This is a repository copy of Party membership and social movement activism: A macro-micro analysis. White Rose Research Online URL for this paper: http://eprints.whiterose.ac.uk/140269/ Version: Accepted Version Article: Giugni, M. and Grasso, M.T. orcid.org/0000-0002-6911-2241 (2019) Party membership and social movement activism: A macro-micro analysis. Party Politics. ISSN 1354-0688 https://doi.org/10.1177/1354068818823446 © The Author(s) 2019. This is an author produced version of a paper subsequently published in Party Politics . Uploaded in accordance with the publisher's self-archiving policy. Reuse Items deposited in White Rose Research Online are protected by copyright, with all rights reserved unless indicated otherwise. They may be downloaded and/or printed for private study, or other acts as permitted by national copyright laws. The publisher or other rights holders may allow further reproduction and re-use of the full text version. This is indicated by the licence information on the White Rose Research Online record for the item. Takedown If you consider content in White Rose Research Online to be in breach of UK law, please notify us by emailing eprints@whiterose.ac.uk including the URL of the record and the reason for the withdrawal request. eprints@whiterose.ac.uk https://eprints.whiterose.ac.uk/ Party membership and social movement activism: A macro-micro analysis Marco Giugni^ and Maria T. Grasso* ^Department of Political Science and International Relations, University of Geneva, Boulevard du Pont-d'Arve 40, Geneva Ch-1211, Switzerland E-mail: marco.giugni@unige.ch *Department of Politics, University of Sheffield, Elmfield, Northumberland Road, Sheffield S10 2TU, United Kingdom E-mail: m.grasso@sheffield.ac.uk 1 Abstract This paper examines the macro-micro dynamics linking party membership with protest participation. We theorise that institutional and extra-institutional engagement are mutually reinforcing and that party membership has a positive effect on party activism. We examine key ideational and structural factors identified in the literature to analyse the relative importance of various factors for party members’ involvement. We look at micro-macrolevel linkages by examining macro-level contextual variables as well as the extent to which these features mediate the individual level effects. Our results suggest that party members support social movements for a variety of ideational and structural reasons but that strategic reasons are also important. Moreover, we find that contexts marked by more open political opportunities close the gap in social movement activism between party members and nonmembers suggesting that contexts with higher public spending as well as crises can be capitalised on to engage the wider public into political activism. 2 Introduction The literatures on institutional and movement politics have historically followed different paths. More recently, however, scholarship has urged for the investigation of the relationship between these two spheres for understanding the dynamics of participation and engagement (Císa and Navrátil, 2015, della Porta et al., 2017, Heaney, 2013, Heaney and Rojas, 2015, Kriesi et al., 2012, Kriesi, 2014, McAdam and Tarrow, 2010, McAdam and Tarrow, 2013, Norris et al., 2015). While most of this literature to date focuses on aggregate-level processes of party-movement interaction, in this investigation we aim to break new ground and turn to examining the macro-micro dynamics linking party membership to social movement activism. While most scholarship aiming to link electoral and protest politics to date has focused on movement-party interactions, we aim to bridge the individual-level and political process perspectives by examining the extent to which individuals engage in protest and parties and how this is affected by the wider structures of political opportunities, in particular those opened up by the recent economic crisis and the austerity policies enacted by European governments. This in turn allows for the paper to provide important answers for many scholars as the approach allows us to analyse structural, cultural, and individual factors at the same time. This is of particular relevance since these factors are interlinked and interact. As such, if we neglect macro-micro linkages we miss something important about the relationship between party membership and movement mobilization. In turn, employing this type of analysis allows us to make sense of various factors at the same time and to more fully understand the democratic underpinnings of European and other industrial nations. 3 Times of crisis can be seen as times of shifting alignments where parties in government are particularly under pressure as they navigate the conflicting demands of being responsive to their core constituencies while at the same time balancing the budget sheets and fulfilling the expectations of supra-national bodies and financial markets (Giugni and Grasso 2018). As mainstream parties have increasingly moved to the centre-ground and espoused centrist policies, this has opened up political space for the populist right and other types of institutional and extra-institutional challengers. This has changed the relationship between parties and various movements, both on the Left and on the Right of the political spectrum. Most importantly, this is likely to have changed the ways in which people relate to both parties and movements. In this context, given the lack of attention to macro-micro linkages in the study of the party-movement nexus, we aim to bridge the study of parties and protest at the individual and contextual level by analysing the relationship between party membership and social movement activism in the context of economic crisis and austerity policies. The broader aim is to address questions pertaining to the linkages between different types of institutional and extra-institutional forms of activism in Western Europe as well as calls for studies analysing the links between parties and protest. We apply random intercept models and control for key socio-demographic characteristics to examine the relative importance of different types of factors for party members’ participation in social movement activities i.e. protest. We also control for key features of the political system that vary between countries as well as analysing the extent to which these features mediate the individual-level effects. We show that party membership has a positive effect on protest participation. We also show that individual unemployment has a negative effect on that person’s protest 4 participation, but societal unemployment increases participation in protest. These are important findings which contribute to the literature linking level institutional and extrainstitutional participation, showing also the impact of macro-level context, with particular