From Divergence to Convergence: Degrees of Formality in the Connections Between Two Flemish Progressive Political Parties and Organised Civil Society

• 1 Introduction

  The connections between political parties and civil society are central to ensuring that the concerns and preferences of ordinary citizens are reflected in the decisions of their governments (Panebianco, 1988; White & Ypi, 2016). Connections matter to both parties and CSOs. For parties, their connections to organised civil society provide pathways to large numbers of voters, sources of policy advice and expertise, and sounding boards during discussions about coalition formation (Allern et al., 2021a; Romeijn, 2020). And for CSOs, parties are ‘gatekeepers’ to public office, providing channels for CSOs to influence policymaking and realise the policy goals of their members (Berkhout, 2013).
  For much of the 20th century, analysis of connections between political parties and civil society focused on the relationships between parties with large memberships and CSOs representing economic interests (Bartolini, 2000; Bartolini & Mair, 1990). Most connections between parties and CSOs were formalised in statutes of political parties or through individual membership of linked organisations (Allern & Verge, 2017; Poguntke, 2002).
  In recent decades, party systems in Western Europe have undergone extensive changes. The weakening of traditional cleavages that structured party systems around class, religion and geography, and the emergence of new cleavages around sociocultural dimensions, (Bornschier, 2009; Ford & Jennings, 2020) have fragmented party systems. The number of parties contesting elections has increased, and elections have become increasingly volatile (Emanuele, 2015). And the increased salience to European electorates of issues, such as the environment and immigration, and the emergence of radical populist parties have resulted in greater polarisation of party systems (Borbáth et al., 2023).
  At the same time, civil society has undergone extensive changes. On the one hand, there has been an increase in the prominence of professionally run and state-funded CSOs in place of membership-based organisations (Maloney, 2012). On the other hand, recent decades have witnessed the mobilisation by new social movements that articulate a wide range of new claims in areas such as environment, globalisation and identity politics (Fuchs & Rucht, 2018) and an increased polarisation of civil society as new conflicts are politicised by radical actors (Grande, 2022).
  These changes have sparked a revival of scholarly interest in the relationship between political parties and organised civil society, suggesting that they have changed in form and structure. While formal connections have declined (Allern & Verge, 2017), a wide range of informal contacts between parties and CSOs have developed that vary widely in their degree of routine or regularity (Allern, 2010; Allern, et al., 2021b). Other studies have compared the different approaches to renewing links of parties within the same party family (Lisi, 2013) or have analysed party strategies for establishing connections in a new democracy (Verge, 2012).
  However, less is known about how parties and CSOs from diverging historical and organisational traditions have navigated their relationships in a time of change. A shift from more formal and exclusive to informal and nonexclusive connections may not mean that connections have become less relevant for important political outcomes such as the openness and legitimacy of public policymaking and the responsiveness of political elites to new ideas and inputs from civil society. The study of how parties and CSOs shape connections in a changing context can, therefore, help us understand how patterns of interest representation are evolving and throw light on challenges to their legitimacy.
  To explore this question, I draw on the growing literature in this field to propose a comprehensive typology of connections between parties and CSOs comprising formal connections and informal connections. My case study considers patterns of connection from the perspective of both parties and CSOs. I focus on two progressive political parties in the Belgian region of Flanders: (1) Vooruit, a party with extensive government experience and a history of ties to pillarised structures of interest representation (Van Haute & Wauters, 2019) and (2) Groen, a challenger party that emerged from new social movements. And I consider the perspectives and strategies of a wide range of Flemish CSOs on the form and evolution of their connections with both these parties and other parties in the Flemish party system. I combine data from a novel dataset, party statutes and semi-structured interviews with political and civil actors in Flanders.
  My study of how patterns of connection between parties and CSOs have been shaped at the system level is guided by two general expectations. The first is that changes in the Flemish political system and civil society will have shaped a pattern of connections dominated by informal rather than formal types of connection. The second is that organisations from the region’s traditional ‘pillar’ structures will retain stronger formal connections than organisations that have emerged outside these structures, reflecting path dependencies rooted in ideology and origins (e.g. Gunther, 2005; Rasmussen & Lindeboom, 2013).
  The findings are supportive of the first expectation, and there is also some support for the second expectation. Departing from patterns of connection shaped by different organisational and ideological traditions, the two parties and CSOs have converged on a very similar structure of relationships, even though this structure retains legacies of the regions’ system of pillarisation. The contemporary structure of relationships between parties and civil society is dominated by non-exclusive informal contacts of varying degrees of regularity that are initiated by both civil and political actors. I explain this convergence in terms of the politicisation of public policymaking in Flanders and the pragmatic approach of mainstream civil society to representing the claims of their constituencies.
  This article makes an important contribution to our understanding of how parties and CSOs navigate their interactions in a time of change, identifying two important factors that play a significant part in shaping relationships: (1) the influence of policymaking institutions and (2) the strategies of both parties and civic groups in pursuing their goals within those institutions. And by identifying a full typology of connections between parties and CSOs and exploring the pattern they take at the level of a political system, this article provides a framework for further comparative research into how patterns of connection may be shaped in political systems characterised by different histories and structures of interest representation.
  This article is structured as follows. In the first section, I review theoretical contributions on the subject of this article and propose a full typology of formal and informal connections between parties and CSOs. In the second section, I detail the research design, introduce the Flemish case and describe the methods and data used to explore the research question. The results of the study are presented in the third section, and the final section discusses conclusions and identifies areas for future research.

• 2 Theory and Literature

  The theoretical discussion of connections between political parties and CSOs has, until recently, been framed by classical cleavage theory and, in particular, by how connections reflected the shared interests of organisations representing opposing sides of prevailing cleavages of social class (Kitschelt, 1994) and religion (Duncan, 2015). The vertically structured mass party which emerged in the early 20th century was particularly well suited to the function of integrating and mobilising constituencies represented by organised civil society (Mudge, 2018). In the era of the mass party, the connections between parties were membership based and formal in nature, and party statutes defined relationships between parties and a wide range of ‘ancillary’ organisations (Poguntke, 2002). In countries such as Austria, Belgium and the Netherlands, whose political and sociocultural life was structured by pillarisation, parties and associated CSOs were key to the creation of ideational sub-cultures (Van Haute & Wauters, 2019).
  The transformation of the socioeconomic structure of Western European democracies provided the backdrop for the evolution of cleavage theory (Bornschier, 2009). De-industrialisation, the rise of the service sector and the expansion of higher education weakened traditional cleavages centred around class, religion and geography and led to the emergence of new cleavages centred around sociocultural issues and dimensions (Bornschier, 2009; Ford & Jennings, 2020). A rise in issue-based voting further complicated the task faced by parties in connecting to and representing social groups; for example, by weakening voting based on economic interests (Kayser & Wlezien, 2011). In traditionally pillarised countries, political and social changes were instrumental to a process of ‘de-pillarisation’ in which institutions and parties associated with different pillars lost influence and electoral support (Hellemans, 2020).
  In the last two decades, the study of connections between parties and CSOs has mirrored the evolution of cleavage theory. Socioeconomic changes in Western Europe not only weakened the structural basis to these connections but also led to change in the organisations that constituted connections. Trade union membership fell throughout Western Europe (Kelly, 2015), party membership also declined (Van Biezen et al., 2012) and there was a growth in the number of CSOs representing an increasingly diverse range of interests and claims (Beyers et al., 2008). The scholarly focus shifted to the analysis and explanation of informal connections (Allern, 2010; Allern et al., 2021b), ‘movement’ parties that fuse the identities and action repertoires of parties and social movements (Della Porta et al., 2017), and the strategies of particular party families in constructing links to civil society (Blings, 2020; Lisi, 2013).
  The historical origins and traditions of different party families influenced the form and intensity of their connections to CSOs. Socialist parties in Western Europe were founded in the late 19th and early 20th centuries to represent the interests of a growing industrial working class (Bartolini, 2000). For parties in the socialist party family, connections closely resembled the close, formal and membership based patterns typical of the mass party (Mudge, 2018). In contrast, while parties in the Green party family emerged from the 1970s to give political expression to the claims of environmental and antinuclear movements, these movements remained comparatively distant from Green parties (Rootes, 2000).
  On the basis of this growing literature, I propose that the connections between political parties and CSOs can be grouped into two principal categories: formal connections and informal connections. Formal connections are comprised of relationships defined by individual membership and affiliation of linked organisations and by party statutes. In contrast, informal connections are not customarily codified in party documents or written agreements between parties and CSOs but are expressed in patterns of interaction between officials and shared forms of political action.

