Neither existing nor emerging: euro stabilization by means of European wage coordination

ABSTRACT Since the 1990s and even more so since the introduction of the euro, European trade unions have been committed to the transnational coordination of wage demands. This commitment has prevailed until today although, under conditions of persistent institutional heterogeneity, coordination attempts have continuously failed. Keeping the commitment nevertheless alive aims at retaining the option of effective wage coordination for the long-term future. In the short to medium term, in contrast, no evidence suggests that transnational wage coordination will be able to correct the distorted real effective exchange rates that have emerged since the introduction of the euro.


Wage coordination as a functional requirement for currency unions
The introduction of the euro has fundamentally increased the functional pressure on wage policies to contribute to transnational economic stabilization. This is because differences in price inflation can no longer be adjusted using nominal de-and revaluations. If such disparities cumulate, they contribute to the emergence (or persistence) of current account imbalances. 1 In order to accept this, we do not need to assume the real exchange rate to be the only driver of such imbalances. Especially in the short run, asymmetrical demand shocks affect the current account at least as much as the real exchange rate (Sanchez, Diaz, and Varoudakis 2013;Storm and Naastepad 2015). Also, we do not need to assume that the demand for all products is equally price elastic (Giordano and Zollino 2016;Athanasoglou and Bardaka 2017;Bayoumi, Harmsen, and Turunen 2011). All we need to assume is that the real exchange rate matters for international trade in the mid to long run, especially in an economically deeply integrated area such as the eurozone. The ability to synchronize national inflation rates in the medium term is therefore a necessary condition for a well-functioning currency union. Given that wage inflation is among the determinants of price inflation, 2 synchronization of price inflation requires some form of transnational wage coordination.
The lack of formal coordination mechanisms is only a problem if homogenous labor cost increases do not occur of their own accord. If the eurozone were a uniform economic area where participating countries share the same business cycle and have, in particular, similar institutions and wage negotiation practices, the synchronization of wage inflations could in principle occur without coordination. But the eurozone is not such a uniform area. As the comparative capitalism literature has pointed out, it consists of different varieties of capitalism (Hall 2014, 355-371;Nölke 2016, 141-161). Some euro members are export-oriented economies, others have small export sectors (Baccaro and Pontusson 2016, 175−207). And in particular, there are considerable differences between the wage-setting mechanisms in the eurozone, in terms of the prevalence of industry-wide collective agreements and declarations on the general obligation towards collective agreements, for instance (Hancké 2013;De Grauwe 2018, 20-21). Under such conditions, the synchronization of wage increases is unlikely to occur spontaneously and active transnational coordination may be required.
Is such transnational coordination existing or emerging? No expert in European labor relations would expect European-wide umbrella organizations to decide about and enforce wage policies which national unions would later have to implement. But the undisputed fact that transnational wage coordination in this very strict sense neither exists nor is in the pipeline does not imply that wage coordination is entirely impossible.
In this article, we will use the term transnational wage coordination in a less ambitious way. In a very general manner, coordination can be defined as making activities relate to each other with regard to a certain objective. It is therefore a behavioural pattern. Hierarchical steering, however, is not the only possible way to relate actionsin our case: national actions that shape wage increasesto each other. A less ambitious and therefore more realistic scenario would be the mutual commitment on the part of the trade unions to accept a wage rule as a medium term guideline and to regularly evaluate the results. Such a type of wage coordination would lack mechanisms for sanctioning "flawed" wage policies. It would fully depend on the willingness of the participating partners to continuously adjust their demands in order to contribute to non-distorted real exchange rates.
The status quo with regard to such a type of transnational wage coordination in the euro area is examined in the following sections from various perspectives. More specifically, our contribution will be structured as follows: Section 2 shows that a field of action in which actors engage in transnational wage coordination has actually emerged since the 1990s at latest. In Section 3, we demonstrate how these attempts at coordination are accompanied by a research literature that takes a cautiously optimistic view on such coordination. In Section 4, we use macroeconomic data on 11 of the 12 founding eurozone countries for the years 1999-2017. These data point to the non-existence of effective wage coordination during the period analyzed.
