THE AUSTERITY EFFECT. PARTY ACTIVISM IN EUROPE DURING THE GREAT RECESSION

1. INTRODUCTION

A range of party politics literature has focused on party membership and discussed its evolution over time (Mair and van Biezen 2001Mair, P. and van Biezen, I. 2001. ‘Party membership in twenty European democracies, 1980-2000’, Party Politics, 7(1), pp. 5-21. https://doi.org/10.1177/1354068801007001001 ; Scarrow and Gezgor 2010Scarrow, S. E. and Gezgor, B. 2010. ‘Declining memberships, changing members? european political party members in a new era’, Party Politics, 16(6), pp. 823-843. https://doi.org/10.1177/1354068809346078 ; van Biezen et al 2012Biezen, I. van, Mair, P., Poguntke, T. 2012. ‘Going, going,...gone? The decline of party membership in contemporary Europe’, European Journal of Political Research, 51(1), pp. 24-56. https://doi.org/10.1111/j.1475-6765.2011.01995.x ; Ignazi 2017Ignazi, P. 2017. Party and democracy : the uneven road to party legitimacy. Oxford: Oxford University Press.). Similarly, another strand has focused on explaining party activism, generally conceived as the involved layer of members who not only pay their fees and express active support to the party as ‘ambassadors in the community’ (Scarrow 1996Scarrow, S. E. 1996. Parties and their members. Organizing for victory in Britain and Germany. Oxford: Oxford University Press. ) but also assume some of the labour and internal tasks necessary for the party’s existence (Whiteley and Seyd 2002Whiteley, P. F. and Seyd, P. 2002. High-Intensity participation: The dynamics of party activism in Britain. Ann Arbor, MI: The University of Michigan Press.; Bale, Webb and Poletti 2020Bale, T., Webb, P. D. and Poletti, M. 2020. Footsoldiers : political party membership in the 21st century. Abingdon and New York: Routledge.; Demker, Heidar and Kosiara-Pedersen 2020Demker, M., Heidar, K. and Kosiara-Pedersen, K. 2020. Nordic party pembers. Linkages in troubled times. London & New York: Rowman & Littlefield International / ECPR Press.).

The relationship between party membership and party activism has been less explored by the literature (Duverger 1954Duverger, M. 1954. Political parties: Their organization and activity in the modern state. London: Methuen.; Heidar 1994Heidar, K. 1994. ‘The polymorphic nature of party membership’, European Journal of Political Research, 25(1), pp. 61-86. https://doi.org/10.1111/j.1475-6765.1994.tb01201.x , 2006Heidar, K. 2006. ‘Party membership and participation’. Pp. 301-315. in Handbook of party politics, edited by R.S. Katz and W. J. Crotty. London: Sage Publications.; van Haute 2009van Haute, E. 2009. Adhérer à un parti. Aux sources de la participation politique. Brussels: Editions de l’Université de Bruxelles.; Scarrow 2014Scarrow, S. E. 2014. Beyond party members: changing approaches to partisan mobilization. Oxford: Oxford University Press.). While parties have accepted the importance of inactive members that never intended to participate in party activities (Scarrow 1996: 146Scarrow, S. E. 1996. Parties and their members. Organizing for victory in Britain and Germany. Oxford: Oxford University Press. ), those members prone to adopting a higher intensity in their involvement have gained relevance in parallel to recent transformations of political parties (Whiteley and Seyd 2002Whiteley, P. F. and Seyd, P. 2002. High-Intensity participation: The dynamics of party activism in Britain. Ann Arbor, MI: The University of Michigan Press.). This suggests that party membership and party activism might have followed different trends: membership might have been in decline, but activism could have remained steady or even increased (Scarrow 2000Scarrow, S. E. 2000. ‘Parties without members? Party organization in a changing electoral environement’, in Dalton, R. J. and Wattenberg, M. P. (eds) Parties without partisans: Political change in advanced industrial democracies. Oxford: Oxford University Press, pp. 79-101.; Heidar 2006Heidar, K. 2006. ‘Party membership and participation’. Pp. 301-315. in Handbook of party politics, edited by R.S. Katz and W. J. Crotty. London: Sage Publications.). Not only that, political parties have started to embrace new kinds of affiliates, even if they do not necessarily establish traditional formal linkages with the organisation as they used to in the past (van Haute 2009van Haute, E. 2009. Adhérer à un parti. Aux sources de la participation politique. Brussels: Editions de l’Université de Bruxelles.; Scarrow 2014Scarrow, S. E. 2014. Beyond party members: changing approaches to partisan mobilization. Oxford: Oxford University Press.). This suggests a wider variation in party activism than what has been captured in party research so far. Indeed, the blurring of the boundaries between sympathisers and new types of non-formal members is paramount at a time when parties attempt to open their machinery to the wide grassroots membership through intra-party democratic innovations (Young 2013: 65).

The aim of this article is twofold. On the one hand, it provides a new typology to help conceptualise different forms of party affiliation in political parties and presents some preliminary evidence of their patterns over time. On the other hand, we also aim to observe whether such evolution, particularly in terms of party activism, has been affected by contextual factors such as the Great Recession. Recent research suggests an important demobilisation effect on voluntary associationism (Cameron 2021Cameron, S. 2021. ‘Civic engagement in times of economic crisis: A cross-national comparative study of voluntary association membership’, European Political Science Review, pp. 1-19. https://doi.org/10.1017/S1755773921000060 ) but to the best of our knowledge this has barely been discussed in relation to party activism. Using data from the European Social Survey rounds 1 to 9 (2002-2018), this article explores the effect of some political, social and economic variables aiming to test the consequences for party activism of the austerity policies implemented in some European countries.

The article is organised as follows. The second section depicts a general overview of contemporary party membership and activism research in Europe. The third section discusses the conceptualisation of party membership and party activism and suggests a new typology of party affiliation that helps to disentangle both concepts. This section also reviews the main explanations of party activism and points out the need to assess the relevance of the socio-economic context as one of its drivers. The fourth section describes our research design. The following two empirical sections describe evidence of the different modes of party affiliation in contemporary European democracies between 2002 and 2010, and test the effect of austerity policies and other drivers on party activism. Finally, we present our conclusions and indicate avenues for future research.

