How the Eurozone disempowers trade unions: the political economy of competitive internal devaluation


 The marginalization of trade unions was a notable feature of the sovereign debt crisis in the Eurozone periphery. However, governments have recently imposed liberalizing reforms against union protests in the Eurozone core too. We argue that organized labour loses influence across the core-periphery divide because the ‘new economic governance’ puts national governments under enhanced pressure to compete against each other on wage and labour market flexibility—a process known as competitive internal devaluation. The article illustrates this argument through comparative quantitative indicators of liberalization and qualitative process-tracing in three core countries. Whereas Germany’s outstanding competitiveness position allowed its unions to extract significant concessions, their counterparts in France and Finland faced unprecedented defeats from governments aiming to restore economic growth by closing down the competitiveness gap to Germany. Our findings highlight the class power implications of the Eurozone’s reliance on the labour market as the main economic adjustment variable.


Introduction
The impact of the Eurozone crisis on the trade unions of the periphery has been adverse and deep. Faced with austerity and competitiveness demands, governments of virtually all partisan colours pursued reform strategies that rejected union demands to facilitate labour market liberalization against all odds (Armingeon and Baccaro, 2012;Culpepper and Regan, 2014). These periphery countries, so the logic goes, have faced not only an asymmetric shock to their disadvantage (Hall, 2014;Johnston and Regan, 2016), but also display trade unions that have become too weak to influence public policy (Culpepper and Regan, 2014). Comparative studies, therefore, emphasize that while trade unions of the Northern European 'core' continued to receive policy concessions from the state during the crisis, their counterparts in the Southern European 'periphery' lost any significant influence over policy change (Marginson, 2015;Lehndorff, 2015;Lehndorff et al., 2017).
In this article, however, we argue that trade unions of the Eurozone core are not insulated from such defeats in policy-making, which we define as instances of labour market reform that contradict the union confederations' publicly stated policy preferences and provide no single concession to them. While the previous literature highlights the exclusion of union preferences in periphery countries during the crisis period from around 2010 to 2014, we observe similar patterns of union marginalization in core countries during the post-crisis period since 2015. 1 We attribute the demise of union influence in policy-making across the core-periphery divide to the demands posed by the Eurozone's institutional framework to stimulate the economy through a downwards adjustment of wages and thus prices. By reinforcing the emphasis on cost competitiveness in the availability of domestic growth strategies, the EU 'new economic governance' puts national governments under enhanced pressure to compete against each other on wage and labour market flexibility-a process known as competitive internal devaluation. Whilst not exclusive to Eurozone countries, we contend that this process is more pronounced therein, due to institutionalized constraints on fiscal spending and wage growth in a common currency union. Whereas the long-term economic prospects of such a concerted strategy are questionable in functional terms (IMF, 2012(IMF, , 2016, the political implication of internal devaluation is the disempowerment of trade unions in the interest of generating short-term business confidence. This article uses comparative quantitative indicators and qualitative case studies of France, Finland and Germany to illustrate the impact of competitive pressures on the policymaking influence of trade unions in the Eurozone core since the global financial crisis of 2008, with a particular focus on the reform activity of the past 5 years. We select cases that display variation in the degree of competitive pressures as a source of structural union power (i.e. our explanans) and thus union influence in policy-making (i.e. our explanandum), while providing grounds to address alternative 'domestic politics' explanations at the same time. Whereas German unions received concessions from the state by benefitting from the structural advantage of an outstanding competitiveness position, trade unions in France and Finland suffered defeats from governments imposing liberalizing reforms in an effort to close down the competitiveness gap to Germany. Our case selection puts our argument to a hard test, as the process of internal devaluation against union protests took place under both right and left governments across different varieties of capitalism and trade unionism.
1 With the term 'crisis period', we essentially refer to the period by which the global financial crisis from 2008 onwards turned into a sovereign debt crisis in the Eurozone periphery. In our view, this crisis period ended in around 2013-2014, as the ECB's (2012) previous shift in 2012 to a policy of 'cheap money' had materialized in sustainably reducing the spread between the sovereign bonds of the Eurozone periphery on the hand, and Germany on the other. We thus call the subsequent 5 years from 2015 to 2020 the 'post-crisis period'.
Despite these differences, trade unions lost influence in similar ways when cost competitiveness deteriorated in a context of economic slowdown (France and Finland), whereas they (re-)gained influence when cost competitiveness improved (Germany). In other words, the Eurozone's institutional framework has turned the competitive position of a country into a crucial source of structural union power. Our findings highlight the class power implications of the 'new European economic governance', which rules out wage-led growth strategies and directs political attention on the labour market as the main economic adjustment variable. Our argument about competitive internal devaluation is in line with recent accounts of the post-Keynesian (Stockhammer, 2016) and critical political economy literature (Erne, 2015;Bieler et al., 2019;Jordan et al., 2020) in pointing to the interconnectedness of domestic reform struggles in an integrated European political economy. Although political contestation over labour market and welfare state reform remains primarily national in character (hence our primary focus on domestic-level developments), the new economic governance links the fortunes of trade unions and working people across the core-periphery divide.
We proceed as follows. In Section 2, we outline the theorized relationship between competitive internal devaluation and union influence in the Eurozone, followed by the presentation of comparative quantitative indicators on policy measures in support of our claim. Section 3 outlines our case selection strategy and methodological approach. In Sections 4 to 6, we show through process-tracing how competitive pressures put governments under pressure to dismiss trade union demands in France and Finland, whereas this pressure was absent in Germany over the entire period, leading to divergence in the direction of labour market reform. A final section concludes.

The political economy of competitive internal devaluation
The creation of the Eurozone rested on monetarist assumptions of economic convergence through central bank independence and fiscal discipline. Both conditions would incentivize the adoption of supply-side labour market reforms to ensure sufficient wage flexibility in a monetary union of diverse varieties of capitalism. Following these assumptions, preparations to Economic and Monetary Union (EMU) membership during the 1990s required governments to limit inflation and public debt levels as part of the 'Maastricht convergence criteria', while the design of the European Central Bank (ECB) followed the German Bundesbank model by restricting its mandate to the objective of price stability (excluding employment stability).