How is this discrepancy between an established field of wage coordination and the absence of actual, effective coordination to be interpreted? In Section 5, an example from practice, IG Metall in the federation of trade unions IndustriALL, 3 is taken to explore how the key stakeholders resolve the incompatibility of the external demands placed on them. Discussing the overall implications (Section 6), we will argue that European-level wage coordination currently servesand will continue to do so for the foreseeable futureto maintain a setting where effective wage coordination could be achieved in the longer term and under far more favourable circumstances. In the short to medium term, however, the capacity for the transnational synchronization of wage increases remains modest at best.
2 European wage coordination: the emergence of a field of action Discussions among the European trade unions over the cross-border coordination of wage demands date back to the 1960s (Pen 1963, 581-85). In the 1970s, fuelled by increasing competition between European business locations, the discussions culminated in the first actual initiatives: metalworkers' trade unions decided not only to establish informal exchange but also to make some initial attempts at actual coordination (Glassner and Pochet 2011, 11). The aim of these initiatives was to prevent competition-driven wage dumping. Owing to the high level of trade union organization and focus on exports in this sector, the discussions and the initiatives predominantly centred on wage policy in the metal and electrical industry, although this did not lead to any actual formal wage rules.
Over the course of the 1990s, on the initiative of the European Metalworkers' Federation (EMF), representatives of the member unions developed what was referred to as the "coordination rule." The EMF coordination rule set down that nominal wage increases were to be at least equivalent to the sum of productivity growth and inflation. Note here that the EMF coordination rule could be interpreted in at least two possible ways, a matter which later became relevant in the monetary union. Productivity growth refers to each country's forecasted productivity gains for the national economy in the given year (at least we are not aware of any diverging interpretation). The problem is the reference to inflation. If the national inflation rate is meant the rule fails to meet its potential objective of preventing transnational imbalances. This is because, against the backdrop of heterogeneous inflation rates, employing this guideline would result in each participating country fixing its current inflation pathprecisely what must not happen if nominal exchange rate adjustments are ruled out.
More functional in terms of preventing transnational imbalances would therefore be to interpret the coordination rule with respect to the European Central Bank's (ECB) inflation target. This is precisely the interpretation that succeeded. In general, the wage rule became more and more interpreted in the light of its implications for real exchange rate distortions within the EMU. According to a position paper adopted by IndustriALL's Executive Committee in 2012, one of the explicit objectives of wage coordination is the prevention of "imbalances in foreign trade." In a subsection on "European Collective Bargaining Policy" published in its European Memorandum in 2009, the German metalworkers' union IG Metall refers to the distinctive requirements in single currency areas (IG Metall 2009, Section 4.2). Since the 2013 collective bargaining round, the IG Metall has used the ECB's target inflation rate as justification for its wage demands.
The adoption of the coordination rule by the EMF marked a turning point. While, up until that point, collaboration on collective bargaining at the European level had primarily fallen under the remit of the international secretaries of the participating trade unions, now a "collective bargaining committee" took on the responsibility. The committee comprised representatives of metalworking unions from Germany, Austria, Sweden and Belgium, as well as one EMF employee. At a meeting in Frankfurt in 1996, the committee made a decision that bound all participating organizations to the implementation of the EMF coordination rule in their local collective wage.
In order to facilitate the coordination of wage bargaining in accordance with this rule, the European metalworkers' unions set up a database that provided an overview of the collective bargaining situation in the various member states (the Eucob@n-System). Apart from providing information on the specifics of the national systems of wage determination, the primary purpose of the database was to document the results of the local wage rounds (Schulten 2004, 292). In addition to this form of record keeping, the collective bargaining committee's resolution also provided for reciprocal visits by national delegates to the wage negotiations of their sister organizations in other European countries as well as for major European-level conferences on wage policy. As we see, much effort went into the wage coordination attempts. But the approach adopted wasand could only beone based on "soft coordination."

