life Balance Directive ended up being a partly watered-down version of the Commission’s initial proposal (Copeland 2022). Still, the directive extended employees’ rights and benefits substantially compared with its predators, just as the revised Posting Directive did.

In sum, the newest studies indicate that although the Commission also appears to be the strongest actor in these cases, the Council and especially the Parliament seem to be more influential than they were in the formulation of the EPSR. The role of European social partners is less illuminated. However, in the Posting Directive, they seem to have gained some influence through their lobby activities. The aim of the article is to address, through a critical case study, the question of ‘who governs’ amongst the main EU-level actors by analysing the roles and relative importance of the key actors in a specific recent Social Europe working rights-oriented initiative to access who the main driver has been and the extent to which the European social partners have been able to influence processes and outcomes. This focus excludes an evaluation of the impacts of these initiatives when implemented in the member states. Moreover, because areas of EU regulation other than the social dimension are beyond the focus here, whether the initiatives analysed represent a real balance to the EU’s economic regulations will not be discussed.

State of the art in research

The theoretical framework includes approaches from both the industrial relations and the Europeanization traditions, which normally do not interact much. However, regarding Social Europe-related decision-making processes, both have something to offer.

The classic Europeanization approaches in old and new variations

Theoretically, the aim of this article relates to the classic debate in EU-level studies between neo-functionalist (Lindberg 1963) and inter-governmentalist (Keohane and Hoffmann 1991) approaches. In recent years, both inter-governmentalist and supranationalist approaches have been refined.

The most influential of these approaches has been ‘new inter-governmentalism’, which finds that member states have taken a much stronger leadership role than previously. To do so, member states establish new bodies and institutions outside EU institutions (for instance, a President of the Council) and place themselves where they were not well represented previously (for instance, the governing body of the European Central Bank) (Hodson 2015), instead of delegating power to the Commission. For decisions to be respected and consensus sustained, constant processes of consensus generation are needed at all levels to allow for EU regulation, including after the decision-making processes have officially ended (Puetter 2012, 2014).

The ‘new supranationalists’ argue, contrary to this, that actors such as the Commission and the European Central Bank have been able to use their power—ironically sometimes granted to them by inter-governmentalist actors—to obtain more influence than in the past (Schmidt 2020a). Focusing primarily on financial and economic policies, Bauer and Becker (2014) found that the Commission continues to be a powerful player in EU economic governance. Although its agenda setting role has diminished, most decisions still depend on the Commission to make them work, and it has been granted greater implementation competencies.

Traditionally, neither intergovernmentalists nor supranationalists have paid much attention to the European Parliament. However, the power of Parliament seems to have increased since the Maastricht treaty, and in research on European integration, the ‘parliamentarians’ have subsequently been strengthened. The Parliament has increased its power in several ways. One is by playing a very active role in approving (and vetoing) candidates for top positions such as President of the Commission and the EU’s Ministry of Foreign Affairs; another is the increasingly important role of ‘trialogues’ (including the Parliament, the Council and the Commission) in the final stages of decision-making processes. However, the Parliament has also aimed to increase its power through informal contacts with other actors. Moreover, other EU institutions often aim to legitimise their initiatives by getting the support of the Parliament in processes where this is not formally needed, which also contributes to the Parliament’s influence (Heritier et al. 2016; Schmidt 2020a; Hix and Høyland 2013).

Studies of European social dialogue

Although there are no theoretical approaches comparable to those focussing on the ‘main’ European institutions, a rich debate of what is often labelled ‘European social dialogue’ (ESD) has taken place within the industrial relations tradition.

