Transnational Trade Union Action in Europe: The significance of national and sectoral industrial relations

In November 2012 the European Trade Union Confederation (ETUC) mobilized simultaneous strikes in four countries (Italy, Greece, Portugal, and Spain), in addition to the demonstrations and other activities it regularly mobilizes on its action days. In total, 40 trade unions from 23 countries were involved in various activities on this European Action Day, which was a protest against austerity measures. Although this was not the first time trade unions had organized contentious action on a transnational level in Europe, transnationally coordinated strikes are not as common as less contentious actions such as demonstrations or writing joint petitions or open letters. There have been other well-known strikes, such as the Renault strikes of 1997 and the recurrent conflicts at General Motors Europe, both of which were organized by their respective European Works Councils (EWCs) (Imig and Tarrow 2000; Greer and Hauptmeier 2009).

Some researchers have found that such activities, as well as cooperation between trade unions and NGOs and social movements, suggest an emerging Europeanization of a democratically based political mobilization (Bieler and Schulten 2008). Such action is often seen as important for the success of interest-group lobbying toward political institutions – not least in transnational arenas (Gajewska 2008; Hyman 2011). It may also encourage a transnational labor identity: ‘By enabling union action, unions empower otherwise isolated individuals to engage with governments, EU institutions, and firms’ (Erne 2008: 186). However, recent research into European trade unionism has discussed whether the opposite tendency, towards ‘union diplomacy’ and a ‘logic of influence’ is so strong that unions are largely disinterested in more democratically oriented and contentious action at the European level, based on a ‘logic of membership’ (Hyman 2005; Erne 2008). Union diplomacy, with its top-down logic driven by union executives and expertise, is often seen as an adjustment to the technocratic mode of governance in the EU (Bieler and Schulten 2008; Gajewska 2008; Hyman 2011). This tendency has already been inscribed in the construction of the consultation processes within the EU, specifically the macro-economic dialogue, the European Economic and Social Committee, and the national social pacts related to the open method of coordination. Union diplomacy also seems to be a central approach in the strategies of the ETUC, and even in the European Industry Federations (EIFs)¹ and the EWCs (cf. Taylor and Mathers 2002; Schroeder and Weinert 2004; Platzer and Müller 2011: 33ff.; Glassner and Vandalele 2012; Keune 2012).

Because of national competition and the tendencies towards technocratic adjustment among unions, there is a great deal of pessimism related to the prospects of a campaigning transnational labor movement on the European level (cf. Pianta and Gerbaudo 2012). Even so, the cases discussed in the introduction show that the mobilization of transnational contentious action in Europe is also connected to the institutionalized structures of the European trade union organizations, including ETUC, the EIFs, and the EWCs. In addition to more lobby-oriented activities and negotiations as social partners, these organizations also pursue campaign activities that mobilize their members in demonstrations and strikes (Gajewska 2008; Pianta and Gerbaudo 2012).

The present article analyzes the attitudes towards transnational action of more and less contentious forms among trade unions in Europe, as well as the degree to which national unions actually partake in such action, with a particular focus on the production of joint statements and petitions, demonstrations and boycotts, and strikes. However, the article does not discuss other kinds of transnational bottom-up cooperation and action between unions, such as the exchange of information and observers and coordination of negotiations (e.g., Traxler and Mermet 2003; Glassner and Pochet 2011; cf. Larsson 2012). Empirically, the present article is based on survey data collected in 2010–11 from unions affiliated to the ETUC as well as from unions below peak level in 14 European countries. The analysis aims to explain the similarities and differences between unions in different countries and regimes in transnational trade union action, but also the impact of sectoral differences. This aim is based on the theoretical assumption that differences in industrial relations regimes and traditions between countries and sectors are significant for trade unions' attitudes and strategies regarding transnational trade union action. The paper begins by discussing these empirically grounded theoretical assumptions, before presenting the data and some methodological considerations and then turning to the analyses.

