European integration, social democratic Europeanism and the competitiveness discourse: a neo-Poulantzian approach to discursive policy analysis

This article addresses the so-called institutional deficit in poststructuralist discourse theory by articulating a recent development of such a theory, the so-called discursive policy analysis as mainly elaborated by David Howarth and Jason Glynos, with neo-Poulantzian state theory, especially as re-worked by Bob Jessop. It then employs the resulting framework for analysing the internal socio-political logic of the European integration process, and asks why this process has not been challenged by any robust counter-hegemonic alternative despite the regressive consequences that its neo-liberal design has triggered. The argument developed in the article is that the increasing support of European social democracy, once Euro-critical, to the European project played a crucial role in this development. Nevertheless, such a thesis raises the following research question: how the forces leading the European integration process were able to gain and maintain the active support of European social democracy to that process despite the lack of opportunities for social democratic policies at the EU level? To address this question, the article engages with the main policy proposals developed by the European social democratic parties between the mid-1980s and early 2000s, paying special attention to the impacts of these proposals on EU’s institutional architecture. The textual analysis of both the social democratic and EU policy discourses reveals that the widespread appeal of a narrative structured around the “empty signifier” of “competitiveness” served to integrate the social democratic identity within the neo-liberal integration project, organizing a range of discursive logics that successfully marginalized critical worries about European integration. Nevertheless, the article ends by hypothesizing that the technocratic mode of European integration that this historical development has instituted, constitutes the main condition of possibility for the current emergence of Eurosceptic populist forces throughout the EU, so that in the medium term the neo-liberal-oriented European construction might lead to the destablization of the very foundations of the European order.


Introduction
M ost scholarly work on European studies treats the European integration process as a directionless process, which seeks EU expansion as an end in itself. Nevertheless, since the late 1980s, the European integration process has constituted a motor of neo-liberalization of European society (Fitoussi and Saraceno, 2013). Despite the regressive consequences of this process in terms of increasing economic inequality, loss of social rights and decreasing democratic legitimacy, the European project had not been challenged by any robust counter-hegemonic alternative until the recent outbreak of right-wing populism throughout the EU. The argument presented in this article is that the increasing support of European social democracy, once Euro-critical, to the integration project has played a crucial role in this historical development. Social democratic Europeanism has been an essential link in the EU hegemonic chain, since it has enabled the integration of large segments of the popular classes into its neo-liberalizing discourse. In the same manner, that the "Keynesian compromise" (Duménil and Lévy, 2011) reached between labour and capital was essential for the stability of post-war socioeconomic order, the neutralisation of the effects of the neo-liberal breaking up of that settlement was an even more important task for ensuring EU neo-liberal-oriented construction.
In fact, during the post-war decades the social content of the European integration process was not neo-liberal at all. The "Monnet mode of integration" sought to foster a form of Europolity functional to Western European Keynesian models of development (Jessop, 2007: 211). The notion of "embedded liberalism" (Ruggie, 1982) captured well the kind of liberalization promoted during that period, in which market logics were politically controlled by state regulations. Accordingly, at that stage, the integration process sought the establishment of a supranational mode of economic cooperation, limiting the liberalizing tendencies to trade and leaving "enough space for governments to pursue social democratic policies" (McGowan, 2001: 75).
Nevertheless, during the seventies, this "first project of European integration" (Ryner, 2007) entered into crisis, in conjunction with the crisis of the Fordist mode of accumulation at the national level. The crisis of the 1970s provoked divergent responses in the different member states, and consequently, serious intergovernmental frictions at the EU level. But it also generated key stimuli for relaunching the integration process (Moravcsik, 1991). It was thus in the aftermath of this crisis that the "second project of European integration" (Ryner, 2007) was born. With it, during the eighties, the socio-economic content of the integration process progressively took a different direction and eventually neo-liberalism came to determine its subsequent steps.
However, it was precisely during this second period that previous Euro-critical social democrat parties became enthusiastic supporters of the integration process. Before the eighties, most social democratic parties outside the founding member states were Eurosceptic (Dimitrakopoulos, 2011: 2). For instance, in the early seventies, the social democratic Prime Minister of Sweden Olof Palme criticized the European Community (EC) for its "four Cs": capitalism, clericalism, conservatism and colonialism (Moss, 2005: 6). In the six founding states, social democratic parties were more pro-European integration, although usually maintaining a faltering Europeanism, lacking any type of consistent project for Europe and always leaving the leadership in this issue to their main rivals, the Christian Democrats (Camia, 2012). However, this situation changed in the course of the eighties, when all previously Eurosceptic social democratic parties (the parties from Sweden, the United Kingdom, Greece, Denmark, Finland, Norway and Ireland) carried out an Europeanist turn (Featherstone, 1988: 3;Holmes and Lightfoot, 2007), while the social democrats of the Six ostensibly increased their Europeanist profiles (Sassoon, 1996;Pierson, 2001). As Bailey (2009: 3) points out, given that the neo-liberalising effects of the European project made social democratic policies less likely to be implemented across the member states, this "social democratic turn to Europe" 1 seems to constitute a puzzle: why and how were the actors leading the European integration process able to gain and maintain the active support of European social democracy to that process despite the lack of opportunities for social democratic policies at the EU level?
Existing explanations of this puzzle can be broadly divided into two groups. A first group stresses the importance of institutional factors, emphasizing how the increasing relevance of European institutions stimulated social democrats to shift their attitude towards the EU (see, for example, Ladrech, 1997Ladrech, , 2000Douglas, 2002;Diamond, 2012;Escalona et al., 2013). Alternatively, a second field of literature remarks the importance of the role of ideas in the shaping of social democrat rationalities, situating the "turn to Europe" in the context of the deep erosion of core traditional social democratic principles during the late twentieth century (see, for example, McNamara, 1998;Worth, 2007). For this second set of accounts, socialist parties dropped their opposition to the EU because they sidelined any conception of wealth redistribution and state interventionism, thus engaging in instead of challenging neo-liberalism.
Although these explanations are important in describing the pro-EU social democratic move, both types of interpretations leave important issues unresolved. Firstly, both obliterate the agential responses to the contextual changes they describe, naturalizing in a teleological way the Europeanist move. In this sense, it remains unclear why socialist forces opted to answer these contextual changes in the manner they did, rather than, for instance, attempting to reverse them (Bailey, 2009: 14). Secondly, these explanations take the institutional and/or ideational environments as being formed outside the realm of politics, thoroughly constraining social democratic forces, but never being constrained by them. Nevertheless, nor the redefinition of the boundaries of political economy away from the Keynesian emphasis on redistribution, nor the emergence of Europe as a political space of reference were apolitical phenomena. And finally, these accounts are unable to explain why social democratic forces remained loyal to the EU once it appeared to systemically curtail progressive policies.
To overcome these deficits, this article deploys a theoretical framework based on the articulation of post-structuralist or discursive policy analysis (Howarth and Griggs, 2012;Glynos et al., 2015) with neo-Poulantzian state theory (Poulantzas, 2000(Poulantzas, [1978). According to the former, public policies constitute radically contingent discursive articulations whose signification is the outcome of multiple struggles carried out by different actors tussling to overdetermine them (Foa Torres, 2015: 376). This Laclaudian perspective is hence very helpful in explaining the processes of constitution and dissolution of political identities that underpin public policies. However, because poststructuralist policy analysis lacks a strong theorization of state power and infrastructural conditions, we turn to Poulantzas' sociology of the modern state, especially as reworked by Jessop (2007), to better grasp the functions of the state in relation to the political economy of capitalism.
This article is structured as follows: the second section outlines the basic arguments of the above-mentioned two strands of the theory and establishes a dialogue between them based on the centrality of the Gramscian concept of hegemony. The following three sections reconstruct the institutional dynamics that underpinned the development of the European project in relation to the different integration strategies followed by the two central discourses competing within the European political arena. Finally, the last section summarizes the analytical productivity of the discursive approach used in the article by raising some conclusions about the state of the European integration project starkly different from those most commonly found in mainstream integration theory. A main assumption developed throughout the article contends that, despite the fact that Europeanization was obviously not the only factor pushing towards the neoliberalisation of social democratic identity, since such a neoliberalisation was a global phenomenon that in several EU countries pre-dated the neo-liberal re-launch of European integration, the EU, nonetheless, "took the lead in enjoining its members to conform to a new world where Keynesianism was anachronistic and welfare states and industrial relations needed serious reforms" (Ross, in Moschonas, 2009b: 11).