Until the 1970s, virtually no Social Europe-relevant regulations were introduced, apart from free movement regulations (Larsen and Taylor-Gooby 2004; Threlfall 2007). In 1986, the Single European Act created a legal basis for the development of ESD. In 1991, the European social partners adopted a joint agreement calling for mandatory consultation of the social partners (tripartite social dialogue) on legislation within the social dimension and for the possibility that the social partners could negotiate framework agreements at the Community level (bipartite social dialogue). This request was acknowledged in the Agreement on Social Policy annexed to the Maastricht Protocol on Social Policy in 1993 and fully included in the Amsterdam Treaty in 1997. Hence, studies of ESD in the 1990s were generally optimistic about the scope of the dialogue. Some observers even saw the gradual development of a ‘corporatist policy community’ with strong and stable relations and privileged access for the European social partners, which were operational and able to produce binding agreements, although treaty-based limits regarding social policies existed (Falkner 1998, 2003).

Later, the perspective became more pessimistic, and during the Barroso Commissions 2004–14, it was common to see the dialogue as a kind of crisis. The following Juncker Commission’s intention to ‘restart’ the dialogue did not fully overcome the crisis (Tricart 2020). The position of the Commission is very important for ESD, which de facto takes place in ‘the shadow of hierarchy’, including the signal ‘if you don’t agree on regulation, we will regulate’ (Scharpf 1997; Sørensen et al. 2022). During the Barroso Commission, the shadow was weak and employers felt no pressure from the Commission to engage seriously in the dialogue. During the Juncker Commission, the pressure was seemingly higher, but this Commission showed a great willingness to get regulation adopted through the community procedure rather than working patiently with the social partners. This approach has created disincentives for employees to engage seriously in social dialogue—and incentives for 'regime shopping'—as they have anticipated better results from allowing processes to go through the community procedure rather than social dialogue.

With regard to legally binding regulation (‘hard law’), the outcome of ESD was not numerically impressive. Only three directives derive directly from bipartite dialogue, as they first appeared as European social partners’ so-called framework agreements. These are the directives on parental leave (1996), part-time work (1996), and fixed-term work (1999). In addition, four framework agreements have the status of so-called ‘autonomous agreements in that they have not been transposed into directives, and the social partners themselves are responsible for their implementation. These includes the framework agreements on telework (2002), work-related stress (2004), harassment and violence at work (2007), and inclusive labour markets (2010). Apart from these intra-sectoral directives and framework agreements, ESD—including the dialogue in the numerous sectoral committees—has resulted in numerous texts of various, but mostly legally non-binding, types (Clauwaert 2005; European Commission 2016).

The ESD studies do not address the question of the relative importance of actors as directly as the Europeanization studies. However, the inability of the European Trade Union Congress (ETUC) and other European trade union associations to convince their employer counterparts to enter into bargaining can be seen as trade union weakness, but also questions the power of European social partners in general. It appears that the European social partners are increasingly just organisations that lobby other actors instead of developing regulations themselves.

Policy entrepreneurship

Since this article aims to access if the European Commission is the main driver of Social Europe policy initiatives (see below), the theoretical framework will include an approach within the family of supranationalism: policy entrepreneurship. There are different approaches to studying policy entrepreneurship. De la Porte and Natali (2018) focus on individuals and distinguish between intellectual entrepreneurs (academics or experts active in defining problems in the early part of the decision-making process), bureaucratic entrepreneurs (active in transferring knowledge from the previous phase and framing consensus building), and political entrepreneurs (active in consensus building during the time-limited window of opportunity). Crespy and Menz (2015) focus on collective actors and emphasise the need for policy entrepreneurs to possess four key abilities: displaying social acuity (making use of policy networks and responding to important ideological changes), defining problems, building teams, and leading by example. Copeland’s above-mentioned study (2022) applies Crespy and Menz’s approach.