As discussed by Hyman (2001), the perspectives and strategies of national trade unions differ across Europe against the background of divergent historical traditions and contexts. While these traditions have some stability and vary between the various regions of Europe (cf. Glassner and Vandaele 2012), any clear-cut distinctions between national traditions appear to exaggerate national homogeneity and undervalue variation in ideology and strategy (Hyman 2001). Nonetheless, it is common for comparative research on industrial relations in Europe to discuss the existence of a smaller set of regimes or models that cluster countries on the basis of similarities in national industrial relations (Vos 2007; Bieler and Schulten 2008: 235ff.; Visser et al. 2009: 49ff.; Vandaele 2011; Meardi 2012).

A common trait of these empirically grounded typologies is that they crudely single out the same set of geographic clusters of countries: Central Western European Countries (CWECs), Southern European Countries (SECs), Central Eastern European Countries (CEECs), Western European Countries (WECs), and the Nordic countries. As research on the subject has stated repeatedly, these typologies are simplifications of a diverse reality and detailed examination reveals many internal deviations from the models. Having said that, the typologies still seem to be important for singling out some typical differences. Accordingly, I will provide a brief overview of how the five regimes of industrial relations in Europe can be characterized. My point of departure is the clustering of countries in Visser et al. (2009: 49ff.), which appears to be the most elaborate attempt to produce such a typology.

The CWECs – such as Austria, Germany, and Belgium – are characterized by a social partnership tradition. This is based on corporatist tripartite relations between the employers, trade unions and the state, which have a relatively low level of conflict and are therefore termed a social partnership regime of industrial relations. Historically, the political influence of the unions has been quite strong, and for some unions the cooperative approach is anchored in the tradition of a religious organizational base. Union density is moderate (although quite high in Belgium) in the European context, and bargaining coverage is high. Employee representation is typically accomplished through a dual system that includes both strong sectoral bargaining and local works councils. Social partners have a high degree of autonomy in relation to the state, and collective bargaining has high coverage rates because of legal extension mechanisms (Hyman 2001; Bieler and Schulten 2008; Visser et al. 2009; Glassner and Vandaele 2012).

The SECs – such as France, Portugal, and Spain – are gathered in a polarized/state-centered regime of industrial relations. This is related to the historical tradition of hostility from employer organizations, as well as a more fragmented trade union movement divided around different political and religious orientations, with weak relations between different levels of bargaining. There is both single- and dual-channel representation, and more variation in the principal level of bargaining than in the partnership regime. In comparison to the partnership regime, SEC countries also experience more frequent state intervention into wage and working standards setting and a more politicized involvement of social partners. Union density is generally lower than in the partnership regime, and strikes are seen as an important means of political protest to influence the state to take action (Hyman 2001; Bieler and Schulten 2008; Visser et al. 2009; Vandaele 2011).

The Nordic countries – such as Denmark, Norway, and Sweden – are said to belong to a organized corporatism regime. This regime is characterized by strong unions and employer associations negotiating collective agreements with a high degree of autonomy from the state, and with wide bargaining coverage. Peak-level unions are basically organized on the basis of occupations/class, as opposed to the religious and ideological divergences in other regimes, and union representation is single-channel. Unions have had an influence on policy making and there is a tradition of social partnership. The strength of trade unions is based on a relatively high union density and the historical alliance with social democracy. The bargaining system and the strong position of unions reduces the need for major strikes and, as in a social partnership regime, the level of conflict is relatively low (Dølvik 2007; Visser et al. 2009; Jochem 2011; Larsson et al. 2012).

The WECs – with the UK as a typical example – are characterized as belonging to a liberal pluralist regime, which implies that there is less state intervention in industrial relations than there is in the SECs and CWECs. There is relatively little legally established standard setting, but also relatively low coverage of collective agreements. Union involvement in policy making is more ad-hoc in nature. Trade union organization is more fragmented than in the corporatist and partnership regimes, and bargaining is more decentralized, with company bargaining being the most important level, and union density is moderate compared to Europe overall (Hyman 2001; Bieler and Schulten 2008; Visser et al. 2009; Glassner and Vandaele 2012).