Trade union wage coordination as portrayed in the research
There are good reasons for and against optimism with regard to the capacity of the European coordination attempts. In terms of the possible strategic direction for the trade unions, two distinct options are now evident. First, in order to guarantee job security for their members, the unions could attempt to safeguard the competitive position of their export sectors through relative nominal wage restraint, causing real devaluations that can no longer be subsequently corrected through exchange rate adjustments. From a comparative capitalism point of view, the incentives for such action should be larger the more export-oriented the respective economy is. Secondand in line with what are now conceivable coordination initiatives -, the unions could strive towards wage development coordination that precludes real competitive devaluations.
Large parts of the literature nevertheless have an optimistic undertone and tend to answer the question about an emerging effective transnational coordination with a cautious "yes." This optimism is particularly evident in the research during the early days of wage coordination. Hoffmann and Hoffmann argue that the European trade unions are "not far from a convergent and coordinated wage strategy and a coordination of national negotiations, which would be a necessary and appropriate answer to the processes of Europeanization and globalization" (Hoffmann and Hoffmann 1997, 25). Similarly, Weinert (2001, 333-36) points, even after the establishment of the currency union, to the "gradual elimination of diplomacy" (our translation) from the procedure which, until that point, had lacked binding commitments. Broadly speaking, the relevant literature identifies two driving forces behind this development, the acknowledgement that wage internationalization is a functional requirement and the presence of a certain degree of solidarity as a shared value.
Papers with mainly conceptual focuses prefer the sectoral level as the stage for wage coordination. Here, according to Traxler (2000), for instance, "nascent 'soft' networklike institutions" could, under certain conditions, be sufficient for effective coordination. He goes on to say that wage coordination would thus be possible "on the basis of non-hierarchical, networked coordination mechanisms" (ibid., 106; our translations). One argument in favour of the sectoral level is the fact that the sectoral interests of trade unions are easier to reconcile. Contributions from this area of research point towards the importance of cultural commonalities and shared values within sectors. In their secondary analysis of all wage policy coordination initiatives in the national and European industrial and metalworkers' unions, Glassner and Pernicka (2014, 1-18), for example, emphasize the presence of cultural-cognitive institutions which impart shared values within the framework of international co-operation (through exchange with other delegates at wage policy conferences, for instance).
If we continue to follow the relevant research literature over time, a decline in the aforementioned optimism can be observed. 4 First, the breakthroughs that had been predicted or at least hoped for as a result of the creation of the euro failed to materialize. Second, the euro crisis in the years after 2009 illustrated the weight of the problem triggered by the absence of effective coordination (Erne 2008). 5 Here, Henning (2013, 163) maintains that the EMF's coordination initiatives pursue a soft coordination goal only and that the level of commitment to wage coordination is derived from a "voluntary moral obligation" (ibid., our translation); also, he maintains that the strategies explicitly focus on the long-term future (we will return to this in more detail later).
While, against this backdrop, the optimists initially repeatedly voiced teleological assessments - Glassner and Pochet (2011, 8), for example, refer to "an emerging (but still very fragmented) EU framework of industrial relations"these were also recently relativized: "When it comes to assess theseindeed patchy and polycentricprocesses of horizontal Europeanization for their potential to contribute to the emergence of a European field of wage policy, there is doubt that transnational pattern bargaining based on common wage norms will spill over to other regions and sectors." (Glassner and Pernicka 2014, 14) Overall, current research shows a tendency, declining over time but by and large still distinct, to acknowledge the existence of wage coordination or at least to recognize that coordination as nascent.

The heterogeneity of wage inflations in the euro area
As we have seen, there is no room for doubt that a field of action for trade unions' transnational wage coordination has existed at the very least since the preparatory phase for the euro. However, we can only determine whether effective wage coordination has actually taken place within this field by comparing labor cost increases in the eurozone. This is something we will turn to now. Table 1 shows the development of nominal unit labor costs for 11 of the 12 founding members of the euro (not including Luxembourg) for two periods of time: first, for the first ten years of the euro between 1999 (the year the exchange rates were irrevocably fixed) and 2008 (the last year before the outbreak of the financial and euro crisis) and second, the nine years 2009-2017 (since the crisis set in). To begin with, we will concentrate on the first three columns after the country names (columns [2][3][4]. What target values can be used for evaluation? If we take into account that nominal wage increases are neutral in terms of their impact on the price level, provided they are consistent with productivity development (= constant unit labor costs), and if we also take into consideration the ECB's target inflation rate of two percent, the "golden" wage rule for the first ten euro years implies a target of 21.9 percent (column 2). Columns 3 and 4 illustrate that on average, the wage rule seems to have worked pretty well. The average nominal increase in unit labor costs of 24.1 percent (column 3) is just 2.2 percentage points (column 4) away from the target value.
