The fundamental distinctions of ordoliberalism: ordering policies (on the legal framework) vs regulatory interventions (in the economic process)

Against the simplistic view of what they called the ‘paleoliberalism’ of the Manchester School,Footnote ³⁵ ordoliberals distinguish between the ‘framework’ (Rahmen) and the ‘economic process’ (Wirtschaftsprozeß).Footnote ³⁶ The former is the set of social structures and legal institutions that underpin and regulate the economic order. The latter comprises the market mechanisms through which individual decisions are formed and coordinated, ie, competition and the price system. The aim is to systematically develop institutions and rules (the framework) that ensure that competition and price coordination (the process) cannot be distorted by either public authorities or by private interests.

This is the core idea of an economic constitution. To use the ordoliberal terminology, the state must engage in ‘ordering policies’ (Ordnungspolitik),Footnote ³⁷ ie, it must establish the legal framework under which market mechanisms can be fully deployed. According to the liberal tradition, however, it is in principle not the role of the state to intervene in the economic process, otherwise it runs the risk of distorting free competition. But the (early) German ordoliberals believe that the market, as a human institution, is prone to error and abuse. They therefore accept that the state should intervene in the process to correct such malfunctions: it then engages in ‘regulatory interventions’ (Prozeßpolitik). However necessary they may be, these interventions must remain ‘dynamically consistent’ with the market model and remain subsidiary. This means that interventions must be limited in scope and time, and they must be undertaken only when adjustments to the rules are not sufficient.Footnote ³⁸ They must be both targeted at the market malfunction, and temporary, until the deeper problem can be resolved by revising the legal framework to bring it back into line with the needs of the market.Footnote ³⁹

The content of the ordoliberal economic constitution: constituent and regulatory principles

The distinction between acting on the framework and intervening in the process also applies to the content of the economic constitution. It is divided into two categories: the ‘constituent principles’ (konstituierenden Prinzipien), on one hand; the ‘regulative principles’ (regulierenden Prinzipien), on the other.Footnote ⁴⁰ Among the former is a Grundprinzip: the need for a fully competitive price system (vollständige Wettbewerb) as a mechanism for the functioning of the economy. The second ‘constituent principle’ is equally important: monetary stability. This is a sine qua non, if a price coordination system is to remain viable and efficient over time. Then there are the principles of free access to markets, consistency of economic policy, private ownership over the means of production, freedom of contract and, finally, the unlimited liability of economic actors.

These seven ‘constituent principles’ are the necessary conditions of the ‘private law society’.Footnote ⁴¹ However, they may not be sufficient to ensure a stable and fair market, since private powers may always seek to circumvent competition in order to achieve ‘dominant positions’. It is therefore necessary to provide regulatory principles that will allow the State to intervene, but in a way that is consistent with the market economy. Eucken identifies five principles.Footnote ⁴² Three seem less important: (very moderate) tax progressivity; the fight against certain negative externalities; and the correction of different ‘counter-cyclical’ behaviour. But Eucken’s first and last principles are crucial: an anti-monopoly policy must be implemented to combat anti-competitive behaviour; and a relatively ‘automatic’ and price-stability-oriented monetary policy must be pursued.

Notwithstanding this quest for automatic governance, some ordoliberals, such as Röpke and Rüstow, insist on the possible need for direct intervention in the event of profound structural changes in order to bring about the inevitable adjustments – and thus force a rapid return to equilibrium.Footnote ⁴³ This is one of the original features of ordoliberalism, which echoes Eucken’s idea of ‘dynamic consistency’ mentioned above. In their conception of the economy, the ‘real’ market is a social and legal institution. This institution must of course be constantly refined in order to come as close as possible to a hypothetical ‘perfect’ market. But the ‘real’ market never corresponds to the ‘perfect’ market. The latter remains an ‘abstract and hypothetical standard’ that can never be fully achieved.Footnote ⁴⁴ What matters most is not so much that markets behave perfectly, but that economic actors internalise the logic of competition – if necessary, under the constraint of public institutions. In other words, they should act as if they were in a competitive situation, ie, as if markets were functioning ‘perfectly’. Leonhard Miksch, in particular, has systematised this idea through the concepts of ‘complete competition’ (vollständige Konkurrenz) and ‘as-if politics’ (Als-Ob Politik).Footnote ⁴⁵ The idea is that public authorities can force private agents to act in a way that is consistent with market competition. From this perspective, these public authorities are in a sense the ‘guardians’ of the competitive order, which they impose on the ‘real’ market. As we shall see below, this perspective is particularly fruitful for analysing the interventions of European institutions in the wake of the Eurozone crisis.Footnote ⁴⁶

The ordoliberal epistocracy or the central role of scientific expertise in designing and implementing the legal framework of the economy

In order to impose a particular economic order in the long run, it is necessary to understand how it works. Hence, the central role of ‘expertise’ in the ordoliberal vision and the defence of independent technocratic institutions, especially in the areas of competition and monetary policy. On this second point, ordoliberals were initially less insistent on the (statutory) independence of central banks – although some of them, such as Röpke, openly advocated this option.Footnote ⁴⁷ However, it is clear from the outset that in their view price stability requires the neutralisation of monetary policy, which can no longer be used as an economic instrument at the service of public authorities, in order to reduce unemployment or improve growth. Instead, it should be entrusted to technical experts with the necessary knowledge.Footnote ⁴⁸

However, the centrality of expertise is not limited to the implementation of the economic constitution once it has been adopted, but also extends to its design.Footnote ⁴⁹ Böhm expressed this very clearly in the introduction to his famous dissertation, where he explicitly endorsed the physiocrats’ ambition to ‘translate the doctrinal edifice of economics into the language of law’ in order to ‘proclaim natural law as the law of the state in the field of economic life’.Footnote ⁵⁰ Enlightened by scientific knowledge, the public authorities must therefore both organise the market framework ex ante and intervene ex post to regulate malfunctions caused by the concrete behaviour of economic agents. In other words, the ‘rule of law’ should be extended to the economy – or, more precisely, the ‘rule of economics’ should be imposed by law.Footnote ⁵¹

The market as ‘catallaxy’ and its social consequences

Hayek’s starting point is the critique of planning, which is said to be at the root of the crisis of the liberal order in the interwar period. He contrasted the constructivist notion of ‘plan’ with the concept of ‘order’. Like the ordoliberals, Hayek used this word as one of his main Kampfbegriff against the socialist thought. Unlike the latter, however, it is not an order imposed from above by a great architect (Taxis), but an order that comes from below, spontaneously founded and self-organised according to abstract general laws (Kosmos). This is based on his redefinition of the competitive market as an (albeit limited) process of discovery and knowledge, which he calls ‘catallaxy’.Footnote ⁶² The market here represents the networking of the knowledge of all economic agents, which is crystallised in the prices of all goods and services exchanged on the market. The real market may not be ‘perfect’ (in the neoclassical sense of atomistic competition with total transparency of information), but it is necessarily ‘optimal’. No ‘external intervention’ by any public or private body could ever perform better than the real market, since its knowledge is, in this theory, inherently inferior to the sum of the knowledge accumulated in the market. Hence the rejection of the idea of ‘social justice’, which requires ‘artificial’ intervention from outside the market (by public authorities) and would therefore distort the catallactic process of the market.

Besides, this process of discovery through competition would also apply to the legal phenomenon: the general and abstract rules of just conduct would emerge from a process of natural selection through competition. All the judge would have to do is observe the prevailing norm of conduct in order to recognise and gradually shape the spontaneous law of society.Footnote ⁶³ In this respect, the ‘common law’ would capture the true nature of this Nomos. By contrast, ‘law’ in the modern sense of ‘legislation’, as the expression of the conscious will of the legislator (Thesis), would only be (conceptually and historically) secondary to Nomos. Its function would be to implement these general and abstract rules, and only very rarely to modify or adapt them. But the combination of constructivist rationalism (even more developed in the civil law systems) and ‘unlimited democracy’ would have undermined this adaptive function and led to more and more intervention in the general and abstract process of the market.

The content of Hayek’s (neo)liberal model constitution

Against this trend, Hayek was firmly convinced of the need to revive the idea of the rule of law and to reopen the institutional project of (neo)liberal constitutionalism.Footnote ⁶⁴ Apart from some puzzling institutional proposals (voting rights only from the age of 45, a single term of 15 years for members of the legislative assembly, etc), his ‘model constitution’ is most interesting for its content. Even if the Nomos were supposed to emerge spontaneously, Hayek considered it necessary to ‘creat[e] an adequate framework for a functioning competitive market’.Footnote ⁶⁵ To this end, he made five key proposals.

First, competition law should be enshrined in the constitution. However, it should focus only on the market barriers and foreclosure and should be enforced by the courts – not on the general fight against all monopolies by independent administrative authorities, as favoured by the ordoliberals. Second, the public sector should be fiscally constrained by two mechanisms: the principle of ‘balanced budgets’ (the State cannot spend more than it takes in) and the division of fiscal powers. Third, public services should be opened up to competition, ie, liberalised. Fourth, this liberalisation should be extended to the monetary sphere, since allowing the private sector to offer alternative currencies would oblige the state to guarantee price stability so that it remains ‘attractive’. Finally, the sovereignty of the State would have to be ‘dismantled’ through (internal) federalisation and integration into supranational structures (internationalisation or external federalisation).Footnote ⁶⁶ In this way, the rules of the different legal systems would compete with each other, and economic agents would be able to go where the legal framework is most appropriate, ie, the most business-friendly. This would lead to the selection of the most ‘efficient’ rules. In other words, the authorities’ quest for economic attractiveness would act as a self-disciplining mechanism.