Although the CEECs are quite varied, they are often described as belonging to a fragmented/state-centered regime of industrial relations. Trade unions are weak and fragmented in most countries, with relatively low union density and low identification of workers with trade unions. There is little tradition of contentious action, or even of identification of opposing interests between employers and employees, and several countries also limit the right to strike. Bargaining is decentralized, with weak development of the sectoral level. The weakness of the social partners means that they have relatively little autonomy, and the states have been more concerned with accomplishing the transition to market economy than intervening in bargaining (Bieler and Schulten 2008; Hassel 2009; Visser et al. 2009; Vandaele 2011).

As noted, this typology is a simplification and there is a lot of internal variation between countries and between sectors within countries. As shown by Bechter et al. (2011), there is, with some exceptions (primarily the Nordic countries), more divergence in industrial relations across sectors within a country than within sectors in the EU 27, which makes it possible to speak of sectoral regimes of industrial relations in Europe. The basis for such a perspective is that different industries are characterized by different production processes and work organizations. Some industries, particularly in manufacturing, are highly exposed to international competition since their product markets are highly integrated and their work organization enables a high degree of transferability of production location. Together with industries that have undergone liberalization (such as civil aviation and telecommunications) and construction because of the Posting of Workers Directive, these are the industries that have had the greatest cause to collaborate across borders since they are affected by common industrial policies in the EU (Pulignano 2009; Müller et al. 2010; Glassner and Pochet 2011). The more ‘sheltered’ industries, such as public services like education and healthcare and private services like hairdressing, are less exposed not only to international competition in their consumer markets, but also to transferability of production location (Bechter et al. 2011; Glassner and Pochet 2011: 13). They have been under less pressure to develop cross-border cooperation and have therefore been latecomers in terms of European cooperation and sectoral dialogue (European Commission 2010: 7).

This section can be summarized by saying that transnational trade union cooperation in Europe must be understood against the background of differences in both national industrial relations and sectoral industrial relations. These factors may be expected to have an effect on both the general degree of engagement in European trade union cooperation and the more specific balance between the ‘logic of influence’ and the ‘logic of membership’. As noted by Bieler (2005: 479), ‘unions which still enjoy considerable access to decision-making at the national level, are less likely to support the establishment of a European industrial relations system’. One could also suspect that they are less willing to advocate and engage in transnational trade union action in Europe. Moreover, industrial relations regimes in which trade unions have a tradition of pursuing their influence through demonstrations and contentious action, rather than corporatist or partnership strategies, may also be expected to be more affirmative in pursuing such strategies at the European levels (Mitchell 2007; Vandaele 2011).

The data for the following analyses was collected through a web-based/postal survey that was sent out in 2010 to all ETUC member organizations, as well as to trade unions below the peak level in 14 European countries. We targeted organizations with approximately 10,000 members or more in order to exclude minor trade unions, and sent the survey to 512 unions. The overall response rate was 49%, although this was much higher in most countries (76% excluding Poland and France, for which we had great difficulty reducing the number of responding organizations and also received low response rates). In order to produce good validity, respondents at central positions in the organizations were selected. As a result, just over half of the returned questionnaires were filled out by the secretary-general, the president, or the vice president, and one-quarter by an international secretary/correspondent.

In this article, I analyze the responses to two sets of survey questions. The first set concerns attitudes towards trade union cooperation in Europe, and the other deals with whether the respondent's organization had participated in different kinds of union action within Europe during the last 3 years. The analyses are presented in the form of cross-tables with percentages and in multivariate logistic regressions in order to control for sectoral variation and the size of organization in the analyses of differences between industrial relations regimes.

