# **Political Parties and the Challenge of Disintermediation**

**Rhetoric and Practices of Organisational Change in Italy**

https://doi.org/10.5771/9783748912644, am 28.09.2022, 11:13:25 Open Access – - http://www.nomos-elibrary.de/agb

Cecilia Biancalana

# **Political Parties and the Challenge of Disintermediation**

**Rhetoric and Practices of Organisational Change in Italy**

The prepress was supported by the Swiss National Science Foundation (SNSF).

© Coverpicture: ©karandaev – stock.adobe.com

**The Deutsche Nationalbibliothek** lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at http://dnb.d-nb.de

ISBN 978-3-8487-7249-0 (Print) 978-3-7489-1264-4 (ePDF)

## **British Library Cataloguing-in-Publication Data**

A catalogue record for this book is available from the British Library.


## **Library of Congress Cataloging-in-Publication Data**

Biancalana, Cecilia Political Parties and the Challenge of Disintermediation Rhetoric and Practices of Organisational Change in Italy Cecilia Biancalana 228 pp. Includes bibliographic references.

ISBN 978-3-8487-7249-0 (Print) 978-3-7489-1264-4 (ePDF)

1st Edition 2022

© The Author

Published by Nomos Verlagsgesellschaft mbH & Co. KG Waldseestraße 3 – 5 | 76530 Baden-Baden www.nomos.de

Production of the printed version: Nomos Verlagsgesellschaft mbH & Co. KG Waldseestraße 3 – 5 | 76530 Baden-Baden


Onlineversion Nomos eLibrary

This work is licensed under the Creative Commons Attribution – ShareAlike 4.0 International License.

*In loving memory of Emilia Fazzi Contigli and Mauro Fiori*

https://doi.org/10.5771/9783748912644, am 28.09.2022, 11:13:25 Open Access – - http://www.nomos-elibrary.de/agb

# Table of Contents


## *Table of Contents*



https://doi.org/10.5771/9783748912644, am 28.09.2022, 11:13:25 Open Access – - http://www.nomos-elibrary.de/agb

# Introduction

It is commonplace to start a book on parties by stating that parties are said to have been in crisis for a long time. Many causes have been attributed to this crisis, and various events have been recognised as its manifestations. On the one hand, mainstream parties are said to have lost their links with society and have become part of the state, fostering disaffection and a loss of legitimacy. On the other, non-mediated or unmediated forms of political action, such as online participation or social movements, emerge and become strengthened. From this perspective, a part of the crisis seems to consist in the fact that parties act in an environment characterised by the refusal of intermediate bodies. According to Urbinati (2015), some recent transformations in European democracies testify to a revolt against the way in which political parties and professional journalism, that is, the two intermediary bodies which made representative government work, have been organised in the past decades.

Actually, this can be found in other social fields too: it is a process that is said to have its roots in the social and political changes—such as individualisation and cognitive mobilisation—that make people more confident of their own competence and more and more sceptical towards the authority of expert knowledge, including political knowledge, and it is fostered by technological changes such as the spread of the internet. It seems that, thanks to the web, nowadays people can organise their lives autonomously, without any need for intermediaries: Amazon has changed the way we think of commerce, Twitter has transformed journalism, and so on.

Where this revolt against intermediate entities comes from is, of course, a matter of debate. In any case, this work will use it as a starting point. My aim is to study the strategic responses and adaptations of parties to this supposedly changing environment, particularly the responses and adaptations that concern their organisation. One response is what I call disintermediation. The term, which was first used in the financial and economic sector to refer to the opportunities given by the internet to directly link supply and demand (Chircu and Kauffmann 1999), generally describes the process of removal of intermediaries from a supply chain, a transaction, or any set of social, economic, or political relations (Chadwick 2007). Here, conversely, I consider disintermediation to be a response strategy put in place by parties to counter some pressures and external stresses.

There are two main trends that the literature on parties' organisational change has highlighted in recent decades. On the one hand, parties offer their members more opportunities for direct participation. On the other, they strengthen their leadership. My hypothesis is that these two trends are connected by the weakening of a party's intermediate structure and the attempt to create an unmediated connection between its leader and followers. Whether disintermediation does indeed provoke new forms of intermediation, or the permanence of previous ones, still needs to be verified. The distinction between disintermediation rhetoric and practices is obviously decisive in this matter: to what extent does disintermediation correspond to a real change both in the organisation and in the distribution of power within the party? To what extent should disintermediation be mostly considered a top-down process, which gives greater power to the leader in decision-making processes, or rather a bottom-up movement, which allows the opening of decision-making processes to members and voters? In light of these questions, I will consider two case studies in this work: the Partito Democratico (Democratic Party, from now on PD) and the Movimento 5 Stelle (Five Star Movement, from now on M5S). These are two very different parties, a mainstream party and a new party, which we can expect to behave in partially different ways. They will be observed from three points of view in connection to each other.

The first one is organisation. Through an analysis of the organisational history of the two parties, I will first observe their organisation at a national level, in order to understand whether and how they make use of disintermediation strategies, and which dimensions (rhetoric or practices; members' empowerment or concentration of power in the hands of the leadership) prevail. The study of the two organisations will be conducted through the analysis of documents and interviews with privileged witnesses, using the indicators developed in intra-party democracy scholarship. Secondly, considering that technological change is said to be one of the most important drivers of the transformations that push people to bypass intermediaries, I shall examine in depth the role of the internet in the disintermediation strategies of the two parties, and in particular the tools that potentially allow members to directly influence the life of the party: the online platform *Rousseau* for the M5S and the *circoli online* (online sections) experience and the mobile app *Bob* for the PD. Finally, I shall focus on what happens at the local level. Against this backdrop, how do parties adapt at this level? To this end, I conducted fieldwork research during the 2016 campaign in the municipality of Turin, observing the mobilisation of volunteers aimed at the re-election of the incumbent mayor of the PD Piero Fassino called Noi Siamo Torino (We Are Turin, from now on NST) and the participation practices of a local group of the M5S.

The work will be organised as follows. In chapter 1, I will focus on disintermediation as a strategic response of parties in times of crisis, outlining its various dimensions and the possible ways to analyse it. Chapter 2 will be dedicated to the research design, discussing the selection of the cases, the methods and the data source used. The following six chapters (3–8) are dedicated to the examination of the two parties chosen for the analysis. Chapters 3, 4 and 5 are dedicated to the PD; chapters 6, 7 and 8 to the M5S. For each party, I considered party organisation (chapters 3 and 6), the use of the internet (chapters 4 and 7), and members' and activists' participation during a local electoral campaign in Turin (chapters 5 and 8). In chapter 9, I will summarise the findings of my research, answer the questions that underpin my work and outline similarities and differences between the two parties in the three "arenas" considered: the national, the virtual and the local one. How do two different parties adapt to a context in which citizens appear to reject all kinds of intermediate bodies? Do disintermediation strategies produce new forms of intermediation or a permanence of previous ones? Finally, in the conclusions I will trace some avenues for future research and try to understand the implications and consequences of the phenomena under investigation on the future of political parties and, more generally, of representative democracy.

*This book originates from a PhD thesis that was conceived and written between 2015 and 2018, and defended in June 2019. Many things changed between 2015 and 2018, and many others happened in 2019 and 2020. Italian politics in general, and Italian parties in particular, change very quickly, and it is difficult to keep pace with them. I updated the manuscript until the end of 2020, accounting for all the relevant changes in the two cases examined in this work doing my best to analyse the latest events in the light of the framework that guided my research. However, unfortunately, for the writing of this newer version, I couldn't conduct new interviews and I relied on secondary sources and data.*

*I want to thank André Mach, Michele Sorice and Emilie van Haute—the members of my jury—for the time and energy that they devoted to my work; Alfio Mastropaolo, Oscar Mazzoleni and Franca Roncarolo for their constant and affectionate support. The merit is theirs and the mistakes all mine.*

#### Disintermediation as a Response to the "Crisis of Parties"? 1.

#### *Introduction 1.*

Political science scholars seem to agree that "the age of party democracy has passed" (Mair 2013, 1). This diagnosis does not primarily concern party resources, the centrality of parties in the processes of representative democracy at the level of national legislatures and executives, since "representative government remains very much a partisan affair" (Scarrow and Webb 2017, 3), or their role as institutionalisation agencies (Pizzimenti 2020), but their legitimacy and their connection with society (Ignazi 2004; 2017). It is a thesis that has been recognised for decades (Katz and Mair 1995) and has become almost common sense (Invernizzi-Accetti and Wolkenstein 2017): following some cultural–societal changes, mainly linked to individualisation, parties failed to perform their representative function of forming a link between the citizens and the state.

More generally, some have envisaged a true "revolt" against intermediary bodies in our age (Urbinati 2015): in politics like in other fields, especially following the massive spread of the internet, people seem to want to do without intermediaries. This is also testified to by the increasing success of the term, and of the concept, of disintermediation. According to the definition given by Chadwick (2007), disintermediation means removing intermediaries from a supply chain, a transaction, or, more broadly, any set of social, economic, or political relations. The term was first used in the financial and economic sector, and then became popular with the spread of the internet: through the web, demand and supply can meet directly, making (at least potentially) intermediaries useless: "Internet communication networks reduce the need for those who have some traditional claim to expert knowledge or market dominance" (*ibidem*, 232).

But disintermediation seems not to be limited to commerce and business or to the effects of the internet. More generally, the elimination of intermediaries is an increasingly relevant phenomenon in contemporary societies: the success of Amazon and Twitter is only the tip of the iceberg. For instance, when we think of the changes that have occurred in journalism and communication, it is possible to notice a growing trend in the loss of importance of previously relevant intermediaries. Digital and social media have changed the journalistic profession and the structure of the public domain (Chadwick 2013). On the one hand, every citizen can become an information producer, thus making journalism potentially redundant (Castells 2009); on the other, specifically in politics, digital and social media can create a direct communicative relationship between leaders and supporters, fostering personalisation processes (Ceccobelli 2017).

The literature so far seems then to have been more concerned with disintermediation as a process caused by recent social and technological transformations, to which actors are subject. My focus will be partially different, as I will consider the responses and adaptations of parties to this changing environment and label them as disintermediation strategies. Looking at disintermediation from this perspective, I will be able to consider the parties' internal dynamics and characteristics, which could shape the parties' strategies in differing ways. However, with regard to the general concept of disintermediation, two questions arise. The first is whether or not this alleged process of elimination of the intermediaries is a genuine one. In this regard, Chadwick (2007, 232) states that:

[…] it is by no means clear that intermediaries are being undermined by new information and communication technologies. The claim needs to be assessed alongside an appreciation of broader institutional concentrations of power. *Old intermediaries* have found their skills highly relevant to the internet age. They have at their disposal forms of knowledge, expertise, and wealth that are not distributed evenly throughout society. In some areas, *new intermediaries* are mushrooming (emphasis added).

On this subject, Chircu and Kauffmann (1999), writing almost twenty years ago about the emergence of new technologies for electronic commerce on the internet, developed the so-called intermediation–disintermediation–reintermediation (IDR) cycle. These scholars analysed the evolution of firm strategies when electronic commerce innovation occurs and discovered that traditional firms have access to a range of strategies that enable them to avoid disintermediation, hence becoming more powerful in the long run. Reintermediation is indeed the process by which a competitor that has been disintermediated is able to re-establish itself as an intermediary<sup>1</sup> . If old intermediaries can therefore re-establish their power, and new intermediaries can arise, disintermediation appears to be—more

<sup>1</sup> It is interesting to note that the fact that disintermediation always involves a process of reintermediation has been understood and seems very clear since the end of the 1990s in the field of commerce. On the contrary, as we will see, political

than intermediary elimination—intermediation of a different sort. Indeed, intermediary elimination does not automatically mean absence of intermediaries. Disintermediation can hence be considered a transformation or change of intermediaries, or of the forms of intermediation<sup>2</sup> .

The second question regards the specificity of disintermediation in the political field. Speaking of mediation or intermediation in politics equates to speaking of political representation (Pitkin 1967). Scholars (Manin 1995; Saward 2010; Urbinati 2013) have long claimed that we are witnessing a "crisis of representation" that manifests itself in growing dissatisfaction with and mistrust in politics and politicians and in lower turnout rates, but also in the rejection of intermediate entities and in the emergence and strengthening of new forms of non-mediated or unmediated forms of political action. According to Tormey (2015), we are moving from a "vertical" to a "horizontal" mode of politics, and this reflects the kind of society in which representation emerged and in which we are living now. Representative politics as we know it is a practice that emerged under particular historical conditions, in a particular time and place. We can therefore imagine that the practices of representation might change under different conditions.

For a long time, parties have been at the centre of representative democracy (Manin 1995). According to Urbinati (2015, 480–1) "it is impossible to understand representation without understanding the meaning, role and function of the political party". In contrast, nowadays we are witnessing deep dissatisfaction with the way in which they have been organised in recent decades, embodied both by the electoral rejection of traditional parties and the emergence of protest movements and new parties which have profited from the distrust towards mainstream parties. It seems that some social, political and cultural changes have made parties "unfit" (Ig-

actors have used the term rather uncritically as a rhetorical tool to legitimate their strategies.

<sup>2</sup> Analyses such as Chircu and Kauffmann's remind us of Pierre Bourdieu's studies on the concept of field (2000). According to Bourdieu, a field is a relatively autonomous microcosm within which struggles take place to define a "dominant view" and power relationships. When new actors enter the field—in the case of commerce, for instance, intermediaries in electronic commerce—the structure of the field is modified: some actors (e.g. old intermediaries) can adapt and try to re-establish their power; others (new intermediaries) can be incorporated into the field and accept its norms, contributing both to the modification of those norms and to their own institutionalisation.

nazi 2020) for postmodern societies. How are parties adapting to these new circumstances? Obviously, it is not a stimulus–response dynamic.

According to Harmel and Janda (1994), parties change in response to internal or external stimuli. Change can be considered the effect of an external stimulus (environmental—the environment being other parties too—or technologic) that joins forces with internal factors (that were autonomously shaking the party's internal power structure, e.g.: leadership change, dominant coalition change). Harmel and Janda's theory relies on Panebianco (1988), who stated that environmental stimuli may act as an important *catalyst* for the process that ultimately results in change, accelerating the transformations to the power structure, whose preconditions were already existent. However, in order to not to fall into a deterministic approach, it is also important to note that parties modify their environment too, and that their leadership is free to decide whether or not adapt to the changes, whatever consequences might result (Deschouwer 1992). The context does not *determine* the choices made by party actors. The critical actors within the parties must *perceive* environmental changes, and their probable effects on the party, in order for the environmental change to cause the party change. "Perception is the intermediate variable that has to be placed between objective facts and the reactions of the party" (*ibidem*, 17).

Recent research on organisational change has specified that the drivers of party change can be located at three different but intertwined levels: the level of the political system, where the general political norms and practices are located, the level of the party system—that is, the interactions and "contagion" between parties—and the level of the political parties, the level at which party members, party officials and party leaders act (Barnea and Rahat 2007; Gauja 2017). However, the aim of this study is not to propose a causal framework for the emergence of disintermediation strategies. As also underlined by Pizzimenti, Calossi and Cicchi (2020), disintermediation is considered in this work a heuristic tool used to jointly frame a number of changes in party organisations; in the following chapters my goal will be to test the heuristic validity of the concept. Nevertheless, bearing in mind that parties, with their actions and discourses, can also modify the environment, for instance influencing citizens' opinions and attitudes, and that their perceptions of environmental changes are just as important as the environmental changes themselves (Gauja 2017), to better contextualise the rise and success of disintermediation rhetoric and practices, in the next section I shall present some social and political trends common to most Western democracies that can influence party change towards disintermediation.

#### *A Quest for Unmediated Relationships 2.*

Two connected trends that can affect party politics and influence party change have been outlined by the literature: refusal of hierarchy and mistrust in politics. A trend that scholars have been investigating in recent decades is the decline of deference (Nevitte 1996; 2014), defined as a positive orientation towards authority. Studies have demonstrated that, in the last forty years, orientations towards authority became less deferential in the family, the workplace and the polity, and that orientations towards authority are connected across these different domains. According to these data, we are witnessing a continuing shift away from obedience to authority towards more individual autonomy in shaping family, work and social relations. The decline of deference is part of the broader processes of social modernisation (Inglehart 1990), individualisation (Giddens 1990; Bauman 1999; Elias 2001; Beck and Beck-Gernsheim 2002; Corcuff, Jon and de Singly 2005) and the shift away from tradition that characterise the so-called second modernity and might have political consequences.

The subtitle of a recent book edited by Russell Dalton and Christian Welzel (2014), *From Allegiant to Assertive Citizens*, summarises well the shift that the two scholars notice in citizens' attitudes and behaviour in dozens of countries around the world. According to these scholars, an allegiant political culture consists of orientations that tie citizens loyally to their society and its institutional order. It is possible to distinguish between three manifestations of such allegiant orientations: institutional confidence, philanthropic faith (trust in others and in democracy, interest in politics) and norm compliance. On the other hand, assertive orientation is a posture that encourages people to be critical and to voice shared concerns. The three elements of the assertive orientation are: individual liberties, equal opportunities and *people's voice*, that is precisely "the belief that people should have a voice in collective decisions on various levels, so that these decisions reflect what most people want". According to Dalton and Welzel, we are witnessing a transition from allegiant to assertive cultures. People in mature post-industrial democracies have become sceptical towards authority and institutions and are now more willing to assert their own views and to confront elites with demands from below, even in politics. The growth of suspicious attitudes towards experts during the coronavirus outbreak is a further sign of a shift in that direction.

In fact, mistrust in politics is also a subject that has been deeply investigated (Norris 1999; Hay 2007; Tormey 2015) and in turn has its roots in individualisation, cognitive mobilisation (Inglehart 1977), de-ideologisation and the decline of party identification (Dalton and Wattenberg 2002). Data show that citizens—also due to a public discourse that contributed to fostering anti-politics phenomena (Mastropaolo 2012)—have become more distrustful of electoral politics, institutions and representatives. As a consequence, voter turnout is decreasing, party membership is dropping (van Biezen, Mair and Poguntke 2012) and so is the level of trust and confidence in political parties in general. For the people that decide to vote, the option of the so-called populist or anti-establishment parties, which openly contest the organisation and ideology of traditional parties, is more and more appreciated. On the other hand, we are witnessing the emergence of new forms of political activism, online and offline, and the creation of deliberative and participatory experiments that try to involve citizens directly within institutions' decision-making processes (Della Porta 2013). To sum up, these studies argue that in recent decades individualisation, communicative abundance, the spread of digital media<sup>3</sup> , and the decline of collective identities, fostered by the shift from Fordism to post-Fordism (Tormey 2015) have made it hard for citizens to accept authority—in particular, the authority of politicians—, as well as the respect for hierarchy that is at the basis of political organisations such as parties and, in general, the concept of representation.

However, a recent study by Foa and Mounk (2017) seems to contradict this line of reasoning, showing that more and more American citizens are open to non-democratic types of regimes. In particular, the percentage of respondents who replied that it would be "good" or "very good" to have a "strong leader" who doesn't have to "bother with parliament and elections" is rising. Among all age cohorts, the share of citizens who believed that it would be better to have a "strong leader" who does not have to "bother with parliament and elections" rose over time: in 1995, about 24

<sup>3</sup> Both the decline of deference and mistrust in politics can be related to the changes that have occurred in media systems. On the one hand, communicative abundance makes information easy to retrieve, so the need for expert knowledge is considered less important, as is the authority of experts. On the other, the opportunities given by the web 2.0 can give citizens the impression that their voice as common citizens is important and will potentially be heard.

per cent of respondents held this view; by 2011, that figure had increased to 32 per cent. Meanwhile, the proportion of citizens who approved of "having experts, not government, make decisions according to what they think is best for the country" grew from 36 to 49 per cent<sup>4</sup> .

The decline of deference and support for non-democratic types of regimes seem two trends in contradiction, but it is possible to argue that they follow the same logic: the bypassing of intermediary entities and the creation of unmediated relationships between citizens and power. These unmediated relationships may be constituted by forms of direct, bottomup requests to the elites (the "people should have a voice") or through the decisions of a strong leader that directly embodies "what most people want". In both cases, there seems to be a will to bypass intermediate entities and the processes and procedures of representative politics intended in the last century.

[…] we are witnessing, on the one hand, the re-emergence of charismatic leaders […] and, on the other, a process going in quite the opposite direction as it claims to promote the diffusion of politics in view of reaffirming ordinary citizens' voice against the established political elites (Urbinati 2015, 477).

As a consequence of these broad social and political changes, we can hypothesise that political actors such as parties are trying to adapt to this new context. Nevertheless, we know that in order to analyse party change it is also necessary to pay attention to the internal factors that could shape party strategies. For instance, traditional and new parties could adapt differently to the same external conditions. Disintermediation strategies can be thus considered the parties' reaction to those broad social and political changes and to their perception of them, mediated by intra-party mechanisms and inter-party competition.

