Varieties of trade union protest

Höltmann, Gesine; Borbáth, Endre; Hutter, Swen Article — Published Version Varieties of trade union protest Socio-Economic Review Provided in Cooperation with: WZB Berlin Social Science Center Suggested Citation: Höltmann, Gesine; Borbáth, Endre; Hutter, Swen (2025) : Varieties of trade union protest, Socio-Economic Review, ISSN 1475-147X, Oxford University Press, Oxford, Vol. 23, Iss. 1, pp. 75-96, https://doi.org/10.1093/ser/mwae056 This Version is available at: https://hdl.handle.net/10419/310916 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Article Varieties of trade union protest Gesine H€ oltmann 1, �, Endre Borb� ath 2 and Swen Hutter 1 1 Freie Universit€ at Berlin & WZB Berlin Social Science Center, Berlin, Germany and 2 RuprechtKarls-Universit€ at Heidelberg & WZB Berlin Social Science Center, Berlin, Germany �Correspondence: [email protected] Abstract The article examines varieties of trade union protest across industrial relations regimes, using protest event data for 27 European countries between 2000 and 2021. We present a large-n analysis of how the level and ‘movement character’ of union protest covaries with the strength and institutional settings of union movements across regimes. We show that unions remain important protest actors and that union protest in the public sphere notably outweighs workplace-related strikes. Furthermore, we find an inverse relationship between union institutionalization and the ‘movement character’ of union protest: While strong union movements in highly institutionalized regimes display a strike-heavy repertoire, weaker union movements in contexts of low institutionalization rely heavily on protest actions beyond the workplace strike. With these findings, we provide a novel empirical assessment of what unions do in the protest arena and how institutional settings can be conducive to strike-heavy versus protest-heavy union tactics. Key words: trade unions, social movements, industrial relations, strikes, civil society, Europe 1. Introduction Trade unions have largely departed from their social movement origins and have become insiders throughout the 20th century: They have been recognized as negotiating partners by both government and business, granted rights and privileges in wage bargaining and gained influence in social policymaking (Streeck and Hassel, 2003; Hyman, 2015). However, this development has not progressed to the same extent across European democracies. Hyman (2001, p. 2) identifies three trade union ideologies ranging between the poles of market, class and society: (a) ‘Business unionism’, where unions act primarily as interest groups and are mainly concerned with labour market issues and collective bargaining; (b) unions as ‘schools of war’ and agents of class conflict, where protest via extra-institutional channels is the primary mode of action; and lastly (c) ‘social partnership’, where unions act as economically stabilizing forces in society and advocate for social justice more generally. While none © The Author(s) 2024. Published by Oxford University Press and the Society for the Advancement of Socio-Economics. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https:// creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. Socio-Economic Review, 2025, Vol. 23, No. 1, 75–96 https://doi.org/10.1093/ser/mwae056 Advance Access Publication Date: 24 September 2024 Article of these ideologies exist today in their ideal-typical form, trade unions across Europe can be found on this spectrum, balancing negotiation and militancy, as well as workers’ interests and broader societal issues. This balancing act between interest group and social movement actor in trade unions’ repertoires has been neglected by the social movement and industrial relations literatures. Industrial relations scholars place a strong focus on declining strike levels and largely disregard unions’ protest repertoire beyond the workplace strike (Kelly, 2015a). While individual case studies have begun to examine whether other forms of union protest might be displacing workplace strikes in Sweden (Jansson and Uba, 2023) and in the UK (Gall and Kirk, 2018), Jansson and Uba (2023) note that ‘the literature lacks a systematic investigation of unions’ broader protest repertoire’ (p. 256). The social movement literature, on the other hand, has mainly focused on union activities beyond the classical strike, including literature on social movement unionism as a strategy of union revitalization (Grote and Wagemann, 2018; Chesta et al., 2019; Pilati and Perra, 2020, 2022; Portos and Della Porta, 2020), case studies on union movements in individual countries (Chesta et al., 2019; Pilati and Perra, 2020, 2022) and in particular moments in time, such as the austerity protests in the wake of the Great Recession (Portos and Della Porta, 2020). In this article, we want to shed a new light on unions as protest actors by uniting the industrial relations and social movement perspectives: We ask how union movements balance workplace strikes and protest beyond the workplace, and how this varies with the strength of union movements and institutional settings across industrial relations regimes. We believe that by jointly considering workplace strikes and protest beyond the workplace as integral parts of the union protest repertoire, we can learn under what circumstances unions primarily target employers and challenge working conditions (workplace strikes), and when they challenge the political and public realm (protest beyond the workplace). This adds a new dimension to the debate on declining strike rates by empirically contrasting strikes with all other union activities in the protest arena (e.g. Gall and Kirk, 2018; Jansson and Uba, 2023). By showing how the structural and historical setting of the industrial relations regimes affects unions’ repertoires in positions of strength versus weakness, we contribute to ongoing debates on when and under what circumstances unions mobilize more generally (Korpi and Shalev, 1979; Brandl and Traxler, 2010; Baccaro et al., 2003; Lindvall, 2013) and when they resort to movement tactics more specifically (Della Porta, 2006; Fairbrother, 2008). We set out to describe (a) how often unions engage in protest activities relative to other actors in the protest arena (level of union protest), (b) to what extent unions engage in classical strikes versus union protest beyond the workplace (what we call the movement character of union protest) and (c) how the level and movement character of union protest vary with the strength of the union movement and their institutional settings across industrial relations regimes. We do so by