2. PARTY MEMBERSHIP AND ACTIVISM RESEARCH IN 21^ST CENTURY EUROPE

The comparative literature on political parties has intensively studied changes in the levels of party membership in Europe between the late 1980s and the 2010s based on party registers (Widfeldt 1995Widfeldt 1995. ‘Party membership and party representatives.’ Pp. 134-182 in Citizens and the State, edited by H.-D. Klingemann and D. Fuchs. Oxford: Oxford University Press, pp.134-182.; Mair and van Biezen 2001Mair, P. and van Biezen, I. 2001. ‘Party membership in twenty European democracies, 1980-2000’, Party Politics, 7(1), pp. 5-21. https://doi.org/10.1177/1354068801007001001 ; Scarrow and Gezgor 2010Scarrow, S. E. and Gezgor, B. 2010. ‘Declining memberships, changing members? european political party members in a new era’, Party Politics, 16(6), pp. 823-843. https://doi.org/10.1177/1354068809346078 ; Biezen et al. 2012Biezen, I. van, Mair, P., Poguntke, T. 2012. ‘Going, going,...gone? The decline of party membership in contemporary Europe’, European Journal of Political Research, 51(1), pp. 24-56. https://doi.org/10.1111/j.1475-6765.2011.01995.x ; Krouwel 2012Krouwel, A. 2012. Party Transformations in European Democracies. New York: State University of New York Press.). While a declining trend on party membership seems to be constant over time, it has not been homogeneously reproduced in all European countries and certainly not for all party families. A similar picture emerged from the scattered comparative data on population surveys (Schmitt and Holmberg 1995Schmitt, H. and Holmberg, S. 1995. ‘Political parties in decline?’. Pp. 95-133 in Citizens and the State, edited by H. -D. Klingemann and D. Fuchs. Oxford: Oxford University Press.; Scarrow and Gezgor 2010Scarrow, S. E. and Gezgor, B. 2010. ‘Declining memberships, changing members? european political party members in a new era’, Party Politics, 16(6), pp. 823-843. https://doi.org/10.1177/1354068809346078 ). More recent and fine-grained accounts of party membership figures have also reported that the phenomenon is not as universal as previously thought, and tends to be strongly related to countries’ cultural and institutional peculiarities (van Haute and Gauja 2015van Haute, E. and Gauja, A. 2015. Party members and activists. London and New York: Routledge. ; Kölln 2016Kölln, A.-K. 2016. ‘Party membership in Europe: Testing party-level explanations of decline’, Party Politics, 22(4), pp. 465-477. https://doi.org/10.1177/1354068814550432 ; Ignazi 2017Ignazi, P. 2017. Party and democracy : the uneven road to party legitimacy. Oxford: Oxford University Press.; van Haute, Paulis and Sierens 2017van Haute, E., Paulis, E. and Sierens, V. 2017. ‘Assessing party membership figures: the mapp dataset’, European Political Science, pp. 1-12. https://doi.org/10.1057/s41304-016-0098-z ; Gherghina, Iancu and Soare 2018Gherghina, S., Iancu, A. and Soare, S. ed. 2018. Party members and their importance in Non-EU countries. A comparative analysis. Abingdon: Routledge).

The ESS data agrees with the common assessment of party membership being in decline. Not only is there an aggregate low level of membership (not more than 4% of the population are party members) but this aggregate level has also been decreasing over time. As Figure 1 shows, the levels of party membership across Europe have experienced steady decline since 2004, moving from 4% to 3% in 2010.

Regarding party activism, the seminal studies were based on the results of party members’ surveys of the UK main political parties (Whiteley et al. 1994Whiteley, P. F., Seyd, P., Richardson, J. and Bissell, P. 1994. ‘Explaining party activism: The case of the British Conservative Party’, British Journal of Political Science, 24(1), pp. 79-94. https://doi.org/10.1017/S0007123400006797 ; Whiteley and Seyd 1998Whiteley, P. F. and Seyd, P. 1998. ‘The Dynamics of Party activism in Britain: A spiral of demobilization?’, British Journal of Political Science, 28(1), pp. 113-137. https://doi.org/10.1017/S000712349800012X , 2002Whiteley, P. F. and Seyd, P. 2002. High-Intensity participation: The dynamics of party activism in Britain. Ann Arbor, MI: The University of Michigan Press.; Seyd and Whiteley 2004Seyd, P. and Whiteley, P. 2004. ‘British Party Members: An Overview’, Party Politics, 10(4), pp. 355-366. https://doi.org/10.1177/1354068804043903 ). This approach and research strategy became quite influential and many other scholars followed their steps on how to measure and explain activism through party surveys (Heidar 2006Heidar, K. 2006. ‘Party membership and participation’. Pp. 301-315. in Handbook of party politics, edited by R.S. Katz and W. J. Crotty. London: Sage Publications.; Scarrow 2007Scarrow, S. E. 2007. ‘Political Activism and party members.’ Pp. 636-654 in The Oxford handbook of political behavio, edited by R. J. Dalton and H. -D. Klingemann. Oxford: Oxford University Press. ; Correa, Rodríguez-Teruel and Barberà 2021Correa, P, Rodríguez-Teruel, J., Barberà, O. 2021.: ‘Farewell to the jack-of-all-trades activists? Clustering party activism in Spain’, Revista Internacional de Sociología, 79(4).). Both, international comparative research projects, such as the Members and Activists of Political Parties (MAPP) led by Emilie van Haute and the increased use of online surveys have promoted a considerable growth of party members and activists’ research based on party surveys during the last decades (van Haute and Gauja 2015van Haute, E. and Gauja, A. 2015. Party members and activists. London and New York: Routledge. ). However, this research strategy presents several limitations for cross-country and, particularly, longitudinal comparisons. One of the limitations is the scarcity of adequate comparative data previous to the 2000s, which is often focused on party delegates’ demographics and attitudes (Reif, Cayrol and Niedermayer 1980Reif, K., Cayrol, R. and Niedermayer, O. 1980. ‘National political parties’ middle level elites and European integration’, European Journal of Political Research, 8(1), pp. 91-112. https://doi.org/10.1111/j.1475-6765.1980.tb00740.x ; Niedermayer 1986Niedermayer, O. 1986. “Methodological and practical problems of comparative party elite research: The EPPMLE Project.” European Journal of Political Research 14(1-2): 253-259. https://doi.org/10.1111/j.1475-6765.1986.tb00833.x ; Pierre, 1986Pierre, J. 1986. “Attitudes and behaviour of party activists.” European Journal of Political Research 14(4): 465-79. https://doi.org/10.1111/j.1475-6765.1986.tb00845.x ; Reif, Niedermayer and Schmitt 1986Reif, K., Niedermayer, O. and Schmitt, H. 1986. ‘Quantitative survey attitudes issue stands ideology?’, European Journal of Political Research, 14(4), pp. 685-690. https://doi.org/10.1111/j.1475-6765.1986.tb00858.x ).