  2.1 Formal Connections

  Formal connections are made up of three types: statutory links and organisational density, which are strongly associated with the ‘thick’ organisation characteristic of the mass party (Gunther & Diamond, 2003), and multispeed memberships, a newer type of connection that is evolving in response to changes in party organisation and changing patterns of communication between parties and sympathisers (Scarrow, 2019).
  The first type of formal connection between parties and civil society is statutory links. Party statutes are “constituting documents that present an account of who parties are and how they operate” (Scarrow et al., 2023). Statutory links comprise the definition in party statutes of formal relationships between a party and external organisations such as trade unions that share a party’s goals, and internal sub-groups that represent important social demographic constituencies of parties such as young people and women (Allern & Verge, 2017). By specifying substantive rights for such organisations over internal party matters such as the nomination of delegates to decision-making bodies, statutory links bring representatives of organised social groups into the heart of party deliberation on questions of strategy and policy formation (Lisi, 2013; Poguntke, 2002). Because they are formally codified in party documents, statutory links are durable and show great stability over time (Poguntke, 2002).
  A second type of formal connection is organisational density (Bartolini & Mair, 1990), which describes the extent to which groups of voters are linked to parties by organisational membership and other “agencies of electoral mobilisation” (1990, p. 214). In Bartolini and Mair’s formulation, organisational density is based on two types of individual membership – of trade unions and of political parties – and measured as a composite index of the ratio of party membership to the electorate and trade union density (1990, p. 255). The principal function of organisational density is to strengthen the partisan attachment of social groups and secure their electoral loyalty.
  In the period to the late 1980s, studies suggested that both these types of formal connection played a part in constraining electoral volatility; stronger statutory links by extending the organisational reach of party elites and facilitating communication between parties and large groups of voters (Poguntke, 2002), and organisational density by strengthening the partisan attachment of large groups of voters (Bartolini & Mair, 1990). More recent studies (e.g. Martin, et al., 2022a) have suggested that while trade union membership still stabilises party vote shares, party membership has declined to such an extent that it no longer performs this function.
  The third form of formal connection – multispeed membership – is a response to this decline of party membership in Western Europe (Van Biezen et al., 2012). The multispeed membership organisation combines expanded participation and decision-making rights for members with new forms of affiliation for party supporters (Scarrow, 2019). Multispeed membership takes a wide range of forms in Western European parties, including the establishment of friends’ and supporters’ networks that encourage the participation of non-members in party activity (Gauja, 2015), and the opportunity for non-members to register with a party and take an immediate part in the election of leaders and selection of party election candidates (Scarrow, 2014). The main functions of the multispeed organisation include attracting additional organisational resources and connecting to new groups of supporters, but the extent to which it enhances connection to organised civil society is unclear (Kosiara-Pedersen, et al., forthcoming).
  Informal connections are comprised of three types: (1) regular contacts, (2) connective density and (3) shared action repertoires. Regular contacts take many forms, from joint commissions on policy issues to various forms of unofficial contact between individual representatives, some of which are routine in nature (Allern, 2010; Allern et al., 2021b). They perform two main functions, the exchange of information and resources of benefit to both parties and CSOs (Otjes & Rasmussen, 2017) and the development of public policy (Allern et al., 2021a). While influenced by shared policy preferences, regular contacts are not exclusive – CSOs initiate contacts with a wide range of parties, especially when policy issues are highly salient to the public (Statsch & Berkhout, 2020). Widespread regular contacts between parties and organised civil society have been documented in country specific studies (Allern, 2010; Verge, 2012) and extensive cross-country surveys of CSOs and parties (e.g. Otjes & Rasmussen, 2017). Contacts intensify during election periods when CSOs provide information to their members on party programmes (Blings, 2020) and during coalition formation discussions, when parties and CSOs engage in close dialogue over policy trade-offs (Romeijn, 2020).
  The candidates that parties select for elections constitute the core of parties’ elites, forming the pool from which members of parliament and government are drawn (Rahat, 2007). Connections between party elites and membership CSOs are conceptualised as parties’ connective density (Martin et al., 2022a) and measured as party election candidates’ membership of CSOs using data from the Comparative Candidates Survey (CCS, 2016, 2019).¹x More details on the operationalisation and measurement of connective density are provided in the following section and in Appendix C. The main functions of connective density are securing programmatic alignment between parties and membership based CSOs and the stabilisation of party support at elections. Connective density stabilises the vote share of parties of all ideologies and origins (Martin et al., 2022a) and is highest in parties with government experience and in parties with a left-wing ideology (Martin et al., 2022b).
  The third form of informal connection sees the identity, roles and action repertories of parties and CSOs as distinct agents of democratic representation, replaced by a shared action repertoire (Della Porta et al., 2017). Action repertoires comprise conventional initiatives to promote legislative change within representative institutions, conventional forms of protest, such as demonstrations or boycotts, and less conventional forms of interest aggregation, such as citizen policy forums or ‘solidarity’ initiatives (Della Porta et al., 2017). A shared action repertoire exists when parties and CSOs jointly conceive, plan and implement joint political actions in diverse settings, such as the joint mobilisation of social and professional groups in opposition to austerity measures in Southern Europe in the 2010s (Della Porta et al., 2017). The main functions of the shared action repertoire are amplification of the shared voice of parties and CSOs and the mobilisation of unrepresented social claims (Sitrin & Azzellini, 2014).

  2.2 Expectations

  My study is guided by two general expectations. The first is that extensive change in the political system and society of Flanders will have shaped a pattern of connections in which informal connections predominate. Second, I expect to find evidence of path dependency in the extent of formal connections to CSOs, with the social democratic Vooruit retaining more extensive statutory links and higher organisational density than Groen because of Vooruit’s origins and traditional links to the region’s pillars.
  Table 1 presents a schematic summary of formal and informal forms of connection between political parties and CSOs.