This mean, however, hides substantial variance (a standard deviation of ± 14.4 percentage points). Only in a very small number of countries -Belgium, France and the Netherlandswere the unit labor cost increases roughly in line with the specifications set forth in the wage rule. In another group of countries, wage increases were substantially higher than the target. This group includes Greece, Spain and Ireland and, to a lesser extent, Italy and Portugal. In Germany, Austria and Finland, in contrast, there was wage restraint. The number for Germany, indicating even a decline in nominal unit labor cost, is particularly noteworthy. If there were attempts to effectively synchronize wage inflations across the participating countries during the first decade of the euro, then these have quite clearly failed. The crisis made meeting the demands placed on wage policy even more of a challenge. It was no longer enough to adopt the simple "golden" wage rule which stipulated that unit labor costs had to increase by an annual two percent in all the participating countries. Instead, the objective now also had to be to correct the exchange rate distortions that had cumulated as a result of the diverging wage increases during the first decade of the euro. Column 5 provides an overview of the target values. The data are derived from the ECB's target inflation rate, cumulated over nine years (adding up to 19.5 percent), and factor in the distortions during the first ten years of the euro. In the case of Greece, for instance, this "extended" wage rule would have required roughly constant nominal unit labor cost increases (increases of 3.9 percent only) because Greece had to correct much overshooting wages from the previous period. In the case of Germany, it is the other way around: the country's nominal unit labor costs would have had to increase by a considerable 43.2 percent in the years concerned because of its previous exceptional undershooting.
But what actually happened? In contrast to the first decade of the euro, on average, all participating countries failed to meet the wage rule targets (columns 6 and 7). Rather than the stipulated average increase of 17.3 percent, nominal unit labor costs only increased by three percent in the group of countries shown, which implies that wage policy sent deflationary impulses. What is even more remarkable is that the standard deviation to the mean is no less pronounced than the differences in wage policy outcomes in the first decade of the euro: the country data display an average 16.5-percentage-point variance to the three percent average increase rate. A clearly identifiable pattern emerges, where all the crisis countries pursued policies of disinflationoften under pressure from the European troika, of courseto an even greater extent than was actually required.
Yet the countries that had pursued a policy of wage restraint in the previous period did not assist the crisis countries by means of inflation-increasing wage policies. This particularly applies to Germany and Austria which were only able to produce effects that approximately corresponded with the targets of the "simple" wage rule (18.9 percent and 18.1 percent, respectively). 6 In stark contrast to how the problem is conveyed in Germany, the real exchange rates within Europe are not ailing ore due to the misdemeanours of the southern European countries participating in the euro anymore. These countries have effectively corrected their previous overshooting. Today, the inner-European real exchange rates are distorted due to the behaviour of the countries in the former "DM zone" with Germany at the centre.

A look at the actual practice of wage coordination
The findings outlined in this contribution so far have left us with a puzzle in the literal senseit is unclear how the individual pieces should be put together to make a coherent whole. Evidently there is no transnational wage coordination in the eurozone, yet, despite the constant failure to achieve actual coordination, there is nevertheless still a field of action where efforts are made to coordinate wage demands. How can we reconcile these two findings? In the following section, we will open the "black box" of wage coordination initiatives, showing by way of a particularly prominent example how practitioners deal with the failure to achieve actual effective coordination. This example focuses on Germany's key players in national collective bargaining IG Metall and the European trade union federation IndustriALL. In Section 4 we saw that in no other founding member of the eurozone has a greater need for adjustment in terms of meeting the wage rule targets developed than in Germany. 7 At the same time, the coordination initiatives in the sectors of industry brought together by IndustriALL have come farther than any other. Our next steps should therefore be particularly helpful in shedding light on the weird-looking coordination-field-without-actualcoordination constellation.
