The General Framework or Primacy of the Social Dimension (Principles)
The Lisbon Treaty, the Viking and Laval Judgments and the Financial Crisis: In Search of New Foundations for Europe’s ‘Social Market Economy’
The aim of these texts is to consider the relationship between the economic and social dimensions of European integration in the light of a number of recent legal and institutional developments, including the Viking and Laval judgments,1 the changes to EU law made by the Lisbon Treaty and the unfolding financial crisis. Viking and Laval, and the subsequent case law clarifying and extending those rulings, radically altered the nature of the relationship between social policy and internal market law by asserting that national labour law rules were liable, in and of themselves, to distort competition within the internal market, and as such had to be justified by reference to a strict test of proportionality. Prior to Viking and Laval, strong national labour law systems, setting standards above a basic floor of rights guaranteed by a combination of Treaty provisions and Directives, had been not just accepted, but actively encouraged as providing a counterweight to the effects of market integration. With the emergence of open coordination methods, differences in levels of regulation between systems had been seen as providing a basis for experimentation in the social policy field. By associating the application of the labour laws of the Member States with the concept of distortion of competition,Viking and Laval undermined these approaches to social policy and threatened to initiate a race to the bottom between national systems.2
The Lisbon Treaty does not amend the core social policy provisions of the EC Treaty, now contained in Title X of the TFEU, but it does make a number of changes which could have an impact on the evolution of European labour law in the aftermath of Viking and Laval. Three in particular are relevant: (i) a new reference to ‘social market’ goals in the objectives of the EU; (ii) a restatement of the competences of the Union and the Member States in the social policy field; and (iii) greater recognition of the legal effects of fundamental social rights within the EU legal order.3
These changes are already starting to have an impact on the way the Court is approaching the issue of balancing social policy and internal market freedoms post-Viking and Laval. There are some signs that economic freedoms will no longer be placed in a clearly hierarchical position with regard to social rights, and that a more flexible approach to the issue of proportionality may be possible in future.4 How far the Court will row back from Viking and Laval is not easy to assess. The significance of the Lisbon Treaty changes will be addressed here by placing Viking and Laval in the context of the long-run evolution of EU law and in particular by considering the significance of successive Treaty amendments for the EU’s ‘economic constitution’.
The idea of the ‘economic constitution’ refers to the ways in which a given polity views the relationship between the organs of the state (including the legal system) and the market (or economic system).5 These texts will argue that there is no single, predominant conception of the state–market relationship in EU law and that, instead, several different versions of this idea are present in the corpus of law made up by the Treaties, secondary legislation, case law and so on. In evolutionary terms, the EU moved from an ‘ordoliberal’ conception of the economic constitution in the 1950s to a ‘neoclassical’ one in the 1990s and 2000s.6 Ordoliberal elements remained, but, as exemplified by the case of the Treaty’s freedom-of-movement provisions, were increasingly reinterpreted by reference to a neoclassical logic. The Viking, Laval, Rüffert7 and Luxembourg8 judgments embody this approach in context of the application of free movement rules to social policy. As such they constitute, without doubt, a marginalisation of labour law within the EU legal order. However, EU law contains a third, emergent conception of the state–market relationship, which will be referred to here as a ‘human-developmental’ one, from which assistance can be drawn in countering the logic of Viking and Laval. It will be argued in these texts that advancing a developmental conception of EU social policy will prove more useful, as a strategic approach, than attempting to limit Viking and Laval by reverting to the ordoliberal logic of the Rome Treaty, particularly in the conditions created by the current financial crisis.