Both these studies focus empirically on Social Europe areas, but the present study has found more inspiration in Ackrill and Kay’s (2011) approach to ‘commission entrepreneurship’, which focuses on collective actors beyond Social Europe. Ackrill and Kay find that policy entrepreneurs can be directly involved in decision-making and that the capacity of Commission officials to act as political entrepreneurs depends on the presence of institutional ambiguity, which the authors define as a political environment lacking a clear hierarchy. In the case of high ambiguity, the Commission is in a good position to act as an entrepreneur. Ackrill and Kay suggest that their approach is particularly useful when studying agenda setting within broad ‘governance architectures’ where several Directorate Generals are involved because the institutional ambiguity is high. The Commission can also make policies in one area spill over to other areas and respond to changes in policy signals by developing new policies or adapting existing policies to new circumstances. The concepts of institutional ambiguity and governance architecture are useful for the present study and will be addressed below.

Case selection, expectations and methods

The literature review has provided arguments for focussing on each of the main Social Europe actors—the Commission, the Council, the Parliament and the European Social partners—and suggestions for mechanisms and types of regulation that will bring them to the forefront and secure their influence. Drawing on this review, the presented recent empirical studies, and the reflections in the introduction on the main actors’ development, the expectation is that the Commission will also appear as the most influential actor in wage- and working condition-related initiatives, whereas the European social partners—due to the content of these initiatives and their treaty-based right to bargain—will be expected to show some impact.

In the case study that follows, the focus will both be on agenda setting and the formal decision-making process. Indicators of influence will be seen as related, first, to who initiated the initiative. The ability to introduce an initiative is in itself an indication of strong influence. However, the strength of this influence depends on the concessions given through the decision-making process. Hence, second, the fewer the number of concessions and the less these are related to core issues in the initiative, the stronger the influence of the initiator. For non-initiating actors, the reverse is the case: The greater the number of issues within the initiative they are able to change during the decision-making process, and the more these are related to core issues within the initiative, the stronger their influence.

Methodologically, the article will centre on a case study of the decision-making process of a recent major Social Europe initiative, which focuses on key industrial relations issues. The case is the decision-making process during 2016–19 leading to the Directive on transparent and predictable working conditions (in the following, the ‘Working Conditions Directive’ or ‘WCD’), which started as a revision of the Written Statement Directive, but developed into a new directive. This initiative was one of two original directive initiatives under the European Pillar of Social Rights, and it became the first labour law directive adopted for many years. The case is critical in relation to the influence of the European social partners in the sense that its content and context maximises the probability of a certain outcome (e.g. Flyvbjerg 1991)—here, a strong social partner impact. This is due to, firstly, the social partners’ treaty-based right to bargain on the issue; secondly, that the case represents the first EU-level opportunity in many years to improve working conditions; and, thirdly, that the issues covered are core matters for the social partners and they therefore could be expected to make a great effort to maximise their influence. Because the case is critical, it is unlikely that the social partners will have a stronger or even as strong an impact on other types of Social Europe regulation.

Directives are important—including for the European social partners—because they are part of the EU’s ‘hard-law’ (legally binding) rules. There are basically two types of European hard law: directives and regulations. Unless they are judged at not correctly transposed, directives do not have a direct impact on the member states. They have to be implemented in national legislation/collective agreements to come into force. Typically, EU member states have a couple of years to do so. Contrary to directives, regulations are legally binding on member states as soon as they are adopted at the EU-level.

The case study consists of a decision-making process analysis, including: (1) interviews with key decision-makers from the Directorate General of Employment (‘DG Employment’), selected Council members, selected Parliamentarians, and European social partner organisations (see Annex A); and (2) analysis and comparison of draft and final versions of key documents within the three decision-making processes. Unless otherwise stated, the source of information is the interviews.

The directive on transparent and predictable working conditions

The forerunner of the Working Conditions Directive was the Written Statement Directive of 1991 (Directive 91/533/EEC), which required employers to provide information to employees on the terms of their employment in a contract within the first two months of employment.

The decision-making process will be analysed below in chronological order: (1) an agenda setting phase, which included the decision to revise the Written Statement Directive; (2) a consultation phase orchestrated by the European Commission, clarifying amongst other issues the European social partners’ willingness to bargain; (3) the Council’s (member states’) reactions to the Commission’s proposal; (4) the European Parliament’s positioning and report; and (5) the endgame in the form of the ‘trialogue’ and the outcome (the adopted directive).

