Abstract
Social policy at the EU level remains a notably limited enterprise. The Union appears unable to address social inequalities without treading into politicised areas of national decision-making, since intervention in the social domain touches upon core questions of national identity, sovereignty and democracy. In this article, I explore the fraught relationship between EU social policy and the principle of subsidiarity, which has frequently served to legitimate support for as well as opposition to supranational action in the social domain, even when and where the Union has well-defined competences to legislate. The article engages in a conceptual analysis of the implications of subsidiarity for EU social policy and labour law, illustrated with a case study that examines usages of the principle by different institutional actors in recent debates on EU social directives. While it clearly provides a convenient rhetorical instrument to political actors, I argue, the concept of subsidiarity fails to serve as an organising principle for the multilevel governance of social and labour policy. Scholarly and policy debates on the constitution of a more social Europe are therefore best oriented away from the language of subsidiarity and should focus instead on effective and democratically legitimate ways of delivering the social promises of both the EU and its Member States.
Introduction
Social policy in the European Union (EU) falls between two stools. On the one hand, critics argue that ‘the EU is not doing enough to tackle inequalities’ and compensate for the negative effects of economic integration. 1 On the other hand, the Union's capacity to intervene in social affairs is constrained by the sensitive political nature of this policy field, which is strongly associated with the nation-building activities of states. 2 Over the last century and a half, the ‘institutionalization of solidarity’ at the domestic level has ‘strengthen[ed] that link between territories, cultural identities, and participatory institutions on which […] the European nation state ultimately rests.’ 3 Through this ‘bounded structuring’, social policy has become interlaced with the political, legal and cultural devices of national sovereignty, identity and citizenship.
This legacy has complicated the creation of social and labour policies at the supranational level since the advance of European integration in the 1950s. To be sure, the EU has adopted a raft of policy measures including minimum standards in employment regulation, social funding and legal instruments to regulate access to social protection for citizens from other Member States. Nonetheless, social integration has lagged behind in comparison to many other fields of policymaking in which the EU has acquired a larger role and impact on national politics. The institutional reality of the EU remains far removed from the kind of ‘social Europe’ with harmonised social policies that some federalist politicians and social movements envisioned. The exercise of Union competences has been circumscribed because national governments have been strongly reluctant about, or outright opposed to, further integration in terms of social policy and regulation.
Faced with this dilemma, politicians, bureaucrats and intellectuals alike have sought a guiding principle to strike a balance between levels of responsibility, and have found one concept to be particularly useful for this task: subsidiarity. Simply put, this term denotes the idea that authority in a political system should be exercised at the ‘lowest’ level at which an objective can be reasonably fulfilled. 4 The principle was incorporated in the Maastricht Treaty to allay fears about the ‘competence creep’ that might result from the pursuit of ever-closer union. 5 Article 5 TEU states that ‘in areas which do not fall within its exclusive competence, the Union shall act only if and in so far as the objectives of the proposed action cannot be sufficiently achieved by the Member States […] but can rather […] be better achieved at Union level.’ Over three decades, Granat notes, subsidiarity has ‘emerged as the pre-eminent organising principle for the allocation of powers among the EU's different levels of government.’ 6
At the same time, subsidiarity is a treacherous concept that has long bewildered EU law and politics – and their students. Legal scholarship has foregrounded the ambiguous meaning and implications of the principle as well as a reluctance by the European Court of Justice (ECJ) to interpret it extensively. 7 In parallel, scholars of EU politics have sought to unpack the conflictual history of subsidiarity and its weaponisation as a rhetorical device. 8 In this context, social policy can be said to constitute a political minefield, as noted by authors who point out that subsidiarity has served as ‘an instrument of national actors to retain social policy as an exclusive domain.’ 9 Still, its implications for EU social policy have rarely been spelled out systematically in existing scholarship. A notable exception in this regard is Spicker, who has concluded that taking the subsidiarity principle seriously would undermine ‘the case against European intervention at the social level.’ 10 Yet his intervention dates back to the pre-Maastricht era. This leads me to inquire how the relationship between the subsidiarity principle and social policy has been settled in the contemporary EU, and whether subsidiarity is (still) conceptually fit for purpose in this present context: is its meaning sufficiently clear in order to provide a yardstick with which we can assess the (im-)propriety of EU action in the field of social policy?
The article is structured as follows. First, building on existing work on subsidiarity in EU law and politics, I briefly discuss how the principle appeared in the Treaties and elaborate on the conflicting – economic and democratic – rationales that have undergirded its interpretation. Since the rich debate which subsidiarity has elicited among lawyers and political scientists can by no means be captured in this article, I focus on its implications for the debate on EU social policy, and the adoption of social regulation in particular. The enactment of directives that set minimum standards in the domain of working conditions and social protection falls in an area of shared competence between the EU and the Member States, where subsidiarity stricto sensu is relevant. Moreover, since such supranational (re-)regulation is more likely to conflict with entrenched domestic interests than social funding or policy coordination, contestation over subsidiarity compliance will be more pronounced. But the argument should bear upon broader questions of the division of social policy responsibilities in the EU's multilevel legal and political system.
The section that follows adds empirical flesh to this conceptual analysis and examines how different institutional actors have invoked the subsidiarity principle in the debates on three cases of EU social legislation: the revised Posting of Workers Directive, the Work-life Balance Directive and the Transparent and Predictable Working Conditions Directive. Third and finally, on the basis of these empirical observations, I critically reflect on the usage and meaning(s) of the principle and assess its (lack of) analytical leverage for debates on EU social policy. At the end of the day, I conclude, subsidiarity fails to provide a useful guideline or yardstick that we can employ to deliberate about the limits and opportunities for the organisation of social policies in a multilevel polity. In this sense, the search for a ‘subsidiary’ social Europe can only prove illusory.