The Shift From ‘Ordoliberal’ To ‘Neoclassical’ Conceptions Of The Market In EU Law
Ordoliberal Theory, the Rome Treaty and the Original ‘Social Market’ Economy
An ordoliberal conception of the state–market relationship is one in which the legal system has the task of establishing the conditions for a functioning market order.9 This process involves the active construction, through the legal order, of an economic-political environment within which market relations can develop. The legal system must do more than recognise and enforce, through private law, the property and contract rights which form the subject matter of economic exchange. Going beyond private law, the legal system must be deployed so as to ensure that competition is maintained by, for example, regulating monopolies and cartels through the mechanisms of antitrust law. At the same time, ordoliberalism does not envisage a state directed economy. A complementary role of the legal system is to constrain the arbitrary use of governmental power. Thus ordoliberalism requires limits on state ownership of industry and constraints on direct government intervention in, or planning of, economic processes.
The ordoliberal model of the legally-constituted market order heavily influenced the approach taken in the Treaty of Rome to the process of economic integration.10 It was given concrete expression in the Rome Treaty’s provisions on competition and freedom of movement. It also informed the concept of ‘distortion of competition’, which is referred to in the Spaak Report (1956) on which the Treaty was, in part, based.11 The creation of a market free of ‘distortions’ implied a transnational trading regime in which the factors of production could move across borders in response to economic signals, to the same extent as they would do within the unitary jurisdictional space of individual Member States. It also presupposed markets which were not divided or partitioned along national lines. This could have been read as implying a need for the harmonisation of regulations (including those of labour law) affecting the factors of production, with the goal of establishing uniform costs within the common, or later, internal market.
That view was rejected at the time of the Rome Treaty, in a way which expressed acceptance of an active role for the state in the social policy field, but at national level.
The acceptance of a role for social policy within the broadly ordoliberal orientation of the Rome Treaty turned on a particular analysis of the relationship between means and ends in the construction of a transnational market order. Functional convergence of regulatory norms and practice, and hence of costs, was expected to follow from the removal of the more visible obstacles to inter-state trade. This approach made formal convergence unnecessary: on the one hand, it ruled out a comprehensive European labour code or transnational floor of rights in social policy; on the other, it preserved Member State autonomy in the social policy field. Member State autonomy over social policy was not simply the result of political obstacles to agreement between the Member States, initially or at successive stages of enlargement, on the content of a European labour code. There was indeed no political consensus but, at the same time, the decision to leave competence in social policy mostly in the hands of the Member States was presented in structural terms as a part of the integration process. The Ohlin Report (1956), commissioned from the ILO prior to the Rome Treaty, argued that strong labour law systems at national level were understood to be necessary in order to provide a counterweight against the dislocating effects of the economic growth that was expected to follow from market integration. The ‘strength of the trade union movement in European countries and … the sympathy of European governments for social aspirations’ were, in this sense, necessary preconditions for the creation of a common (or internal) market.12 While the Member States might not have agreed on the content of a common labour code, they were all in agreement-up to the mid-1970s-on the need for strong labour laws at national level. They also supported complementary labour market institutions in respect of social insurance, and governmental promotion of full employment through ‘Keynesian’ techniques of demand management, during this period. In addition, most of the Member States had made constitutional guarantees of one kind or another regarding social rights in the immediate postwar period. The commitment of the drafters of the Rome Treaty to strong labour market institutions at national level can be seen in their willingness to accept the need for social policy harmonisation in those areas where, for one reason or another, they did not expect a spontaneous process of ‘levelling up’ to occur.13 This principle of selective intervention at transnational level was the origin of the Treaty of Rome’s provisions on equal pay, which were to have far-reaching significance in the development of European Union labour law, as well as those relating to working time, whose influence turned out to be minimal.