With regard to the country comparison, the percentage tables separately present the seven countries for which the response rate was above 50%, as well as the EIFs. The remaining responses are presented in two categories: ‘other Western European Countries’ and ‘other Eastern European Countries’. In the regression analyses I have followed the categorization of regimes of industrial relations in Visser et al. (2009: 49ff.). The organized corporatism (Nordic) regime (n=102) consists of my analyses of Denmark (20), Finland (16), Iceland (3), Norway (24), and Sweden (39). The social partnership (CWECs) regime (n=46) consists of Austria (7), Belgium (11), Germany (16), Lichtenstein (1) Luxembourg (1), The Netherlands (1), Slovenia (1), and Switzerland (8). The polarized/state-centered (SECs) regime (n=35) consists of France (18), Italy (2), Portugal (1), and Spain (14). The liberal pluralism (WECs) regime (n=27) consists of Cyprus (3), Ireland (5), and the UK (19). Finally, the fragmented/state-centered (CEECs) regime (n=31) consists of Bulgaria (2), Croatia (1), Estonia (1), Hungary (3), Latvia (1), Lithuania (1), Poland (20), Romania (1), and the Slovak Republic (1). The nine responses from EIFs are excluded in these analyses, which results in a total of 241 cases. Some countries are over-represented, which is a result of our decision to send the questionnaire to trade unions below peak level only in 14 countries, a majority of which are often viewed as ‘typical’ countries for their respective regime.

With regard to the sectoral classifications, one would ideally use a theoretically stringent sectoral division according to the approach of Crouch (1999), for instance. However, since unions organize according to different principles (Visser 2012), it was difficult to perform such clear-cut classifications. Therefore, I have approached the question of sectoral distinctions more pragmatically, on the basis of both the classification of economic activities in NACE and the occupational classifications of ISCO-88 (COM). The analyses are based on seven categories: professions (n=63) consists of unions that mainly organize occupations at levels 1 or 2 in ISCO-88 (COM), such as lawyers, engineers, doctors and nurses, dentists, psychologists, teachers, and social workers; services (57) consists of unions that organize services occupations at lower levels of ISCO-88; manufacturing (40) includes unions that organize in industries such as chemicals, mining, and forestry, albeit with a main interest in the manufacturing industries; construction (14) is separated from the other sectors since it has been under increasing competitive pressure due to the increase of posted workers following the inclusion of new EU member states in 2004 and 2007; transportation (11) is separated from services in general, since transportation generally has a higher degree of internationalization than services; peak-level confederations (44) are the national member organizations of the ETUC; finally, cross-sectoral unions (12) is a category for those unions that could not be classified since they organize in more than one of the above sectors.

In order to provide a background for the following analyses, I will first present some data concerning attitudes towards trade union cooperation and action in Europe. I will start by relating to the discussion of whether transnational cooperation is seen as important by trade unions in Europe, even though tendencies towards trade union nationalism might have been strengthened during the recent financial and economic crises (Glassner and Vandaele 2012). The first two statements in Table 1 indicate an established consensus across Europe that trade unions should cooperate more transnationally. Only a few of the trade unions in the survey expressed the opinion that they should engage less in the European level. Although these include unions that represent all sectors, even peak level organizations, there is a slight overrepresentation of unions from the Eastern European and Nordic countries (data not shown).

The last three statements in Table 1 relate to the tension between the ‘logic of influence’ and the ‘logic of membership’ (Hyman 2005). As the table shows, European trade unions tend to recognize the importance of pursuing both a more technocratic logic of influence and a democratically oriented logic of membership, including a mobilization of trade union action at the European level (cf. Glassner and Vandaele 2012). Nevertheless, more than one-third of the unions questioned the importance of engaging in contentious cross-national action, in terms of mobilizing joint demonstrations, boycotts, overtime bans, or strikes. It is here that we see the most diverging attitudes among trade unions in Europe. This is unsurprising given that there are different traditions and strategies among unions in different parts of Europe in terms of mobilizing demonstrations and initiating conflicts at the national level (Vandaele 2011; Peterson et al. 2012b).