In the Eurozone's institutional framework, the governments' dominant instruments of economic adjustment have thus been wage and labour market flexibility ever since. Yet, the idea of domestic supply-side adjustment was complicated by the 'stickiness' of labour market institutions in diverse European growth models. Broadly, previous research in comparative political economy distinguished between the export-led regime in the Northern 'core' (Austria, Belgium, Finland, Germany, the Netherlands and, to a lesser extent, France) and the demand-led regime in the Southern 'periphery' (Greece, Italy, Portugal and Spain), each underpinned by historically evolved wage-setting regimes and distinct political coalitions (Hall, 2014;Johnston and Regan, 2016;Regan, 2017;Hö pner and Lutter, 2018). 2 2 Scholars of Eurozone integration are broadly unanimous in classifying France as a 'core' country (cf. Johnston et al., 2014;Regan, 2017). However, France is more reliant than other core countries on According to this literature, whilst the former group of countries possesses the coordinating capacities to ensure wage restraint for successful export performance, the latter group does not. It follows that the core has a structural advantage compared to the periphery in adjusting to the competitiveness demands of a hard currency union (Regan, 2017). Faced with this adjustment problem, recent contributions highlighted that the periphery could sustain its dependence on domestic demand through reduced costs for external borrowing and cross-border banking loans coming from savers in the core (Jones, 2015). Aggravated by weak productivity growth in the periphery, these trends reinforced long-standing patterns of 'uneven and combined development' within the EU political economy (Bieler et al., 2019), and resulted in growing current account imbalances that were at the heart of the Eurozone crisis.
However, as the nature of the crisis shifted from an economic downturn to a sovereign debt crisis, EU policy-makers in tandem with the German government attributed its origins to a lack of fiscal discipline and cost competitiveness, and thus reinforced the Eurozone's reliance on internal devaluation as the main adjustment mechanism (Armingeon and Baccaro, 2012;Pérez and Matsaganis, 2019). The first step in this direction came with the introduction of the so-called 'European Semester' in 2010, whereby the European Commission (EC) issues annual country-specific recommendations to monitor and coordinate domestic economic policies across the EU. The member states must then incorporate these recommendations into national 'reform programs' that are again assessed by the EC. Secondly, in 2011, the introduction of the so-called 'Six Pack' implied tightened surveillance mechanisms through the 'Macroeconomic Imbalance Procedure' (MIP). This is based on two instruments: (a) an alert mechanism for the early detection of macroeconomic imbalances based on a 'scoreboard' of economic indicators, which highlight unit labour costs (ULCs) developments and (b) financial sanctions in case of non-compliance with the recommendations issued by the alert system. When the EC identifies an excessive macroeconomic imbalance in member states, the concerned government has to present a 'corrective action plan', which typically includes competitiveness-enhancing structural reforms of the labour market (Schulten and Mü ller, 2013;Erne, 2015). Finally, the introduction of the 'Fiscal Compact' in 2012 requires that the balanced budget condition of the Stability and Growth Pact be enshrined in constitutional law and the structural deficit (adjusted for cyclical effects) be no >0.5% of GDP, implying tightened constraints on wage growth in the public sector. Overall, these governance innovations and the renewed focus on external competitiveness as a benchmark of imbalances enhanced 'European interventionism in the area of wage policy' (Schulten and Mü ller, 2013, p. 201).
Non-compliance with tightened fiscal rules and wage surveillance mechanisms not only bears the potential threat of financial sanctions (Erne, 2015). It may also prove detrimental to the economic credibility of a member state and thus enhance pressure from international markets by causing higher interests for debt servicing and the abstention of mobile capital internal demand, sustained historically by public spending and investment (Lux, 2015), as well as on state-led support for promoting export performance (Amable, 2017). These characteristics make France a sui generis political economy model, with some fragilities in its long-term economic performance, triggered by the slow liberalization of its state-led model of coordination since the mid-1980s (Levy, 2017). Due to these features, some authors characterize it as located on the 'perimeter' of the core (cf. Magone et al., 2016)-a characterization that we subscribe to. from foreign direct investment. In other words, the degree to which governments conform to the policy demands of the new economic governance sends signals to managers, credit rating agencies and international finance about the risks of government bonds, capital flows and future investments.
The outcome of tightened fiscal rules and wage surveillance mechanisms in a common currency union is heightened cost competition. While the new economic governance implied union-hostile interventionism in return for financial assistance in the Eurozone periphery during the sovereign debt crisis, it has started to put trade unions of the Eurozone core under pressure when their relative cost competitiveness standing deteriorated. More specifically, within the common currency framework of the Eurozone, trade unions cannot escape the cost competition exerted by Germany's aggressive wage restraint, given the absence of fiscal and monetary tools to offset possible losses of cost competitiveness. As high value-added goods tend to be produced in the Eurozone's Northern core countries, their industrial structures and export baskets have faced direct competition from Germany's internal devaluation period during the 2000s (Dølvik and Marginson, 2018, p. 411).
Given its dominant position in intra-European trade, German labour costs have become the benchmark for the Eurozone's wage surveillance mechanisms and domestic collective bargaining (CB), especially in export manufacturing. As a result, the German 'employment miracle' in the wake of its internal devaluation strategy influences the discourse around the 'requirements' of labour market reform among its trading partners (Lux, 2018). Recommendations issued under the European Semester to countries with negative growth and public finance outlooks routinely cite the imperative of restoring external competitiveness through greater wage and labour market flexibility. Employer associations, in turn, legitimize their demands for liberalizing reforms by pointing to Germany's enormous current account surpluses, which serves as a 'blueprint' for a successful economic adjustment (Lehndorff, 2014).
Whilst governments are the main enactors of internal devaluation, it is important to note that enhanced cost competition in a context of tightened fiscal rules and wage surveillance mechanisms has taken place against the backdrop of a neoliberal re-orientation of employer associations, especially driven by the preferences of large export-oriented employers. Whereas contributions in the Varieties of Capitalism tradition emphasized employers' dependence on union cooperation in skill development, a more recent literature has highlighted that manufacturing employers have turned into supporters of labour market liberalization across diverse coordinated market economies (Paster, 2012;Emmenegger, 2015;Baccaro and Howell, 2017). In the German case, for example, Baccaro and Benassi (2016) attribute the neoliberal outlook of manufacturing employers to the enhanced pricesensitivity of the machinery and transportation sectors under the conditions of globalized competition. The neoliberal re-orientation of employer associations implies a preference for weakening the bargaining power of trade unions as a way of reducing labour costs. The deregulation of employment protection legislation (EPL), tightened activation demands in a 'workfarist' direction, the reduction of unemployment benefits and CB decentralization are clear policy choices in this direction, as they all contribute to lowering labour costs directly or indirectly.