One question not properly addressed by this literature has to do with the evolution of party activism over time. The behavioural dimension linked to party activism has made it difficult to gather comparative cross-national or longitudinal data based on party records or party surveys. This might have been addressed by the above-mentioned population surveys, but party membership research based on this data has focused on other topics, such as the factors shaping members and activists’ political participation (Whiteley 2011Whiteley, P. F. 2011. ‘Is the party over? The decline of party activism and membership across the democratic world’, Party Politics, 17(1), pp. 21-44. https://doi.org/10.1177/1354068810365505 ; Ponce and Scarrow 2014Ponce, A.F., and S.E Scarrow. 2014. “Which members? Using cross-national surveys to study party membership.” Party Politics 22(6): 679-90. https://doi.org/10.1177/1354068814550435 ), or their ideological congruence (Kölln and Polk 2017Kölln, A. K. and Polk, J. 2017. ‘Emancipated party members: Examining ideological incongruence within political parties’, Party Politics, 23(1), pp. 18-29. https://doi.org/10.1177/1354068816655566 ; Lisi and Cancela 2019Lisi, M. and Cancela, J. 2019. ‘Types of party members and their implications: Results from a survey of Portuguese party members’, Party Politics, 25(3), pp. 390-400. https://doi.org/10.1177/1354068817722445 ). In fact, by the 2000s we still knew very little about the main trends of party activism. In her seminal contribution based on data from thirteen countries included in the World Values Survey between 1981 and 1990, Scarrow pointed out that party activism and party membership were not always moving in the same direction (Scarrow, 2000: 95-96Scarrow, S. E. 2000. ‘Parties without members? Party organization in a changing electoral environement’, in Dalton, R. J. and Wattenberg, M. P. (eds) Parties without partisans: Political change in advanced industrial democracies. Oxford: Oxford University Press, pp. 79-101.). By the end of that decade, the literature remained mostly inconclusive on this issue. Party membership surveys on several Anglo-Saxon countries and the Netherlands seemed to highlight declining levels of party activism, but that wasn’t the case of Norway, where they remained quite stable (Heidar 2006Heidar, K. 2006. ‘Party membership and participation’. Pp. 301-315. in Handbook of party politics, edited by R.S. Katz and W. J. Crotty. London: Sage Publications.; Scarrow 2007Scarrow, S. E. 2007. ‘Political Activism and party members.’ Pp. 636-654 in The Oxford handbook of political behavio, edited by R. J. Dalton and H. -D. Klingemann. Oxford: Oxford University Press. ). More recent contributions seem to have confirmed such inconclusive trends, arguing that they might be shaped by party types or the institutional design of the political system (Ponce and Scarrow 2014Ponce, A.F., and S.E Scarrow. 2014. “Which members? Using cross-national surveys to study party membership.” Party Politics 22(6): 679-90. https://doi.org/10.1177/1354068814550435 ; Bale, Webb and Poletti 2020Bale, T., Webb, P. D. and Poletti, M. 2020. Footsoldiers : political party membership in the 21st century. Abingdon and New York: Routledge.; Demker, Heidar and Kosiara-Pedersen 2020Demker, M., Heidar, K. and Kosiara-Pedersen, K. 2020. Nordic party pembers. Linkages in troubled times. London & New York: Rowman & Littlefield International / ECPR Press.).

ESS data on party activism (capturing those having worked in a political party or a political organisation) seems to offer a less distinctive evolution, as shown in Figure 1. Although it paralleled the membership decline of the first decade, after 2008 it experienced a general recovery, achieving a new peak in 2015, followed by a new decrease. This general evolution hides a more heterogeneous picture of the evolution experienced by individual countries. Figure 2 shows different clusters of countries according to their amount of party activism and the overall trend of evolution over two decades. Most of the countries have seen their levels falling between these two time points. Interestingly, those with the highest level of activism at the beginning of the 2000s - Austria, Norway - as well as those with the lowest levels - like Hungary or Italy - are also the countries with the more pronounced decrease in their figures.

Above them, a larger, more heterogeneous group of countries have, on the whole, remained stable with a decreasing trend along time (placed close to the dotted line). We could only detect some increase in Germany or Slovenia, while the rest have reduced their proportion of political activism, probably in line with an equally declining level of grassroots membership. In contrast, a third cluster of countries that traditionally had lower levels of political activism, such as Portugal or Spain, has seen its figures rising, matching the former mentioned countries with higher levels of activism. This heterogeneous evolution indicates that some contingent factors may have produced distinctive effects on the country evolution of activism figures. As we will argue in the next pages, the Great Recession could have played a particularly significant role in those countries experiencing the worst social consequences.

3. DISENTANGLING AND EXPLAINING PARTY MEMBERSHIP AND ACTIVISM VARIATIONS

As already pointed out by Duverger’s seminal work, by the 1950s - the golden age of mass parties - most Western political parties shared a common definition of party members as those fee-paying affiliates registered in the party census and holding a party membership card (Duverger 1954Duverger, M. 1954. Political parties: Their organization and activity in the modern state. London: Methuen.). Such definition set clear boundaries between the party and its environment. Duverger suggested a first typology of party membership that included behavioural elements blurring such distinction. In fact, his concentric model of party membership included other categories, such as party supporters, that were not formally affiliated to the party. Duverger’s typology has shaped the academic and political conceptions of party membership for several generations, particularly in Europe.

During the last decades a growing number of political parties have adopted new organisational strategies in order to reverse declining membership figures. Such reforms have opted for more permeable boundaries between political parties and their environments and more inclusive and plebiscitary decision-making procedures, often relying on the use of new technologies and emulating social media’s affiliation practices (Scarrow 1994Scarrow, S. E. 1994. ‘The “paradox of enrollment”: Assessing the costs and benefits of party memberships’, European Journal of Political Research, 25(1), pp. 41-60. https://doi.org/10.1111/j.1475-6765.1994.tb01200.x ; Dalton and Wattenberg 2000Dalton, R. J. and Wattenberg, M. P. 2000. Parties without partisans. political change in advanced industrial democracies. Oxford: Oxford University Press.; Scarrow, Webb and Poguntke 2017Scarrow, S. E., Webb, P. D. and Poguntke, T. ed. 2017. Organizing political parties. Representation, participation and power. Oxford: Oxford University Press. ; Achury et al. 2020Achury, S., Scarrow, S. E., Kosiara-Pedersen, K. and van Haute, E. 2020. ‘The consequences of membership incentives’, Party Politics, 26(1), pp. 56-68. https://doi.org/10.1177/1354068818754603 ). As a result, the concept of party member and the whole typology in which it is embedded has been substantially redefined. On the other hand, the substantial growth of cross-country and longitudinal comparative studies has also highlighted the need to problematise the operationalisation of such concepts and typologies (Katz et al. 1992Katz, R. S., Mair, P., Bardi, L., Bille, L., Deschouwer, K., Farrell, D., Koole, R., Morlino, L., Muller, W., Pierre, J., Poguntke, T., Sundberg, J., Svasand, L., Velde, H., Webb, P. and Widfeldt, A. 1992. ‘The membership of political parties in European democracies, 1960-1990’, European Journal of Political Research, 22(3), pp. 329-345. https://doi.org/10.1111/j.1475-6765.1992.tb00316.x ; Mair and van Biezen 2001Mair, P. and van Biezen, I. 2001. ‘Party membership in twenty European democracies, 1980-2000’, Party Politics, 7(1), pp. 5-21. https://doi.org/10.1177/1354068801007001001 ; Scarrow 2014Scarrow, S. E. 2014. Beyond party members: changing approaches to partisan mobilization. Oxford: Oxford University Press.; van Haute and Gauja 2015van Haute, E. and Gauja, A. 2015. Party members and activists. London and New York: Routledge. ). So far, the most updated and influential conceptualisation is based on Scarrow’s concept of multi-speed membership that relies on formal, behavioural and technological criteria (Scarrow 2014Scarrow, S. E. 2014. Beyond party members: changing approaches to partisan mobilization. Oxford: Oxford University Press.). In her book, she explored how the rise of the new multi-speed party model has enlarged opportunities for party affiliation beyond the traditional formal (fee-paying) membership. In turn, Scarrow’s conceptualisation opens the floor to new forms of political activism without formal linkages or membership to political parties (Gauja, 2015Gauja, A. 2015. ‘The individualisation of party politics: the impact of changing internal decision-making processes on policy development and citizen engagement’, The British Journal of Politics & International Relations, 17(1), pp. 89-105. https://doi.org/10.1111/1467-856X.12035 ).