reference to deteriorating economic conditions in times of crisis. Previous research and hypotheses Recent scholarship has begun to inquire into complex relationships between social movements and party politics, but mainly at the macro and meso level. Work on the relationship between protest and parties has focused on movements and has tended to argue that if parties articulate a certain discourse then that reduces the political space for movements and there is less need for mobilisation on a particular issue (Tilly, 1999, Tarrow, 1996). The literature has traditionally noted how political space delimits patterns of mobilisation by different actors. For example, research examining the extreme right has shown evidence that where established or moderate right-wing parties articulate a radical agenda, the space for radical parties decreases as potential supporters support the established party (Koopmans and Olzak, 2004, Koopmans et al., 2005). Research has further shown that the interaction of party and protest fields needs to be understood in terms of economic political values but also needs to take into account the importance of secondary conflict axes over cultural values that are becoming increasingly prominent (Kitschelt, 1986, Kitschelt, 1988). In the US, McAdam and Tarrow (2013) found that governments opened up opportunities for their movement allies. On the other hand, for 5 Eastern Europe there is evidence of countervailing tendencies where right-wing governments stimulate protest but leftist ones do not (Císa and Navrátil, 2015). Political process theorists, in particular, have historically considered the broader context of mobilisation central for understanding social movement activism. Factors such as the degree of openness or closedness of the political system and presence of institutional allies (Tarrow, 1994, Kriesi et al., 1995) or favourable discursive political opportunity structures (Koopmans et al., 2005) were seen as central factors for understanding mobilisation. However, despite these developments with respect to meso- and macro-level theorising, there has been little examination of the ways in which parties and protest relate to each other at the individual level including how this is conditioned by the wider political context and relevant political opportunities available. One notable exception is in the recent work of Heaney and Rojas (2015), who examine the macro-micro linkages between parties and movement by looking at individuallevel partisan and movement identities. They argue that partisan identities lead partisans to support social movements but also to drop movement participation when this no longer looks to benefit the party. In our investigation, we look at the impact of party membership on protest participation and how this varies based on different characteristics of the wider political context. While the individual level political participation literature has tended to look at institutional participation and protest participation separately, we argue that institutional engagement (party membership) is related to extra-institutional engagement (protest). We suggest that institutional activism is actually reinforcing of extra-institutional participation rather than belonging to a different sphere of action: H1: Party membership has a positive effect on protest participation. 6 But why does party membership spur protest participation? With respect to political process, the literature suggests a number of hypotheses linking party and protest activism. Based on previous research, three types of factors in particular can be understood to explain the circumstances under which party members might support and/or become involved in social movement activities (Piccio, 2016). While these types of factors have typically been examined at the aggregate level and often taken separately, in the first step of our analysis, we want to test the extent to which they impact on individual-level participation. These can be summarised in three main strands. For each, we develop a specific hypothesis. First, party members might protest because they share social movement’s goals and ideological leanings. In this respect, research has shown that for reasons of identity coherence individuals will tend to participate in social movement activities only when they feel a certain affinity with their goals (Kriesi et al., 1992). As such, if party members participate in protest for purely ideological reasons, we expect that: H2: The effect party membership on protest participation is no longer significant when controlling for left-right and libertarian-authoritarian values. Moreover, party members might participate in protest because they are immersed in organisational networks which act as catalysts for passing on information about events and increasing the possibility that members might be asked to participate by other organisational members. We argue that, since membership in political parties is likely to be linked to participation in other organisations (Norris, 2002), then party members are more likely to become involved in social movement activities since they are more likely to find out about them and get recruited in to them by their wider networks (Schussman and Soule, 7 2005). As such, if party members participate in protest mainly due to network effects, we expect that: H3: The effect party membership on protest participation is no longer significant when controlling for the degree of involvement in other organisations. Moreover, while the literature has identified both ideational and structural factors which might explain the link between party membership and protest participation, we expect that party members also act strategically and support social movements when they think that this will benefit their parties. Vulnerability in the parties’ electoral environment has been seen as favouring interaction with social movements since parties are understood to employ strategies that are beneficial for their organization i.e. electoral support (Kriesi and Wisler, 1996, Goldstone, 1991). In times of economic crisis and shifting alignments, strategic reasons might thus become particularly relevant for party members to engage in protest activism. As such, if party members participate in protest mainly for strategic reasons, we expect that: H4: The effect of party membership on protest participation remains significant even when controlling for ideational, network and when all the other relevant controls discussed in the literature on protest participation are accounted for. This hypothesis is based on a process of elimination logic. Once all the other relevant mechanisms accounting for the link between party membership and protest are controlled for, any residual effect of party membership on protest participation would strongly suggest strategic reasons for movement