Hayek’s ordoliberal compatibility and the ‘Hayekian turn’ of ordoliberalism

Hayek’s views are here ‘compatible and even complementary’ to the ordoliberal theses.Footnote ⁶⁷ Admittedly, their starting points are different: the Freiburg School was developed from an internal reflection on liberalism, whereas Hayek’s theory is first and foremost a defence of liberalism against its external enemies. And of course, there are divergences,Footnote ⁶⁸ especially on competition policy. But in the end, they are clearly united in their denunciation of the politicisation of the economy and the overburdening of parliamentary democracy by social demands – and in the need to devise an appropriate constitutional framework for the market order.Footnote ⁶⁹ Similarly, where the legal constructivism of the ordoliberals barely conceals the persistence of an economic natural law, Hayek’s ‘catallactic spontaneism’ not only reveals an (implicit) immanent economic natural law, but also a (paradoxical) rationalist constructivism.

This probably explains the ease with which the second and third generation of ordoliberals achieved the ‘Hayekian turn’ initiated by Ernst-Joachim Mestmäcker.Footnote ⁷⁰ His work spans more than half a century,Footnote ⁷¹ but can be summed up (broadly speaking) in two basic propositions. First, Mestmäcker used an evolutionary re-reading of competition to relativise the ordoliberal competitive interventionism inherent in Als-Ob Politik.Footnote ⁷² Second, he integrated the idea of an economic constitution into the supranational structure proposed by Hayek in order to develop an analytical framework adapted to the concrete experience of European integration.Footnote ⁷³

The Virginia School: from public choice to constitutional political economy

The origins of the Virginia School can be traced back to the publication in 1962 of James Buchanan and Gordon Tullock’s seminal book The Calculus of Consent, in which the two authors sought to elucidate the ‘logical foundations of constitutional democracy’.Footnote ⁷⁹ This book marked a veritable revolution in the way political decision-making was conceptualised – hence the other name of the Virginia School: ‘Public Choice Theory’. According to Buchanan and Tullock, an economic analysis of the state would provide a systematic and rational explanation of both the establishment of the constitutional ‘rules of the game’ by the consent of citizens and the way in which public decision-makers, the administration and interest groups operate within this constitutional framework.

At the first level, the setting of constitutional norms in a democracy can be analysed, in economic terms, as an aggregate function of citizens’ choices aimed at ensuring the long-term stability of the rules of the game while minimising the impact of these rules on their private sphere and individual freedom – and thus as a ‘meta-norm of coordination’ of individuals. Between a unanimity rule (too difficult to achieve) and a simple majority rule (too vulnerable to demagoguery), the qualified majority rule would be the rational choice.

At the second level (the ‘public choice’ within the constitutional framework), Buchanan and Tullock applied neoclassical economic rationality to public decision-makers in order to analyse the functioning of the state as a ‘political market’. Since politicians would seek only to maximise their individual utility, ie, their power, and since in a representative democracy, power is won through elections, elected representatives would tend to increase social spending to meet the expectations of various interest groups, especially at the end of their term, on the eve of new elections. Through the mechanism of public debt, this spending will only have a constraining effect with a certain time lag and could therefore also backfire on electoral rivals who would come to power at a later date.Footnote ⁸⁰ Moreover, the economic analysis of public choices would apply in the same way to the (bureaucratic) public administrationFootnote ⁸¹ : civil servants would tend not to seek economic efficiency but to increase their operating budget, in order to derive maximum benefit from it (ease of execution of their mission, extension of their scope of action, etc).

The coup de force of the Virginia School was to turn the ‘market failure’ argument against the advocates of public interventionism: the ‘political market’ would have the same defects … but worse. Two remedies were proposed: the introduction of competition mechanisms into public administration (New Public Management); and the restriction of the discretion of political decision-makers by constitutionalising rules that would force them to remain efficient. It was this second dimension, first introduced in the field of taxation,Footnote ⁸² that Buchanan and his colleagues began to explore in greater depth from the mid-1980s onwards. This new research agenda is known as constitutional economics Footnote ⁸³ or constitutional political economy.Footnote ⁸⁴

In fact, Buchanan and his disciples have sought to give neoliberal economic constitutionalism a ‘democratic’ basis by renewing the understanding of the idea of a ‘social contract’ on the basis of Hayek’s intuitions about the ‘epistemic uncertainty’ into which individuals are plunged when they enter into relationships. But this does not prevent them from laying down in advance the constitutional rules on which citizens are supposed to agree. These ‘rules of the game’ cover several areas of the economy. The first concerns taxation and, more specifically, the need to limit the taxing power of the state.Footnote ⁸⁵ Various constitutional reforms are envisaged, including: the introduction of a qualified majority voting for the adoption of tax measures; the setting of maximum tax rate limits; the capping of total tax revenues and expenditures by linking them to certain shares of national product or income. Another mechanism had long been proposed: fiscal federalism.Footnote ⁸⁶ Echoing Hayek, he defends the idea of giving economic agents the power to ‘vote with their feet’,Footnote ⁸⁷ ie, to go where the (para-)fiscal standards are the most favourable, in order to force the authorities to reform the legal framework to improve its attractiveness.Footnote ⁸⁸ Hence his interest in European integration, which he saw as an opportunity to create such an effective ‘constitutional’ framework.Footnote ⁸⁹

Setting fiscal limits is not enough, however, if public policymakers can use debt indefinitely. Here too, Buchanan soon pointed out the problem of chronic fiscal imbalances, ie, the profligacy of public authorities that shifts the burden of debt onto future generations.Footnote ⁹⁰ To prevent such a situation, the possibility of a budget deficit should be removed from majority rule. In other words, the constitution should be amended to include a balanced budget clause.Footnote ⁹¹ Conversely, the central bank should be prohibited from financing public deficits by creating money (and thus inflation). Monetary policy should therefore be constitutionalised, as Buchanan argued in 1981Footnote ⁹² and reiterated during the Great Recession of 2008.Footnote ⁹³ This means that the definition of the quantity of money in circulation should be enshrined as one of the constitutional rules of the game, rather than as a parameter that can be adjusted as the game progresses. In addition, the independence of the monetary authority should be guaranteed in order to set monetary policy once and for all, as empirical studies would confirm.Footnote ⁹⁴

The new classical macroeconomics: ‘rules rather than discretion’

Buchanan was implicitly referring here to the writings of the ‘new classical macroeconomics’. Based on Milton Friedman’s work on monetary stabilityFootnote ⁹⁵ and John Muth’s model of rational expectations of economic agents,Footnote ⁹⁶ the members of this school of thought sought to defend the thesis of the efficiency of the competitive economy and insisted on the necessary neutralisation of money as a precondition for the stable and efficient functioning of the market. Broadly speaking, all their work can be seen as an attempt to refute the scientific validity (and hence the political relevance) of the expansionary fiscal and monetary policies inherited from the Keynesian paradigm.Footnote ⁹⁷ In their view, economic agents will anticipate the fact that, on one hand, public borrowing is a kind of deferred taxation and, on the other hand, the increase in money supply will ultimately lead to inflation and thus to no increase in their personal net wealth. Consequently, and contrary to what Keynesianism claims, they will (logically) refuse in advance to increase their spending and thus (artificially) ‘stimulate’ consumption.

In addition to these two specific criticisms, there is the more general argument of the time inconsistency of (fiscal and monetary) discretionary policy, developed by Kydland and Prescott. The very nature of discretionary policy decisions implies that they can be reversed (as a result of democratic change or simply by a change in the opinion of the decision-makers). This creates a very high risk of incoherence, while these economic policies would need a minimum of time to be effective. The solution would therefore be to impose binding rules on politicians in order to limit their discretion. This is expressed in the title of Kydland and Prescott’s famous article, which has become the motto of new classical macroeconomics: Rules rather than Discretion. Footnote ⁹⁸ Kenneth Rogoff adds that monetary policy must be made autonomous by entrusting it to an independent and ‘conservative’ central bank, ie, one geared to price stability.Footnote ⁹⁹ Finally, Alberto Alesina and Lawrence H Summers then provided empirical support for the (now contested)Footnote ¹⁰⁰ thesis of the economic efficiency of independent central banks.Footnote ¹⁰¹

Despite different methods and assumptions, the new classical macroeconomics inherited from Friedman and Buchanan’s constitutional political economy are therefore very similar in terms of the normative implications they draw from their respective theories.