Agenda setting

After almost 25 years, the Written Statement Directive went through the Commission’s so-called ‘Refit’ evaluation. The evaluation pointed out that non-standard employment had grown substantially since the directive was adopted and that a common European definition of an ‘employee’ (as opposed to a self-employed person) could be considered. The evaluation also described a number of weaknesses in the directive, such as the extent to which member states had implemented it. Importantly, the evaluators also argued that most of the report’s proposals did not necessarily require a revision of the directive (Ramboll Management Consulting 2016).

Therefore, the refit evaluation did not create expectations of a major revision of the directive. However, in May 2016, shortly after the end of the evaluation, the Juncker Commission presented its preliminary ideas to the Pillar in a communication sent for open consultation.

From consultation on a negotiating option to the Commission proposal

The Commission’s proposal for the Pillar—published in April 2017-included a revision of this directive as one of only specific initiatives. Thus, the directive had become 'big politics', as one of the interviewees described it (interview 1).

The hearing of the social partners was launched in April 2017 at the same time as the Commission’s proposal for the Pillar was published, but according to interviewees, the Commission had made the decision that there would be a revision long before. After this hearing, the social partners had to decide whether they were willing to negotiate a proposal for a directive. The European social partners did not, however, enter into such negotiations. The interviewees mentioned two (non-exclusive) explanations for why this did not happen. The first is that whilst BusinessEurope stated its readiness to work towards a negotiation, the ETUC did not because it was counting on the Commission’s ‘social turn’ and the Pillar to get a better revision through a Commission-driven process (interview 1, 2, 3, 6, 7). Such speculation has been seen before in connexion with the revision of EU labour law directives, e.g. when the directive on European Works Councils was to be revised (Mailand 2013). The second interpretation suggests that the Commission was the actor preventing a negotiation by prematurely concluding that the social partners would not negotiate (interview 1, 3, 7). Accordingly, the Commission did not want to wait for the social partners as the end of the Juncker Commission’s term of office (November 2019) was approaching, and a completed revision of the directive before then was a high priority of the President of the Commission (interview 1, 2, 3, 4, 6). He needed to demonstrate an ability to ‘walk the talk’ concerning the social dimension.

In the document presented for the Commission’s first phase consultation of social partners (Europe Commission 2017), the directive is still referred to as the Written Statement Directive. However, it was clear that the Commission wanted a broader revision. It should not only focus on improving the level of information about terms of employment and ensure the implementation of the regulation but also introduce minimum rights for everyone in an employment relationship, including the estimated 2–3 million non-standard employees who are not covered by the Written Statement Directive. The rights were intended to be related to the ‘rights and principles’ within the European Pillar of Social Rights. According to interviewees, the Commission’s wish for a broader directive should also be seen in the light of its inclusion under ‘the Pillar’. As one of the few legally binding initiatives, it should have a certain scope (interview 1, 3, 6, 8).

The ETUC and other European employee organisations generally welcomed the ideas contained in the first phase consultation, including the reduction in the two-month limit regarding employment contracts and the introduction of a broader range of rights. However, the ETUC also stressed that it would like to see even more rights and the inclusion of the self-employed. BusinessEurope and other European employers’ associations were generally opposed to extending the scope of the directive and the idea of introducing more rights as well as shortening the two-month limit for issuing employment contracts. Regarding the negotiating aspect, the European employers declared themselves willing to initiate talks with the ETUC, whereas the European employee organisations opposed it. Thus, the Commission concluded that neither party was ready to initiate a genuine negotiation process (European Commission, 2017c). This conclusion was, as mentioned, controversial.

The Parliament made their main priority clear when they adopted a resolution on the Pillar in May 2017, which called for a revised directive to cover more employees, amongst them non-standard employees (European Parliament 2017).