The subsidiarity principle and its relation to EU social policy
The political history and legal foundations of subsidiarity in the EU
The principle of subsidiarity has not always been part of the supranational political and legal order, even if some authors argue that hints of it could be found in the Treaty of Rome; after all, the EU and its predecessors were built on an (originally rather implicit) division of legislative competences and implementing powers. 11 The principle surfaced more explicitly in debates about EU institutional reform in the 1970s, as government leaders, the Commission and the Parliament sought ways to reconcile further integration with effective and democratically legitimate decision-making. 12 This was visible, for instance, in the Commission's contribution to the Tindemans Report on European Union, on which Altiero Spinelli and his federalist ideas exercised considerable influence. 13
It was in the field of social policy that subsidiarity would first find a specific institutional expression. The figure of Jacques Delors, and his push to include a social rights component in the project of market integration, played a key role in this development. 14 In Delors’ philosophy, subsidiarity in its ‘vertical’ and ‘horizontal’ dimensions – i.e. between levels of government and between the state and civil society and, in particular, the social partners – was central to an effective and legitimate organisation of social policy. In this vein, the principle was invoked in the 1989 Community Charter of the Fundamental Social Rights of Workers. Its preamble stated that ‘by virtue of the principle of subsidiarity, responsibility for the initiatives to be taken with regard to the implementation of these social rights lies with the Member States or their constituent parts and, within the limits of its powers, with the European Community.’ While the principle was left undefined, a follow-up report by the Commission clarified that subsidiarity had to be understood in the sense that ‘binding measures should only be taken at a Community level where absolutely appropriate.’ 15
Following this episode, the subsidiarity principle emerged in 1992 as the ‘word that saved Maastricht’, as it was laid down in the Treaties to address fears that the EU would be covertly transformed into a federal political system. 16 Specifically, Article 5 TEC (now Article 5 TEU) stipulated that ‘in areas which do not fall within its exclusive competence, the Community shall take action, in accordance with the principle of subsidiarity, only if and in so far as the objectives of the proposed action cannot be sufficiently achieved by the Member States and can therefore, by reason of the scale or effects of the proposed action, be better achieved by the Community.’ As discussed below, this left ample room for interpretation. In its post-Maastricht jurisprudence, however, the Court has been reticent to develop a substantive doctrine of subsidiarity, showing ‘considerable deference to the subsidiarity assessment conducted by the Union legislature. Broadly speaking, the Court restricts itself to examining only whether the legislature's decision to exercise competence […] is supported by the available facts. In other words, the Court simply checks whether or not the reasons set out by the Union legislature add up.’ 17 Thus, judicial review of the subsidiarity principle was largely constrained to a procedural matter.
To reaffirm the gravity of the subsidiarity principle and increase its practical relevance, the signatories of the Lisbon Treaty introduced additional safeguards, contained in Protocol No 2 to the TFEU on the Application of the Principles of Subsidiarity and Proportionality. As Article 5 laid down, ‘any draft legislative act should contain a detailed statement making it possible to appraise compliance with the principles of subsidiarity and proportionality.’ Under the innovative political mechanism of the ‘early warning system’ (EWS), detailed in Articles 6 and 7 of the Protocol, national legislatures are invited to scrutinise legislative proposals along these lines. If they find that the subsidiarity principle has been violated, parliaments may issue a ‘reasoned opinion’ laying out their concerns. A sufficient number of complaints will trigger a ‘yellow’ or ‘orange card’, forcing the initiator – typically the Commission – to reconsider its proposal. To date, this has only occurred three times. Two yellow cards had a clear connection to Social Europe, namely, those issued for the ‘Monti II Regulation’ on the right to strike and the revised Posting of Workers Directive. 18 The Monti II proposal was withdrawn by the Commission in response to the political concerns it triggered, even if it did not evidently violate the subsidiarity principle. 19
In late 2017, demands for a better control of subsidiarity compliance led Commission President Jean-Claude Juncker to create a task force that delivered its final report in July 2018. 20 It concluded that subsidiarity remained a key tool in assessing the added value of EU action and needed to be taken more seriously, e.g. by using a ‘model grid’ throughout the decision-making process ‘to foster a common understanding and create a more effective application’ of the principle. 21 This suggests that EU institutions, at least discursively, continue to regard subsidiarity as a pivotal criterion in defining the legitimate scope of action. 22 The Court, meanwhile, has remained ‘extremely reluctant towards the use of the principle of subsidiarity in the exercise of its powers of judicial review.’ 23 In the domain of social policy, this is well-illustrated by the action for annulment of the revised Posting of Workers Directive, discussed in the analysis below.
The conflicting meanings of subsidiarity
Despite profligate attempts to disambiguate the meaning of subsidiarity, many have observed that it remains notoriously difficult to pin it down substantively. 24 The historical roots of the notion are most often traced back to Catholic social thought and Pope Pius XI's 1931 encyclical Quadragesimo Anno, but Cahill argues that ‘it is a mistake to think of subsidiarity either as an exclusively Catholic idea or as a recent modern invention’; 25 as she and others have emphasised, the idea has ancient and medieval precedents and parallels in Protestant doctrine and federalist thought. 26 One can trace a semantic fil rouge that runs through the intellectual history of subsidiarity, insofar as it is understood to refer to the legitimate allocation of authoritative functions between social bodies of a (semi-)public and private nature at different levels of governance. In this way, Fabbrini has noted, ‘subsidiarity embodies values that are typical of federal regimes – providing a constitutional method to allocate authority in a multilayered architecture in which several jurisdictions, each endowed with legislative powers, overlap and interact.’ 27
However, beyond this ideal typification of the purpose of subsidiarity, ambiguity is rife. 28 In spite of its (limited) codification in the EU legal order, it remains unclear and highly contested what exactly the principle should entail in substantive terms. In this context, we can discern two prevalent interpretations that contradict one another in their core assumptions and implications. 29 The ‘economic’ notion of subsidiarity stipulates that political authority should be exercised at the organisational level at which an objective can be accomplished most efficiently. The ‘democratic’ version of the principle, in contrast, entails that binding decisions ought to be taken at the level that is nearest to those who will be affected, so as to respect their right to self-government. As Bartl has pointed out, ‘[t]hese two sets of normative concerns have enjoyed unequal legal status’ in the EU polity. 30 The sparse definition that the Treaties give favours the former, economic conception of subsidiarity. 31 This emerges clearly from the traditional tools for operationalising the principle, the ‘national insufficiency’ and ‘comparative efficiency’ tests, which hold that a measure can be considered subsidiarity-proof if the objective that it pursues cannot be fulfilled sufficiently at the national level or can rather be better achieved at the EU level.