The ‘great synthesis’ of neoliberal economic constitutionalism by Viktor Vanberg

Viktor Vanberg transformed this convergence into a ‘great synthesis’ of neoliberal economic constitutionalism. Born in Germany, he only came to ordoliberalism after a detour through the theories of Hayek and Buchanan.Footnote ¹⁰² As early as 1981, in Mannheim, he attempted to unify different neoliberal paradigms and explain the social order by combining the (individualist) spontaneism of the former with the (individualist) contractualism of the latter. At the University of Virginia, he worked intensively on integrating into a single ‘meta-theory’ the theses of Buchanan and Hayek,Footnote ¹⁰³ and then of the first ordoliberals.Footnote ¹⁰⁴ But it was in 1994, at the University of Freiburg im Breisgau and in the chair of political economy formerly occupied by Eucken and Hayek, that he really brought together the three main strands of neoliberal economic constitutionalism.Footnote ¹⁰⁵

In a nutshell: with the idea of Ordnungspolitik, ordoliberalism took the crucial step of rethinking the legal framework of liberalism in order to find the right ‘rules of the game’; Friedrich Hayek’s ‘catallactic’ understanding of market functioning tempered the ‘interventionist’ tendency of ordoliberalism; and James Buchanan’s constitutional economics provided a ‘democratic’ argument for the need to ‘constitutionalise’ the market order – and to opt for ‘rules rather than discretion’, according to the motto of new classical macroeconomics. Beyond the epistemological complementarities (but also divergences) between these currents, Vanberg insists on their common ‘normative dimension’,Footnote ¹⁰⁶ which includes the points mentioned above: the constitutionalisation of the competitive market orderFootnote ¹⁰⁷ ; the independence of central banks with a strict mandate to ensure price stabilityFootnote ¹⁰⁸ ; the limitation of fiscal powers and the discipline of public spending through the balanced budget ruleFootnote ¹⁰⁹ ; and the federalisation and internationalisation of the state to ensure normative competition.Footnote ¹¹⁰ As Vanberg summed up in a popularisation paper in 1997, ‘there are two ways of forcing the state to act as an honest broker: constitutionalise the most fundamental economic decisions […]; institutionalise elementary economic decisions by entrusting them to independent bodies that have no subordinate links with politicians’.Footnote ¹¹¹

Against this ‘normative’ background, one might ask whether this hard core of neoliberal economic constitutionalism – what Agustín José Menéndez has called ‘neo-ordo-liberalism’ –Footnote ¹¹² does not overlap to a large extent with the European integration project.

Fundamental freedoms of movement: an instrument of Europeanisation and liberalisation

The common (and then internal) market is at the heart of the European project and of the idea of a ‘European economic constitutionalism’.Footnote ¹¹⁶ It was on the basis of the fundamental freedoms of movement that the Court of Justice established the autonomy and primacy of European law over the Member States.Footnote ¹¹⁷

However, the type of market organisation initially remained partly open: the first Treaties appeared as a (relative) compromise between the German (ordo)liberal conception and a more interventionist one, sometimes described as an ‘indicative planning’ à la française.Footnote ¹¹⁸ Hence the criticism of the early ordo-liberals, who were concerned about the effects of this dirigiste tendency.Footnote ¹¹⁹ The European Coal and Steel Community (ECSC) aimed to create a common energy market through planning. Euratom extended this idea, while the European Economic Community (ECC) pushed the economic integration in a more liberal direction. Although it included some interventionist aspects (Common Agricultural Policy, relatively high external tariff, etc), the hard core of the EEC was the common market project, based on a customs union, a policy of ‘undistorted’ competition and fundamental freedoms of movement, under the supervision of a semi-technocratic body (the ‘High Authority’, renamed the ‘Commission’ in 1957) and a supranational court, the ‘Court of Justice of the European Communities’ (renamed the ‘Court of Justice of the European Union’ in 2009).

While the interventionist parts were gradually mitigated, the common market underwent a remarkable development. In the 1960s, the Court of Justice used it to give the new European legal order a truly coherent economic doctrine. As the French liberal economist Jacques Rueff put it shortly after the end of his term as a judge at the ECJ (1958–1962), the Court succeeded in subjecting the interventionist provisions to a more general economic logic in line with ‘Soziale Marktwirtschaft’ (ie, ordoliberalism).Footnote ¹²⁰ This ‘liberal’ interpretation also serves to justify the Court’s claims to autonomy, as the Van Gend en Loos judgement of 5 February 1963 perfectly illustrates. In that case, the autonomy of European law in relation to national law was based on the hard core of the common market: the prohibition of customs duties. Moreover, the direct effect recognised for these provisions and the subjective right conferred on individuals gave the Court a particularly effective means of compelling States to comply with their obligations.Footnote ¹²¹ The Court then extended the claims to autonomy by asserting the ‘primacy’ of European law,Footnote ¹²² even over national constitutional law,Footnote ¹²³ and by claiming the ‘power of the last word’, ie, the competence to decide on conflicts of competence (Kompetenz–Kompetenz).Footnote ¹²⁴

In turn, the alleged ‘primacy’ served to further extend the scope of the common market rules, in particular the prohibition of obstacles to the free movement of goods, both at the ‘fiscal’ level (customs duties and taxes having equivalent effect) and at the ‘material’ level (quantitative restrictions introducing an import or export quota or measures having equivalent effect). In this respect, the concepts of ‘goods’ and ‘taxes having equivalent effect’ were already interpreted very broadly in 1968.Footnote ¹²⁵ Less than a year later, the ECJ clarified the absolute nature of the prohibition, ie, irrespective of the minimal nature of the imposition or the absence of a protectionist objective.Footnote ¹²⁶

In the 1960s and early 1970s, however, this movement remained restrained. The ECJ appeared reluctant to use these new (self-empowered) legal means to further ‘expand’ the realm of economic freedoms. This was likely due to the lingering ‘interventionist’ tendency among Western European public decision-makers, who remained (partly) committed to the ‘Keynesian’ referential. Indeed, until the mid-1970s, there were several attempts at defining not only a common European social policy,Footnote ¹²⁷ but also a common reflationary policy in response to the first oil shock.Footnote ¹²⁸ However, almost simultaneously, the shift that ended the Trente Glorieuses was underway.Footnote ¹²⁹ The increased focus on monetary stability led to a revival of free-market thinking and a greater rejection of state interventionism. In this context, the fundamental freedoms of movement became key drivers for dismantling the ‘barriers’ to the market created by state regulation. In other words, while monetary stability began to take precedence over the objectives of growth and full employment, an extensive interpretation of economic freedoms emerged.

In the Dassonville judgement of 11 July 1974,Footnote ¹³⁰ the ECJ gave an extremely broad definition of the concept of ‘measures having equivalent effect to quantitative restrictions’ (MEERQ). The criterion is the obstacle to trade, even if this obstacle is only potential or indirect. It can be justified only on the basis of one of the grounds listed exhaustively in Article 36 of the Treaty establishing the European Economic Community (TEEC) (now Article 36 of the Treaty on the Functioning of the European Union (TFEU)). In the same vein, during the momentous months of 1978/1979,Footnote ¹³¹ a few weeks after the European Monetary System was finally agreed upon,Footnote ¹³² the Cassis de Dijon judgement of 20 February 1979Footnote ¹³³ confirmed this trend: the Court’s approach is based on market access (rather than the less stringent criterion of discrimination).Footnote ¹³⁴ As a result, the scope of the MEERQs has been extended almost indefinitely – even though at the same time new justifications have been declared admissible (‘mandatory requirements in the public interest’). The ECJ has thus moved from a principle of equal treatment to ‘mutual recognition’.Footnote ¹³⁵ It is no longer enough for Member States to treat economic operators equally; they must also apply the rules of the country of origin of the goods. This has created a process of ‘competition between jurisdictions’: in order to attract companies (and their capital), it may be in the (short-term) interest of States to reduce the legal requirements for trade; and since these companies benefit from freedom of establishment,Footnote ¹³⁶ they are likely to choose the State that offers the best conditions; this ‘law shopping’ then leads to a downward adjustment of legal regimes, which in turn risks creating a ‘race to the bottom’.Footnote ¹³⁷ In other words, the principle of mutual recognition has an almost automatic ‘deregulatory effect’Footnote ¹³⁸ – as Germany had argued in Cassis de Dijon to justify the validity of the contested measureFootnote ¹³⁹ … but in vain, since the Court has simply ignored this argument.

It is true that ‘mandatory requirements in the public interest’ to some extent prevent this deregulatory effect. However, it is the Court’s responsibility to weigh the ‘non-economic’ objectives of the contested measure against the market barrier it creates. Through the proportionality test, the judges actually apply a cost-benefit analysis that indirectly and implicitly reflects the metaphysics of efficiency.Footnote ¹⁴⁰ Since the Court are thus seen as the custodian of a hypothetical objective scale of values that would allow the objectives and effects of the contested measures to be rationally weighed,Footnote ¹⁴¹ it logically becomes the legitimate ‘neutral and impartial’ arbiter of the social choices made by the Member States (and their elected parliaments).