participation by partisans. In terms of the other relevant controls that need to be accounted for before we can speak of a residual, strategic effect of party membership in protest, previous research has clearly shown that not 8 all sectors of society engage in social movements and protest activism evenly. Research has shown that on the whole participants in social movement activities tend to be younger and male (Schussman and Soule 2005). However, the class bases of protest are understood to have become increasingly diffuse. While on the one hand early scholarship saw protest as “weapons of the weak” (Scott, 1976, Tilly, 1978), others have since argued that sociocultural specialist (Kriesi, 1989) should have by now become predominant in this repertoire. Most recently, Standing (2011) and della Porta (2015) have suggested that the precariat could potentially form a new class basis for contention in the context of the current economic crisis. Education is normally understood to play a very important role for spurring protest, particularly through its cognitive liberation functions (Dalton, 2008). Moreover, political interest and political efficacy - both internal and external - are all normally understood as key spurs for protest (Schussman and Soule 2005). Moreover, our analysis aims to tackle the link between the micro and macro-level, and the broader context is also understood to have a role to play for protest mobilisation. In particular, times of crisis and the pressures they put on governments can be understood in terms of the classical political opportunity structure apparatus as signalling a moment of shifting alignments where parties might look for new allies, opening up opportunities for protest (Tarrow, 1994). Moreover, the importance of different types of political issues may wax and wane due to both more long-term processes (e.g. the rise of a new integrationdemarcation cleavage resulting from globalization) or the emergence of external shocks (e.g. the current economic crisis). The latter type of event in particular may serve as triggers for the opening up of political space for new challengers in both the electoral and protest arena and as such spur activism in society more generally. In particular, major crises are 9 understood to undermine support for mainstream political actors and therefore to open up the opportunities for protest (Grasso and Giugni, 2016). These types of shocks can be seen as catalyst events pushing political actors to react to wider circumstances and spurring wider societal mobilisation. This follows from the idea that crises can provide the space for party members to mobilise the wider public into protest action. As such we expect that: H5: The opening up of political opportunities linked to the crisis (i.e. high unemployment, low GDP growth) will narrow down the gap between party members and non-members ofparties in terms of their social movement involvement. At the same time, the way in which these opportunities are exploited will also be conditioned by the wider context such as economic set ups or national features such as social-democratic arrangements (Grasso and Giugni, 2016). On this basis, we expect that: H6: Contexts marked by social-democratic arrangements (e.g. higher social spending) provide a more open space for mobilising non-party activists. Data and methods To test our hypotheses questions, we employ data from an original cross-national survey (N=18,370) conducted in 2015 in the context of the [REMOVED FOR PEER REVIEW]. The survey was conducted in each of the nine European countries included in the project: France, Germany, Greece, Italy, Poland, Spain, Sweden, Switzerland and the UK by a specialised polling agency -YouGov - using online panels with the methodologies available in each country and quota balanced in order to match national population statistics in terms of region, sex, age, and education level (Grasso and Giugni 2016). These countries offer good variation in terms of political context and in terms of the extent to which they were 10 affected by the economic crisis, allowing for a certain degree of external validity of our results. We also include macro-level data from 2014 from the World Bank on unemployment and GDP growth as well as from the OECD on government social spending and tax wedges. Descriptive statistics for all the variables are available in Table 1. The final sample is 16,925 (Ns for each country are reported in Table 2), once all missing values are removed. INSERT TABLES 1-2 Our dependent variable is a dummy variable for whether someone engaged in protest in the last 12 months. Studies have shown that protest participation has distinct features to other types of political action often classed in the “unconventional” realm (Grasso, 2014) so we do not create scales but chose to focus on this “modal” expression (Tarrow, 1996) of social movement activism, also as exemplified by the anti-austerity demonstrations taking place in the wake of the crisis. Additionally, in line with our aim of wanting to capture the impact of the economic context in 2014 on participation in 2015, we only look at participation in the last year. The variable for party membership is also a dummy. Table 2 shows the proportion of individuals that said they protested in the last twelve months as well as that saying they are members of a party in each country. While the levels are relatively high, this is likely to be due to the fact that the measure in the survey asks people if they are members of a party so some individuals are likely to understand membership in less formalised terms, giving the higher levels (Morales, 2009). However, this is not effectively an issue for the analysis since we are interested in the relationship between party membership (even if some individuals might see themselves as 11 members of a party even when they do not have up to date membership cards) and protest participation. We include a control for subjective feelings of relative deprivation retrospective to the last five years that has been shown to be an important determinant of protest participation in times of crisis (Rüdig and Karyotis, 2013, Bernburg, 2015, Grasso and Giugni, 2016). The fact that the indicator is retrospective to the last five years is useful so the deterioration of conditions relative to expectations should have at least begun to occur prior to protest participation in the last 12 months and as such this means that the timeordering respects the requirements of causality. The question asks respondents whether they felt that the economic situation of their household was much better or much worse than five years ago. We dichotomised this measure following previous research (Rüdig and Karyotis, 2013) resulting in a dummy for whether individuals felt their household economic situation had deteriorated. Table 2 also shows the proportion of individuals who said the economic situation had deteriorated in each country. To capture the contextual aspects in our analysis we rely on four measures which have proved useful in a previous study (Grasso and Giugni 2016). Our macroeconomic variables aim to examine both negative and positive indicators of economic context. First, high unemployment levels can be seen as perhaps the most pernicious consequence of the current economic crisis in Europe. Countries such as Greece and Spain, where unemployment is highest, are those that have suffered the most. We also examine GDP growth as perhaps the clearest measure that a country is doing well and emerging from recession. Both variables are for 2014 to examine conditions prior to participation but not too long before. 