To be sure, employers' organizations are not monolithic in their orientations. For instance, recent contributions (cf. Paster et al., 2019;Bulfone and Afonso, 2020) have highlighted the preferences of Dutch and South European employers' organizations for preserving the coverage of sectoral CB, primarily to prevent wage dumping from outsiders and secure their own organizational survival. However, as these authors themselves recognize, this selective defence of specific 'coordinating' institutions constitutes a partial exception to otherwise positive attitudes towards liberalizing interventions aiming at internal devaluation-especially among export-oriented employers. Indeed, in all our cases, the policy preferences of large employers resonated with the EC's Directorate-General for Economic and Financial Affairs (2012a, pp. 103 and 104), which states that 'employmentfriendly reforms'-ranging from minimum wage reduction and CB decentralization to employment deregulation-need to 'result in an overall reduction in the wage-setting power of trade unions'.
While heightened demands for liberalization and the corresponding loss of union influence in policy-making are by no means exclusive to Eurozone countries (Baccaro and Howell, 2017;Rathgeb, 2018), we argue that tightened fiscal rules and wage surveillance mechanisms in a common currency union aggravate such pressures. Figures 1 and 2 present indicators supporting this claim. They suggest a trend towards competitive internal devaluation in the Eurozone-11 countries relative to two groups of comparable advanced capitalist economies outside the Euro area: non-Eurozone social democratic countries (Denmark, Iceland, Norway and Sweden) and non-Eurozone liberal countries (UK, USA, Australia, Canada and New Zealand). We consider the liberalization of labour market institutions as the observable manifestation of reform strategies aimed at an internal devaluation of the economy. Using the 'Liberalization database 1970-2013' developed by Armingeon et al. (2019), we compare the depth of liberalization in the main policy fields of relevance for organized labour [EPL, industrial relations (IRs), unemployment benefits and active labour market policies (ALMPs)] in three periods: 1990-1998, 1999-2007 and 2008-2013. We use the 'liberalisation index' score of Armingeon et al. (2019), which measures for each country-year the incisiveness of liberalizing reforms on the basis of depth, encompassingness and impact of the policy change introduced. Higher scores indicate more far-reaching liberalization, and vice versa. Figure 1 shows the period average of the liberalization index by policy field and country grouping.  These results bolster our account: the incisiveness of liberalization in EPL and IRs (i.e. all regulations affecting the bargaining relationship between employers and employees, such as bargaining levels, right to strike and union organizing) is consistently higher in Eurozone countries, increasing markedly after 2008. This comparison suggests that the enhanced reliance on internal devaluation since the global financial crisis in 2008 is distinctive of the Eurozone countries, rather than a general development across the advanced capitalist countries. As the 'Liberalization database' stops in 2013, as does the OECD EPL strictness index, we use the ICTWSS database (version 6.0) for indicators on CB up to 2017. It is important to capture this recent period in order to evaluate whether policies of internal devaluation gradually proceed from the periphery to the core. Our focus on CB arguably captures the most pertinent area for reform strategies aiming at internal devaluation (Dustmann et al., 2014). Figure 2 thus charts change over time in the determinant level of wage bargaining across country groupings, customarily used as an indicator of CB liberalization (cf. Baccaro and Howell, 2017;Meardi, 2018 ). We distinguish between the Eurozone's periphery and core countries to examine whether the push for internal devaluation is restricted to the countries hit hardest by the sovereign debt crisis.
The analysis highlights a trend towards greater decentralization of CB, affecting most markedly the Eurozone periphery, but to some extent also the Eurozone core countries from 2013 onwards. This is consistent with our claim that trade unions in the core are not insulated from the competitive pressures exerted by tightened fiscal rules and wage surveillance. Germany France Finland Eurozone-11 Eurozone-11 core Eurozone-11 periphery Source: Authors' calculations based on Eurostat (tipslm20).
While CB decentralization was more pronounced in the Eurozone core than in the periphery during the post-crisis period, the picture in the two comparator country groups is one of stability over the whole period of investigation. Again, the Eurozone seems to be distinctive in shifting the main costs of economic adjustment on to trade unions. Whilst an analysis of the distributive outcomes of liberalization lies beyond the scope of this article, further support for our argument can be found in the evolutionary trajectory of the wage share (see Figure A1 in the Appendix). In contrast to growing rates in Nordic and especially Anglo-Saxon non-Euro countries post-2008, the wage share declines markedly in the Eurozone periphery and stagnates in the Eurozone core since the crisis onset.
Nominal unit labour costs (NULCs) and real exchange rates (REERs), both part of the MIP's economic scoreboard, can shed light on variations in the degree of competitive pressures as a source of structural union power, which we claim are related to the degree of union influence in policy-making (Figures 3 and 4). The picture is clear-cut: while NULCs and REERs have remained low in Germany since the introduction of EMU, the cost competitiveness of France and Finland gradually deteriorated relative to the Eurozone average. The periphery's wage freezes in response to the sovereign debt crisis contributed to the deteriorated competitiveness standing of France and Finland within the Eurozone. In fact, NULC growth flatlined in the Eurozone since 2008, even though productivity growth was sluggish too. Although Germany experienced NULC growth post-2008 relative to the Eurozone average, it maintained its comparative competitive standing, underlined by its outstanding  REER indicator. We may thus assume that trade unions in Germany are in a stronger position to influence the policy-making process than their counterparts in France and Finland. These figures certainly do not 'dictate' the fortunes of trade unions in the functionalist sense that 'there is no alternative' to wage competition for countries struggling to stimulate the economy. In fact, even the IMF (2012, 2016) contends that austerity and structural reform failed to generate stable growth, especially when implemented in a concerted effort across countries (see also Pérez and Matsaganis, 2019). The economic failure of internal devaluation supports instead our class-based analysis because it suggests that the 'new economic governance' primarily intensifies class discipline over organized labour, generating political pressures rather than functional demands for liberalization.