Theoretical framework Discursive policy analysis. Firmly rooted in Laclau and Mouffe's (1985) post-structuralist discourse theory, and taking their lead from the so-called "argumentative turn" (Fischer and Forester, 1993) in public policy studies, post-structuralist policy analysis' most fundamental ontological assumptions are condensed in the theoretical games born of the coupling between two provocative Laclauian aphorisms: the one stating "the rhetorical foundations of society" (Laclau, 2014) and the other affirming "the impossibility of society" (Laclau, 1990). To begin with the first, poststructuralists contend that the foundations of social reality are in a crucial sense rhetorical: all objects and subjects are discursively constructed by signifying practices to the extent they only acquire meaning when they are articulated within relational systems of signifying rules. Discursive articulation, understood "as any practice establishing a relation among elements such that their identity is modified as a result of the articulatory practice" (Laclau and Mouffe 185: 105), therefore acquires ontological primacy: an independent material world outside the discursive does exist, but the elements of this world do not have meaningful identity automatically or by themselves, but only when they are contingently placed in concrete semantic systems.
In this sense, in contrast to interest-based approaches to policy analysis, a discursive perspective understands that not only agents' interests are discursively constructed via articulatory practices, but also that the agents themselves are political products whose identities are dependent upon their relation to other identities (Griggs and Howarth, 2000: 55). Identities appear therefore as "points of temporary attachment to the subject positions which discursive practices construct for us […] they are the result of a successful articulation or 'chaining' of the subject into the flow of the discourse" (Hall, 1996: 6).
In fact, a concrete discourse is defined as hegemonic when it effects the joining of the subjects in its structure of meaning, showing in consequence its schema of interpretation of the world as a commonsensical universal truth. In this manner, the practices of institution of hegemonic discourses, to the extent that delimit which political identities are possible in each social order, also crystallize power relations by drawing lines of inclusion and exclusion. In this sense, the definition of what is included in the hegemonic formation and thus what comes to shape the "common worldview" constitutes not only a purely political question but also a fiction: the common worldviewwhich prefigures the "general interest"-is always the expression of a specific identity that gets to present its particular vision as the most logical for most of the social actors. For this reason, the weaving of a hegemonic discourse always consists in "practices of coalition building" (Howarth, 2013: 199), that is, practices whereby a particular actor links different demands and identities into a single political project capable of presenting its discourse as representing the interests of the whole society.
Nevertheless, given that every hegemonic articulation is only a partial fixation of the social world, there is always some meaning that escapes the complete structuring intended by the hegemonic discourse: in fact, "it is a logical impossibility that a particular identity, constituted through difference, can fully represent all identities" (Koechlin, 2013: 96). This constitutes hence the second basic axiomatic premise of the post-structuralist discourse theory, that is, the idea that any system of rules is marked by a radical impossibility, where impossibility refers to the inherent incompleteness and precariousness that uncovers its unnecessary character. As Derrida (1967) pointed out, any structure of meaning fails to institute a complete fixation of the field of discursivity since there is always an irreducible possibility of the drifting of meaning. In less ontological terms, once instituted, a hegemonic discourse has no guarantee of its maintenance over time: in fact, its reproduction depends on its capacity to prevent subjects from translating dislocatory experiences into challenges to its predominance. Therefore, in second place, hegemony consists also in practices of governance (Howarth, 2013: 202), that is, ruling practices that incorporate and accommodate, in an institutional way, some of the demands of the ruled sectors, while dispersing and marginalizing others.
Nevertheless, although any hegemonic discourse always has to face the impossibility to construct an ultimate closure that fixes ad eternum the meaning of all identities, a minimal degree of fixation is necessary within every system of meaning, since otherwise we would have something which amounts to a state of "social psychosis" (Marchart, 2014: 277). Two conceptual elements carry out this function according to poststructuralist discourse theory. Firstly, what Laclau and Mouffe (1985) called "nodal points", that is, those privileged signifiers that stabilize the meanings of a discursive system by partially unifying such a system: these are those words "to which 'things' themselves refer to recognize themselves in their unity" (Žižek, 1989: 96). Secondly, despite the institution of nodal points, the impossibility of universal ground persists. But at the same time, its need does not disappear. "Empty signifiers come to conceptualize such an impossible but necessary task. They are the symbolic means of representing the absent fullness "through the presence of its absence" (Laclau, 1996: 53). For example, "In a situation of radical disorder 'order' is present as that which is absent; it becomes an empty signifier, as the signifier of this absence" (ibid: 44). This is the essence of an empty signifier: a signifier that, either because an underdetermination or an overdetermination of signifieds that prevents it from having a fixed signified, becomes detached from any concrete meaning to embody the very incompleteness of society (Wullweber, 2015: 81). These symbols, however, do not emerge arbitrarily, but come into existence because actors succeed in emptying certain signifiers of any reminder of particularity, so that such signifiers can become points of identification for multiple different identities. At the same time, however, this implies that empty signifiers constitute empty places where different actors compete in their efforts to represent a specific sense of universality, that is, to fill them by promising the achievement of the absent fullness.
Accordingly, discursive policy analysis develops a kind of policy analysis that focuses on the discursive practices that seek to mobilize and articulate shared meanings around certain issuesinstead of others-to fix the discursive fields that institute and sustain institutional orders. It is an approach therefore that emphasizes the centrality of certain terms in giving identity to a policy field and develops the possibility of studying the political strategies through which these terms are developed in policy paradigms.
Interlude: The institutional deficit in post-structuralist discourse theory. Because of its stress on the contingency of any social reality because of the ontological primacy of the discursive, critics of poststructuralist discourse theory have frequently argued that this perspective suffers from a kind of institutional or sociological deficit, in the sense that it is unable to account for the complex relationships between structural (socio-economic, institutional) arrangements and concrete practices (Geras, 1990;Lewis, 2005). In Mouzelis' (1988: 11) words, "because their excessive fear of reifying institutional structures, [Laclau and Mouffe] go to other extremes and analyse practice in an institutional vacuum". Consequently, poststructuralist discourse theory is accused of lacking the tools for apprehending "the materiality and structurality of the social" (Chouliaraki, 2002: 94), ending in an underestimation of the "conditions of reception of discourse in favour of the discursive mechanisms of its production" (Jessop, 1990: 298). For instance, many Marxists have criticized post-structuralist discourse theory's incapacity to conceptualize how the contradictions arising from the capitalist accumulation process dislocate existing hegemonic regimes (Rey Araújo, 2015), while others have signalled how in Laclau's work the "institutional system" is deployed as an anonymous black box that allows eschewing the necessary fine-grained analysis of the array of forces operating within institutions (Boucher, 2008;Anderson, 2016: 82). Nevertheless, at the ontological level, post-structuralist discourse theory has never defended the idea that social relations are in a state of continual flux, since it has precisely coined concepts like "nodal points" or "sedimentation" for referring to the temporary structuring of social relations. Likewise, this theory has never conceptualized discursive practises as occurring in contexts freed from structural constraints. In fact, to the extent that poststructuralist discourse theory, as its name indicates, draws its basic ontology from post-structuralism, it remains attached to the de-centring of the subject characteristic of it, so that it avoids any kind of agency-centred model of social explanation. For poststructuralist discourse theory, in fact, the only space for the autonomy of the subject is the space left free by the dislocation of the structure (Laclau, 1990: 60).
In this sense, contrary to what is sustained by some of its voracious critics, the so-called institutional deficit of poststructuralist discourse theory does not arise from inherent deficiencies in its ontology, like for instance a putative semiotic monism (Sum and Jessop, 2013: 133) or subjective ahistoricism, but from areas of analytical neglect, a fact which calls for further developments of its theoretical potential. In fact, Glynos, Griggs and Howarth's efforts during the last decade for articulating an approach to policy analysis on the basis of poststructuralist discourse theory's ontological commitments can be read as an attempt to work on this deficit, and thus to overcome the traditional self-limitation of discourse theorists, long ago noted by Torfing (2005), of focusing on "soft" topics of social science research, such as ethnic identities or political ideologies, leaving "hard" topics like institutional analysis, political economy or public administration to mainstream neo-positivist approaches. Nevertheless, as Howarth himself admits, in post-structuralist styles of theorising "structures are still deemed to be too abstract or general, whilst their radical contingency and fluidity renders them unhelpfully indeterminate in critically explaining particular regimes and practices in specific historical conjunctures" (2013: 208; Hansen, 2008). It is for this reason that this paper articulates poststructuralist policy analysis with the conceptual apparatus provided by neo-Poulantzian state theory.