12 On the policy side, we include both government expenses for social policies (as a percentage of the GDP) and the tax wedge (as a percentage of labour cost). They reflect a definition of austerity policies as reducing government spending, particularly in the social realm, and increasing taxation, predominantly on labour. This matters for our analysis since our survey was conducted in a period of economic crisis and austerity policies in Europe. These measures are meant to capture the output side of political opportunity structures a dimension generally neglected in the literature (Meyer and Minkoff, 2004, Meyer, 2004). Both variables are for 2014. We also include the usual socio-demographic controls in our models. These include age, gender, education level (low), occupation (8-categories, see Table 2) and employment status (whether the respondent is unemployed). We know that being unemployed is an important variable within biographical availability theories; while some unemployed people are resource-poor some are more resource-rich (Dunn et al., 2014). We also include controls for political attitudes and resources that are mainstay in the literature on participation and protest activism (Schussman and Soule, 2005): political interest, internal and external political efficacy, left-right values, libertarian-authoritarian values, and number of organisational memberships (distributions by country are provided in Table 2). Our dependent variable is measured at the individual level, but we have independent variables at both the individual and the country level. Moreover, we are interested in the interactions between these two levels since our argument refers to differences in how individual level party membership relates to individuals’ protest behaviour according to country-level economic and political context. For this reason, we specify multilevel models with random intercepts to take into account the two-level nature of the data (country and 13 individual). This type of model is useful to correct for the within-country dependence of observations (intraclass correlation). Since our dependent variable is dichotomous, we estimate logistic multilevel models with a Gaussian link function. Findings Table 3 shows ten models specified. Model 1 is the empty model. Model 2 includes the main individual-level independent variable measuring party membership and only the individual-level control variables. Model 3 includes unemployment rate in 2014; Model 4 GDP growth in 2014; Model 5 social spending in 2014; Model 6 tax wedge in 2014. Models 7-10 include, in turn, each of the four macro-level variables from Models 3-6 with the relevant cross-level interaction with party membership. INSERT TABLE 3 As we can see, model fit improves with the inclusion of the individual level variables as noted with the reduction in Log Likelihood. Moreover, there is improvement with the inclusion of the macro-level factors and the cross-level interactions, particularly unemployment in Model 3 and Model 7. Providing evidence for H1, we can see that across Models 1-6 (Models 7-10 include cross-level interactions) party membership has a positive effect on protest participation. Model 2, which includes all the individual level controls, shows no negative effect for low education level. Fitting in with expectation, relative deprivation has a significant and positive effect across the models. In terms of social bases, we find that managers and foremen are significantly less likely relative to socio-cultural specialists to be involved in social movement activities. On the other hand, unemployed status has a negative and significant effect across the models. Thus, at the individual level 14 there is little evidence for a rising precariat at least as captured by this potentially narrow measure and also against the predictions of some biographical availability theories: unemployment reduces the chances of protest. In line with the resource and civic voluntarism model (Verba et al., 1995), greater political interest, stronger internal and external efficacy (the scale is negative so the effect of external efficacy is also positive) all have positive effects. Moreover, for H2, being more left-wing (relative to right-wing) and also being more libertarian (relative to more authoritarian) all have a significant and positive effect on protest but the effect of party membership remains significant, suggesting that party members do not purely engage in protest for ideological reasons. With respect to H3, we also see that the density of organisational networks also has a strong positive effect, but the effect of party membership remains strong and significant event when controlling for all the relevant predictors and particularly network and ideological variables. This suggests that party members engage in social movement activities for a whole series of ideational and structural, but also strategic, reasons, which we identified with the residual effect of party membership on protest once all other variables are controlled for. Next, we turn to considering the macro-level economic factors. When unemployment is included in Model 3, there is a positive and significant effect of this macroeconomic context variable on protest. Individuals in countries with higher unemployment are more likely to have demonstrated in the last 12 months. However, the inclusion of this variable in Model 3 does not remove the individual-level effect of party membership we saw before. Party members are still more likely to become involved in social movement activities regardless of whether individuals are in countries with higher or lower levels of unemployment. 