The legal concept of subsidiarity, then, serves as a ‘calculus’ 32 that ‘does not provide […] a substantial indication of how to distribute competences between public (and [semi-]private) actors in various layers of society.’ 33 As Bartl argues, this conception of subsidiarity presupposes a functional understanding of the EU and its authority toward the Member States, assuming that shared competences should only be exercised insofar as it would spur the ‘effective realisation of uncontroversial goals.’ 34 This functional framing of subsidiarity, in turn, is historically bound up with an idea of European integration as the construction of an ‘economic constitution’: a legal-political order geared to guaranteeing economic freedoms and setting rules to facilitate market integration via technocratic decision-making. 35 In this institutional context, public intervention is to be used sparingly and must be assessed objectively, i.e. on the basis of technical and economic knowledge about its effects in light of functional aims, rather than through democratic voice and political contestation. 36 As I argue below, this functional reading of subsidiarity has problematic implications for the exercise of competences in the domain of social policy.
EU social policy and regulation through the lens of subsidiarity
Due to its substantive vacuity, the application of a functional conception of subsidiarity to salient political questions – including those that emerge in the field of social policy – is bound to raise problems of assessment. It should not be controversial to argue, on the basis of the Treaties and legal scholarship, that EU intervention in the social domain would have to clear a higher threshold of justification than comparable action at the national level and that this analysis would have to take stock of the (comparative) efficiency of the proposed measure. 37 However, even if we take a functional rationale for granted, we may end up with two antithetical interpretations of the way(s) in which subsidiarity should delimit the supranational enactment of social policies.
One way of conceptualising the ramifications of subsidiarity for supranational decision-making in the social field resonates with the aforementioned idea of the EU as an economic constitutional order. Following this functionalist understanding, social policy has historically been ‘treated as a categorically distinct subject [from market integration]. It was the domain of political legislation and, thus, had to remain national. The social embeddedness of the market could, and should, be accomplished by the Member States in differentiated ways.’ 38 The reason why social policy has been conceived of as too political to fall under the functional competence of the EU lies in the view that it seeks to resolve questions about the equity of market outcomes at the domestic level, rather than increasing the efficiency of said market processes. 39 In this conception, only measures that improve the operation of transnational markets, such as the coordination of social security systems, could clear the threshold of subsidiarity; in the same vein, the efficiency criterion would rule out EU intervention in redistributive and protective policies.
A contrasting but prima facie equally tenable view has been put forward by Spicker, who has argued – with a hint of irony – that ‘the case against European intervention at the social level becomes very much weaker’ if one approaches subsidiarity through a functional lens. 40 Leaving aside the argument that social policies may themselves be efficiency-increasing, 41 we would here consider social policy as an instrument to give (binding) expression to widespread social values. As suggested by the notion of a ‘European social model’ 42 and the inclusion of such values in the Treaties and ancillary texts such as the Social Charter and the European Pillar of Social Rights, one can legitimately entertain the argument that these values may be realised more effectively by pursuing certain social policies at the Union level. 43 Supranational enactment and enforcement of policies in this domain might be more efficient, e.g., in terms of being cost-effective, preventing a race to the bottom and ensuring adequate protection for all EU citizens, than fully retaining these powers and responsibilities at the national level.
In its 1993 Green Paper on social policy, the Commission effectively adopted this second approach, which connects social rules to market integration without subsuming the former to the latter. 44 Deakin and Wilkinson have argued that this view of social policy ‘provides a much more plausible rationale for extensive Community action’ and a ‘potential basis for the elaboration of the doctrine of subsidiarity in the social field.’ 45 From their perspective, ‘the threat of destructive competition over standards’ meant that ‘the member states [could] not in principle achieve alone the establishment of common “social ground rules”.’ 46 But this justification of harmonisation runs into another conceptual and argumentative problem. As Davies noted in his classic critique, once the objective of (minimal) legal uniformity is accepted, a self-fulfilling prophecy will ensue: ‘Member States clearly cannot achieve this alone, [so] subsidiarity no longer applies.’ 47
As a result, it remains difficult – if not impossible – to implement this efficiency-oriented calculus in an ‘objective’ manner that is detached from the contestation of political and cultural values, as other scholars have emphasised before. 48 The question of whether a social objective, e.g. an acceptable work-life balance or stable employment, can be sufficiently realised at the national level is political and cannot be avoided through a recourse to technical knowledge. At stake are conflicting interpretations of concepts such as ‘adequacy’, ‘fairness’ and ‘equality’. The resulting indeterminacy implies that ‘[t]he determination of which level of government is best suited for regulatory tasks […] will be coloured by what the Member States are willing to accept in terms of the degree of regulatory control in any particular area’ in the first place. 49
If we approach subsidiarity from a democratic vantage point, we can circumvent some of these issues, but different problems emerge. The essential question is when policy decisions can be legitimately enforced on unwilling parties – that is, how ‘close’ to the affected constituency does such authority have to be? According to Scharpf, majority decision-making in a salient field like social policy requires a sufficiently thick common identity that, at present, exists at the national level, but not at the European level ‘above’ it. 50 Following this reasoning, one may want to advance social integration in a subsidiarity-proof way by pursuing legal differentiation between Member States, as Scharpf and others have suggested. 51
The Commission foresaw this option in its 2017 reflection paper on the future of the social dimension, which envisaged a scenario in which ‘those who want more do more.’ 52 Such initiatives in the field of social policy and regulation could take place, for example, through the mechanism of enhanced cooperation. Under this procedure, Member States that are prepared to go further in terms of social integration would be enabled to do so, without forcing others to cede rule-making authority. Yet, as the Commission itself has noted, this would not be an easy route; ‘the indispensable coherence between the economic and social aspects’ 53 of European integration is bound to complicate moves toward differentiation, e.g., by inducing regulatory competition on social standards or by impeding future membership of the Eurozone. Beyond its implications for both efficiency and legitimacy, one may ask what the added value of subsidiarity is in this light, since uniform action at the supranational level would only be admissible if agreement between (a group of) Member States were present in the first place – eliminating the very need to assess the acceptability of such intervention.
