The relationship between a party leadership and party activists is often illustrated by the use of John May’s "Law of curvilinear opinion structures in political parties", i.e. the idea that mid-level party activists are assumed to be more radical than both party sympathisers and the party elite. This tension between party leadership and party activists can be assumed to lead to a restriction on the party leadership’s freedom of action. However, in recent years we have seen a number of examples where political parties have made major realignments of their own policies without any substantial internal conflicts. Given the role assigned to conditions for involvement in the formulation of May's Law it appears reasonable to look in that direction when trying to understand these changes. This study covers the Moderate Party, the Social Democratic Party and the Green Party in Sweden. The offer made by the parties to their members can be understood in terms of collective goods, such as ideological and political community, and selective goods, such as elected offices. This thesis shows how the political involvement for the interviewed party activists is intimately linked to selective goods. This finding can help us to understand ideological and political changes in present-day parties. If, as a party activist, one is attracted by the opportunities to become an elected representative or the opportunities to exercise political power, it is reasonable to imagine that one will be more inclined to support actions that increase the opportunity to enjoy these goods. Since both the party leadership and the party activists are attracted by the selective goods offered by the party organisation, the conflict between these groups decreases, as both groups are focused on winning elections and exercising political power. The result of this study therefore provides grounds for questioning the ideas of a tension between party activists and party leaderships.

In this doctoral dissertation, I put party members and activists back in context. I stress theimportance of two contextual dimensions, often overlooked in the scientific literature. On theone hand, I put forward social network explanations of party membership and activism,emphasizing the importance of social interactions, relations and structures, which were scarcelyexplored as potential triggers. Like other forms of participation, party membership haspredominantly been portrayed through individuals as isolated, independent units, disconnectedfrom their micro social environment. On the other hand, local party branches as research objecthave often been ignored in the literature on party organisations, dominated by research on thenational mother party. Yet, we know little about how these local groups organise, how theycontribute to the recruitment and retention of members, and, more broadly, how the local partylevel copes with the challenges faced by the national organisation.With contribution on both supply and demand side of party membership studies, my dissertationis interested in questioning party membership and activism through network triggers and localparty organisations. Therefore, the final work comprises and articulates three empiricalchapters, offering their own sets of empirical analyses developed on original data.Inspired by network theory of political participation, the first chapter highlights the role ofpersonal networks as micro-contexts. Relying on original survey data gathered among a quotasample of 2,800 Belgian citizens, the chapter raises the question how the structure andcomposition of citizens’ social networks influence their probability to be party members.Regarding the structure, I demonstrate that the chance of joining a party increases alongnetwork size and density, with a stronger effect when the network is operationalised throughpolitical discussion. Regarding the composition, I show a major, positive effect of politicalattributes (attitudes of others: satisfaction and party closeness; and behaviours of other: otherforms of participation and party membership) and homophily (congruence in the network onthose political attributes) on the chances of joining a party, and, furthermore, the dominant partyin the network. Besides, social composition and homophily calls for further investigations.Overall, the chapter shows the effect of social context: how individuals behave and think towardpolitical parties is intimately linked with the features of their close social environment.The second chapter looks at party membership in local context. Framed by a functionalapproach of local party organisations, the chapter is dedicated to a qualitative assessment oflocal party branches as social and political groups. I ask how they organise, how they function,and, ultimately, discuss how they are impacted by party change. For this, I rely on a qualitativeanalysis of various empirical materials: hours of ethnographic observation within 11 local partybranches active in Brussels, complemented by interviews with their local presidents and adocument analysis of party statutes and local party rules and procedures, complemented withother internal and public party documents (leaflets, emails, posters, invitations, minutes,agenda, etc.). I show, first, that the organisation of local parties can be read through the trypticdeveloped to apprehend their national organisation (on the ground, in central and public office).Actually, local branches are divided into smaller groups, which all adopt different formal rulesand informal practices. Second, I stress that local party branches fulfil specific functions and anessential role of organisational and democratic maintenance, even if national membershipfigures are dropping. Local parties are involved in a complex dynamic between their role intheir municipality and as part of a larger party organisation. Nonetheless, this second chapterconcludes by pinpointing evidence of local party change: ascendency of the central and publicoffice, concentration of power, gap between members and officials, focus on nominationfunctions, etc.The third and last chapter puts party members in social and local party context. Bringing backnetwork theory, the chapter questions to what extent party activism might be triggered by thesocial networks built by members with their local branch fellows, or with higher party actors.The chapter consists in a quantitative analysis of survey data collected among local partyinformants belonging to 5 of the 11 local party branches. Respondents completed a shortquestionnaire adopting a similar operationalisation of social networks as in the first part of thedissertation. I demonstrate that the central place occupied by members in the network of theirlocal party branch (network centrality) is a good predictor of involvement at both the local andother organisational levels. This finding provides evidence to the “strength of strong ties” theoryin context of party organisation. In other words, the more a member is connected to othermembers of the branch aside from formal party activities, the more probability for this memberto be a party activist. I also show that extensive contacts with elected officials enhance thechances of joining a local party branch. Finally, I highlight that party members motivated bysolidary incentives are less likely to take part in higher participatory opportunities proposed bypolitical parties.More generally, my dissertation explores contexts in which party members and activists areembedded. It goes beyond individualistic considerations of traditional political behaviour andbrings back the local level into the organisational analysis of party membership.Methodologically, I extensively rely on Social Network Analysis to provide original relationalinsights on phenomena deeply collective by nature. Empirically, I focus on Belgium, a casehighly relevant to test new, “social” perspective on party membership. / Doctorat en Sciences politiques et sociales / info:eu-repo/semantics/nonPublished