15 When including GDP growth in Model 4, there is a negative and significant effect of this economic context variable on demonstrating. Individuals in countries with lower GDP growth are more likely to have protested in the last 12 months. The inclusion of this macro-level variable in Model 4 does not remove the individual-level effect for party membership found before. Party members are still more likely to protest regardless of GDP growth levels in the countries. With respect to the macro-level political factors, a similar pattern as with economic context is observed. Social spending in Model 5 has a significant positive effect at the 10 percent level on demonstration activities. We interpret this as a sign of more open political opportunity structures. The inclusion of this macro-level political factor does not remove the effect of party membership, meaning that the latter still matters regardless of spending levels in the countries where respondents are living. When including tax wedge in Model 6 there is also a significant (at the 10 percent level) effect. This finding supports Bermeo and Bartels’ (2014) hypothesis that people react to austerity policies rather than directly to the negative effects of the crisis. However, the inclusion of this macro-level factor also does not change the effect of party membership: the latter plays a role regardless of this contextual control. One of the key aims of this this research, developing on previous work, was to allow us to examine the individual and macro-level perspectives on protest in times of crisis combined. Models 7-10 include cross-level interactions between each of macro-level variable and party membership. With respect to H5, Model 7 shows that in countries with higher unemployment rates, the effect of party membership on protest participation is greater. Therefore, in worse economic conditions the protest gap between party members 16 and non-members becomes smaller. As such, it seems that more open political opportunities in the wider national economic context spur the mobilisation of wider sections of society beyond those individuals that are already committed to political parties. A negative economic context may thus be understood in terms of increasing the chances that members of the public will become mobilised to protest. Thus, while party membership has a positive effect on protest participation regardless of the economic context, at higher levels of unemployment the effect of being a committed party member is lower relative to contexts with lower unemployment. This therefore suggests that this type of contextual factor may serve to politicise individuals that do not participate outside of periods of crisis and therefore that crises may be seen to open up opportunities for mobilisation. Supporting H5, the results for GDP growth (Model 8) provide a similar narrative in that where economic growth is slower, differences between party members and nonmembers become smaller, whereas when GDP growth is higher the gap between party activists and others is larger. This conditional effect is also found for the political context with respect to the political opportunity structure offered by social democracies i.e. welfare states, when testing for H6 in Model 9. However, the tax wedge (Model 10), while it has a direct effect on demonstrating, does not condition the effect of party membership. Moving on we can see from Model 7 with the cross-level interaction for party membership and unemployment that the coefficient for party membership is 0.79 and significant. This means that when the unemployment rate is 0 party members are more likely than non-members to protest. The negative and statistically significant coefficient for the interaction term between party membership and unemployment (-0.03) furthermore 17 suggests that the gap between non-members and party members becomes smaller as unemployment goes up. For every percentage point rise in unemployment rate, the gap in the log-odds of protesting falls by 0.03. The estimates suggest that party members catch up with party members when unemployment reaches a level of 26.3% (0.79 /-0.03=26.3). As such, it would seem that particularly negative economic contexts can act to mobilise the wider members of the public, opening up opportunities for social change. In Model 8 we can see that the coefficient for party membership suggests that when GDP growth is 0 there are no differences between party members and non-members in protesting. The positive and statistically significant coefficient for the interaction term between party membership and unemployment (0.30) suggests that the gap between party members and non-members rises as GDP growth does. For every percentage point increase in unemployment rate, the gap in the log-odds of protesting rises by 0.30. These estimates suggest that even very slow economic growth is enough to differentiate the two groups (0.02/0.30=0.06). Thus, it looks as though improved economic conditions have a demobilising effect for everyone, but this is particularly true of the wider population relative to party members. Again, this would support the idea that crises are conducive to opportunities for social movements to mobilise the wider public. In Model 9, the coefficient for party membership shows that when social spending is very low only party members protest. This supports the idea that in contexts with more closed opportunity structures, only the committed engage. The negative and statistically significant coefficient for the interaction term between party membership and social spending (-0.04) suggests that the gap in protesting between members and non-members falls at higher levels of social spending, where opportunities for protest are presumably 18 perceived as more open, since the wider context is more favourable to demonstrating and social movements. For every percentage point increase in social spending, the gap in the log-odds of protesting between the two groups falls by 0.04. We can see that non-members start protesting at higher levels than party members when social spending is particularly generous i.e. over 38.25% of GDP (1.53/-0.04=38.25). This shows that party members are more active than non-members in social movement activities but that more favourable political contexts can narrow the gap in participation between these groups, once more suggesting that context can allow for wider opportunities for societal-level mobilisation and chances of movement success. Figures 1, 2 and 3 plot the cross-level interaction effects between party membership and, in turn, unemployment, GDP growth and social spending to allow for visualisation. The graphs show that the effect of party membership for mobilisation is greater where there are more closed political opportunities. Political opportunity theory allows us to interpret these findings. We consider that economic crisis and thus high unemployment or slow GDP growth open up the political space for social problems to be understood at a more collective level. In turn we suggest that this can then form the basis for political mobilisation even of those members of the public that are less politically committed i.e. non-members. We see a similar process operating for higher social spending in that here we consider these contexts as more