drawing on a large-n dataset of protest events for 27 European countries and the years 2000–2021, coded based on English-language newswires. We make an empirical contribution by introducing a novel indicator of unions’ movement character, which sets in relation classical strikes and other union protest beyond the workplace. We rely on the industrial relations regimes proposed by Visser (2009) and Ebbinghaus and Visser (1997) as our main analytical lens and distinguish between corporatist, social partnership, state-centred, liberal and the heterogenous transitional regimes (from here 76 G. H€ oltmann et al. on: Central and Eastern European (CEE) regimes). Combining these regimes with theories on union mobilization (power-resource hypothesis and institutional-access hypothesis), we formulate expectations as to how present unions are in the protest arena (levels of union protest) and the extent to which they engage in protest beyond strike at the workplace (movement character of union protest) across the different regimes. Overall, we find that unions are part of, on average, a third of all protest events across the regimes examined. Notably, for all but one industrial relations regime, workplacerelated strikes make up a minority of all union mobilization in the protest arena. While we find little variation across industrial relations regimes when it comes to the relative presence of unions in the protest arena, the regimes do play a significant role in explaining the varying movement character of union protest. We find that while the most powerful unions in the corporatist regimes are least likely to resort to protest action beyond the workplace, the structurally weakest unions of the CEE regimes display the strongest movement character. By contrast, our results demonstrate a surprisingly strong convergence between the social partnership, liberal and state-centred regimes, which all display intermediate levels of movement character. We therefore present initial evidence that the movement character of union protest is inversely related to the strength of the union movement and its institutional access: in other words, stronger union movements with high institutional access rely more heavily on strikes over other forms of protest, whereas weaker union movements in contexts of low institutional access primarily mobilize outside of the workplace. 2. Theoretical framework 2.1 Trade unions as protest actors: uniting perspectives from social movement and industrial relations literature The role of trade unions as protest actors has received little attention in both the social movement and industrial relations literature. With the emergence of new social movements and the demise of the ‘social question’, trade unions have increasingly disappeared from social movement research. Only recently, in the wake of the Great Recession, have social movement scholars rediscovered trade unions as protest actors: The role of unions in antiausterity protests since the onset of the global financial crisis has led to a call to re-embed unions in social movement research (e.g. Chesta et al., 2019; Pilati and Perra, 2020, 2022; Portos and Della Porta, 2020). Overall, these case studies observe more and more labor mobilization from positions of structural weakness and with a reorientation towards movement tactics (Portos and Della Porta, 2020). Chesta et al. (2019) describe worker mobilization in the precarious gig economy, Pilati and Perra (2020) show that union mobilization was especially successful where unions maintained alliances with social movement actors, and Pilati and Perra (2022) point to the fragmentation of labor interests in the streets between traditional unions, non-working groups and precarious workers. The industrial relations literature on the other hand has a narrower, almost exclusive focus on strikes and union mobilization at the workplace (Kelly, 2015a; Jansson and Uba, 2023). This branch of research documents a continuous decline in strike action since the 1980s (Piazza, 2005; Gall and Allsop, 2007; Kelly et al., 2013; Kelly, 2015b). However, over the same period, the number of general strikes directed at governments has remained stable and even increased in certain countries (Gall and Allsop, 2007, pp. 62–65; Kelly et al., 2013). This has led Kelly (2015a,b) to suggest that ‘the collapse in strike activity [ … ] Trade union protest 77 may significantly understate the continuing ability of unions to mobilize members and citizens in anti-government protests and campaigns’ (p. 538, also see Hamann et al., 2013). There is thus reason to believe that protest in the streets beyond routine strikes remains an important—if not increasing—part of the contemporary repertoire of trade union mobilization. This so-called displacement theory has been examined in two case studies on Sweden (Jansson and Uba, 2023) and the UK (Gall and Kirk, 2018), which both find that, on the whole, the decline in strikes does not seem to be compensated by a rise in other forms of union protest (Gall and Kirk, 2018; Jansson and Uba, 2023). In this article, we set out to present a novel perspective on trade unions as protest actors by taking a large-n comparative perspective on union protest activities and painting a differentiated picture of union protest beyond workplace strikes. In doing so, we answer the call of social movement scholars to re-embed unions in the protest arena and to examine union protest not just in moments of economic crisis (Portos and Della Porta, 2020). We also seek to extend the perspective of industrial relations scholars to the realm of union protest beyond the workplace strike (Kelly et al., 2013; Jansson and Uba, 2023) and by asking how industrial relations settings affect unions’ protest repertoires. 