The national and regime differences in industrial relations are important in terms of union strategy and also union power and density (cf. Hyman 2001). For example, the power of the Nordic unions is based on their autonomy in bargaining as well as the ‘implicit use of numbers’ following from their high density, as compared to the more ‘overt displays of numbers’ through demonstrations in which non-members may partake in countries with lower union density, such as France (cf. Lindvall 2011; Peterson et al. 2012a). With regard to strike levels, even though the Nordic countries have institutionalized forms of conflict resolution in the labor market, most of them still have higher levels of strikes than many CWECs (although not higher than Spain, France, Italy, and Belgium) (Vandaele 2011: 11). In addition, general or mass strikes are most frequent in the south, which is related to the fact that strikes are used in southern Europe and France as a means of political protest and demonstration, directed at the state rather than directly at the employers. One could also note that, for the reasons discussed above, strike levels are low in the CEECs. There is a strong path dependency in these issues, as noted by Vandaele (2011: 39): ‘the rank order [of strike activity] between European countries shows remarkable stability over a 20-year period – albeit a tendency towards convergence …’.

Of course, differences in union traditions and strategies and in industrial relations at the national level may influence the attitudes towards union action at the European level. I will analyze such attitudes towards transnational action in greater detail later. However, I will begin by analyzing the actual degree of participation in such joint action in Europe, since this also provides crucial information about the impact of regime differences.

Trade unions in Europe are involved in varying types of cross-national union action. Here, I will only offer a short overview in order to illustrate the broad range of actions relating to the kinds focused on in this article.² Starting with the ETUC, they produce a great number of resolutions every year, but also occasional petitions (such as the one on public services in 2007), and open letters (such as the ones to Sarkozy and Barosso in 2008 on the financial crisis, and the one to the European Council in 2008 regarding how to help Europe recover from the depression). ETUC has also organized euro-demonstrations during the actual period, both in specific cities such as Ljubljana and Strasbourg in 2008 and Brussels in 2010, and coordinated across different cities, such as the ‘Fight the crisis’ demonstrations in Berlin, Brussels, Madrid and Prague in 2009 and the ‘European Day of Action’ that occurred all around Europe in 2010. However, the ETUC did not organize strikes before the 2012 action mentioned in the introduction. According to the ETUC, that remained a matter for national trade unions since there are no clearly defined rights to strike at the EU level. Nevertheless, Fabbrini (forthcoming) claims that the right to strike has been explicitly recognized at the EU level at least since the Viking and Laval cases, albeit with significant limitations.

The palette of union action in the EIFs varies between sectors and between sections within the EIFs. All EIFs produce statements, resolutions, or position papers, as well as some joint declarations, common demands, manifestos, and open letters and petitions. Other statements include responses to the EU commission, calls for solidarity action or letters of solidarity with national unions, and ‘lobbying tools’ for affiliates. Some EIFs, or sections thereof, organized action days during the period; examples include The European Transport Workers' Federation (ETF), The European Metal Worker's Federation (EMF) (now a part of IndustriAll), and The European Federation of Public Services Unions (EPSU).

Organizations other than the ETUC and the EIFs may also pursue transnational trade union action. Other trade union-based organizations also exist, such as the Eurocadres, as well as regional and bilateral cooperation networks such as the Inter Regional Trade Union Councils (IRTUCs), the cross-border network called the Doorn group, the Council of Nordic Trade Unions (NFS), and the Baltic Sea Trade Union Network (BASTUN), all of which produce statements but, as far as I can tell, do not organize demonstrations or strikes (although some of them have participated in ETUC-organized actions). Transnational trade union action has also been organized through European Works Councils (EWCs), the best known case of which was the Renault strikes of 1997 (Imig and Tarrow 2000; Mitchell 2007).