open opportunity structures with respect to protest. Where social spending is lower we expect there to be greater neoliberal approaches including aspects of individualisation of poverty being more normalised creating greater barriers to social movement involvement. INSERT FIGURES 1-3 19 Discussion and conclusion We hope that our study has served to show how the individual level relationship between protest and parties is contextualised within wider political opportunities. We show that party members are more likely to take to the streets even when controlling for relevant characteristics that would be understood to account for various ideational and structural factors, suggesting that strategic reasons, which we associate with the residual effect, are also important. This suggests that party members act strategically and support social movements where they think that this will benefit their party. This shows that the institutional and extra-institutional domains are clearly linked and that activists tend to participate in both. More so, committed institutional activism is found here to be reinforcing of extra-institutional participation. Furthermore, in linking the individual level to the wider macro or economic and political context, we find that not only does party membership have direct effect on the propensity to have engaged in demonstrations in the last year, but that party members are also more likely than non-members to engage in social movement activism regardless of more open or closed political opportunities. More open political opportunities in the context of the crisis or social-democratic arrangements were further found to narrow the gap between members and the wider public in their likelihood to protest. We interpreted these findings with respect to political opportunities for protest. Political opportunity theory tends to emphasise contextual variables for understanding mobilisation. Our results for the crosslevel interactions between party membership and macro-level factors show that context also conditions the extent of the effect of party membership on participation. Moreover, strong commitment to a political cause such as would be signified by party membership was 20 shown to become less determinant of engagement when political opportunities are more open, allowing for the mobilisation of non-party members. In this way, we showed that political opportunities are important for our understanding political action both in terms of their impact on the individual level mobilisation of party members but also for the wider opportunities for mobilisation beyond them. We showed how their dynamic interaction can explain differential protest behaviour. We showed how higher unemployment reduces the gap between party members and nonmembers so that crisis may allow for wider opportunities for mobilisation. We understood in this respect higher levels of unemployment as providing a context where politicisation and protest against the government is more likely to occur (Grasso and Giugni 2016). Unemployment can be seen as a wider social ill affecting society and leading individuals to see it as a shared collective social problem and thus supporting mobilisation processes. A similar argument could be made for economic growth in that slower growth contexts have similar mobilising impacts on non-members. This suggests that the wider environment and worsening economic contexts could allow for the development of collective action. We also showed differences in social movement mobilisation between party members and non-members is reduced where we find greater welfare. We understand these contexts has having more open political opportunities leading also non-members to engage. This suggests that party members may suffer from an “optimistic bias” (Gamson and Meyer, 1996: 286) when it comes to political opportunities where they “just keep trying and sometimes succeed in engaging a broader public” (Meyer and Minkoff, 2004: 1464). In the words of Meyer (2004: 104), party members are “consistent champions” whereas the wider public are more akin to the model of “strategic respondents.” 21 More widely, we have shown how those situated within social democratic arrangements are more likely to react politically, but that everyone is more likely to engage in more open opportunity contexts. These results support the idea that welfare state provisions and citizenship rights can be critical resources for groups organising for collective action (Grasso and Giugni 2016). Case studies have shown the mechanism relates to specific movements e.g. unemployed mobilisations are linked to unemployment benefits (Giugni, 2008), immigrants’ mobilisations are linked to the type and level of citizenship rights (Giugni and Passy, 2004, Koopmans et al., 2005). Here we can show that more generally higher levels of social spending, encourage mobilisation. Finally, our study goes to show that analysing macro-level economic and political context alongside ideational, structural and strategic factors is important for understanding the link between party membership and protest. Studies have shown how mobilisation is more likely when economic conditions deteriorate (Baglioni et al., 2008, Piven and Cloward, 1977) and we have supported this by showing that moving from lower to higher levels of GDP growth demobilises non-members at a faster rate than it does for party members. We also show how in times of crisis, members and non-members becomes more similar in their protest behaviour, thus suggesting movements should capitalise on these contexts to aim to mobilise the wider public beyond party activists. We thus show the extent of mobilisation depends on political membership and commitment but also on the wider environment and the framing of opportunities (Gamson and Meyer, 1996). We therefore argue that widening of political opportunities brought about by crisis can mobilise wider sectors of the public, including those that are not ordinarily members of political parties. 22 Table 1: Variable descriptive statistics mean sd min max Demonstrated Party membership Relative Deprivation Age Male Education (low) 0.11 0.12 0.45 45.35 0.48 0.24 0.31 0.33 0.50 14.53 0.50 0.43 0 0 0 18 0 0 1 1 1 88 1 1 Class Unemployed Political interest Internal political efficacy External political efficacy Left-right Libertarian-authoritarian Memberships 4.00 0.11 0.65 0.50 0.48 5.25 4.47 1.24 2.27 0.31 0.48 0.40 0.36 1.85 1.88 2.35 1 0 0 0 0 0 0 0 8 1 1 1 1 10 10 12 Unemployment GDP growth Social Spending Tax wedge 11.79 1.55 25.14 39.70 7.72 1.12 3.89 8.51 4.5 -0.4 19.4 22.3 26.5 3.4 31.9 49.3 N 16925 23 Table 2: Variable distributions, by country N Protest participation (%) Party membership (%) Relative deprivation (%) Age (mean) Male (%) Unemployed (%) Occupation (%) Socio-cultural specialists Managers Clerical Routine non-manual Foremen and supervisors Skilled manual Semi/Unskilled manual Other Education (low) (%) Political interest (%) Internal political efficacy 0-1 (mean) External political efficacy 0-1 (mean) Left-right values 0-10 (mean) Libertarian-authoritarian 0-10 (mean) Organisational memberships (0-12) (mean) Unemployment rate 2014 (%) GDP growth 2014 (%) Social spending 2014 (%) Tax wedge 2014 (%) All 16,925 11.1 12.3 45.4 44.8 47.2 11.7 France 1,861 14.2 9.0 52.6 48.7 44.1 9.1 Germany 1,905 8.4 9.3 27.3 44.2 51.9 4.0 Greece 1,916 23.2 14.7 84.6 39.9 46.7 27.5 Italy 1,766 12.0 17.6 55.7 44.2 47.5 17.2 Poland 1,909 6.1 7.0 41.8 41.5 44.5 11.5 Spain 1,852 17.8 11.5 54.3 43.0 50.1 18.8 Sweden 1,869 7.3 16.6 22.5 47.2 46.0 5.4 Switz. 