2.2 The movement character of union protest: balancing strikes versus other union protest Trade unions have various channels and modes of action to choose from (Baccaro et al., 2003; Silver, 2003; Pilati and Perra, 2020). From the perspective of unions as interest groups, Streeck and Hassel (2003, p. 344) attribute three main institutional channels of influence for trade unions—preferential party linkages (historically mainly with leftist parties), quasi-state institutions that concede varying levels of social policy influence to unions, and tripartite councils. In some cases, unions have even launched their own candidates in elections or mobilized in direct democratic campaigns (e.g. in Switzerland) (Ebbinghaus and Visser, 2000, p. 6). However, when these institutional channels fail, strikes are unions’ classical form of action. Unions are the only protest actors with the powerful leverage of commanding work stoppages (Ebbinghaus and Visser, 2000, p. 14). At the same time, unions’ action repertoire in the extra-institutional realm goes far beyond the classical strike. Trade unions are traditionally present in the protest arena, regularly mobilizing for broader societal issues beyond just labour interests: Amongst others, unions took to the streets in the name of international peace and abortion rights, against the war in Iraq in the early 2000s, and, most recently, played an important role in the anti-austerity movements in the wake of the global financial crisis (e.g. Streeck and Hassel, 2003; Accornero and Ramos Pinto, 2015; Della Porta, 2015; Peterson et al., 2015; Hunger and Lorenzini, 2020). The literature on social movement unionism also emphasizes that unions’ antiinstitutional struggle includes engaging for broader societal issues and beyond one’s immediate constituency (Bacccaro et al., 2003; D€ orre, 2008; Fairbrother, 2008). Fairbrother (2008) suggests that the four distinguishing elements of social movement unionism are (a) encompassing grassroots mobilization, (b) building alliances with other societal actors, (c) embracing broader emancipatory claims and transformative societal visions and (d) sponsoring collective action beyond strikes or mobilization at the workplace (p. 214). Focusing in particular on this last element of social movement unionism, we want to draw a conceptual distinction between workplace strikes on the one hand, and other forms 78 G. H€ oltmann et al. of union protest on the other hand. We suggest that a central dividing line between strikes and other forms of union protest lies in whether unions primarily address employers—in the form of workplace strikes—or whether they also address the state and the public, through protest action beyond the workplace on public economic, political and other societal issues. While workplace strikes are the primary action form to address grievances with employers, demonstrations and other protest forms beyond the workplace may be better suited to gain the attention of elected decision-makers and the general public, especially when demands are related to broader economic or societal issues (Brandl and Traxler, 2010; Peterson et al., 2012). We thus propose to define the extent to which unions maintain a ‘movement character’ in terms of protest forms and issues that go beyond ‘classical’ strikes at the workplace. In our analysis, we therefore distinguish between workplace strikes on the one hand versus union actions beyond the workplace (such as general strikes, demonstrations, blockades or symbolic protest forms) on the other hand. To what extent unions engage in strike versus other protest forms reveals whether unions primarily focus on conflicts with employers (strike-heavy) or on conflicts with government, society and the economy at large (protestheavy). In the following, we discuss how this balance may depend on the institutional structures that unions are embedded in. 2.3 Theories of union mobilization: when do unions mobilize for strike and other forms of protest? While research discussing the factors that determine the occurrence and overtime decline of workplace strikes is extensive, there is very little work when it comes to explaining other forms of union mobilization. Exceptions are for example Brandl and Traxler (2010), who have written about factors determining the occurrence of strike and lockouts as well as Lindvall (2013) and Kelly et al. (2013) who have examined factors determining political strike. We therefore draw on theories that explains workplace and political strike and hypothesize how they might travel to union protest more generally. As Kelly (2015a) writes, the factors that lead to strikes can be grouped into economic, political and institutional factors. Here, we focus on institutional factors, because we are interested in cross-country patterns rather than changes over time. From an institutional perspective, how frequently unions move into the protest arena is affected by their associational power on the one hand (power-resource hypothesis) and their access to institutions on the other hand (access hypothesis) (Brandl and Traxler, 2010). Associational power in the form of union membership (union density) is thought to provide unions with the basic capacities and resources for mobilization (Shalev, 1992; Piazza, 2005; Lindvall, 2013; Kelly, 2015a; Yang and Kwon, 2019). In contrast, institutional access and corporatist institutions more generally are thought to reduce the likelihood for protest and increase the chances that labour conflicts are resolved by negotiation (Korpi and Shalev, 1979; Brandl and Traxler, 2010). This can materialize in terms of high bargaining levels or structures that provide unions with access to institutions or influence in the policymaking process such as tripartite councils. Limited institutional and political access should therefore make unions more likely to take it to the streets (Della Porta, 2006; Gall and Allsop, 2007). There is mixed evidence in support of these broad mechanisms: Brandl and Traxler (2010) find evidence that union density relative to business power has a mobilizing effect. For political strikes, Kelly et al. (2013) find union density to be less relevant in predicting Trade union protest 79 political strikes than unions’ inclusion in policymaking. Moreover, union density does not necessarily have a linear relationship with union mobilization. Rather, several studies propose and find a non-monotonic relationship: Tsebelis and Lange (1995) theoretically introduce and empirically demonstrate a curvilinear relationship between union density and strike levels. Piazza (2005) also finds a non-monotonic relationship between union density and strike levels. Lastly, Lindvall (2013) is able to extend these findings to the occurrence of political strikes. His argument for a curvilinear relationship is essentially a synthesis of the power-resource and access-hypotheses: Both in the case of very powerful unions that can gain concessions from employers or governments by the mere threat of strikes (and therefore do not need to resort to militancy), as well as in the case of weak union movements that are not in a position to mobilize because of a lack of resources and members, political or economic strikes are unlikely. As a result, union movements of medium associational power should be most likely to resort to political strikes, given that the negotiating situation is dominated by uncertainty about the other sides’ bargaining power (Tsebelis and Lange, 1995; Lindvall, 2013). We would expect that the curvilinear argument should apply not only to the frequency of economic and