  Table 1 Degrees of Formality: A Typology of Connections Between Parties and CSOs

  Type	Sub-types	Functions
  Formal	Organisational density	Strengthening partisan attachment of social groups; securing voter loyalty
  	Statutory links	Granting substantive organisational rights to linked groups
  	Multispeed membership	Attracting additional resources and connecting to new groups of supporters
  Informal	Regular contacts	Exchange of information and resources beneficial to parties and CSOs; development of public policy
  	Elite connections – connective density	Programmatic alignment between parties and membership based CSOs; stabilisation of party support at elections
  	Shared action repertoires	Amplification of shared voice of parties and CSOs; the mobilisation of unrepresented social claims

• 3 Research Design and Case Description

  To explore the research question, I conduct a comparative case study of two progressive political parties in the region of Flanders and consider the perspectives of a wide range of Flemish CSOs on the form and evolution of their connections with these and other parties in Flanders. Flanders represents an interesting context for the study of patterns of connection between parties and organised civil society for several reasons. First, the region represents a good example of a corporatist system in which larger organisations cooperate with each other and with public authorities to make and implement public policy (Lehmbruch, 1979). A key feature of the region’s corporatism are extensive consociational practices of policy formation and interest representation designed to contain and accommodate conflicts based around social, religious and linguistic cleavages (Andeweg, 2019).
  Second, the region has a history of strong formal links between parties in the Christian, Liberal and Socialist party families and institutions in civil society. For decades, Flanders’ political and civic life has been structured by pillarisation – institutions of political parties and networks of civic associations – that integrates Belgian citizens into ideational subcultures (Van Haute & Wauters, 2019). Each of the three main ‘pillars’ – socialist, Christian and liberal – has associated trade unions in both parts of the country, and, until the 1980s, pillarised parties dominated the country’s elections.²x A process of de-pillarisation has weakened the link between membership of pillar organisations and voting behaviour (Hellemans, 2020). Between 1985 and 2019, the combined vote share of socialist and Christian parties halved – from 58% to 29%.
  I focus on two progressive parties within the Flemish party system: Vooruit and Groen. These parties make an interesting comparison because they share a similar ideological profile and compete largely for the same subgroup of progressive Flemish voters³x Vooruit and Groen also contest elections in the predominantly French-speaking Brussels-Capital region. (Walgrave et al., 2020). Both parties have shared experience, between 1999 and 2003 and since 2019, as members of coalition governments at the federal level.⁴x Between 1999 and 2004, Vooruit shared office at the regional level with Groen, the latter under its original party name, Agalev. However, the two parties differ substantively in respect of their organisational traditions and historical relationship to civil society. Vooruit is the recently re-branded Flemish Socialist party that was founded in 1978 on the dissolution of the Belgian Socialist Party (BSP) into separate regional organisations. Historically, the party was firmly based in the socialist pillar with strong relationships with pillar institutions in civil society, but it has distanced itself from traditional allies in recent decades (Martin, 2023). Groen emerged in the 1980s from new social movements outside the pillarised system of interest representation. While Vooruit owes its origins to older cleavages based on social class, Groen is an example of a younger party giving expression to newer social cleavages centred on sociocultural dimensions (Ford & Jennings, 2020). The two parties are thus a most likely case for finding patterns of ‘path dependency’, which would generate differences between them in their respective patterns of connection to civil society.

  3.1 Data Collection

  The principal aim of this article is to describe the contemporary pattern of connections between political parties and CSOs at the level of the political system and to understand how that pattern has been shaped by political and civil actors navigating extensive change. To this end, I employ a mixed-methods approach to data collection and combine data from a novel dataset, party statues and semi-structured interviews with political and civil actors in Flanders.
  First, I operationalise statutory links as the number of formal links or associations between parties and either internal subgroups or external organisations defined in party statutes (Scarrow et al., 2023). To collect this information for Vooruit and Groen, I analysed key documentation, particularly current party statutes and party websites, for both parties – Groen last updated its party statutes in 2019 (Groen, 2019) and Vooruit in 2022.
  Second, to calculate estimates for organisational density and connective density, I draw from a novel dataset developed specifically for my research. The dataset comprises information on 152 political parties contesting 30 elections in 14 West European countries between 2005 and 2017. The dataset has a total of 36 variables at the level of political parties and 19 variables at the country-election level and draws extensively on a wide range of comparative datasets and statistical sources.
  I operationalise organisational density as an additive measure of two separate variables First, party-trade union support measuring the percentage of a party’s election vote from trade union members sourced from the European Social Survey (ESS, 2020), and second, the party membership ratio expressing party membership as a percentage of each party’s electorate at each election. Measures for both variables were standardised to have a mean of 0 and standard deviation of 1 and then added to form a composite measure. The mean value of organisational density for all parties in the dataset is 0.06 with a minimum of −2.08 and a maximum of 6.04.
  I operationalise connective density as the average number of CSO memberships of election candidates aggregated to the party level. I calculate the measure with data on candidates’ memberships of CSOs from the Comparative Candidates Survey (CCS, 2016, 2019). Equal weight in the measure was given to CSOs such as trade unions historically associated with left-wing parties and CSOs such as business associations historically associated with right-wing parties. The measure for connective density is a count variable with values ranging from 0 to 4. Parties exhibited considerable variation on the measure – the mean value for connective density was 1.09, and the maximum value for any party was 2.09.
  Next I operationalise multispeed memberships as evidence, in party documents and practices, of parties providing different formal means, such as membership and registered supporter status, for individuals to participate in the internal organisational life of the party. To collect information on this type of connection, I analysed party statutes and websites and drew on semi-structured interviews with party officials.
  Finally, I operationalise shared action repertoires as the joint deployment by parties and CSOs of forms of political action such as legislative activity within formal political institutions and both conventional and unconventional actions such as protests and policy forums that take place outside such institutions. My principal source of information on the presence of shared action repertoires was material from semi-structured interviews with political and civil actors in Flanders, supported by analyses in primary and secondary publications.

  3.2 Semi-Structured Interviews

  To reach a full answer to my research question, I collected rich data by means of semi-structured interviews. I sought a representative balance of current and former officials of both parties; of representatives of a diverse range of CSOs, including those linked to each party; and of country experts and commentators. I identified interviewees initially by means of information in the literature on country experts and on the websites of the two political parties. This initial sample was supplemented by snowball sampling and the identification of additional respondents through secondary research material. Interviewees were contacted by means of an email communication, describing the aims of the study, indicating the likely length of an interview and confirming anonymity and confidentiality of data collected in the interview. Interviews were conducted in English and lasted for between 30 and 90 minutes. Because of restrictions related to COVID-19, most interviews, 22, were conducted online. An anonymised list of the interviewees is produced in Appendix A.
  In Appendix D to the manuscript, I provide summary information on the number of people contacted, the response rate achieved and the composition of the final sample. I worked with a standard set of core questions but customised interview schedules to capture data on the specific experience and perspective of participants. In Appendix D, I reproduce schedules for interviews conducted with respectively party officials and representatives of CSOs.

  3.3 Analysis of Qualitative Data

  The interview transcripts ran to over 23,000 words, and, to structure my analysis of this qualitative data, I used thematic analysis (Braun & Clarke, 2021). Thematic analysis is well suited to this task because it helps capture the essence of rich interview data bringing out both commonalities and differences among participants. I conducted a two-stage coding of my interview data. First, I coded interview transcripts inductively identifying over 150 initial codes. Second, I reviewed all the initial codes and searched for themes that would help address my research question. A codebook reporting, describing and illustrating the themes is provided in online Appendix B.