Although the Keck jurisprudenceFootnote ¹⁴² seemed for a time to mark a return to the criterion of discrimination in selling arrangements, the Court has been careful to limit the scope of this new categoryFootnote ¹⁴³ – and even seemed inclined to abandon it: sometimes it invoked it without applying it, sometimes it ignored it altogether.Footnote ¹⁴⁴ At the same time, the market access criterion has been extended to the other fundamental freedoms of movement: servicesFootnote ¹⁴⁵ and freedom of establishment,Footnote ¹⁴⁶ workersFootnote ¹⁴⁷ and capital.Footnote ¹⁴⁸ Moreover, the deregulatory dynamic has been further reinforced by the reduction of the ‘purely internal situations’ that were supposed to be excluded from the scope of European law,Footnote ¹⁴⁹ but which the Court of Justice has gradually reintroduced by means of various abstract reasoning.Footnote ¹⁵⁰

It is true that, since the 2000s, this liberalisation of trade through the resolute action of the ECJ (negative integration) has gradually been replaced by the activity of the European legislative institutions (positive integration).Footnote ¹⁵¹ This may have helped to promote certain non-economic concerns,Footnote ¹⁵² but it has also (and perhaps above all) made it possible to combat disparities that persist despite the principle of mutual recognition,Footnote ¹⁵³ in particular barriers that were declared to be justified but which give rise to what economics calls ‘transaction costs’ (information costs, costs of adapting the product or service, etc).Footnote ¹⁵⁴ In this respect, the Court of Justice ensures that harmonisation legislation contributes to the establishment of the internal market – and thus to the liberalisation of trade.Footnote ¹⁵⁵ Hence the early adoption of the ‘new approach’ to technical harmonisation,Footnote ¹⁵⁶ followed by the ‘new governance’ approachFootnote ¹⁵⁷ and the ‘reflexive harmonisation’ approach,Footnote ¹⁵⁸ which aim to promote procedural mechanisms that enable economic actors to develop their own methods of regulation in a context of continuous adaptation. However, these new approaches have complemented rather than replaced the mechanisms of regulatory competition.Footnote ¹⁵⁹ This is also true of the approach adopted for financial servicesFootnote ¹⁶⁰ and of the ‘open method of coordination’ (OMC),Footnote ¹⁶¹ which was developed to identify ‘best practices’ following a comparative analysis of national (or regional, local, etc) regulations.

The fundamental importance of positive harmonisation in trade liberalisation becomes even clearer at two levels. On one hand, important sectors of the economy, which were still partially or totally excluded from the competition induced by the free movement of goods (due to mandatory requirements or overriding reasons in the general interest invoked by the Member States), have been explicitly subjected to liberalisation by EU secondary legislation, in particular in the network industries (energy, telecommunications, public transport or postal services).Footnote ¹⁶² On the other hand, the so-called ‘Bolkestein’ Directive of 12 December 2006 led to a general harmonisation of the rules applicable to the provision of services,Footnote ¹⁶³ which went much further than negative integration had already allowed – even if some exceptions had to be reintroduced in response to the wave of opposition the initial draft had provoked.Footnote ¹⁶⁴

The subordination of fundamental (social) rights to fundamental economic freedoms

Nevertheless, one might well wonder whether this deregulatory dynamic has not been counterbalanced, or at least greatly weakened, by the gradual emergence of ‘social’ considerations at the EU level. Initially, social considerations were almost entirely absent from the EU Treaties,Footnote ¹⁶⁵ except for the idea that economic integration through the common (internal) market would automatically bring social progress to all classes and citizens.Footnote ¹⁶⁶ However, the almost unlimited extension of the scope of internal market law raises the question of the existence of a possible ‘social’ space outside the market that could impede the free movement of goods, capital, services and persons. ‘Social’ objectives have gradually been incorporated into the Treaties or put forward by certain Member States, so that the ECJ has had to articulate the economic freedoms of the internal market with these social concerns.

The Court’s position can be summed up as follows: absolute refusal to exclude certain activities of a ‘social’ nature from the scope of the freedoms of movement,Footnote ¹⁶⁷ but acceptance of social considerations as justification for barriers (protection of workers,Footnote ¹⁶⁸ maintenance of a balanced medical and hospital service accessible to all,Footnote ¹⁶⁹ etc), subject, however, to the conditions of necessity and proportionality. In other words, it is the ‘rule-exception’ scheme that applies here. And via the proportionality test, the ECJ once again assumes the role of arbiter of the social choices made by states and elected representatives. This was by no means self-evident: some authors have argued that it was just as possible to derive a kind of ‘reservation of sovereignty’ in social matters on the basis of Article 118 TEEC (ie, Article 137 of the Treaty establishing the European Community (TEC) and 153 TFEU).Footnote ¹⁷⁰

The EU Charter of Fundamental Rights, adopted in 2001 and made binding by the Lisbon Treaty, could have changed the dynamic, if the ‘social’ provisions (Title IV of the Charter) had not been relegated to the background. Most of the economic and social rights enshrined have only the status of ‘principles’, with virtually no justiciability, rather than full ‘rights’ (Article 52 of the Charter).Footnote ¹⁷¹ Besides, even those that have been given the status of ‘rights’ are considered subordinate to the fundamental freedoms of movement. The landmark Viking and Laval cases on the right to strike and the freedom to provide services are of course paradigmatic expressions of this subordination.Footnote ¹⁷² The defence of a ‘balance’ between the ‘economic purpose’ of the freedoms of the internal market and the ‘objectives pursued by social policy’ of the EUFootnote ¹⁷³ hardly conceals the primacy of the former over the latter. The right to collective bargaining (including strikes) must still be justified against economic freedoms. According to the ECJ, strict control of its necessity, appropriateness and proportionality sensu stricto even includes checking whether the unions have exhausted all means of collective bargaining before taking strike action.Footnote ¹⁷⁴ It is difficult to express more clearly the subordinate relationship between social rights and fundamental economic freedoms.

Three points seem particularly significant here. First, the Court turned the floor (minimum threshold) of social protection for posted workers, derived from Directive 96/71, into a ceiling (maximum threshold) of restrictions on the free movement of their employers.Footnote ¹⁷⁵ Second, the decisions reveal a variable application of the principle of non-discrimination (Articles 18 and 45(2) TFEU), since the Court did not seem to be moved by the existing difference in treatment between posted workers and national workers – on the contrary, it even imposes it.Footnote ¹⁷⁶ Third, it considered that even the protection of human dignity must be ‘reconciled’ with the freedom to provide services.Footnote ¹⁷⁷ The proportionality test shows here once again how much it tends to make the most incommensurable things commensurable … but according to a market economy logic: social rights may exceptionally stand in the way of the market freedoms, but the latter remain the rule.Footnote ¹⁷⁸

Intermediary lessons: from ‘common market’ law to ‘common law’ of the market

This does not mean, of course, that the development of internal market law is the linear story of pure and perfect legal implementation of neoliberal theories. The case law of the Court of Justice and the actions of the Commission (and other EU institutions) can only be understood if other political (neo-functionalism) or legal (primacy vs sovereignty) analytical grids are also taken into account. But there is also an economic rationale that the actors have pursued, albeit implicitly. From the outset, as Jacques Rueff has pointed out, the judges of the Court of Justice have sought to give it unity by ‘bringing [liberal] economic theory into legal doctrine’.Footnote ¹⁷⁹ The Treaties already contained an important decision in favour of market competition, but they nevertheless left open the possibility of moving in a more interventionist direction. The ECJ, supported by the Commission, has endeavoured to close down this possibility.

Despite its slight social temperament, the model of integration of the internal market thus reflects some key tenets of neoliberal economic constitutionalism: improving the legal framework to create the legal conditions for the proper functioning of the market; leaving competition to operate within this framework (‘mutual recognition’ as a means of competition between jurisdictions). Hence the fact that, after the initial reluctance of ordoliberals, some of them have soon defended the ‘eco-constitutional’ nature of the Treaties.Footnote ¹⁸⁰ The acquis communautaire of the internal market is particularly consistent with the idea of ‘integration by framework activities’, ie, by Ordnungspolitik.Footnote ¹⁸¹

Ordinary competition law: controlling cartels, tackling dominant positions

Ordinary European competition law regulates the behaviour of economic agents, ie, ‘undertakings’, on the market in order to ensure that they act ‘fairly’ in their struggle for market share. Its scope therefore depends on the meaning of the term ‘undertaking’. Yet, the ECJ has given an increasingly broad definition to this concept,Footnote ¹⁸³ so that it now includes ‘every entity engaged in an economic activity, regardless of the legal status of the entity and the way in which it is financed’.Footnote ¹⁸⁴ Similarly, ‘economic activity’ is defined as any activity that can be carried out on the market: the fact that profits can potentially be made from the activity is sufficient. The absence of a profit motive is therefore irrelevant.Footnote ¹⁸⁵ Ultimately, only activities relating to the ‘exclusively sovereign’ functions of the State (diplomacy, security, etc)Footnote ¹⁸⁶ or ‘exclusively social’ activitiesFootnote ¹⁸⁷ are, in principle, excluded from the scope of competition law. Moreover, the broadening of the concept of ‘undertaking’ has been accompanied by a significant relativisation of the condition of ‘affecting trade between Member States’. Since 1967, it includes direct or indirect, actual or potential influence on the trade between Member States.Footnote ¹⁸⁸

From an historical point of view, cartel control has been the driving force behind competition law. As early as the first Hallstein Commission, Competition Commissioner Hans von der Groeben’s main objective was to adopt a regulation implementing Articles 85 and 86 of the EEC Treaty in order to establish competition policy as the cornerstone of the new European legal order. The ‘ordoliberal citadel’ of Directorate-General for Competition (DG IV) succeeded in pushing through a particularly ambitious project: Regulation 17/62.Footnote ¹⁸⁹ It established the principle of prohibiting cartels affecting intra-Community trade, unless the Commission had given its express consent after compulsory notification. The Commission also received the monopoly on deciding on infringements of Articles 85 and 86 of the EC Treaty, with the possibility of imposing fines and periodic penalty payments, under the supervision of the ECJ. However, due to administrative difficulties, reforms were introduced in 1965 (Regulation 19/65), to allow for block exemption regulations, and in 2003 (Regulation 1/2003), to replace the ex ante notification system with an ex post legal exception system (more decentralised) where the Commission could focus on the most serious restrictions. At the same time, the Court of Justice gave substance to the idea that the prohibition of cartels covers agreements ‘which have as their object or effect the prevention, restriction or distortion of competition within the internal market’, by applying the cartel control to tacit participation in cartelsFootnote ¹⁹⁰ or professional associations.Footnote ¹⁹¹