The use of national labour law regimes as a counterweight to transnational economic integration gave expression to one of the elements of ordoliberal thought, namely the idea that the market was not a natural order but was constituted by, and embedded in, legal and political processes. In the 1950s and 1960s, if not all then at least some adherents of ordoliberal economic theory were able to accept the case for minimum wages, collective bargaining and social insurance as mechanisms complementing the operation of product and financial markets.14 Wealth creation-and societal wellbeing in a wider sense-were seen as ultimately dependent on the harnessing of economic resources through private enterprise, and on a market-oriented division of labour. Social cohesion was understood, in large part, to be the consequence of economic growth, but for that growth to be sustainable, it was also accepted that elements of the social fabric had to be maintained. This was the essence of the ‘social market’ philosophy which influenced German economic policy debates of the 1950s and 1960s. The ‘social market’ idea promoted market forces and required limits to state intervention which were not compatible, for example, with macroeconomic policies aimed at achieving full employment through the management of demand which the Member States mostly followed at that time. To that extent it operated more as a model than as a representation of the practice of the Member States, Germany included.15 Nevertheless, in so far as EU labour law could be said to have had a guiding principle during this period, the concept of the ‘social market’ economy provided it. The Keynesian model of redistributive and growth-oriented economic policy which held sway at Member State level was not embedded at EU level.
The Impact of Viking and Laval on Member State Autonomy in Social Policy
The decisions in Viking and Laval involve a rejection of the compromise struck at the time of the Rome Treaty between economic integration at transnational level and strong labour law institutions at national level. Prior to Viking and Laval, the Court had accepted the broad terms of that compromise, and had translated them into legal principles which more or less held the deregulatory tendencies of free movement law at bay. In Rush Portuguesa16 it had endorsed the principle of the territorial effect of national labour laws in the face of the Treaty-based provisions protecting the free movement rights of service providers. In Albany17 it insulated national-level labour law and collective bargaining from the full application of competition policy rules. These decisions, and others like them, were seen by their critics as carving out artificial and unjustified exemptions to the operation of internal market rules, but they were arguably consistent with, and followed from, the institutional architecture of the Rome Treaty. That architecture was nevertheless coming under increasing strain as a result of the changes set in train by the Single European Act and the Maastricht and Amsterdam Treaties in the 1980s and 1990s. The ‘old’ ordoliberal economic constitution was never formally revoked, and its core provisions on free movement and competition remained the same. However, the assumptions on which the old constitution had been based were being eroded by the deepening of the internal market programme and by the process of economic and monetary union. Viking and Laval are part of a wider realignment of the Court’s case law which took place alongside this process.
The common idea underpinning Viking, Laval and the subsequent case law in the same line is that national-level labour law rules are capable of constituting a distortion of competition within the internal market and, as such, must be justified by reference to a strict test of proportionality.
Although ostensibly expressed in terms which are familiar from free movement case law, the principle put forward in Viking and Laval goes far beyond the Court’s earlier decisions on the application of internal market law to social policy. Neither of the two tests conventionally relied on in free movement cases-the older and limited ‘discrimination’ test and the more recent, and more interventionist, ‘market access’ or ‘restriction test’-was clearly satisfied in these cases.18 There was no discrimination on the grounds of nationality in either case. In Viking, the reflagging of the vessel and not the nationality of its owner was the issue. In Laval, the collective agreements which the unions were seeking to enforce were intended to apply to all employers engaged in construction work on Swedish territory, on an equal basis. On one view, the Lex Britannia was discriminatory in its treatment of overseas collective agreements, but even this is not clear,19 and the Court’s broad interpretation of Article 56 (ex-49) did not in any event turn on this aspect of the Lex Britannia.
In its earlier decision in the free-movement decision of Säger,20 the Court had moved away from a requirement of discrimination on grounds of nationality in free-movement cases, and it relied in Viking and Laval on this wider concept of a restriction on market access to trigger the application of Article 56. It is difficult to see, however, that even the market access test was met. The issue in Viking was not access to Estonian markets, but-an entirely different proposition-access to Estonian jurisdiction as a low-cost alternative to the Finnish one. Similarly, the service provider in Laval had access to the Swedish construction market; its complaint was that it could not operate in Sweden while simultaneously taking advantage of the less stringent labour laws that would have applied had the work been done in Latvia.