  3.4 Triangulation

  I triangulated the data from interviews with a comprehensive review of literature and other secondary documentation. The review included scholarly literature, articles and commentary in newspapers and blog sites, party and CSO documents and websites, and other political commentary. Details and examples of sources used to triangulate my interview data are provided in online Appendix E.

  3.5 Timeframes

  The timeframes covered by data sources for different types of connection are not completely coterminous. For example, data on statutory links between parties and CSOs are sourced from the most recently updated statues of Vooruit and Groen, and measurements for both organisational density and connective density are calculated are based on data for the Belgian general elections of 2007, 2010 and 2014. Respondents who were interviewed for the study were invited to describe the contemporary pattern of connections between Flemish parties and CSOs, to compare them with dominant patterns in the past and to explain how contemporary forms of connection evolved in response to change.

• 4 Results and Case Analysis

  I first present the results of a mapping of the two parties’ statutory links, organisational density and connective density. I then present my findings on evidence of multispeed party membership and shared action repertoires. I complete this section of the article with a detailed thematic analysis of the qualitative data collected through semi-structured interviews.

  4.1 Statutory Links

  The scale and scope of the statutory links of both parties are similar and higher than the average for their respective party families (Allern & Verge, 2017). Vooruit has statutory links with three sub-organisations (women, young people and senior citizens) and Groen has statutory links with two (young people and senior citizens, but not women). Until November 2022, when it revised party statutes, Vooruit guaranteed representation for the socialist trade union the ABVV and mutual insurance provider Solaris on its central party board.⁵x The change was made because of a fear that infiltration of the ABVV and Solaris by militants from the radical left Partij van der Arbeid (PVDA) would result in members of the PVDA sitting on the Vooruit central board (De Tijd, 22 November 2022). Vooruit also funds an international solidarity organisation with partners from the socialist pillar and has two ‘sister’ organisations in the cultural field. These data are consistent with expectations (see Allern & Verge, 2017) that social democratic parties retain more statutory links while newer parties, including Green parties, have less frequent statutory links and, in particular, fewer links with women’s sub-organisations.

  4.2 Organisational Density

  Table 2 summarises data for organisational density and shows that Vooruit’s organisational density remained very stable and above the mean for all parties in the dataset throughout this period, reflecting a continuing high level of support from individual members of trade unions. By contrast, the organisational density of Groen fell across each election to a level substantially below that of the mean for all parties in the dataset.

  Table 2 Organisational Density: Vooruit and Groen 2007-2014

  Party	2007	2010	2014
  Vooruit	1.16	1.10	1.20
  Groen	0.60	0.10	−0.58

  Note: Organisational density has a mean value of 0.064 for all parties in the dataset.

  4.3 Connective Density

  Table 3 summarises data for connective density and shows that at the elections of 2007 and 2010 both parties had nearly identical levels of connective density. However, their connective density fell sharply by the 2014 election, by a third in the case of Vooruit and by more than half in the case of Groen. Analysis of the types of organisation that make up connective density (Appendix C) indicates that the fall in connective density for Vooruit was entirely due to falling connections to socioeconomic groups, as the party distanced itself from traditional allies (Martin, 2023), while Groen experienced falling connections to all categories of CSO.

  Table 3 Connective Density: Vooruit and Groen 2007-2014

  Party	2007	2010	2014
  Vooruit	1.04	0.98	0.65
  Groen	1.05	1.07	0.47

  Note: Connective density is a count measure with values ranging from 0 to 4. The mean value of connective density is 1.09 for all parties in the dataset.

  In sum, I find that Vooruit has stronger formal connections to CSOs than Groen reflecting its origins in the socialist pillar. However, it exhibits a low level of connective density, its elites having progressively withdrawn from membership of socioeconomic organisations. While Groen maintains substantial statutory links, it too has a low level of connective density reflecting the steady withdrawal of its candidates from organised civil society.

  4.4 Multispeed Membership and Shared Action Repertoires

  My analysis of primary and secondary source material provided limited evidence that either party has developed forms of multispeed membership. Vooruit offers the option of becoming an ‘Online Ambassador’, though this confers no organisational status and is limited to receiving digital communications from the party. Similarly for Groen, recent changes have been confined to expanding members’ voting rights in elections for party President to include online participation alongside members voting in person at the party Congress. While studies suggest that new social movements in Belgium have drawn large numbers of new participants into the protest arena, these actions have not taken the form of shared repertoires (e.g. Wouters et al., 2022). Neither party directly co-organises protests with CSOs nor participates in citizens’ forums or solidarity initiatives.

  4.5 Thematic Analysis

  In this section, I present the detailed thematic analysis of material collected through semi-structured interviews. I present five specific themes that illustrate different elements of the typology of parties’ connections to CSOs introduced in this article. In particular, the focus is on the form and substance of informal contacts, strategies guiding relationships between civil and political actors, and mechanisms shaping an evolved pattern of connection between parties and CSOs in Flanders.
  The first theme – older and new social movements – describes the impact that the emergence of new social movements (NSM) from the 1980s had on the connections between parties and CSOs. For the founders of NSMs in Flanders, traditional CSOs that were closely linked to pillar structures were perceived as “too close to the parties … and only interested in power, they were conservative and did what the parties wanted them to do” (In. 11). Newer civic and multicultural associations introduced a new dynamic into the relationship between CSOs and parties. Accustomed to a tradition of “incorporating CSOs into exclusive relationships” (In. 8), Flemish political parties faced the challenge of establishing relationships with more openly critical and independent civic groups.
  In respect of retaining connections to traditional pillar institutions, both parties were cautious and wary. For Vooruit, its associations with traditional pillar organisations were partly responsible for the party’s electoral decline, prompting it to distance itself from traditional allies. The party is ambivalent about a continuing association with trade unions, with its attitude depending, as one senior trade union official put it, “… on who you talk to within Vooruit. Some want strong connections, but others have a third way perspective and attach blame to trade unions” (In. 15). For Groen, while the party has edged towards a closer relationship to trade unions, the relationship is characterised by mutual wariness. For the party, there is an enduring sense that unions are “focused on keeping their traditional power base” (In. 11), while for unions, the perception is that Groen’s closeness to unions’ positions varies with their government status (In. 15). Notwithstanding these tensions, one trade union official observed that “More trade unionists are joining and active in the Greens giving us a de facto link and influence over their programme” (In. 28).
  The emergence of regionally based social movements pressing new claims (Willems et al., 2021) has prompted the two parties to adopt a more open attitude towards engagement with CSOs and to deploy fewer exclusive channels of interaction. While this may have been a relatively straightforward task for Groen because, as one former official put it, “We are ourselves a product of de-pillarisation and have never expected exclusive relationships, and recognise that CSOs will talk to other parties” (In. 8), Vooruit have followed suit. As one Vooruit official indicated, “In developing policy we try to be open and participative … we try to reach new groups and not just CSOs that are formally constructed” (In. 1). There has been a change in the party’s style of engagement with CSOs from structured formal ties to a more discretionary approach to contacts with local CSOs aimed at “stimulating issue-based engagement on common areas” and, in doing so, “outside the old structure of local party committees for fear of driving them [local CSOs] away” (In. 19).
  Respondents reported a very wide range of informal contacts, including exchanges of research reports (In. 10, 14) and policy proposals (In. 2, 5 and 13), invitations to events and conferences (In. 2 and 8) and meetings to draft legislation and parliamentary motions (In. 6 and 10). Exchanges were initiated by both parties and CSOs and centred largely on policy issues. During election periods, these contacts shift to advocacy and information exchange focused on “exploring and influencing party positions and influencing party manifestos” (In. 6). During the 2019 general election, one large CSO organised Facebook live events at which Party Presidents were invited to “talk about their election programmes … with the aim of getting our members and followers to know about their positions” (In. 6). These processes continue post-election into negotiations over coalition formation with CSOs “writing lobby letters” on behalf of their members (In. 10) and working directly with party officials to get “proposals adopted and included in the coalition agreement” (In. 5). Personal relationships and policy alignment are perceived as important by both parties and CSOs. For the official of a CSO working on migration policy, general political alignment with Groen and Vooruit means, “we are in more of a real working relationship with these parties, where we are working to strengthen each other” (In. 10).
  For several decades, representatives of economic interests in Flanders enjoyed privileged status as social partners in government policymaking. More recently, CSOs who emerged out of NSMs have become important players in policymaking institutions, and the second theme – policy structures and competences – describes how the two parties and mainstream CSOs are now closely connected through the framework of government policymaking. The focus of mainstream CSOs on advocacy and lobbying directed at ministerial cabinets has become key to their connections with parties: “We mainly start with a report or policy letter or summary with our positioning that is directed to cabinets, responsible ministers and their vice-chairs” (In. 13). CSOs follow up on these initiatives by “meeting with Cabinet staff” (In. 12) and “talking to them extensively online” (In. 5). The government status and ministerial responsibilities of parties clearly influence the frequency and substance of connections. As one CSO official described it, “Being in government is important for contacts as there is a clear hierarchy of cabinets, parliament and administration” (In.13).
  The entry of new civic actors into policymaking institutions has strengthened the tendency of the two parties to seek connections beyond those organisations with whom they were traditionally close. This has presented both parties with challenging trade-offs. For Groen, frequent institutional interaction with mainstream environmental CSOs risks alienating more radical environmental movements that may see the party as “not natural translators of their demands” (In. 8). For Vooruit, new relationships within policymaking institutions have accelerated “its withdrawal from the social partners, the party preferring instead consultation with external experts” (In. 28). This has made Vooruit vulnerable to the charge, levelled, for example, by a senior national figure in the party, that in substantive terms, it “no longer connects to its base in civil society … its officials moving in social circles that don’t connect with ordinary people” (In. 21). Established CSOs also acknowledge that their ‘institutionalisation’ may also mean that, as one trade union official put it, “We have missed the boat on the issues they [new protest movements] bring forward” (In. 28).
  A third theme – pragmatism and independence – summarises the approach that contemporary civil society takes to connecting with all Flemish political parties. Officials of CSOs stressed that their approach to connections with parties was to remain independent. Despite policy closeness, an official with a leading environmental CSO commented:

  This does not mean we will align our message exclusively with Groen but also with other parties when they have similar positions to us … on nature conservation for example we are active in talking to the [nationalist] N-VA. (In. 6)

  One consequence is that established CSOs will almost never engage in joint protests or campaigns with a single party despite having shared interests. As one officer in an environmental CSO put it, “We would never work with a political party on a shared campaign as we don’t want an automatic link” (In. 6).
  The approach of CSOs to seeking policy influence through connections with parties is highly pragmatic. As agents of policy influence, CSOs regard themselves as “not serious if we don’t engage with the Liberals and Conservatives who hold ministerial positions” (In. 13). Connections are guided by the goal of making an impact through “parties accepting concrete proposals. We applaud small steps in the right direction” (In. 10). The two parties have accepted and accommodated CSOs’ desire to remain independent of any single party. When it comes to cooperation on campaigns and communication with CSOs, for example, Vooruit acknowledge that CSOs want to keep their identity hidden. As one Vooruit official commented, “they [the CSO] will want to keep their logo out of the material, they want to keep their distance” (In. 1). And similarly, for Groen, a party official observed:

  It is a habit for Groen politicians to have regular meetings with CSOs and to organise events to which CSOs are invited and speak. CSOs know that Groen’s doors are always open. But the relationship is not an exclusive one. (In. 8)

  The inclusive approach of CSOs to relationships with parties does not, however, extend to radical challenger parties. All CSO interviewees confirmed that contacts with political parties explicitly excluded the radical right Vlaams Belang. While a cordon sanitaire did not apply to the radical left PVDA, CSO officials were wary of getting too close to the party because “they reject compromise and this rules out progress” (In. 13).
  The approach of the two parties to connections with civil society constitutes the fourth theme – party strategies. The declared strategy of Vooruit is to “bypass older CSOs to enable smaller CSOs and voters to participate in party decision-making” (In. 16). A key tool in this strategy is the party’s Big Shift, a campaign combining open online voting to determine key programmatic issues and ‘hackathons’ to explore the detail of chosen themes. For many CSO actors, the party’s strategy is ambiguous reflecting the aim of the party’s President Conner Rousseau to ‘cover all the bases’ (In. 6). A strong digital focus promoting “brand Conner in comparison to the unpopular brand socialist” (In. 18) is contrasted with the work of M.P.’s on policy dossiers who “are interested in our themes but are working on their own initiative rather than with party backing” (In. 14).
  For both parties, the challenge of balancing commitments to coalition government with centre-right parties and the progressive aspirations of civil actors impacts the closeness of their connections. For Vooruit, while the party is very “proud of being a government party focused on good policy that benefits everyone” (In. 27), being in government has necessitated making “a lot of agreements that it has to defend even if they would be critical if outside government” (In. 28). For Groen, these challenges are particularly acute. The party has, for example, taken responsibility for government decisions to reverse plans to close nuclear plants, drawing intense criticism from CSOs. As one former Groen government minister put it, “My focus was on building the party in government, but we forgot what the movements were thinking.… We had become wrapped up in the institutions” (In. 9).
  With some exceptions, for example, the occasional use of Vooruit Party President Conner Rousseau’s popular Instagram account to promote posts by CSOs (In. 1), and the recent extension of voting rights for the election of Groen co-Presidents to members voting online (In. 3), I found limited evidence that either party used multispeed memberships and digital activity as part of a strategy of connecting to organised civil society as opposed to simply as a communication tool.
  Alongside stressing independence in its relationships with the two parties, an increasingly diverse civil society has also evidenced a trend over the last decade for interest representation to be channelled through umbrella organisations. This constitutes a fifth theme – coalitions of representation. These coalitions endeavour to increase collaboration and unity between established and newer CSOs.