However, agreements between undertakings are not the only form of behaviour likely to restrict competition: if an economic operator has enough market power, it can unilaterally impose on its partners, competitors or consumers conditions that it would not have been able to obtain in a ‘normal’ competitive situation. This is why the Treaties also enshrined the principle of prohibiting the abuse of a dominant position.Footnote ¹⁹² The fact that only abuses are punished has its theoretical background in the ordoliberal conceptual distinction between competition ‘on the merits’ (Leistungswettbewerb) and ‘impediment’ competition (Behinderungswettbewerb).Footnote ¹⁹³ This distinction was implicitly taken up by the ordoliberal Ernst-Joachim MestmäckerFootnote ¹⁹⁴ to counter the idea, defended by René Joliet, that only exploitative abuses against a company’s trading partners, suppliers or customers would be prohibited, but not practices designed to exclude or impede the entry and expansion of competitors on the market.Footnote ¹⁹⁵ Against the future judge at the Court of Justice, Mestmäcker defended the ordoliberal conception of the abuse of a dominant position. In particular, he explained the provision in the light of the general Treaty objective of ‘undistorted competition’ enshrined in Article 3(f) TEEC (Article 3(3) TEU read in conjunction with Protocol No 27) and emphasised that the ordoliberal ‘structural’ concept of abuse was in no way in contradiction with the consumer protection defended by Joliet, since the protection of competition per se indirectly ensured the consumer’s freedom of choice. Shortly afterwards, the ECJ ruled in favour of Mestmäcker in the famous Continental Can case.Footnote ¹⁹⁶ The judges upheld the Commission’s decision, prepared by none other than DG IV’s special adviser: Ernst-Joachim Mestmäcker. The result is that the ‘competitive structure’ of the market is ‘constitutionally’ guaranteed.Footnote ¹⁹⁷

But how does the Court understand the concepts of ‘dominant position’ and ‘abuse’? The former was clarified in the United Brands judgement of 14 February 1978, with emphasis on the concept of ‘economic power’ (wirtschaftliche Machtstellung or puissance économique), in a clearly ordoliberal perspective: ‘The dominant position referred to in this article [Article 86 TEEC] relates to a position of economic strength enjoyed by an undertaking which enables it to prevent effective competition being maintained on the relevant market by giving it the power to behave to an appreciable extent independently of its competitors, customers and ultimately of its consumers’.Footnote ¹⁹⁸

The latter, ie, the ‘abuse’ (of a dominant position), was defined in the Hoffmann-La Roche judgement of 13 February 1979, which introduced the criterion of ‘normal competition’.Footnote ¹⁹⁹ In other words, the company in a dominant position cannot behave as it pleases. It must behave ‘as if’ there were competition.Footnote ²⁰⁰ Albeit the terms are translated in English by the expression ‘competition on the basis of quality’, the Court even explicitly enshrined the concept of ‘competition on the merits’ (Leistungswettbewerb or concurrence par les mérites) in the French and German version of the decision.Footnote ²⁰¹ In another case, the ECJ stressed more explicitly the ‘special responsibility’Footnote ²⁰² of undertakings in a dominant position, which ‘may be prohibited from conduct which is legitimate where it is carried out by non-dominant undertakings’.Footnote ²⁰³

This is very much in line with the ‘ethics’ of competition advocated by ordoliberalism.Footnote ²⁰⁴ This ‘ethic of responsibility’ for the structure of competition has been confirmed in the GlaxoSmithKline case.Footnote ²⁰⁵ The ECJ openly rejected the EU General Court (EGC)’s attempt to introduce an ‘effects-based’ approach, according to which the lawfulness of a market practice should be assessed (solely) on the basis of its net effects on consumer welfare. Despite pressure from the Commission, it maintained its traditional position – demonstrating, according to Josef Drexl, ‘a tremendous ordoliberal stability against the influence of newer economic theories’.Footnote ²⁰⁶ Eventually, however, the Court of Justice has shifted its traditional position to partially embrace the so-called ‘more economic’ approach.Footnote ²⁰⁷ It has counterbalanced its defence of the competitive structure of the market with a ‘Darwinian’Footnote ²⁰⁸ or ‘Schumpeterian’ propositionFootnote ²⁰⁹ that Hayek would certainly not have disavowed: ‘not every exclusionary effect is necessarily detrimental to competition […]. Competition on the merits may, by definition, lead to the departure from the market or the marginalisation of competitors that are less efficient and so less attractive to consumers from the point of view of, among other things, price, choice, quality or innovation’.Footnote ²¹⁰ Hence the fact that the Court agreed, in the Intel judgement of 6 September 2017,Footnote ²¹¹ to accept the ‘as-efficient-competitor test’ proposed by the Commission in 2009.Footnote ²¹²

Nevertheless, the concentration of economic power on the market remains a very real concern for the European institutions, even if the Treaty is a priori silent on the subject. In 1971, following an expert report on the subject,Footnote ²¹³ Mestmäcker prepared the Commission’s decision in the Continental Can case mentioned above,Footnote ²¹⁴ before defending his theory of ‘structural abuse’ in a double academic paper.Footnote ²¹⁵ Although the ECJ annulled the Commission’s decision because of inaccuracies in the data used, it endorsed the theory of ‘structural abuse’Footnote ²¹⁶ : since the dominant undertaking by its very existence jeopardises the structure of competition, it can be considered abusive, even without any element of fault.

In doing so, the Court indirectly took a stance on an issue that had been the subject of intense debate among Member States and within the Commission: the legal status of industrial policy in relation to competition law. As early as the 1960s, a conflict had emerged between the (German) ordoliberal defenders of a free and undistorted competition policy and the (French and Italian) proponents of a European industrial policy aimed at creating ‘continental champions’ capable of competing in the global economy.Footnote ²¹⁷ In addition to coordinating national sectoral aid policies for both declining sectors (steel, shipbuilding, textiles) and cutting-edge industries (aeronautics, IT, etc), the idea was to facilitate and encourage the concentration of ‘European’ companies. Although the legal means under consideration were not dirigiste but primarily involved the removal of legal and administrative barriers, the members of DG IV quickly opposed this alternative economic approach. The Commission Memoranda on the Community’s industrial policy of 1967 and 1970 (the latter known as the ‘Colonna Memorandum’) bear witness to this ideological confrontation.Footnote ²¹⁸ An action programme in the field of technological and industrial policy was then adopted in 1973,Footnote ²¹⁹ but its implementation stalled due to disagreements between Member States on the degree of interventionism to be deployed and, above all, the reluctance of certain governments to transfer new powers to the Community.

However, the development of national industrial policies to support sectors affected by the sharp recession paved the way for the revival of the idea of a common industrial policy in 1977. This ambition was reflected in the super-portfolio allocated to Etienne Davignon, responsible for ‘Industrial Affairs, Internal Market and Customs Union’ in the Jenkins Commission (1977–1981) and then for ‘Industrial Affairs and Energy’ in the Thorn Commission (1981–1984).Footnote ²²⁰ Davignon adopted a specific approach based on the close involvement of the Member States and the major European industrialists, gathered in the Union of Industrial and Employers’ Confederations of Europe (known by its French acronym UNICE) and since 1983 in the famous ‘European Round Table of Industrialists’.Footnote ²²¹ In this way, he managed to implement several important concrete sectoral aid measures for declining industries – and to some extent for advanced industries.Footnote ²²² But this ‘golden age’ of Community industrial policy, especially in the steel sector, was also made possible by relying on a specific legal basis: the state of ‘manifest crisis’ as defined in Article 58 of the ECSC Treaty.Footnote ²²³ Yet, this ambitious industrial policy, which involved the creation of a ‘crisis cartel’ and a system of production quotas, clearly interfered with the competition and state aid rules of the ECC Treaty.Footnote ²²⁴ Germany’s reluctance and DG IV’s refusal to extend these exceptional measures to other sectors under article 85 (3) EEC Treaty made it therefore difficult to consolidate this interventionist industrial policy in the long term.Footnote ²²⁵ From the Delors Commissions onwards, the ‘single market’ project was rarely linked to the idea of a common industrial policy but to the dismantling of the non-tariff barriers to the four fundamental freedoms of movement (Single European Act) and to monetary unification (Economic and Monetary Union)Footnote ²²⁶ – although the Commission’s approval of several Important Projects of Common European Interest (IPCEIs) under Article 107(3)(b) TFEU in recent years suggests a revival of the idea of an EU ‘industrial policy’Footnote ²²⁷ in response to the neo-mercantilist turn of the global economic order.Footnote ²²⁸