As stated below, all the seminal typologies of party membership take into account behavioural elements in order to distinguish between different levels of engagement (Duverger 1954Duverger, M. 1954. Political parties: Their organization and activity in the modern state. London: Methuen.; Heidar 1994Heidar, K. 1994. ‘The polymorphic nature of party membership’, European Journal of Political Research, 25(1), pp. 61-86. https://doi.org/10.1111/j.1475-6765.1994.tb01201.x ; van Haute 2009van Haute, E. 2009. Adhérer à un parti. Aux sources de la participation politique. Brussels: Editions de l’Université de Bruxelles.; Ponce and Scarrow 2014Ponce, A.F., and S.E Scarrow. 2014. “Which members? Using cross-national surveys to study party membership.” Party Politics 22(6): 679-90. https://doi.org/10.1177/1354068814550435 ; Scarrow 2014Scarrow, S. E. 2014. Beyond party members: changing approaches to partisan mobilization. Oxford: Oxford University Press.). In fact, since Michel’s study on the SPD, a key finding of the empirical literature has been that only a minority of party members remain active (Duverger 1954Duverger, M. 1954. Political parties: Their organization and activity in the modern state. London: Methuen.; Whiteley and Seyd 1998Whiteley, P. F. and Seyd, P. 1998. ‘The Dynamics of Party activism in Britain: A spiral of demobilization?’, British Journal of Political Science, 28(1), pp. 113-137. https://doi.org/10.1017/S000712349800012X , 2002Whiteley, P. F. and Seyd, P. 2002. High-Intensity participation: The dynamics of party activism in Britain. Ann Arbor, MI: The University of Michigan Press.; Scarrow 2000Scarrow, S. E. 2000. ‘Parties without members? Party organization in a changing electoral environement’, in Dalton, R. J. and Wattenberg, M. P. (eds) Parties without partisans: Political change in advanced industrial democracies. Oxford: Oxford University Press, pp. 79-101., 2007Scarrow, S. E. 2007. ‘Political Activism and party members.’ Pp. 636-654 in The Oxford handbook of political behavio, edited by R. J. Dalton and H. -D. Klingemann. Oxford: Oxford University Press. ; Heidar 2006Heidar, K. 2006. ‘Party membership and participation’. Pp. 301-315. in Handbook of party politics, edited by R.S. Katz and W. J. Crotty. London: Sage Publications.). Hence, the distinction between active and passive members along with their main causes and practical implications has become a recurring problem for the literature. That said, the definition of activism relies heavily on the political context and the controversial notion of membership, which makes it difficult to assess over time and through different political cultures.

The studies analysing party activism tend to classify party members according to their level of activism defined by the intensity with which members engage in a certain range of pre-defined activities or the time they dedicate to party activity. As such, these studies rely on behavioural measures collected through different sources, but mostly on party members’ surveys. In their seminal research on this topic, Whiteley, Seyd and their colleagues built an operational definition of party activism applied to party members, labelled as high-intensity participation, through an index based on several activities performed (or not) by party members, which they applied when testing their General Incentives Model (Whiteley et al. 1994Whiteley, P. F., Seyd, P., Richardson, J. and Bissell, P. 1994. ‘Explaining party activism: The case of the British Conservative Party’, British Journal of Political Science, 24(1), pp. 79-94. https://doi.org/10.1017/S0007123400006797 ; Whiteley and Seyd 1998Whiteley, P. F. and Seyd, P. 1998. ‘The Dynamics of Party activism in Britain: A spiral of demobilization?’, British Journal of Political Science, 28(1), pp. 113-137. https://doi.org/10.1017/S000712349800012X , 2002Whiteley, P. F. and Seyd, P. 2002. High-Intensity participation: The dynamics of party activism in Britain. Ann Arbor, MI: The University of Michigan Press.; Seyd and Whiteley 2004Seyd, P. and Whiteley, P. 2004. ‘British Party Members: An Overview’, Party Politics, 10(4), pp. 355-366. https://doi.org/10.1177/1354068804043903 ).

Following a similar approach but going one step forward, Ponce and Scarrow (2014)Ponce, A.F., and S.E Scarrow. 2014. “Which members? Using cross-national surveys to study party membership.” Party Politics 22(6): 679-90. https://doi.org/10.1177/1354068814550435 built two measures of party activism based on several behavioural items and going beyond formal membership. This behavioural measure allowed them to avoid problems of inconsistency when analysing party activism over time and across countries. Combining different measures of partisan-related activity, they distinguished two categories of behaviourally-qualified members. On the one hand, core members were defined as those individuals reporting party activism and, simultaneously, declaring three other political activities: having voted in the last election, discussing politics, and working in a non-party organisation. On the other hand, partisan activists were those reporting either party activism or all three political activities mentioned above.

Building on Scarrow’s research, Table 1 provides an alternative way to assess party activism and to disentangle it from formal membership. The table presents a four-category typology of party affiliation based on whether an individual has a formal connection to a political party (e.g. is at least registered in their census as a member) and whether they are also politically active or not. This typology allows us to classify individuals along different party affiliation modes that account for different levels of participation (including active and passive options). One of the poles of this axis corresponds to the Duvergerian notion of militants, those party members with high-intensity participation that dedicate part of their time to party activities. On the opposite side, inactive citizens (at least, with regards to party activity) do not engage in political action connected to political parties. Between them, two additional categories take on intermediate roles. Party grassroots are those formally enrolled members not presenting any recent political activism, corresponding to the traditional role of passive members. In contrast, supporters are those civic activists engaged in political action (run by parties or other related collective associations) but without being full party members (Gauja 2015: 241Gauja, A. 2015. ‘The individualisation of party politics: the impact of changing internal decision-making processes on policy development and citizen engagement’, The British Journal of Politics & International Relations, 17(1), pp. 89-105. https://doi.org/10.1111/1467-856X.12035 ). This classification allows us to include categories that have fallen outside previous classification (i.e., while concepts of militants and supporters are close to Ponce and Scarrow’s abovementioned core members and partisan activists, our classification as activists is less demanding).