  We gather in organisations from many sectors of civil society. On a series of topics, where there has not been alignment … the coalition is trying to forge joint positions and a joint memorandum. (In. 6)

  Umbrella organisations not only represent CSOs directly to political parties but also seek to resource and enhance the efforts of their members to influence parties locally and at the federal level. “We want to be a voice for the people … connecting grassroots organisations who did not know about each other so that they can be a strong force together” (In. 12).
  Many umbrella coalitions have a broad organisational membership, and the goal of attaining greater balance and credibility in the representation of members’ interests was stated explicitly: “Our new organisation means we are more independent than in the past when we were dominated by the unions” (In. 14) The broad base of umbrella coalitions combined with their independence – “We would definitely never make recommendations to our members and supporters about who to vote for” (In. 6) – makes it easier for the parties to maintain connections with them (also In. 5 and 14).
  Table 4 summarises my analysis of the pattern of connections between parties and CSOs in the region of Flanders. I find that despite some differences between the two parties in respect of connections to CSOs that reflects a legacy of the region’s system of pillarisation, Vooruit and Groen have navigated change by converging on similar patterns of, and perspectives on, connections to CSOs. For both parties, connections are based predominantly on informal contacts focused on policymaking and information exchange, and both parties accept the nonexclusive basis of these connections. Similarly, while established CSOs are often closer to the two parties in terms of values and policy positioning, they adopt a standard playbook in pursuing objectives within the politicised system of institutional policymaking. The strategy of CSOs is instrumental and focused on securing marginal policy gains. And the power status of Vooruit and Groen as government partners with a strong presence in the region’s institutions of public policymaking strengthens the parties’ connections with organised civil society at all points of the electoral cycle.

• 5 Discussion and Conclusions

  I ask how parties and CSOs navigate their relationships in the context of profound change in party systems and civil society. To answer this question, I develop a typology of formal and informal party-societal connections and undertake a comparative case study of the connections between two progressive parties in Flanders and a cross-section of CSOs. I find that these connections have converged on a very similar structure of relationships with organised civil society and explain this convergence in terms of the politicisation of public policymaking in Flanders and the pragmatic approach of mainstream civil society to representing the claims of their constituencies.

  Table 4 Patterns of Connection Between Parties and CSOs in Flanders

  Type	Data Source	Main Findings
  Formal
  Parties’ statutory links	Party statutes; party websites	Extensive for both Vooruit and Groen; exhibit the legacy of the region’s history of pillarisation
  Organisational density	Comparative Candidates Survey; novel dataset	High and above average for Vooruit; declining and below average for Groen
  Multispeed membership	Party documents and websites; semi-structured interviews	No evidence of development beyond the use of digital tools for communication and election of party President (Groen)
  Informal
  Regular contacts	Semi-structured interviews	Extensive nonexclusive informal contacts between CSOs and parties within policymaking institutions and at bilateral level; accentuated by expanding coalitions of representation; excludes radical civil and political actors
  Connective density	Comparative Candidates Survey; novel dataset	Below average and declining for both Vooruit and Groen
  Shared action repertoire	Semi-structured interviews; secondary documentation	No evidence that mainstream civil and political actors share action repertoires

  I find that while both parties retain substantial statutory links and Vooruit has a legacy of relatively high organisational density, for both parties connections to CSOs have become dominated by informal contacts of varying degrees of regularity. In substantive terms, these contacts facilitate the exchange of information and resources that are beneficial to both parties and CSOs. Through regular contacts, both types of actors exchange research information and policy ideas and information on each other’s positions and aspirations that helps align parties with members of CSOs.
  These informal connections occur both within the highly politicised institutions of policymaking and in the form of direct bilateral interaction. Contacts are initiated by both civil and political actors and are nonexclusive – CSOs seek contacts with all mainstream parties and are highly pragmatic in their aims. Institutional structures of interaction between Flemish political parties and CSOs have adapted over recent decades to incorporate both newer social movements in the field of the environment and migration and new political competitors from green and nationalist party families. However, new protest-based social movements and challenger parties of the radical right and left largely fall outside this framework.
  Vooruit and Groen are both governing parties, and my study confirms findings (Martin et al., 2022b) that power status influences the intensity of their connections to CSOs in a number of ways. The extensive presence within public policymaking institutions of ministers and advisers from both parties incentivises mainstream CSOs to maintain routine and regular communication with party elites on the policy concerns of their members. The entry of new parties and CSOs into policymaking institutions in recent years has, however, weakened the traditionally privileged position of social partners and contributed to the non-exclusive nature of contacts between parties and CSOs. I found ample evidence from semi-structured interviews that CSOs make pragmatic decisions to connect to ministers from liberal, conservative and nationalist parties. It is sometimes also the case that the two parties’ responsibilities to coalition partners drawn from the centre and centre-right result in strains in the parties’ relationships with those CSOs to whom they are closely aligned programmatically.
  The central roles played by informal contacts and institutionalised relationships and the determination of CSOs to remain independent of any single party are responsible for the near absence of any shared repertoires of action and the comparatively weak and declining ties (connective density) of party elites to civil society. I found limited evidence that, despite very high levels of social media expenditure,⁶x In 2021, Belgian parties spent €4.71 million on Facebook advertising. Excluding parties contesting general elections in other countries in 2021, three Flemish parties – N-VA, VB and PVDA – spent more on Facebook adverts than any other party in the EU (https://adlens-be.medium.com/). the parties deploy different forms of multispeed membership to connect with organised civil society.
  This article makes an important contribution to our understanding of how parties and CSOs navigate their interactions in a time of change and has a number of important implications. First, by providing insights on the mechanisms that have shaped the pattern of connections at the level of the regional political system. The weakening of structures and cleavages associated with pillarisation and industrialisation; the emergence of new, and often regionally based, social movements; and the incorporation of new social and civic actors within policymaking institutions have each contributed to a pattern of relationships structured around regular and informal interactions. And this article has identified two additional factors that play a significant part in the pattern of connections observed: - the influence of policymaking institutions and the strategies of both parties and civic groups in pursuing their goals within those institutions.
  Second, by identifying and mapping a full typology of connections at the level of the Flemish political system, my study provides a framework for further comparative research into how patterns of connection may be shaped in political systems characterised by different histories and structures of interest representation. This typology of connections is also helpful for understanding the form and resilience of interest representation in a political system. I found a vibrant pattern of informal contacts between parties and CSOs that faced upwards to institutions of policymaking and downwards to grassroots CSOs that are integrated into representative structures through coalitions of interest representation. Respondents from both parties and CSOs were positive about the outcomes of this system and its capacity to respond to emerging social claims.
  Third, my study suggests that the evolution of pragmatic, informal and institutionally focused connections by mainstream actors while integrating many social actors into policymaking processes has also created ‘outsiders’ – new protest-driven CSOs and challenger parties of the radical right and left. The exclusion of the growing constituencies of these parties and radical social movements from structures of policy formation raises interesting questions about these structures’ legitimacy and stability. Further it suggests that the challenges of forming government coalitions based on compromise between political actors – a compromise that underpins the instrumental value of institutional structures to mainstream actors – are likely to intensify.
  My study has a number of limitations. First, my study focuses on just one political sub-system with a highly specific set of institutional structures built on a history of pillarised interests and consociational conflict resolution (Andeweg, 2019). While other countries share some of the features of the Flemish system, features such as strong parliaments, pluralist systems of interest representation and unitary state governance are more common in Western European countries. Second, while my interviews with CSOs demonstrated their engagement with all mainstream parties, my detailed analysis focused on the role of just two progressive parties with a history of closeness to social movements of different types. This may have resulted in a skewed picture of the pattern of connections I was able to describe.
  I can begin to address these limitations by suggesting some topics for future research. First, the utility of the typology of connections should be extended, for example, by research into patterns of connection in systems with pluralist systems of interest representation and strong roles for parliaments in legislative policymaking. Exploratory case studies could also consider the pattern of connections through the lens of parties of the centre-right and the developing action repertoires of parties that have emerged from social movements in the last decade, bringing to light other forms of connection prevalent in specific political systems.
  Second, I observed that the comparative absence of some forms of connection in Flanders, such as shared action repertoires, are a consequence of the prevalence of other forms of connection between both mainstream political and civil actors. This suggests that adopting an ecological approach to the study of patterns of connections could be instructive. An ecological approach would explicitly consider ways in which different combinations of relationship act on, and interact with, each other to create a dynamic pattern of connections within a political system. Describing and comparing the ecology of connections in different political systems could lead to new theoretical insights on the form and resilience of interest representation in modern democracies.
  Third, my study of the Flemish case describes patterns of connection that are largely well adapted to traditions of compromise between political interests and to the goal of incremental policy change. However, mainstream political and civic actors in Western Europe face increasingly strong challenges from radical actors who express impatience with incremental change and intensify conflict around issues that are less embedded in the agenda of institutional insiders. How do challenger parties with growing constituencies that reject, or are rejected by, mainstream actors develop their organisation and strategy to represent parts of civil society and mobilise emerging interests? And how best can mainstream actors respond to these challenges to ensure that systems of interest representation adapt to emerging issues and political claims, and retain legitimacy with citizens?
  The form of party cartelisation described by Mair (2013) is one in which the goals of parties become self-referential as they retreat into the state while citizens head “for the exits of the national political arena” (2013, p. 43). In contrast, my study of the Flemish political subsystem points to extensive contacts and relationships between parties and organised civil society that channel representative claims from both regional and grassroots CSOs and ensure that the policy system responds, at least partially, to new claims. To retain their relevance and legitimacy in the face of persistent economic, political and environmental challenges, systems of representation will need to find new ways to adapt and respond. I suggest that the study of the diverse patterns of connection between parties and CSOs within political systems has the potential to generate many new insights on how both sets of actors contribute to meeting these challenges.