In the next section, I examine how these conceptual tensions have been settled in practice. In particular, I illustrate how national actors make use of subsidiarity controls to lend credence to their political objections to EU social policymaking. The symbolic and legal value attached to the principle has provided opponents of social integration with a powerful means to legitimate their scepticism of legislative initiatives – even if such concerns have little to do with the efficiency of the latter, but rather reflect a desire to retain ‘sovereignty’ in this domain. While these challenges have – with notable exceptions – not directly disputed the Union's right to regulate, they have imposed high limits on the (co-)promotion of social rights at the EU level. This should curb any remaining hopes about its potential as a guiding principle for the EU social dimension.
Subsidiarity in practice: the contested revival of social regulation after the crisis
From a conceptual vantage point, the subsidiarity principle has ambiguous implications for the multi-level organisation of social policy in the EU. While the concept can be construed in such a way as to promote the effective delivery of social objectives by using supranational authority, it can equally be framed as a safeguard against the transfer of decision-making powers in the social domain, which are argued to be inherently political (and thus should remain national) and cannot be justified in terms of efficiency. In the history of European integration, the balance has generally tilted towards the latter view: limited use has been made of the Union's (already narrow) social policy competences, reinforcing the ‘asymmetry’ between the EU's economic and social dimensions to which Scharpf and others have drawn attention. 54
In the section that follows, I illustrate how different institutional actors have invoked the subsidiarity principle in social policy debates. I focus on three directives in the field of social and labour regulation that were proposed and adopted under the Juncker Commission (2014–2019). Through its advocacy for a European Pillar of Social Rights, among other initiatives, Juncker and his colleagues have sought to revitalise the social dimension of the EU, bolster its legitimacy and mitigate the fallout the Eurozone crisis. 55 At the same time, as noted before, this College put considerable emphasis on the question of subsidiarity control as part of its ‘Better Regulation agenda’. 56 The analysis examines the invocation of the subsidiarity principle by the Commission, through the compliance assessment that it submits for every legislative proposal; the national legislatures that are invited to scrutinise the subsidiarity of legislation under the EWS; and Member State representatives in the Employment, Social Policy, Health and Consumer Affairs (EPSCO) Council, who may also express their views on subsidiarity in internal deliberations. I leave the European Parliament out of this consideration, as it has not been especially proactive in debates on the subsidiarity of social regulation.
Revised Posting of Workers Directive: disputing the re-regulation of transnational work
In March 2016, the Commission published a proposal 57 for the revision of the Posting of Workers Directive (PWD), ultimately adopted as Directive (EU) 2018/957 of the European Parliament and of the Council of 28 June 2018 amending Directive 96/71/EC concerning the posting of workers in the framework of the provision of services. 58 Based on the internal market provisions of Articles 53(1) and 62 TFEU and thus falling outside the scope of social policy stricto sensu, 59 this legislative act nevertheless aimed to ensure that the process of economic integration would not (further) undermine domestic welfare regimes. In response to the Laval quartet rulings by the ECJ, which cast doubt on the ability of Member States to impose social and labour policy requirements on posted work, 60 the Commission wanted to ‘enhance legal certainty and strike a more appropriate balance between social rights and free movement rules in the EU.’ 61 However, the ‘Monti II regulation’ on the right to strike in cross-border settings that it proposed in 2012 prompted 12 national parliaments to submit a reasoned opinion and foundered on the opposition of national governments and social partners. 62
Directive 2014/67/EU of the European Parliament and of the Council of 15 May 2014 on the enforcement of Directive 96/71/EC concerning the posting of workers in the framework of the provision of services, and amending Regulation (EU) No 1024/2012 on administrative cooperation through the Internal Market Information System, was adopted to combat fraud, exploitation and non-compliance with the original posting directive, imposing requirements on public information provision, monitoring, penalties and cooperation between national authorities. But since the rules applicable to posting remained unchanged, several Member States found that the initiative did not sufficiently address the risk of social dumping and pressured the Commission to overhaul the directive. 63
The ensuing Commission initiative notably proposed to include a definition of long-term posting (after 24 months, when the posted worker would fall under the labour-law regime of the host Member State); to replace the core right to ‘minimum rates of pay’ with the broader notion of ‘remuneration’, including overtime rates; to require the application of universally applicable collective agreements to posted work (also outside of the construction sector); to enable Member States to impose certain terms and conditions of remuneration on posting by subcontractors; and to oblige the Member States to ensure the equal treatment of workers posted by a cross-border temporary agency and domestic temporary workers. For the Commission, there could be little doubt that the initiative was compatible with the subsidiarity principle. Its assessment was brief and formalistic, asserting that ‘[a]n amendment to an existing Directive can only be achieved by adopting a new Directive.’ 64 Indeed, the proposed revision would only clarify and broaden the ‘hard core’ of rules to be applied, while the task to define the content of those rules would be left to the Member States and their social partners. 65
Quickly, however, 14 legislative chambers in 11 countries submitted a reasoned opinion about the alleged subsidiarity deficit of the proposal. Three lines of argument recurred in their complaints. Most frequently, they singled out the lack of positive justification and ‘proper substantiation’ of the need for EU action. The Bulgarian National Assembly and the Polish Sejm, among others, found that the Commission ‘did not state any reasons justifying that the objectives of the proposed amendments to the Directive cannot be better achieved through regulations at the national level.’ 66 The transnational nature of the problem identified and the deficiency of national action might be clear, as the Romanian Chamber of Deputies admitted, yet there was no proper analysis of the value of the newly introduced and revised legal provisions. 67 This was especially problematic, according to the parliaments, since an enforcement directive had only been adopted two years previously, and could not be adequately evaluated yet. In this light, the Slovak National Council and the Czech Senate regarded the intervention as ‘premature’. 68 In addition, the Latvian Saeima noted that the pre-legislative consultation process had been flawed: only the positions of some (receiving) Member States had been taken seriously, resulting in a biased proposal. 69
Another concern related to the ramifications of the amendments for the internal market. Invoking proportionality arguments under the guise of subsidiarity, these parliaments sought to defend the business model of lightly regulated posting. The Hungarian National Assembly opined that the proposed revision violated the subsidiarity principle because it ‘would significantly limit the freedom to provide services’ and ‘distort competition’ in a way that was not warranted by the severity of the problem. 70 The Romanian Chamber of Deputies and the Estonian Riigikogu also expressed concerns about competitiveness and economic freedom, maintaining that ‘[d]ifference in pay is a legal way for service providers to ensure their competitive advantage,’ 71 while the National Assembly of Bulgaria worried that the ‘intention to equalize payments between the seconded and the local workers will remove the existing competitive advantages of certain service providers.’ 72 According to the Slovak National Council, there was no real problem in the first place, since ‘enterprises profiting from the wage disparities between Member States and the posting of workers as a way of using these disparities has never been known as discriminating.’ 73
A last grievance concerned the legal basis of the proposal and the balance of competences in the domain of social regulation. In the eyes of the Hungarian National Assembly, the provision on remuneration violated the conferral of competences, since it ‘aim[ed] to artificially equalize the wage levels of Member States.’ 74 The Bulgarian legislature, meanwhile, reasoned that the legal basis on which the Commission had relied was unsuitable in light of the ‘social’ aims of the amendment. 75 Moreover, like the Croatian Parliament, it complained about the way in which the proposal infringed on the autonomy of the social partners. 76 From an altogether different angle, the Danish Folketing also touched upon the competence question. 77 Even though it supported the reinforcement of the rules on posted work, it worried about the lack of explicit recognition of the national prerogative over the definition of pay and the terms and conditions of temporary work.