Together with this tension between industrial policy and competition rules, Member States’ reluctance to cede any part of their economic sovereignty to the Commission helps to explain the failure of the 1973 regulation on the control of abuses of dominant positions, prepared by the Commission but rejected by the Council,Footnote ²²⁹ and the extreme difficulty of reaching a consensus on this issue until the end of the 1980s. The deadlock was finally broken in 1987 when the ECJ confirmed the possibility of controlling mergers, this time on the basis of the prohibition of cartels.Footnote ²³⁰ In order to avoid ex post control, which would have created considerable difficulties, the Council finally agreed to adopt a regulation on concentrations (4064/89).Footnote ²³¹ A system of prior notification and authorisation of concentrations with a Community dimension was introduced. Authorisation was granted on the basis of a ‘competitive assessment’ of the merger, to the detriment of the ‘economic assessment’ favoured by the Chicago School.Footnote ²³² This ordoliberal approach culminated in the General Electric/Honeywell decision of 3 July 2001, in which the Commission enshrined the theory of the efficiency offence: far from offsetting the anti-competitive effects, the efficiencies gained by the new entity as a result of the merger are even more detrimental to competitors and thus to the competitive structure of the market.Footnote ²³³

This ordoliberal approach was mitigated in later years. Amended for the first time in 1997,Footnote ²³⁴ the merger regime underwent a real overhaul in 2004, with the adoption of various regulations (n°139/2004; n°802/2004), guidelines, notices, etc. Without denying the need for a ‘competitive assessment’, the latter was finally counterbalanced by an ‘economic assessment’. The ordoliberal objective of disempowering the market is thus complemented and tempered by the efficiency criterion favoured by the Chicago School.Footnote ²³⁵ But in practice, the Commission has continued to reject mergers that are deemed too detrimental to the competitive structure of the market.Footnote ²³⁶ Moreover, some authors have pointed to a ‘return of ordoliberalism’ in the increasingly dominant digital sector,Footnote ²³⁷ which is characterised by an extremely high concentration of market power. The important decisions in the Google Shopping and Android cases,Footnote ²³⁸ as well as the Platform to Business regulationFootnote ²³⁹ and, above all, the Digital Market Act and the Digital Services Act,Footnote ²⁴⁰ have demonstrated a clear retreat from the ‘more economic approach’ in favour of the defence of the competitive structures of the market.Footnote ²⁴¹

‘Public’ competition law: liberalising public monopolies, controlling State aid, marketising public services

However, one of the main threats to the free play of competition in the market remains the public authorities. The basic rule of ‘public’ competition law is laid down in the Treaty of Rome: the principle of equal treatment between private and public undertakings, which means that the latter are subject to the common rules on competition (Articles 90 TEEC, Article 86 TEC and Article 106 TFEU). Consequently, State aid is in principle prohibited (Article 92 TEEC, Article 87 TEC and Article 107 TFEU). However, the letter of the Treaty did not seem to fit very well with the national realities of the Trente Glorieuses era, which was characterised by an expanding public sector. Therefore, until the 1980s, the existence of public monopolies was de facto tolerated at the European level,Footnote ²⁴² on the basis of the above-mentioned neutrality with regard to the national systems of property ownership.Footnote ²⁴³

But as Keynesianism was challenged by the stagflation of the 1970s, the tolerance of European law towards the public sector began to crumble. First financial transparency requirements were introduced in the early 1980s.Footnote ²⁴⁴ Contested by some Member States, the system was upheld by the Court of Justice.Footnote ²⁴⁵ In the meantime, the ECJ had also developed a broad understanding of the concept of ‘State aid’. On one hand, it includes ‘not only positive benefits, such as the subsidies themselves, but also interventions which, in various forms, mitigate the charges which are normally included in the budget of an undertaking’.Footnote ²⁴⁶ On the other hand, the public interventions are assessed on the basis of their effects and not on the basis of their causes or objectives.Footnote ²⁴⁷ Finally, the concept of ‘State’ has been broadened: it covers all public authorities (federal, regional, local, etc),Footnote ²⁴⁸ but also all (public or private) companies owned by the public authorities,Footnote ²⁴⁹ regardless of their decision-making autonomy.

In the 1980s the Commission also developed the idea that State aid should be controlled according to a specific standard, namely that of the private investor in a market economy. This market economy investor principle was first enshrined in a small number of directives,Footnote ²⁵⁰ then extended by the Commission,Footnote ²⁵¹ and finally endorsed by the Court of Justice in 1986.Footnote ²⁵² For any public measure likely to favour an economic operator or to distort competition, it is thus the ‘economic rationality of the State’s conduct’Footnote ²⁵³ and its ‘leverage effect’Footnote ²⁵⁴ that is assessed by the Commission, under the supervision of the Court of Justice.

As mentioned above, the major public monopolies in the network industries (telecommunications, postal services, railways and energy) began to undergo a dual process of dismantling and liberalisation in the 1990s and 2000s, through the combined action of the Court of Justice ^{Footnote 255} and the Commission.Footnote ²⁵⁶ Of course, this opening up to competition has been accompanied by the maintenance of certain ‘public service’Footnote ²⁵⁷ or ‘universal service’Footnote ²⁵⁸ requirements, which reflect the idea behind the SGEI exception.Footnote ²⁵⁹ In order to manage the ensuing tension between the logic of competition and public service missions, a new legal figure has emerged: the ‘independent regulatory bodies’.Footnote ²⁶⁰ Their proliferation testifies to the change in the role of the State vis-à-vis the market: it no longer acts against the market – opposing it with an alternative logic – but for the market, as a ‘neutral third party’ and guardian of the competitive mechanism.Footnote ²⁶¹

SGEIs are not the only exceptions to the general prohibition of State aid. Block exemption regulations have been adopted,Footnote ²⁶² while guidelinesFootnote ²⁶³ and case law have provided certain criteria for the control of State aid. One of these criteria is necessity, which may take the form of the ‘market failure’ test. According to this test, State aid can benefit from the derogations only if ‘the Commission can establish that the aid will contribute to the attainment of one of the objectives specified in the derogations, which under normal market conditions the recipient firms would not attain by their own actions’.Footnote ²⁶⁴

This does not mean that the European institutions do not tolerate massive intervention by public authorities in the economy, especially in times of acute crisis when the very survival of the system seems to be at stake, such as the ‘Great Recession’ of 2008–2010 or the ‘Great Lockdown’ of 2020–2021. In the first case, however, the Commission maintained a strict line of conductFootnote ²⁶⁵ aimed at limiting aid to the ‘minimum necessary’, establishing ‘safeguards against undue distortions of competition’Footnote ²⁶⁶ and ensuring, during its monitoring, that unviable operators were eliminated or that certain sectors were opened up to competition,Footnote ²⁶⁷ through its power to impose ‘conditionality’ on aid in order for the State to benefit from the derogation.Footnote ²⁶⁸ The 2020 health crisis, on the other hand, was an even more exceptional situation. It led to the creation of a State Aid Temporary Framework,Footnote ²⁶⁹ which has extended the key idea of saving the system ‘whatever it takes’,Footnote ²⁷⁰ while minimising the anti-competitive effects.Footnote ²⁷¹ Only ‘damage due to and directly caused by the COVID-19 outbreak’ benefits from the automatic exemption (Article 107[2][b] TFEU). Otherwise, the Commission maintains its strict control of state aid granted to ‘remedy serious disturbances in the economy’. In doing so, it confirms that the normative primacy of the market over any non-economic objectives of the Member States has been maintained – albeit one might wonder whether this primacy has not been circumvented in practice by the most powerful States, such as Germany and France.Footnote ²⁷²

What is certain, however, is that the normative primacy of the market is reflected in the way public services are understood. Most of these public services have been requalified as SGEIs and have to fulfil several conditions in order to benefit from the exemption provided for in Article 106(2) TFEU.Footnote ²⁷³ The most important of these are the demonstration of a ‘market failure’,Footnote ²⁷⁴ ie, the inability of the market to provide certain services deemed to be of the general interest under satisfactory conditionsFootnote ²⁷⁵ and the fact that the harm to competition must be strictly proportionate.Footnote ²⁷⁶ Since the successive waves of liberalisation of the network industries and the drastic reduction in the ‘exclusive rights’ granted to the legal monopolies to compensate for the less profitable sectors, most SGEIs have been financed by direct subsidies to the undertakings entrusted with a public service mission to offset the costs of providing it. Once again, the ECJ has intervened to limit the Member States’ room for manoeuvre: if they do not want the financing to be classified as ‘State aid’, they must comply with the conditions laid down in the Altmark jurisprudence,Footnote ²⁷⁷ in particular the requirement of objective, transparent and pre-established parameters and the obligation to apply public procurement procedures – or, failing that, the imposition of a ‘comparable efficient undertaking’ test (ie, the ‘typical undertaking, well run and adequately provided with means of transport’). In other words, the ‘market economy investor principle’ is also applied to SGEIs. The Monti-Kroes package, adopted in 2005,Footnote ²⁷⁸ made the regime more flexible for a while, but the Almunia package, adopted in 2011–2012,Footnote ²⁷⁹ brings back stricter control of the ‘efficiency’ of SGEIs.Footnote ²⁸⁰

This has not prevented some authors from developing an ‘economic critique’ of the SGEI regime: clinging to the idea of ‘market failure’, the existing system would be blind to the many ‘government failures’ that could be observed when measured against market standards.Footnote ²⁸¹ Yet national public services have been profoundly reshaped by European law: far from establishing a public domain outside the market, they have been ‘marketised’ through their incorporation into competition and State aid law. The ‘enshrinement’ of access to SGEIs in the EU Charter of Fundamental Rights (Article 36) does not change anything.Footnote ²⁸² The idea remains that the market is the main instrument for allocating goods and services. Imposing a contrary logic based on non-economic objectives can only be done in exceptional and limited circumstances. As mentioned above, the only areas that are genuinely outside the scope of competition law are ‘exclusively sovereign’ functions and ‘exclusively social’ activities, which are of course very strictly defined.Footnote ²⁸³ In contrast to the quasi-hegemonic criterion of ‘economic activity’,Footnote ²⁸⁴ these very rare exceptions testify to the particularly vast territory of the competitive market order.