References

• Allern, E. H. (2010). Political Parties and Interest Groups in Norway. ECPR Press.

• Allern, E. H., Kluver, H., Marshal, D., Otjes, S., Rasmussen, A., & Witko, C. (2021a). Policy Positions, Power and Interest Group-Party Lobby Routines. Journal of European Public Policy, 29(7), 1029-1048. https://doi.org/10.1080/13501763.2021.1912148.

• Allern, E. H., Otjes, S. P., Poguntke, T., Hansen, V., Saurigger, S., & Marshall, D. (2021b). Conceptualising and Measuring Party-Interest Group Relationships. Party Politics, 27(6), 1254-1267. https://doi.org/10.1177/1354068820949393.

• Allern, E. H., & Verge, T. (2017). Still Connecting with Society? Political Parties Formal Links with Social Groups in the Twenty-First Century. In S. E. Scarrow, P. Webb & T. Poguntke (Eds.), Organising Political Parties: Representation, Participation and Power (pp. 106-135). Oxford University Press.

• Andeweg, R. B. (2019). Consociationalism in the Low Countries: Comparing the Dutch and Belgian Experience. Swiss Political Science Review, 25(4), 408-425. https://doi.org/10.1111/spsr.12361.

• Bartolini, S. (2000). The Class Cleavage: The Political Mobilisation of the European Left 1860-1980. Cambridge University Press.

• Bartolini, S., & Mair, P. (1990). Identity, Competition and Electoral Availability: The Stabilisation of European Electorates 1885-1985. ECPR Press.

• Berkhout, J. (2013). Why Interest Organizations Do What They Do: Assessing the Explanatory Potential of ‘Exchange’ Approaches. Interest Groups & Advocacy, 2, 227-250. https://doi.org/10.1057/iga.2013.6.

• Beyers, J., Eising, R., & Maloney, W. (2008). Researching Interest Group Politics in Europe and Elsewhere: Much We Study, Little We Know? West European Politics, 31(6), 1103-1128. https://doi.org/10.1080/01402380802370443.

• Blings, S. (2020). Niche Parties and Social Movements: Mechanisms of Programmatic Alignment and Party Success. Government and Opposition, 55, 220-240. https://doi.org/10.1017/gov.2018.18#.

• Borbáth, E., Hutter, S., & Leininger, A. (2023). Cleavage Politics, Polarization, and Participation in Western Europe. West European Politics, 46(4), 631-651. https://doi.org/10.1080/01402382.2022.2161786.

• Bornschier, S. (2009). Cleavage Politics in Old and New Democracies. Living Reviews in Democracy, 2009, 1-13. https://doi.org/10.5167/uzh-26412.

• Braun, V., & Clarke, V. (2021). Thematic Analysis: A Practical Guide. Sage Publications.

• Comparative Candidates Survey (2016). Comparative Candidates Survey Wave I - 2005-2013 [Dataset - Cumulative File]. FORS. https://doi.org/10.23662/FORS-DS-620-3.

• Comparative Candidates Survey (2019). Comparative Candidates Survey Wave II - 2013-2018 [Dataset - Cumulative File]. FORS. https://doi.org/10.23662/FORS-DS-886-3.

• Della Porta, D., Fernandez, J., Kouki, H., & Mosca, M. (2017). Movement Parties against Austerity. Polity Press.

• Duncan, F. (2015). Preaching to the Converted? Christian Democratic Voting in Six West European Countries. Party Politics, 21(4), 577-590. https://doi.org/10.1177/1354068813487118.

• Emanuele, V. (2015). Dataset of Electoral Volatility and Its Internal Components in Western Europe (1945–2015). Italian Center for Electoral Studies. http://doi.org/10.7802/1112.

• European Social Survey Cumulative File, ESS 1-9 (2020). Data file Edition 1.0. Sikt - Norwegian Agency for Shared Services in Education and Research, Norway, Norway - Data Archive and distributor of ESS data for ESS ERIC. www.europeansocialsurvey.org/data/?round=9.

• Ford, R., & Jennings, W. (2020). The Changing Cleavage Politics of Western Europe. Annual Review of Political Science, 23, 295-314. https://doi.org/10.1146/annurev-polisci-052217-104957#.

• Fuchs, D., & Rucht, D. (2018). Support for New Social Movements in Five Western European Countries. In C. Rootes & H. Davis (Eds.), Social Change and Political Transformation (pp. 86-111). Routledge.

• Gauja, A. (2015). The Construction of Party Membership. European Journal of Political Research, 54(2), 232-248. https://doi.org/10.1111/1475-6765.12078.

• Grande, E. (2022). Civil Society, Cleavage Structures and Democracy in Germany. German Politics, 32(3), 420-439. https://doi.org/10.1080/09644008.2022.2120610.

• Groen (2019). Statuten. www.groen.be/partijstatuten.

• Gunther, R. (2005). Parties and Electoral Behaviour in Southern Europe. Comparative Politics, 37(3), 253-275. https://doi.org/10.2307/20072889.

• Gunther, R., & Diamond, L. (2003). Species of Political Parties: A New Typology. Party Politics, 9(2), 167-199. https://doi.org/10.1177/135406880300920.

• Hellemans, S. (2020). Pillarization (‘Verzuling’). On Organised ‘Self-Contained’ Worlds in the Modern World. The American Sociologist, 51, 124-147. https://doi.org/10.1007/s12108-020-09449-x.

• Kayser, M. A., & Wlezien, C. (2011). Performance Pressure: Patterns of Partisanship and the Economic Vote. European Journal of Political Research, 50(3), 365-394. https://doi.org/10.1111/j.1475-6765.2010.01934.x.