Neo-Poulantzian approach to state power. The starting point of Poulantzas' conceptualization of the state neatly connects to the last insight we raised from the poststructuralist understanding of public policies, namely that policies constitute products "shaped by the overall balance of political forces in society" (Griggs and Howarth, 2013: 22). Thus, in a very similar manner that also echoes the Gramscian notion of integral state (Buci-Glucksmann, 1980), Poulantzas (2000Poulantzas ( [1978) understood the state as a social relation that condenses in a concrete institutional ensemble the balance of forces existing in a given society as well as the various contradictions underlying it. Hence, in a neo-Poulantzian perspective, public policies and institutions constitute crystallizations of historical processes, for which their functioning reflects the interests of victorious political actors in past social conflicts, or alternatively, the institutionalised compromises reached between the different contending agents. In this sense, according to this perspective, the state does not constitute an autonomous neutral window as in some forms of pluralism, nor a mere tool in the hands of some other agent as in some versions of Marxism, but an asymmetrical "terrain of struggle" (García Linera, 2015) embodied with certain "strategic selectivities" (Jessop, 1990) that predispose it to reproduce certain logics over others or to provide advantages to some groups in pursuing their objectives while imposing constraints on others. Furthermore, these structural selectivities are understood as having strategic meaning for social struggles, so that they modulate the strategies of the agents that try to influence the state, becoming state institutions constitutive of actor's identities. Given the discursive ontology exposed above, here we can understand these strategic selectivities as the outcome of the "sedimentation" (Torfing, 1999: 69) of concrete discourses that institutionalize specific policy narratives.
Neo-Poulantzian state theory understands the state as a "distinct ensemble of institutions whose socially accepted function is to define and enforce collectively binding decisions on a given population in the name of their 'general will' " (Jessop, 2007: 9), following the metonymical logic of hegemony explained above according to which any universality is always a particular political orientation capable of gaining a general consensus. At this point, the distinctive contribution of post-structuralist discourse theory to state theory probably resides in thinking of institutions as not only hegemonic sedimentations, and so as radically political, but also as constitutively barred or contingent (Foa Torres, 2015: 362). Policies appear therefore as signifying systems aimed towards the temporary suspension of the radical contingency which haunts any social formation: the concern is not only on the instances of discursive institution of any policy regime, but also on the ways through which such regime excludes alternative possibilities.
Understanding the state as a social relation enables a vision of how its power is dependent on social relations outside itself. In this sense, for any state project to become hegemonic it must be linked to a feasible accumulation strategy capable of securing the social reproduction of the economic order. The concept of "accumulation strategy" hence refers to those economic aspects of any hegemonic discourse, that is, those state strategies intended to solve the problems of capital reproduction (Jessop, 1990: 159). An accumulation strategy is hegemonic when it can sustain the institutionalised correlation of forces represented in the correlated state project, producing in this way a "general economic interest" able to delimit within certain parameters the interests, strategies and spatial-temporal horizons of the different social ARTICLE PALGRAVE COMMUNICATIONS | DOI: 10.1057/palcomms.2017.60 forces as well as to ensure the social cohesion of the whole (Jessop, 2002). In this sense, changes in the correlation of forces and/or accumulation strategies might lead to spatial reorganizations of state's strategic selectivities (Poulantzas, 1974). When analysing the dynamic of EU integration, this implies understanding this process as a process that constructs a new institutional form of European governance coupled with a specific socio-economic content (Van Apeldoorn, 2002).
In sum, the theoretical framework articulated in this article directs its analytical focus on the production of policy discourses as a strategic-discursive moment of the production and reproduction of hegemonic regimes. Hence, it is a kind of policy analysis that always implies political analysis, that is, analysis of the correlation of forces within a society as well as of the dynamic contradictions underpinning such a correlation.
European social democracy's role in the establishment of the Single European Market As exposed in the introduction, European social democrats started to identify in an explicit way with the Europeanist discourse during the 1980s, when the integration process was being relaunched through the single market initiative. To a large extent, such a revitalization of the integration project was made possible by the deep socio-economic dislocations that European countries were facing during the 1970s, which involved, among other elements, stagflation, chronic declining growth levels, fiscal crises, decreasing rates of productivity or intensified international competition. Nevertheless, as it arises from the theoretical framework discussed above, the nature of a crisis is not simply given by its dislocations or effects. Instead, some actors impose on others their diagnosis of "what has gone wrong" and thus "what is to be done" (Blyth, 2002: 10). Therefore, in periods of crisis, it is imperative to attend to the policy discourses that key actors develop.
Among the multiple interpretations struggling to hegemonise the reading of the thorny conjecture of the 1970s, a shared consensus progressively emerged that the crisis responses would have to imply, in one way or another, a deepening of the European integration project. For the social democratic forces, the narrative that came to formalise such an intuition was the Delorist project, opposed, as we will see in the following subsection, by a neo-liberal discourse. Nevertheless, the Delorist discourse, understood as the first comprehensive approach to the integration project articulated by European social democracy, was in turn the product of the failure of a previous and alternative social democratic exit to the crisis of the 1970s, markedly more national in character: the Mitterrand project for a democratic socialism. Therefore, to understand the former, we need to grasp what the latter-and especially its defeat-implied for the European left.
The Mitterrand experience. The socialist François Mitterrand was elected president of France in June 1981 relying on the "Common Programme of the Left" (supported by the French communists) that called for a "rupture with capitalism" and a reformist transition towards a democratic socialism. Nevertheless, his left-wing government was faced from the beginning with the contradiction of maintaining the French commitments with the EC, a clear capitalist project, or pursuing a program of socialist transformation in France. The Partie Socialiste (PS) exposed in its ideological manifesto of 1980 its confidence in this double aim: The realization in France of a socialist project will be a shock within our European environment, which will cause it to be less marked by liberalism and Atlanticism. Engaged in the construction of a socialist society, France will contribute to the democratization of the Community, she will utilize the institutions to favour the convergence of the social struggles against unemployment, for the reduction of the working time, for the control of multinational companies, for the defence of liberties and for the extension of democracy (PS, 1980). After Mitterrand's victory, the French government started a vigorous Keynesian strategy of reflation of the economy coupled with a huge reformist project involving, among other elements, nationalisations, industrial relations legislation that implied worker control, welfare expansion, diminishing of weekly working hours, industrial planning and a large increase in subsidies granted to public enterprises (Kesselman, 1984). In Moss's (2005: 131) words, "it was once the most radical and credible alternative to neo-liberalism in Europe", and because of that, threatened the future of the EC.
Nevertheless, Mitterrand's intentions to pursue a reflationary strategy within the EC started to run into serious difficulties by 1982: the domestic economy failed to respond in the desired manner, high US interest rates squeezed financial markets and other EC countries failed to take complementary action 2 . In relation to this latter point, the pressures arising from the commitments involved in staying in the European Exchange Rate Mechanism (ERM)-the fixed exchange rate monetary regime introduced in 1979 by the European Economic Community that, in the image of the Deutschmark block, was aimed at the maintenance of low levels of inflation (Hancké, 2013)-played a crucial role in halting the expansionary programme of Mitterrand. In the ERM's fixed exchange rate system, unlike what happened in the previous Bretton Woods' fixed but adjustable exchange rate system, the costs of price inflation could not be adjusted through devaluation, but impacted directly on the balance of payments. This led the French government's counter-cyclical economic policy to cause a tremendous trade imbalance that ruined the franc and triggered a harsh budgetary crisis (Clift, 2005). By March 1983, the only alternative available for the Socialist-Communist government to follow the programme of socialist transformation was a deepening in the reflationary strategy through a substantial devaluation of the franc (coupled by an important revaluation of the mark) accompanied by the implementation of protectionist trading measures, the lowering of interest rates and a purposeful nationalization of industry 3 . However, this strategy implied their exit from the ERM, since both the devaluation of the franc permitted within the ERM as well as the revaluation of the mark accepted by the Germans were a long distance from what France needed to neutralize the accumulated inflation differentials. The other option was remaining within the ERM, turning towards contractionary fiscal and monetary policy measures that would reduce the rate of inflation and implementing only a small devaluation.
The U-turn of 1983 marked the shift towards the second option, implementing a set of austerity measures coupled with a new hyper positive stance vis-à-vis European integration. In 1984, Mitterrand declared that France's new "national goals" were "modernization" and "global competitiveness" (quoted in Kuisel, 1987: 30). The new "economic rigueur" framework implied not only a shift of priorities in macro-economic policy, but it also constituted the springboard for an assault on the French dirigiste model of capitalism, with reforms de-indexing wages from prices, "restructuring" the steel, automobile and shipbuilding sectors, deregulating the financial market or implementing an initial change in the Welfare State away from universalistic solidarity and towards the Anglo-Saxon scheme (Ross, 1987: 211-213).