To begin with, the structural parameters of the German system of wage determination within which the key wage coordination actors operate will be looked at below (Section 5.1). The quotes included in Section 5.2 originate from a research project on the development of common policy positions in European trade union organizations. The data is comprised of a total of 20 interviews with the representatives of various national trade union organizations as well as European umbrella organizations. 8

Structural perspectives
In the comparative capitalism literature, the German economy (particularly its industrial sector) is described as a paradigmatic case of nationally coordinated wage determination. This is another way of saying that wage developments in Germany can be steered to a comparatively high degree. If Germany were to noticeably deviate from wage development in the eurozone, one might therefore be tempted to rashly ascribe this to a lack of will rather than ability on the part of the wage leader IG Metall to meet wage targets.
Following this perspective, the story would basically have the following content: The 2.3 million members of IG Metall that come from the iron and steel, skilled crafts and trades, wood and plastics, metalworking and electrical, and textile industries are exposed to international competition to a particularly significant degree, a fact which also has an impact on IG Metall's representative work. Rüb (2009, 144), for example, describes IG Metall's political orientation as tending towards a reformist economic democracy which the organization pursues by safeguarding competitiveness and productivity gains. The related sectoral corporatism therefore complicates adoption of a formal wage rule. Thelen (2014, 36) summarizes this as follows: "Rather than conflict, we observe an intensification of cooperation between labor and capital in the interest of what remains a very competitive export sector." To say that labor relations in the German export sector are characterized by "competitive corporatism" is an undoubtedly accurate description. It would nevertheless be a mistake to infer that wage restraint in a constellation like this can be solely ascribed to will, which logically implies that the actors could behave differently. The corporatism literature on the "German model" of labor relations evolved between the 1970s and 1990s, in an environment where the capacity to act in the area of wage policy was always discussed with reference to the capacity to pursue wage restraint, in order to provide fiscal and monetary policy with a stability-compatible room for manoeuvre. 9 The fact that this worked out well in Germany does not imply, however, that they would also be in a position to do the opposite, i.e. to allow wage policy to trigger inflation to correct the exchange rate distortions that had emerged in the euro area.
As an absolute minimum, in our view, the following structural aspects must be taken into consideration: Although the drop in trade union membership has not had the same impact on the export sectors as the service sectors, the responsibility for export sector wage policy has nevertheless been increasingly transferred to the company level. In general, with the introduction of a new work hours and employment model in Germany in the 1980s, the use of opening clauses in industry-wide collective wage agreements in the early 1990s, and the Pforzheim agreement 10 in 2004, an increasing shift of the class conflict to the firm level took place (Lehndorff 2016, 169-196). This translates into pressure to actually use the new scope created by decentralized wage determination practices. Thus, the further away from the macroeconomic arena wage policy is applied, the more difficult it is to gear policy towards macroeconomic (let alone European) targets.
By the end of the 1990s, and further accelerated by the Hartz reforms later, we witnessed an institutional change in the relationship between the state, the economy and the society (Busemeyer and Trampusch 2013, 291-312;Hassel 2017, 360-79;Scharpf 2018a, 27-46). In terms of wage policy, this was reflected in the growth of the low wage sector, an indirect drop in wage costs resulting from the increase in precarious forms of employment such as temporary work or service contracts, as well as the structural weakening in the negotiating position of trade unions due to higher subjective costs of unemployment (Baccaro et al.,Chap. 6). Overall, therefore, it can be said that even if the German trade union leaders were to turn their back on any form of strategic wage restraint and were crystal clear in their commitment to using wage policy to achieve inflation in order to eliminate the exchange rate distortions that had accumulated in the eurozone over time (the "will"), they would still come up against unfavourable organizational conditions (the "ability").