The Commission’s proposal

The Commission’s proposal for a directive followed in December 2017 (European Commission 2017).

According to interviewees, a sign of the high priority given to the proposal was that the compulsory impact assessment was carried out rather quickly. Contrary to the norm, the regulatory scrutiny board did not send the proposal back for adjustments (interview 6).

The title of the directive had changed since the consultations with the WCD. The main proposed changes in relation to the existing directive were as follows: (1) extend coverage to include a number of previously uncovered groups of non-standard employed (though not self-employed persons); (2) specify the definition of an employee and update the minimum requirements for what proof of employment must include; (3) introduce a number of minimum rights (including a maximum probationary period of six months); (4) reduce the deadline for proof of employment from two months to the first day of employment; and (5) introduce an option to derogate from minimum requirements through collective agreements, so long as the total level of protection is not reduced.

The ETUC was mostly positive in its response in December 2017, but nevertheless expected a stronger proposal. The elements of the proposal they emphasised as positive were the requirement for proof of employment from day one of recruitment and that platform workers should also be covered by this requirement. However, the ETUC found that the measures were insufficient in safeguarding against the worst forms of precarious employment, including zero-hour contracts and contracts with too low an hourly rate of pay (ETUC 2017).

BusinessEurope maintained its criticism from the first consultation phase. In their press release in February 2018, they argued that the proposed directive was disproportionate to the problems to be addressed and not in accordance with the principle of subsidiarity. In addition, it highlighted that the directive as proposed would impose large administrative burdens on companies that the EU should not interfere in how the concept of an employee is defined and that it was inappropriate that the proposed definition included self-employed people. BusinessEurope further emphasised that the proposed minimum rights should be regulated nationally and in collective agreements (BusinessEurope 2018).

The Council’s consideration and response of the social partners and the Parliament

The Council discussed the Commission’s proposal at a number of meetings up to June 2018. The Council proposed, inter alia, to reduce the ‘exception limit’ of a maximum of eight working hours per week to five; to only reduce the deadline for employment contracts to one working week for ‘essential’ information and one month for other information; and to delete all references to the definition of an employee, so only the references to decisions of the European Court of Justice remained (Kiss 2019).

Five of the most controversial issues in the decision-making process were the definition of an employee, the question about the coverage of non-standard employees, time limits, the opportunity to derogate from the directive via collective agreements, and the enforcement chapter (interview 1–8).

The issue concerning time limits concerned the two-month limit for receiving adequate information on the employment relationship as well as whether the directive should continue to apply only to employment of at least eight hours a week. As mentioned, the Commission proposed that proof be provided on the first day of employment, but the employers and a number of member states considered it too soon. With regard to the eight-hour limit, the employers and a number of member states wanted a kind of ‘triviality limit’ at very low hours to avoid unnecessary administrative burdens on employers. It developed into a tug of war where the boundary moved back and forth between different levels below eight hours before a compromise of three hours was agreed late in the decision-making process (interview 2, 5).

The question of the possibilities for derogating through collective agreements was something that particularly occupied Danish and Swedish actors especially. The Temporary Agency Directive (from 2008) includes a provision that opens up the possibility of implementation through collective agreements, and the Danish actors wanted a revision of the Written Statement Directive to ensure the same opportunity. Already in the agenda setting phase, Danish and Swedish actors had pushed for such an opportunity, which was subsequently included in the Commission’s proposal, but the article was short and open for interpretation, and according to interviewees, the Commission’s ideas were not along the lines of the derogation model from the Temporary Agency Directive. In particular, the Swedish actors were not satisfied and ‘raised a yellow card’ (for disrespecting the subsidiarity principle) (interview 1, 6). However, one of the interviewees described the DG Employment officials as very responsive to the Danish position and willing to defend the relevant section in the proposed directive if it came under pressure (interview 1). This occurred when, in early 2018, the Parliament’s Committee on Employment and Social Affairs (EMPL) began working on their report, which forms the basis of the Parliament’s position. According to some of the interviewees, defending the derogation was a challenge as both the rapporteur and the shadow rapporteur were southern European (Spanish), and as the Spanish collective bargaining system was weakened, there were some concerns amongst the report's rapporteurs that derogation via collective agreements would lead to a watering down of the directive (interview 1, 3).