#### *Parties' Disintermediation Strategies 3.*

In order to analyse parties' disintermediation strategies, it is first necessary to specify that parties are not unitary entities: on the contrary, they are multifaceted (Katz and Mair 1993) and multilevel (Deschouwer 2006; Detterbeck 2012) institutions that perform many different actions. They

<sup>4</sup> Responses to Foa and Mounk's article can be found at https://journalofdemocracy. org/online-exchange-democratic-deconsolidation/.

structure the electoral competition, recruit political personnel, aggregate demands and interests, create public policies, organise members' participation and mobilisation (Bartolini and Mair 2001). However, parties are first of all organisations: if we focus on this aspect, we can ask how do parties respond to the changing environment by modifying their internal organisation.

#### Disintermediation in Party Organisations 3.1

The topic of party organisational change has been debated by political sociologists and political scientists since the dawn of their disciplines (Duverger 1951). In particular, in the last fifty years, various works have investigated the evolution of party models (for a critique, see Webb, Poguntke and Scarrow 2017). These works assume that the mass party, founded on party membership and typical of the industrial era, whose characteristics have often been considered "mythological" (Lefebvre 2013; Scarrow 2014), is vanishing, and they use it as an explicit or implicit term of comparison for the other party types that emerged over time. These studies take into consideration several dimensions of party change; here, I'm interested particularly in organisational changes<sup>5</sup> . From Kirchheimer's catch all-party (1966), to Panebianco's electoral-professional party (1988), and finally to Katz and Mair's cartel party (1995), this stream of literature outlines a series of trends.

The starting point, as we have already pointed out, is the gradual detachment of parties from society and their move towards the state "to compensate for the deficits that emerged with their delinking with society" (Ignazi 2020, 10). This, in turn, can be linked to one the greatest transformations in party organisations in recent decades, which is the decline of party membership (van Biezen, Mair and Poguntke 2012). The move of parties towards the state can be considered a cause of the decline of party membership, because citizens came to perceive them as self-referential, affluent and resourceful actors, interested only in maintaining their privileges. But the move towards the state is also a consequence of the drop in the figures of

<sup>5</sup> As regards the catch-all party, in reality only two out of five characteristics of this party model are related to party organisation (the lesser importance of party membership and the increased importance of party leadership). Even the cartel party model is not based primarily on organisational characteristics: individualisation of the membership and centralisation are only corollaries.

members: parties had to somehow replace the resources (e.g. money and workforce) that previously came from members, who were less and less stably linked to a party due to processes such as de-ideologisation.

Anyhow, the decline of party membership considerably changed the shape of party organisations. The loss of parties' legitimacy and the decline of their membership can be considered the drivers that led parties to open up their decision-making processes to members. The fact that the so-called party on the ground is declining in number and becoming apparently more powerful at the same time can indeed be seen as "paradoxical" (Scarrow 2014). In reality, it can be considered as an attempt by parties to revitalise themselves and to regain legitimacy by giving individual members a say in internal party decisions. Indeed, party members, and in some cases also supporters, have become more and more involved in the selection of a party leader (Pilet and Cross 2014), candidate selections (Hazan and Rahat 2010) and, in some cases, also in policy decisions (Gauja 2015) through direct votes. It is the well-known trend of an increase in intra-party democracy (Cross and Katz 2013) that has affected parties in the last few years. The declared aim of these reforms is to give relevance to party membership, to recover the relationship with citizens and, more generally, to democratise parties.

However, besides the desired effects, this trend towards the opening of decision-making processes might have other outcomes. The first one is the individualisation of participation within parties: the direct appeal to members would in fact promote an unmediated and atomised kind of participation. The second one is the marginalisation of the so-called middle-level elite: it has been demonstrated that empowering members causes the bypassing and weakening of the intermediate levels of party organisation. Finally, the third one is an increase in the power of the leadership. Indeed, some scholars have argued that the trend towards the internal democratisation of parties can be understood as a deliberate strategy by the party leadership to have more control over the organisation (Mair 1994, 16–17). In this respect, Mair hypothesises that:

[…] parties are actually making a careful and conscious distinction between different elements within the party on the ground, in the sense that the process of intra-party democratization is being extended to the members as individuals rather than to what might be called the *organized* party on the ground. In other words, it is not the party congress or the middle-level elite, or the activists, who are being empowered, but rather the "ordinary" members, who are at once more docile and more likely to endorse the policies (and candidates) proposed by the party leadership and by the party in public office […]. Ordinary members, often at home, and via postal ballots, are increasingly being consulted by the party leadership, and are increasingly involved in legitimizing the choices of the party in public office […]. In a related vein, it might also be argued that the process of intra-party democratization is often meaningless and/or illusory.

More generally, the attempt to reverse the oligarchic tendencies of parties (Michels 1911) runs counter to the internal dynamics of party organisations. Party types that have emerged with the transformation of societies have not been able to overturn Michels' "sociological law" (Carty 2013), because the devolution of power to individual members foster a plebiscitary approach to politics and create an unmediated relationship between leader and followers which substantially benefits the former. This increased relevance of party leaders is favoured not only by the changes in party organisations but also by long-term trends towards leadership personalisation (Blondel and Thiébault 2010) and the presidentialisation of politics (Poguntke and Webb 2005). As part of the growing relevance of individual political actors at the expense of parties and collective identities (Karvonen 2010), leaders are increasingly resourceful within executives and parties, and in electoral processes (Calise 2010). Transformations in the structure of political communication, starting from the growing role of television in politics, have had a crucial role in these processes: electronic media allow the transmission of an unmediated message from the leader to the electorate, without the need of an intermediate organisation or the groundwork of party members.

To sum up, what emerges in this stream of literature, against the backdrop of the detachment of parties from civil society, are two separate but interrelated trends: on the one hand, the empowerment of individual party members; on the other, a concentration of power in the hands of party leaderships. The concept of disintermediation can be used to summarise the two trends that have occurred in recent decades, as it holds together the two directions of the weakening of the party's intermediate organisation, which are often considered separately. Indeed, what unites these two trends is the attempt to bypass the party's intermediary structure and to create an unmediated relationship between leader and followers.

In an increasing number of situations, as we will see with the analysis of the two cases examined in this work, we witness the simultaneous presence of elements of horizontality (direct participation from below) and verticality (concentration of power in the hands of the leader): the concept of disintermediation can then be helpful in understanding these apparently contradictory phenomena. Moreover, disintermediation does not simply mean the lack or the elimination of intermediaries. On the contrary, it appears as a *transformation* of the forms of intermediation. So, the use of this concept allows us to describe not only the trends towards the creation of an unmediated relationship between leader and followers, but also the presence of new forms of intermediation or the persistence of previous ones. The idea of the existence of an IDR cycle can also be useful in the case of party organisations: the analysis of parties' disintermediation strategies should also encompass the evaluation of the emergence of new forms of intermediation and the persistence of traditional forms of intermediation, that is, a reflection on disintermediation as a process which takes place over time. Finally, the alleged "illusory" nature of members' empowerment highlighted by Mair urges us to stress the difference between disintermediation rhetoric and practices and to evaluate the actual distribution of the decision-making power within a party.

#### Disintermediation Rhetoric and Practices 3.2

We have observed how in a general context characterised by some social and political trends, parties have changed their organisations by developing unmediated relationships between leaders and supporters: this is the essence of the concept of disintermediation. At this point, an important distinction that has to be made is the one between rhetoric and practices of disintermediation. A request for an unmediated relationship, or the acknowledgement of that request by parties, does not automatically mean that actual unmediated relationships are established. It is then important, in the first place, to distinguish between rhetoric and practices (Kittilson and Scarrow 2003), or between the symbolic and substantive aspect of disintermediation (on substantive and symbolic party change see Harmel 2002; Gauja 2017). Parties' disintermediation strategies can be only rhetorical or symbolic, not involving any actual change in the distribution of their internal decision-making power, but only the creation of a discourse or a narrative on it.

Symbolic change is that which is largely formal and ceremonial, without creating any corresponding alteration in political practice. Change may be symbolic because it codifies an existing practice, or because it primarily seeks to change attitudes rather than enforce a particular type of behaviour (Gauja 2017, 1).

However, this doesn't mean that the narrative is not important. Writing about party reforms that are "intentional and publicized changes that are made to a party's structures and practices in order to improve them", Gauja (2017) acknowledged that the symbolism of change can be just as important as the substance—including whether or not the reform initiative actually succeeds in changing established party practice. The discourse surrounding party reform can be even more important than its actual implementation, having an impact on both citizens' perceptions and parties themselves.

We know that in recent years the term and the concept of disintermediation entered the public debate. As we have seen, it has been used by scholars as an analytical or heuristic tool, but it has also been used by political actors as a strategy to legitimise their actions. If, on the one hand, parties adapt to social and political changes, it is also true, on the other hand, that they can also strategically use those changes in their rhetoric and with their practices in a way that is convenient for them, influencing, in turn, citizens' opinions and attitudes. This can make the study of disintermediation slippery ground and urges us to carefully consider the rhetorical or symbolic side of party strategies. This problem has already been recognised, for instance, in the case of the category of populism and, more generally, in the complex relationship between the political scientist, the object of his or her research and the categories used for analysis; a relationship that also involves the political, scientific and cultural context in which research is carried out.

Political scientists are faced with "labels" and "tags" conveyed by their own colleagues, but also by actors themselves (representatives of the party, opponents, various commentators, etc.). Labels (markers, classifications) which claim to be "scientific" *de facto* adopt a terminology, a language which refers to and makes sense in universes which go far beyond the scientific sphere […]. For this reason, the work of definition, and first of all labelling, should include a critical reflection on the categories used, in particular on their conditions of production and reception. These categories contribute to ensuring that a particular party, at any given time, can be defined in one way rather than another. The act of labelling and of scientific definition thus participates, voluntarily or not, in the construction of the public image of the party, in making it "exist", promoting for example its "centrality" or its political "marginality" (Mazzoleni 2007, 18, my translation).

As regards the two Italian parties considered in this work, we can say that the term and the concept of disintermediation are present in both their rhetoric and their practices. As regards the M5S, we can say that disintermediation, intended as the creation of a direct link between citizens and power, is the very core of its political message and its idea of democracy. We can see it in this quotation from a post published on Beppe Grillo's blog in 2013:

The M5S wants to achieve direct democracy, disintermediation between state and citizens, the elimination of parties, initiatives without quorum: the citizen in power (Post *Il M5s non è di destra né di sinistra*, 11/1/13).

In an interview quoted in Gerbaudo (2021), Roberto Fico, the current speaker of the lower house, states that the original idea behind the M5S "was making politics as direct as booking tickets on Ryanair or booking a room on AirBnb". This idea, as we will see, is mirrored in the M5S's internal organisation, as well as in its use of the internet. We can say that the aim of this newly established party is to overcome representative democracy through forms of internet-mediated participation. But the term disintermediation has been used by representatives of the PD too. An article written by Lorenzo Guerini, chief of the party's national organisation during Renzi's first mandate as party secretary, is revealingly titled *The Political Party in the Era of Disintermediation*. Guerini (2014) states that "[mainstream] parties have not managed to change themselves, adapting to social changes" and asks which party model can fit the new demands in the era of disintermediation. According to him, party primaries, "can be understood as a democratic interpretation of disintermediation"<sup>6</sup> . Through primaries the citizen-elector can identify "his preferred programmatic proposal, embodied by a democratically legitimated leadership". Furthermore, the term disintermediation was widely used by Matteo Renzi, party leader of the PD between 2013 and 2018, as well as by scholars and journalists in order to describe Renzi's attitude towards communication and trade unions (Cuono 2015).

<sup>6</sup> It is interesting to note that Guerini adds the adjective democratic to the term disintermediation. It seems that he perceives disintermediation as a rather negative trend, which the PD is making positive with party primaries. In contrast, we can note that, for the M5S, the concept of disintermediation is a completely positive one and seen as the very goal of the party.

#### Disintermediation from Below and from Above 3.3

Parties' disintermediation practices in party organisations can be observed through the examination of the distribution of the party's internal decision-making power. We can expect that an increase in the decision-making power of its leadership and of its individual members would weaken the party's intermediate organisation as a consequence. We can thus distinguish between two different sub-dimensions of disintermediation, which correspond to the two separate trends identified in the literature on parties' organisational change, and question whether there is one that prevails in their rhetoric or practices, and which one it is: from below (i.e. inclusiveness, thus increasing decision-making power for members and supporters); and from above (i.e. the increasing autonomy of the party leadership)<sup>7</sup> .

Inclusiveness is the core of the concept of intra-party democracy (von dem Berge and Poguntke 2017). Intra-party democracy (IPD) is a broad term used to describe a wide range of methods to include party members in intra-party deliberations and decision-making (Scarrow 2005). So, in order to analyse inclusiveness (and exclusiveness) in party organisations, we can rely on the dimensions and indicators developed in this field of research. According to Scarrow (2005, 6), inclusiveness regards "how wide the circle of party decision makers is". In exclusive parties, the main decisions are made by a small number of party actors, that is, the party leadership. In contrast, in inclusive parties, a large number of party members make decisions on the central issues of the party. Von dem Berge and Poguntke (2017, 140) define inclusiveness in the following terms:

(i) the higher the number of party members involved in intra-party decision-making (relative to party size), (ii) the more open the election and composition of party organs (e.g., absence of ex officio seats), and

<sup>7</sup> Pizzimenti, Calossi and Cicchi (2020) further developed the concept of disintermediation applied to party organisations ("internal" disintermediation). They propose using four indicators to analyse disintermediation in party organisations, namely: the opening of boundaries of party organisations (for instance, also opening the organisation to "friends" and/or "sympathizers"); the dismissal of the party's collateral organisations, the decrease in the number of party layers between the highest executive body and the party congress, and a greater presence of representatives of the party in public office in the party's executive organs; the expansion of the rights and functions of the party leader.

(iii) the more the party leader shares power with other, more inclusive party organs or actors, the more inclusive […] a party is.

Scholars have recognised three main areas to investigate IPD (von dem Berge et al. 2013). The first is the decision-making power of members in formulating and implementing policies. The second one is the decision-making power of members in deciding on party personnel. Here, two intra-party processes are relevant: leadership selection and candidate selection. The third regards the formal distribution of power within the organisation. Consequently, in this work, together with a general analysis of party structure (number, role and characteristics of the party's intermediate bodies), three main aspects will be taken into consideration in order to analyse disintermediation practices<sup>8</sup> : the selection of the leader and his/her role within the organisation; the selection of candidates; the determination of policies. Moreover, one last aspect to consider is the role of party membership, a dimension in part already included in the three areas outlined above (e.g. whether or not members have the power to select leaders, candidates, policies), but that it is also worth to analyse separately.

The topic of transformation of party membership is crucial for the study of party organisations. On the one hand, traditionally intended members' participation is said to have been replaced with an unmediated relationship between leaders and individual supporters. On the other hand, recent research (Scarrow 2014) accounts for a more nuanced picture, as parties can open new channels of partisan engagement and communication that can complement or substitute traditional party membership. At the same time, other scholars tell us that party members still have important functions within their party: on the one hand they are an important organisational resource (for instance, in electoral campaigns or in selecting candidates), but they are also an important source of legitimacy for parties (van Haute and Gauja 2015). Analysis of the role of party membership is also relevant since, drawing on previous research, we know that disintermediation strategies could involve both the weakening of the middle-level elite and the decline of the role of party members within parties' organisational structures.

To conclude, when importing or developing a new concept, it is appropriate to explain why existing concepts are not fit to describe or explain

<sup>8</sup> I must specify that my goal is not to analyse these aspects quantitatively, but to use these dimensions as a guide for the qualitative analysis of the two party organisations.

the phenomenon under investigation. In my case, the essential reference is the aforementioned Katz and Mair's cartel party thesis. The cartel party thesis served as a starting point for the development of my framework, but since the focus of my research is partially different, I detached my work from it in various respects. In the cartel party thesis, organisational changes are seen as the by-product of the increasing proximity of parties to the state. The focus, then, is on party–state relationships and on inter-party collusion, aspects that in my research are kept in the background. The premises are thus similar; in a context characterised by a declining level of participation and by individualisation, state resources are used to maintain parties' positions within the political system, IPD is used as a tool of control by the party elite, and the autonomy of the leadership increases in the end. But, on the other hand, my framework is detached from the cartel party thesis in three main respects.

The first one, as I said, is the different focus (inter-party in the case of the cartel party thesis vs. intra-party in this research study). Organisational changes, and in particular the unmediated relationships established within the party, are my main focus. Starting from this different focus, in my framework I enriched and made more complex some ideas present in Katz and Mair's piece of research. In the first place, I analytically divided the two sub-dimensions (inclusiveness and autonomy of the party leadership) that, often considered separately, are kept together within my definition of disintermediation. In the second place, I distinguished between practices and rhetoric, valuing the symbolic dimension of party change. In the third place, as we have seen, the concept of disintermediation presented in this study, as opposed to existing concepts, allows us to also take into consideration the process of the emergence of new forms of intermediation. Finally, I found the concept of disintermediation useful with respect to the existing concepts because it represents the decline in politics of phenomena that, as we have seen, are present in other fields too. By using this concept, I do not intend to deny the specificity of political phenomena; on the contrary, my aim is to show their distinctiveness against the backdrop of broad social processes.

#### The Role of the Internet 3.4

Disintermediation is not limited to the internet, but it is true that the web plays an important role in these processes, so it is interesting to consider the relationship between the internet and parties specifically: how do parties use the internet and what is its impact on them? This is relevant because not only can the web transform the way in which parties communicate and organise themselves (Barberà et al. 2021), but also because new technologies could have a broader impact on the relationship between the represented and the representatives in contemporary democracies (De Blasio 2014). The internet allows faster exchange of information, potentially without geographical boundaries. Moreover, especially following the advent of the so-called web 2.0, friendlier interfaces have made bidirectional communication suitable for all. Each user has the opportunity to become not only a receiver but also an information producer, and this has consequences on different fields, including the political one. On the one hand, institutions, parties and social movements have adapted to this new environment; on the other, citizens now act in a public sphere in which new technologies play an increasing role.

With regard to parties<sup>9</sup> , the pioneering study by Gibson, Nixon and Ward (2003) identified three main areas in which to investigate the use and the impact of the internet on parties: party competition and online campaigning; internal democracy; and the role of parties in contemporary democracies. This classification has the trait of not limiting the web's impact on parties only to communication: parties can use the internet not only to communicate with citizens and supporters and to receive feedback from them, bypassing journalism and creating a direct link with public opinion, but the web may also be an organisational infrastructure for activism (as it allows communication in real time, without sharing physical space) and potentially a tool for deliberating and deciding, thus innovating decision-making processes. Digital technologies can thus redefine parties' structures and practices from various and different points of view, with repercussions on the mechanisms of political representation.

In order to investigate the role of the internet in parties' disintermediation strategies, it is necessary to specify and distinguish between the possible uses of the web by parties and their impacts, building a classification of parties' uses of the internet. In order to create it, we can ask ourselves two questions: Who is the recipient of the party's digital communication? What function or goal does it perform? As far as the first question is concerned, we can distinguish between an external recipient (citizens, the media, etc.) and an internal one (party members, activists, party personnel, etc.). As regards the second one, we can identify three different goals of

<sup>9</sup> I must specify that here I'm dealing only with the use of the internet by parties, and not by leaders or candidates.

parties' digital communication: communication, action, direct democracy. Crossing these two dimensions, we can build a classification of the possible uses of the internet by parties and of their digital communication tools (Table 1.1).


*Uses of the internet by parties Table 1.1*

In the case of an external recipient, the internet can be used by parties to communicate with citizens (top-down dimension) or to receive feedback from them (bottom-up dimension). The tools that can be included in this category are the party website or its social network accounts, through which the party can inform citizens (also directly, with direct messages such as e-mails), but also collect feedback from them (e.g. through comments or parts of the website dedicated to feedback). The external use of the internet mainly has a communicative function, intended as a transfer of information and sharing of interpretations between actors.

The internal use of the internet can have two main goals. The first is aimed at action, which is the achievement of a purpose. The web can help groups to coordinate their action (horizontal dimension) and/or organise participation and mobilisation (vertical dimensions). The difference between coordination and organisation is relevant here. The use of the internet for organisation purposes is a top-down one, with a centre from which the communication flow starts. In contrast, the use of the internet for coordination purposes is horizontal, as it allows groups to coordinate themselves in the absence of a single centre or leader. Examples of coordination could be chats, mailing lists or groups through which party members (e.g. the party's local groups or sub-groups) can organise local meetings or other initiatives<sup>10</sup>. On the other hand, examples of use of the web for organisation purposes could be tools such as dashboards, used by parties to mobilise members and supporters during electoral campaigns. The difference between coordination and organisation is that in the first case there is not an official hierarchy, while in the second case it is the party that organises the online mobilisation from above.