Party membership and activism changes have generally been explained by demand (parties) and supply (citizens) factors (Scarrow 1996Scarrow, S. E. 1996. Parties and their members. Organizing for victory in Britain and Germany. Oxford: Oxford University Press. ). Demand explanations highlight the relevance of organisational variables to understand the phenomenon. Several influential but also contested theories have been stated over the last decades following this approach. Epstein’s controversial argument of the contagion from the right suggested that the new mass media (TV) would eventually displace members’ functions, hence explaining their decline (Epstein 1967Epstein, L. D. 1967. Political parties in Western democracies. London: Pall Mall.; Whiteley 2011Whiteley, P. F. 2011. ‘Is the party over? The decline of party activism and membership across the democratic world’, Party Politics, 17(1), pp. 21-44. https://doi.org/10.1177/1354068810365505 ). In a similar vein, changing party models and party families have also been linked to differences in party membership figures (Duverger 1954Duverger, M. 1954. Political parties: Their organization and activity in the modern state. London: Methuen.; Gaxie 1977Gaxie, D. 1977. ‘Economie des partis et rétributions du militantisme’, Revue française de science politique, 27(1), pp. 123-154. https://doi.org/10.3406/rfsp.1977.393715 ; Krouwel 2012Krouwel, A. 2012. Party Transformations in European Democracies. New York: State University of New York Press.; Kölln 2016Kölln, A.-K. 2016. ‘Party membership in Europe: Testing party-level explanations of decline’, Party Politics, 22(4), pp. 465-477. https://doi.org/10.1177/1354068814550432 ). The same might be extended to other organisational factors such as differences on party branch activity, size, staff, financial resources, etc. (Scarrow 2000: 95Scarrow, S. E. 2000. ‘Parties without members? Party organization in a changing electoral environement’, in Dalton, R. J. and Wattenberg, M. P. (eds) Parties without partisans: Political change in advanced industrial democracies. Oxford: Oxford University Press, pp. 79-101.; van Haute and Gauja 2015van Haute, E. and Gauja, A. 2015. Party members and activists. London and New York: Routledge. ; Achury et al. 2020Achury, S., Scarrow, S. E., Kosiara-Pedersen, K. and van Haute, E. 2020. ‘The consequences of membership incentives’, Party Politics, 26(1), pp. 56-68. https://doi.org/10.1177/1354068818754603 ). Finally, other theories link (declining) party membership figures to the competition of other political organisations, such as interest groups or NGOs that increasingly compete with political parties for the recruitment and mobilisation of politically motivated citizens (Lawson and Merkl 1988Lawson, K. and Merkl, P. H. 1988. When parties fail. Emerging alternative organizations. Princeton, New Jersey: Princeton University Press.; Ramiro and Morales 2012Ramiro, L. and Morales, L. 2012. ‘Examining the “demand” side of the market for political activism: Party and civil society grassroots activists in Spain’, Party Politics, 20(4), pp. 506-520. https://doi.org/10.1177/1354068811436057 ). Party level approaches to explain party membership shifts mostly account for cross-party, not cross-country variations.

Supply explanations of party membership have also relied on classic individual-level models of participation. The civic voluntarism model argues that the social status of individuals (income, education) largely shapes their political participation. The cognitive engagement model points out how post-industrial transformations of society (education and communication changes) have shifted political participation practices. Finally, the social capital model emphasises the relevance of trust and civic-voluntary networks for political participation (Whiteley 2011: 27-29Whiteley, P. F. 2011. ‘Is the party over? The decline of party activism and membership across the democratic world’, Party Politics, 17(1), pp. 21-44. https://doi.org/10.1177/1354068810365505 ). Such models have mostly been applied to cross-country comparisons, not to longitudinal studies. On the other hand, institutional factors (type of regime, size, etc.) and short or medium-term effects (transitions to democracy, electoral or government cycles) have mostly been employed to assess cross-country and longitudinal differences on party membership (van Haute and Gauja 2015: 5van Haute, E. and Gauja, A. 2015. Party members and activists. London and New York: Routledge. ).

The demand side literature has stressed the relevance of the rational choice theories to understand party membership and activism but has only partially acknowledged the relevance of the economy as a factor explaining cross-country and longitudinal changes. Whiteley (2001) suggested, for example, that high regulatory practices could eventually discourage both membership and party activism across countries (Whiteley 2011: 34-35Whiteley, P. F. 2011. ‘Is the party over? The decline of party activism and membership across the democratic world’, Party Politics, 17(1), pp. 21-44. https://doi.org/10.1177/1354068810365505 ). However, to the best of our knowledge there is no comparative literature yet that considers the effects of economic crises on party activism. This is quite puzzling because the last Great Recession prompted a surge on protest activities in several European countries and, eventually, quite striking changes on several party systems where new parties emerged while others declined or eventually died out (Bermeo and Bartels 2014Bermeo, N. and Bartels, L. 2014. Mass politics in tough times: opinions, votes and protest in the Great Recession. Oxford: Oxford University Press.; Quaranta 2015Quaranta, M. 2015. Political protest in Western Europe: exploring the role of context in political action. Cham: Springer.; Morlino and Raniolo 2017Morlino, L. and Raniolo, F. 2017. The impact of the economic crisis on South European democracies. Cham: Springer International Publishing. ; Kriesi et al. 2020Kriesi, H., Lorenzini, J., Wüest, B. and Häusermann, S. ed. 2020. Contention in Times of Crisis: Recession and Political Protest in Thirty European Countries. Cambridge: Cambridge University Press.). Trying to answer a similar question on the effects of an economic crisis on participation in voluntary associations, Cameron suggested two competing hypotheses supported with previous partial empirical evidence (Cameron 2021Cameron, S. 2021. ‘Civic engagement in times of economic crisis: A cross-national comparative study of voluntary association membership’, European Political Science Review, pp. 1-19. https://doi.org/10.1017/S1755773921000060 ). The retreat hypothesis suggested, in line with the civic voluntarism model, that the economic crisis would demobilise citizen participation. The mobilisation hypothesis pointed out that people would join certain types of civic organisations to advocate for their policy interests or provide support to others. Using WWS data before and after the crisis in 14 democratic countries, Cameron only found support for the retreat hypothesis, and discarded any specific positive effect over politically oriented voluntary associations.