Intermediary lessons: the constitutionalisation of competition and the market norm hegemony

Driven mainly by the advocates of ordoliberal theories (Müller-Armack, von der Groeben, Mestmäcker, etc) – despite the existence of a competing ‘Keynesian conception of competition policy’ –,Footnote ²⁸⁵ competition law has been implemented and strengthened by the European institutions, which have been particularly favourable to the Ordnungspolitik it presupposes, since it gives a predominant place to the legal framework – which in turn legitimises the existence of the Commission as the ‘guardian of the Treaties’ and of the Court of Justice as the ‘authentic interpreter’ of the Treaties. Admittedly, competition law has evolved,Footnote ²⁸⁶ notably as a result of the internal debates within neoliberalism, between the defenders of the competitive structure of the market (ordoliberals), the adepts of a natural selection of ‘viable’ firms (à la Hayek) and the proponents of ‘economic efficiency’ (Chicago approach). However, its core and fundamental assumptions about the functioning of the market and the place of competition as the central pattern of social organisation remain largely unchallenged.

This metaphysics of competition law has been extended to the domain of the state, through the subjection of public enterprises to the logic of the market, the control of State aid and the recoding of public services into the grammar of ‘services of general economic interest’. The ‘market economy investor principle’ and the imposition of ‘efficiency’ as a criterion for judging interventions motivated by non-economic reasons (social, environmental, etc) demonstrate the overflow of market logic into the entire social sphere, apart from the rare exclusively social activities and exclusively sovereign functions. All this does not mean that State intervention in the economic process is receding, but that it is changing: it is no longer acting because of or against the market, but for the market, as Michel Foucault foresaw.Footnote ²⁸⁷

The responses to the crises of 2008 and 2020 are particularly consistent with the criteria of ‘dynamic conformity’ and ‘subsidiarity’ theorised by Walter Eucken to distinguish between appropriate and necessary ‘regulatory interventions’ and those to be avoided. The conditional authorisation of State aid was targeted at market failures, and, moreover, temporary – until the deeper problem could be solved by redesigning the legal ‘framework’ to adapt it to the new needs of the market. This was the path taken, with the adoption of a ‘European System of Financial Supervision’Footnote ²⁸⁸ and progress towards a ‘banking union’ based on a single supervisory mechanism (SSM) and a single resolution mechanism (SRM).Footnote ²⁸⁹ On the other hand, some have speculated that the COVID-19 crisis may mark the ‘end of the “ordoliberal” model of market regulation’.Footnote ²⁹⁰ However, neither State aid law nor competition law in general has been structurally changed. The European institutions have certainly shown flexibility, but without ever denying or questioning the underlying logic of the competition rules.Footnote ²⁹¹ Faced with an existential crisis, they have accepted massive State intervention in the economy in order to save the market order, not to overturn it. The Member States have not, for example, carried out large-scale nationalisations: their intervention has been marked by the seal of subsidiarity. This is illustrated by the German government’s refusal to activate its constitutional clause on ‘socialisation’ (Sozialisierung; Article 15 of the Basic Law) and its subsequent decision to resort only to ‘emergency state intervention’ (Notverstaatlichung), with the explicit aim of saving the market economy.Footnote ²⁹² State aid control has therefore hardly been removed. ‘Conditionality’ has helped to maintain the criteria of necessity and proportionality of measures, which are still perceived as infringements of the competitive market norm. The latter thus retains normative primacy over the State and its possible non-economic objectives.

It is true, however, that this primacy might mask a shift in the state-market relationship as conceived in the neo-ordoliberal ‘Great Synthesis’, highlighted by the rise of what Daniela Gabor calls the ‘derisking state’, ie the fact that the state increasingly ‘enlists private capital into achieving public policy priorities by tinkering with risk/returns on private investments in […] social infrastructure (schools, roads, hospitals and houses, care homes and prisons, water plants and natural parks) and most recently, green industries’.Footnote ²⁹³ This derisking state certainly implies a greater public intervention in the market process, as it seeks to correct market failures that generate significant risks by redirecting private investment towards strategic (and safer) financial sectors. Conversely, it also involves a deep penetration of the market logic and the search for profitability within the state as well as a commodification of social policies, as evidenced by the increasing use of public-private partnership mechanismsFootnote ²⁹⁴ and the gradual privatisation and financialisation of old pay-as-you-go pension systems.Footnote ²⁹⁵ In this respect, the way in which the European pillar of social rights is being implemented is actually mostly compliant with the market logic – and can therefore hardly be seen as a victory for the proponents of the democratic and social state.Footnote ²⁹⁶

The initial architecture of EMU: monetary neutralisation and fiscal discipline

In order to respond to possible asymmetric shocks, but also to take account of the Member States’ intent to retain a certain degree of fiscal autonomy, two pillars were separated when EMU was set up: the monetary pillar is centralised at the EU level (Article 119[2] TFEU); the economic pillar remains decentralised (Article 119[1] TFEU) and is therefore in principle the responsibility of the Member States, although economic and fiscal policies are coordinated by the EU institutions.

At the first (monetary) level, new centralised institutions have been created to ensure that the single currency is backed by a single monetary policy: the European Central Bank (ECB) and the European System of Central Banks (ESCB) – in practice superseded by the ‘Eurosystem’, ie, the ECB and the national central banks (NCBs) of the countries that have adopted the euro. Moreover, the NCBs are hierarchically subordinate to the ECB (Article 14, §3 ESCB/ECB Statute), so that the ECB and its decision-making bodies (governing council and executive board) are ultimately responsible for the monetary policy in the Eurozone (Article 3 and 8 ESCB/ECB Statute). However, the room for manoeuvre of these institutions is (in theory) limited by a mandate strictly geared to price stability. Support for the Union’s general economic policies (economic growth, competitiveness, pursuit of full employment, etc) is subordinate to this primary objective and should be provided in accordance with the ‘principle of an open market economy with free competition’ and the ‘following guiding principles: stable prices, sound public finances and monetary conditions and a stable balance of payments’ (Articles 119, 127[1] and 282[2] TFEU and Article 2 of the ESCB/ECB Statute). Against this backdrop, the second objective (support for general economic policies in the Union) tends to be interpreted in a monetarist sense, thus excluding any Keynesian expansionary policy.Footnote ³⁰³ The Maastricht Treaty therefore makes price stability ‘the very essence, the raison d’être of EMU’, granting it ‘a pre-eminence unparalleled in legal history’, which has decisive legal consequences because ‘the canonisation of a single economic parameter dramatically transforms the whole hierarchy of normative values not only within the Community system, but also within the constitutional law of Member States’Footnote ³⁰⁴ . Besides, the ECB and the national banks have been granted a very high degree of independence (Article 130 TFEU and Article 7 ESCB/ECB Statute). This independence, which is seen as an essential condition for monetary stability,Footnote ³⁰⁵ is defended on the basis of scientific arguments, both in the Emerson report and subsequently by the ECB.Footnote ³⁰⁶

Once the monetary option of price stability had been established, a coherent legal framework for economic and budgetary policies still had to be created.Footnote ³⁰⁷ To avoid the risk of ‘free riders’, the economic and budgetary pillar of EMU was built around a ‘stability order’ or a ‘stability community’ (Stabilitätsgemeinschaft), as the German Constitutional Court put it at the time.Footnote ³⁰⁸ Article 119 TFEU certainly provides for ‘close coordination of the economic policies of the Member States’ to serve the (economic, social and environmental) objectives mentioned in Article 3 TEU. But once again, this ‘close coordination’ may only be pursued in compliance with the general principle of ‘an open market economy with free competition’ and the ‘guiding principles’ mentioned above.

These principles are conveyed and supported by the specific provisions of EMU, which together form a three-dimensional governance model.Footnote ³⁰⁹ The first dimension covers the individual financial responsibility of Member States, through the triple prohibition on inter-state solidarity (no bail-out clause: Article 125 TFEU), monetary financing (Article 123 TFEU) and privileged access to financial institutions (Article 124 TFEU). As a result, the issuance of public debt now necessarily passes through the financial markets, making them the judge of the ‘soundness’ of national economic and fiscal policies. The second dimension concerns the coordination of national economic policies (Article 121 TFEU) within the ECOFIN Council and the ‘Eurogroup’, through the establishment of ‘broad guidelines for the economic policies of the Member States and the Union’ (BEPGs). The orientation of the structural reforms advocated here testifies to the primacy of the objective of competitiveness and productivity over social or labour issues.Footnote ³¹⁰ Finally, budgetary discipline has been introduced (Article 126 TFEU) and provides for limits on debt (60 per cent) and public deficit (3 per cent) of the Member States (Article 1 Protocol No. 12 on the Excessive Deficit Procedure). This discipline is supposed to be implemented by the Stability and Growth Pact (SGP), through both a preventive arm (surveillance of budgetary positions)Footnote ³¹¹ and a corrective arm (excessive deficit procedure).Footnote ³¹²

The ‘new economic governance’: European institutions as guardians of market discipline

However, the structure of EMU soon showed its weaknesses. In 2002, Germany and France were not penalised for exceeding the 3 per cent deficit limit – leading to considerable tensions between the Council and the Commission, which had to be resolved by the Court of Justice.Footnote ³¹³ In response, a first (minor) revision took place in 2005.Footnote ³¹⁴ But it was not until the Eurozone crisis of 2010–2012 that a thorough reform was adopted. Financial assistance under strict conditionality was introduced, through bilateral loans and various mechanisms (European Financial Stability Facility and European Financial Stability Mechanism), which were then made permanent in 2012 in the ‘European Stability Mechanism’ before being validated by the Court of Justice.Footnote ³¹⁵ At the same time, the ‘Six Pack’ and the ‘Two Pack’ introduced a number of innovations in the coordination and surveillance of national economic and budgetary policies: the European Semester, the excessive macroeconomic imbalance procedure, greater powers for Eurostat, etc.