Case selection and methodology
Following a most different systems design, our case selection strategy tests our argument about the impact of competitive pressures (i.e. our explanans) on union influence in policymaking (i.e. our explanandum), while simultaneously controlling for alternative 'domestic politics' explanations. Our cases differ in important theoretical respects but are similar in the theorized relationship between cost competitiveness and trade union influence (Table 1). To enhance the validity of our claim, we briefly discuss how our case selection helps us to rule out alternative 'domestic politics' explanations based on partisanship, union power and varieties of capitalism.
First, the fortunes of organized labour could be conceptualized as a product of partisan preferences (Hä usermann et al., 2013). Whilst we do not exclude that right-wing partisanship may be associated with union defeats in policy-making, it does not explain the observed variation. Indeed, the German red-green government imposed painful labour market reforms on the unions prior to the Eurozone crisis (Agenda 2010), and French unions experienced similar defeats under socialist governments since 2015. In contrast, conservative-led grand coalitions provided unions with significant concessions in Germany, but not in Finland.
A second alternative explanation would be that unions lost influence over policy-making because they lost members, unity or legitimacy in public opinion (Culpepper and Regan, 2014). Finnish unions, however, suffered defeats since 2015 despite their exceptionally high density of nearly 70% and their high levels of public trust. Whilst internal divisions facilitate union marginalization, they are nevertheless a long-standing feature of French IRs, which cannot account for recent changes in the extent of union influence. Thirdly, and finally, the observed variation cannot be associated with differences in varieties of capitalism and firmlevel arrangements, as all three cases have been classified as CMEs, albeit France is a 'special' case due to its legacies of state-led economic adjustment. Taken together, our case selection has several methodological advantages in examining the relationship between Eurozonerelated competitive pressures and trade union influence.
Methodologically, we use a 'systematic process analysis'-often called process-tracingto test the hypothesized causal mechanism linking competitive pressures and union influence in policy-making against the rival 'domestic politics' explanations described above (Hall, 2006;Trampusch and Palier, 2016). In this mode of causal analysis, the temporal sequence of events is consequential for the outcome of interest, which requires the researcher to specify the hypothesized process in conceptual, operational and sequential terms.
As stated in the introduction, we consider the absence of any social compensation for CB decentralization, employment deregulation, unemployment benefit cuts and tightened activation demands against union protests a case of 'union defeat'. Our definition is in line with the neo-corporatist concept of 'political exchange', which refers to a situation whereby trade unions deliver wage restraint in return for union-friendly government policies that expand welfare state arrangements (Pizzorno, 1978;Regini, 1984). The absence of any concession excludes this political exchange in favour of a unilateral reform strategy that marginalizes trade union demands. Operationally, we would observe a 'union defeat' if governments rejected trade unions from the policy-making process altogether or trade unions legitimized their consent to a particular reform purely for 'damage limitation' despite the absence of a single positive concession to them. We derive empirical evidence for union influence from the initial policy demands of trade unions, the subsequent reform negotiations between governments and trade unions, and the degree to which the final reform corresponds to their initial policy demands.
So how, then, do we identify the causal mechanism leading to union defeats in policymaking? First, we expect a deteriorating competitiveness situation in a context of low economic growth to cause recommendations for the stimulation of internal devaluation through the European Semester and the MIP, which facilitates the agenda setting of employer associations, bond traders and credit rating agencies. We thus expect to find that official documents from the EC converged towards a strategy of internal devaluation to the detriment of unions in France and Finland, but not in Germany. The downgrading of a country's economic credibility from credit rating agencies would enhance market pressures for the stimulation of an internal devaluation, especially in case of non-compliance with the EC's policy prescriptions.
Secondly, to link government action to the Eurozone's institutional framework, we need to find official statements where governments legitimize their reform plans against union protests with reference to competitive constraints. We also use public opinion data to evaluate whether government's reform plans contradict the policy preferences of electoral majorities. Thirdly, trade unions should subsequently respond with negative reactions in press statements or street demonstrations and industrial action. Our causal mechanism expects the final policy output to contradict union preferences and provide no positive concession if competitive pressures are high (France and Finland). In the absence of competitive pressure, however, we expect a more balanced distribution of class power, which creates opportunities for trade unions to enhance their policy influence (Germany).
We ensure validity by combining evidence from different sources about the sequence of labour market reform: official documents, parliamentary debates, media interviews, survey data and secondary sources. The consulted primary sources from the EC and the European Council on country-specific recommendations, MIP and excessive deficit procedures (EDPs) are listed in the Appendix. To contextualize the degree and direction of domestic labour market reforms, we draw on secondary sources and official country reports. Understanding the dynamic of trade union disempowerment requires taking into account the sequence of competitive internal devaluation over time, which initiated in today's paradigmatic 'core' country during the 1990s.

A German Sonderweg: the rise of union influence-after its fall
The case of Germany illustrates how the absence of competitive pressures creates opportunities for trade unions to extract policy concessions (Table 2). Unlike the 'periphery' countries, the German economy faced serious cost competitiveness problems prior to the Eurozone crisis. As of the 1990s, high non-wage labour costs and rigid wage determination placed a burden on private-sector employment, while growing debt levels made increases in public-sector employment fiscally unsustainable. At the same time, the internationalization of supply chains led to the mass offshoring of low-skilled manufacturing jobs (Baccaro and Benassi, 2016). Germany became thus emblematic of the 'welfare-without-work' problem (Iversen and Wren, 1998). In response, employers' associations, especially in the metal industry (Gesamtmetall), leveraged the threat of relocation to deviate from industry-wide standards (Dustmann et al., 2014) by withdrawing from sectoral agreements and pushing the state for liberalizing labour market and welfare reforms. This economic environment in combination with high unemployment led the Red-Green government under Gerhard Schrö der (1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006) to pursue unilateral reforms that aimed to reduce non-wage labour costs, deregulate employment at the margins, incentivize faster job placement and enhance pressure on the (long-term) unemployed to take up jobs deemed 'suitable' (Agenda 2010). Therefore, a social democratic-led government no longer secured unions a privileged position in the reform process. The above internal devaluation period coupled with fixed exchange rates greatly enhanced Germany's cost competitiveness relative to its Eurozone neighbours (see also Figures 3 and 4). As a result, in 2012, Germany's current account surplus was roughly as high as the combined current account deficits of Greece, Italy, Portugal and Spain (Armingeon and Baccaro, 2012, p. 259). In line with our expectations, given this competitive advantage, the pressure for union exclusion in Germany was largely absent during the crisis period. On the contrary, the grand coalition actively reached out to the unions to pre-empt political conflict before the 2009 federal election (Urban, 2012). The ensuing revival of tripartite consultations-also known as 'crisis corporatism' (Urban, 2012)-led to labour-inclusive crisis policy responses. Mainly, unions agreed on economic stimulus programmes and the stabilization of employment in the exposed manufacturing sector (Dribbusch and Birke, 2014). In this respect, the expansion of short-time work, which provides subsidies to employers who reduce workers' hours rather than laying them off, contributed to prevent a translation of GDP contraction into growing unemployment, thus fuelling a 'job miracle' (Krugman, 2009). While labour-inclusive 'crisis corporatism' (2008)(2009)) was a temporary phenomenon, the subsequent Conservative-Liberal government (2009-2013) also refrained from political conflict with organized labour.