In what sense can it be said, in the context of posting, that the application of the normal labour law rules of the ‘host state’ in these cases constitutes a ‘restriction’ on cross-border trade? There are two possibilities. One is that the restriction arises from differences in labour law rules and practices between the home and host states (a comparative standard); the other is that labour law rules, once they exceed a certain (unclearly defined) level of strictness, are inherently restrictive (an absolute standard). Laval and Rüffert have mostly been interpreted as turning on the application of a comparative test: from this point of view, the critical issue was that Latvian and Polish labour laws were less ‘restrictive’ than, respectively, Swedish and German laws. The problem with this reading of the decisions arises from the lack of any obvious connection between the establishment of a company (which determines its nationality for free-movement purposes) and the labour law rules which it observes in relations with employees and other labour market actors. A company established in a given Member State may, as a matter of law and practice, observe the labour law of that state in its relations with employees who work there, but this does not follow from the company’s establishment; it results from the choice of law made by the parties to the employment contract, supplemented by the territorial effect of any mandatory labour law rules of the state concerned. If a company’s establishment does not determine which labour law rules apply to it in the absence of a supposed ‘restriction’, the converse is also true: the applicable labour law rules do not determine the establishment. Thanks to the Court’s Centros21 decision and the increasing move away from the ‘real seat’ principle at Member State level, in favour of freedom of incorporation, a company’s main seat of business or operations can, if its controllers choose, be detached from the jurisdiction in which it has its establishment.
It is also relevant to bear in mind that the parties whose free-movement rights had apparently been infringed in Laval and Rüffert were not even established in the low-cost states whose laws they were trying to access. The industrial action in Laval was targeted against a Swedish company which was a subsidiary of the Latvian parent which appeared as a party before the Court, while the company whose bankruptcy triggered proceedings in Rüffert was a German construction firm that had employed a Polish company as one of its sub-contractors. For these various reasons, there is a case for seeing Laval and Rüffert as turning on an absolute, as opposed to a comparative, notion of what constitutes a ‘restriction’. Labour law rules which impose undue regulatory costs on service providers are, in this view, seen as inherently restrictive. To trigger the free-movement principle, there must be a transnational dimension to the dispute. But this is very easily satisfied, as the facts of Laval and Rüffert make clear: a home-state parent of a host-state subsidiary (Laval) and a host-state client of a home-state supplier (Rüffert) can equally well invoke it. It is not out of the question that the principle applied in those cases will operate to disallow the application of labour law rules to service providers established in the host state and predominantly operating there, where they can show that they are competing in that state with overseas providers who can take advantage of lower cost labour law regimes to which they have some connection. The Court was invited to make just such a ruling in the Omalet case, which, however, it declined to take up on that occasion.22 Viking can also be understood as a case based on an absolute as opposed to a comparative approach to defining a ‘restriction’. On one view, a restriction arose in Viking because of differences between Estonian and Finnish labour law and collective agreements which made the former more attractive to the employer. On another view, the Court was articulating a rule according to which strike action becomes subject to review when it weighs on any decision of the employer which has transnational elements. Relocations involving cross-border flows of assets or resources clearly enough come into this category, but once the principle is extended, as it was in Viking, to virtual relocations, such as the reflagging of a vessel, it becomes potentially much more open ended and its link to the original rationale of the free movement rules becomes tenuous, because there is no sense in which there is either discrimination or a denial of market access.
The purpose of raising these points is not simply to highlight the doctrinal murkiness of Viking and Laval, but to stress how radically they depart from the ordoliberal assumptions which informed the foundation of the EU legal order, and hence by extension, from the social market model, at least in its original ordoliberal sense. This is so whether we take a narrow view of the decisions as turning on a comparative approach to defining what a restriction is, or a broader one based on an absolute standard.
If the former interpretation is correct, the decisions are incompatible with the Spaak Report’s rejection of the argument that a common (or internal) market requires uniform rules in the area of labour law (or any other area of regulatory policy). If the latter interpretation is correct, they are incompatible with the idea of national-level labour law regimes as a necessary counterweight to the effects of economic integration.