The definition of an employee was controversial because the Commission’s proposal opened up a common European definition based on European Court of Justice case law. This idea received support from a number of member states and parts of the European trade union movement, whereas employers, other member states (including Austria, Germany, Sweden, the Netherlands and the United Kingdom) and parts of the European trade union movement (including the German unions) opposed it. The German government, according to interviewees, was amongst the leading voices against the proposal (interview 1, 6). The sceptics’ resistance was partly motivated by a fear of being forced into accepting an intermediate category between employee and self-employed, which is recognised in some European countries, and partly by concerns that a pan-European definition would be too rigid when the European labour market is undergoing rapid change. However, resistance was also related to problems with fitting such a definition into existing national labour codes. According to the opponents, the concept of an employee should be defined at the national level, but with respect to the European Court of Justice’s EU case law. The opposition to an employee definition was pronounced, and during the Council’s discussion of the proposal it was, as mentioned, abandoned (interview 5, 6, 7).

The coverage of non-standard employees, including zero-hour work contracts, was also a focus. Actions in this regard would normally be opposed by the UK government, with the backing of a number of other member states. However, according to one interviewee, the so-called Taylor report (Taylor et al. 2017) had put the Conservative government under pressure to do something about the large number of non-standard employees in the UK. This, perhaps together with the Brexit referendum, led the UK to be surprisingly supportive of the Commission’s proposal (interview 6).

According to some of the interviewees, an interesting difference in opinion appeared within the Council. The Council working group, staffed with civil servants, turned out to be more sceptical of the proposal than the ministers. The difference had to do with the focus on non-standard employees. Contrary to the meetings in the working group, meetings with the participation of the ministers are public, and none of the ministers wanted to be cited as not supporting better working conditions for non-standard employees (interview 1, 6).

Finally, the directive included an enforcement chapter, which is normally not found in such an extensive form in labour law directives. Interviewees found that some member states (including Austria, Germany, Hungary and Estonia) wanted virtually the entire chapter deleted, but there was not a blocking minority behind this wish and it survived with a few changes (interview 6).

In October 2018, the EMPL committee presented its report. It contained a large number of amendments, including that trial periods should be limited to six months that all information should be included in the evidence and provided within one week of the start of the employment relationship and that social partners should be involved in implementing the directive in member states. Based on the report, a large majority voted in November 2018 for the Parliament to enter the trialogue process (Kiss 2019).

Some interviewees found the changes proposed by the Parliament more extensive than usual. An explanation might be that the centre-right European Peoples’ Party (EPP) group in the Parliament was split, thus allowing more amendments to pass (interviews 5, 6, 7). This created a position in which the Parliament and Council were far removed from each other from the outset. However, the trialogue process was not very long. None of the main actors wanted the directive blocked, and all were therefore willing to compromise.

The final directive and the main actors’ influence

In February 2019, the trialogue was completed. The Commission’s proposal for derogation through collective agreements survived with minor linguistic changes, but the Commission had to promise to check that derogation would not lead to inadequate implementation of the directive in member states. A number of other conditions and formulations were also adjusted.

The main points of the adopted directive (compared to the Written Statement Directive) are:

  • The employer must offer an employment contract no later than one week after the employment relationship has begun (previously up to two months). Employees who work less than three hours per week, on average, during a four-week period, may be exempted if the member state chooses to do so (previously it was eight hours per week).

  • The employer is obliged to inform the employee of a number of rights, e.g. on social security schemes and continuing education.

  • In the event of changes in the employment relationship, the employee must be informed accordingly as soon as possible—and no later than the same day as the change is implemented.