The number of reasoned opinions sufficed to initiate a yellow card procedure. In its reply, the Commission rebutted the various complaints about breaches of the subsidiarity principle. 78 It reasoned that the durability of the posting sector demanded a consistent application of core rules and conditions by all Member States, which could evidently only be achieved through EU action. As reflected in its choice of legal basis, the proposal sought ‘to facilitate the correct functioning of the Internal Market, in particular to facilitate the freedom to provide services, while ensuring a better level-playing field between national and cross-border service providers, an adequate protection of posted workers and clarity and predictability in the legal framework applicable to posted workers.’ 79 Unilateral national action, in contrast, would threaten to fragment the posting market and create legal uncertainty. Since there was no objective to achieve wage equalisation, nor were there any provisions to regulate remuneration, the division of competences was respected. Based on this brief justification, which was ostensibly in line with ECJ case law and spelled out in the recitals of the directive, the Commission decided to maintain the proposal.
Opposition to the initiative was also manifest in the Council, where the governments of sending countries raised similar concerns about the need for – and effects of – new regulation. 80 Nonetheless, during interinstitutional negotiations on the proposal, even more space was carved out for the Member States to impose social policy requirements on posting. 81 Besides reaffirming the national competence on pay and clarifying the calculation of remuneration, the directive as adopted expanded the list of core rights to include the conditions of workers’ accommodation as well as allowances for travel, board and lodging expenditures. It also created room for Member States to apply to collective agreements to posted workers, even if these have not been declared universally applicable. The threshold for long-term posting was reduced from 24 to 12 months, with a possible extension of six months. And while the provision on subcontracting was removed, the directive reinforced the rules on equal treatment for temporary agency work, adding an information obligation for the user undertaking in respect of its terms and conditions of employment and the ‘double posting’ of workers in third Member States.
In the final vote on the directive, Croatia, the UK, Latvia and Lithuania abstained, while Hungary and Poland voted ‘No’. 82 The latter governments maintained that the amended directive would ‘render the legal institution of posting itself meaningless’ and ‘be highly detrimental to the competitiveness of the Union as a whole.’ 83 They subsequently brought two cases before the ECJ, challenging the legal basis of the directive and alleging a discriminatory and disproportionate restriction on the freedom to provide services. In both disputes, however, the Court ruled that all pleas had to be rejected, and the EU legislature had legitimately exercised its regulatory power. 84
Work-life Balance Directive: controversies on paid leave and flexible working time
A proposal for a Work-life Balance Directive 85 (WLBD) was published by the Commission in April 2017 and adopted as Directive (EU) 2019/1158 of the European Parliament and of the Council of 20 June 2019 on work-life balance for parents and carers and repealing Council Directive 2010/18/EU. 86 As part of a broader initiative to address the balance between work and private life for parents and carers, this legal act would replace Council Directive 2010/18/EU of 8 March 2010 implementing the revised Framework Agreement on parental leave concluded by BUSINESSEUROPE, UEAPME, CEEP and ETUC and repealing Directive 96/34/EC, while building on Council Directive 92/85/EEC of 19 October 1992 on the introduction of measures to encourage improvements in the safety and health at work of pregnant workers and workers who have recently given birth or are breastfeeding, as well as the broader acquis on equal treatment in employment and occupation. The draft Directive provided for a right to paternity leave of at least ten working days and carers’ leave of at least five working days a year; fourth months of non-transferable parental leave; adequate allowances (amounting to at least the level of payment for sick leave); time off from work for urgent family reasons on grounds of force majeure; a right to request flexible work arrangements for care purposes (for carers and workers with children up to the age of 12); and protection from dismissal when exercising these rights.
According to the Commission, its proposal quite evidently complied with the subsidiarity principle. First of all, it built on an existing acquis, which ‘illustrate[d] the common agreement that EU-level action in this area is necessary.’ 87 To achieve the goals of work-life balance and gender equality in employment, however, further EU intervention was needed to overcome the hesitance of the Member States, which would be concerned about short-term costs. Left to their own devices, the Commission worried, ‘variations between Member States with regard to length and the generosity of the conditions could result in an unbalanced level of rights, an unequal protection for EU citizens across the EU and differences in the functioning of labour markets.’ 88
Yet the legislative initiative immediately evoked an adverse response from policymakers in several Member States. A number of legislatures flagged their concern with the subsidiarity compliance of the envisaged law, in a way that suggested an understanding of the subsidiarity principle which was at the same time broad, in the range of political concerns that it covered, and narrow, in the policy space that it admitted as open to common EU action. As opposed to the more procedural and proportionality-based arguments raised against the revision of the posting directive, national parliaments here disputed the very possibility of European intervention in this domain.