I have defined the second internal use of the internet by parties as "direct democracy", and it involves intra-party democracy. Indeed, the web does not only allow the organisation of social action without the sharing of the same physical space. From the very beginning, it was thought that the internet could foster direct democracy experiences, especially in the so-called cyber-optimistic vision of the relationship between the web and democracy. One major critique of direct democracy in modern societies is that it is impossible to realise, due to the impossibility of managing deliberations and votes involving a large number of people. Theoretically, the internet could solve this problem, and that is true not only at the political system level, but also within parties. With direct democracy tools, such as participatory platforms, parties could indeed empower members, giving them a say in party decisions and then potentially altering the distribution of the internal decision-making power in their favour.

However, this varies according to the actual power that parties are willing to grant to their members. Using the typology by De Cindio and Stortone (2013), we can outline three different types of use of the internet for this purpose, depending on the actual transfer of power to the base and on the commitment of the party to take into consideration members' stances. These three types are: discussion, consultation and decision. Discussion means that the party neither commits to collecting members' ideas nor to taking them into consideration. Tools falling into this category could be discussion forums: spaces in which there is an exchange of ideas among members, completely unlinked to the party's decision-making process. In the case of consultation, the party commits to collecting and considering members' ideas, but in a non-binding manner. Finally, decision means that the party commits itself to pursuing the decisions taken by the members online11. Tools falling into this category could be participation platforms in which it is possible to organise binding consultations.

<sup>10</sup> This can call to mind the use of the internet by social movements (Castells 2012).

<sup>11</sup> Here, we can also distinguish between initiatives (members have the power to request a consultation) and referenda (consultations come from below).

The impact of the use of the internet on parties is therefore different depending on the recipient and the goal. Communication is not the main focus of this work. With regard to action, we can say that the internet can have an impact on members' and activists' participation practices, allowing both the autonomous and horizontal coordination of groups and top-down organisation of individual activism, and on the party's structure. With regard to direct democracy, the main impact is on party organisation, as digital tools can potentially change the distribution of the decisionmaking power within the party, for instance in the case of participatory platforms<sup>12</sup>. Not to mention that the very fact of creating an online tool regardless of its effectiveness—can have an impact on the image of the party in the public sphere, and then on citizens' and members' perceptions of the party.

However, if parties can empower their members with digital tools, online participation can be highly individualised. In addition, it is important to pay attention to the architecture of participatory platforms, as they can hide new concentrations of power: technology is not neutral, and the ownership of the online participation tools and the identity of the subject who holds the authority to set the rules of participation are two key issues to be investigated<sup>13</sup>. The risk is empowering the centre instead of the base, as online decision-making processes could result only in the ratification of choices taken elsewhere.

In this regard, together with Pedersen and Saglie (2005, 362), we can ask ourselves: "What would happen to party organisations if traditional party activities were replaced by electronic participation?". These two scholars envisage three scenarios. In the first one, the new ICTs undermine the power of the party leadership in favour of the empowerment of party membership. In the second one, the individual participation fostered by ICTs weakens the deliberative aspect of party organisations and thus give the leaders, and in general the party elites, more power. In the third, the impact of ICTs on power, democracy, and participation is "limited". This does not mean that the new media are unimportant, but that, for instance,

<sup>12</sup> Some authors (Gerbaudo 2019) have claimed that parties that rely on this kind of tools in their internal decision-making processes can be considered a new party type, the so-called digital party. For a critique, see: Passarelli, G., *Il partito digitale: un'ipotesi per parlare di politica, ma senza prove*, https://www.che-fare.com/partito-d igitale-gerbaudo-senza-prove/, October 2nd, 2020.

<sup>13</sup> At a broader level, we can also consider companies such as Google and Facebook new intermediaries, as they are acting more and more as gatekeepers. This is particularly true for parties that don't rely on proprietary software.

increased access to them does not necessarily create greater interest in political participation.

Drawing on the work of Pateman and Verba, we can imagine a fourth scenario according to which ICTs perform mainly a *symbolic function*: a point that resonates with the rhetoric dimension of disintermediation. According to this hypothesis, digital tools are not used to give members more power, but to give them the *impression* they can influence the decision-making processes of the party. It's what Pateman (1970, 69; see also Verba 1961) called pseudo-participation. Pseudo-participation is a situation in which no participation in decision-making actually takes place: for decision-makers, the concern is to create a *feeling* of participation and openness, while retaining power in their own hands. Pateman, dealing with participation in industries, defines it as a situation in which participation is limited to an endorsement of a decision made elsewhere; for instance, a situation in which the supervisor, instead of merely telling the employees of a decision, allows them to question him about it and to discuss it—yet without changing the desired outcome.

In order to understand the role of the internet in the disintermediation strategies of the two parties, for the analysis of parties' direct democracy tools, which will be the main subject of the part of this work dedicated to the parties' use of the web, I shall focus on three dimensions: the architecture of the platform and its affordances (Dahlberg 2011), that is, the features present in the digital tool and the activities users are encouraged to perform; the transfer of power from the top to the bottom, and specifically members' rights and powers within the digital tool and their ability to influence the "rules of the game"; and the consequences on party organisation, in particular on the internal distribution of power.

#### *New and Mainstream Parties' Disintermediation Strategies 4.*

Against this backdrop, the aim of this study is to answer three main questions. The first one concerns how parties adapt and change in response to social and political changes, and in particular to a context characterised by the perception of a refusal of intermediate bodies, especially with regard to their internal organisation. I hypothesise that disintermediation strategies are the parties' answers to such changes; following what has been said in the previous pages, I can define disintermediation strategies as *rhetoric or practices developed by parties in order to stage or deliver an unmediated* *relationship between leader and followers, which happens with the weakening of the party's intermediate organisation*.

The second question regards whether parties' disintermediation strategies produce new forms of intermediation. Is the process of intermediary removal a genuine one? As I have stressed in the previous pages, the literature on disintermediation is scarce, but the existent contributions underline that both *new intermediaries can arise and that old intermediaries can avoid disintermediation*. In the first case, Chadwick (2007) refers to the birth and strengthening of the new intermediaries in the internet age; in the second one, Chircu and Kauffman (1999) identify the IDR cycle and argue that traditional firms can avoid disintermediation and become more powerful in the long run through reintermediation. So, starting from the idea that disintermediation would imply a *transformation* of intermediaries rather than their removal, in analysing disintermediation strategies I expect to find new forms of intermediation and/or the persistence of old ones.

The third issue is to understand whether different parties adapt in different ways. Parties change in response to external stimuli, but internal factors count too. For this reason, different parties respond to the same stimuli in a different way. In particular, we can distinguish between "old" (traditional or mainstream) parties and "new" parties. But, what is a new party?

Newness is not easy to define (Chiaramonte and Emanuele 2016). According to Bartolini and Mair (1990), a party is new when it does not derive from the structure of an existing party. This is a criterion that takes into consideration the party's structure, but new parties are also innovative in other ways. Deschouwer (2008), for instance, points out that there are three different dimensions of newness to consider: the age of the party, its ideology and the type of party organisation. Indeed, we know that starting from the 1970s, party and electoral competition has increasingly been structured by a diversity of policy issues, rather than on long established societal cleavages. The emergence of new political issues has been followed by the rise of new parties, both on the left—ecologist, libertarian parties and on the right—radical right parties. The emergence of these new parties can be seen as a reaction to traditional parties that were part of the establishment. It is a reaction that concerns not only mainstream parties' policies and ideologies but, especially in the case of the parties that can be placed on the left of the political spectrum, also their organisation. Also, in recent years some Western European countries, especially those hit hardest by the 2008 recession (Chiaramonte and Emanuele 2016), experienced an increase in new parties which question the ability of existing parties to cope with the effects of the crisis.

It is not the aim of this work to provide a clear definition of what a new or a mainstream party is. For the purpose of our analysis, we can say that new parties are political actors that do not derive from existing parties and that oppose them both with respect to their policies and organisation. Existing parties to which new parties oppose themselves can be defined as mainstream. According to Meguid (2008), these are actors that are located on the left–right political dimension, that have the electoral dominance of that bloc, and that are widely considered governmental actors.

How do these two types of parties adapt and change? As far as mainstream parties are concerned, we know that they are conservative organisations, which will not change simply for the sake of change (Panebianco 1982). Parties are not willing to simply give up part of their power, so we can expect these parties to open up their decision-making processes in order to gain legitimation, in a context in which they are accused of being disconnected from society (Ignazi 2020): we can then assume that they will use rhetoric characterised by disintermediation from below and practices marked by disintermediation from above. Moreover, like in the case of traditional firms (Chircu and Kauffman 1999), we can expect them to try to re-establish themselves as intermediaries (reintermediation).

In contrast, starting from the 1970s, new parties (such as in the case of Green parties, or the so-called New Left parties and more recently in the case of movement parties, Kitschelt 2006; Della Porta et al. 2017) oppose mainstream parties and challenge them both regarding their policies and organisation. These new parties are frequently "intraparty democracy maximizers" (Harmel and Janda 1994), and their goal is to empower members' participation. However, even these new parties tend towards institutionalisation and centralisation (Poguntke 2002; Frankland, Lucardie and Rihoux 2008), and to implement some organisational changes, following the new functions that they will have to enact with their internal complexification (Pedersen 1982). Thus, we can expect that, as far as their rhetoric is concerned, they will enhance disintermediation from below, but that, over time, they will increasingly employ disintermediation practices from above and develop new forms of intermediation.

In order to answer these questions, I will examine how two Italian parties (a mainstream party, the Partito Democratico, and a new party, the Movimento 5 Stelle) have adapted to the changing context. In the next chapters, after having presented the design of my research, I will ask whether the two parties employ disintermediation strategies at the national level, how they are interpreting them and which dimensions among those outlined (rhetoric or practices; members' empowerment or autonomy of the leadership) prevail in each party. Furthermore, I will investigate the role of the internet in the disintermediation strategies of the two parties and what consequences it has on the party's organisation. Against this backdrop, I will observe what happens at the local level, especially with regard to members' and activists' participation during electoral campaigns. Finally, I will summarise my findings and outline similarities and differences between the two parties in the three "arenas" considered (the national, the virtual and the local one).

#### A Roadmap to the Study of Disintermediation 2.

#### *Italy, Disintermediation and Party Change 1.*

This study is focused only on one country and one political system: Italy's. This will allow me to take into consideration the country's institutional context, and to carry out an in-depth analysis of the two case studies. Italy is an interesting country for the study of political parties and party change, in general, and of the topic of disintermediation, in particular. On the one hand, in this country parties have always played an important and peculiar role. At first, they facilitated the democratic transition, and their strength heightened their tragic collapse in the nineties. At the same time, anti-political (Mastropaolo 2000) and anti-party (Lupo 2013) sentiments are deeply rooted in Italian society. On the other hand, the Italian case befits Western democracies' general trend towards the increasing role of leaders and the will of citizens to make their voices heard directly.

Historically, we can say that Italian "democratic consolidation" in the post-war period, which was the passage from an authoritarian to a democratic regime, happened thanks to the actions of parties (Morlino 2003); they served as "anchors" for civil society, for instance through growing membership figures and the links between parties and collateral organisations. The centrality of parties formed the base of the functioning of the so-called First Republic, i.e. the historical and political period between the end of WWII and 1992, and allowed parties to be the main actors in the stabilisation of the democratic regime. At the same time, however, this centrality, the so-called *partitocrazia*, produced some negative effects and created the conditions for the collapse of the party system in the early nineties, because of the excessive accumulation of power and parties' penetration into the socio-economic system of the country (Grilli di Cortona 2007).

According to Morlino and Tarchi (2006), Italy has always been characterised by a chronic and widespread dissatisfaction with the political system, which was present even in the immediate post-war period, but only fully manifested itself in the early nineties because of the convergence of a series of circumstances: namely, the collapse of the Berlin Wall and the processes of de-ideologisation and secularisation. Although mistrust and anti-political attitudes might derive primarily from a public discourse spread by political actors themselves (Mastropaolo 2000), this doesn't mean that it has not had consequences on the Italian political system. Against this backdrop, the event that triggered a major change in the Italian party system was *Tangentopoli*. Indeed, at the beginning of the nineties, a series of scandals, dubbed *Tangentopoli*, uncovered a vast network of political corruption and shook the Italian political system. This, together with the collapse of the Berlin Wall, marked the end of the most important Italian political parties and the beginning, despite the absence of constitutional changes, of the so-called Second Republic. Moreover, the move from a proportional to a mixed majority electoral system changed the dynamics of party competition. During the so-called Second Republic, the Italian party system was then characterised by a bipolar competition dynamic. On the one hand, there was the centre-right coalition, led by Silvio Berlusconi; on the other, a centre-left one, which lacked an unchallenged leader.

It was especially in the period after *Tangentopoli* that Italy was defined as a "populist paradise" (Tarchi 2015) and anti-party sentiments among citizens increased dramatically (Lupo 2013). Fostered by the new electoral system and by other societal and technological changes, such as the growing popularity of commercial television, as well as by some Italian peculiarities, such as the role of Berlusconi as a media entrepreneur, the Italian political system became more and more personalised (Mazzoleni and Sfardini 2009). Furthermore, if it is true that "personalization has affected political parties in many contemporary democracies", it is also true that "Italy plays a leading role in this scenario" (Musella 2015, 241– 242). At the same time, some scholars (Calise 2010) have pointed out that in this period Italy was characterised by a peculiar form of protest against parties, called *direttismo* (directism). Directism consists in bypassing party intermediation, in order for citizens to be able to directly influence politics. According to Calise, Italy, more than any other European country, became the cradle of directism (see also Sartori 2006), and this was also due to the importance of the referendum in the Italian political system (Barbera and Marrone 2003). In addition to this, the direct election of the executive at the the municipal, provincial and regional levels, which occurred starting from 1993, contributed to giving citizens the impression that their electoral choice was a deciding factor.

Finally, the Italian Partito Democratico was the first European party to organise open primaries for the election of the party leader, and Italy is a country in which primaries for the selection of candidates are widely used (Sandri and Venturino 2020. For a reflection on primaries and directism, see Melchionda 2005). Finally, the extraordinary results of the M5S at the 2013 and 2018 general elections (Chiapponi 2017; Biancalana and Colloca 2018) testified to the deep crisis in traditional Italian parties, but also the fact that Italian citizens were still keen to participate in elections and to vote for a party that lists the creation of a sort of direct democracy among its promises.

#### *The Partito Democratico and the Movimento 5 Stelle 2.*

This study will focus on two parties: the Partito Democratico and the Movimento 5 Stelle. The two parties share some similarities but are also very different in many respects. As we will see in the rest of the book, in both cases it seems that the two parties wanted—albeit in two partially different interpretations—to bypass the internal party organisation and to create an unmediated relationship between their leader and followers.

If we simply look at their dates of birth, both the PD and the M5S seem to be new parties. The first was born in 2007 and the second in 2009. But, in reality, their histories are very different. The PD can be considered a mainstream party (Meguid 2008). A centre-left party, it is the heir of the two most important mass parties of the so-called First Republic (Democrazia Cristiana and Partito Comunista Italiano), as it was founded in 2007 through the merger of two existing parties which, in turn, derived from those parties. Over time, it has moved more and more to the centre of the political spectrum (Natale and Fasano 2017) and, as other European centre-left parties, has gradually lost contact with the popular classes<sup>14</sup> . Its peculiar organisational structure, which has been defined as "open" (Vassallo and Passarelli 2016), presents a very nuanced distinction between party *supporters*, who have many rights, including the opportunity to vote for the party secretary (and not just for candidates) during party primaries, and *members*, who have few additional rights with respect to supporters and, perhaps also for this reason, have sharply decreased in number over time.

The M5S, in contrast, is a genuinely new party, founded in 2009 by the former comedian Beppe Grillo, a well-known showman who has long been engaged in civil and political battles, and an entrepreneur with no

<sup>14</sup> De Sio, L., *Il ritorno del voto di classe, ma al contrario (ovvero: se il PD è il partito delle élite)*, https://cise.luiss.it/cise/2018/03/06/il-ritorno-del-voto-di-classe-ma-al-co ntrario-ovvero-se-il-pd-e-il-partito-delle-elite/, March 6th 2018.

previous political experience, Gianroberto Casaleggio. Contrarily to the PD, the M5S can be considered anti-establishment and is labelled by some (Tarchi 2015; Chiapponi 2017) a populist party. The M5S is the most significant political innovation of the last decade in Italy. It was officially born as a national political subject in October 2009, but its history dates back to 2005 when Grillo created a blog that was to become the core of the organisational structure of the Movement (Ceri and Veltri 2017). The opening of the blog was followed by an invitation to citizens interested in the issues dealt with by Grillo (the defence of the environment and the fight against corruption) to meet using the online platform *Meet-up*. Thus, in various Italian cities, people, often without prior political affiliations, started to meet on the basis of a common interest in these issues and to form the first local groups (Biorcio 2015).

Over the years, the Movement has undergone many important metamorphoses. The 2013 general elections marked an exceptional success for the M5S. In a context characterised by the economic crisis and strong delegitimisation of politics and politicians, a party made up of ordinary citizens who engage in politics for a fixed period of time—a maximum of two terms are allowed by party rules—and who present themselves as the spokespersons of the people against the politicians' "caste", obtained 25.5 per cent of the vote at its first national test and the election of 161 MPs, upsetting the Italian political system. This sudden electoral growth triggered some transformations in the party's organisation and communication. As we will see, the main feature of the M5S's structure was precisely the absence of a traditional party structure, which had been replaced by the internet. This has partially changed over time. After five years in opposition, at the 2018 general elections the M5S became the first national party with 32.8 per cent of the vote and entered the national government, forming a coalition government with Matteo Salvini's Lega, experiencing the constraints of national government for the first time and breaking some taboos (Bordignon and Ceccarini 2019). After the end of the coalition with the Lega, from September 2019 to the end of 2020 the M5S governed the country with the Partito Democratico.

Despite this alliance, the two parties always had an ambiguous relationship. At first, it can be stated that the base of the nascent M5S and its electorate were oriented towards the left of the political spectrum. To begin with, the core issues of the M5S were those dear to the left, such as environmentalism and the fight against corruption (Vignati 2015a). Initially, voters mainly came from anti-establishment parties and from areas in which local social movements were stronger (Colloca and Marangoni 2013). Indeed, there were some attempts by Grillo to support the leftist coalition: at the 2006 general elections he pushed his supporters to vote for Prodi, the leftist candidate, against Berlusconi. Interestingly, in 2009 Grillo tried to participate in the primaries of the Partito Democratico, but the party rejected his application. Piero Fassino, then an MP, famously said: "If Grillo wants to do politics, he has to found a party and compete in elections. We will see how many votes he will get". Over time, the PD has increasingly been considered an enemy by the M5S. The PD is seen by the Movement as part of the establishment and, with the M5S entering into electoral competition, since according to the party rules the M5S could not form alliances with other parties at that time, also as an electoral adversary. After the 2013 elections—in which the M5S obtained 25.5 per cent and the PD 25.4 per cent of the vote (Chamber of Deputies)—Bersani, then head of the leftist coalition Italia Bene Comune and secretary of the PD, tried to form a coalition government with the M5S, but representatives of the party, backed by Grillo, refused the offer. The M5S then spent the legislature (2013–2018) in opposition, and the relationship with the PD, especially when Renzi became party secretary in December 2013, became more and more conflictual. In 2014, for the second time, the M5S refused to form a government with the PD, led by Renzi. The first meeting with Bersani and the second one with Renzi were famously streamed, according to the M5S's principles of transparency.

At the general elections of 2018, the M5S obtained 32.8 per cent and the PD 18.8 per cent of the vote. After long negotiations, the M5S put aside its reservations on alliances and formed a government with Salvini's Lega, the so-called "yellow–green" government<sup>15</sup>, led by Giuseppe Conte, a former law professor close to the M5S (Biancalana and Colloca 2018). After one year, in summer 2019, the Conte government fell over the TAV issue, in an attempt by Salvini to provoke early elections (Biancalana 2020) and to capitalise on his popularity after the particularly good results of the 2019 European elections (Biancalana and Colloca 2019). In September 2019, the fear of early elections and of the creation of a government led by Salvini, together with economic concerns, such as the approval of the 2020 budget legislation to avoid a rise in VAT, spurred the two former enemies to form a new coalition government, the second Conte government, dubbed "yellow–red"<sup>16</sup>. The creation of the coalition government with the PD, as had happened with the Lega, was approved on the M5S's part by an

<sup>15</sup> Yellow and green are the colours of the M5S and the Lega, respectively.

<sup>16</sup> In this case, red symbolises the PD.

online consultation on the online platform *Rousseau*. The government was supported by the M5S, PD, Liberi e Uguali and Matteo Renzi's new party: Italia Viva, a party with low public support but indispensable if the coalition was to have a majority.