political strikes, but also to the level of union protest overall. We would therefore expect that both very strong and very weak union movements should have the lowest presence in the protest arena, while union movements of intermediate power should mobilize most frequently. However, would we also expect this interplay between resources and institutional access when it comes to the movement character of union protest? It is important to note that this indicator captures not the level, but the composition of union protest: in other words, the balance between engaging in strikes versus other forms of protest in the streets. We argue that while the level of union protest is very much related to a combination of resources and institutional access, the composition of union protest should not necessarily be related to union resources. It is not a question of how many protests a union can stage, but whether it focuses on a strike-heavy or a protest-heavy repertoire. Returning to the arguments about power resources and institutional access, we expect the degree of institutional access to play the central role for the composition (movement character) of union protests. We argue that precisely because they lack access to the institutional channels necessary to address their grievances, weaker union movements may need to engage in non-strike mobilization and movement-oriented actions to expand their membership and develop crosssectoral solidarity. By contrast, if unions are safely embedded in an industrial setting that provides them with sufficient institutional access, such as access to policymaking, high bargaining power and coverage, the incentive to resort to broad mobilization or militancy should be strongly reduced. We therefore expect to find the lowest movement character (strike-heavy repertoire) in settings in which unions are strongly institutionally embedded, while we expect a high display of movement character (protest-heavy repertoire) in settings which offer low institutional access, forcing unions to the streets instead of the negotiating table. 2.4 Varieties of union protest across industrial relations regimes Now that we have discussed the institutional factors that shape union mobilization, we introduce the industrial relations regimes and formulate expectations for the level and movement character of union protest in each. We opt for a regime perspective because the industrial relations regimes bundle several variables of institutional union settings, but also 80 G. H€ oltmann et al. because they capture a certain historical path dependency of union movements, which we deem highly relevant for understanding unions’ protest behavior. For the same reasons, the social movement literature often examines protest patterns from a regional perspective (e.g. Borb� ath and Gessler, 2020; Kriesi et al., 2020a). Capturing different settings of union density, bargaining levels and institutional access, Visser (2009) based on Ebbinghaus and Visser (1997) suggests five industrial relations regimes: organized corporatism (Scandinavia/Northern Europe); social partnership regimes (Continental Europe); state-centred regimes (Southern Europe including France); liberal regimes (Anglo-Saxon countries); and the CEE regimes (Central-Eastern Europe). Combining the industrial relations typology with the power-resource and institutionalaccess hypotheses, we can formulate expectations as to (a) the level and (b) the movement character of union protest. The regimes can roughly be placed on a scale ranging from high to low levels of union movement strength. While the organized corporatism of the Scandinavian countries champions the highest levels of union density, bargaining levels and institutional access to this day, the heterogenous regimes of Central and Eastern Europe can be placed on the low end of this scale. The social partnership, state-centred and liberal regimes lie in between these two extremes, where the social partnership countries share more similarities with organized corporatism than the latter regimes. Starting with our expectations for the level of union protest, we expect a curvilinear protest pattern where both very strong and very weak union movements display low levels of union protest, while union movements of intermediate strength should be most likely to engage in protest. We would therefore expect that the high levels of institutional access, higher union densities and a strong cooperative relationship with employers should encourage relative labour quiescence in the corporatist and social partnership regimes. The state-centred and liberal regimes, by contrast, are characterized by respectively lower union density (particularly in the liberal regimes) and less institutional access (such as sectoral to companylevel bargaining or lower bargaining coverage), making conflictual labour relations more likely. Especially in southern Europe’s state-centred regimes, trade unions have, despite low membership numbers, ‘shown a tremendous capacity to mobilize people in strikes and protests well beyond the ranks of their own membership’, a prime example being France, with very low union density but very high bargaining coverage (Kelly, 2015b, p. 538). Lastly, union movements in the heterogenous regimes of Central and Eastern Europe score very low in terms of all of these indicators, with bargaining practices described as ‘acquiescent’ (Visser, 2009, p. 49). Given the structural weakness of unions in the CEE regimes, we therefore expect unions’ presence in the CEE protest arena to be especially low. Moving on to our expectations on the movement character of union protest, we expect the decision to engage in protest actions beyond the workplace to be linked mainly to a lack of institutional access, as argued in Section 2.3. Here, we would hypothesize that stronger union movements with more institutional access are less likely to resort to movement tactics than unions in weaker industrial relations settings with less institutional access. In Northern Europe, and under organized corporatism we observe the strongest institutional access of all the industrial relations regimes: In that context, unions almost fully rely on the political process to reach their ends (Streeck and Hassel, 2003, p. 345), to the extent that they are sometimes referred to as ‘labour market parties’ (Hyman, 2001). This level of institutional access should render movement-oriented protest forms far less necessary. Trade union protest 81 However, with decreasing institutional access, which can manifest in terms of lower bargaining levels, the absence of tripartite councils or involvement in policymaking, we expect to see an increasing share of protest actions beyond the workplace strike. We therefore expect to see a higher movement character in the social partnership, state-centred and liberal regimes compared to organized corporatism. Lastly, institutional access is lowest in the CEE regimes, leading us to expect that they invest primarily in union protest beyond the workplace. This expectation is also in line with scholarship on unions’ protest repertoire in the CEE regimes: Since the end of communist rule, unions have primarily organized against the state not against capital (Ost, 2002) and have continued to display tactics oriented towards social movement unionism (Varga, 2015; Dolenec et al., 2021). Table 1 summarizes our expectations with regard to the level and the movement character of union protest across the different regimes. 