Building on this argument, while Cameron’s retreat hypothesis suggests those years of dramatic social turbulences caused by the economic crisis should have seen a decline in party activism over time, we argue that some party-related factors may have produced the opposite effect. Previous studies have found disruptive political outcomes following the Great Recession. These are the years of the financial crisis that saw an increase of political dissatisfaction and distrust in political parties. In this critical context, protest mobilisations spread in those countries more affected by austerity policies, and party systems experienced turbulences and electoral realignments (Bosco and Verney 2016Bosco, A. and Verney, S. 2016. ‘From electoral epidemic to government epidemic: the next level of the crisis in Southern Europe’, South European Society and Politics. 21(4), pp. 383-406. https://doi.org/10.1080/13608746.2017.1303866 ). New political parties made their breakthrough in critical elections, enlarging the opportunities for participation within political parties and new voluntary organisations. These new actors often opposed austerity policies implemented by governments and claimed political reforms and social protection. In other cases, new parties simply opposed traditional ‘old politics’ and defended more authoritarian values. In sum, we should expect a net mobilisation effect on party activism in those countries affected by the austerity policies.

4. METHODS AND DATA

As stated in section 2, several comparative studies based on party registers from Western countries have highlighted that membership figures have been decreasing since the 1960s. These studies have pointed out some fluctuations based on countries and party families. The alternative way to approach party membership and activism comes from the analysis of several waves of international comparative population surveys such as the International Social Survey Program (ISSP), the World Values Survey (WWS) or the European Social Survey (ESS). Such surveys allow for cross-national and longitudinal comparisons based on behavioural and self-reported measures of membership and activism through one or two questions.

Building on the self-reported approach to studying party membership and activism, we will adopt subjective definitions of party membership and party activism to study their evolution in recent years across democratic Europe. This subjective approach based on opinion surveys helps to avoid problems of unreliability and heterogeneity that are usually related to the comparative analysis of objective party registers’ data (Ponce and Scarrow 2014Ponce, A.F., and S.E Scarrow. 2014. “Which members? Using cross-national surveys to study party membership.” Party Politics 22(6): 679-90. https://doi.org/10.1177/1354068814550435 ). Hence, this self-perceived notion of party membership and party activism guarantees the similarity of the measures across parties and countries and allows cross-time comparisons. The ESS dataset employed in our analysis provides two basic indicators to measure those concepts: a self-definition of party membership, and a self-declaration of having worked in a political party or a political organisation in the last 12 months¹The wording of the first measure is: “Are you a member of any political party?” For party activism, the survey asked: “There are different ways of trying to improve things in [country] or helping prevent things from going wrong. During the last 12 months (…) have you worked in a political party or action group?”.

This strategy might also involve issues related to question wording, consistency and meanings (e.g. activity within vs. activity for the party) over time and different political contexts (Whiteley 2011Whiteley, P. F. 2011. ‘Is the party over? The decline of party activism and membership across the democratic world’, Party Politics, 17(1), pp. 21-44. https://doi.org/10.1177/1354068810365505 ; Ponce and Scarrow 2014Ponce, A.F., and S.E Scarrow. 2014. “Which members? Using cross-national surveys to study party membership.” Party Politics 22(6): 679-90. https://doi.org/10.1177/1354068814550435 ; Achury et al. 2020Achury, S., Scarrow, S. E., Kosiara-Pedersen, K. and van Haute, E. 2020. ‘The consequences of membership incentives’, Party Politics, 26(1), pp. 56-68. https://doi.org/10.1177/1354068818754603 ). The self-definition as a party member can be understood as an equivalent measure of formal party enrolment. While the behavioural measure of party activism is probably intended to capture close organisational figures such as electoral coalitions or cartels, its ambiguous formulation cannot totally exclude some responses from individuals that were involved in citizen action committees or pressure groups (Ponce and Scarrow 2014: 4Ponce, A.F., and S.E Scarrow. 2014. “Which members? Using cross-national surveys to study party membership.” Party Politics 22(6): 679-90. https://doi.org/10.1177/1354068814550435 ). Therefore, we should not consider them just as a sub-group within the wider pool of party members. Nonetheless, and in line with our typology (Table 1), it allows us to capture party affiliation and involvement beyond formal membership.

As we mentioned in the previous sections, the literature on party membership and activism has stressed the relevance of country or regional patterns related to their different institutional, political and cultural background. In this article, our empirical question aims to test the political effects of the financial crisis on party activism. To test our hypothesis, we employ the OLS technique estimated with panel-corrected standard errors (PCSEs), a panel regression that accounts for the heterogeneity of panel data, considering the possibility of contemporaneous correlations and potential problems of heteroscedasticity (Beck and Katz 1995Beck, N. and Katz, J. N. 1995. ‘What to do (and not to do) with Time-Series Cross-Section Data’, American Political Science Review, 89(3), pp. 634-647. https://doi.org/10.2307/2082979 ). As abovementioned, we use the ESS nine rounds from 2002 to 2018. We have built a time-series cross-section data although not all the countries included in the research participated in all the rounds. In order to avoid this problem, we estimate the panel regression with pairwise selection, in order to include only available observations with non-missing pairs. Only half of the rounds asked about self-defined partisanship (until 2010) preventing us from extending our hypotheses to all modes of party affiliation beyond that time point, also limiting the scope of our empirical analysis.

Hence, our sole dependent variable in the explanatory section is party activism, which includes the militant and supporter modes of party affiliation discussed in the theoretical section and it is captured by the behavioural measure of party activism from the ESS surveys. Because our interest is focused on cross-national evolution over time, we count this variable in aggregated terms, measuring the proportion of activism per country in each year/round. The main independent variable is bailout, a dummy capturing those time points when a country’s economy was under strict austerity policies linked to the bailout in the years after the Great Recession. We also include here Spain that got a European bailout to save its bank system. Additionally, our model controls for other political variables at the country level that were employed by previous studies. These include electoral fragmentation (through the effective number of electoral parties), government effectiveness, and regulatory quality (Kaufmann index) (Whiteley 2011Whiteley, P. F. 2011. ‘Is the party over? The decline of party activism and membership across the democratic world’, Party Politics, 17(1), pp. 21-44. https://doi.org/10.1177/1354068810365505 ; Ponce and Scarrow 2014Ponce, A.F., and S.E Scarrow. 2014. “Which members? Using cross-national surveys to study party membership.” Party Politics 22(6): 679-90. https://doi.org/10.1177/1354068814550435 ). Social factors (education, inequality) are measured with the proportion of school enrolment in tertiary education, and the Gini index. Besides, the effects produced by the economic structure are captured through the use of GDP growth, and the level of national debt²The data on tertiary education (SE.TER.ENRR), the Gini index (SI.POV.GINI), and GDP growth (NY.GDP.MKTP.KD.ZG) are from the World Bank. The figure for the General government debt comes from the OCDE..