The practitioner's perspective
What is now remarkable is that neither Germany's competitive wage restraint nor the unfavourable structural conditions for a far more expansive wage policy resulted in any reduction in actual commitment to transnational wage coordination. As IG Metall states in its European Memorandum on this topic, "European mainstreaming" (IG Metall 2009, section 5) aims to establish a European awareness among union members that "must become second nature across all levels of the organisation" (ibid., our translation). Here, and this is supported by a number of publications (IG Metall 2005, the European wage coordination initiative is the main indicator of IG Metall's commitment to the European dimension. This agenda is also shared by IndustriALL Europe. In terms of wage policy, the organization has formulated a series of objectives including compliance with the pay formula, an increase in coordination initiatives, and an improvement in practices used to monitor the effectiveness of coordination initiatives (IndustriALL 2012). The fact that this umbrella organization has no mechanisms in place for hard sanctions limits it to the role of moderator who can provide the member unions with a forum to discuss their national wage demands. Despite this non-binding character, in almost all interviews, the trade union collective bargaining officers emphasize the practical relevance of the coordination rule. According to one interviewee, this common frame of reference is "not a guideline" but definitely "a regulation" (our translation).
In light of the failure of this rule, IG Metall has found itself subject to heavy criticism. Deppe (2012, 94), for instance, views the German trade unions as playing the "role of a 'junior partner' in Germany's policy of hegemony in the EU" (our translation). IG Metall's international sister unions have also accused it of intentional wage restraint. They describe IG Metall's restrained wage demands as a "strategy of choice." Such allegations are firmly denied by the representatives of the regional departments for collective bargaining as well as by IG Metall's head office in Frankfurt.
All the IG Metall collective bargaining officers we interviewed attribute Germany's exceptional wage development since the adoption of the euro to the general conditions outlined in Section 5.1. Against this backdrop, the collective bargaining representatives from IG Metall's head office are extremely pessimistic about the prospects of success of effective wage coordination in the narrower sense of the word: "The problem here is simply, well one of the fundamental problems with this wage coordination is: how on earth is it supposed to work? We have no formula for this and, at this point, we have to point out that we have attempted to do this in the last few years" (our translation).
According to the collective bargaining officers we interviewed, the difficulties that stand in the way of the successful implementation put the applicability of a wage rule into perspective. It is virtually impossible to communicate the need to comply with the rule: "So you cannot stand there during a token strike and announce this is something we are doing because our European friends need it" (our translation).
His colleague from another area of collective bargaining describes the situation in similar terms: "Europe is not as visible as you might think from the brochures or you might imagine in reality. As I said, there is no one sitting in a back room with their European computer open typing things in and saying what is allowed and what isn't. It's more the case that we say this is something we have to justify, this is something we have to include. But in my day-to-day work, it's not as though I have a wage round 2015 folder with a section on Europe" (our translation).
From the statements by the IG Metall representatives interviewed we can conclude that there is an unresolved contradiction between the nominal commitment and the perception of actual practicability of wage coordination.
If we now extend our analysis to beyond the pure percentages of wage increases, we can gain more perspective on the effectivity of the coordination attempts. Here, it is striking that IG Metall's wage demands up to the 2013 wage round were not based on the ECB's target inflation but rather on Germany's lower than average inflation rate (see Section 2). Furthermore, collective bargaining officers from the sister unions abroad were not seconded to the collective wage negotiations, something which was a core component of the coordination initiative. In a similar vein, IndustriALL (2014, 10) reported that the provision of data by the national member organizations was unreliable.
This discrepancy between nominal commitment to wage coordination and actual implementation in practice is put into perspective in one of the interviews with a collective bargaining officer, who underlines the largely symbolic character of coordination initiatives. In his view, wage coordination does not mean actual synchronization of wages (or at least wage demands), but rather that policy agendas call for that harmonization. Another respondent describes the situation as follows: "If all I do is base everything on practice and say that the all-important common idea is of no interest to me, that won't work either. But this very conflict is what I think makes it so remarkable, and it's also why I am now defying practice a little, too. Or perhaps I won't. Perhaps I will try to do better in future. But I can't just stay on one side or the other. I have to consider the bigger picture" (our translation).