  • In addition to rules on employment contracts, the Directive introduces a number of minimum rights relating to employment conditions more generally. These include, e.g. a maximum probationary period of six months, a right to take up employment with more employers, minimum standards of predictability during working hours, the right to inquire about other types of employment with the same employer, and restrictions on contracts without a guaranteed minimum working time.

  • The scope of the directive was broadened to include a large number of non-standard employees (European Union, 2019).

  • Member states are allowed to derogate from the directive through collective agreements.

As was the case with the EPSR addressed in the introduction, the Commission cannot as clearly be pointed to as the sole initiator, and the revision process was less centralised and directed by the Commission. However, it was the Commission that initiated the Refit evaluation and decided to choose the directive as one of four specific initiatives to put flesh on the bones of the Pillar. Although helped by the Parliament’s resolution, the Commission also wanted a broader directive than its forerunner. All of this illustrates commission entrepreneurship. Moreover, the Commission’s eagerness to get the revision done before the end of its term might also have played a role in sidelining the European social partners, who seem to have played a mainly reactive role despite the directive being a critical case in which one could expect their influence to be at its strongest. Finally, despite some important changes, the adopted directive is very similar to the Commission’s proposal. (Table 1)

Table 1 Summary table of issues of contention and their outcomes
Full size table

The Council has also been important, as the tug of war on many of the new directive’s dimensions illustrates. Moreover, individual member states—alone or in coalitions with, for instance, social partners at the national level—had managed to add or change specific provisions in the directive, as the Danish and Swedish actors did in relation to the derogation via collective agreements. Most importantly, resistance in the Council was key to getting the Commission proposal of a European definition of an employee abandoned.

Still, the most important actor apart from the Commission might have been the Parliament. The Parliament’s resolution helped to broaden the scope of the directive so that it also covered non-standard employees. Furthermore, it proved difficult for the Parliament to endorse the compromise between the Council and the Commission. Finally, the Parliament got essential elements of the draft directive amended. In sum, the Commission stands out as the most influential actor in this case, though other actors had a stronger influence in this case than in e.g. the case of the EPSR.

Discussion, conclusions and perspectives

This case study has provided findings of relevance for the aim of the article, which, as stated in the introduction, was to analyse the roles and relative importance of the main actors in EU-level Social Europe regulation.

During the period since 2015, commissions have been very active in the field of Social Europe, and they have so far been able to get a large majority of their initiatives adopted. This article has attempted to delve into one of the decision-making processes to see if the Commission, as hypothesised, stands out as the dominant actor whilst the European social partners are reactive secondary actors, as they were, e.g. in the case of the meta-governance initiative, the EPSR.

The case study, to some extent, confirms the expectation about Commission supremacy through entrepreneurship. Supporting the hypothesis is that the Commission initiated and timed the revision, elevated it high on its political agenda, and was able to broaden its scope considerably. Moreover, comparing the Commission proposal with the adopted version of the directive shows extensive similarity, although a few important differences can also be found.

However, the study also shows the limits of commission entrepreneurship. One reason that the expectation is only supported to some extent is the Council’s success in removing an important element in the WCD—the definition of a worker—due to resistance from member states.

Another reason why the expectation can only partly be confirmed is that the case shows what has generally been said by the ‘parliamentarians’ within Europeanization studies: the Parliament has become more active and influential. The high activity level is shown both in issuing the resolution early in the decision-making process and in the extent and content of the amendments that the Parliament proposed during the trialogue late in the decision-making processes.

The European social partners seem not to have had a strong influence in the case of the WCD. They could have had a strong influence if they had entered and finalised the bargaining path option, but they did not. Hence, this case reflects the crisis in European social dialogue, including the trust crisis between European social partners and the Commission, as described by Tricart (2020) and others.

As the WCD is a critical case, a hypothesis for further research could be that European social partners are reduced to secondary actors, even in the formulation of a new Social Europe regulation on their core areas of interest.