5 THE EVOLUTION OF PARTY AFFILIATION TYPES IN EUROPE DURING THE 2000S

The first part of the empirical section is devoted to the descriptive analysis of the ESS country evolution data of three different measures of party affiliation (grassroots, militants and supporters) until 2010. This will allow us to better capture the evolving relationship existing between membership and activism within parties³In Figure A1 (see Appendix), we can observe the overall distribution of those profiles of party affiliation across countries in the ESS dataset for 2002-2010. The significant differences among countries suggest, in line with previous studies, are from the World that country-variance is an important aspect when trying to capture trends of evolution of the levels of membership and activism..

As we noted in the second section, while party membership decline has been well documented, Figure 1 also reflected an interesting diversion of party activism at the end of the decade. Both measures - membership and activism - evolved in parallel until 2008. The diverted evolution after 2008 might be related to the wide definition of activism contained in the ESS measure, including also political action not strictly developed within political parties. Hence, some authors have referred to the rise of new forms of civic activism as an explanation of the party decline, as people would increasingly prefer non-partisan channels for social involvement (Whiteley 2011Whiteley, P. F. 2011. ‘Is the party over? The decline of party activism and membership across the democratic world’, Party Politics, 17(1), pp. 21-44. https://doi.org/10.1177/1354068810365505 ). In the ESS dataset for all countries between 2002 and 2010, half of the 3.5% individuals reporting party membership also declared having being involved in party activism in the last 12 months (we label them as militants). Furthermore, we find a similar proportion (45%) of party members among all those declaring party activism. If the new activism argument could explain this divergence, we should observe a common path of decline for grassroots (non-active members) and militants, in contrast to supporters (active non-militants). However, Figure 3 indicates a different evolution. After a common decreasing path for all types of party affiliation, the most interesting feature of this portrait is the divergent change of trends between activists and grassroots: while militants remain stable in 2010, supporters go up in parallel to a steady decline of grassroots that reach the lowest point of the decade.

Unsurprisingly, the evolution of this trend for both activists and grassroots is not consistent across countries. As seen in Figure 4, we can distinguish at least four clusters of countries based on their patterns of evolution of party affiliation between 2008 and 2010. On the one hand, there are countries like Spain, Switzerland or the Netherlands, including to a lesser extent the Czech Republic, Portugal or Croatia, where party activism increases while grassroots membership goes in the opposite direction. In most of those countries, the result is a total increase of party activism based on an extension of the activist network of political parties. Belgium stands as a peculiar case with a slight increase of both militants and grassroots. A second cluster shows a very different pattern, following the same trend expected by the literature, with a decline in both party activism and grassroots membership. Interestingly, in most of those countries non-member activism also has a negative trend. The UK, Denmark and Ukraine are the only countries where supporters are on the rise in contrast to militants.

A third cluster forms around Scandinavian countries (Sweden, Norway, Finland), defined by a small increase of grassroots and a decrease in both militants and supporters. These countries reflect a decline of party activism, although this does not erode the partisanship base of the organisations. Only France seems to follow a similar pattern, although it would be better located among those other countries forming a fourth cluster of countries where militants keep a stable proportion, while one of the other groups (grassroots or supporters) follow increasing, decreasing or opposing trends. As Figure 4 shows, this fourth cluster is mostly composed by north and central European countries such as Germany, Slovenia, Bulgaria, Estonia -or even Poland. Its main distinctive point is to fill a gap between those other main clusters.

Unfortunately, our data does not allow us to track the evolution of these types of party affiliation after 2010. But, as the most recent literature based on party registers has also highlighted (van Haute and Gauja 2015van Haute, E. and Gauja, A. 2015. Party members and activists. London and New York: Routledge. ; Kölln 2016Kölln, A.-K. 2016. ‘Party membership in Europe: Testing party-level explanations of decline’, Party Politics, 22(4), pp. 465-477. https://doi.org/10.1177/1354068814550432 ; Ignazi 2017Ignazi, P. 2017. Party and democracy : the uneven road to party legitimacy. Oxford: Oxford University Press.; van Haute, Paulis and Sierens 2017van Haute, E., Paulis, E. and Sierens, V. 2017. ‘Assessing party membership figures: the mapp dataset’, European Political Science, pp. 1-12. https://doi.org/10.1057/s41304-016-0098-z ; Gherghina, Iancu and Soare 2018Gherghina, S., Iancu, A. and Soare, S. ed. 2018. Party members and their importance in Non-EU countries. A comparative analysis. Abingdon: Routledge), these findings indicate a more complex portrait against the argument of party membership decline. While almost all European democracies were experiencing a general decline of party grassroots by 2010, half of them were simultaneously keeping stable or increasing trends of party activism, within or outside political parties. In this respect, different forms of stronger political involvement could be replacing traditional soft partisanship. In the next section, we test the drivers of party activism, especially the influence of the Great Recession.

6. EXPLAINING PARTY ACTIVISM 2002-2018

To assess the role of the Great Recession on party activism, we use the PCSEs technique with pairwise selection, clustering the standard errors on the country level to estimate the relevance of several factors on the national proportion of party activism across time. Figure 5 shows the average marginal effects of the different factors included in our full model. Our main expectation is confirmed, as those countries that got bailout observed an increase in their levels of party activism during those years under the austerity policies, associated with that bailout. In the years under bailout, countries had on average 5.2 percent of party activists, increasing by 1.4 compared to countries without bailout (see the plotted predicted margins in Figure 6). In that sense, our results indicate the Great Recession acted as a mobilisation agent in those countries that were more severely affected. Contrarily to Cameron (2021)Cameron, S. 2021. ‘Civic engagement in times of economic crisis: A cross-national comparative study of voluntary association membership’, European Political Science Review, pp. 1-19. https://doi.org/10.1017/S1755773921000060 findings on joining voluntary associations, when it comes to party activism and not simply membership, the economic consequences of the Great Recession manifested in the bailout acted as a trigger to enhance individuals’ chances to become more active in political parties. To check the robustness of this result, we conduct an alternative strategy, running a multilevel mixed-effects linear regression, clustering data at the country level and controlling by time. The bailout effect remains the same. Hence, the mobilisation hypothesis receives empirical support, pointing in the same direction as other studies reporting that the recession brought new actors and citizens into the political scene as a reaction to the huge social costs produced by the crisis (Aslanidis 2016Aslanidis, P. 2016. ‘Populist Social movements of the Great Recession’, Mobilization: An International Quarterly, 21 (3): 301-321. https://doi.org/10.17813/1086-671X-20-3-301 ; Rodríguez-Teruel, Barrio and Barberà 2016Rodríguez Teruel, J., Barrio, A., & Barberà, O. 2016. Fast and Furious: Podemos’ quest for power in multi-level Spain. South European Society and Politics, 21(4), 561-585. https://doi.org/10.1080/13608746.2016.1250397 ).