Table 2 shows the overall involvement in union action in Europe according to the union representatives in our survey. These activities include some that are not specifically related to the ‘logic of membership’, such as writing statements, petitions or open letters, and more contentious action that builds upon the mobilization of members, such as demonstrations, boycotts, and strikes. As the table shows, 80% of the unions surveyed have participated in writing statements, petitions or open letters, and over half of them stated that they have participated in demonstrations or boycotts. The largest part of these participations is related to actions organized by the ETUC and the EIFs, although some have other organizers. These are somewhat crude measures, not least since ‘partaking’ in action such as demonstrations can cover everything from organizing a seminar to mobilizing great masses on the streets (Mitchell 2007), however, they provide at least a starting point for analyzing differences in the involvement in union action in Europe.

Almost one-fifth of the respondents stated that they have participated in overtime bans, strikes or blockades, or sympathy strikes. This is somewhat surprising given the unclear rights to such action at the EU level, and the figures may be exaggerated because of the vagueness of the survey question, which could also be interpreted as including national trade union action. However, many of these actions are said to have been organized by the ETUC or the EIFs, which suggests that they are not national activities. In fact, it may be difficult to delimit some union action as clearly belonging to the national or transnational level, so it may be interpreted as transnationally coordinated action (cf. Imig and Tarrow 2000). For example, Hammer (2010: 361) discussed a solidarity action in Hungary relating to an Austrian strike in which an IRTUC network was used. The EMF mentions a connection between youth members and a flash mob action in Hungary (EMF 2011: 47). The best known cases bordering between national and transnational action are the Viking and Laval cases, which concerned transnationally oriented strikes, blockades, and sympathy strikes (cf. Fabbrini forthcoming).

These results may be given more stability by comparing them with a previous study of the Nordic unions (Nergaard and Dølvik 2005). In that study, the most common kind of transnational sympathy action was signing a petition or similar (69%), while it was less common to participate in manifestations (35%) or boycotts/blockades (21%). There are three possible reasons for the slightly lower figures in the first two questions (compared to the present survey). Firstly, it could be an effect of Nergaard and Dølviks' survey being older and the fact that the level of action has increased since then. Secondly, it could be because Nergaard and Dølvik included smaller trade unions than we did, and they participated to a lower degree than unions. Thirdly, the lower figures could be related to a difference in the level of participation between the Nordic countries and the rest of Europe. This last possibility is examined in the next section.

Table 3 provides an overview of participation in union action in the seven countries in which we received a response rate of above 50%. In addition, the EIFs are presented separately and the remaining responses are presented in the two categories of ‘Other Western European Countries’ and ‘Other Eastern European Countries’. Some tendencies can be discerned from the results. The first is that the EIFs have a high degree of participation in activities, as have the continental European countries of Austria, Germany, and Spain. There is a particularly high degree of participation in overtime bans and strikes in Germany and ‘Other WECs’. Much of this action occurs in the manufacturing and services sectors, and may be related to EWC company-level action, but it is not possible to know this for certain based on the data. Secondly, the Nordic countries score quite low on the two more contentious classes of action. Thirdly, the Eastern European trade unions score low in terms of participating in producing joint statements, petitions, and open letters.

These analyses indicate that regime differences may have an effect on participation in European trade union action. In addition, it can be presumed from the above discussion that there are sectoral differences and that the size of the organization does matter. The analysis in Table 4 is an attempt to test the effect that these factors have on the item in which the greatest differences between countries were found – that is, participation in demonstrations and boycotts. The main result is that industrial relations regimes have a strong and significant effect on the participation in European demonstrations and boycotts among trade unions, even when controlled for sector and size of organization. The highest degrees of participation can be found in continental and Southern Europe. The Eastern European trade unions participate to a somewhat lower degree, but still to a greater extent than unions in the Liberal pluralism and particularly the Nordic countries, which had the lowest level of participation in European demonstrations and boycotts.