Mitterrand's experiment had collided with the limits of the strategic selectivities inscribed in the political economy of the EC, in which the cycle "strong currency-increasing unemploymentlow interest rates-low costs" was favoured by the institutional framework. This macroeconomic framework favoured the interests of the supply-based Northern economies, led by Germany, which were (and still are) inflation-averse to the extent that their growth came from exports. Thus, in 1983, under the danger of bankruptcy, Mitterrand, although obviously aware that an important cause of the French inflation was the mark's undervaluation, came to accept that "leaving the monetary system was impossible" (Mitterrand, in Cameron, 1996: 70). The implicit assumptions behind this decision were that a long-run austerity approach to fiscal and monetary policy was required to bring about a strong price and wage deflation as the means to increase national competitiveness and stimulate investment, and especially, that this was the only feasible strategy towards boosting growth. Within the ERM, the traditional enlarged popular coalition around the working class constructed by social democratic parties (Moschonas, 2002), because of their constitutive reliance on strategies pursuing full employment and protection against market commodification tendencies, was no longer a viable hegemonic articulation. Indeed, we can read this process from the second face of the concept of hegemony theorised above, according to which politics consist of weaving a terrain of struggle biased towards the constitution of discursive coalitions favourable to one's own interests. In this sense, replacing the 1981 program of "social democracy in one country" by another ordoliberal in character perhaps was not the only available option for Mitterrand 4 , but it did constitute the most "appropriate" decision (Olsen, 2008) in the web of intersubjective practices in which European actors were embodied.
As many observers have noted (Featherstone, 1988: 129;Haywood, 1993: 274), in the 1983 turnabout Mitterrand picked up the European project as a substitute collective horizon of future for rationalizing his neo-liberal turn while retaining a degree of socialist credibility. In the aftermath of the dislocation of the statist discourse, in which the identities and interests of the French actors were deeply contested, Mitterrand opted for presenting the European project as the new platform for defending French national interests, Europeanizing consequently the identity of the socialist party (Risse et al., 1999: 170).
Importantly, such a correlation between a transition towards a "new" social democracy and a turn towards a Europeanist stance was a common trend of all European social democratic parties in the 1980s (Bailey, 2009: 107). In this sense, the joint manifesto of social democratic parties for the 1984 European elections mirrored this shift and represented a qualitative step towards a more pro-EC stance with respect to the 1979 minimalist manifesto (Lodge, 1984: 41). Thus the manifesto of the Confederation of Socialist Parties of the European Community (CSPEC) assumed European integration as a central instrument for the realization of the social democratic goals of social justice, job creation, solidarity and freedom (CSPEC, 1984). Nevertheless, the manifesto emphasized the need for coordinated efforts rather than integrated common policies, an approach that reflected the British and Danish reluctances to further supranationalisation (Lodge, 1984: 41). In contrast, in the French case, Mitterrand's Europeanism was deeper and implied the assumption that "the future of France goes via the renewal of Europe" (in Featherstone, 1988: 129). From that moment, the advance of the European project would be presented not only as equivalent to the advance of France but also to the defence of the French collectivist tradition: The future of French exceptionalism, the future of the social achievements which we have attained, passes through Europe […] Understand that this Europe is a necessity today for the forces of progress, for the values which represent the Republic (Mitterrand 1986, in Risse et al., 1999. Two opposite discourses on the Single European Market. It was in this context that European social democratic parties changed their attitude from "ambiguity to promotion" (Holmes and Lightfoot, 2002) with regards European integration. As Escalona and Vieira (2014: 23) explain, European social democracy saw in Mitterrand's U-turn both the defeat of the "last chance" of national Keynesianism as well as the right response to such a defeat that sought to renovate the social democratic model at the European level. Mitterrand, in alliance with Helmut Kohl, became from that moment onwards one of the leading figures in the development of the new stage of the integration process, the socalled Single European Market.
Mitterrand's pro-integration discourse was mainly articulated by his former Minister of Finance and from 1984 to 1995 President of the European Commission Jacques Delors. For Delors, the revitalization of the integration process constituted the only feasible opportunity for European societies for, in the face of the challenges entailed by the "globalization of markets and new technologies", recapturing control of the economy to defend the "European model of society" and thus "to master our destiny" (Delors, 1988). At the same time, however, Delors (1985b) was not reluctant to emphasize the need for substantial changes to adapt to the growing international competition. In this sense, according to Delors the first step towards the revitalization of the integration process was the completion of the internal market, that is, the abolishment of all internal barriers to the free movement of goods, services and investment within the EC. Nevertheless, the Delorist approach to the completion of the single market envisaged the elimination of "internal frontier controls in their entirety" (Commission of the European Communities (CEC), 1985: 9) to be complemented, in the short term, by (1) the institution of a single currency, (2) the development of some kind of social union and (3) the establishment of mechanisms of supranational governance (especially in the field of industrial relations) (Delors, 1989a). It was therefore a "balanced" (Ziltener, 2000) approach to the single market, in line with a conception of the European integration process as being guided by the formula "competition that stimulates; co-operation that reinforces; and solidarity which unites" (Delors, 2013: 174): The European Community will be characterized by cooperation as well as competition. It will encourage individual initiative as well as solidarity.
[…] A large market of 320 million will increase competition. It will benefit the consumer, and allow European industry to compete on a world scale. It will create new job opportunities and contribute to a better standard of living. These benefits will only be fully achieved with increased cooperation (Delors, 1988).
As reflected by the 1985 White Paper on Completing the Internal Market, cornerstone of the Delorist project, Delors' intention was to use cooperation regards market integration to build a momentum toward more European expansion: The free movement of goods is an important, valuable and possibly indispensable step which has to be taken before economic integration can be achieved. But it is not the ultimate goal.
[…] Just as the Customs Union had to ARTICLE PALGRAVE COMMUNICATIONS | DOI: 10.1057DOI: 10. /palcomms.2017 precede Economic Integration, so Economic Integration has to precede European Unity (CEC, 1985: 55) The nodal point that came to provide to the social democratic forces with a point of attachment to the Delorist discourse was that of the European Social Model (ESM), which drew inspiration from the 1973 Social Action Programme's idea of a "Social Europe". The basic idea behind the notion of an ESM was the existence of a European model of capitalism, different from Anglo-Saxon pure-market capitalism, capable of simultaneously promoting sustainable economic growth and social cohesion (Jepsen and Serrano, 2005). European integration had therefore to protect such a social model against the potentially destructive tendencies of globalization. In a detailed economic survey presented in 1992, Delors came to defend this vision and showed how regulated integration could maintain Europe's international competitiveness whilst ensuring proper working conditions and wage levels (Delors, 1992). In the Delorist discourse, the creation of a large market must go hand in hand with the organization of a European social area. The beneficial effects of a large market would be dissipated if some Member States were to seek a competitive advantage by sacrificing social achievements (Delors, 1985a: 30).
But in opposition to the Delorist discourse, which was progressively being endorsed by the different European social democratic parties, a neo-liberal approach to the single market was articulated by the right-wing governments of the EC, the more export-orientated fractions of capital and global financial institutions (Van Apeldoorn, 2002). The master signifier which came to structure this interpretation of the crisis was that of "Eurosclerosis", coined at the beginning of the 1980s by the president of the Hayekian Mont Pèlerin Society Herbert Giersch (1985). According to Giersch, the stagflation of European economies was a consequence of the "institutional rigidities" provoked by government over-regulation, trade union abuses and excessively generous social benefits draining resources from the private sector (ibid.). The result was decreasing levels of competitiveness of the European economies. The institution of a common market was therefore imagined as the removal of trade barriers to intensify competition between member states and market forces and thus accelerate technological innovation, eventually improving the international competitive position of the EC. As expressed by Thatcher (1988) in her speech to the College of Europe: The aim of a Europe open to enterprise is the moving force behind the creation of the Single European Market. By getting rid of barriers, by making it possible for companies to operate on a European scale, we can best compete with the United States, Japan and other new economic powers emerging in Asia and elsewhere.
Therefore, in contrast to the Delorist balanced approach, the ordoliberal "systems competition" (Ziltener, 2000: 180) was based on a narrow interpretation of the single market, purely aimed at the reduction of political intervention in the market sphere by enabling European capitals to exploit differences among national regulations. As expressed again by Thatcher (1988) The institution of competitiveness as master policy paradigm. The strategic goals and compromises established in the Single European Act (SEA) of 1987 were mainly shaped by the neoliberal discourse, so that the national economies were forced into the treadmill of unfettered competition. As Grahl and Teague (1989) explained, the key mechanism that ensured the elimination of all national barriers was the institution, following the proposal of the 1985 White Paper, of the method of mutual recognition-instead of harmonization-as the preferential mechanism for building the single market: according to this principle, in those areas without specific European-wide regulation member states had to accept without restriction products fabricated in the other member states. Such a policy change represented a crucial break in the mode of market governance within the EC, and, coupled with the qualified-majority voting system in the European Council that made the drafting of EUwide harmonization policies incredibly difficult, generated important downward pressures on national regulatory standards, since goods commercialized according to lower standards could be sold just as easily as higher-standard ones. Nevertheless, the SEA allowed member states to maintain stricter legislation without mutual recognition in especially sensitive areas (Schmidt, 2007). In any case, in most policy areas, the combination of free movement, no internal frontiers and mutual recognition led to intense exposure of national product markets to competition. As implied in the Cecchini Report (CEC, 1988: 6), it was expected that the single market would intensify competition not only in the economic field but also in the regulatory field between states.