An explanation as to why the case analysed shows Commission dominance to a lesser extent than the EPSR could be that the institutional set-up around the decision-making process was more or less fixed and therefore did not include institutional ambiguity, which, according to Ackrill and Kay, facilitates Commission entrepreneurship. However, the WCD-case illuminates another of Ackrill and Key’s mechanisms for Commission entrepreneurship, spill over. However, it did not take the form where the Commission facilitates spill over from one policy area to another, but where a broader initiative (the EPSR) spill over to facilitate and legitimise the following initiatives, amongst them the WCD.

Combining the higher level of activity and influence of the Parliament with another observation from empirical research helps to explain how a Commission-driven ‘social turn’ has been possible and why other main actors have not blocked it. The observation is that when it comes to Social Europe-related issues, the Parliament tends to take ‘employee-friendly’ and ‘citizen-friendly’ positions. Several interviewees shared this observation (interview 1, 4, 5, 6, 7). Moreover, it is interesting that this orientation in Social Europe issues does not reflect the Parliament’s partisan composition on a one-dimensional left–right scale. The voting of the centre-right party group EPP might be key to understanding this phenomenon. Although it is a conservative party group in many respects, many of the group’s members share a social partnership ideology with party groups to the left (see also Copeland 2022). In Social Europe-related matters, the EPP group often either votes with the centre-left S&D group or is split. Although the EPP and S&D groups together no longer have a majority in the Parliament, the axis is still key in gaining support for Social Europe-related initiatives.

The Parliament’s increased importance and tendency to lean towards the ‘workers side’ gives the ETUC good opportunities for parliamentary lobbyism, which this case also illustrates, as ETUC and the Parliaments Employment Committee discussed their report (interview 6). Hence, trade union influence in cases where the European social partners’ formal roles are limited could be underestimated if informal lobbying activities are not considered.

Despite its status as critical, generalisations from a single case study should be made with care. Apart from the limits that follow from the low case numbers, the case clearly belongs to the labour law part of Social Europe, which its particular actor constellations. These cannot easily be generalised to other parts of Social Europe. However, by including the six other important initiatives adopted since 2017 mentioned in the introduction, it is possible to broaden at least some of the findings from this case study to other parts of the labour market related to Social Europe. With regard to the role of the European social partners, it is notable that in five of these initiatives they had a treaty-based right to bargaining and subsequently could have had their autonomous agreements transformed into directives, but the social partners could not agree and did not even try to bargain on any of the six initiatives. Hence, although Social Europe regulations currently provide workers and citizens with numerous new rights and possibilities, contrary to Social Europe’s golden period in the 1990s, the social partners themselves are not in the driver’s seat, not even with respect to regulation on core industrial relations issues. Their main roles now are to influence other key actors through agenda setting, lobbying, and consultation processes. Unsurprisingly, the European employers’ organisations find it problematic to be burdened with the cost of these new regulations, whereas the trade unions, for the time being, to some extent can rely on the Commission pushing the social agenda forward.

Another limitation of the findings is that the strengthening of the EU’s social dimension argued for in this article is not driven by a strengthening of the trade unions and their European associations, but rather by a combination of learning from austerity-related politics and a political wish to use Social Europe to counteract the EU’s legitimacy crises and Brexit, as described in the introduction. Of these three possible explanatory factors, only Brexit can be expected to have a lasting effect. In other words, it is certainly possible that the strengthening of Social Europe will not last. Commissions come and go, and the latest developments could easily turn around again in the near future if political circumstances bring about a Social Europe-unfriendly commission presidency. Still, in the longer run, it is possible that Social Europe will have some resilience for another reason, which has been ignored in research: Social Europe initiatives are never really dismantled when they first are adopted, and proposed directives tend to be adopted, although, as described above, often with small or large changes. The blocking of initiatives is rare and difficult, and even when blocking occurs, the initiatives are seldom really dead. They tend to reappear on the agenda, although in some cases, it might be years before this happens.