Even though industrial relations regimes differences are important for the variation in the degree to which a trade union participates in European demonstrations and boycotts, the sector and size of the organization are also significant. It is clear from Table 4 that cross-sectoral and peak level unions tend to participate to a higher degree than most of the more narrowly organizing sectoral or industrial unions. This may seem natural, as they represent broader sets of members and partake in activities in more than one sector. However, the strongest sectoral effect on participation is found in the transportation sector (cf. Sweeney et al. 2011: 52).

Another finding that is worth mentioning, despite not being statistically significant, is the tendency for unions organizing in the services sector and the professions to participate less than the other sectors. Finally, the size of the unions is also important for participation in European demonstrations and boycotts, at least in the sense that the smaller unions tend to participate to a lower degree than the larger ones.

Similar analyses to that in Table 4 were performed on the two other items in Table 3. The analysis of participation in overtime bans and strikes produced results similar to Table 4, albeit with somewhat weaker tendencies (data not shown). A specific difference was that the transportation sector was closer to the others in terms of odds-ratios. The analysis of participation in writing joint statements, petitions, and open letters produced less valuable results. The only significant effect was that the smaller unions tend to participate to a lower degree than the larger ones (data not shown).

With the above analysis of the participation in trade union action in mind, it is time to return to the attitudes toward such participation, as this may provide clues about future strategies. Table 5 provides a detailed view of the attitudes towards participation in transnational demonstrations, boycotts, overtime bans, or strikes. Again the seven countries in which we received a response rate above 50% are presented separately, as are the EIFs. As Table 5 shows, most EIFs have a strongly affirmative attitude towards transnational action. With regard to the national trade unions, the great divide is between the central and south-western European countries, on one hand, which are quite positive to such action, and the more skeptical Nordic countries on the other – with the UK somewhere in between.

As Table 6 shows, the effect of industrial relations regimes holds even when controlling for sectoral divisions and size of organization. The analyses show that trade unions in the continental social partnership regime are the ones most affirmative of increased engagement in transnational trade union action, followed by the SECs and the WECs. The unions in the CEECs seem to be almost as skeptical as their Nordic counterparts.

With regard to sectoral differences, Table 6 confirms that there is relatively little interest in such action from unions organizing in services or professional workers compared to other sectors, particularly manufacturing. We can also see that, in terms of attitudes, transportation does not really differ from other sectors under competitive pressure. Finally, the size of the organization seems to be of little importance for the attitudes towards transnational trade union action, unlike its importance for the actual participation in such action.

This paper has analyzed the attitudes towards transnational trade union action in Europe, as well as the degree to which national trade unions actually partake in such action. I have focused particularly on action in terms of the production of joint statements or petitions, demonstrations and boycotts, and overtime bans and strikes. My aim was to show the similarities and differences between different countries and industrial relations regimes in such activities, but also the degree to which sectoral differences matter. The first conclusion touches upon the discussion of whether the tendency towards ‘union diplomacy’ is so strong that there is little interest in more democratically oriented union action at the European level (Erne 2008). In accordance with the conclusions drawn from Glassner and Vandaele (2012), my results suggest that European trade unions tend to recognize the importance of pursuing both a more technocratic ‘logic of influence’ and a democratically oriented ‘logic of membership’, including a mobilization of contentious trade union action at the European level (cf. Hyman 2005). This conclusion needs to be elaborated in relation to another important result, however, since there are diverging attitudes towards and engagement in member mobilizing trade union action among trade unions in Europe.