3. Data and methodology We now turn to our empirical material and strategy of data analysis. One of the principal reasons why the industrial relations literature tends to focus on workplace strikes is the lack of publicly available datasets that collect data on union protest action beyond the strike (Kelly, 2015a). Further, official national strike statistics suffer from serious shortcomings when analysed in a comparative manner: between countries, measures of industrial action often differ in their definition and sources (Brandl and Traxler, 2010; Kelly, 2015a). In this article, we rely on an updated version of the recently published PolDem Protest Dataset, covering a total of 30 European countries based on the coding of 10 English language newswires (Kriesi et al., 2020b; Lorenzini et al., 2022). Our updated version of the dataset allows us to cover the years from 2000 to 2021. The underlying method, protest event analysis (PEA), has been one of the main approaches in the study of longitudinal and cross-sectional evolution of protest (Hutter, 2014). While protest event data has its own shortcomings, it has the major advantage of allowing us to set strikes and other protest events in relation, given that they are drawn from the same data source and systematically collected in the same manner. We further expand on our decision to use PEA vis-� a-vis official strike statistics in Supplementary Appendix D and present a comparison of the two data types. In our dataset, union sponsorship of a protest event includes instances when unions (co- )organize, take part in and/or call for participation in a protest event. We adopt this broad understanding of sponsorship because, as Rucht (1998, p. 41) highlighted some time ago, Table 1 Expectations: level and movement character of union protest by regimes Organized corporatism (North) Social partnership (West) State-centred (South) Liberal (Anglo-Saxon) CEE (Central and Eastern) Level of union protests Low Low High High Low Movement character of union protests Low Medium Medium Medium High 82 G. H€ oltmann et al. In the final step of our analysis, we test the effect of the union regimes on the level and movement character of union protest in a multilevel logit regression. We do so to identify the effects of industrial relations regimes on top of cross-national heterogeneity and contextual factors, namely, unemployment rates, fluctuations in government cabinets and the financial crisis. Table 3 presents the results. Starting with the role of the industrial relations regimes in these two models, we find that while the regimes appear to have no (significant) association with the relative presence of unions in the protest arena (DV1/Model 1), we do observe a significant relationship between the regimes and the movement character of union protest (DV2/Model 2). Corroborating our previous descriptive findings, the margins plot in Figure 5 shows that the corporatist and CEE regimes stand apart with the lowest and highest movement character respectively. The remaining regimes cluster relatively close together in the middle. However, it should be noted that while there is a substantive difference in movement character between the state-centred and CEE regimes, this difference is not significant, suggesting that the two regimes are in fact similarly movement oriented. Supplementary Appendix Table A1 provides an overview of the same regression with rotating reference categories. While there is no strict linear relationship between the (lack of) institutional access and movement character, our findings do strongly suggest that in contexts of high Table 2 Union protest repertoire by industrial relations regimes Corporatism (North) Social partnership (West) State-centred (South) Liberal (Anglo-Saxon) CEE regimes Issues Econ Private 77.5 39.5 22 40 19 Econ Public 16 48.5 59 44 63.5 Political/Cultural�6 10 14.5 9.5 11.5 Other issue 1 2 5 6.5 6 Action Forms Strikes 79.5 58.5 52 68 38.5 Demonstrations 12 32 35.5 25.5 44 Petition/symbolic 2 3 2 2.5 8.5 Confrontations/blockades 3.5 4.5 6.5 3 6 Violent protest 1.5 0.5 2.5 0.5 0.5 Other protest 1.5 1.5 1 0.5 2.5 Cooperation with other protest actors Unions ‘alone’ 67 62 70 73.5 74.5 w. occupational social groups 25 29 20.5 20 19 w. other social groups 1 1.5 2.5 1.5 1.5 w. political parties 0.5 2 2 1.5 2.5 w. other actors 2 4 2.5 2 1.5 Notes: � The category ‘Political/Cultural’ includes issues coded by Kriesi et al. (2020a,b) as political, culturally liberal, culturally conservative or xenophobia. Percentages are calculated for an average country in each region. The difference to 100% is due to rounding errors. Trade union protest 89 institutionalization and union power, the movement character of union protest is low, while the opposite is the case for contexts of very low institutional access and union power. Moving on to the specific characteristics of union protest, the regression supports the tendencies presented descriptively: unions are more likely to sponsor protest on private or public economic issues, while they are less likely to participate in events related to political or cultural issues (Model 1). Further, the most likely action form for unions is the strike, followed by demonstrations. Importantly, we find that when unions mobilize beyond the Table 3 Level of union protest (DV1) and movement character of union protest (DV2) Model 1 Model 2 Union sponsorship of protest (¼1) Movement character of union protest (¼1) Region (Ref: Corporatist: North) Social-partnership (West) 0.07 (0.29) 1.95 (0.57)��� State-centred (South) 0.36 (0.30) 2.04 (0.59)��� Liberal 0.18 (0.33) 1.31 (0.65)� CEE regimes 0.13 (0.27) 2.89 (0.54)��� Protest Issues Private economic (0,1) 2.25 (0.07)��� Public economic (0,1) 2.27 (0.06)��� Political (0,1) −1.09 (0.07)��� Cultural (0,1) −1.32 (0.08)��� Action Forms Demonstration (0,1) 0.90 (0.06)��� Strike (0,1) 2.49 (0.07)��� Protest Actors Occupational social group (0,1) −3.12 (0.06)��� −0.07 (0.09) Other social group (0,1) 0.99 (0.12)��� 2.62 (0.48)��� Parties (0,1) −1.06 (0.09)��� 3.00 (0.52)��� Participation rate 5.06 (0.94)��� 12.47 (3.07)��� Crisis (Ref: Normal Times) Shock period 0.31 (0.13)�0.73 (0.23)�� Euro crisis −0.09 (0.11) 0.08 (0.18) Post-crises years (after 2015) −0.97 (0.11)��� 0.29 (0.20) Political & economic context Cabinet left-right −0.03 (0.19) −0.03 (0.32) Unemployment (quarterly) 0.24 (0.32) 1.89 (0.58)�� Num. obs. 25711 6638 Num. groups: str_label 589 479 Num. groups: iso2code 27 27 Note: ��� P<.001; �� P<.01; � P<.05. The dependent variables are binary and are defined in the following way: DV1: Levels of union protest; 1 ¼Union-sponsored protest events; 0 ¼All other protest events; DV2: Movement character of union protest: 1 ¼Non-standard union protest; 0 ¼Workplace strikes. We use multilevel logit regressions with random intercept, with protest events nested in country�year contexts, nested in countries. Protest issues and action forms appear in model 1, but not in model 2, since the second dependent variable is defined by the combination of action forms and protest issues. 90 G. H€ oltmann et al. workplace strike, they are more likely to cooperate with non-occupational social groups or parties than to protest alone (Model 2). Exploring the role of additional contextual factors, we find that the political orientation of the government does not play a role for unions’ presence in the protest arena, or for the movement character of union protest. Contrary to the literature on the relationship between unemployment and union protest, our findings show that unemployment rates do not significantly affect unions’ relative presence in the protest arena (DV1). At the same time, higher unemployment seems to increase the likelihood of unions to move beyond the workplace strike and mobilize in the streets, even when controlling for the crisis years (DV2). Lastly, we perform three robustness tests on our regime effects. First, to ensure that our results are not only due to the prolonged economic crisis, we follow Kriesi et al. (2020a) in exploring a crisis effect, and run a split sample regression for the pre-crisis years (2000– 2007) versus the crisis years (2008–2015) (see Supplementary Appendix Table B1). Second, to strengthen the case for our regime argument, we test to what extent the effect of union regimes is robust to controlling for intra-regional heterogeneity in industrial relations (union density, bargaining levels, tripartite councils and routine involvement) and whether the effect of these variables holds explanatory power beyond the clusters of union regimes (see Supplementary Appendix Tables B2 and B3). Here we also test whether the regime Figure 5 Predicted probabilities of industrial relations regimes on unions’ movement character (Model 2; DV2). Note: Figure 5 presents predicted probabilities for the effect of the industrial relations regimes on unions’ movement character. The results are calculated based on Model 2 (DV2) in Table 3 above. Trade union protest 91 differences are explained by differential exposure to the economic crisis, the so-called refugee crisis or the Coronavirus disease 2019 (Covid-19) pandemic. Third, to account for possible heterogenous overtime and cross-national characteristics, we change our modelling strategy and, despite their limitation (Imai and Kim, 2021), employ a two-way fixed effects model with regime and year dummies and two-way clustered standard errors. Starting with the split sample regression, we find that for both of these time periods, unions were equally present in the protest arena compared to other protest actors (DV1). Unions were therefore not disproportional carriers of protest during the crisis, but were also not underrepresented. In a similar vein, we do not find significant differences for the movement character of the different union regimes (DV2) when comparing the pre-crisis with the crisis years (with the exception of the liberal regime as in Supplementary Appendix Figure 2B shows). Interactions between the crisis periods (normal times, shock period and Eurozone crisis) and the union regimes do not yield significant results for either DV1 or DV2. We therefore conclude that the regime effects on both dependent variables are robust to the period effect of the Eurozone crisis. Turning to our second robustness test, we compare classical institutional union variables against the industrial relations regimes. We find that in settings of decentralized bargaining (company and industry) as well as with regular involvement in policymaking (full levels of concertation), unions are less present in the protest arena. In contrast, higher union density significantly increases the presence of unions in the protest arena in a linear, as opposed to a curvilinear fashion (DV1). In comparison, none of the institutional union variables are significant in explaining the movement character of union protest until the union regimes are introduced (DV2). In conclusion, the industrial relations regimes outperform classical institutional variables in explaining the movement character (DV2), but not the relative presence of unions in the protest arena (DV1). This suggests that there is a (possibly cultural) regime effect in explaining the movement character of union protest beyond just institutional factors. Controlling for differential exposure to economic circumstances (unemployment & GDP per capita), asylum seekers or Covid-19 cases does not change the substantive conclusions we draw on the relative differences between the regimes in predicting the movement character of trade unions. Lastly, we run our analyses using regime and year fixed effects. The results are substantively similar to the conclusions we presented above. We find that the regimes only differ in the movement character of unions, but not in the level of mobilization. In terms of the movement character of union protest (DV2), the difference between the state-centred south and the CEE regime shrinks, but the relative rank order largely follows the expected pattern. 