An IndustriALL representative, in response to the question of whether, given the immediate problem, the largely symbolic coordination seemed frustrating, stated that to his mind the representation of interests meant providing continual support in the long-term pursuit of objectives: "You want early retirement at 58 in your sector. And you ask it from the employers. And they don´t give it to you. In two years´time, we come back and we want early retirement with 58. And if we don´t succeed, we come back and want early retirement at 58. Until finally, we succeed. It is the objective in collective bargaining." He describes this approach as "working towards the future and guiding it": current wage coordination initiatives serve to gradually establish a usable European platform for negotiation in the indeterminate futureand not for the coordination of wages in the present. The objective is to validate your own capacity to act through appropriate sustained efforts. In this sense, the failure to achieve coordination constitutes a problem when measured against our own expectations but not one that needs solving immediately.
6 Conclusion: fixed exchange rates without effective wage coordination The starting point of our analysis was that the functional need for effective transnational wage coordination became more pronounced when most inner-European nominal exchange rate became irrevocably fixed in 1999 (Section 1). In accordance with that, the past 25 years and especially the years after the introduction of the euro have led to the development of a transnational field of action where the key protagonists work on the coordination of wage demands (Sections 2 and 3). But European-level coordination attempts are one thing, the will and ability of implementation another. The data presented in Section 4 have left no room for doubt that effective wage coordination in the heterogeneous eurozone has not emerged.
This outcome makes sense if we consider the structural environment in which the coordination attempts operate. We have taken the failed wage coordination in the metal and electrical sector (in which the coordination field is more developed than elsewhere) and the German behaviour within it (a case of extraordinary undershooting) as an example (Section 5.1). 11 Here, adverse conditions include the decentralization of the collective bargaining system and the increasingly conceivable risk of relocations, both of which make it difficult to align national wage policy with macroeconomic, let alone European targets. In general, effective wage coordination requires an extent of steering capacity that is hardly given. The participating actors are, as we have seen in Section 5.2, all too aware of this.
Nevertheless, as Section 5.2 has also shown, wage coordination has remained an integral part of the philosophy in place at IG Metall and IndustriALL. From the perspective of a rational action model based on the concept of the targeted pursuit of manifest interests as a stable mode of conduct, the juxtaposition of the dramatic failure of wage coordination and continuous commitment seems absurd. Why would rational actors behave in this way, and, for over 15 years at that? A cynical interpretation of this constellation would argue that European wage coordination is pure lip service without practical content. But this would not do justice to the incompatible external demands and constraints placed on the key players in the field of wage coordination. A fair assessment of the status quo in wage coordination would have to duly acknowledge the genuine forward-looking character of political mobilization.
Let us take into account the tendency organizations have to meet contradictory external expectations by de-coupling talk from action (Brunsson 1989, 174). What might superficially appear to be systematic deception is often actually a (temporary) interim outcome made up of different, interwoven lines of action. From a perspective informed by Pragmatist action theory, creative problem-solving takes place by actors (further) developing purposes and goals in the context of examining specific circumstances (Beckert 2016, 64) or when, as Festenstein remarks, those actors "define the specific problem that the situation presents and . . . re-establish in accordance with human purposes the provisional equilibrium which held." 12 Thus, what may have started out as a de-coupling of talk and action, can, over the course of continuous negotiation, in fact come to constitute a move towards convergence. In order to maintain their political capacity to act, the key actors reinterpret the purpose of wage coordination. Therefore, although the capacity to actually coordinate wage policies in the eurozone has remained low until today, we believe that the commitment to wage coordination among European trade unionists is more than just lip service. By sticking to their commitment, the trade unions also contribute to the "promise-oriented legitimacy" (Beckert 2019) of the EU and, as far as the Europeanlevel trade unions are concerned, to their own legitimacy.
But all this fairness shall not distract our attention from the fact that in the short to mid run, the field of action analyzed here does not provide the capacity to correct the inner-European exchange rate distortions shown in Section 4. In this sense, euro stabilization by means of European wage coordination is, as we have put forward in the title of this article, "neither existing nor emerging." Let us think through the policy implications of this insight. The most obvious of them is that this instrument will not bring about wage increases which exceed the "golden rule" in Northern Europe, especially Germany, in order to correct its previous undershooting.