The coalition between the PD and M5S was an unusual alliance between two parties that used to be competitors, in which the fear of early elections was the main stabilising factor. The unease of the two parties with this alliance is testified to by the fact that, in most cases, at the local and regional levels the alliance has not been repeated. Finally, at the time of writing (early 2021), Italian politics has managed to surprise us once again. The country is currently governed by a coalition including almost all parties and led by the economist Mario Draghi. The developments of this strange government will surely be the subject of future publications.

#### *Turin: Deindustrialisation, Local Politics and Local Conflicts 3.*

The research dedicated to the local level was conducted through fieldwork in the city of Turin. Turin is a large city (900,000 residents) located in the north-west of Italy and is the capital of the Piedmont region. For decades, Turin has been identified in the collective imagination, like other European and American metropolises (Pizzolato 2013), as a "factory city" or a "Fordist city" (Bagnasco 1990). In fact, the presence of FIAT, a company founded in 1899, strongly influenced the development of the city. The presence of a large manufacturing industry, in particular the automotive one, had, in the years of Fordism, an impact on the economic development of Turin, on its demography—with the great flow of immigration of workers from the south and north-east of the country, and the resulting urban expansion—as well as in social and political fields.

The magnitude of the power held by the large manufacturing industry was paralleled by the lack of autonomy in the political field with respect to economic actors and processes. During the years of Fordism, factories were central to the politics of the city, in particular that of the left. For instance, the Partito Comunista Italiano was characterised by a marked attention to the problems of the working class (Hellman 1988); unions and big companies acted not only as interlocutors, but also as actors in local politics. However, since the 1970s, and especially in the 1980s, just like other similar cities in other parts of the world, Turin has undergone a process of deindustrialisation linked to the transition of models of economic development towards what has been called "flexible specialisation". FIAT began a process of relocation, fragmenting the urban working class and favouring changes to the physical and social identity of the city. Consequently, from the beginning of the 1990s, Turin tried to focus on development models based on culture and knowledge. But, while de-industrialisation changed the social and political identity of the city, the transition to the new model of development happened only in part, despite the narrative carried out by the governing elite (Berta 2016).

But it was not only the process of deindustrialisation which, by changing the identity of the working class and therefore of the parties which aspired to represent it, affected the politics of the city. Indeed, national factors were also at play here. We can highlight two: on the one hand, *Tangentopoli* and the collapse of the traditional parties at the beginning of the 1990s; on the other hand, the reform of the electoral system at the local level, with the direct election of the mayor (law 81/1993). As we have seen, the *Mani Pulite* investigation, which entered public discourse under the name of *Tangentopoli*, revealed a vast system of political corruption which led to the dissolution of the main parties that had dominated the Italian political scene and to the growth of anti-political and anti-party sentiments already present in the country (Morlino and Tarchi 2006; Lupo 2013). In addition, the reform of the local authorities' electoral systems sanctioned the direct election of the mayor by the citizens. Mayors gained visibility and, sometimes, they became national figures. For these reasons, civil society gained in importance during the local elections of 1993. It's what has been called the "spring of mayors"; moreover, mayors were no longer just professional politicians and were recruited in areas such as business and culture.

To sum up, it can be said that in Turin's transition from a Fordist city to a new model of development, local politics played an increasingly important role. In 1993, within the city's Chamber of Commerce, a group of notables from the fields of economics and finance formed a group called Alleanza per Torino (Alliance for Turin), which aimed to influence the choice of the mayor, in order to promote a new model of territorial development. This proposal came from the world of industry and finance, and it was supported by intellectuals, professionals, entrepreneurs, representatives of Catholic and liberal associations, and former politicians of the PCI. The candidate mayor who was proposed and who also obtained the support of the moderate left (Partito Democratico della Sinistra, a party born from the ashes of the PCI) was Valentino Castellani, a professor at the Polytechnic of Turin (Belligni and Ravazzi 2012). Although not totally alien to local politics, Castellani can be seen as one of the representatives of the rise of Italian civil society. Castellani's victory in 1993 marked a period of discontinuity for Turin, due to the advent of a political class alien to the world of the parties which had previously dominated the politics of the city. This rupture led to the establishment of a regime that was to last several years: the same ruling coalition elected, after Castellani, who was mayor from 1993 to 2001, the mayors Sergio Chiamparino (2001–2011) and Piero Fassino<sup>17</sup> (2011–2016), who, unlike Castellani, are professional politicians.

We can say, then, that from 1993 to 2016 the city was ruled by a group with three main characteristics (Belligni and Ravazzi 2012). The first is that it was a small group, not open to external solicitations, not sociologically representative of the inhabitants of the city, and which shared a kind of ideological unity induced by social affinities. The second characteristic is related to the method of recruiting the municipal elite. In the new cycle of government, party membership mattered less than belonging to the same professional, family, or religious network. Finally, the third characteristic is that it was a group closed to the idea of citizen participation in political choices. In this sense, the relationship of the ruling elite with the citizens has been defined as "paternalistic", a circumstance favoured by the crisis in local parties and by the lack of functioning of participatory tools, such as district councils, thus making it difficult to connect the administration with the population. For these reasons, the ruling elite has been critically referred to as the *Sistema Torino* (Turin System) by political opponents, and especially by the local M5S.

Due to significant local environmental conflicts, Turin and Piedmont have always been of great importance for the M5S (Biancalana 2019). The M5S has been present in the region since the very beginning, both in connection with local mobilisations and with an early presence in the institutions. In 2008, the second V-Day—a big gathering of the M5S's supporters—took place in Turin. The first local groups linked to Beppe Grillo took part in provincial elections in 2009, even before the official founding of the M5S. In 2010, two regional councillors were elected to the regional council—the first ones, together with those from the Emilia-Romagna region. In 2011, the Movement obtained about 5 per cent of the votes and the election of two city councillors, and finally, in 2016, Chiara

<sup>17</sup> Fassino is an important figure on the Italian left. A minister on several occasions, he was national secretary of the Democratici di Sinistra (the party which followed the PDS in the evolution of the parties born from the PCI) between 2001 and 2007.

Appendino, a former city councillor, became mayor of the city, marking one of the most important electoral victories of the M5S. Moreover, in recent years, in Turin and Piedmont (especially in Val di Susa) a strong local conflict against the high-speed train lane Torino–Lione (TAV, Treno ad alta velocità) arose, and consequently a social movement emerged (the No TAV movement). The M5S has always supported the No TAV movement, and this had consequences both on the profile of the first activists and on the results of local electoral competitions (Mosca 2015; Biancalana 2020).

#### *The National Party Organisation 4.*

The two parties under investigation will be observed from three points of view. These three main topics will also constitute the three separate chapters dedicated to each party. They are: national party organisation, the use of the internet for direct democracy purposes and a local electoral campaign.

The aim of chapters 3 and 6 is to analyse the national organisation of the Partito Democratico and of the Movimento 5 Stelle, respectively. In order to complete this task, together with a general examination of the parties' internal structures, I will investigate the three IPD areas outlined by von dem Berge and colleagues (2013), namely the internal distribution of power with respect to the selection of the leadership, the selection of candidates and the determination of policies. The aim of chapters 3 and 6 is thus to investigate the two parties' organisations in order to assess whether they employ disintermediation strategies, how they implement and interpret them, and which dimensions amongst those outlined prevails. The analysis includes both the examination of party documents, such as party statutes and rules, and in-depth interviews with privileged witnesses, such as members of party personnel able to give valuable information on their party's internal organisation. As highlighted by Katz and Mair (1992), party statutes are relevant sources, but they tell only one, minimal, side of the story of the internal distribution of power within that party: the official one. Actual power relations could be partially different. Moreover, we said that, in order to understand party change, it is necessary to comprehend how critical actors *perceive* the stimuli that lead to change. So, interviews are useful both in order to have a different account of the state of power relations within a party, which is different to that shown in the official documents, and in order to comprehend how critical actors perceive change and the reasons for some strategic decisions. In this regard, interviews with privileged witnesses will allow us to grasp the discursive surroundings and narrative on party organisational change. In both cases, I analysed the evolution of each organisation from its foundation to the end of 2020. This allowed me to analyse the organisation not only in a static way but to observe the changes that occurred over time. This is particularly important for the M5S, a new party for which the evolution of the organisation equates with its institutionalisation, but also for the PD, in which leadership changes have shaped the party in different ways over time—although there haven't been substantial changes in the structure of the party.

The analysis of the M5S's organisational history will be carried out using Pedersen's "lifespan model" (1982), as adapted by Rihoux (2001). The lifespan model is a tool that has often been used to analyse the evolution of new parties, allowing researchers to examine the different phases of party development in terms of their links with the parties' organisational evolution. According to Pedersen, during their evolution, political parties pass through a number of phases or "thresholds". Not all thresholds are necessary for a political party to exist, and not all thresholds have the same weight, but there is a given sequence in the parties' evolution, and crossing some thresholds is considered to be particularly critical for the parties' development, since the new functions that the parties face will instigate important organisational changes. Adapting Pedersen's model, Rihoux identifies five thresholds: declaration (the decision to participate in national elections); authorisation (the acquisition of the formal requirements to be allowed to compete in national elections); representation (entering the parliament); institutionalisation (the first national electoral confirmation); and government participation (entering the government). We can expect that, with the passing of these different thresholds, the internal structure of the M5S would become more complex, possibly developing new forms of intermediation. In contrast, as the PD's formal organisation has remained substantially unchanged over time, in the case of the Partito Democratico we can identify four phases: a genetic phase—the phase that precedes the party's foundation—and four phases that coincide with the four party leaders that succeeded each other in the time span 2007–2020.

#### *The Use of the Internet 5.*

The use of the internet is a fundamental topic to be analysed when dealing with parties' disintermediation strategies. Consequently, two separate chapters here are dedicated to the analysis of the use of the web by the two parties. In reality, insights into the role and use of the internet by the two parties are also located in other sections of this work. For instance, in the chapters dedicated to members' and activists' participation, we can observe how the web changed the practices of participation of these actors. Moreover, in chapter 6 we will see that the internet has always been very important for the organisation of the M5S, both in terms of rhetoric and structuring the party's organisation. Even in chapter 3 we will see how the broad technological and societal changes brought by the spread of the internet have shaped the PD's organisation, in particular in the period when Matteo Renzi was the party leader.

Besides that, two separate chapters (chapters 4 and 7) are dedicated to the two parties' use of the internet. In these chapters, I will analyse the tools that can be included in the category of direct democracy, i.e. digital tools that could change the internal distribution of the parties' decision-making power in favour of their members. In particular, I will consider the online platform *Rousseau* for the M5S, analysing the online votes held in the time span 2012–2020, and the *circoli online* (online circles) experience and the mobile app *Bob* for the PD. The questions that chapters 4 and 7 endeavour to answer are related to the parties' use of the internet as a disintermediation tool potentially able to empower members, but also to strengthen existing or to create new hierarchies. The aim is then to assess whether or not these tools changed the power distribution within the parties and their impact on their organisation.

#### *Members' and Activists' Participation 6.*

The parties' disintermediation strategies consist in the creation of an unmediated connection between leader and followers, through the weakening of the party's intermediate organisation. In order to assess the consequences of this strategy at the local level, and the differences with respect to the national one, I focused specifically on members' and activists' participation practices, especially during a local electoral campaign. Against the backdrop of a collection of literature on parties that identifies the fall of party membership as the main sign of party decline—or better, the decline of the "mass" type of party—focusing on members' and activists' participation could seem pointless. In contrast, we know that party members continue to have an important role, mainly as organisational resources and a source of legitimacy (van Haute and Gauja 2015), and the picture of party decline is much more nuanced than it appears at first glance (Scarrow 2014). Indeed, van Haute and Gauja (2015, 6) acknowledged that "aggregate figures can't really provide an understanding of the changes that are occurring in membership, nor how people experience it". Furthermore, research on party membership at the individual level has typically focused on who joins parties, why, and what opinions they hold rather than on the activities they have undertaken.

For this part of my research, I conducted fieldwork in Turin. Fieldwork research, that is, an in-depth investigation of a few cases conducted through direct or participant observation and thick description (Cefaï et al. 2012) cannot lead to generalisations but can offer innovative insights into how members' and activists' participation unfold and change. Ethnography allows an understanding of "what is political participation, in the classic sense of the term, by reporting it as it happens and not as it is portrayed in interview accounts or in survey questionnaires" (*ibidem*, 10). Accordingly, the study was conducted in a particular place and in a particular moment. The place was the city of Turin and the moment was the 2016 electoral campaign for the election of the mayor. Electoral campaigns are indeed a relevant moment, if not the most relevant moment, in which to analyse participation within parties and can tell us a lot about parties' inner workings. Most research on electoral campaigns seems to be dedicated to the effects of electoral campaigns on voters. In contrast, in line with Agrikoliansky, Heurtaux and Le Grignou (2011), the aim of this work is to study the electoral campaign *au concret* (concretely), as a *fait politique total*. Studying electoral campaigns as "total political facts" means analysing them as being revealing of the logic and practices of politics. Indeed, electoral campaigns analysed "concretely" allow us to see and understand processes that are usually hidden.

More specifically, we can say that electoral campaigns are the field in which parties are said to have "externalised" their functions the most. The voluntary work of members and activists is said to have been replaced with that of professionalised campaigners, and the relationship between party members and citizens with an unmediated leader–supporters relationship enacted through television or social media. An in-depth analysis of an electoral campaign at the local level would allow us to verify these assumptions, at least with respect to the cases under investigation. "One of the challenges of the ethnographic perspective is to capture precisely what is happening at the local level, the complexity and richness of participation situations as they occur and participants' experience, with its innovations, its ambiguities and its paradoxes" (Cefaï et. al. 2012, 9).

The 2016 electoral campaign in Turin saw two main competing fields: on the one hand, the M5S as a challenger and outsider party; on the other, the incumbent mayor Piero Fassino, member of the PD (then in the national government) and supported by a centre-left coalition. If the M5S represented the need for radical discontinuity, the figure of Fassino, an important personality on the Italian left and in the PD, symbolised in contrast a form of continuity, perceived as the guarantee of safeguarding the interests of groups that had participated in the management of the city for thirty years (Berta 2016). Chiara Appendino's victory, the candidate of the M5S, who explicitly presented herself as an alternative to this system of power (the slogan of her campaign was *L'alternativa è Chiara*, The Alternative is Clear), testified to the choice of the city for change<sup>18</sup> .

With regard to the Partito Democratico, my focus won't be precisely on the party's local campaign. I shall instead analyse an innovative practice, that is, the employment of "volunteers" for the electoral campaign. Noi Siamo Torino (We Are Turin, NST) is the name of the organisation that managed this part of the PD's local electoral campaign: it is a managerial and professionalised organisation, external and separate from the party, that aims to recruit volunteers in order to persuade citizens to vote for the mayoral candidate, not in the name of party identification but by enhancing their experience as lay citizens. In chapter 5, I shall offer an in-depth analysis of this experience and of its relationships with the Partito Democratico. As regards the Movimento 5 Stelle, I will focus on examining how a local M5S group functions. Indeed, the analysis won't be focused only on the electoral campaign, but also on the organisation of this actor at the local level, which is partially different to the national one.

The national context is usually considered the primary institutional context that shapes party politics. However, in line with Deschouwer (2006, see also Detterbeck 2012), it is important to question and problematise the different institutional contexts (national, regional and local) in which political parties play their role. The context—made up of formal institutions, electoral systems and cycles, and societal heterogeneity—must be considered one of the crucial variables related to the strategic and organisational choices of political parties. Parties that, like the M5S, operate

<sup>18</sup> In the first round, Fassino won 41.8 per cent and Appendino 30.9 per cent of the vote. In the second round, Fassino 45.4 per cent and Appendino 54.6 per cent. For an interpretation of electoral results that consider the territorial dimension of the vote and the divide between the centre and the periphery, see Cepernich, Cittadino and Pellegrino (2018).

in a multi-level system, have to face problems of vertical and horizontal integration: they have to coordinate both their action on different levels (national and local, for instance) and the homogeneity of the various local or regional organisations. The intertwining of these different levels affects parties' strategies, organisation and also members' participation.

For instance, while for the M5S the most important form of participation at the national level takes place online, face-to-face participation is considered fundamental at the local level. This fact goes against the image of the M5S as a "cyber" (Margetts 2006) or "digital" (Gerbaudo 2019) party. Also in this case, the image is more nuanced. And for this reason too, it is interesting to conduct an in-depth analysis of this player at the local level, a level often overlooked by scholars (for the analysis of the M5S at the local level, see Biorcio 2015; Biorcio and Sampugnaro 2019). The method of participant observation, and, more in general, qualitative research seems particularly fitting for an analysis of the M5S at the local level, not only because it allows us to take the context into account, but also because the absence of official structures at the local or regional level makes it difficult to analyse, at least with the analytical tools commonly used for the study of party organisations, such as the examination of statutes and other documents. Being unable to rely on such documents, a researcher can rely only on the reports of interviewees and on participant observation, i.e. on a point of view that is internal to the organisation.

To sum up, chapters 5 and 8 are dedicated respectively to the organisation that managed a part of Fassino's electoral campaign and to a local M5S group. As a consequence of disintermediation strategies, the classic role of party membership could be substituted by an unmediated relationship between leader and individual supporters. The aim of the chapters is an in-depth analysis of participation practices of party members at the local level in order to assess if this assumption is true, and of the observation of parties' responses and adaptations at this level.

#### *Data and Methods 7.*

As it encompasses different topics and subjects, this study relies on various types of data. Besides the analysis of party documents (statutes, rules, regulations, codes that will be mentioned in the text), the data used include: 42 interviews, the results of a fieldwork analysis, and an examination of the parties' participatory platforms.

The interviews conducted can be classified into two categories. The first type of interviews is those with the so-called "privileged witnesses", that is, members of party personnel or other key informants who were well informed of the internal dynamics of the party and able to give me information on the internal distribution of the decision-making power within the organisation, through which it was also possible to uncover the discursive surroundings of the two parties' organisations<sup>19</sup>. The interviews of this kind were focused mainly on three aspects: party organisation; the role of members and activists within the party; and communication (in particular the use and role of the internet). In order to understand the motivations behind the creation and the functioning of online circles, the founders of the circles of Bologna<sup>20</sup> and Turin have also been interviewed. The complete list of interviewees can be found in the Appendix.

The second type of interviews is that with party members and activists. These interviews were conducted in parallel with the activity of participant observation in Turin. After a short self-presentation, the interviewees were questioned on their perception of their role within the organisation and of the internal dynamics of the local and national organisations. Moreover, questions were asked regarding the activities carried out within the party and their participation practices. For NST, in a partially different vein, the interviews had the purpose of investigating the social and political identity of the volunteers, their motivation for joining the party and their perception of the electoral campaign. As regards privileged witnesses, the criterion for choosing the interviewees wasn't the representativeness of the sample but the relevance of the information that each interviewee could give; as regards members and activists, however, I tried to maximise the diversity of opinions and biographies (age, faction, how long they had been enrolled, etc.).

The second type of data are the results of a fieldwork study conducted in Turin between March and December 2016. In that period, and in particular between March and June, that is, the electoral campaign period, I took part in the activities of three different groups: a local section (*circolo*) of

<sup>19</sup> I interviewed the PD's and M5S' privileged witnesses, but I decided not to interview privileged witnesses for NST. In that case, my close and prolonged contact with the organisers before and after the electoral campaign gave me all the information that I needed. Moreover, I had the opportunity to speak informally with the organisers during the writing of this work.

<sup>20</sup> The online circle of Bologna was the first one to be opened and one of the most important ones in Italy.

the Partito Democratico<sup>21</sup>, a district group of the M5S, and NST. During the electoral campaign I took part in meetings, rallies, flyer hand-outs, but also closed or open meetings, and I continued to meet with the PD and M5S groups even a few months after the end of the electoral campaign. My activity of observation was openly declared, meaning that I declared my identity as a researcher to the groups.

As for the PD, I entered the local section introduced by an MP well known to the members of the local section. The MP introduced me to an old member, and I began to take part in the activities of the section with his "protection". This man was highly respected by the group, and my participation posed no problem to them; on the contrary, everyone seemed to be happy to have me around. From the beginning, I clearly stated that my interest in the party was only a scientific one. Nevertheless, I had the impression that the PD members thought that I was a sympathiser of the party, and everyone was accommodating and kind with me. The interviewees were totally open, and members shared their concerns with me, the things that they believed wrong or unfair in the party (at the national and local level), as well as their personal sympathies and aversions within the local section and so on. The impression that I had was that they wanted and needed to be heard, and that they perceived my presence as a way in which they could express their concerns regarding the party.