The Dutch Senate, in its reasoned opinion, found that the subsidiarity principle had been violated because the initiative did not target ‘a transnational problem or a subject that could only be realised by means of measures at [the] EU level’ or ‘that could not be solved by Member States’, suggesting an extremely prohibitive understanding of the comparative efficiency test, in which any effective national action precludes the EU from taking complementary or supporting measures to address a problem. 89 Reflecting a functional understanding of the EU, the Dutch parliamentarians argued that it was ‘unclear what the benefit [was] of equalising regulations on work-life balance between the Member States, in view of the functioning of the internal market.’ Indeed, the Senate and the Chamber of Deputies noted that ‘the social domain […] is a national concern.’ Work-life balance was regarded as an issue of individual choice and national regulation, which reflect ‘specific national, social and cultural considerations.’ 90
The Polish Sejm expressed a similar criticism of the proposed legislation, framed in more proceduralist terms. Although the Commission had formulated a legitimate objective, it had failed to substantiate why this goal could be better achieved at the EU level; the Sejm contended that, in principle, ‘EU regulation should not interfere with national legal measures which effectively ensure an adequate level of family protection in the area of parental care and child upbringing.’ 91 Because the Commission had not considered Polish law in its memorandum, it had violated the subsidiarity principle. In other words, the question of comparative efficiency need not even be asked, and EU legislative action would be precluded if the social policy of one Member State would – by its own judgment – pass the national sufficiency test.
The Senate of Poland shared the view that the WLBD would ‘excessively interfere in the legal systems of Member States by imposing numerous detailed regulations, without taking into account the existence of national systemic solutions aimed at achieving the same purpose.’ 92 Moreover, it identified a breach of subsidiarity in terms of the horizontal balance between state and society: EU measures which assumed that women's participation in the labour market would translate into gender equality, e.g. increasing the non-transferable component of parental leave, would interfere excessively in the private lives of citizens and working families.
While the Danish Folketing did not submit a reasoned opinion to the Commission, it sent a ‘political opinion’ expressing its concerns about the WLBD. A majority of parties found that ‘an extension of the existing rights must be adopted on the basis of agreement between the social partners or nationally in the individual Member States’ and that there would be ‘no need for any further EU regulation of parental leave and earmarking of this.’ 93 The far-right Danish People's Party appealed to a democratic understanding of subsidiarity, arguing that ‘[d]ecisions regarding labour market conditions and questions concerning the distribution of leave in families should be made as close to the citizens as possible in order to give consideration to national differences and individual wishes.’ 94 Like the Dutch Senate, the Red-Green Alliance maintained that work-life balance was ‘not a matter of a cross-border problem which cannot be solved by national Parliaments’ and that the Commission had failed to explain ‘why it [was] necessary that the regulation of rights to leave should take place at European level.’ 95
In the Council, various governments expressed reluctance about the proposal. Most of their objections related to the legal clarity and effectiveness of the WLBD. But some ministers and officials indicated that they were anxious about the compliance of the proposed legislation with the subsidiarity principle, echoing the complaints raised by their parliaments – even if their concerns were not always couched in the language of subsidiarity. The strongest opposition was reserved for the provisions on an adequate minimum payment for leave-taking, which many governments refused to countenance. It was argued that these measures ‘encroached on Member States’ competence for defining and organising their social security systems and might impact on their sustainability.’ 96 A number of delegations were more fundamentally sceptical of any move to broaden the material rights of workers at the EU level; like their parliamentary counterparts, they perceived this as a national prerogative.
When the Permanent Representatives Committee (COREPER) reached an agreement on the WLBD, several derogations had been included and legal obligations, including on allowances, were curtailed. Some governments nevertheless voted against the final directive, ostensibly on the grounds of a subsidiarity problem. The Dutch Minister for Social Affairs explained that his government had voted against ‘not because of its content, but mainly because of the principle that it [should] be left up to the Member States and [is] not a European responsibility,’ in line with the negative opinion of the national legislature. 97 Hungary explained that, in family policy, subsidiarity entailed that ‘it is only appropriate [for the EU] to designate basic principles for the Member States, leaving the necessary room for manoeuvre for them to set up the tools, timing and priorities at national level.’ 98 Arguably, this was precisely what the WLBD sought to do, i.e. to set minimum standards while leaving the exact means to Member States, but it is nonetheless indicative of the recurrent deployment of the subsidiarity principle – framed in such a way as to put a brake on EU social policymaking and safeguard ‘sovereign’ prerogatives.
Transparent and Predictable Working Conditions Directive: delimiting workers’ rights
The Commission published its proposal for the Transparent and Predictable Working Conditions Directive (TPCWD) in December 2017, 99 enacted as Directive (EU) 2019/1152 of the European Parliament and of the Council of 20 June 2019 on transparent and predictable working conditions in the European Union. A report for the Commission found that Council Directive 91/533/EEC of 14 October 1991 on an employer's obligation to inform employees of the conditions applicable to the contract or employment relationship had failed to keep up with flexibilisation and other changes in the world of work, leading to gaps in the coverage of workers. 100 The draft Directive proposed to widen the scope of the law to include everyone with an employment relationship of at least eight hours a month (instead of eight hours a week) or undefined working hours; include information requirements on probation, training entitlement, overtime, variable working schedules and social security; reduce the deadline for information provision to the first day of employment or whenever working conditions are modified; limit probationary periods to six months; prohibit exclusivity clauses; establish a minimum predictability of work, including a reference period within which one may be required to work; and guarantee that employers provide legally required training cost-free.