5. Discussion and conclusion In this article, we have built bridges between social movement studies and industrial relations research by examining how the level and movement character of union protest varies in the context of different industrial relations regimes. We make a theoretical contribution to the social movement and industrial relations literature by extending theories on economic and political strike occurrence to union protest beyond the workplace (Korpi and Shalev, 1979; Brandl and Traxler, 2010; Lindvall, 2013). We make an empirical contribution by examining workplace strikes and protest beyond the workplace jointly as part of 92 G. H€ oltmann et al. unions’ protest repertoires (Jansson and Uba, 2023) and by offering a quantitative operationalization of the ‘movement character’ of union protest to debates on social movement unionism (Della Porta, 2006; Fairbrother, 2008). We also present the first large-n comparative assessment of union protest across industrial relations regimes. First, countering claims by social movement scholars, we are able to show that unions are still highly relevant protest actors, accounting for about a third of all protest events, even independent of crisis moments like the Great Recession. Further, our analysis shows that the composition of union protest across Europe is much more movement-oriented than expected: Beyond the corporatist North, we see that union protest in the public sphere notably outweighs union protest at the workplace. Public economic, political and other issues characterize union protest action more than private economic issues. Contrary to what one might expect in environments of high social movement unionism, we find that in the overwhelming majority of cases, unions sponsor protest on their own and relatively rarely ally with other civil society actors. Second, we find that unions’ movement orientation strongly depends on the industrial relations regime: To what extent unions resort to strike-heavy versus protest-heavy repertoires depends on the institutional setting and the related strength of union movements. Our regression analyses show that while there is no strict linear relationship, highly institutionalized union settings tend to display low levels of movement character, while we observe higher degrees of movement orientation in settings with low institutional access. It is especially notable that these regime effects on unions’ movement character are robust to classical institutional union variables as well as to the years of the Great Recession and Eurozone crisis. What are the general implications of our findings? We observe remarkable vitality of unions as protest actors at the beginning of the 21st century. This also suggests that viewing unions as pure interest groups or focusing exclusively on strike patterns does not do justice to their varied mobilization repertoire. On a larger scale, union engagement beyond the workplace may be a reflection of their position in a hyper-globalized economy. In this context, protest beyond strikes may be both the result of as well as a means to compensate for declining institutional and associational power. In line with the social movement unionism literature, broader protest action can be both a sign of structural weakness but also a tool and strategy for organization and revitalization. Examining long-term changes over time with data covering earlier periods of union strength and subsequent decline can give us more insight into whether our findings are a sign of union revitalization or document a struggle from a position of weakness. Supplementary material Supplementary material is available at Socio-Economic Review journal online. Funding Funding support for this article was provided by the Volkswagen Foundation (Lichtenberg Professorship, grant number 93598). Conflict of interest statement. The authors report there are no competing interests to declare. Trade union protest 93 References Accornero, G. and Ramos Pinto, P. (2015) ‘Mild Mannered’? Protest and Mobilisation in Portugal under Austerity, 2010–2013’, West European Politics, 38, 491–515. https://doi.org/ 10.1080/01402382.2014.937587 Baccaro, L., Hamann, K. and Turner, L. (2003) ‘The Politics of Labour Movement Revitalization: The Need for a Revitalized Perspective’, European Journal of Industrial Relations, 9, 119–133. Bewernitz, T. and Dribbusch, H. (2014) ‘“Kein Tag Ohne Streik”: Arbeitskampfentwicklung Im Dienstleistungssektor’, WSI-Mitteilungen, 67, 393–401. https://doi.org/10.5771/0342-300X2014-5-393 Borb� ath, E. and Gessler, T. (2020) ‘Different Worlds of Contention? Protest in Northwestern, Southern and Eastern Europe’, European Journal of Political Research, 59, 910–935. https:// doi.org/10.1111/1475-6765.12379 Brandl, B. and Traxler, F. (2010) ‘Labour Conflicts: A Cross-National Analysis of Economic and Institutional Determinants, 1971-2002’, European Sociological Review, 26, 519–540. https:// doi.org/10.1093/esr/jcp036 Chesta, R. E., Zamponi, L. and Caciagli, C. (2019) Labour Activism and Social Movement Unionism in the Gig Economy. Food Delivery Workers Struggles in Italy (1.0) [dataset], Lecce, Italy, University of Salento. https://doi.org/10.1285/I20356609V12I3P819 Della Porta, D. (2006) ‘From Corporatist Unions to Protest Unions? On the (Difficult) Relations between Organized Labour and New Social Movements’. In Crouch, C. and Streeck, W. (eds) The Diversity of Democracy, Corporatism, Social Order and Political Conflict, Cheltenham, Edward Elgar Publishing, pp. 71–97. Della Porta, D. (2015) Social Movements in Times of Austerity: Bringing Capitalism Back into Protest Analysis, Cambridge, Polity Press. Dolenec, D., � Sirini� c, D. and Balkovi� c, A. (2021) ‘Resisting the Great Recession: Social Movement Unionism in Croatia and Serbia’, European Journal of Industrial Relations, 28, 105–121. https://doi.org/10.1177/09596801211023270 D€ oring, H. and Manow, P. (2022) ‘Parliaments and Governments Database (ParlGov): Information on Parties, Elections and Cabinets in Modern Democracies’, ParlGov, accessed at https://www.parlgov.org D€ orre, K. (2008) ‘Die Strategische Wahl Der Gewerkschaften—Erneuerung Durch Organizing’, WSI-Mitteilungen, 61, 3–10. https://doi.org/10.5771/0342-300X-2008-1-3 Ebbinghaus, B. and Visser, J. (1997) ‘Der Wandel Der Arbeitsbeziehungen Im Westeurop€ aischen Vergleich’. In Hradil, S. and Immerfall, S. (eds) Die Westeurop€ aischen Gesellschaften Im Vergleich, Germany, Leske þBudrich, pp. 333–376. Ebbinghaus, B. and Visser, J. (2000) Trade Unions in Western Europe since 1945, London, Palgrave Macmillan UK. https://doi.org/10.1007/978-1-349-65511-3 Fairbrother, M. (2014) ‘Two Multilevel Modeling Techniques for Analyzing Comparative Longitudinal Survey Datasets’, Political Science Research and Methods, 2, 119–140. https:// doi.org/10.1017/psrm.2013.24 Fairbrother, P. (2008) ‘Social Movement Unionism or Trade Unions as Social Movements’, Employee Responsibilities and Rights Journal, 20, 213–220. https://doi.org/10.1007/s10672008-9080-4 Gall, G. and Allsop, D. (2007) ‘Labor Quiescence Continued? Recent Strike Activity in Western Europe’, LERA for Libraries, 51–71. Gall, G. and Kirk, E. (2018) ‘Striking out in a New