As we can therefore observe, the empty signifier that came to structure the dominant policy discourse on the single market was that of "competitiveness". In fact, many scholars have argued that it is precisely the placing of the logic of competition as the guiding principle to reform and inform both the state and the society the element that better indexes a shift towards a neo-liberal discourse (Cerny, 1997;Jessop, 2002;Foucault, 2008). Thus, while the embedded liberalism of the post-War era assumed the need to insulate certain areas of social life from market competitive logic while at the same time promoting competition in other areas, neo-liberalism defines competition as the general principal of societal coordination, a move that implies the extension of market logics throughout the social body. Competition and the inequality it brings are valued positively within the neo-liberal discourse because they are seen as boosting an entrepreneurial ethos through which innovative knowledge is best pursued. As expressed by Hayek (2002Hayek ( [1968: 19): the necessary changes in habits will occur only when those who are ready and able to experiment with new procedures can make it necessary for the others to imitate them, with the former thereby showing the way; but if the majority is in a position to prevent the few from conducting experiments, the necessary discovery procedure will be frustrated. The fact that competition not only shows how things can be improved, but also forces all those whose income depends on the market to imitate the improvements, is of course one of the major reasons for the disinclination to compete.
The work of Schumpeter, the first theorist of entrepreneurial competition, was crucial for the development of this specific neo-liberal approach to competition (Davies, 2014). Schumpeter (1934) had theorzsed entrepreneurial competitiveness as the outcome of individual and collective engagement in permanent innovation in the face of competition. Especially important, he understood that the success of such a search for innovation was dependent on a wide range of extra-economic factors rather than solely on economic costs, as traditionally assumed by neomercantilism. Analogous to the two traditional strategies of business competition, costs leadership and product differentiation, countries can compete on costs but also on other elements that are not related to price (Fagerberg, 1988).
In this sense, and this constitutes a capital difference between neo-liberalism and classical laissez-faire liberalism, competitiveness is seen by neo-liberals as a quality that does not emerge in isolation within markets, but that needs to be actively cultivated between the market, society and individuals. This theoretical move thus expanded the arena of competitiveness to include a wide range of cultural, technological, sociological, and especially important here, institutional factors. Regarding the links between the latter and competition, the work of another Schumpeterian economist, Michael Porter, was very influential in European policy-making circles. In The Competitive Advantage of Nations (1990), Porter wrote that while it is firms which compete in the market, it is state policies which cultivate market conditions to enable them to exploit each economy's competitive advantage. Thus, under the new Schumpeterian understanding of competitiveness, it became accepted that institutions set the framework for the competitive advantages of firms and individuals, so that the competitiveness discourse came to identify more and more policy areas as sources of competitive advantages or disadvantages. In the last instance, the point became no longer to defend markets as competitive arenas, but to project societies-as-a-whole as large competitive arenas (Davies, 2014). As phrased by President Clinton, each nation is "like a big corporation competing in the global marketplace" (quoted in Krugman, 1994: 29). The paradox was that this process did not lead to a simple "roll-back" of state interventions, but to an actual "rollout" of them to enforce the institution of competitive logics (Peck, 2010). In Thatcher's (1988) words, what was needed was "action to free markets, action to widen choice". However, to make this possible, international competitiveness had to have been previously problematized as a central policy concern for European political actors. This is precisely what we observe from the early 1980s and especially during the single market debate, when globalization was increasingly constructed as an unavoidable constraint that made inefficient any type of state intervention at the national level (Hay and Rosamond, 2002), and when the decreasing competitiveness of Europe vis-à-vis Japan and the United States was established as a permanent threat to Europe's leading economic role (see, for example, Delors, 1989a: 5;Delors, 1989b: 116). For instance, the manifesto of the Party of European Socialists (PES) for the 1994 European elections affirmed the need to "face the economic challenge from both America and Asia ensuring the European model of social democracy" 5 . In this manner, the imagination of an indomitable process of globalization was intimately linked with the perception of the need to manage competition at the EU level, as seen for example in the joint declaration of the leaders of the CSPEC in 1990: The ever-increasing internationalization of the economy and interdependence of our societies at every level means that it is increasingly difficult to respond on a national level to the new challenges which arise. Democratic control of the future remains possible, provided that those elements of sovereignty which can no longer be exercised in a purely national framework are pooled (quoted in Ladrech, 1993: 207) But once competitiveness is instituted as the key object of governance, it also becomes a "mode of meta-governance" (Jessop, 2015): how to link a wide range of factors to the demands of economic competition becomes the guiding aim of state activities, so that there is a shift in governmental priorities. This is the reason why the competitiveness discourse raises major questions about the sustainability of welfare states. As many have pointed out, important trade-offs exist between the requirements imposed by the aim of achieving higher levels of competitiveness and the deployment of the range of social services associated with the welfare state: for instance, while the development of the welfare state increases the retribution of labour power, it also hampers comparative advantages associated with low labour costs; or while the institution of social rights protect citizens from adverse conditions, it also limits the flexibility with which agents respond to market changes (Alonso, 2007: 559-560). At a more general level, while rising competitiveness implies assuming the need to ensure the full adaptation of a society to market exigencies, defending social protection involves the establishment of mechanisms that tame market competitive logic.
Consequently, to the extent that the competitiveness discourse prevailed in the design of the single market, the social policy content of the SEA was minimal. It represented in fact a "byproduct" rather than a realization of the social dimension that had been imagined by the Delorist discourse (Ziltener, 2000: 159). Thus, although some "flanking social policy measures" (Bieler, 2007: 114) were included in the SEA, particularly regarding the regulation of occupational health and safety policies, the creation of regional policy funds or the institution of social dialogue between management and labour at the European level, such measures did not represent any significant advance towards the construction of the social dimension. Delors did not disagree with this evaluation and defined the social content of the SEA as "purely symbolic" (quoted in Ziltener, 2000: 163).
Nonetheless, as explained in the Venturini Report, the single market was seen as a positive step by the Delorist discourse to the extent that it was thought of as promoting economic and social convergence within the EC and thus enabling a future transition towards a monetary, social and political union (Venturini, 1988: 26). In our view, if the integration-cum-competitiveness discourse incarnated in the SEA succeeded in holding as credible such a future horizon, it was not thanks to a simple elite operation of depiction and manipulation, nor to any kind of social democratic betrayal, but thanks to the capacity of the competitiveness discourse to draw a collective imaginary that successfully interpellated not only neo-liberals but also social democrats, thus coming to represent the general interest. In this sense, the social democratic turn to Europe was not favoured only from the more defensive form of hegemony, the one that consists of reproducing the status quo by disarming the challengers to its maintenance, but also from its more offensive face, the one that implies active persuasion of competing forces and the building of new collective identities.
In this aspect, at least three elements played an important role. In first place, some supranationalising reforms were included in the single market programme, ranging from the legally binding treaty form of the SEA (instead of a mere intergovernmental agreement), to the increased powers given to the European Parliament 6 or the inclusion in the SEA's preamble of the need to develop a monetary union, all of them fiercely opposed by the more Thatcherite actors. Secondly, the wave of electoral defeats that most of social democratic parties suffered during the 1980s provoked an important dislocation of the social democratic identity and a subsequent search for new discourses (Wilks-Heeg, 1996). And last but not least, the Europeanization-cumcompetitiveness discourse was highly attractive for being the ARTICLE PALGRAVE COMMUNICATIONS | DOI: 10.1057DOI: 10. /palcomms.2017 final selection of such a search for at least two reasons. Firstly, because as exposed above, during the 1980s the European social democrats progressively took as self-evident the existence of a relentless globalization process that could not be tamed at the national level. In second place, because the interpretation of what "to be competitive" meant was not a straightforward question. Competitiveness constitutes one of those "floating signifiers" (Laclau, 2005: 95) widely used in the political debate without being defined in a clear way. It is an ambiguous and elusive concept that can be invested with very different meanings, while paradoxically, at the same time, it seems rather obvious and intuitive, with a seemingly universal meaning freed from any necessary qualification. In this sense, thanks to such a productive fuzziness, because of this polyvalent but instinctive meaning, each actor can see in it what he needs to see, so that it becomes a very useful rhetorical resource for articulating different visions of values while retaining some appearance of overall coherence. Hence, what does competitiveness actually mean-who does compete? How and through which mechanisms? How is competitiveness related to other economic and social logics?-is not a self-evident question, but a matter of intensive discursive struggle.
And nevertheless, at the same time, competitiveness contains an implicit or minimal normative content useful for grounding political practices from a neo-liberal point of view. When we think of competitions, we think of situations where participants are free to compete and have some vaguely equal opportunity to win. Furthermore, we assume unequal benefits for the different participants at the end of the competition. If we speak about countries, and given that one can only compete with competitors, it is in the external sector where a country finds its competitors, so that a country's competitiveness entails necessarily its external competitiveness. In this manner, to demand international competitiveness is to demand that nations prove themselves relative to one other.