The conservative-led grand coalition (since 2013) fulfilled then major policy demands from the German union confederation (DGB): the introduction of a statutory nation-wide minimum wage (2013), the temporary re-introduction of early retirement (2013), and the re-regulation of temporary agency work (2016) and fixed-term contracts for which there is no 'valid' reason to incentivize permanent employment (2018). All these steps helped to deviate from the previous internal devaluation path, albeit to a moderate extent (Lehndorff, 2016). The minimum wage of e8.50 gross per hour (2015)-about 50% of the median wage and at a medium level from an EU-comparative perspective-aimed to curb low pay in a context of declining CB coverage. Similarly, the re-regulation of temporary agency work targeted the margins of the (manufacturing) workforce by re-introducing a maximum duration of 18 months and equal pay after 9 months.
These policy concessions conform to our hypothesized connection between the intensity of competitive pressures and unions' relative policy influence. Specifically, the absence of competitive pressures deprived German employers of the capacity to block union demands. First, record-high current account surpluses and employment rates called into question the warnings of the employer confederation BDA against the minimum wage's potential 'burden' on job creation (BT-Drucksache, 2014a, pp. 10 and 11; European Council, 2014, p. 23). As we would expect according to our theorized causal chain, the government invoked the absence of competitive pressures as a reason for supporting the minimum wage. The Christlich Demokratische Union Deutschlandsspokesperson for labour market and social affairs pointed to changed economic conditions to legitimize his party's support for the minimum wage, arguing that 'the situation in 2014 is different from the one in 2002 and 2003' (BT-Drucksache, 2014b, p. 4096). Secondly, Germany's outstanding competitiveness also facilitated unions' efforts to influence public opinion in their preferred direction, by undermining the political discourse supporting aggressive wage moderation. In 2009, for example, public support for a minimum wage increased to 85% in nation-wide opinion polls while the amount of respondents perceiving economic conditions as 'unfair' increased up to 70% in 2013 (Marx and Starke, 2017).
Overall, competitive pressures were no longer a credible threat to block union demands. Rather, their absence facilitated a more even level-playing field between labour and capital. As Lehndorff et al. (2017, p. 215) argue, the 'gradual comeback' of German organized labour cannot be understood without recognizing the 'improved structural power' of trade unions and their successful public campaigns after a decade of pronounced internal devaluation. To be sure, these measures have not substantively shifted Germany's trajectory away from its competitiveness strategy. Rather, post-crisis union achievements in Germany should be seen as moderate deviations that help to sustain political support for Germany's exportled growth model (cf. Mabbett, 2016). Nonetheless, they indicate how the fortunes of trade unions are enhanced when the issue of cost competitiveness cannot be politically mobilized to push for liberalization against union protests.

Catching up with Germany: French trade unions after the crisis
France is exemplary of a core country where the gradual deterioration of competitiveness vis-à -vis Germany led to trade union defeats in the aftermath of the Eurozone crisis (Table 3). Whilst the impact of the 2008-2009 financial crisis had been comparatively small, unemployment growth and the initial counter-cyclical crisis response aggravated long-standing fiscal deficit problems. As the EC activated an EDP from 2009, the French state had limited capacity to use fiscal spending in order to stimulate internal demand for economic recovery. By 2012, France also experienced a growing trade deficit and decreasing export market shares (OECD, 2015), with key industries such as car manufacturers Peugeot-Citroen suffering particularly.
In line with our expectations, the deterioration in external competitiveness vis-à -vis Germany generated demands from the EU Commission and employers for internal devaluation (cf. Figures 3 and 4). The EU Semester recommendations in 2012 and 2013 flagged up the divergence of French and German ULCs (European Council, 2012Council, , p. 27, 2013a and recommended measures to reduce labour costs (especially employers' social security contributions), expand firm-level flexibility in wage-setting and working conditions and restrain minimum wage growth. The implementation of 'comprehensive structural reforms' was an explicit condition for France to obtain fiscal leeway and an extension until 2015 to rectify its excessive deficit situation (European Commission, 2013a, p. 6). The Commission's recommendations bolstered the requests of domestic employers' confederation Medef, which repeatedly voiced demands over 2012-2013 for a 'competitiveness shock to help the exposed sectors' (Financial Times, 2012a), advocating for cuts in labour taxes to close the 'gigantic gap' with Germany's labour costs (Financial Times, 2012b). Pressure from the international financial markets piled on when the rating agency Moody's downgraded France's bond rating in November 2012, citing declining export competitiveness and excessive EPL rigidity as key risk factors (Moody's, 2012).
Against this backdrop, the government's attention shifted to export-promotion as the privileged avenue of economic recovery, making external competitiveness the focus of political action. Initially, Hollande attempted to pursue a negotiated reform strategy by involving the unions in tripartite discussions. Social dialogue resulted in two agreements, signed only by the moderate unions CFDT and CFTC: an inter-confederal concession bargaining agreement (ANI) for 'maintaining employment' ('sécurisation de l'emploi') in 2013, which created a framework for firms in financial difficulties to derogate temporarily from sectoral agreements (Rehfeldt and Vincent, 2018, p. 160); and a 'responsibility and solidarity pact' (pact de responsabilité) in 2014 that reduced employers' social contributions to lower labour costs (Lux, 2015). Employer associations and the EC, however, considered these measures as excessively timid because they stopped short of addressing established 'rigidities' in labour market regulation.