The model also shows the importance of political and social control factors, in line with previous studies. Starting with the quality of the regulatory system, our model supports previous findings indicating a negative relationship with party activism (Whiteley 2011Whiteley, P. F. 2011. ‘Is the party over? The decline of party activism and membership across the democratic world’, Party Politics, 17(1), pp. 21-44. https://doi.org/10.1177/1354068810365505 ). As Whiteley (2011)Whiteley, P. F. 2011. ‘Is the party over? The decline of party activism and membership across the democratic world’, Party Politics, 17(1), pp. 21-44. https://doi.org/10.1177/1354068810365505 points out, this could lead to over-regulation of political parties and act as a barrier for party activism (p.35). On the other hand, the effectiveness of the governance model of the political system (capturing perceptions of the quality of public services, civil service, policy formulation and implementation among others) shows a positive relationship suggesting a robust democracy paves the way for intense party activism, within or outside parties. Finally, the level of education in a country (captured with the level of school enrolment in tertiary education) shows the importance of cognitive resources for party activism, as we commented in the theoretical section (Whiteley, 2011Whiteley, P. F. 2011. ‘Is the party over? The decline of party activism and membership across the democratic world’, Party Politics, 17(1), pp. 21-44. https://doi.org/10.1177/1354068810365505 ; Ponce and Scarrow, 2014Ponce, A.F., and S.E Scarrow. 2014. “Which members? Using cross-national surveys to study party membership.” Party Politics 22(6): 679-90. https://doi.org/10.1177/1354068814550435 ).

A descriptive cross-country analysis can help us to show a more detailed view of the connection between the crisis and the evolution of activism. If we observe the overall evolution of political activists over time in different countries, two distinct patterns are clearly defined. We observe a first pattern followed by those countries with larger variations in some periods (see Figure 7). It is interesting to note that the common feature of these cases is the existence of variations in the levels of party activism of almost 2 or even more points. These variations are often sustained in at least two rounds, producing significant increases or decreases in the proportion of party activists. Although most of them differ in the form of the general trend, particularly during the first years, all these countries have experienced substantial increases of activism between 2010 and 2014, or afterwards. See the case of Spain, where party activism decreased during the first decades, and was then followed by an increasing trend until 2014, although the last round (2018) has fallen again to the lowest point in ten years. In this period of recovery for party activism, the Spanish party system was in a deep turmoil, with the rise of new successful challenger parties that contributed to the rise of political involvement. This political transformation also coincides in time with the implementation of new forms of party mobilisation that did not necessarily entail the traditional inactive role of affiliated members. At the same time, the effect of the economic recovery and the access of challenger forces to parliament could have eroded the potential for party activism in Spain. The result is an ‘albatross’ trajectory of activism figures, with two ‘unfolding wings’, one per decade. Portugal is a different case, where low levels of political activism remained very stable between 2004 and 2012, but since then activism has been on the rise, reaching a peak of mobilisation in 2018, the highest since 2000. However, in most of the countries with increasing figures of party activism, the positive trend stopped by the end of the decade. While in some of these cases the Great Recession helped to temporarily boost party activism, the most recent ESS round suggests this effect has started vanishing.

On the other hand, there is a cluster of countries where the predominant trend in party activism is stability or smooth decline, in line with the general pattern of demobilisation of grassroots membership (see Figure A2 in the Appendix depicting the evolution of the more typical countries in this group). All these countries share the same pattern of stability, where there have not been important changes across time, meaning the general absence of inter-round variations higher than 1 point, with the exception of some significant variations related to the 3^rd round (2006), particularly in Poland, Denmark or Belgium. Countries with similar patterns are Switzerland, Hungary or Finland, as well as several others (although not so well documented as a consequence of the lack of data in some rounds). The existence of these two different patterns generally matching those countries most affected by the Great Recession highlights the relevance of accounting for the economic context when studying party activism.

7. CONCLUSION

The main purposes of this study were to provide comparative evidence of the recent evolution of party affiliation modes in European democracies and to explain the role of the Great Recession in the levels of party activism. Apart from the well documented membership decline, we expected to observe different patterns of evolution for different types of party affiliation. We also expected to find different country patterns, particularly among those countries more affected by the Great Recession. Building on the ESS cross-section data for most of European countries, we show that party activism and grassroots membership have followed a similar declining pattern but this pattern has diverged in the second decade in parallel to the economic crisis. Additionally, our results also show an important country variance in the evolution of party activism with two main distinct patterns: one with remarkable stability along time, and another one with significant variation defined by ups and downs, particularly in the second decade. When trying to explain these different country patterns, our results show the positive effect the bailout and government effectiveness have had on party activism and the potential drawbacks of over-regulation. Importantly, our inferential analysis sustains the relevance of accounting for systemic and contextual factors and their changes over time when analysing party activism.

Furthermore, even if our results are curtailed by the lack of full comparative data in the second decade, we can conclude that the levels of party membership and party activism have not necessarily evolved following the same path. Instead, country context may favour new forms of activism, particularly one without formal affiliation with political parties. This, in turn, opens new chances for individuals and political parties to boost party activism despite the shrinking formal membership. Indeed, the increasing weight of party activists (as formal militants or as informal supporters) compared to the grassroots members reflect the changing nature of party membership. The co-existence of different modes of party affiliation and their different evolution patterns inform about the challenges and offsets that political organisations might face when managing external crisis and when opening their machinery to more democratic internal participation. This evolution is consistent with the notion of multi-speed membership parties established in recent literature (Scarrow 2014Scarrow, S. E. 2014. Beyond party members: changing approaches to partisan mobilization. Oxford: Oxford University Press.) and the development of new tools for participation designed for non-members (Gauja 2015Gauja, A. 2015. ‘The individualisation of party politics: the impact of changing internal decision-making processes on policy development and citizen engagement’, The British Journal of Politics & International Relations, 17(1), pp. 89-105. https://doi.org/10.1111/1467-856X.12035 ). It also raises questions about the risks for political parties regarding increasing discontinuity of party activism, which could become more sensitive to the evolution of the social and individual context.

For a better and more robust understanding of this transformation, our findings should be complemented with supply and demand analysis accounting for individual drivers of activism and party measures to encourage certain types of party activism. Although the party literature has tended to apply the same explanatory models for both party members and party activists, the existence of new forms of high-intensity party engagement without a necessary formal attachment require paying further attention to the potential explanatory capacity of these models and their shortcomings. In this regard, the analysis of subjective measures of party activism needs to be complemented with more objective data about party activists, as we may find in other works in this special issue. As we have observed in our article, there may be good times to come yet for research on party activism.