1,936 6.0 11.7 33.0 43.8 47.2 6.5 UK 1,911 4.3 13.8 34.8 51.2 46.8 4.5 16.6 10.8 24.0 14.4 4.6 8.0 12.3 9.4 24.1 64.3 .49 .48 5.2 4.5 1.3 11.8 1.5 25.2 39.8 12.6 8.4 26.4 14.5 7.2 9.4 8.9 12.7 28.9 56.3 .39 .40 5.5 4.8 1.0 10.3 0.2 31.9 48.4 10.6 15.1 30.1 15.2 4.4 7.9 9.3 7.5 17.5 71.0 .59 .48 5.2 4.0 .7 5.0 1.6 25.8 49.3 22.7 10.0 17.6 19.2 3.6 4.7 12.3 10.0 13.4 62.9 .49 .35 4.8 4.7 1.6 26.5 0.8 24 40.4 15.4 6.7 30.7 10.1 3.6 8.8 13.0 11.7 32.8 59.9 .48 .57 5.4 4.8 2.0 12.7 -0.4 28.6 48.2 15.1 7.3 19.9 14.5 6.2 7.1 19.6 10.3 15.0 73.7 .51 .65 5.0 5.4 1.1 9.0 3.4 20.6 35.6 16.2 9.9 25.4 14.1 4.2 9.2 12.5 8.5 38.5 57.7 .45 .49 4.8 3.7 1.2 24.4 1.4 26.8 40.7 18.7 8.6 20.7 14.1 3.9 4.9 19.2 10.0 26.8 63.9 .40 .46 5.5 3.7 1.6 8.0 2.3 28.1 42.5 11.2 13.7 24.9 17.3 5.1 12.9 6.5 8.5 18.2 58.4 .48 .44 5.4 4.5 1.3 4.5 1.9 19.4 22.3 27.1 16.9 20.5 10.6 2.9 6.7 9.7 5.8 25.8 75.5 .66 .49 5.7 4.5 .8 6.1 2.6 21.7 31.1 24 Table 3: Multi-level logistic regression models predicting protest participation (last 12 months) (2) 0.41*** (0.08) (3) 0.41*** (0.08) (4) 0.41*** (0.08) (5) 0.41*** (0.08) (6) 0.41*** (0.08) (7) 0.79*** (0.14) (8) 0.02 (0.11) (9) 1.53*** (0.47) (10) 0.98** (0.35) 0.19** (0.06) 0.19** (0.06) 0.19** (0.06) 0.19** (0.06) 0.19** (0.06) 0.18** (0.06) 0.18** (0.06) 0.19** (0.06) 0.19** (0.06) Age -0.01*** (0.00) -0.01** (0.00) -0.01*** (0.00) -0.01*** (0.00) -0.01*** (0.00) -0.01** (0.00) -0.01** (0.00) -0.01*** (0.00) -0.01*** (0.00) Male 0.01 (0.06) 0.01 (0.06) 0.01 (0.06) 0.01 (0.06) 0.01 (0.06) 0.00 (0.06) 0.01 (0.06) 0.01 (0.06) 0.01 (0.06) Unemployed -0.21* (0.08) -0.21* (0.08) -0.21* (0.08) -0.21* (0.08) -0.21* (0.08) -0.21* (0.08) -0.21* (0.08) -0.21* (0.08) -0.21* (0.08) 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 (.) (.) (.) (.) (.) (.) (.) (.) (.) 2.Managers -0.23* (0.10) -0.22* (0.10) -0.23* (0.10) -0.23* (0.10) -0.23* (0.10) -0.23* (0.10) -0.23* (0.10) -0.23* (0.10) -0.23* (0.10) 3.Clerical -0.10 (0.08) -0.10 (0.08) -0.10 (0.08) -0.10 (0.08) -0.10 (0.08) -0.10 (0.08) -0.11 (0.08) -0.10 (0.08) -0.10 (0.08) 4.Routine non-manual -0.21* (0.10) -0.20* (0.10) -0.21* (0.10) -0.21* (0.10) -0.21* (0.10) -0.20* (0.10) -0.21* (0.10) -0.21* (0.10) -0.21* (0.10) 5.Foremen & supervisors -0.15 -0.15 -0.15 -0.15 -0.15 -0.14 -0.15 -0.15 -0.15 (0.14) (0.14) (0.14) (0.14) (0.14) (0.14) (0.14) (0.14) (0.14) -0.20 -0.19 -0.20 -0.20 -0.20 -0.19 -0.19 -0.19 -0.20 Party Membership Relative Deprivation 1.Socio-cultural specialists (Ref) 6.Skilled manual (1) 25 (0.12) (0.12) (0.12) (0.12) (0.12) (0.12) (0.12) (0.12) (0.12) 7.Semi/unskilled manual 0.03 (0.10) 0.03 (0.10) 0.03 (0.10) 0.03 (0.10) 0.03 (0.10) 0.04 (0.10) 0.04 (0.10) 0.04 (0.10) 0.03 (0.10) 8.Other -0.11 (0.11) -0.10 (0.11) -0.11 (0.11) -0.11 (0.11) -0.11 (0.11) -0.10 (0.11) -0.11 (0.11) -0.11 (0.11) -0.11 (0.11) Education (low) -0.05 (0.07) -0.05 (0.07) -0.05 (0.07) -0.05 (0.07) -0.05 (0.07) -0.05 (0.07) -0.05 (0.07) -0.05 (0.07) -0.05 (0.07) Political interest 0.73*** (0.07) 0.73*** (0.07) 0.74*** (0.07) 0.73*** (0.07) 0.73*** (0.07) 0.74*** (0.07) 0.74*** (0.07) 0.74*** (0.07) 0.73*** (0.07) Internal political efficacy 0.73*** 0.73*** 0.73*** 0.73*** 0.73*** 0.73*** 0.73*** 0.74*** 0.73*** (0.08) (0.08) (0.08) (0.08) (0.08) (0.08) (0.08) (0.08) (0.08) External political efficacy -0.16 -0.16 -0.16 -0.16 -0.16 -0.15 -0.14 -0.15 -0.16 (0.08) (0.08) (0.08) (0.08) (0.08) (0.08) (0.08) (0.08) (0.08) Left-right -0.17*** (0.02) -0.17*** (0.02) -0.17*** (0.02) -0.17*** (0.02) -0.17*** (0.02) -0.17*** (0.02) -0.17*** (0.02) -0.18*** (0.02) -0.17*** (0.02) Libertarian-authoritarian -0.18*** (0.02) -0.18*** (0.02) -0.18*** (0.02) -0.18*** (0.02) -0.18*** (0.02) -0.18*** (0.02) -0.18*** (0.02) -0.18*** (0.02) -0.18*** (0.02) Memberships 0.14*** (0.01) 0.14*** (0.01) 0.14*** (0.01) 0.14*** (0.01) 0.14*** (0.01) 0.14*** (0.01) 0.14*** (0.01) 0.14*** (0.01) 0.14*** (0.01) Unemployment GDP growth 0.07*** (0.02) 0.07*** (0.02) -0.40** (0.14) -0.48*** (0.14) Social spending 0.09+ (0.05) 26 0.10* (0.05) 0.04+ (0.02) Tax wedge Cross-level interactions Party membership X Unemployment 0.04 (0.02) -0.03*** (0.01) Party membership X GDP growth 0.30*** (0.06) -0.04* Party membership X Social spending (0.02) -0.01 Party membership X Tax Wedge Constant N Log lik. AIC BIC Sigma u Rho (0.01) -2.22*** (0.20) 16925 -5599.58 11203.15 11218.62 0.59 0.10 -1.49*** (0.27) 16925 -4861.16 9764.33 9926.79 0.65 0.11 Standard errors in parentheses + p ≤ 0.10,* p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001 -2.28*** (0.30) 16925 -4856.91 9757.83 9928.03 0.40 0.05 -0.87** (0.31) 16925 -4858.27 9760.55 9930.75 0.47 0.06 -3.76** (1.22) 16925 -4859.64 9763.28 9933.48 0.55 0.08 27 -3.02** (0.92) 16925 -4859.88 9763.76 9933.97 0.56 0.09 -2.39*** (0.30) 16925 -4850.98 9747.96 9925.90 0.41 0.05 -0.80* (0.31) 16925 -4845.64 9737.28 9915.22 0.47 0.06 -4.03** (1.23) 16925 -4856.65 9759.30 9937.24 0.55 0.08 -3.17*** (0.93) 16925 -4858.53 9763.06 9941.00 0.56 0.09 Figure 1: Plot of the cross-level interaction between party membership and unemployment (adjusted predictions Model 7) Predicted Mean, Fixed Portion Only .05 .1 .15 .2 Adjusted Predictions 4.5 7 9.5 12 14.5 17 unemployment partym=0 28 19.5 partym=1 22 24.5 Figure 2: Plot of the cross-level interaction between party membership and GDP growth (adjusted predictions Model 8) 0 Predicted Mean, Fixed Portion Only .05 .1 .15 Adjusted Predictions -.4 .1 .6 1.1 1.6 gdpgrowth partym=0 29 2.1 partym=1 2.6 3.1 Figure 3: Plot of the cross-level interaction between party membership and social spending (adjusted predictions Model 9) .04 Predicted Mean, Fixed Portion Only .06 .08 .1 .12 .14 Adjusted Predictions 19 20 21 22 23 24 25 26 sspending partym=0 30 27 28 29 partym=1 30 31 32 References Baglioni, S., Baumgarten, B., Chabanet, D. & Lahusen, C. (2008) Transcending Marginalization: The Mobilization of the Unemployed in France, Germany, and Italy in a Comparative Perspective. Mobilization 13, 323–35. Bermeo, N. & Bartels, L. (2014) Mass Politics in Tough Times. IN Bermeo, N. & Bartels, L. (Eds.) Mass Politics in Tough Times: Opinions, Votes and Protest in the Great Recession. Oxford, Oxford University