As growth performance remained sluggish over 2014 and France struggled to meet its deficit-reduction targets by 2015, pressures to continue with labour market liberalization intensified. The MIP in-depth reviews (European Commission, 2014a, 2015a) again flagged up cost competitiveness as a primary issue to be addressed to rectify France's excessive imbalances; and numerous CSRs issued between 2014 and 2016 reinstated the priority of reducing labour costs and rigidities (Clauwaert et al., 2016). In 2014, the Commission restated the importance of France's relative competitiveness standing, highlighting that although French ULC developments had been mostly aligned with Eurozone trends, 'the difference between ULC developments in France and in Germany since 2000 has resulted in a deterioration of the relative cost competitiveness of France vis-à -vis Germany' (European Commission, 2014a, p. 27). The Commission reinforced its recommendations by threatening in 2015 the activation of an excessive imbalance procedure in case of insufficient progress in structural LMP reforms (European Commission, 2015b). Again, these pressures contributed to empowering domestic employers' demands. For instance, in December 2014, public protests led by the SME employer confederation CGPME called for lowering the tax burden on companies, simplifying the labour code and introducing firm-level flexibility to derogate from the 35-hours working week (Financial Times, 2014).
Our theoretical argument would expect governments to respond to these competitive pressures by implementing liberalizing reforms aimed at reducing labour costs. This was indeed the case. From 2015, Hollande subscribed to the idea that fostering external competitiveness through supply-side reforms was the only available pathway to re-launch growth within the Eurozone's constraints. He thus adopted a strategy of unilateral labour market liberalization, stating that 'any measure using the budget to relaunch activity would increase our public debt and worsen our foreign trade [deficit]' (Financial Times, 2014). To be sure, the implementation of liberalizing LMP reforms was aligned with the preferences of the Socialist Party's 'modernist' wing, which had embraced Schrö der's Agenda 2010 as the example to follow in order to reinstate competitiveness (Amable and Palombarini, 2018, pp. 148-151). However, in the absence of mounting external competitive pressures, it is unlikely that the Parti socialiste would have pursued such a politically risky path.
First, in 2015, the government pushed through, against joint union opposition, measures deregulating rules on Sunday trading (Eurofound, 2015). Secondly, it introduced the 'Macron Law' and 'Rebsamen Law' to liberalize criteria for collective redundancies (Clauwaert et al., 2016) and tighten the conditionality of unemployment benefits and social assistance (Milner, 2017). Thirdly, in 2016, the government unilaterally implemented the controversial 'El Khomri Law' (loi travail), without parliamentary debate. The reform increased the scope for company-level bargaining to derogate from labour law and sectoral agreements in working time regulation, overtime pay and wage setting, partly reversing the hierarchy of the CB framework and allowing agreements to be concluded even by minority unions representing 30% of the workers (Amable, 2017, p. 228); and widened the scope for economic dismissals (Amable, 2017, p. 231).
These reforms were instances of union defeats, as no compensation was offered to organized labour even though mass opposition ensued. In the face of joint union resistance, the government eventually retracted initial proposals for capping maximum dismissal severance payments (Rehfeldt and Vincent, 2018, p. 162). However, FO and CGT remained strongly opposed to the reform in its entirety (Eurofound, 2016a(Eurofound, , 2017a, alongside the bulk of public opinion (Rehfeldt and Vincent 2018, p. 162). Several strikes and mass demonstrations (the so-called 'Nuit Debout' movement) were insufficient to halt the government's intervention (Amable, 2017, pp, 232 and 233). While the El Khomri Law fits into a long-term trajectory of using decentralized bargaining to revise labour law, its degree of liberalization exceeded all prior reform attempts in French IRs (Baccaro and Howell, 2017, p. 94). Moreover, as Amable (2017, pp. 228-233) argues, it also represented a break in French party competition because it was the first time that the Socialist Party actively pushed for labour market liberalization-traditionally the prerogative of Gaullist governments.
Ensuing developments also conform to our theorized causal sequence. As France's competitiveness position remained negatively deviant from the Eurozone average post-2015, pressures to continue with internal devaluation persisted. Whilst praising the El Khomri Law (European Commission, 2017, p. 11), the 2017 MIP in-depth review restated the imperative of consolidating labour costs reductions by implementing decentralized firm-level bargaining; revising minimum wage indexation; and reforming the unemployment benefits system to strengthen work incentives (European Commission, 2017, pp. 11 and 19). Simultaneously, rising bond yields reached in early 2017 Eurozone-crisis levels. In this context, labour market liberalization became again a test for the French government's credibility, functional to reinstate market confidence (Connolly, 2017).
The newly elected centrist President Macron, therefore, implemented another farreaching labour market reform in September 2017. This went further than the El Khomri Law as it directly reduced EPL by capping the maximum amount of dismissal compensation (in line with Medef's long-standing demands); facilitated firm-level bargaining without unions in SMEs; and expanded the scope for firm-level agreements to derogate from sectoral bargaining and labour law. Even though the reform was opposed by two-thirds of the electorate (Financial Times 1/09/2017), Macron's government prioritized compliance with the EC and employers' preferences for fostering external competitiveness-explicitly citing these imperatives as the primary motivation (Eurofound, 2017b).
Unions remained critical and exercised opposition against the reform, although with different strategies (Eurofound, 2017b). The second-largest radical confederation CGT organized mass strikes in September 2017, whilst the largest reformist confederation CFDT did not join in, adopting instead a 'pragmatic' attitude (Connolly, 2017). The third-largest union, FO, also joined in the protests in November 2017 (ETUI, 2017). This belated broadening of the union opposition front did not succeed in halting the reforms, which were enacted through executive order in September 2017 and included not a single compensatory concession to unions.