As for the M5S, I came to the local section introduced by an interviewee, who was in charge of the electoral campaign of Chiara Appendino. There are 8 district groups in Turin—one for each district of the city—and he decided to "send" me to one of these. Following his advice, I contacted the district elected representative (that, as we will see in chapter 8, is a sort of local coordinator); she told me the date of the first meeting and I started to participate. In contrast to the case of the PD, here my presence was never totally accepted, at least by some members. In general, although I tried to look innocuous, my presence was almost always seen as a threat, perhaps for fear that I would spill their "secrets" to the public. In general, with respect to the majority of activists, the climate was of one of suspicion. There, I felt that they thought that I was not a sympathiser of the M5S. While some members considered me an activist, because I took part with continuity in the activities of the group, others avoided talking to me and also talking to others when I was present. My access to mailing lists and chats, as well as to important meetings, was denied. This attitude

<sup>21</sup> Data regarding the local section of the PD has been used only marginally in this work.

seems rather paradoxical for a party that makes transparency one of its trademarks, and can be considered, together with the rest, a result of the analysis. Likewise, interviews with activists were difficult to obtain and, despite the promise of anonymity, pretty standard; therefore, in order to hear some different voices, I had to talk to former members, who had moved away from the M5S.

As far as NST is concerned, I took part in the activities of the group for four months, from the beginning of March to the end of June 2016. During this period, I attended the weekly staff meetings and I took part in the activities of the volunteers on the ground, mainly, but not exclusively, in a particular district, a semi-peripheral and mainly residential one. I also took part in some public events of the electoral campaign in which the group was involved. I had full access to field operations, thanks to the organiser, Dr. Cristopher Cepernich. I had complete access to the field, to staff meetings, and to the WhatsApp chat of the group. The purpose of my presence in the field was declared and was always up to date. Moreover, the similar background that I shared with the staffers facilitated the task. My relationship with them was friendly, and everyone was keen to share their impressions on the campaign with me.

I didn't conduct interviews with the electoral campaign staff, as the interactions with them took place mainly "on the ground", while they were doing their job. However, I interviewed the greater part of the campaign volunteers. Constantly taking part in the campaign activities and interacting in a context of great trust with the participants gave me a privileged point of view of NST's internal functioning mechanisms. As regards the type of activities I took part in, during the electoral campaign I participated in all the electoral campaign activities (rallies, flyer hand-outs…), and in meetings. After the electoral campaign, I took part in some meetings of the PD and in weekly stands and district council meetings with the M5S. The number of times I took part is listed in the Appendix. The aim of my participant observation was mainly to observe member and activist participation as it unfolded. While doing that, I also tried to focus on the organisation and communication practices of the groups.

Both the M5S and the PD have sections/groups in each district of the city. Until 2015 there were 10 districts in Turin, and in 2016 they became 8. The city has about 900,000 inhabitants, and each district has 80,000 to 140,000 inhabitants. It is clear, then, that each district, and thus each group, has its peculiarities. This is particularly true for the PD, because local sections (*circoli*) are often a continuation of the old Italian Communist Party sections (*sezioni*), and each section is rooted in the very different history of the neighbourhoods of the city. But this also holds true for the M5S: each district has its problems and peculiarities; in each district the M5S partially attracted different people; finally, in a new organisation, without a structure, the quality of the people belonging to the group counts enormously. I could have interviewed one person for each district—the local section secretary for the PD, or the district elected representative for the M5S—but I decided to concentrate my efforts in order to carry out an in-depth analysis instead.

Like all choices, this came at a cost. In this case, the costs are related to the peculiarity of each district, which I will briefly describe here. The section of the PD that I studied is located in a central neighbourhood of Turin. The neighbourhood is characterised both by being an immigrant neighbourhood (the presence of immigrants, mainly from sub-Saharan Africa is high) and being one of the centres of the city's nightlife. The section was said to be one of the most active and leftist, but a change of the section secretary and the departure of some old militants following the election of Renzi as party leader created a sort of deadlock for the group. At the time of my observation, there were about 100 members, but about 25 were the most active. The local group of the M5S, in contrast, was in a peripheral district known for having been the site of a large car factory. The group didn't have official headquarters but rented a meeting room from time to time. About 25 activists made up the group, but about 15 of them, predominantly men of different ages and social background, were the most active ones. In the case of the M5S, for reasons that we will see in the following chapters, it is impossible to know the exact number of official members of the M5S in the area.

Finally, the last type of data used in my research is the analysis of the parties' participatory platforms and, more generally, of the online tools that can be classified in the category of direct democracy. These are the participatory platform *Rousseau* for the M5S and the *circoli online* (online sections) and the mobile app *Bob* for the PD. Most *circoli online* are no longer available online for consultation, so I had to rely on interviews for their examination. In contrast, at the time of my analysis the app *Bob* was still online and therefore I managed to examine its architecture (sections, available functionalities, affordances, etc.). I also tried to interview the person that various sources indicated to me was the only one in charge of the mobile app *Bob*, but after our first positive contact the person in question disappeared and so I had no opportunity to interview him. With regard to *Rousseau*, besides having conducted a general analysis of the architecture of the platform, I collected all the results of the online votes held from 2012 to 2020, creating a complete database. All the votes were classified by territorial level, turnout, as well as content and available options. Online votes are analysed in chapter 7, together with a broad assessment of the sections and functionalities present on the platform.

## The Partito Democratico and Its Organisational Innovations: The Consequences of an Open Party 3.

#### *Introduction 1.*

Mainstream parties in contemporary democracies are affected by several social and political changes that are likely to undermine their legitimacy. What we defined as disintermediation strategies could be seen as parties' responses, on the organisation's side, to the challenges of this new context. The aim of this chapter is to analyse whether the Partito Democratico (PD) employs disintermediation strategies, how it implements and interprets them and which dimension amongst those outlined prevails. The PD is a party that was born in 2007 from the merger of two existing organisations with deep roots in the history of Italian parties; it can be considered a relevant case to examine because it allows us to analyse the organisational innovations pursued by a mainstream party.

In this chapter, the organisational history of the PD will be examined from its foundation in October 2007 to the end of 2020. Five dimensions will be taken into consideration: the selection and role of the leader; the role of party members; the selection of candidates; the determination of policies; and the role and characteristics of the party's intermediate bodies. To analyse the organisational history of the party, it can be divided into five phases. The first is the so-called genetic phase, i.e. the phase preceding the foundation of the party. The other four phases correspond to the mandates of the four party leaders that succeeded each other in the time span 2007–2020<sup>22</sup> .

Two types of sources will be compared: on the one hand, party statutes and documents; on the other hand, interviews with privileged witnesses and party members. As highlighted by Katz and Mair (1992), party statutes

<sup>22</sup> Officially, there were seven party leaders in the time span 2007–2020, but three of them (Dario Franceschini, Guglielmo Epifani and Maurizio Martina) were elected by the national assembly (NA) of the party after the party leader's resignation, and their only task was to lead the party to the next congress. For this reason, they won't be taken into consideration. In early 2021, Nicola Zingaretti resigned, and the PD is currently led by Enrico Letta, elected as secretary by the NA in March.

tell only one side of the story, the official one, and then it is useful to compare them with the accounts of party members and party personnel regarding the actual distribution of power within the party. Moreover, we have seen that it is crucial to understand how actors perceive the stimuli that lead to change and the reasons for some strategic choices, as well as to grasp the discursive surroundings and narrative on party organisational change: interviews are the most appropriate tool to fulfil this goal.

#### *The Genetic Phase: The 2005 Primaries as the "Founding Myth" 2.*

The so-called genetic phase is fundamental in order to understand a party's development, as it can influence its organisation in the years to come (Panebianco 1988). Initial decisions can indeed determine the frame of constraints and opportunities within which the party will act in the future. For this reason, it is relevant to analyse the period that precedes the foundation of the PD. The PD was officially born on October 14th, 2007, from the merger of two existing parties: Democratici di sinistra (DS) and Democrazia è libertà – La Margherita (DL). These two parties are the heirs of the two most important Italian mass parties, the Partito Comunista Italiano (Italian Communist Party, PCI) and the Democrazia Cristiana (Christian Democracy, DC).

The idea of creating a major reformist party capable of uniting the progressive forces of social-democratic and Catholic-democratic inspiration and of transforming the structure of the political system in a bipolar sense, had its roots in the previous years (Salvati 2003) but only became a reality in 2007. The PD is therefore formally a new party, but in reality, it is a merger party (Ventura 2018) deeply rooted in the history of Italian parties. Bordandini, Di Virgilio, and Raniolo (2008) have indeed observed that the most important representatives of the new party come from the two major parties of the First Republic, and that militants, middle-level elite, and national leaders maintain close relations with the identities of the past.

The definition of the rules regarding the constituent process of the new party was related to a committee composed of 45 personalities from both the two parties and the so-called civil society. The committee had the task of deciding how to elect a constituent assembly (CA), which was also in charge of drafting the party statute, and the first secretaries, at both the national and regional levels. Two alternative models were proposed (Vassallo 2006): on the one hand, given that the new party derived directly from two existing parties, the CA could be composed by the delegates of the founding parties; on the other hand, the CA could be elected by the citizens according to the principle "one head, one vote".

These kinds of choices are not neutral and, as we will see, will have some consequences on the future of the party. The committee decided for the second option: the CA of the new party was elected directly by citizens with a proportional system. The party's first national secretary was also elected directly; each candidate was linked to closed lists for the CA. The selectorate was as broad as possible: all Italian citizens over the age of 16 and also foreign residents who committed to sign a charter of values and pay the minimum contribution of one euro could vote. This was an absolutely relevant novelty in the European political scene: it was the first time that the secretary of a party was elected not only by its members but, potentially, by the whole electoral body<sup>23</sup> .

Numerous studies have highlighted a trend towards the adoption of inclusive methods for the election of the party leader (Pilet and Cross 2014), but the innovation initiated by the Partito Democratico is a radical change in the method of his or her selection. In Italy, with the exception of some rare cases, this process has always been managed by party elites (Musella 2015, 231). Indeed, as pointed out by Bordandini, Di Virgilio, and Raniolo (2008, 315), these initial choices highlight a tension between "oligarchical dynamics", namely the selection of the members of the assembly using closed lists, and "plebiscitary dynamics", the direct election of the secretary. In this way, what was essentially the merger of two parties is covered by a halo of popular participation.

In fact, there existed "the real danger that this operation would become only a sort of «cold fusion», that is, that the reorganisation between post-Christian Democrats and post-Communists corresponded to nothing but the unification of the ruling groups of the two founder parties" (Natale and Fasano 2017, 7). As pointed out by a senior executive of the Partito Democratico, the foundation of the PD "could appear a political class agreement and instead we decided to turn it into a popular fact" (Int 16). According to another, "the crisis of politics needed an external and stronger response, and not for parties to close-in on themselves" (Int 2).

<sup>23</sup> For instance, the French Socialist Party inaugurated the use of open primaries in France during the 2012 presidential election (see Lefebvre and Treille 2016). It should be noted, though, that these are primaries in the strict sense, that is, for the selection of candidates for elective offices. Italy is one of the two European countries that use open primaries for the selection of party leaders, the other being the Greek Pasok. For a comparison between Italy and France, see De Luca and Venturino 2015; Giannetti and Lefebvre 2015.

There is also a specific event that contributed to making this choice to make the new party's constitution coincide with the direct election of the secretary<sup>24</sup> by such a wide-ranging selectorate—almost inevitable: the primaries for the selection of the 2005 prime minister, defined as the "founding myth" of the party and by some scholars (Vassallo and Passarelli 2016) as "the open party laboratory", as this primary election strengthened the building process of the Partito Democratico as an "open party". The primaries held on October 16th, 2005, legitimised Romano Prodi, the candidate of the centre-left coalition, as the prime minister candidate of the coalition who would run in the 2006 general elections<sup>25</sup>. These primaries were a crucial event for several reasons. According to one of their major promoters, the sociologist<sup>26</sup> Arturo Parisi, it was the first time "that the fundamental choices concerning the government are entrusted *directly* to the citizens" (emphasis added).

In fact, it was an open primary: those who had the right to vote at the next elections could vote, after having signed a political document and having paid at least one euro. Four million three hundred people voted, while there were about one million members of centre-left parties (Vassallo and Passarelli 2016). The 2005 primaries had a strong symbolic importance in two respects: on the one hand, they became a symbol of renewed and strong popular participation; on the other hand, it was, in reality, a poorly competitive primary election because, despite the presence

<sup>24</sup> The election of the secretary of the Partito Democratico is commonly called a primary election even if this is not completely correct. The term primary indicates, in fact, the selection of a candidate for public office. The use of this term can be partly justified by the fact that, at least until 2019, the party secretary was automatically the party's prime minister candidate. In the following pages, in line with the common use of the word by PD members, supporters, elected representatives, and commentators, we will call "primary" the direct election of the party secretary.

<sup>25</sup> After some limited local experiences in the late nineties, primaries have been organised with some continuity in Italy starting from 2004. These primaries have three characteristics: they are asymmetrical (they are held only by centre-left parties), they are coalition primaries, and, above all, they are open primaries (Venturino 2009). In addition, we can note that the success of the 2005 primaries has had the secondary effect of increasing the use of primaries in Italy (Venturino 2015).

<sup>26</sup> It is interesting to note that university professors (in particular political science and political sociology scholars) seem to have a relevance in the choices regarding the organisation of the new party. As I said, Arturo Parisi is a sociologist (as well as a politician). In the same way, Salvatore Vassallo (author of the paper on the organisational form of the party quoted above) is a political science professor.

of seven candidates, Prodi obtained about 74 per cent of the votes. The goal of the election was, in fact, more about giving popular legitimacy and strength to a "natural candidate" rather than actually entrusting the choice to citizens. This event also had some consequences on accelerating the founding process of the new party and on the configuration of its organisational model. As recalled by an interviewee strongly involved in the foundation process of the new party:

There are various reasons why the 2005 primaries were held. Some have to do with the idea of the open party that took shape then, which has different roots. The idea of forming a party with strong leadership, a party not fragmented into factions, had a particularly high risk since it was formed by the aggregation of two previously structured organisations and therefore meant creating a party that was able to present itself with a clear line, etc. […] There was precisely the need for the centre-left to replace the absence of a natural, strong leader, who has control of the organisation, with an additional investiture. But the other reason was also that […] we had a conviction that the willingness to participate in political life, in the life of the parties, was paradoxically much more widespread than the willingness to register, to become a member (Int. 4).

We can therefore establish some points regarding the genetic phase of the Partito Democratico, the marks that affected its evolution in the years to follow. On the one hand, the birth of the new party was essentially the merger of two existing parties, and therefore an elite operation. On the other, in a context in which anti-party sentiments were strongly rooted, the birth of the party was ratified with the direct election of the secretary open to all citizens, following the great success of the primaries of 2005, which are considered the "founding myth" of the Partito Democratico.

Both the 2005 and the 2007 primaries were not competitive, since their goal was to give what is called an "additional investiture" to an already selected leader, and to give citizens the idea that they were directly participating in the life of the party. It was therefore an operation with a strong symbolic rather than a practical value: opening the decision-making processes to the citizens, showing that the new party was not the result of a political agreement, but was open to citizens. Moreover, the direct election of the secretary aimed to create a strong leadership, legitimised directly by a popular vote. And through this tool of direct citizen participation, the goal was to encourage participation outside the traditional party structures.

#### *Open Party, Light Party, Liquid Party 3.*

With the primaries of October 14th, 2007, the Partito Democratico was officially born. The birth of the party was ratified by a primary election through which citizens (including 16-year-olds and foreigners with a regular residence permit) elected the members of the CA and the party secretary, at both the national and the regional level. In 2007 there were five candidates for the national secretariat, and Walter Veltroni (the former party leader of the DS and mayor of Rome) seemed to be the favourite, also because the majorities of the two co-founding parties supported him. Participation was lower than in 2005, but it was still high: three and a half million voters. Veltroni's victory, as in the case of Prodi, appeared then more like a legitimation of the natural candidate rather than the result of a real competition: in fact, Veltroni obtained 75,8 per cent of the votes (Pasquino 2009). The CA elected in 2007 was made up of roughly 2,800 delegates, and about 73 per cent of them came from the two founding parties, so there didn't seem to have been a substantial change in the organisational structure or ruling coalition, with the foundation of the new party. The CA had the task, through a special commission, of writing the statute, which was published at the beginning of 2008.

The model of the Partito Democratico has been defined by its promoters as the "open party" (Vassallo 2006). In fact, we can state that the fundamental feature of the organisational model, which makes it original compared to other political parties, is the opening of internal decision-making processes (selection of leaders, candidates for public offices and policies) to a very broad and inclusive public: not only the members, but those who recognise themselves as "voters" of the party. As we can read in the speech with which Veltroni presented his candidacy to the public, the so-called *Lingotto speech*, the "openness" that characterises the new party is pitted against "closed politics". A second fundamental characteristic of the party is its will to be a party that aims to have strong leadership.

The Partito Democratico, an open party that proposes, because it wants and needs it, to fascinate the millions of Italians who […] find politics closed, and when they try to get closer to it, it is easier for them to come across the request to join a faction or a power group, rather than an idea or a project […]. In the Partito Democratico everyone will and must be, right from the start, at the same level as the others. This is why we wanted the principle "one head, one vote" […]. One method, direct election, was chosen, certainly knowing what the consequences are for having it as an internal life model. I had, and I still harbour many doubts, but so it is. Strong leadership must exercise all its prerogatives, none excluded, and must know how to do so by listening and sharing (Walter Veltroni, *Lingotto speech*).

The greatest organisational innovation of the Partito Democratico lies in article 1.2, dedicated to the principles of internal democracy, which states that "the Partito Democratico is a federal party *consisting of voters and members*" (emphasis added). The Partito Democratico, in fact, "entrusts the fundamental decisions concerning the political direction, the election of the most important internal offices, the selection of candidacies for the main institutional positions to its voters" (article 1.3). Therefore, there are two "subjects of the internal democratic life: members and voters" (article 2.1).

The voters (article 2.3) are those who "declare themselves willing to recognise themselves in the proposals of the party, to support it in elections and agree to be registered in the public register of voters". The voters have the right to (article 2.4) participate in the definition of the political direction of the party through the direct election of secretaries and assemblies at the national and regional levels, participate in primary elections for the selection of party candidates to the main institutional offices, propose their candidates should hold institutional positions, take part in thematic forums, vote in the referenda open to the voters, and take part (without voting rights) in the assemblies of the local sections (*circoli*).

In addition, members (article 2.5) can participate in the direct election of secretaries and assemblies at the levels lower than the regional one, be consulted on the selection of candidates, vote in referenda reserved for members, participate in the formation of the party's political proposals, apply for the party's governing bodies and endorse candidacy proposals. In short, "members can apply for the party's governing bodies, endorse candidacies, and they are the only ones to vote for local and provincial governing bodies; the voters can take part, in addition to the primaries in the proper sense, in those for the election of the national and regional secretaries" (Int. 4). As noted by some commentators on the statute (Floridia 2009; 2019), the boundaries between the rights of the two subjects, voters and members, are very blurred. Indeed, there seems to be a supremacy of the former over the latter—who are also, significantly, listed first in the statute. As stated by an interviewee who is part of the national organisation office of the party:

The Partito Democratico gives shares of its sovereignty to the voters: through the involvement of voters we elect our national secretary, and then we choose our prime minister candidate; through voters we choose our mayors or candidates for monocratic offices; it is clear that voters have a very significant role, which is different from the "normal" tradition in which members are those who hold the power to choose, to make this type of choices. This has also led, and it is a discussion that exists within the party, to the consideration of members playing a marginal role, because… Why do I have to register if I am able to exercise my powers, the most important ones, even by being a simple voter? (Int. 17).

Such a configuration of membership, so different from the role traditionally granted to party members, has different motivations. The first involves "the crisis of parties as channels of political participation" (Vassallo 2006).

The moment in which the party is born and its fundamental documents are written, it is given this definition of membership. There was an awareness that parties' traditional organisational structure was no longer holding firm because people do not enrol anymore, because people do not see the neighbourhood as the place in which politics happen anymore. Because membership in a party can also be membership related only to some issues, not necessarily an entire ideology that explains the whole world. There is also a certain rejection of mediation in favour of more immediate models of direct participation. Do not forget that the Second Republic was born from referenda, from the referendum movements (Int. 3).

It is not a question of replacing today's members with the "people of the primaries", but the primaries teach us that political participation can be considered attractive by a wide and heterogeneous group of people if it does not imply too demanding (all-encompassing) a form of "belonging" and if it has, in the perception of the participant, an immediate, recognisable, relevant effectiveness […]. Therefore, if the Partito Democratico really wants to be open it must provide a form of individual membership that is easy, simple, immediate, and user-friendly, to paraphrase computer jargon. With an adhesion that obviously does not exclude (and indeed is perhaps a prelude to) more intense and stable militancy, it is clear that there will be different intensities of participation and exercising the rights connected to membership […]. Membership must imply a right to participate directly in the main choices regarding the party's political direction and in the selection of its governing bodies (Vassallo 2006).