In this sense, the competitiveness discourse, despite being constitutively amorphous, entailed a terrain of struggle biased towards economism and neo-liberal policies. When trade unions and labour parties accepted that the welfare of a country is largely determined by its success in world markets (Krugman, 1994: 30), they also came to internalize the view that institutions constitute mechanisms for seeking competitive efficiencies instead of places of political struggle. Nevertheless, as the following section explains, the new routes opened by such a hegemonisation of the EU policy-making field by the competitiveness discourse were not, admittedly, entirely neo-liberal: they contained both new possibilities as well as important threats to the development of a social democratic approach to Europeanization. How social democrats addressed these potentialities is another issue.
The development of the EMU Two different models of monetary integration. In 1990, 3 years before the year of completion of the Single European Market, the first stage of the European and Monetary Union (EMU), which consisted in the abolition of exchange controls, was launched by the European Council. Although the basic reform envisaged by the EMU, namely the institution of a common European currency by 1999, was defended by both the Delorist and the neoliberal discourses on the base that it would reduce uncertainty and encourage stability, these discourses nevertheless raised different visions on how this reform should be institutionally developed.
For the Delorist discourse, monetary integration constituted "the trigger for political union" (Degryse et al., 2013: 9): the economic convergence that the common currency would bring about would help EU member states to grow into a more politically integrated region, thereby relegating potential political frictions between them and producing common interests (Delors, 1992). The institution of the monetary union was going to be implemented on the basis of formal selection criteria, so that the pursuing of these criteria would foster convergence of economic structures (Delors, 1989b: 13). In turn, European macroeconomic, tax and social policies were expected to develop gradually, since they were understood as necessary supports for economic integration (PES, 1996). For Delors (1989a: 82), "a substantial development of the budgetary function of the EC in the case of monetary integration" was necessary to correct regional imbalances.
In contrast, the neo-liberal discourse assumed that it was not feasible to build a social and fiscal union out of the heterogeneity of national interests. Thus, by drawing from the German ordoliberal tradition, which understands that market equilibrium depends upon the existence of a legal framework that strips out political discretion from the policy-making process to constrain both economic policies and market functioning under concrete pre-fixed rules, it turned need into a virtue and defended a model of "monetary union without political union" (Degryse et al., 2013: 10). According to this model, monetary policy should be managed by an independent central bank at the European level, while social and economic policies should remain within the competence of national governments (James, 2016). Nevertheless, some kind of fiscal policy coordination at the European level-in the form of binding rules on budgetary policy-was seen as indispensable to ensure price stability, the chief aim of the EMU for this paradigm. The reason was, in the words of Alexander Lamfalussy, president of the European Monetary Institute, the need "to avoid tensions arising from excessive differences between public sector borrowing requirements of individual member countries" (Lamfalussy, 1989: 93). Or in other words, ordoliberals demanded strict supranational fiscal rules-consisting of limits to public deficits-because they feared that in a monetary union perverse incentives for individual states towards excessive debt financing would emerge, since interest rate convergence provoked by monetary integration would enable cheap credit in the poorer states. The ordoliberal discourse therefore entailed a form of procedural governance at the European level, that is, governance by formal rules and indicators devised by experts, or in other words, "with only extremely limited margin for political interpretation" (Degryse et al., 2013: 10).
In contrast, European social democrats did not expect economic convergence, and hence successful monetary integration, to be developed spontaneously through market stimuli, but understood both processes as dependent on political debate and discretionary decisions. For European socialist leaders, the EMU was a "strategic objective", but was "inconceivable without strong political institutions to guide it" (PES, 1995). Thus, the leaders of the CSPEC declared in 1990 that monetary integration needed to be complemented by democratic mechanisms of control, instruments allowing for anticyclical fiscal policies and provisions against tax dumping (CSPEC, 1990). Furthermore, they rejected fixed limits for national debts although accepting the necessity of some guidelines for guaranteeing monetary stability (ibid.).
A regime of competitive austerity. EMU's institutional set-up as implemented by the Maastricht Treaty (1992) constituted a compromise between the neo-liberal and social democratic discourses, even though it clearly contained an important ordoliberal bias that took tangible form on the basis of three pillars. Firstly, in the field of macro-economic policy-making, the EU role was limited to monitoring national budgetary policies and loosely "coordinating" national macro-economic policies. Secondly, the Europeanised monetary regime was left in the hands of the European Central Bank (ECB), a central bank with a particularly strong degree of independence to the extent that, unlike the Fed or the Bank of England, it enjoys full independence with respect to instruments and targets within a strict mandate for price stability (at the expense of other goals such as growth or full employment). And thirdly, articles 121 and 126 of the treaty and the subsequent 1997 Stability and Growth Pact (SGP) established a fiscal surveillance regime that placed stringent constraints on budget deficits and public debt and hence left little room for discretionary demand management at the national level. Furthermore, to verify compliance with the rules, quasi-automatic mechanisms of European intervention were instituted, consisting of a monitoring procedure for preventing excessive deficits (known as the "preventive arm" of the SGP) and a corrective procedure for redressing them (the "corrective arm" of the SGP).
Nevertheless, such a rule-based framework was attenuated by the fact that the application of the corrective arm of the SGP required a discretionary decision in the ECOFIN (the Council of Ministers in its composition of economic and finance ministers) with a qualified majority vote in favour, a political obstacle therefore to the implementation of automatic sanctions 7 . In fact, before the 2008 financial crisis, even 44 times (out of 177 possible occasions) the deficit or debt ceilings were breached with relative impunity (Calmfors Wren-Lewis, 2011), leaving in consequence SGP's bounding capacities with little credibility.
In any case, the separation of fiscal policy from monetary policy implied that the monetary and budgetary tools available before to manage economic problems at the national level were gone without previously development at the supranational level. Essentially, less competitive countries lost the single mechanism that allowed them to periodically eliminate trade deficits, restore competitiveness and ease budgetary pressures, namely currency devaluation. Thus, without the capacity to adjust the exchange rates, without any kind of European coordinating mechanism (such as an EU treasury) between state fiscal policies and in a context of free capital mobility and no customs barriers, the remaining macro-economic tools used by modern governments, tax and spending policies, also became seriously constrained by, in the first case, the tax competition that emerged between states in the struggle for attracting mobile assets (Oates, 2001), and in the second, the fiscal discipline required by monetary integration itself.
Consequently, domestic cost reductions as an instrument for correcting competitiveness imbalances became privileged in the strategic selectivities of the new EU architecture. This fact was problematic not only because it was against the Schumpeterian discourse on competitiveness that the European Commission had promoted (and that defended, as explained above, not less public investment but a different one), but also because the correlation between lower wages and increased competitiveness is rejected by several economists (see, e.g., the classical study of Kaldor (1978)). Despite these facts, strategic voices within the neo-liberal discursive coalition, like for instance BussinessEurope, the biggest confederation of European employers' organizations, overtly defended that the main determinant of the evolution of competitive advantages resides in wages. As expressed by its president, George Jacobs, in the 2003 Macroeconomic Dialogue with Social Partners: Europe must accelerate its effort at home to improve its competitiveness. It is therefore important to improve the labour market's ability to react, allowing wages to reflect productivity differences, promoting mobility and improving incentives to take a job. Furthermore, let me emphasize that the Commission should also propose measures promoting entrepreneurship (Jacobs, 2003) In this sense, the ordoliberal paradigm on European integration propelled a pathway towards monetary integration that assumed as inevitable the painful nature of this process. It was in fact a regime of "competitive austerity" (Albo, 1994: 147), that is, a system in which each country reduces its domestic labour costs, and thus its internal demand, and thus its economic growth 8 . Not surprisingly therefore, this discourse saw as productive the exclusion of politics from the pressures arising from democratic economic policy-making. Economic convergence was expected to occur as a consequence of the market pressures arising from monetary integration itself and in this sense it can be qualified as a market-driven model of monetary integration. Given these sets of elements, it would not have been surprising if Europe's social democrats had taken a negative stance towards the EMU. Nevertheless, they supported the project once again. How then were the social democratic aspirations articulated within the prointegration discourse?
The post-Maastricht scenario: searching for a new European social democratic strategy The re-articulation of the ESM. As many scholars have observed (Lightfoot and Wilde, 1996: 10;Bieler, 2007: 114), European social democracy supported the EMU process with a "yes, but". Although the Maastricht treaty did not include any significant step towards the building of a social or fiscal union, since in the social policy field the only outcome was a modest Social Protocol, Kohl, Mitterrand and Delors agreed that these unions would be developed at some point in the near future (Marsh, 2011: 154). Therefore, once again, social democrats assumed that while most of their demands were not feasible in 1992, the integration in the monetary domain would foster functional spill-overs towards the rest of policy areas. For the European socialist parties, the Maastricht Treaty therefore constituted "a step in the right direction" (PES, 1994).