Direction? Strikes and the Displacement Thesis’, Capital & Class, 42, 195–203. https://doi.org/10.1177/0309816818770611 94 G. H€ oltmann et al. Grote, J. and Wagemann, C. (2018) Social Movements and Organized Labour, Passions and Interests, Abingdon, Routledge. Hunger, S. and Lorenzini, J. (2020) ‘All Quiet on the Protest Scene’. In Kriesi, H., Lorenzini, J., H€ ausermann, S. and W€uest, B. (eds) Contention in Times of Crisis: Recession and Political Protest in Thirty European Countries, Cambridge, Cambridge University Press, pp. 104–127. Hutter, S. (2014) ‘Protest Event Analysis and Its Offspring’. In Della Porta, D. (ed) Methodological Practices in Social Movement Research, Oxford, Oxford University Press, pp. 335–367. Hyman, R. (2001) Understanding European Trade Unionism: Between Market, Class and Society, Thousand Oaks, CA, SAGE. Hyman, R. (2015) ‘Austeritarianism in Europe: What Options for Resistance’. Social Policy in the European Union State of Play, Brussels, Belgium, European Trade Union Institute, pp. 97–119. https://etui.org/sites/default/files/Austeritarianism%20in%20Europe%20Richard%20Hyman.pdf Imai, K. and Kim, I. (2021) ‘On the Use of Two-Way Fixed Effects Regression Models for Causal Inference with Panel Data’, Political Analysis, 29, 405–415. Jansson, J. and Uba, K. (2023) ‘Trade Unions’ Protest Cycles in Sweden, 1980–2020’. In Jørgensen, J. and Mikkelsen, F. (eds) Trade Union Activism in the Nordic Countries since 1900, Berlin, Germany, Springer International Publishing, pp. 255–278. https://doi.org/10. 1007/978-3-031-08987-9_13 Kelly, J. (2015a) ‘Conflict: Trends and Forms of Collective Action’, Employee Relations, 37, 720–732. https://doi.org/10.1108/ER-06-2015-0102 Kelly, J. (2015b) ‘Trade Union Membership and Power in Comparative Perspective’, The Economic and Labour Relations Review, 26, 526–544. https://doi.org/10.1177/ 1035304615614445 Kelly, J., Hamann, K. and Johnston, A. (2013) ‘Unions Against Governments: Explaining General Strikes in Western Europe, 1980–2006’, Comparative Political Studies, 46, 1030–1057. https://doi.org/10.1177/0010414012463894 Korpi, W. and Shalev, M. (1979) ‘Strikes, Industrial Relations and Class Conflict in Capitalist Societies’, The British Journal of Sociology, 30, 164–187.https://doi.org/10.2307/589523 Kriesi, H., Lorenzini, J., W€uest, B. and H€ ausermann, S. (eds) (2020a) Contention in Times of Crises: Comparing Political Protest in 30 European Countries, 2000–2015, Cambridge, Cambridge University Press. Kriesi, H., W€uest B., Lorenzini J., Makarov P., Enggist M., Rothenh€ ausler K., Kurer T., H€ ausermann S., Wangen P., Altiparmakis A., Borb� ath E., Bremer B., Gessler T., Hunger S., Hutter S., Schulte-Cloos J., and Wang C. (2020b) PolDem-Protest Dataset 30 European Countries, Version 1. Lesch, H. (2005) ‘International Comparison of Labour Disputes and Structural Change’, CESifo Forum, Ifo Institut F€ ur Wirtschaftsforschung an Der Universit€ at M€ unchen, 6, 42–52. Lindvall, J. (2013) ‘Union Density and Political Strikes’, World Politics, 65, 539–569. https://doi. org/10.1017/S0043887113000142 Lorenzini, J., Kriesi, H., Makarov, P. and W€uest, P. (2022) ‘Protest Event Analysis: Developing a Semiautomated NLP Approach’, American Behavioral Scientist, 66, 555–577. https://doi.org/ 10.1177/00027642211021650 Ost, D. (2002) ‘The Weakness of Strong Social Movements: Models of Unionism in the East European Context’, European Journal of Industrial Relations, 8, 33–51. https://doi.org/10. 1177/095968010281003 Peterson, A., Wahlstr€ om, M. and Wennerhag, M. (2012) ‘Swedish Trade Unionism: A Renewed Social Movement’, Economic and Industrial Democracy, 33, 621–647. https://doi.org/10. 1177/0143831X11425257 Trade union protest 95 Peterson, A., Wahlstr€ om, M. and Wennerhag, M. (2015) ‘European Anti-Austerity Protests—Beyond “Old” and “New” Social Movements? ’, Acta Sociologica, 58, 293–310. https://doi.org/10.1177/0001699315605622 Piazza, J. A. (2005) ‘Globalizing Quiescence: Globalization, Union Density and Strikes in 15 Industrialized Countries’, Economic and Industrial Democracy, 26, 289–314. https://doi.org/ 10.1177/0143831X05051518 Pilati, K. and Perra, S. (2020) ‘Trade Unions and Labour Conflicts: Social Movement and Radical Political Unionism in France and Italy’, Italian Political Science, 14, 167–187. Pilati, K. and Perra, S. (2022) ‘The Insider–Outsider Divide and Contentious Politics: The Tripartite Field of the Italian Labour Movement’, West European Politics, 45, 1283–1309. https://doi.org/10.1080/01402382.2022.2030593 Portos, M. and Della Porta, D. (2020) ‘Social Movements in Times of Inequalities: Struggling against Austerity in Europe’, Structural Change and Economic Dynamics, 53, 116–126. Rucht, D. (1998) ‘The Structure and Culture of Collective Protest in Germany since 1950’. In Meyer, D. and Tarrow, S. (eds) The Social Movement Society: Contentious Politics for a New Century, Lanham, MD, Rowman & Littlefield, pp. 29–57. Shalev, M. (1992) ‘The Resurgence of Labour Quiescence’. In Regini, M. (ed.) The Future of Labour Movements, Thousand Oaks, CA, Sage, pp. 102–132. Silver, B. J. (2003) Forces of Labor: Workers’ Movements and Globalization Since 1870, New York, Cambridge University Press. Streeck, W. and Hassel, A. (2003) ‘The Crumbling Pillars of Social Partnership’, West European Politics, 26, 101–124. https://doi.org/10.1080/01402380312331280708 Tsebelis, G. and Lange, P. (1995) ‘Strikes around the World: A Game Theoretic Approach’. In The Workers of Nations: Industrial Relations in a Global Economy, Oxford, Oxford University Press, pp. 101–125. Varga, M. (2015) ‘Trade Unions and Austerity in Central and Eastern Europe: Did They Do Something about It’, Transfer: European Review of Labour and Research, 21, 313–326. https://doi.org/10.1177/1024258915585943 Visser, J. (2009) The Quality of Industrial Relations and the Lisbon Strategy, Luxembourg, Office for Official Publications of the European Communities, pp. 45–72. https://doi.org/10. 2767/54876 Visser, J. (2016) Database on Institutional Characteristics of Trade Unions, Wage Setting, State Intervention and Social Pacts, 1960–2011 (ICTWSS), Amsterdam, Netherlands, University of Amsterdam. Yang, J. and Kwon, H. Y. (2019) ‘Union Structure, Bounded Solidarity and Support for Redistribution: Implications for Building a Welfare State’, International Political Science Review, 42, 277–293. https://doi.org/10.1177/0192512119887535 © The Author(s) 2024. Published by Oxford University Press and the Society for the Advancement of SocioEconomics. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. Socio-Economic Review, 2025, 23, 75–96 https://doi.org/10.1093/ser/mwae056 Article 96 G. H€ oltmann et al.