The Ohlin and Spaak reports were written over 50 years ago, and there may be a case for seeing them as having diminished significance in today’s environment. There is indeed a sense in which the analysis they advanced has been superseded by changing circumstances, but not in a way which is capable of justifying the Court’s approach in Viking and Laval. In making the argument against a comprehensive European labour code, the Ohlin Report emphasised the inherent flexibility provided by separate national currencies and adjustable exchange rates. In the same way, the Spaak Report argued that differences in exchange rates could be expected to cancel out the potentially distorting effects of cross-national differences in nominal wage rates. Where the flexibility provided by exchange rates was thought to be absent, as in the case of the advantage enjoyed by employers in states with weak sex discrimination legislation, both Spaak and Ohlin accepted the logic of harmonisation of labour laws at transnational level, hence the incorporation of the equal pay principle into the Rome Treaty. By the same logic, the advent of the single currency, and the restrictions placed on national economic policymaking by the stability pact and associated measures, constitute a case for strengthening, not weakening, national-level labour law regimes, if necessary through further harmonisation.23
1 Case C-438/05 ITF v Viking Line ABP  ECR I-10779; Case C-341/05 Laval v
Svenska Byggnadsarbetareförbundet  ECR I-1176
2 S Deakin, ‘Regulatory Competition after Laval’ (2008) 10 Cambridge Yearbook of European Legal Studies 581.
3 P Syrpis, ‘The Treaty of Lisbon: Much Ado… But About What?’ (2008) 37 Industrial
Law Journal 219.
4 C Barnard and S Deakin, ‘European Labour Law after Laval’ in I Ulasiuk (ed), Before and After the Economic Crisis: What Implications for the ‘European Social Model’? (Cheltenham: Edward Elgar, 2011).
5 C Joerges and F Rödl, ‘Informal Politics, Formalized Law and “Social Deficit” of European Integration: Reflections after the Judgments of the ECJ in Viking and Laval’ (2009) 15 European Law Journal 1.
6 R Salais, ‘Employment and the Social Dimension of Europe: What Constitutive Convention of the Market?’ in R Rogowski, R Salais and N Whiteside (eds), Transforming European Employment Policy: Labour Market Transitions and the Promotion of Capability (Cheltenham: Edward Elgar, 2011)
7 Case C-346/06 Dirk Rüffert v Land Niedersachsen  ECR I-1989.
8 Case C-319/06 Commission v Luxembourg  ECR I-4323.
9 See C Joerges and F Rödl, ‘“Social Market Economy” as Europe’s Social Model?’ EUI Working Paper LAW No 2004/8 and C Joerges, ‘What is left of the European Economic Constitution?’ EUI Working Paper LAW No 20045/13, for accounts of ordoliberal thought and analyses of its relevance to the development of EU law.
10 Joerges and Rödl, ‘Social Market Economy’ (n 9).
11 Rapport des Chefs de Délégation aux Ministres des Affaires Étrangères (‘Spaak Report’) (Brussels, Comité Intergouvernemental créé par la conférence de Méssine, 1956).
12 ‘Social Aspects of European Economic Cooperation’ (‘Ohlin Report’) (1956) 74 International Labour Review 99, 112.
13 Spaak Report (1956) p 60.
14 Joerges and Rödl, ‘Social Market Economy’ (n 9) pp 14–17.
15 Ibid pp 18–19.
16 Case C-113/89 Rush Portuguesa  ECR I-1417, para 18.
17 Case C-67/96 Albany  ECR I-5751.
18 See Deakin, ‘Regulatory Competition’ (n 2).
19 Joerges and Rödl, ‘Informal Politics’ (n 5) p 16.
20 Case C-76/90 Säger  ECR I-4221.
21 Case C-212/97 Centros Ltd v Erhvervs-og Selkabsstryrelsen  ECR I-1459.
22 Case C-245/09 Omalet v Rijksdienst voor Sociale Zekerheid, Judgment of 22.12.2010.
23 See next text.