The Commission again reasoned that the prior acquis demonstrated the importance of EU intervention in this area. Further regulation would be needed since ‘action solely by Member States in response to new and non-standard forms of employment would not necessary have the same level of protection in terms of transparency and predictability and would risk increasing divergences between Member States with potential competition on the basis of social standards.’ 101 Next to ensuring such a level playing field, EU action would have the potential to ‘create a momentum for Member States to advance together towards better outcomes, supporting upwards convergence’ and to ‘contribute to improving internal market mobility.’ 102
The TPWCD proposal elicited critical reactions among national parliaments, however. The concern about subsidiarity was expressed most acutely by the Swedish Riksdag. Its reasoned opinion recounted that Sweden's relationship to the EU was founded on the understanding that ‘questions in the area of labour law that lack a clear cross-border dimension should, in the first place, be handled at the national level.’ 103 Since the directive dealt with an issue of employment conditions that was regarded as primarily domestic, the new material rights and the definitions of ‘worker’ and ‘employer’ that it laid down would be overly intrusive. The Riksdag echoed the concerns expressed by the social partners that the TPWCD would disturb Sweden's established model of industrial relations, ominously concluding that ‘[t]he consequences of the proposal [would be] hard to oversee.’ 104 Subsidiarity, then, was not understood as an organising principle based on the relative efficiency of regulatory intervention at different levels of governance, but as a firewall that would safeguard national traditions against perceived competence creep.
The Austrian Bundesrat also assessed the proposal negatively, although it steered clear of submitting a reasoned opinion. It found the scope of the directive to be too broad, in particular as the alleged coverage of self-employed workers ‘would intervene very heavily in the respective national legal orders.’ 105 The new information requirements and reversal of the burden of proof were also considered to leave too little latitude for national considerations and specificities. Like their Swedish counterparts, the Austrian legislators maintained that their established system of social partnership made EU solutions superfluous. The Bundesrat considered the horizontal obligation on Member States to ensure that individual and collective agreements would be amended in line with the directive to be an ‘unusual and not acceptable’ constraint on the prerogative of national political and social representatives. 106
In the Council, the draft TPWCD was ‘considered by most delegations in line or to some extent in line with the principles of proportionality and subsidiarity,’ but the proposal met with considerable resistance from some quarters, which questioned the added value EU action or its compatibility with diverse national laws and practices. 107 While most governments expressed support for the objectives of the directive, there was dissent over specific provisions, e.g. with regard to the coverage of civil servants and deadlines for the provision of information to workers. These concerns were (again) met by introducing derogations and less stringent provisions that created more room for manoeuvre for the Member States in implementing the directive.
Yet some governments expressed deep-seated objections that matched the scepticism of the Riksdag: the law sought to impose obligations in a sensitive area of policymaking while riding roughshod over national specificities and the autonomy of the social partners. The main object of their critique was the proposal to include a definition of ‘worker’, based on the criteria of the ECJ case law, for the purposes of the directive. This provision was regarded by Germany, the Czech Republic and other Member States as an attack on the ‘basic structures’ of a policy domain that was especially coloured by ‘tradition’ and ‘historical roots’ and therefore best left to domestic decision-making. 108 Acknowledging the sensitive nature of this debate, the Presidency referred the issue to COREPER – where division persisted – and then to the Council, where the critics were successful: the contested article was deleted, while retaining an ambiguous reference to ECJ jurisprudence in the recitals. In parallel, other provisions were also amended to provide more latitude to governments and employers.
Subsidiarity, sovereignty and the limits of EU social policy
As the cases above make clear, there is barely any common ground to be found in the array of meanings and interpretations that different institutional actors have attached to subsidiarity as an organising principle in debates over EU social policy. The Juncker Commission, on the one hand, has quite consistently invoked subsidiarity from an efficiency-oriented perspective, arguing that a minimum of social protection and a ‘level playing field’ for competition in the internal market can only be secured through European legal standards. In doing so, it has relied on a justification of social regulation through minimum harmonisation that Davies rightly criticised as tautological, since the Member States by definition cannot achieve such uniformity on their own. 109
The use of subsidiarity by national opponents of EU social regulation, on the other hand, diverges from this technocratic, legalistic reading of the principle. The EWS, as recounted above, invites national legislatures to assess the justifiability of EU action in a certain domain without attending to the content of the measures proposed. In line with what earlier studies have found, 110 however, ‘activist’ parliaments have been less concerned with this limited notion of subsidiarity than with two other aspects of legislative proposals: their proportionality and the conferral of competences in the respective area. In this view, the implications of EU social legislation would not be commensurate to the scope of the problem identified, or such action would interfere with responsibilities (e.g., pay) that had been agreed to remain a national prerogative. This conceptual stretching of subsidiarity may well be inevitable, since it is difficult to imagine how one could assess the admissibility of norm-setting without considering its effects and potential implications for future decision-making. Nevertheless, as I elaborate in the concluding section, the messiness of the concept should raise questions about its value as a yardstick for policy debates.
Substantively, this second approach to subsidiarity, i.e. as a defence mechanism against EU-action to promote social rights, differs not only from the efficiency-based conception on which the Commission relies, but also from a democratic interpretation of the principle. The crucial point of reference for such arguments about a subsidiarity deficit, as presented above, is not so much the (lack of) political support for EU social and labour policies among the demoi of Member States, but rather the primacy of the nation-state and its interests in this domain of government. This approach resembles a sovereigntist or ‘state-centric’ conception of subsidiarity, which prescribes, as Føllesdal has noted, ‘immunity for the local units from interference by more central authorities.’ 111 Pride of place is accorded to raison d’état, national tradition and historical fact, which are threatened by the transfer of power to a higher (i.e. supranational) level of governance. Such (over-)regulation would entail a problematic loss of autonomy for the national polity, which has a legitimate interest in – and prerogative over – its favoured model of social protection and employment law (even, some would argue, when its citizens are posted as workers in another Member State!).