In fact, the political scenario that was opened in the late 1990s seemed to provide the long-awaited window of opportunity for developing these social democratic aspirations. By the end of the twentieth century, the predominance of social democrats in European national governments was remarkable, since they were not only governing in twelve out of fifteen EU states, but also in the largest ones (United Kingdom, Italy, France and Germany) and in the richest ones (Austria, Sweden, Denmark, Finland and Netherlands). Such a correlation of forces had never existed before, so that it represented an ideal opportunity for changing the neo-liberal course of European integration.
The common social democratic strategy for reforming the monetary union was settled by the so-called Larsson Report in 1993. This report aimed at developing a renewed social democratic approach to the integration project by articulating in a "two-pillar strategy" both the traditional social democratic demand of more expansionary macroeconomic policies at the European level as well as new proposals on gender equality, active labour market policies and welfare state reform (PES, 1993). The inclusion of these last points constituted a remarkable shift, since it meant the assumption by European social democracy of some of the neo-liberal critiques to the welfare state: We will establish [...] a new productive balance between security and change [...] Therefore we will [...] reform welfare policy so that it gives good incentives for work and employment. Social protection, taxation and ARTICLE PALGRAVE COMMUNICATIONS | DOI: 10.1057DOI: 10. /palcomms.2017 unemployment insurance have to be made coherent to open doors to job opportunities and ways out of poverty (PES, 1993: 8).
Such an approach, which later was defined as a "social investment perspective", rested on the idea that social policy provisions should constitute investments aimed towards enhancing labour skills and human capital rather than just means of providing social protection. In the words of some of its main theorists, it was a perspective that, although largely in agreement with the Keynesian assumption that a strong economy requires a strong state, advocated for an "active welfare state" that had to be capable of ensuring the sustainability of public finances while "upholding a different, knowledge-based, economy" (Vandenbroucke et al., 2011: 6). In policy terms, the option was to use social policy for improving the labour supply rather than for simply repairing damage caused by market failures.
At the European level, this implied a shift from "Eurokeynesianism" (Aust, 2004), understood as an attempt to rebuild welfare institutions at the European level, towards an "innovation-driven economic policy" (Escalona and Vieira, 2014: 32) that perceived European integration as an instrument to raise European competitiveness. The social democratic approach to European integration of the 1980s, based on the building of an ESM as a maginot line of resistance to globalization pressures, was sidelined during the nineties by this new approach highly influenced by Third Way philosophy. Thus, although the "new social democratic" discourse incorporated straightforward critiques to neo-liberalism, especially regarding the latter's approach to unemployment uniquely based on the reduction of labour costs, it clearly also articulated pivotal elements of neo-liberal thinking, particularly regarding its critique to the passive welfare state and its lauding of competition.
The overall approach of the Larsson Report was also displayed by the 1993 Delors' White Paper on Growth, Competitiveness and Employment (European Commission, 1993). This White Paper, which consolidated competitiveness as the empty signifier of EU discourse, settled as the two central objectives of the EU economic policy "helping European firms to adapt to the new globalized competitive situation", and "exploiting the competitive advantage associated with the gradual shift to a knowledge-based economy" (ibid: 62). To achieve these aims, the Commission established a "policy of global competitiveness" based on (1) a "macroeconomic framework which instead of constraining market forces […] supports them", (2) reforms "aimed at increasing the competitiveness of European industry and removing the rigidities which are curbing its dynamism" and (3) "active policies and structural changes in the labour market" (ibid: 62-67).
As we can observe, the enthusiasm with which social democrats endorsed this new approach to social policy is explained, to a large extent, because they saw in it an integral element of the new form taken by European capitalism during the 1990s, which they defined as a "knowledge-based economy" (KBE). Once the Fordist mode of accumulation had run out of steam, social democrats envisaged a "new high-tech Fordism" capable of restoring economic growth and creating the conditions for social inclusion policies (Escalona and Vieira, 2014: 30). As explained by Portuguese socialist Maria João Rodrigues (popularly known as "the mother of the Lisbon Strategy"), the new knowledge economy constituted "a strategy for international competitiveness and social cohesion": With globalization, nations are competing to attract investment, which, on one hand, depends increasingly on the general conditions supporting business competitiveness. On the other hand, business competitiveness depends increasingly on the capacity to answer just in time to the specific needs of the customer. This involves managing a greater amount of knowledge with the intensive use of information technologies. Knowledge is becoming the main source of wealth of nations (João Rodrigues, 2002: 2) Given that a KBE was presumed to be featured by continuous knowledge exploration, mutual learning between organizations and path-breaking innovations, its labour market structures needed to be more porous so as to permit continuous adaptation and insertion of new ideas from the outside. In this manner, the promotion of the social investment perspective on social policy was accompanied by the deployment of the KBE as the dominant economic imaginary and key focus of competitive strategies. In the 1995 Barcelona Declaration, the PES completely endorsed this new discourse, linking, via investment in knowledge and skills, the creation of employment to increases in competitiveness: Investing in knowledge should be a priority objective. The unemployed should have effective possibilities in their search for work by updating their skills, taking temporary employment and developing activities in the local community.
Employment is the priority and competitiveness the instrument. This must be the basis for the defence of and innovation in the welfare state Productivity of businesses has to be improved, their costs contained and their activities orientated towards fields which offer comparative advantages. The operation of the competitive markets must be stimulated, technological innovation must be promoted, equitable salary agreements must be guaranteed and competitively factors must be promoted (PES, 1995) However, as stated above, the European socialist strategy under the EMU also emphasized the need to rectify the ordoliberal architecture of the EU. For the PES (1995), "The implementation of Monetary Union should be accompanied by increased coordination of economic, budgetary, fiscal, employment and social policies". In this front, the social democratic discourse was again mainly articulated by the Prime Minister of the French government, in that moment Lionel Jospin, although now in alliance with the German Finance Minister Oskar Lafontaine. In line with the Delorist model of monetary integration, the socialist parties defended a rewriting of the Stability Pact-re-naming it as "Jobs and Growth Pact"-based on the modification of three core elements of the EMU: the prevalence of sound money policies, the independence of the ECB and the absence of fiscal policy coordination (Howarth, 2002: 357). They also argued for an extension of state activism to the European level through the institution of an economic government (Lafontaine and Strauss-Kahn, 1999), whereby the European Council would provide broad orientations for economic policy.
Governing the EU. Some advances in the building of the social dimension achieved during the late 1990s and early 2000s seemed to give credibility to the new social democratic strategy. What has been called the "social moment" (1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)   of the integration process enjoyed its highpoints in the adoption of the European Employment Strategy (EES) in 1997, by which employment policy became a "matter of common concern" 9 for the EU, in the development of the so-called Cologne Process (1999), through which procedures for a European macroeconomic dialogue between European public authorities and the social partners were put in place, and in the launching of the Open Method of Coordination within the Lisbon Strategy (2000). These measures, together with others like the extension of cohesion funds, the adoption of the Charter for Fundamental Social Rights (2001) or the approval of the 1998-2000 Social Action Programme seemed to be initial steps towards the construction of the largely awaited social dimension of the EU.
Nevertheless, these set of achievements were not only "modest" and "limited" (Manow et al., 2004: 30), given that any transfer of competences in the social policy field was consciously avoided, but also significantly distinct to those envisaged by the social democratic discourse 15 years before. For instance, the EES, accorded in the first Inter-governmental Conference with a majority of social democratic leaders in 1997, reflected the new social democratic discourse on "employability" and thus posited as the primary aim the promotion of labour market flexibility (European Commission, 1997). Its four official pillars were employability, entrepreneurship, adaptability and equal opportunity (ibid.). Through them, the EES attempted to pursue the traditional social democratic aim of full employment, but not by relying on Keynesian macroeconomic instruments but on the development of a competitive, knowledge-based economy. The ambition was to articulate the aim of full employment with the acceptance of supply-side economic policies as the best governmental instruments in the era of globalization.
The new catchword "flexicurity" (coined by the social democratic Prime Minister of Denmark Poul Nyrup Rasmussen (Jørgensen, 2000)) came to synthetically condense this paradigm, being a key nodal point in the re-articulation of the EU's social discourse under the hegemony of competitiveness imaginary (see, for example, European Council, 2006). The notion of flexicurity sought to combine in a "magic triangle" a flexible labour market with few restrictions on employers' hiring and firing of workers, a highly developed activation system that facilitates re-entering the labour market and a generous unemployment insurance scheme (Martín, 2008: 14). Unemployment was therefore linked to a lack of adequate skills to fill the new KBE, so that social policies should invest in training and support labour market fluidity (European Commission, 2007: 3).