In order to nevertheless correct the distorted inner-European exchange rates, reformers will therefore have to identify functional equivalents. As far as we can see, two optionstwo options onlyare available in theory. One is to give up on the idea of horizontal selfcoordination and replace it with pure hierarchical steering. The Troika interventions have shown that this is possible in principle (Schulten and Müller. 2015, 331-64). But this strategy has already reached its limits. First, it directs the entire adjustment burden to those who suffer real overvaluation, while no possible intervention can force undervalued countries to reflate. Second, the loss of internal demand due to nominal wage reductions has increased rather than decreased the problems especially of Greece. Consequently, the last rounds of reform proposals have redirected their focuses from adding more bite to the macroeconomic imbalance procedure to matters such as risk sharing, to which we will turn shortly.
The other functional equivalent is to no longer try to rebalance by means of wage coordination in the short to mid run, but to simulate price adjustment by means of import and export taxes. In 1968, the German government had raised its import tax in order to prevent a formal revaluation of the mark. Scharpf (2016, 38-40) has proposed using this instrument to rebalance the inner-European current accounts. This would mainly imply allowing the Southern European economies to introduce discriminatory import taxes. In the reform debates, however, this option is almost completely absent, for an obvious reason: discriminatory taxation is precisely what is most effectively prohibited in the common market. The reintroduction of such discriminatory measures would mean sacrificing the common market in order to save the euro, an possibility that is widely perceived as being outside the available range of options. 13 The last reform proposals have consequently concentrated on other issues: risk sharing, mainly by deepening inner-European capital markets and thereby breaking up vicious bankstate cycles, and risk absorption, mainly by introducing a European fiscal capacity of yet unclear nature, perhaps in the form of a European unemployment reinsurance scheme. 14 As things stand today, the chances of potential risk distribution measures are much better than those of risk absorption. We do not deny that better risk sharing would make the eurozone more robust. But note that its aim is to prevent macroeconomic imbalances from culminating in financial crises, rather than to prevent current account imbalances as such. Even in a more robust eurozone, the accelerated de-industrialization of countries that suffer real overvaluation is likely to persist. As long as reformers find no solution to that problem, euro exits by single countries such as Italy remain an ongoing threat. In our view, the eurozone should carefully prepare for such eventualities (Scharpf 2018b;Sinn 2014, Ch., 9). Notes 7. Compare also Deutsche Bundesbank (2018, 13-28). As Höpner and Lutter (2018, 71-96) show, the pronounced wage restraint found in Germany is further substantiated if the manufacturing industry is examined rather than macroeconomic data (the manufacturing industry is the best available proxy for the export sector). Here, nominal unit labor costs fell by five percent between 1999 and 2008. Germany is the only one of the 11 founding members of the eurozone where the nominal unit labor costs had been on a downward trend in both the manufacturing industry and the economy as a whole during the first ten euro years. 8. We are drawing on data from a research project on international trade union cooperation in the EU with regards to wage coordination, the debate around the European minimum wage and the freedom of services. With regard to wage coordination, twenty expert interviews with collective bargaining secretaries from European metal sector unions (mostly from IG Metall) and IndustriALL Europe were conducted. See the details in Seeliger (2019, 73-79). 9. On the fundamentals of this, see Scharpf (1991). 10. According to this agreement, with the consent of the bargaining parties, company management and works councils can pay wages that are lower than the agreed levels (this is referred to as a "negative deviation"). 11. A similar story could be told about wage coordination among public sectors. In the year 2000, the European Federation of Public Service Unions (EPSU) agreed on a framework for collective action similar to the one adopted by the European Metalworkers' Federation (EMF, compare Section 2), but effective wage coordination has nevertheless not occurred. What makes the public sector particularly interesting is that public wage coordination could, in principle, benefit much more from the support of the employers, that is, from public authorities. But this has clearly not materialized. We thank Donato Di Carlo for pointing our attention to this initiative. See Johnston (2016, ch., 5) and Di Carlo (2018). 12. Quoted in Ansell (2016). 13. See also Bundesministerium für Wirtschaft und Energie (2019, 15). 14. Among the numerous proposals are the so-called "Saint Nicholas" package of the European Commission (2017); a reform paper by 14 French and German economists, published in January 2018 (Cesifo 2018) and the declarations of the Eurozone summit in June 2018 (Euro Summit 2018).