Press. Bernburg, J. G. (2015) Economic Crisis and Popular Protest in Iceland, January 2009: The Role of Perceived Economic Loss and Political Attitudes in Protest Participation and Support. Mobilization, 20, 231-252. Císa , O. & Navrátil, J. (2015) At the Ballot Boxes or in the Streets and Factories: Economic Contention in the Visegrad Group. IN Giugni, M. & Grasso, M. (Eds.) Austerity and Protest: Popular Contention in Times of Economic Crisis. London, Routledge. Dalton, R. (2008) Citizen politics: Public opinion and political parties in advanced industrial democracies, Washington, DC, CQ Press. della Porta, D. (2015) Social Movements in Times of Austerity: Bringing Capitalism Back into Protest Analysis, Cambridge, Polity Press. della Porta, D., Fernández, J., Kouki, H. & Mosca, L. (2017) Movement Parties against Austerity, Cambridge, Polity Press. Dunn, A., Grasso, M. T. & Saunders, C. (2014) Unemployment and Attitudes to Work: asking the ‘right’ question. Work, Employment, and Society, 28, 904925. Gamson, W. & Meyer, D. (1996) Framing political opportunity. IN McAdam, D., McCarthy, J. & Zald, M. N. (Eds.) Comparative Perspectives on Social Movements: Political Opportunities, Mobilising Structures and Cultural Framing. Cambridge, Cambridge University Press. Giugni, M. (2008) Welfare States, Political Opportunities, and the Mobilization of the Unemployed: A Cross-national Analysis. Mobilization, 13, 297–310. Giugni, M. and Grasso, M.T. (eds) (2018) Citizens and the Crisis: Perceptions, Experiences, and Responses to the Great Recession in Europe.London: Palgrave Macmillan (Palgrave Studies in European Political Sociology) Giugni, M. & Passy, F. (2004) Migrant Mobilization between Political Institutions and Citizenship Regimes: A Comparison of France and Switzerland. European Journal of Political Research, 43, 51–82. Goldstone, J. (1991) Revolution and Rebellion in the Early Modern World, Berkeley, University of California Press. Grasso, M. T. (2014) Age-period-cohort analysis in a comparative context: Political generations and political participation repertoires. Electoral Studies 33 63-76. Grasso, M. T. & Giugni, M. (2016) Protest participation and economic crisis: The conditioning role of political opportunities. European Journal of Political Research, 55, 663-80. 31 Heaney, M. (2013) Elections and Social Movements. IN Snow, D., della Porta, D., Klandermans, B. & McAdam, D. (Eds.) The Wiley-Blackwell Encyclopedia of Social and Political Movements. Oxford, Blackwell. Heaney, M. T. & Rojas, F. (2015) Party in the Street: The Antiwar Movement and the Democratic Party after 9/11, Cambridge, Cambridge University Press. Kitschelt, H. (1986) Political Opportunity Structures and Political Protest: AntiNuclear Movements in Four Democracies. British Journal of Political Science, 16, 57-85. Kitschelt, H. (1988) Left-Libertarian Parties: Explaining Innovation in Competitive Party Systems. World Politics, 40, 194-234. Koopmans, R. & Olzak, S. (2004) Discursive Opportunities and the Evolution of Right-Wing Violence in Germany. American Journal of Sociology, 110, 198230. Koopmans, R., Statham, P., Giugni, M. & Passy, F. (2005) Contested Citizenship: Immigration and Cultural Diversity in Europe, Minneapolis, University of Minnesota Press. Kriesi, H. (1989) New Social Movements and the New Class in the Netherlands. American Journal of Sociology 94, 1078-1116. Kriesi, H. (2014) The Political Consequences of the Economic Crises in Europe: Electoral Punishment and Popular Protest. IN Bermeo, N. & Bartels, L. M. (Eds.) Mass Politics in Tough Times: Opinions, Votes and Protest in the Great Recession. Oxford, Oxford University Press. Kriesi, H., Grande, E., Dolezal, M., Helbling, M., Hoglinger, D., Hutter, S. & Wuest, B. (2012) Political Conflict in Western Europe, Cambridge, Cambridge University Press. Kriesi, H., Koopmans, R., Duyvendak, J. W. & Giugni, M. (1992) New Social Movements and Political Opportunities in Western Europe. European Journal of Political Research, 22. Kriesi, H., Koopmans, R., Duyvendak, J. W. & Giugni, M. (1995) New Social Movements in Western Europe: A Comparative Analysis, Minneapolis, University of Minnesota Press. Kriesi, H. & Wisler, D. (1996) Social Movements and Direct Democracy in Switzerland. European Journal of Political Research 30, 19-40. McAdam, D. & Tarrow, S. (2010) Ballots and Barricades: On the Reciprocal Relationship between Elections and Social Movements. Perspectives on Politics, 8, 529-42. McAdam, D. & Tarrow, S. (2013) Social Movements and Elections: Towards a Better Understanding of the Political Context of Contention. IN Roggeband, C., Klandermans, B. & Van Stekelenburg, J. (Eds.) The Changing Dynamics of Contention Minneapolis, University of Minnesota Press. Meyer, D. & Minkoff, D. (2004) Conceptualising Political Opportunity. Social Forces, 82, 1457-1492. Meyer, D. S. (2004) Protest and political opportunities. Annual Review of Sociology 30, 125-45. 32 Morales, L. (2009) Joining political organisations: institutions, mobilisation and participation in Western democracies, Colchester, Essex, ECPR Press. Norris, P. (2002) Democratic Phoenix. Reinventing Political Activism, Cambridge, MA, Cambridge University Press. Norris, P., Frank, R. & Martínez i Coma, F. (Eds.) (2015) Contentious Elections: From Ballots to Barricades, New York Routledge. Piccio, D. (2016) The impact of social movements on political parties. IN Bosi, L., Giugni, M. & Uba, K. (Eds.) The Consequences of Social Movements. Cambridge, CUP. Piven, F. F. & Cloward, R. (1977) Poor Peoples’ Movements, New York, Vintage Books. Rüdig, W. & Karyotis, G. (2013) Who Protests in Greece? Mass Opposition to Austerity. British Journal of Political Science, Available on CJO 2013 doi:10.1017/S0007123413000112. Schussman, A. & Soule, S. (2005) Process and Protest: Accounting for Individual Protest Participation Social Forces, 84, 1083-1108. Scott, J. C. (1976) The moral economy of the peasant: rebellion and subsistence in Southeast Asia, Yale University Press. Standing, G. (2011) The Precariat: The New Dangerous Class, London, Bloomsbury. Tarrow, S. (1994) Power in movement, Cambridge, Cambridge University Press. Tarrow, S. (1996) States and opportunities: The political structuring of social movements. IN McAdam, D., McCarthy, J. & Zald, M. N. (Eds.) Comparative Perspectives on Social Movements: Political Opportunities, Mobilizing Structures, and Cultural Framings,. Cambridge, Cambridge University Press. Tilly, C. (1978) From Mobilization to Revolution, Reading, MA, Addison-Wesley. . Tilly, C. (1999) From interactions to outcomes in social movements. IN Giugni, M., McAdam, D. & Tilly, C. (Eds.) How Social Movements Matter. Minneapolis, University of Minnesota Press. Verba, S., Schlozman, K. L. & Brady, H. (1995) Voice and Equality: Civic Voluntarism in American Politics, Cambridge, MA, Harvard University Press. 33