• Kelly, J. (2015). Trade Union Membership and Power in Comparative Perspective. The Economic and Labour Relations Review, 26(4), 526-544. https://doi.org/10.1177/103530461561444.

• Kitschelt, H. (1994). Austrian and Swedish Social Democrats in Crisis: Party Strategy and Organization in Corporatist Regimes. Comparative Political Studies, 27(1), 3-39. https://doi.org/10.1177/0010414094027001001.

• Kosiara-Pedersen, K., van Haute, E., & Scarrow, S. E. (forthcoming). Multispeed in Practice: Generating Linkage through Membership and Social Media Following.

• Lehmbruch, G. (1979). Liberal Corporatism and Party Government. In P. Schmitter & G. Lehmbrug (Eds.), Trends Towards Corporatism Intermediation. (pp. 147-184) Sage Contemporary Political Sociology.

• Lisi, M. (2013). Rediscovering Civil Society? Renewal and Continuity in the Portuguese Radical Left. South European Society and Politics, 18(1), 21-39. https://doi.org/10.1080/13608746.2012.757450.

• Mair, P. (2013). Ruling the Void: The Hollowing of Western Democracy. Verso.

• Maloney, W. A. (2012). The Democratic Contribution of Professionalized Representation. In J. W. van Deth & W. A. Maloney (Eds.), New Participatory Dimensions in Civil Society: Professionalization and Individualized Collective Action. (pp.84-96) Routledge.

• Martin, N., De Lange, S.L, & van der Brug, W. (2022a). Holding on to voters in volatile times: Bonding voters through party links with civil society. Party Politics, 28(2), 354-364. https://doi.org/10.1177/1354068820980304

• Martin, N., De Lange, S.L, & van der Brug, W. (2022b). Staying connected: explaining parties’ enduring connections to civil society. West European Politics, 45(7), 1385-1406. https://doi.org/10.1080/01402382.2021.1986784

• Martin, N. (2023). Remaining close or drifting apart? Explaining the divergent paths taken by Belgium’s socialist parties in connecting to civil society. Acta Politica, 59, 289-319. https://link.springer.com/article/10.1057/s41269-023-00293-8

• Mudge, S. (2018). Leftism Reinvented: Western Parties from Socialism to Neoliberalism. Harvard University Press.

• Otjes, S., & Rasmussen, A. (2017). The Collaboration between Interest Groups and Political Parties in Multi-Party Democracies: Party System Dynamics and the Effect of Power and Ideology. Party Politics, 23(2), 96-109. https://doi.org/10.1177/1354068814568046.

• Panebianco, A. (1988). Political Parties: Organisation and Power. Cambridge University Press.

• Poguntke, T. (2002). Parties without Firm Social Roots? Party Organisational Linkage. In K. R. Luther (Ed.), Political Parties in the New Europe (pp. 43-62). Oxford University Press.

• Rahat, G. (2007). Candidate Selection: The Choice Before the Choice. Journal of Democracy, 18(1), 157-170. https://doi.org/10.1353/jod.2007.0014.

• Rasmussen, A., & Lindeboom, G.-J. (2013). Interest Group-party Linkage in the Twenty-first Century: Evidence from Denmark, the Netherlands and the United Kingdom. European Journal of Political Research, 52, 264-289. https://doi.org/10.1111/j.1475-6765.2012.02069.x.

• Romeijn, J. (2020). Lobbying during Government Formations: Do Policy Advocates Attain Their Preferences in Coalition Agreements? West European Politics, 44(4), 264-289. https://doi.org/10.1080/01402382.2020.1755515.

• Rootes, C. A. (2000). Environmental Movements and Green Parties in Western and Eastern Europe. In M. Redclift & G. Woodgate (Eds.), The International Handbook of Environmental Sociology (pp. 319-348). Edward Elgar Publishing.

• Scarrow, S. E. (2014). Beyond Party Members. Changing Approaches to Partisan Mobilization. OUP.

• Scarrow, S. E. (2019). Multi-Speed Parties and Representation: The Evolution of Party Affiliation in Germany. German Politics, 28(2), 162-182. https://doi.org/10.1080/09644008.2018.1496239.

• Scarrow, S. E., Wright, J. M., & Gauja, A. (2023). Party Statutes and Party Institutionalization. Party Politics, 29(2), 217-228. https://doi.org/10.1177/13540688211071066.

• Sitrin, M., & Azzellini, D. (2014). They Can’t Represent Us! Reinventing Democracy from Greece to Occupy. Verso.

• Statsch, P., & Berkhout, J. (2020). Lobbying and Policy Conflict: Explaining Interest Groups Promiscuous Relationships to Political Parties. Interest Groups and Advocacy, 9(1)#, 1-20. https://doi.org/10.1057/s41309-019-00072-x.

• Van Biezen, I., Mair, P., & Poguntke, T. (2012). Going, Going…Gone? The Decline of Party Membership in Contemporary Europe. European Journal of Political Research, 51(1), 24-56. https://doi.org/10.1111/j.1475-6765.2011.01995.x.

• Van Haute, E., & Wauters, B. (2019). Do Characteristics of Consociational Democracies Still Apply to Belgian Parties? Politics of the Low Countries, 1, 6-26. https://doi.org/10.5553/PLC/258999292019001001002.

• Verge, T. (2012). Party Strategies towards Civil Society in New Democracies: The Spanish Case. Party Politics, 18(1), 45-60. https://doi.org/10.1177/1354068811422648.

• Walgrave, S., van Erkel, P., Jennart, I., Lefevre, J., & Baudewyns, P. (2020). How Issue Salience Pushes Voters to the Left or to the Right. Politics of the Low Countries, 2(3), 319-352. https://doi.org/10.5553/PLC/258999292020002003005.

• White, J., & Ypi, L. (2016). The Meaning of Partisanship. OUP.

• Willems, E., Beyers, J., & Heylen, F. (2021). Interest Representation in Belgium. Mapping the Size and Diversity of an Interest Group Population in a Multi-layered Neo-Corporatist Polity. Politics of the Low Countries, 3(1), 32-72. https://doi.org/10.5553/PLC/.000006.

• Wouters, R., De Vydt, M., & Staes, L. (2022). Truly Exceptional? Participation in the Belgian 2019 Youth for Climate Protest Wave. Politics of the Low Countries, 3, 229-249. https://doi.org/10.5553/PLC/.000029.

Noten

• 1 More details on the operationalisation and measurement of connective density are provided in the following section and in Appendix C.

• 2 A process of de-pillarisation has weakened the link between membership of pillar organisations and voting behaviour (Hellemans, 2020). Between 1985 and 2019, the combined vote share of socialist and Christian parties halved – from 58% to 29%.

• 3 Vooruit and Groen also contest elections in the predominantly French-speaking Brussels-Capital region.

• 4 Between 1999 and 2004, Vooruit shared office at the regional level with Groen, the latter under its original party name, Agalev.

• 5 The change was made because of a fear that infiltration of the ABVV and Solaris by militants from the radical left Partij van der Arbeid (PVDA) would result in members of the PVDA sitting on the Vooruit central board (De Tijd, 22 November 2022).

• 6 In 2021, Belgian parties spent €4.71 million on Facebook advertising. Excluding parties contesting general elections in other countries in 2021, three Flemish parties – N-VA, VB and PVDA – spent more on Facebook adverts than any other party in the EU (https://adlens-be.medium.com/).