A second motivation for such a membership configuration is to make allegedly "non-manipulated members" enter the competition, in order to allow the creation of strong and contestable leadership.

It was necessary to introduce non-manipulated members into the competition because this was one of the ways to question the consolidated establishments that were consolidated thanks to intermediation, that is, thanks to the fact that, previously, members were largely manipulated by a group of party officials (Int. 4).

What does this strong and contestable leadership look like? According to article 3.1 of the statute, "the national secretary represents the party, expresses its political direction on the basis of the platform approved at the time of his election and is proposed by the party as the candidate for the office of prime minister". The secretary can stay on a maximum of two mandates of four years, unless if at the end of the second mandate he holds the office of prime minister for the first term. The secretary is elected in a three-step process (article 9, see also Venturino 2015). In the first phase, candidacies are submitted to members in an OMOV (one member one vote) procedure. The first three candidates are admitted to the second phase if they have obtained at least 5 per cent of the votes and at least 15 per cent in five regions. The second phase is the open primaries: voters are called to vote for the leaders and the connected lists for the national assembly (NA). In the third phase, the assembly elects the secretary if he has obtained at least 50 per cent of the votes. Otherwise, a run-off occurs within the NA, between the two candidates who have received the most votes.

The fundamental innovation of the Partito Democratico therefore lies in the fact that the leader draws his support from the outside, and no longer from a more or less restricted group of people, that is, party members. Moreover, the internal bodies (the national assembly and the national directorate) are composed in proportion to the vote for the secretary, thus configuring a balance of internal power that derives from the vote, not by the members, but by the voters. In sum, primaries foster the autonomy of the party leadership *vis-à-vis* the organisation (Sandri, Seddone, and Sozzi 2020). As stated by a senior executive of the party:

[The primaries] have changed everything; before you had to have the support of a narrow group of people, your members; now you must speak to a wider public. You don't win a congress if you convince the members, you win it if you convince the voters and then you need to have tools and methods to speak with wider public opinion and this changes the structure, the dynamics of the internal life of the party (Int. 16).

External legitimation is linked to another fundamental characteristic of the party model of the Partito Democratico: the equivalence of the party leader and the prime minister. In article 18.8 it is indeed stated that "if the Partito Democratico adheres to coalition primaries for the office of prime minister, the only candidature that is admitted among the members of the Partito Democratico is that of the national secretary".

As regards the selection of candidates for public office, the Partito Democratico institutionalises open primaries for the selection of candidates as a method. According to the statute, candidates for the office of mayor, president of the province, and president of the region will be chosen through the use of coalition primaries. If coalition primaries are not held, there will be party primaries. Even in this case, both members and voters can participate, thus conferring great decision-making power, considered a classic function of the parties, to all citizens. With regard to the candidacies of elected representatives to the assemblies, the statute states that the selection takes place at every level using the primary method or, also in relation to the electoral system, other forms of "broad democratic consultation" (article 19.1). However, in its first national electoral test, the general elections of 2008, the Partito Democratico did not use primaries for the selection of its candidates to the parliament: they were chosen by the party's governing bodies.

In 2008, the candidates for parliament were chosen through a selection made first among candidacies emerging from the local sections, then the provincial and regional offices created a first list, then the regional secretary was sent to take part in a national table of candidacies and at this national table we tried to find some points of equilibrium between the national requests, the requests of the factions, the requests of the territories, and at the end those who started in the first positions could find themselves outside the list, and vice versa... Let's say a sort of slaughterhouse... (Int. 2).

Even as far as "the elaboration of the programme" is concerned, the statute of the Partito Democratico confers many rights to its members and voters. In particular, there are three tools listed in the statute: the thematic forums (article 23), the annual programmatic conference (article 26), and the internal referendums (article 27). The thematic forums' aim is to involve citizens (members and voters) in the drafting of programmatic proposals, while referendums are intended to allow their participation in the formation of party decisions. According to the statute, the national secretary, the national leadership with the favourable vote of the absolute majority of its members, 30 per cent of the members of the NA, and 5 per cent of the members of the Partito Democratico can apply for a referendum. According to the statute, referendums can be deliberative or consultative, reserved for members or voters. Despite their innovation ("it represents politics that choose to delegate a decision to members", Int. 2), these tools have never been used.

The internal referenda [have never been used] because in reality despite all the rhetoric, this idea does not make sense in a party. And this is due to the fact that internal referenda, if systematically used (and they are very complicated because… Find me more than four arguments that can be reduced to a yes/no in terms of policy...), they can only have two systematic applications: a plebiscite by the party leader or an internal opposition campaign that aims to put the party leader in difficulty on issues for which he may be in the minority at that particular moment (Int. 4).

As regards the party's intermediate bodies, there are three internal organs that make up the party: the national assembly (NA), the national directorate (ND), and the national secretariat (NS). The NA (article 4.1) is made up of 1,000 delegates and is responsible for the direction of the national party policies, the organisation, and the functioning of the national executive bodies. It lasts for four years and can impeach the party leader. The ND (article 8.1) has the task of implementing the decisions of the NA and is a political body. It meets every two months and is made up of 120 members elected by the NA with a proportional vote, and therefore reflects the balances defined by the voters' vote. The NS has executive functions (article 7.1) and is appointed directly by the party leader. Moreover, the local sections (*circoli*) (article 14) are the basic organisational units, and they can be of three types: territorial, environmental, and, as we will see in the next chapter, online.

What are the fundamental features of the organisational model of the Partito Democratico, from the analysis of its statute and the first secretariat? The most important feature and the greatest novelty is the direct participation not only of its members but of the citizen-electors in the fundamental decisions of the party (hence the label of open party). These fundamental decisions do not only potentially concern the selection of candidates for public office and the selection of policies, but above all the election of the secretary, who is also the candidate for prime minister, and the governing bodies of the party. There is a direct and external legitimation of the leadership, which is therefore potentially more contestable.

We can therefore see that in the party model of the Partito Democratico there are the two dimensions of disintermediation: the opening of decision-making processes even beyond members and the strengthening and greater autonomy of the party leadership. A consequence of this great, even though mainly symbolic, opening of the decision-making processes is the loosening of the party's organisational boundaries. The status of being a member does not give significant additional powers compared to those of a voter, except for the possibility of endorsing candidacies, voting for the governing bodies at the lower level, and becoming a member of the party's governing bodies.

The Partito Democratico thus seems to have responded to the crisis in parties by opening itself up to its voters. However, the voters enjoy individualised participation, which is expressed basically at the time of the elections and does not have mechanisms of accountability, thus potentially leading to greater autonomy of the leadership. In fact, there is a sort of personal mandate for the party leader, an unmediated relationship between leaders and citizens, to the detriment of members and middle-level elites. This pre-eminence of the leader over the organisation is strengthened by the fact that the intermediate organs are established starting from the voters' votes (NA and ND) or nominated directly by the elected leader (NS). To sum up, on the one hand, there is a will to strengthen the party's leadership through the direct and unmediated participation of voters:

[we had] the need to reduce the weight of the intermediaries, that is, their weight and influence on the internal party structure of the network of officials who are perhaps not so useful in a phase like the one we are living in, but certainly useful for the maintenance of positions within the party establishment (Int. 4).

On the other hand, we note that, in the PD, there are intermediate bodies and precise procedures of internal democracy, which limit both the leader's and the voters' power. What is interesting is the tension between persistence and change: despite the organisational innovations, the intermediate bodies are still mainly composed of personnel who come from the old parties. Moreover, against a rhetoric of openness and the direct participation of citizens, the process of leadership selection results in the legitimisation of a natural candidate. The most innovative procedures, for instance those for the selection of policies, are not actually implemented.

#### *The (Apparently) Solid Party 4.*

Walter Veltroni resigned in February 2009, fewer than two years after his election as secretary, following the party's electoral defeat at the 2008 general elections and at the regional Sardinian elections of 2009. There was therefore no time to consolidate either his leadership or the party's organisational model. Dario Franceschini was then elected by the national assembly as *ad interim* secretary and new elections for the leadership of the party were called: the first held in accordance with the rules contained in the statute. As we have seen, the candidates and their political-programmatic platform are first voted for by members using an OMOV method.

The first three candidates and the lists connected to them are then voted for by the voters. The candidates in this case were Pierluigi Bersani, Dario Franceschini and Ignazio Marino. Three million one hundred supporters participated. Bersani won with a much smaller majority than Veltroni (53.2 per cent), and the members' and the voters' votes were almost identical. We can therefore say that these primaries were more competitive than those in 2007. The platform with which Bersani, a long-term politician who comes from the tradition of the PCI, presented his candidacy is called *Identity, Territory and Organization*. Against what was called the "liquid" or "light" party, Bersani's promise was to give greater solidity to the PD project, even from an organisational point of view.

The question we asked ourselves in recent months is not whether we are an "old" party or a "new" party, but are we really a party: a free association of citizens with a recognisable identity, internal organisation, social roots, places of discussion and participation, as well as accepted and shared rules? Not having clarified these fundamental points has weakened the initial path of the Partito Democratico. In the aftermath of the primaries, we have disappointed both those who were linked to more traditional forms of militancy, and those who expected new forms of political participation and social involvement. We have lost an immense asset, cultivating a senseless juxtaposition between voters and members, whilst the voters are asking us for a more organised presence in the territories and in society […]. A party is organised in local sections (*circoli*) present in every municipality or district, in work and study places, in communities abroad, but it can really open up to voters only if it is rooted and recognised in the country […]. The Partito Democratico is *a party of members and voters* who pursue gender equality in political responsibilities. *Sovereignty belongs to the members*, who share it with the voters on the occasions regulated by the statute. Fundamental rights such as the participation in decisions at various levels (including referenda) and the election of governing bodies are acknowledged to the members. The Partito Democratico involves voters, through the primaries, to select candidates for elective offices with particular reference to the elections in which there is no preference vote. It takes part in coalition primaries with a representative chosen by the members and governing bodies. The primaries for the election of the national secretary require new rules inspired by two criteria: they must not be turned into a plebiscite and cannot be distorted by other political forces (Platform *Identity, Territory and Organization*, 2009).

At this stage, despite the emphasis on the organisational dimension, there were no significant statutory changes. What the Bersani secretariat tried to do, and this aspect can be clearly seen in the text on the platform, is to "provide meaning" to the participation of members, thus trying to shift the decision-making power from voters to members. In the absence of statutory changes that redefine the rights of the members, this seems a difficult task to achieve. However, the attempt seems to have been successful, since members seem to have the *perception* that they count more within the party. This perception derives from different elements.

In the first place, there truly is a greater consideration for the local sections and members, who are encouraged to take part in numerous campaigns that come from the national level. And this seems to work, although this participation is, according to the party executives interviewed, an end in itself, with the sole purpose of involving members (participation, eloquently reported as "dig a hole, fill the hole"). Secondly, there is an ideological proximity between the militants who, in the majority of cases, come from the PCI-DS tradition, and the new secretary and governing bodies. Finally, members were enthusiastic about the genuine expectation of winning the 2013 general elections. Although, as we have seen, there are no procedures and channels, apart from the primaries, through which one could directly influence political and party decisions, the interviewees perceived, had the idea, they were being heard.

There has never been a direct channel through which the discussions that you made in the local section came directly to the national level, but there were times when the base was consulted and the party tried to follow this path, to convey the idea—and it was often true—that if you participate in a local section, somehow your participation counts. You are not one of many, but together with the other members of the section you can influence the decisions of the party. Of course, you can't influence decisions on national policies, but with a path through many levels… You bring a discussion into the *circolo*, which is brought to the provincial level, then possibly at the regional level and so on […]. So, it was never the individual member that counted, but you conveyed the idea that there was a path with several steps... The pyramid had the right shape, you start from the base and then you slowly climb until you reach a synthesis that is expressed by a national representative (Int. 7).

There were no statutory changes regarding the election and the role of the secretary in this phase. A substantial novelty, however, can be found in the coalition primaries for the selection of the candidate prime minister (called at the end of 2012 for the 2013 general elections) in which, according to the statute, only Bersani could take part. During these consultations the challenge by a young member of the party, Matteo Renzi, took shape. Matteo Renzi was the mayor of Florence, elected through primaries, and he had been carrying out a strong critique, mainly on a generational basis, of the party's establishment for several years. Renzi's critique of the party was that it was not able to keep its promises of renewal, given the persistence of the ruling coalitions of the two founding parties. The provocative proposal by Renzi was to "scrap" the old ruling class of the party, in favour of a renewed political class, without ties with the political traditions of the past (on the 2012 primary elections, see Corbetta and Vignati 2013).

After a great internal debate, article 18.8 of the statute, the one that states that only the party leader can take part in coalition primaries for the role of prime minister, was suspended, and Renzi was therefore allowed to take part in the consultations. Five candidates took part in these primary elections, two of which (Bersani and Renzi) came from the Partito Democratico. This was a two-round consultation with a run-off. In the first round Renzi got 35.5 per cent, while in the second round, as expected, Bersani won and thus became the prime minister candidate of the coalition. Another novelty was the method for selecting candidates for the parliament. Given that the electoral system established closed lists, the general anti-political climate and the fact that the M5S had already organised online primaries for the selection of its candidates, it seemed clear to the leaders of the Partito Democratico that it was no longer possible to decide on the candidates in "smoke-filled rooms", as before.

We made the choice that seemed natural to us, considering the electoral system present in that phase, with the closed lists. Bersani didn't want to […] seem to be the man who wants to decide on everything and make lists in his image and likeness, but we wanted to leave the choice up to the territory, except for some candidates, chosen by the centre (Int. 2).

These are the so-called *Parlamentarie*, which are used to select and order 90 per cent of the candidates for parliament. 10 per cent of the candidates are nominated by the ND, while the candidates in the first positions are proposed to the ND by the secretary, after having heard the regional secretaries. So, 782 candidates out of 918 are selected by citizens (or rather, by those who had taken part in the primaries for the selection of the prime minister). In reality, we can state that the *Parlamentarie*, although representing an opening to the outside of the party, were highly controlled from above. Firstly, in addition to the number of "reserved" places, it is necessary for candidates, except the incumbent MPs, to collect signatures among the members. Secondly, it is the provincial office that defines a first draft of the list of names. Finally, the regional offices have to assign a precise number of places on the list to the various provinces, and the ND has the task of approving the final lists.

To sum up, in this phase, against a party model defined as open, Bersani's attempt was to strengthen the organisation, and above all the centrality of members with respect to voters, reinforcing the party's boundaries and internal organisation. It is necessary, however, to be careful not to get trapped in the contrast between the so-called "light party" and the "solid party". These seem to be more *narratives* of the party than really different organisational configurations. Moreover, in the previous phase, we noticed how party organisation had not disappeared; on the contrary, there was a persistence of personnel belonging to the two previous parties. Overall, the party model remained substantially unchanged.

We can also note that there was an attempt to restore the centrality of the party members. However, given that the statute did not change, the members' impression of counting derived more from an ideological affinity with the leadership than from a real possibility of influencing political and party decisions. Moreover, we can note that during the Bersani secretariat a further opening took place, both inside and outside the party, with the primaries for the candidate prime minister and with the *Parlamentarie*, where we see that a number of the candidates for parliament were chosen by citizens. This seems to contrast with the rhetoric of the solid party, even though in both cases we note that the procedures and the "rules of the game" were controlled from above.

#### *Renzi: The Normalisation of the Challenge from Outside 5.*

After the defeat of the PD in the 2013 general elections (ITANES 2013) and the failure to elect the president of the republic (Seddone and Venturino 2015), Bersani resigned and the NA elected Guglielmo Epifani as secretary, until the new congress, called for December 2013. Notwithstanding the efforts to create strong leadership, six years after its foundation the Partito Democratico elected its third secretary, which was a sign of instability and poor institutionalisation of the new political entity.

There are two trends that need to be highlighted to understand the context of Renzi's rise to the leadership of the party. The first is the sharpening of the crisis of representation in Italy: the 2013 elections (which were defined by scholars as an "electoral earthquake", Chiaramonte and De Sio 2014) were characterised by a decrease in voter turnout and very high electoral volatility, due to the great success of the M5S. The M5S, a new anti-establishment party at its first electoral test, got 25.5 per cent of the votes and broke the bipolar dynamics that had been established during the so-called Second Republic. A second trend that we can observe, which underpins the transformations of politics in general and of the PD in particular, is the growth in the use and relevance of digital media and social networks (Ceccobelli 2017). It is in this context that Matteo Renzi found room to consolidate his challenge to the party's establishment.

At the 2013 primaries, participation decreased to two million and eight hundred thousand; regarding the results, an interesting datum was the difference between the outcomes of the two voting phases: the one reserved to members and the one open to voters. We see (Table 3.1) that Renzi also won among members, but with very different percentages, and his victory among voters came especially at the expense of Gianni Cuperlo, an exponent of the left of the party. It is worth synthetising the results of the 2009 primaries and those of 2013 in a table. And it is important to remember that the balance within the governing bodies of the party is determined by the votes of the voters, and not of the members: Renzi's success was constructed not within but outside the party.


*Results of the 2009 and 2013 PD primaries: first and second phases Table 3.1.*

Thanks to the constraints and opportunities provided by the party model, Renzi's candidacy sought external support rather than internal support. The party therefore seems to have adapted "to a tendency that was implicit in some original choices" (Int. 4). Renzi seems to have understood and exploited the potential of the so-called open party in order to conquer the party leadership through direct and external support, provided by a broader base than by the party members alone.

When Renzi took part in the primaries against Bersani in 2012 he had the support of three deputies... Three, it's no joke. Of all the parliamentary groups of the Partito Democratico only three MPs supported Renzi. So, it was a minority position that found extraordinary support outside the Partito Democratico (Int. 3).

Renzi was perceived as a foreign body to the party by a part of the political and ruling class. It was, for the first time, a leadership that was imposed, not identified with and shared within the party, but which imposed itself from the outside (Int. 17).

Even in this case we see that there were no relevant statutory changes: the role of the leader on paper did not change, what changed was the style, the way the leadership was played out. Leadership, in fact, "varies according to personalities, depending on the idea that one has of the organisation of the party, and of politics itself" (Int. 2). As regards the role of the leader, Renzi, although positioning himself as a continuator of the Veltroni line, the line of the "light party" with a "majority vocation", represented, according to the interviewees and scholars, a break "in the method, in the communication; a political and even symbolic break" (Int. 17), carrying out unprecedented communication and agenda-setting modes (Natale and Fasano 2017). Other scholars stated that "his rhetoric, his direct approach to the party and supporters, together with a (conscious and capable) use of new media, suggest an approach centred on the leader that has no precedents in the history of the Partito Democratico" (Seddone and Venturino 2015, 487).

What were the characteristics of Renzi's leadership? First, and we also see this from the primaries' data, its approach consisted in seeking support outside the party and outside the boundaries of the traditional members of the party. In addition, there are two features of Renzi's leadership that emerge from the interviews collected and the analysis of the literature (Bordignon 2014; Seddone and Venturino 2015; Bobba and Seddone 2016): direct communication with followers, also through the skilful use of social networks, and strong and personalised leadership, both outside and within the party. With regard to communication, digital tools allow fast, direct and potentially bidirectional communication between a leader and their followers. As stated by a journalist working for the party's website (Int. 1), "the main feature of Renzi's communication is direct communication without intermediation". Therefore, an unmediated relationship between leadership and citizens is established through direct communication and identification with the leader, although this is a predominantly top-down relationship. As stated by a member of the national organisation office of the first Renzi secretariat:

The transformation of the party in recent years has been quite marked [in the sense of] a much more direct relationship between leadership and our militancy and our voters [that currently] is still very much based on the top-down dimension, so it is the leadership that communicates with the base (Int. 17).

Even on the internal side, i.e. the dialectic within the party, the interviewees refer to a more direct and decisive style, also because of the large majority in the party's governing bodies. This is seen as a problem by an executive that was in charge of the national organisation in the Bersani secretariat:

Once members held a power that derived from the fact that their representatives, within the party, proportionally had the power that they [the members] gave them: so, if a secretary got 60 per cent of the votes, there was a 40 per cent that still had power in the representative bodies of the party... If, amongst the members, one secretary gets 40 per cent but later gets 20 per cent among the voters, where is this power transmitted? It is transmitted to the leadership that, without having obtained those votes from the members, will govern the party with a strength that cannot be counterbalanced, because the members will be represented by a 20 per cent in the governing bodies, despite the choice made by 40 per cent […]. The difference that is created between the voters' and the members' data does not turn into power for a broader body, i.e. the voters, but it goes directly to the leadership. And so, there is an improper accumulation of power, compared to the normal situations of a party in the past (Int. 2).

This has had consequences on the internal life of the party, in the direction of greater centralisation of decision-making processes.