North-North competition and Finnish union defeats
With different starting points and institutional legacies, the Finnish case provides an illustration of our theorized mechanism in a social-democratic regime (Table 4). It shows how the rejection of union demands may take place in the wake of a deteriorating competitiveness position despite high levels of associational and institutional union power in tandem with corporatist traditions. First, like in France, competitive pressures emerged in Finland after the crisis due to declining export shares in the wake of intensified North-North competition. Up until 2008, Finland had enjoyed an advantageous competitiveness position within the Eurozone due to wage moderation and a competition strategy centred on technological innovation favouring productivity growth (Kaitila, 2019, pp. 5-9). However, from 2012, its macro-economic problem load intensified. The decline in external demand for key exports accelerated the demise of its core sectors: the forestry industry and the electronics/information and communication technology (ICT) sector around Nokia (Kaitila, 2019, p. 12). Flagging export performance contributed to a deteriorating current account balance and steep decline in overall productivity growth (related to the problems of the high-productivity ICT sector; cf. Kaitila, 2019, p. 12). This, in turn, triggered above-average ULC growth compared with Eurozone trends in general and Germany in particular (van Gyes and Schulten, 2015, p. 159). In theory, as Krugman (2015) argued, Finland could have responded to the structural challenges of intensified North-North competition (cf. Dølvik and Marginson, 2018, pp. 414 and 415) by devaluing its currency like in the early 1990s and expanding fiscal spending to kick-start its economy at times of industrial restructuring. However, in reality, both options were no longer available within the Eurozone's institutional framework.
Consequently, as expected by our theory, the EC supported the employer association in politicizing the problem of competitiveness. First, the EC's MIP in-depth review procedure found Finland to be experiencing macro-economic imbalances for five consecutive years (European Commission, 2012b, 2013b, 2014b, 2015c, and the country-specific recommendations specifically mandated the alignment of wages to productivity developments (e.g. European Council, 2013b). Secondly, these developments empowered employers' requests for internal devaluation. In 2013, the Confederation of Finnish Industries (EK) started advocating for a wage freeze (Eurofound, 2013) and greater flexibility to align wage developments more closely with Germany (van Gyes and Schulten, 2015, pp. 159 and 160).
Against this backdrop, the focus of Finnish policy-makers also turned towards internal devaluation as the main strategy for economic recovery (van Gyes and Schulten, 2015, p. 158;Kaitila, 2019, pp. 13 and 14). The cross-bloc Katainen government (2011)(2012)(2013)(2014) pushed the peak-level employer and union federations to agree in 2013 on a 'Pact for Employment and Growth' (2013), entailing pronounced wage moderation for 2013-2015. As we would expect, the government presented this as necessary to make Finnish labour costs lower than in its direct competitor countries, Sweden and Germany (Eurofound, 2014).
However, despite wage moderation, Finland's growth performance and relative competitiveness standing did not recover. Low growth deteriorated the outlook for public finances: for the first time since the 1990s, in 2015 Finland's public debt was poised to exceed the SGP threshold of 60% of GDP. As we would expect, this triggered intensified pressures from the EC and financial markets for further internal devaluation. The Commission's MIP in-depth review again highlighted excessive ULC growth in non-tradables as responsible for Finland's cost competitiveness deterioration (European Commission, 2014b). In October 2014, the rating agency Standard & Poor's downgraded Finland's AAA bond rating, identifying as motivations a deteriorating external competitiveness standing, declining export market shares and rising labour costs (Reuters World General News, 2014). The EU Semester country-specific recommendations in 2015 and 2016 re-affirmed demands to remove rigidities in wage-setting and increase labour supply (European Council, 2015. The domestic employers' confederation EK also reinstated that in the absence of currency devaluation, further wage devaluation was necessary as the only way to 'close down the gap with competitor countries' (Bloomberg, 2015).
As expected, although the links between wage growth and Finland's competitiveness problems were empirically disputed, the incoming centre-right coalition government of PM Sipilä (comprising the Centre Party, the conservative National Coalition Party and the populist right wing Finns Party) embraced the diagnosis about excessive ULC growth (Kaitila, 2019, p. 16). Hence, from 2015, reducing ULCs to address Finland's 'sustainability gap' became the focal point of the government's reform agenda, to catch up with the internal devaluation pursued by other Eurozone member states and reassure markets. The following quote by the Minister of Economic Affairs, Olli Rehn, illustrates the government's motivation exhaustively: '. . . Case Finland [is] a case in point of a Northern Eurozone country that is going through an internal devaluation in the era of a common currency -and pursuing economic reforms in the spirit of Commission's recommendations from the years 2010-2014' (Rehn, 2016).
In conformance with our theorized causal chain, the pursuit of internal devaluation involved union defeats in policy-making. Breaking with Finland's consensual tradition of corporatist tripartite policy-making, in September 2015, the government used the 'shadow of hierarchy' to force the social partners' hand in wage-setting negotiations. PM Sipilä threatened to intervene unilaterally via legislation to implement far-reaching structural reforms (involving cuts to sick pay, overtime pay, vacation entitlements and employers' social insurance contributions) and spending cuts amounting to 1.5 billion Euro (Kaitila, 2019, p. 17). Unions denounced the proposals as violations of CB freedom and reacted with mass public protests and the first general strike in 20 years, which severely disrupted the country's public services (ETUI, 2016).
The government held back on unilateral intervention on the condition that the social partners would autonomously agree to modify wage-setting institutions to improve relative cost competitiveness. These threats were ultimately successful. In March 2016, the unions eventually agreed on a macro-concession bargaining agreement, the 'Competitiveness Pact'. This involved a wage freeze for 2017, reduced pay for public sector employees, a shift in the liability for social security contributions from employers to employees, an extension in annual working time of 24 hours (without additional compensation), and a preliminary agreement to renounce centralized peak-level CB and increase scope for company-level bargaining (Dølvik and Marginson, 2018, pp. 415 and 416). The end of national-level CB materialized in 2017, when the Confederation of Finnish Industries EK unilaterally terminated all confederation-level national agreements (Eurofound, 2017c).
Although the agreement did not imply a single positive concession, the unions legitimized their consent by claiming that the outcome was still 'better than its alternatives', and that it rescued their future involvement in tripartite policy-making (Eurofound, 2016b). Conversely, the agreement fulfilled long-standing preferences of manufacturing employers for greater wage-setting flexibility through opening clauses and cross-economy wage moderation through industry-led pattern bargaining, benchmarked against German wage costs (Dølvik and Marginson, 2018;Mü ller et al., 2018). Whilst the Finnish CB system remains relatively centralized compared to other Eurozone countries, the demise of national-level bargaining represented a defeat for the union movement, and a path-breaking shift towards industry-level wage-setting-moving Finland closer to its competitors, Germany and Sweden (Dølvik and Marginson, 2018, p. 415;Mü ller et al., 2018, p. 362).