The result of these reforms is a new, genuinely integrated coordination framework, which can be described as three concentric circles. The first circle concerns the identification and prevention of risks of budget deficits (preventive arm of the revised SGP) and macroeconomic imbalances (new procedure). This level is characterised by a high degree of transparency in the transmission of information, both from the Member States to the Union and from the Union to the Member States – transparency that is crucial for the rating of Member States in the financial markets. This basic preventive core is supported by a second circle, which comprises both the corrective arm of the SGP (excessive deficit)Footnote ³¹⁶ and macroeconomic surveillance (excessive macroeconomic imbalances).Footnote ³¹⁷ The common objective is to force the countries concerned to adopt the ‘structural reforms’ deemed necessary by drawing up ‘corrective action plans’Footnote ³¹⁸ and other ‘economic partnership programmes’.Footnote ³¹⁹ The implementation of these reforms is subject to close monitoring and control, which could in principle lead to the imposition of financial sanctions. The ‘structural reforms’ are those that the State should undertake under pressure from the financial markets, so that the EU institutions are, in a sense, the ‘interpreters’ of the market. Finally, if all these mechanisms are not sufficient to control national macroeconomic and fiscal policies, there is still the possibility of subjecting a Member State, if it is in the euro area, to ‘enhanced surveillance’ (the third concentric circle),Footnote ³²⁰ especially in the case of financial assistance.Footnote ³²¹ The triggering of such ‘enhanced surveillance’ has the effect of imposing a specific surveillance regime,Footnote ³²² which amounts to a veritable tutelage of the State by the European institutions,Footnote ³²³ via the ‘conditionality’ attached to the ‘Macroeconomic Adjustment Programme’.Footnote ³²⁴ This new global and integrated framework for the coordination, surveillance and control of Member States was complemented by the adoption of an additional treaty under international law: the Treaty on Stability, Coordination and Governance (TSCG).Footnote ³²⁵ This ‘Fiscal Compact’ aims to enshrine the ‘golden rule’ into national law (and, if possible, in the constitution) as well as to establish an automatic correction mechanism in the event of budgetary slippage.Footnote ³²⁶

While the framework was being reformed, it was still necessary to intervene in the economic process in order to avoid a break-up of the Eurozone. Financial assistance mechanisms were able to play a role in this respect, but, above all, the ECB’s action proved decisive. The ECB intervened ‘whatever it [took]’ ^{Footnote 327} to save the single currency by reassuring the bond markets, while at the same time working to impose ‘conditionality’ on the countries in financial difficulties, even if the various monetary programmes (SMP, LTRO, OMT, PSPP, etc) were always justified on the grounds of official objectives in line with its mandate.Footnote ³²⁸ This monetary intervention gave rise to a veritable ‘guerre du dernier mot’ between the ECJ and the German Constitutional Court.Footnote ³²⁹ Behind the intense legal debates was the question of whether it was possible for a public institution (the ECB) to distinguish between ‘normal’ and ‘irrational’ market financing conditions and to differentiate interest rates according to whether or not they were ‘justified’ in the light of the ‘real’ and ‘objective’ economic and budgetary situation of the Member States. In other words, the question was whether it was possible for public authorities to judge the ‘(ir)rationality’ of the markets and, if necessary, to intervene to modify the actual functioning of the markets in a way that was deemed to be more in line with market logic.Footnote ³³⁰ In the end, the ECJ ruled in favour of the ECB and confirmed the legality of its monetary programmes.Footnote ³³¹

However, this monetary intervention was intensified during the health crisis to such an extent that the ECB’s ‘unconventional monetary policy’Footnote ³³² resembled a thinly veiled monetary mutualisation of sovereign risks.Footnote ³³³ Moreover, the triggering of the misnamed ‘general escape clause’ of the Stability and Growth Pact (SGP)Footnote ³³⁴ may have given the impression of a major change of direction. In reality, however, the consequences of ‘suspending’ the SGP are symbolic: Member States are only allowed to deviate temporarily from the adjustment path ‘provided that this does not endanger fiscal sustainability in the medium term’, while the European Semester continues to impose control by the European institutions.Footnote ³³⁵ A reform of the SGP, initiated by the Commission in 2022Footnote ³³⁶ , was adopted by the Parliament in April 2024 and will be implemented from 2025 onwards.Footnote ³³⁷ However, it does not in any way alter the intrinsic logic of the budgetary framework: it leads to greater individualisation (and slight flexibility) of fiscal adjustment paths, but at the cost of establishing a mandatory minimum effort, which is now quantified.Footnote ³³⁸ Conversely, the adoption of the €750 billion ‘recovery plan’ (Next Generation EU or NGEU), financed by a common loan, is often presented as a ‘Hamiltonian moment’,Footnote ³³⁹ a ‘critical juncture’ towards fiscal unionFootnote ³⁴⁰ or a real ‘game changer’ for EMU.Footnote ³⁴¹ Here too, enthusiasm should be tempered. On one hand, the German Federal Constitutional Court has laid down very strict guidelines to prevent the establishment of a ‘transfer union’.Footnote ³⁴² On the other hand, the Regulation establishing the Recovery and Resilience Facility stipulates that the national recovery plans adopted under the NGEU must take into account the country-specific recommendations issued in the context of the European Semester,Footnote ³⁴³ giving thus the Commission an additional (and particularly effective) lever to impose the expected structural reforms on the Member States.Footnote ³⁴⁴

Intermediary lessons: the ordoliberal ‘ressourcement’ of neoliberal economic constitutionalism

EMU has undergone profound changes in barely thirty years, but its fundamentals have not changed: monetary policy is still, at least officially, entrusted to a central bank with a primary mandate of price stability; economic and budgetary policies remain decentralised, but are subject (and increasingly so) to controls designed to limit public spending and ensure compliance with balanced budget rules. Such a structure is directly in line with the programme of neoliberal economic constitutionalism, from Hayek to Vanberg, via the new classical macroeconomics and the Public Choice School. The intrinsic legal logic of EMU was, is and remains to neutralise expansionary monetary policies, which were disqualified on the basis of rational expectations, and to impose ‘market discipline’ on Member States,Footnote ³⁴⁵ as some ECB researchers have explicitly pointed out.Footnote ³⁴⁶

It is true, however, that the ‘market discipline’ of EMU has been reshaped, as the new economic governance can be understood, in a sense, as an ordoliberal ressourcement of economic constitutionalism. Whereas most neoliberal currents, under the influence of Hayek, placed their trust in markets as they actually functioned, the early ordoliberals maintained a more assertive tutelary role for public institutions in relation to the competitive functioning of the market – seen as susceptible to occasional failures that must be prevented or, failing that, compensated for. However, the post Eurozone crisis reforms give us a glimpse of something of the order of the ordoliberal Als-Ob Politik. The financial and monetary interventions certainly depart from the orthodox principles of EMU,Footnote ³⁴⁷ but on the other hand, they are ‘dynamically consistent’ with the conditions for regulatory intervention as conceptualised by the early ordoliberals.Footnote ³⁴⁸ The ECB’s fundamental role in designing, negotiating and monitoring the structural adjustment programmes helped to impose conditionality on ‘defaulting’ countries, in order to ensure that they (finally) undertook the reforms they should have undertaken under market pressure. Monetary intervention was also ‘subsidiary’, in that it was used only to ensure the Eurozone’s survival in the short term, until the reform of the legal framework could produce its stabilising effects. Finally, it was ‘limited’ in the twofold sense that it was ‘targeted’ mainly at the countries in most difficultyFootnote ³⁴⁹ and declared to be temporary, although the exact duration was not announced in advance. At the same time, the structural reforms of the (new) economic governance are tantamount to a real Ordnungspolitik on the legal framework of the economy, through which the European institutions are now established as guardians of the logic of the market, if necessary even against the actual functioning of the market. Although the market logic is maintained as an abstract hypothetical standard with which the Member States are supposed to comply in implementing their public economic policies, it is thus gradually guaranteed as a last resort by the (European) public institutions, through the new integrated coordination framework, which includes ‘enhanced surveillance’ imposing a veritable EU tutelage over the states deemed to be failing.

The responses to the COVID-19 crisis are even more heterodox, but they do not change the outcome. The European Semester continues to impose its logic of structural reforms. The modification of the SGP will further strengthen the control of economic and budgetary policies in order to ensure a ‘Stability Union’. Finally, the Next Generation EU, which according to some scholars contains the seeds of a transfer union, is also being used by the Commission to impose the desired structural reforms – and might not be sustainable in the long run, given the strict position of the Karlsruhe Court.