In addition to this vertical dimension of state-centric subsidiarity, the critics of EU social policy and regulation have sometimes problematised the horizontal aspect of such action. Conservative political actors, in particular from Central and Eastern Europe, have complained that European legislation would violate the ‘natural’ authority of families over their private lives and the ‘free hand’ of entrepreneurs in running their businesses. From this perspective, subsidiarity would imply that the Union must defer ‘to their status as naturally occurring groups with primary and presumptive responsibility for making decisions for the good of the group.’ 112
Conversely, adherents of this understanding of subsidiarity obstinately refuse to consider whether more effective pathways to reaching social rights-related objectives might exist at the supranational level. The very prerogative of the Member States, after all, renders superfluous the assessment of any potential benefits of common action in light of shared aims and values. This attests to what Streeck already observed in the 1990s: ‘what is conveniently forgotten is the right […] of the “higher level” of governance under the classical concept of subsidiarity to ensure that the outcomes of self-regulation [i.e. by the Member States or business] are compatible with general objectives and norms of social justice.’ 113 Spicker also worried about the restrictive role that subsidiarity would play in debates on EU social policy, as it came to be understood ‘as a limitation which leaves decisions to the member states.’ 114 ‘If this were all that the principle of subsidiarity implies,’ he maintained, ‘it would be equivalent to sovereignty.’ 115
Indeed, such a trend has been consolidated in the way in which national parliaments and other opponents of EU social regulation have leveraged the subsidiarity principle. For Spicker, this meant that subsidiarity ‘loses its moral force’, since ‘one has to be suspicious of a moral principle which seems to limit the actions of the [EU] and almost no one else.’ 116 I concur that this sovereigntist (mis-)interpretation of subsidiarity has little argumentative value for debates on the ‘social dimension’ of the EU. The premise of sovereignty, as Cahill has pointed out, does not enable different parties ‘to engage in a rational reflection about which of the two options is the more reasonable or morally defensible or kindly course of action.’ 117 It only invites tautological reasoning: national actors, waving the banner of subsidiarity, end up arguing that a (sub-)domain of social policy is off-limits for EU intervention simply because the Union is not supposed to interfere there. Contrary to its own promise, subsidiarity fails to offer any positive justification as to how we should choose between different (local, national or supranational) routes to social protection and rights promotion. Rather than providing a useful argumentative device in debates about multilevel social governance, the principle is invoked as a rhetorical firewall by actors who seek to contain the Union's role in providing social rights in the first place. The next section offers some concluding thoughts on what the arguable flaws of the subsidiarity principle portend for future debate on EU social policy.
Conclusion: searching for a democratic and subsidiary social Europe?
Despite widespread pressures and proposals for further integration, the involvement of the EU in the field of social and labour policy has met with deep-seated political objections. Since social policies have contributed to a sense of identity, community and democratic legitimacy that underpins European nation states, national policy-making elites have been reluctant towards the transfer of regulatory and redistributive capacities in the social domain. To navigate this policy minefield and strike the difficult balance between diverse interests and institutional commitments, various academics, officials and politicians – notably including Jacques Delors – have proposed putting the principle of subsidiarity front and centre in the debate on EU social policy and law.
But to what extent has subsidiarity been able to fulfil its promise as an effective organising principle for social policies in the EU, propelling and slowing their (de-)centralisation when and where needed? As illustrated above, usage of the subsidiarity principle in recent policy debates has scarcely exceeded that of a convenient rhetorical instrument, whose analytical content is at the very least equivocal. On one side, the Commission invokes it to justify legislative initiatives by reference to objectives that – given their premise of legal uniformity – can only be effectively achieved through supranational intervention in the first place. In response, sceptics of EU social regulation appeal to the same principle to legitimise their opposition to such measures, asserting the disproportionality of the latter or the national prerogative over social rights and labour law.
Is there anything to be salvaged from the conceptual remains of subsidiarity once we write off its technocratic and sovereigntist manifestations? Bartl argues that a more democratic concept of subsidiarity could help address the social and democratic deficit of EU governance. 118 This demands an explicit recognition that ‘[d]efining at what level a task is better accomplished is primarily a political problem,’ as Dehousse put it. 119 We would no longer need to search for functional, ‘objective’ grounds that tell us whether the EU should act in a matter, obviating the need to ‘proceduralise’ subsidiarity via obscure legal or bureaucratic mechanisms. Such questions must rather be resolved through public debate, in particular in the salient domain of social policy. In this light, Bartl writes, democratic subsidiarity ‘may become a forum’ to broaden and deepen EU policy debates, serving to ‘re-politicise the “forgotten” normative concerns as well as the numerous “assumptions” which have taken the place of a political debate on societally salient matters.’ 120
At the same time, it remains unclear how much such a (re-)reading of subsidiarity would contribute to the ongoing debate on social Europe. This is ultimately a question of priorities: why must we put our argumentative eggs in the basket of subsidiarity? Given the semantic confusion and unproductive debates which the principle has occasioned, it appears to be a Herculean task to reconstruct a useful yardstick out of its conceptual debris. Conversely, there are other ideas and approaches that have proven their value in stimulating debate on the organisation of policy in a multilevel social Europe. One is the idea – invoked by Bartl, Scharpf and others – of democratic legitimacy, which must be conceptually separated from subsidiarity. It implies asking not only whether the pursuit of certain policies at any level of governance reflects popular preferences across the Union, but also more fundamentally whether such a policy arrangement resonates with social norms, values and ambitions in a recognisable fashion. 121 These considerations can then be fleshed out with normative and functional lines of reasoning about the propriety of EU action in specific spheres of social policy and law, e.g., as set out by Vandenbroucke and his colleagues. 122
The less optimistic conclusion, then, is that political, academic and policy debates on social Europe have little to gain from the subsidiarity concept. Subsidiarity has never been up to the job of guiding such intricate debates, unlike what Delors foresaw when he suggested that the principle would enable the European Community to reconcile the supranational protection of social rights through law and collective agreement with the ‘full respect of existing diversities’ in national traditions. 123 Three decades on, his expectation that it would encourage policymakers to strike the right balance between local, national and European involvement in the social domain seems to have been overly rosy. The potential of subsidiarity to offer more than an opportunistic defence for or against EU social policy and labour law has been overestimated. Despite attempts to imbue it with shared meaning, the subsidiarity principle remains little more than an empty shell, thrown around in a dialogue of the deaf. It is so conceptually thin that the search for a social Europe that is truly ‘subsidiary’ can only end in disillusionment.
Footnotes
Acknowledgements
An earlier version of this paper was presented at the EUI Social Investment Working Group in April 2022. I would particularly like to thank Francesco Corti for his comments.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