In 2018, the Sipilä government passed legislation to intensify activation requirements for the unemployed without consulting the unions (Eurofound, 2018a). It also advanced proposals to deregulate fixed-term contracts for young people and reduce individual dismissal compensation in small companies (Eurofound, 2018b). These have triggered strong union opposition and large-scale industrial unrest (Eurofound, 2018c). The relative improvements of Finland's macro-economic position from 2018 and the election victory of the centre-left cabinet in April 2019 may provide the unions with some leeway to reverse this trend-but whether Finland will deviate from previous reforms remains to be seen.

Conclusion
Previous studies have considered the demise of trade union influence in policy-making an exclusive feature of the countries hit hardest by the Eurozone's sovereign debt crisis-i.e. Greece, Ireland, Italy, Portugal and Spain (Armingeon and Baccaro, 2012;Culpepper and Regan, 2014;Marginson, 2015). However, our quantitative evidence on liberalization in the Eurozone and our qualitative findings from France, Finland and Germany suggest that the rejection of union demands in policy-making is neither restricted in space to the periphery nor restricted in time to the sovereign debt crisis. Although the periphery countries face long-standing problems of innovation capacity, skill formation and productivity growth, France and Finland have recently confronted the same source of pressure the periphery had experienced during the sovereign debt crisis; namely, the Eurozone's in-built demand to restore economic growth by closing down the competitiveness gap to Germany. It follows that the Eurozone's institutional framework has substantially enhanced the relevance of cost competitiveness for the structural power of trade unions. Secondly, our findings also call into question constructivist accounts that document a progressive turn in country-specific recommendations towards the promotion of a more 'Social Europe' (Bakker 2017;Zeitlin and Vanhercke, 2018). In our view, the 'Social Europe' hypothesis is difficult to sustain when taking into account how fiscal rules and wage surveillance have affected union power and labour market protections in member states.
We argue that the main reason for the ongoing decline of union influence on reform trajectories across the core-periphery divide lies in the institutionalized reliance on competitive internal devaluation in the Eurozone. In this framework, member states are under pressure to compete against each other on wage and labour market flexibility when faced with an economic slowdown, which creates incentives and opportunities for governments to impose liberalizing reforms on trade unions in the interest of generating short-term business confidence. The German case of union success buttresses this claim in the sense that the absence of competitive pressures created opportunities for trade unions to partly reverse the tide of internal devaluation against employer opposition.
While the evidence of this article is restricted to what we observed in three core cases and through quantitative comparative indicators, our theoretical claims have wider application to the remaining core countries-Austria, Belgium and the Netherlands-and thus open up a new frontier for future research. The hypothesis emerging from this study is clear: export-oriented core countries, which benefited in the past from current account surpluses, are likely to come under similar pressures for internal devaluation if their growth rates slow down and competitive advantages exhaust from cuts in labour costs elsewhere or other adverse shocks. While these countries have traditionally been hard-currency regimes well equipped to strive in the Eurozone, our findings from Finland and France make it reasonable to assume that cost competitiveness may turn into the political focal point if growth slows down and alternative policy options remain constrained by fiscal rules and wage surveillance.
The extension of our argument to the smaller core countries returns us to the case in which the path of competitive internal devaluation started in the first place: Germany. The post-Keynesian literature suggests that Germany holds the key to address the Eurozone's macroeconomic imbalances by pushing for pronounced wage increases and public investment (Stockhammer, 2016). Accordingly, Germany's domestic revaluation would boost demand for imports and thus undermine excessive current account surpluses, which create more leeway for domestic economic reform among its trading partners. We find indeed successes of union politics towards this direction (e.g. minimum wage), but the competitiveness gap is still large from a comparative optic, as shown in the theory section of this article. It remains to be seen whether the German DGB manages to reverse the tide of internal devaluation, given the related threat of job losses among its powerful manufacturing unions and the cross-partisan consensus around the country's export-led growth model. The joint initiative of the main manufacturing (IG Metall and IG BCE) and service unions (ver.di) to extend CB coverage and enhance firm-level co-determination are at least further steps in this direction.
Our findings have important implications for future research in comparative political economy and welfare state research. First, the political economy of competitive internal devaluation highlights that the trajectories of national varieties of capitalism must be placed in a context of international interdependence. The Eurozone is a perfect illustration of how the ever-closer integration of nation states in capitalist world markets translates developments in one country into far-reaching consequences in others. However, much of today's comparative political economy is still based on a 'methodological nationalism' (Ebenau, 2015) that treats nation states as discrete units that can be analysed independently. Future research should thus be more attentive to international interdependencies in order to understand national differences.
Secondly, the institutionalized disempowerment of trade unions might call for a return of theoretical attention to questions of class power (Culpepper, 2015). Against functionalist perspectives, the EU new economic governance did neither solve the problem of uneven development nor produce stable economic growth as such (IMF, 2012(IMF, , 2016). Yet, it was successful in enhancing class discipline over organized labour, in line with the theoretical expectations of a critical political economy perspective (Erne, 2015;Bieler et al., 2019). Even though the state-led rejection of union demands is not new and clearly predated the Eurozone crisis (Rathgeb, 2018), the important point here is that institutional frameworks based on fiscal rules and wage surveillance mechanisms may reinforce government activism at the expense of trade unions while leaving untouched the financialized nature of the Eurozone crisis itself. Seen in this way, institutional arrangements must be conceptualized in light of their class power implications, and not their functional problem-solving capacities.
In concluding, our findings suggest that the European economic and monetary integration process is likely to shape the future fortunes of organized labour and working people. The reform trajectories of labour market and social policies cannot be understood without recognizing the Eurozone's in-built pressure to stimulate the economy through internal devaluation. We, therefore, believe that the political implication of our argument is that the fortunes of labour movements rest to a growing extent on their transnational cooperation against the 'new economic governance' in favour of concerted labour-friendly growth strategies. While differences in national conflict settings and trade union organizations have posed problems for such a transnational strategy in the past (Erne, 2015), it remains the most likely source of union revitalization in the domestic policy-making arenas of an integrated European labour market. The alternative, it seems, is a Eurozone-like race to the bottom.