The minority does not exist today in the Partito Democratico because there are no places to exercise that margin of possibility that the congress has entrusted to those who did not vote for Renzi, because the DN is a voting place; Renzi does not even reply at the end of the debates, Renzi makes the introduction and then there are the discussions—five minutes each—often Renzi doesn't even reply, and we vote... Obviously the majority is taken for granted (Int. 3).

A few months after his election as party leader, in February 2014, strengthened by the popular legitimisation of the primaries and having a large majority in the DN, Renzi proposed that the national directorate of the party withdrew its support for the government led by Enrico Letta (who was also a member of the PD) and became prime minister of the multi-party government formed after the 2013 elections. For the first time, therefore, the roles of the party leader and the prime minister converged, as stated in the statute. This had consequences on the relationship between voters and members, and on the perception of the political effectiveness of the members.

The change that has taken place in the party in recent years has greatly transformed the role that each of us had in mind as party members […]. Now you feel you don't count at all. Because you have a national level in which the proposals, especially in the first phase of the government, arrive very fast, because there is a need for communication... Let's say that Renzi is very good, from this point of view […]. This speed does not allow you to hold a discussion in the *circoli*: if you have the prime minister announcing that in five days' time a law proposal will arrive in parliament, you feel a bit discouraged to organise a meeting in the *circolo* where you talk about that law […]. In an ideal world, from the *circolo* point of view*,* you should have a national secretary who opens a broad discussion on that topic, who lets the *circoli* have their internal discussions on that specific topic, who makes a synthesis at the provincial level and at the higher levels, up to the DN, where you take note or otherwise gather opinions (Int. 7).

But perhaps members' discontent during Renzi's secretariat is to be found more in the lack of political affinity between the leadership and members, rather than in the lack of political effectiveness of the members. As pointed out by a senior executive, in reality the militants "did not count at all even before" (Int 16). In fact, if on the one hand the perception of not being able to change the party line was heightened during Renzi's secretariat, on the other it emerged that the real problem was that the majority of members did not recognise themselves in Renzi's proposal. It seems, however, that Renzi's election as secretary sharpened members' discontent, as they did not feel involved in political and party decisions, and were used as a "labour force" during electoral consultations. If, on the one hand, there is the perception of the breaking of the so-called "transmission belt" between members and party; on the other hand, it seems that the party was more open, through digital media, to external solicitations. These are individualised and unmediated stimuli, such as the collection of inputs via the party website and of comments on social networks.

The horizontal and reticular dimension of social networks is exactly an answer to this [the problem of the relationship between members and party] because it shortens the distances... If today Renzi takes pieces of e-mails that he receives or questions to the *Matteo risponde*<sup>27</sup> and transforms them into pieces of his political discourse as a secretary or suggestions for governmental activity, in this sense, I believe that the organisational and not just the communicative power [of social networks] is strong (Int. 18).

It has been thought that it is enough to write what one thinks on the internet and nothing else, but the discussion and then the creation of a common opinion has a different power compared to when everyone writes their opinion under a post... That is *Matteo risponde*, a way of gathering opinions that is neither right nor wrong, in my opinion, but it does not reflect the decision-making process that a party must have. Because a decision, an idea, is formulated with time, with discussions, with experience, and it cannot be just a Q&A (Int. 11).

The European elections, held in May 2014, certified the very high approval ratings among citizens towards the new prime minister Matteo Renzi: on this occasion, the PD obtained its highest result ever, 40.8 per cent of the votes. Abstention was, however, high. The policy style of the Renzi government has been defined as "founded on leadership" (La Spina 2016, 31), and the pillars of its government (February 2014–December 2016) were institutional, public administration, and labour reforms (Salvati 2016). The issue of institutional reforms was certainly the most relevant. The Renzi government was the promoter of a constitutional reform project whose main objective "was to restructure parliament through a series of changes that would do away with symmetrical bicameralism (a term used to describe a system with two chambers which have the same powers and the

<sup>27</sup> *Matteo risponde* is a live video chat on Facebook where Renzi answers the questions made by users.

same functions). In combination with the new electoral law, the so-called *Italicum*, the reforms would have had a significant impact on Italian institutional arrangements and on the mechanics of the whole political system, producing a clear shift towards a *majoritarian democracy*" (Ceccarini and Bordignon 2017, 281).

In December 2016, a constitutional referendum was called to approve the reform. The strong personalisation that characterised Renzi's leadership in this case proved to be a double-edged sword: the referendum became a vote "on Renzi" and his government and 59.1 per cent of citizens voted against the reform (Pritoni, Valbruzzi and Vignati 2017). Following this defeat, Renzi resigned as prime minister and secretary of the party, but in April 2017 he ran again for the role of secretary of the PD, winning the primaries (Sandri and Seddone 2018).

The primaries of April 30th, 2017, saw three competitors for the phase open to votes from members and supporters: Matteo Renzi, Andrea Orlando (minister of justice in the Renzi government) and Michele Emiliano (president of the Apulia region). Also because of the fact that members from the leftist faction of the party, including Bersani, left the PD to form a new party (Articolo 1 – MDP), in this primary we witness a sharp decline in participation, which stopped at one million eight hundred thousand. In this case, in contrast to the 2013 primaries, the votes of members and the votes of supporters almost coincided (see Table 3.2), meaning that many members dissatisfied with Renzi left the party, but also that in his first term as party leader Renzi managed to strengthen his leadership not only outside, but also within the PD. According to Sandri and Seddone (2018), in this phase the leader is the "organisational glue" of a party that faces a loss of members and participation, and in which the old ruling class has been marginalised.

As usual, these are non-competitive primaries. Renzi's victory, with almost 70 per cent of the votes, confirmed that the "external body" had been absorbed, and that the party—after the split by the leftist faction was united around its leader and almost identified with him. In the face of such figures, we can say that the 2017 primaries served as a sort of re-legitimation of Renzi's leadership within the party, after the defeat of the constitutional referendum.


*Results of the 2013 and 2017 PD primaries: first and second phases Table 3.2.*

What were the characteristics of the Partito Democratico during Renzi's two terms as party secretary? Against the absence of statutory changes, what interviewees and scholars perceive is a marked change in the party towards a more direct and personalised style of leadership in the relationship with citizens and a more centralised style within the party and in the government. The real novelty lies, however, in the fact that, for the first time, in 2013 the secretary's support came more from outside than from within the party. This led to even greater attention on voters with respect to members. In this context, the spread of digital media was perceived as an opportunity to establish a direct relationship with supporters. However, the process of listening is only apparently bidirectional, since the choice of which stimuli to accept or reject always comes from the centre. In the same way, the listening to the members was apparently bidrectional during Bersani's secretariat.

To sum up, we can state that the preconditions for the success of a form of leadership such as Renzi's were already present in the statute of the Partito Democratico. Against the backdrop of a deep crisis of representation, helped by the spread of social media, the strongly personalised leadership of Renzi succeeded in creating a direct relationship with the electorate, more than his predecessors. However, in the absence of statutory changes, the party model remained formally unchanged: the only noticeable changes were in the interpretation of the leader's role, within the constraints and opportunities given by the statute.

The unmediated relationship between leaders and citizens has, in fact, been present in the Partito Democratico's party model since its genetic phase, as has the supremacy of voters over members. It seems that Renzi's leadership exploited and strengthened these dynamics. On the one hand, one can observe strong and more personalised leadership; on the other, the spread of social networks allowed the creation of a direct relationship of a mainly communicative nature with citizens. The spread of digital media seems to have amplified the disintermediation strategies already present in the party, which are facilitated by the general social and political trends that affected Western societies and were strengthened in Italy by the outbreak of the crisis of representation, especially from 2013 onwards.

It is necessary, however, to point out that, despite a major emphasis on the figure of the leader, the Partito Democratico in this phase cannot be considered a personal party but rather a personalised one (Bobba and Seddone 2016). Renzi, in fact, acted within well-defined rules and procedures, through the majority he obtained, according to the statute, in the internal governing bodies. By statute there are procedures for the removal of the secretary, and his office has a defined duration that precludes any form of permanent leadership.

#### *Zingaretti: The Party Strikes Back? 6.*

At the 2018 general elections, the PD obtained its worst result (18,8 per cent) and Renzi resigned as party secretary. Maurizio Martina was then appointed as *ad interim* secretary, in order for the party to elect a new leader, with open primaries. The primaries, the fifth in the short history of the party, were called for March 3rd, 2019. The rules were unchanged; six candidates were proposed, and after the members' voted three were admitted to the phase open to supporters. The three candidates were: Maurizio Martina, Roberto Giachetti and Nicola Zingaretti, president of the Lazio region. The primary election happened after the worst electoral result of the party, and in a phase in which the PD was in opposition. For these reasons, perhaps also because primaries were no longer a novelty in Italian politics, and due to the relatively low profile of the candidates, it did not attract much attention from the media or in public debate.

After a defeat such as the 2018 one, one would have expected a fierce internal battle, with different positions on the future of the party's organisation and on its programmatic profile. Instead, the competition was characterised by the absence of polarising issues and strong programmatic distinctions (Valbruzzi 2019), also because, as in most past primaries, the likely victorious candidate soon became apparent: Zingaretti. As regards participation, it decreased, both regarding members and voters (one million five thousand), testifying to the inability of the PD to broaden its decreasing constituency of supporters. As regards results, in the end Zingaretti was the winner with 66 per cent of the votes. Even in this case, as in 2013, despite the very different profiles of the contenders, we see that the support for Zingaretti came more from supporters than from members of the party (Table 3.3).


*Results of the 2017 and 2019 PD primaries: first and second phases Table 3.3*

What were Zingaretti's ideas about the party? Even if the patterns of the support for the new secretary were similar to those for Renzi in 2014, we can say that after the long experience, and the failure, of the personalised leadership of Renzi, the aim of the new secretary was to break with his style and to give more importance to the collective nature of the party. In his platform for the congress, called *Prima le persone* (People First), Zingaretti denounced the "leaderships who have too often manifested selfishness and myopia", and claimed he would "replace the pride of the ego with the strength of the us". It is exactly the style of Renzi, considered adverse to mediation, that he challenged.

In our field, the adversaries' myths were imposed […[for instance] that intermediate bodies are always useless and harmful, when instead they should be innovated and reformed as a fundamental element of a participatory and strong democracy […]. It is necessary The Partito Democratico leave behind the season of "disintermediation", establishing "consultation agreements" with economic, social, civic forces and with associations, foundations and think tanks with a political orientation close to that of the party (Platform *Prima le persone*, 2019).

With regard to membership, Zingaretti acknowledged the crisis of the party and affirmed that "sovereignty must move towards the base of the pyramid; we must make members, supporters, voters really protagonists". However, it is interesting to note that he made no difference between the three subjects. The new secretary stated that "the role of members and activists has been gradually reduced", and called for organisational reform, leaving behind "the useless and banal opposition between solid and light party". In his motion, Zingaretti called for the end of the identification between the role of national party secretary and that of candidate prime minister, to organise members' consultations on the political and programmatic strategies of the party, also through a new online platform, and to finally implement the statutory provision that provides for the organisation of an annual programmatic conference.

The statute was indeed reformed in November 2019. According to rhetoric that dates back to the foundation of the party, Zingaretti stressed that, with the new statute, the PD would be "a more open party, which opens up to the participation of the people, a more direct one, and one that will make those who are part of it the protagonists". In reality, there were only a few relevant changes to the organisational structure of the party. The structure remained that of an open party that, giving relevant decision-making power to party supporters through the direct election of the secretary, supposedly strengthens the leadership and weakens the party's intermediate structure.

The most relevant change was that the party secretary is no longer automatically the candidate prime minister. For the rest, the PD is still a party "consisting of voters and members" (article 1.4) that entrusts to voters the fundamental decisions concerning the party (article 1.5), including the election of the party secretary. As regards members' rights, the only difference to the first version of the statute is that only members can elect regional secretaries, a decision that before was entrusted to supporters (article 4.5). Therefore, the boundaries between the rights of the two subjects are still very blurred. Despite the ten-year debates on this issue, there has been no change in this peculiar aspect of the party.

In terms of the configuration of the leadership and the leader's rights, the most important difference is that the party leader is no longer appointed automatically as the party's candidate party for the office of prime minister. According to articles 5.2 and 5.3, *when he deems it appropriate for the interests of the country and the party*, the secretary proposes to the ND a different candidate for the office of president of the council of ministers. When the PD joins a coalition and primaries are used to identify the candidate prime minister, the NA establishes the procedures for presenting and selecting any other candidate, in addition to the secretary, who will be admitted to the competition. So, we can see that the party leader still holds a relevant decision-making power on this issue.

How is the secretary elected? The procedure is still articulated in three phases, albeit partially differently from the past. Article 12 is dedicated to the *Choice of political direction through congress and direct election of the secretary and the national assembly*. The congress (a word that was missing in the first statute) is divided into two phases. In the first phase, which ends with the holding of the national assembly, programmatic platforms are discussed. The second phase consists in the vote of the members on the candidates for secretary. Finally, open primaries between the two candidates who have obtained the most votes among members are organised. The candidacies for national secretary are presented together with a list of the candidates for the national assembly; the candidate who obtains the support of the most delegates in the assembly is elected. Furthermore, the secretary has the right to propose to the assembly the holding of an extraordinary congress on a single issue.

Therefore, the fundamental innovation of the Partito Democratico, that is that the leader draws his support from the outside, and no longer from a more or less restricted group of people, remains unchanged. Moreover, as before, candidates for public office are selected through primaries, as are candidates for the representatives elected to the assemblies, at every level. As far as the elaboration of the programme is concerned, the statute of the Partito Democratico continues to list a vast array of tools through which members and supporters can influence the policies of the party: thematic forums (article 30), the annual programmatic conference (article 33) and internal referendums (article 34). However, as we have seen, these tools haven't been used in the past. In addition, the new statute pays more attention to the issue of digital democracy, of which we will speak in the next chapter.

As regards the party's intermediate bodies, there are still three main internal organs that make up the party: the national assembly (NA), the national directorate (ND) and the national secretariat (NS). In the new statute, there is also the national assembly of mayors (article 7), an assembly of PD local administrators. The assembly appoints a delegation of five mayors who, together with their coordinator, are members of the ND. The coordinator is also a member of the NS by right. Like in the first version of the statute, the internal bodies (NA and the ND) are still partially composed in proportion to the vote for the secretary. A novelty is that the two organs are now composed according to territorial representation too. The new NA (article 6) is composed of 600 delegates elected proportionally, plus other representatives of the party in central office and one hundred representatives of the party in public office. Nevertheless, in matters regarding the removal of the leader and the selection of the prime minister, only delegates have the right to vote.

The new ND is made up of one hundred and twenty-four elected members. Half are elected by the NA using the proportional method, and half indicated by the regional levels among local administrators and representatives of the provincial federations and circles. As regards local sections (*circoli*) (article 17), the new statute mentions a new form of online aggregation, *Punto PD*, that can be created by three members belonging to the same place of residence, study or work. We will deal with this novelty in the next chapter. Finally, for what we will see in the chapter on the local electoral campaign in Turin, it is worth mentioning the new article 20 on the *Network of volunteers*. It states that the PD promotes a network of democratic volunteers, a network organised in local communities, who are to be active in the territories through specific actions and mobilisation campaigns.

To sum up, we can say that this organisational reform didn't change the main characteristics of the party. Zingaretti presented his candidacy in opposition to the Renzi experience, but we can see that there are many continuities: the role of the leader strengthened from the outside, and a rhetoric of citizen participation that, in reality, consists only in the participation in primaries. With regard to participation in primaries in particular, Figure 3.1 shows the evolution of the PD's membership figures, compared to the participation numbers in the party's primaries, both in the open phase and in that reserved to members. We see a constant decrease in the three figures, showing that the promise of greater members' and supporters' participation is not paralleled by a growth in participation over time.

*PD's membership and voters in primaries (closed and open phases) Figure 3.1.*

#### *Conclusions 7.*

This examination of the organisational history of the Partito Democratico shows that the promise of an unmediated connection between leader and followers is a fundamental characteristic of the party which has been present since its genetic phase. In fact, there are no significant changes in the organisation of the party over time, at least according to the analysis of official party documents. Renzi's leadership is often considered a rupture in the party's history, towards the creation of a "leaders' party" (Bordignon 2014), but we saw that in reality, despite being the only one who took full advantage of the opportunity given by the model of the open party, his innovative style was finally absorbed and then superseded. In a context of accelerating technological change and a deep crisis of representation, Renzi exploited the opportunities that were already given by the party model more than previous party leaders, increasing the personalisation of its leadership, creating a direct link with supporters and thus developing, more than his predecessors, disintermediation strategies.

The party model of the Partito Democratico is based on the opening of decision-making processes to voters and on the direct legitimisation, by the voters, of the party leadership, through its direct election. Through this sort of "personal mandate", the leader is supposed to be strengthened both on the outside, towards citizens, and potentially also within the party, since the intermediate bodies are established starting from the voters' vote. The promise of the opening of decision-making processes to voters—and not just to members—is a very important novelty: in the statute we see that the voters have several rights, such as to elect the secretary, to vote in the primaries and to define the party's programme. In practice, we see how the greatest decision-making power has been granted to supporters in the case of the selection of candidates through primaries; in contrast, the tools for the determination of policies (e.g. for the definition of the programme) have never been implemented.

We thus find the two dimensions of disintermediation in the Partito Democratico: on the one hand, the leader, strengthened by the direct and personal mandate given by citizenship, is expected to be stronger, both inside and outside the party. On the other, we see that supporters (and not members) seem to increase their power. The leader is therefore stronger because he or she draws his consent directly from the outside, bypassing party members and the middle-level elite. The primary elections for the election of the secretary are central in this mechanism: indeed, they have been defined by Lorenzo Guerini, then chief of national organisation, as a "democratic interpretation of disintermediation" (Guerini 2014).

Finally, we can question which dimensions of disintermediation prevail. In the Partito Democratico, voters have great decision-making power, greater than that of members. We can talk about a blurring of the organisational boundaries and a bypassing of the members by the voters, who thus are potentially empowered (disintermediation from below). However, I have also shown that the voters' power is limited to their participation in primaries, i.e. it is mainly *symbolic* even if relevant, configures individualised participation and lacks accountability mechanisms, to the detriment of the organised party on the ground. On the other hand, we have seen that the PD, although it cannot be defined as a personal party, due to the presence of well-defined rules and procedures, has been built specifically to strengthen the party leadership (disintermediation from above). It is this latter dimension that, in the end, prevails in the party's practices: through voters' participation in primaries, the leader is supposed to be stronger outside and within the party, even because the intermediate bodies are representative of the voters' vote.

We can also question, however, whether new forms of intermediation emerge. In the case of the PD, we can say that, more than the creation of new forms of intermediation, the old party structures have not disappeared: intermediate bodies still play a relevant role. The PD presents an innovative party model, in which open primaries are the backbone. Nevertheless, the party has governing bodies and codified procedures that limit the decisional autonomy of the leader. Even primaries, the most important organisational innovation of the party, appear to be a party affair, and are used consciously by party elites to pursue their goals: to give the impression that the process of the foundation of the party is not only an elite affair, to strengthen and legitimise a natural candidate, to create the image of a party that is open and that offers renewed participation practices, or to conduct an internal challenge against old party elites.

Primaries appear to be, thus, "an elitist instrument behind a plebiscitarian disguise" (Sandri, Seddone and Sozzi 2020). And that is perhaps the reason why, in the end, the Partito Democratico failed to deliver on its promise of durable leadership. Except for Renzi's experience, which was initially seen as an "external body", and ended up being absorbed and normalised, we see that what the PD lacks is precisely strong leadership. The resignation of Zingaretti and the election of Letta as party secretary at the beginning of 2021 testify to this.

## The Broken Promises of Members' Empowerment: Rhetoric and Use of the Internet by the Partito Democratico 4.

#### *Introduction 1.*

The party model of the Partito Democratico is based on the opening of decision-making processes to voters and on the direct legitimisation of the party leadership, through its election through open primaries. Therefore, the promise of an unmediated connection between leader and followers has always been a fundamental characteristic of the party. It is thus interesting to analyse whether or not the internet plays a role in the party's disintermediation strategies.

In this chapter, I will analyse the perception and use of the internet by the PD, and in particular the experience of the so-called online circles (*circoli online*) and the mobile application *Bob*. I will focus in particular on these two tools for two reasons. In the first place, these are proprietary tools of the party. Unlike, for example, in the case of a Facebook page, in this case the party has the power to decide the structure and architecture of the tool, and the features available. For this reason, these tools are a relevant source of information on the party's concept and employment of digital tools. In the second place, these tools can potentially fall within the category that I have defined as direct democracy: they could empower party members, giving them a say in party decisions and then altering the distribution of the internal decision-making power in their favour.