The second main conclusion is that differences between countries and industrial regimes in Europe are of great importance, both for the level of participation in European trade union action, such as demonstrations and boycotts, overtime bans and strikes, and for the attitudes towards such action. Firstly, one may note that EIFs are more affirmative towards and do participate to a higher degree than national trade unions. With regard to the national trade unions, my results confirm the more general results in Glassner and Vandaele (2012) that the Nordic countries in the organized corporatism regime are the ones least engaged in and most skeptical of contentious action. A reasonable cause for this is that these countries pursue more of national strategies of influence based on their high membership rates, and that their traditional bargaining autonomy and corporatist influence on policies still has an effect on their approach to European cooperation (cf. Larsson et al. 2012). At the other extreme are the Central Western European trade unions in the (continental) social partnership regime, and the Southern European trade unions in the polarized/state centered regime. These unions are approving of strengthening such trade union action, and are also those that have already participated to the highest levels. This can be understood against the background that trade unions in these countries generally have been positive towards developing trade union cooperation on the European level. In the case of the SECs, there is also the tradition of using strikes as a means of political protest to influence the state to take action (Hyman 2001; Bieler and Schulten 2008; Visser et al. 2009; Vandaele 2011).

Between these extremes, we find trade unions in the liberal pluralist regime and the Eastern European trade unions of the fragmented/state centered regime. Trade unions in the UK have previously been skeptical about developing a strongly coordinated European cooperation, and even though they have recently recognized the benefits of cooperating at the European level, they seem to be more oriented towards a ‘logic of influence’ than campaigning (Mitchell 2007). As discussed above, unions in the CEECs have less of a tradition of contentious action, but also less identification of workers with unions and therefore less potential for campaigning (Bieler and Schulten 2008; Hassel 2009; Visser et al. 2009; Vandaele 2011). This could partly explain their position on the issues and activities analyzed in this paper.

A third conclusion is that sectoral differences also affect the level of participation in European trade union action, such as demonstrations and boycotts, overtime bans, and strikes, and towards strengthening cooperation at the European level in terms of such action. Unions organizing professionals and in the services sector are the least engaged in and most skeptical of such action. National peak level unions and other cross-sectoral national unions participate to a higher degree than other unions below peak level, with transportation as an exception in terms of reported participation. As discussed above, the need for transnational trade union cooperation has been smaller in services and unions organizing professionals than in manufacturing or in transportation and construction, due to the competitive pressure in these sectors. In addition, one can note that the size of the organization does matter; if not in terms of attitudes then at least regarding the actual participation in union action. In other words, the smaller unions participate to a lower degree than the large ones. This implies that even if power and resources is not important for the attitudes towards transnational trade union action, it is of importance for actually partaking in it.

Another important set of factors that these analyses have not controlled for is the divergent impact in Europe of the recent financial and economic crises. Therefore, there is a need for some reservation regarding the overall interpretation of these results, since some of the regime differences found in the analyses might be caused by economic and contextual factors that are not directly related to differences in regimes of industrial relations. Even if such factors are important for explaining national levels of trade union action, it seems improbable that they would be the major explanation for the willingness of trade unions to participate in transnational action.

To conclude, Bechter et al. (2011) may be correct in their suggestion that there is a convergence in sectors within Europe, which lays a base for analyses of ‘sectoral regimes’, while Hyman (2001) has a point in arguing that the clear-cut divisions between national traditions and strategies among trade unions are dissolving. Even so, the results in the present paper show that national differences still play an important role in the strategies and activities that national trade unions have at the European level. At least when looking at their engagement in and attitudes toward pursuing mobilization of members in transnational trade union action, it seems that Mitchell (2007) is correct that trade unions tend to replicate to their traditional national strategies when approaching the European level.

The research on which this article is based was funded by the Swedish Council for Working Life and Social Research. The author would like to thank Mattias Bengtsson, Bengt Furåker, Kristina Lovén Seldén, and the anonymous reviewer for helpful comments on previous versions. The author is grateful also to Maria Jepsen, Vera Glassner, and Kurt Vandaele at the European Trade Union Institute (ETUI) as well as to Roland Janssen at the ETUC for constructive comments on the questionnaire.

Bengt Larsson is Professor of Sociology at the Department of Sociology and Work Science, University of Gothenburg, Box 720 SE 405 30 Göteborg, Sweden. E-mail: [email protected]. His research interests are in the fields of economic sociology and welfare: auditing, trust, economic crime, consumption and over-indebtedness, and trade union cooperation in Europe.