Hence, in the context of increasing global competition, social democrats perceived in flexicurity the way to preserve the ESM while improving competitiveness (see, for example, Rasmussen and Delors, 2007). Nevertheless, this strategy, given that fostered recommodifying pressures on labour power instead of decommodifying ones, implied an important shift in the social democratic identity. Such a discursive transition is better understood if we take into account that, as explained above, the social democratic point of attachment, the ESM, to the integrationist discourse since the mid-1980s represented a clear subaltern "subject position" within such a discursive field because of the strategic selectivities of the EU's hegemonic articulation. In this manner, it was not surprising that, in the absence of great dislocations, the flow of meaning of the ESM was constantly modified by the neo-liberal radiator core constructed around the empty signifier of competitiveness. Paradoxically enough, the same quality that had enabled the ESM to become a nodal point capable of representing the social democratic discourse and even modifying the neo-liberal one, namely its capacity to resonate with a wide set of interests (Bradanini, 2009), was also the quality that facilitated its openness to the influence of contending discourses. Thus, while references to the ESM steadily increased in official documents of the Commission (Vaughan-Whitehead, 2015: 11), they were progressively deployed for justifying market expansion instead of market correction.
The 2000 Lisbon Strategy, and overarching agenda of socioeconomic reforms for the next decade, deepened in this Schumpeterian discourse. It was launched with the broad aim to make Europe, by 2010, "the most competitive and dynamic knowledge-based economy in the world" (European Council, 2000). According to the Lisbon Agenda, this aim was going to be achieved by "completing the internal market", "stepping up the process of structural reform for competitiveness", "modernizing the European social model", "combating social exclusion", integrating financial markets, regulations and services and stimulating technoscientific innovation (ibid.). In the social policy field, the main reform was the introduction of the Open Method of Coordination, a new governance instrument aimed at extending "best practices" in a range of social policy areas through forms of voluntary coordination.
These sets of social policy measures were not accompanied by any advance towards the building of European macro-economic institutional arrangements, as Delors himself has recently admitted (Delors, 2013: 175). In this dimension, in the 1997 Amsterdam Summit, Jospin only obtained "a face-saving compromise" (Howarth, 2002: 358) that involved the addition of the word "growth" to the stability pact, so that no reformulation of the structure of the EMU was agreed. Although Jospin's socialism and its associated European vision were not especially leftist, and in fact they were known as Réalisme de Gauche (Clift, 2002), they were nevertheless challenging the "there is no alternative" neo-liberal discourse. Hence, European integration remained a form of "negative integration" (Scharpf, 1999) that did not provide a strong legal basis for the reconstruction of market-correcting mechanisms at the European level.
In last instance, this implied that the social investment discourse articulated by the new social democracy of the 1990s was undermined by strategic selectivities of the EU. The ambition of the social investment approach was to combine the contradictory demands of achieving higher levels of employment and at the same time meet the Maastricht principles, the implications of which were deflationary. In this sense, a very important obstacle to the social investment strategy was not only that the social dimension of the integration process relied on the non-binding mechanisms of the Open Method of Coordination, but also and especially that it had to compete with an opposed and stronger ordoliberal monetary framework. Hence, the pressures to reduce public investment coming from the design of the Eurozone, as well as the strongly institutionalised internal market, curtailed the social investment necessary to promote competitiveness. The result was a general retrenchment of the European welfare states and the creation of what Martín (2008) called an "ESM of low cost".
Despite this fact, European social democratic parties remained completely loyal to the EU project. Indeed, the period of social democratic predominance in Europe coincided with the consolidation of the neo-liberal hegemony in the EU: as several empirical studies have shown, during the first years of the new century, a finance-driven accumulation strategy stimulated by the Financial Services Action Plan of the Lisbon Strategy was developed throughout Europe (Birch and Mykhenko, 2013), while both the Lisbon agenda and the EES were used in different countries for justifying tighter forms of labour activation and flexibility (De la Porte and Jacobsson, 2012). Furthermore, the emergence of an egalitarian alternative to neo-liberalism became a remote possibility, and in fact, the intellectual hegemony in Europe became more liberal than ever. Nevertheless, as the concluding section explains, this apparent contradictory situation -the simultaneous wave of electoral successes of social democratic parties throughout the EU and the parallel deepening of ARTICLE PALGRAVE COMMUNICATIONS | DOI: 10.1057DOI: 10. /palcomms.2017 neo-liberalism in Europe-can be read from the theory of hegemony as precisely the best proof of the consolidation of neoliberal discourse.
Conclusions: the dangers of the narrow strategic selectivities of the European integration project The theoretical framework articulated in this article understands that policy analysis cannot be reduced to the description of institutional "data" nor can it take political phenomena as epiphenomenal expressions of economic or technological developments. Instead, it assumes that political analysis must seek to "open" social objectivity and interpret the sum of elements that are discursively articulated in each historical period. In this manner, policy analysis becomes the task of uncovering the meanings and correlations of forces that are condensed in concrete institutional constructions.
In our case study, the neo-liberal "need to improve competitiveness" became a depoliticised and apolitical normative imperative that came to guide the construction of the EU to the extent that it was assumed by those parties claiming to confront neo-liberalism. A discourse is hegemonic when, although defining a terrain of struggle that compels its political adversaries to assume its own values, it is universally accepted as objective within a society. As Thatcher once proudly declared, the best legacy of her mandate was Blair's New Labour. Indeed, this is what hegemonic struggles are about: not about the competition for the imposition of certain policy aims, but about the definition of the rules that govern competition itself. Thus the strategic selectivities sedimented in the institutional architecture of the EU favoured that, to electorally defeat neo-liberal forces, the socialist parties had to look as much as possible like them. As said by one of the right-wing crafters of the post-Pinochet's Chilean constitution, it was "about making a constitution that even when the adversary rules he does it in much the same way as we would" (quoted in Errejón, 2014: 8).
This historical development helps to explain why social democracy has failed to benefit from the crisis that has engulfed the Eurozone since 2008. One on hand, despite the fact that the dramatic forms that the global financial crisis took within the EU uncovered important dysfunctionalities in its ordoliberal design, the crisis management has privileged a solution (supported by social democratic parties) based on the strengthening of the technocratic mode of European integration as well as well as on the implementation of structural reforms aimed at improving competitiveness. On the other hand, not surprisingly, social democratic parties have lost most of the EU governments since 2008. In May 2017, social democracy only governs in 6 of the 28 member states (Italy, Portugal, Sweden, Czech Republic, Slovakia and Malta-the only state where it governs alone), while in some other countries, like Netherlands, United Kingdom or Greece, it has become a residual force. Indeed, from the neo-Gramscian approach outlined in this article, such a process of marginalization can be interpreted as the result of the successful articulation-around the empty signifier of competitivenessbetween the social democratic and neo-liberal discourses, since the conditions that enable the emergence of counter-hegemonic resistance zones are the gaps left inside a hegemonic articulation (Laclau, 1979). We can therefore see in the very deepness of the articulation between social democracy and competitiveness both the condition for the return to power of most European social democratic parties in the years before the crisis as well as the cause behind the incapacity of these parties, once the crisis began, to construct an alternative to the neo-liberal discourse.
Nevertheless, this radical restraining of the range of political choices available within the European institutional order might produce destabilizing tendencies for such an order in the medium term. The technocratic fantasy assumed by the Third Way social democracy of reducing politics to the mere administrative management of non-conflicting affairs might as a result leave a large segment of their popular constituency-probably those more unable to "compete in the market"-orphan of an alternative to neo-liberalism 10 . In fact, cannot we think in this post-political closure of the discussion of issues that most affect social conditions of the subaltern sectors as constituting the main condition of possibility for the emergence of Eurosceptic populist forces throughout the EU? Notes 1 Branded by Escalona and Vieira (2014) as a "Faustian pact". 2 By 1982 socialist parties were only in power in Greece and in Italy. 3 This strategy was defended by Jean-Pierre Chevènement, Ministre d'Etat and leader of the left-wing faction of the PS (Featherstone, 1988: 128). 4 On the ideological nature of the U-turn, see Bliek and Parguez (2008) and Cameron (1996). 5 The comparison with Japan and the United States became really pervasive in the policy documents of the EC during the 1990s. For instance, in the 1993 White Paper on competitiveness one finds up to 58 times the EU compared to Japan (European Commission, 1993), while the 2000 European Competitiveness Report permanently evaluates EU's economic performance in relation to the United States and Japan (European Commission, 2000). 6 Article 6.1, SEA. 7 The automaticity of the penalties was a German demand that was excluded from the final project (Wilsher, 2014: 249) 8 Cycle only countervailed by the incessant accumulation of debt that the soft binding constraints of the SGP allowed. 9 Article 126, Treaty of the European Union. 10 In this sense, sociological analysis support the idea the increasing detachment of lower classes constitutes one of the main factors behind the steady electoral decline of European social democracy (Moschonas, 2009a).