The Partito Democratico, although born in 2007, must not be considered a new party. On the contrary, it is a mainstream party that, unlike the M5S, has had to adapt to a new social, political and technological context. In other words, it is not a "netroot organisation", i.e. one that has been shaped from the very start by digital technology, but a "legacy" one (Karpf 2012). In order to analyse whether and how the internet plays a role in the party's disintermediation strategies, and in particular if it has an impact on party organisation, the direct democracy digital tools of the PD will be observed from two points of view: the description of their use and the assessment of their impact on party organisation. For the analysis of these tools, three dimensions in particular will be observed: the architecture of the tool and its affordances (Dahlberg 2011), its features, and the activities that users are encouraged to perform; members' rights and powers within the digital tool; and the consequences on party organisation and on the internal distribution of power.

In order to do so, in paragraph 2 I will examine the references to digital tools in the statute of the Partito Democratico and in other party documents, and their evolution over time. In paragraph 3, I will analyse the actual implementation of online circles through the voice of the promoters of two of them, now inactive: the one in Bologna, the first one to be opened and one of the most advanced at the time, and the one in Turin. In paragraph 4, I will analyse the mobile application *Bob*, released in 2017 and now abandoned, oberving the architecture of the application. Finally, in the concluding paragraph I will summarise the main findings and try to assess the consequences of the use of these tools on party organisation, in particular regarding members' empowerment.

#### *The Promises of Empowerment Through Digital Participation 2.*

According to its internal documents, the Partito Democratico seems to have grasped the opportunities given by digital technologies from the very beginning. The 2008 statute provides, in fact, two major organisational innovations, related to new technologies. The first concerns the party's local sections, called *circoli* (circles). In addition to territorial circles, linked to a specific territory, and environmental circles, linked to places of study or work, online circles have also been established. The statute also mentions an "IT system for participation" that is intended to encourage internal debate and allow members to "make proposals". However, the 2008 statute does not provide much information concerning these two tools. According to the statute, the circles constitute the basic organisational units through which the members take part in the party's life (article 14). Online circles "are set up on the internet", and it is possible to join them "regardless of place of residence, work or study". In this type of circle, the aggregation of members is therefore based on a criterion that is not of physical proximity.

If then, on the one hand, the party seems to recognise the importance of the internet, and to promote new forms of engagement, on the other, the impression is that it seeks to normalise and limit the potential innovation of online circles. In fact, in article 14.2 we read that "members of online circles can take part in the party's internal political life and the election of the online circle's governing bodies, but must in any case indicate the territorial or environmental circle where they exercise their other rights" (14.2). Basically, it is not possible to be a member only of an online circle: another circle is needed in which to exercise one's rights, especially during the congress. This is a matter related to the monitoring of party registrations, which are potentially less surveilled on the internet.

You become a member of the online circle, but you are not automatically a member of the party […]. [This is] to have a minimal filter on who becomes a member, because Italy is not made up only of "beautiful souls"; there are also the less beautiful souls and so I have to be careful about who enters the party […]. To avoid infiltrations, because otherwise I build an online circle, I put in 400 members and I go to the congress: it does not work! We have to keep a filter […]. If [the members of the online circle] want to elect the party leader, they must register with the party (Int. 17).

With regard to online circles, the 2008 statute refers to a subsequent regulation. This document specifies some organisational aspects: it is possible to become a member of the party online, but the registration becomes effective when the applicant collects the card in a territorial section. The request to establish an online circle must be signed by at least twenty people, and these people must be the expression of a territory of at least four provinces, in order to avoid overlapping with the territorial circles. Each circle can define its organisational and functional model, providing an assembly and a "coordinator"—instead of a secretary. These provisions are almost unchanged in the 2019 statute (article 17) and in the following regulation, except that now only ten people, without any reference to their territory of origin, can form an online circle, and three people from the same place of residence, study or work can form a *Punto PD*, a novel form of online aggregation established by the new statute.

Therefore, we note that a party which has been defined as "open", and which is based on the opening of decisional processes to its electors, is in fact very careful in the control of its organisational boundaries, when it comes to the online world. Members are in fact responsible for the first screening of the candidates for the party leadership, even if the final choice is entrusted to voters through primaries. On the contrary, promoters see the online circles not as a way to recruit more members in order to merely influence the leader's selection, but to create new forms of political participation.

Our aim really wasn't to create more registration cards. I've never been interested in saying "let's make 5,000 cards in order to take the party and change it". But we wanted to give people who are not part of this world and who do not feel involved in this world a key to access politics with. In contrast, the party has always adopted the view "we cannot control who enrols members online and so we are scalable" (Int. 12).

Article 1.10 of the 2008 statute is dedicated to the "IT system for participation". This is supposed to "foster the internal debate and to rapidly circulate all necessary information" and to allow "voters and members to be informed, to participate in the internal debate and to make proposals through the internet", as well as to make accessible "all information on its internal life, including financial statements, and on the meetings and resolutions of the party's governing bodies". The executives and the elected representatives of the party are required to make their activities public through the IT system for participation". The IT system for participation was never implemented, and it was reproposed in the 2019 statute under the name of "online deliberative platform".

In general, with respect to the 2008 one, the 2019 statute increases and deepens the references to digital participation. The eleven years between 2008 and 2019 represented a real revolution for the relationship between citizens, parties, and the online world. If the 2008 statute was ahead of its time—if not in practice, then at least for the declarations it made—it is now almost taken for granted that parties have to deal with the digital world. Moreover, in Italy, the fact that the Movimento 5 Stelle, a party which is based on the internet, gained huge success encouraged other parties to take a position and to adapt even more to the challenges of the online world. For instance, when presenting the new statute and its supposed digital innovations to the press, Zingaretti stated that the PD wanted to "beat the M5S on the web field". When one reads the statute and the leader's declaration, it seems that, on the one hand, the web is seen with suspicion by the PD, as a place of simplification, hatred and distortion of reality. On the other, it is perceived as a tool with which to fill the gap between citizens and politics, and as a tool of participation.

We must also learn to be a community on the web, an unavoidable place today, to open channels of encounters and communication with a large number of citizens of all ages, but especially the young. The net today is often the place of exasperated simplification, hatred and deception. It is often populated by those who use it to spread lies and negative values. But the web can also be the place for new and fruitful communication for informed citizens; it can be the tool for new forms of political and social participation, a way to fill the gap between citizens and politicians (Platform *Prima le persone*, 2019).

The Partito Democratico recognises the potential that digital networks offer for citizens' participation in public life, is aware of the risks posed and opportunities offered by the advent of the digital society and has organised itself to counter any form of falsification and distortion of reality, also through the activity of a national legal office (Article 30, 2019 statute).

The online deliberative platform is described in article 30 of the new statute. In its platform, Zingaretti acknowledges that the tool, already foreseen by the 2008 statute, has remained a dead letter. According to the statute, the new deliberative platform should be a place for "analysis, comparison, information, participation and decision-making, and for the discussion and dialogue phase that precedes and accompanies the decisions taken by the representative and executive bodies of the party". It should be "open to members and voters", developing its functions "through the institutional website and the official party application process". Its functions will include "coordinating PD members and circles, as well as interacting with voters", members and voters will be able to "advance ideas and contributions and report issues; verify the activity of the party and elected representatives; investigate issues of particular relevance thanks to constant access to studies and analyses; adhere to action and mobilisation campaigns; disseminate party activities". Future analyses will tell us whether it will ever see the light<sup>28</sup> .

#### *The Pioneering Experience of Online Circles 3.*

Except for the experience of the application *Bob*, no platform has been developed by the Partito Democratico in the time span considered in this work. When asked about online circles, the PD's party executives state that they "essentially carry out work of political elaboration; they elaborate documents, proposals... Many have a blog in which they exchange ideas; people take part in them... It is a promotional role particularly on the internet, and then on Facebook or on other social media" (Int 17). Accord-

<sup>28</sup> The *Agorà democratiche* project launched by Enrico Letta at the begninning of 2021 includes the creation of a digital platform, whose analysis falls beyond the scope of this work.

ing to this definition, it would be difficult to grasp the difference between an online circle—which, according to our typology, should fall into the category of direct democracy, since the circles are "the basic organisational units through which the members take part in the life of the party"—and a Facebook page or a Facebook group, which, in contrast, would fall into the category of communication or coordination.

Indeed, it is not entirely clear, from party documents, what the role of online circles is in the party model. Looking at practical experiences, we see that the first online circle was born in 2012. In 2012, online circles were able to have a pioneering function; since then, digital communication has changed profoundly, and nowadays every circle or group has the chance to have a "parallel life" online. Although "if a territorial circle has a website, it doesn't automatically become an online circle" (Int. 12), technological changes made the organisational innovation represented by online circles partially outdated<sup>29</sup> .

To understand the motivations for the creation and the functioning of online circles, the experiences of Bologna<sup>30</sup> and Turin—two online circles created in 2012 and 2014, respectively, and which are currently inactive can be analysed. For both, the reason for creating an online circle was twofold. In the first place, the aim was to attract and involve people for whom participating in the life of a party was difficult because of the constraints of work and family. Instead, the online circle was intended to be "a place where it is not the time of people's lives that adapts to politics but the opposite. It is politics that adapts its times to people's lives" (Int. 12). In the second place, there was an attempt to give new value to participation in the party, to make members count and to discuss contents in a context in which participation in the primaries is open to all voters and few additional rights are given to members, who thus cannot significantly influence the decision-making processes of the party.

The online circle is then, in the promoters' conception, something more than a communication tool: it is a tool for participation and deliberation,

<sup>29</sup> To March 2018, only three circles were active: Circolo online PD Campania – DemOnLine; Circolo online PD – Libertà è Partecipazione; and Circolo online PD – CittàMondo. According to a document written by those three circles (*A new way of doing politics: the online circles of the Partito Democratico*), online circles have assumed three configurations over time: territorial, for example corresponding to a region; circle of discussion, within which debates and documents of synthesis are created; and thematic, in which a specific political theme is debated.

<sup>30</sup> The online circle of Bologna was founded in 2012 and had an *ad hoc* derogation to act as a real territorial circle.

which can allow decisions to be made and then the life of the party to be influenced by giving members more participation opportunities. For instance, in the case of Bologna, it was possible to vote online, and "the governing bodies at the provincial level were informed of the results" (Int. 12). But in the absence of a clear regulation, and a lack of interest from the national party, a problem of this kind of participation tool was precisely that of finding a channel through which to pursue proposals, to connect with the party's decision-making process.

Perhaps the point of weakness was precisely what is called the "participatory contract": I take part, we discuss and debate, but it is necessary to actually understand what the outcome is and where it ends... […]. We had imagined that [the outcome] could be some initiatives to be presented in the city council, due to the fact that I'm on the city council, or even to identify local MPs at the national level who could somehow take charge of them (Int. 6).

To sum up, in its statute the PD provides for some innovative digital tools and, at the level of declarations, pays attention to the online world. However, in reality, the national level didn't invest much in online circles, leaving the initiative to individual members and not giving any technological support to promoters. The reason seems to be that online circles are perceived as a threat to the organisational boundaries, as a way to "scale" the party. This clashes with the image of an open party that opens its decision-making processes not only to members but also to voters. Moreover, regarding the transfer of power to members, we have seen that it is difficult for online circles' promoters to develop an actual "participatory contract", that is, to link participation with clear outputs and decisions. For these reasons, the impact of online circles on party organisation appears to be very limited.

#### *Participation or Just Propaganda? The Mobile Application Bob 4.*

Also because of the evolution in digital technologies and the spread of social network sites, online circles tended to close over time. In parallel, especially starting with Renzi's secretariat, the party has paid greater attention to online communication: one of the features of Renzi's leadership style was precisely the direct communication with supporters, also through social networks. The online application *Bob*, released in 2017, seemed to represent a new interest in direct democracy tools by the PD. *Bob* was introduced for the first time by Renzi in March 2017, during the presentation of his candidature for the spring 2017 primaries. The application was defined by the party leader as an "IT platform": the explicit reference was the M5S's platform *Rousseau*. It seemed that, with *Bob*, the Partito Democratico wanted to take up the digital challenge with the M5S.

Our counter-offence on the internet is about to begin. Against the lies of those who led people to believe that politics makes fake news go viral, thus gaining in advertising. The *Bob* project is ready (Renzi's newsletter, May 2nd, 2017).

The application was released in May 2017. It was an application exclusively for smartphones and did not have a linked website. In the description of the application, we can read that *Bob* is "a unique, inclusive, collaborative digital ecosystem and offers those who want it the chance to be a protagonist". The application had six sections: *Volontari* (Volunteers); *Democratica+*, *Conosci* (Know); *Partecipa* (Participate); *Sostieni* (Support); *Video*.

The *Volontari* section contained materials for the electoral campaign for the general elections of March 4th, 2018. Through this section it was possible to register as a volunteer and download materials for the press and social media<sup>31</sup>. The *Democratica+* section presented some information material, especially videos. It consisted of two sub-sections: *Ore nove* (Nine o' Clock) and *Terrazza PD* (PD Terrace). Also, the following section, *Conosci,* was dedicated to information and made up of various sub-sections: *Democratica*, in which there was a link to the online newspaper of the party, *News in evidenza* (News in the Foreground), in which there was brief news about the party, *Team*, in which we could have found the list of the members of the national secretariat, with biographies, and *Territorio* (Territory)*.* In this last section, a list of the physical and online addresses of the regional offices of the party and those of the metropolitan federations was presented.

Since the application was advertised as a platform for participation and decision-making, the most interesting section is the fourth: *Partecipa*. *Partecipa* consisted of five sub-sections: *Le idee di oggi* (Today's Ideas); *Sondaggi* (Surveys); *Condividi* (Sharing); *Magliette gialle* (Yellow T-Shirts); *Eventi*

<sup>31</sup> A volunteer campaign for the 2018 general elections was advertised on the party's website and social network accounts. It seemed that the PD could replicate an experience such as that of NST at the national level, but a campaign based on volunteers was never organised.

(Events). It was possible to access this section with a Facebook or Google account. To enter this section, it was also necessary to respond positively to the question: "Do you accept the *Charter of Values* and intend to vote PD?". This was not, therefore, a section reserved for the members of the party. Like other offline decision-making processes, this section was open to all those who declared they they were voters of the PD.

Let us analyse in detail the architecture and affordances of this section. The sub-section *Le idee di oggi*, contained the news of some results achieved by the party in government with a very brief description. The user could respond to this item with a like (thumb up) or dislike (thumb down). It thus represented a simple expression of agreement or disagreement with the policies carried out by the party in government. In *Sondaggi,* there was a survey on some laws currently under discussion in the Chamber of Deputies. After a brief description of the law in question, five questions were presented to the user, who had three possibility to answer: agree, inclined to agree, and disagree. Even in this case, it is not clear who the recipient of this feedback is, and how this will be interpreted and eventually used. In *Condividi*, there was some news to be shared on Facebook; in *Magliette gialle* some news on the party's volunteering activities and on the themes of civic engagement were listed, but not information on how to take part in this mobilisation; in *Eventi,* a list of some national events, without the opportunity for the user to propose new ones, was presented.

The fifth section, *Sostieni*, consisted in a link to the donations page of the party website. In the last one, *Video*, some video-interviews with the PD's representatives were available. In addition to the six sections mentioned, there was another button through which it was possible to reach a section called *Proponi* (Propose). In this section, it was possible to send a proposal (an event; an idea; a survey) to the application's editorial staff.

What are members' rights and powers? In the first place, we note that *Bob*, like most of the party's decision-making processes, was open to both members and voters. From the analysis of the application's affordances, we can see that the only actions that the user could undertake within *Bob* were: to respond with a like (thumb up) or dislike (thumb down) to some results achieved by the party in government; to agree or disagree on certain predetermined questions regarding a law discussed in the Chamber of Deputies; read news and share it on social media; connect to the party website; and contact the application editorial staff. Therefore, we can state that *Bob* had mainly a communicative top-down function. The application, particularly in the section allegedly dedicated to participation, did not allow users either to decide, debate or create links among themselves and between themselves and the party.

Despite the rhetoric that surrounded it, *Bob* doesn't fall into the direct democracy category. The sections *Le idee di oggi* and *Sondaggi* could be considered part of the sub-category consultation (that is, when the party commits to collecting and considering members' ideas, in a non-binding manner), but there isn't a "participatory contract": the user doesn't know whether and how his or her contribution will be interpreted and taken into consideration. Moreover, we have seen that the contribution that the user can provide is definitely poor: it is limited to an agreement or disagreement with some very simple questions, and it is completely unlinked to the party's decision-making process. The only way in which the user can interact, apart from expressing agreement or disagreement is to send a proposal to the application's editorial staff. To sum up, we can say that there isn't any transfer of power in *Bob*, and consequently no impact on the distribution of the party's internal decision-making power.

#### *Conclusions 5.*

The aim of this chapter was to analyse the role of the internet in the PD's disintermediation strategies, and in particular whether and how the use of direct democracy digital tools has an impact on its organisation. To this end, I have analysed the perception and use of the internet of the Partito Democratico, its evolution and impact on the organisation, in particular regarding two digital tools: online circles and *Bob*. In terms of the evolution of the party's use of the internet, we can say that in a first phase there was an interest in direct democracy tools, which is testified to by the innovations included in the 2008 party statute. The party gave people the opportunity to create online circles, and some local representatives of the party took advantage of this. Nevertheless, apart from this, the PD did not seem interested in the development of such tools, mainly because of its fear of being infiltrated and the loss of control over the organisation.

In a second phase, corresponding to Renzi's secretariat, despite rhetoric on the importance of online participation, we witness a shift to mainly communicative digital tools. Indeed, despite it being presented as a "participatory platform", the second digital tool analysed (*Bob*) is neither a direct democracy nor an organisational tool: it is just a top-down communication tool. To define it as a participatory platform is only a rhetorical attempt to show the party's interest in digital innovation, and to create the perception that members can count and participate online. In reality, the participation that took place via *Bob* was completely detached from the party's actual decision-making processes. In the case of online circles, in contrast, there was an attempt to give more power to members through online participation, but this attempt was limited to the local level.

In sum, the tools examined do not allow a transfer of power from the centre to the base (or vice versa), and so we can say that they do not exert an impact on the party's organisation and on the internal distribution of the decision-making power: the role of the internet in the PD's disintermediation strategies seems then to be only of a rhetorical nature. This hypothesis is confirmed by the fact that *Bob* was soon abandoned, and that in 2019 similar rhetoric and similar tools were reproposed in spite of the fact that no participatory platform had been created by the PD.

It is worth noting that in early 2021 Enrico Letta, the new party secretary, started a process called *Agorà democratiche*. This initiative, according to its presentation on the party's website, is "aimed at expanding and innovating the spaces and forms of democratic participation and at developing, in the most shared and broad way possible, an ambitious programme for the centre-left". The *Agorà* are divided into two phases. On the one hand, there will be some events, both in person and online, in which proposals are made and shared by participants. On the other hand, a new digital platform will be the place in which the proposals will be discussed and through which priority actions will be identified. The platform will be open to all citizens and will require the signing of the party's *Charter of Values* and the payment of a minimum amount of 1 euro. It remains to be seen whether this initiative will be placed in the wake of the preceding ones. So far, the party has resisted the most radical innovations potentially created by digital tools, using them only to create the image of a party whose decision-making processes are open to members and supporters, and in which members can participate online.

## Staging Directness: Ethnography of a Grassroots Campaign in Turin 5.

#### *Introduction 1.*

What happens at the local level, where the party's contact with citizens is closer? Electoral campaigns are the field in which parties are said to have externalised their functions the most. The voluntary work of members and activists is said to have been replaced with that of professional campaigners, and the relationship between party members and citizens with an unmediated leader–electorate connection, carried out through television or social network sites. To answer this question and verify this assumption, this chapter will observe Noi Siamo Torino (We Are Turin), an innovative experience of a grassroots campaign carried out during the 2016 Turin mayoral election.

Noi Siamo Torino (NST) was an electoral campaign inspired by the US model—specifically, by Barack Obama's electoral campaigns—based on volunteers' field mobilisation and micro-targeting that took place in Turin in 2016 and that was aimed at the re-election of the mayor Piero Fassino, a member of the Partito Democratico (PD), supported by a centre-left coalition. NST was a managerial and professional organisation, external and separate from the PD, whose aim was to recruit volunteers with the task of persuading citizens to vote for the candidate Fassino, not in the name of party identification, but using their experience of lay citizens. The goal of the promoters of the campaign was the creation, through the activity of volunteers, of a direct (that is, not mediated by the party structure) relationship between citizens and candidate that was supposed to encourage, as a result, an increased turnout.

Nonetheless, in the second round Fassino lost the election to the M5S's candidate, Chiara Appendino. However, the case of NST won't be analysed here from the point of view of electoral effectiveness (Cepernich 2017, 105–106); on the contrary, through data collected during the direct observation of the campaign, the NST experience and its meaning will be investigated with the aim of detecting the changes in practices and repertoires of campaigning and party membership, particularly regarding a mainstream party, such as the PD, and their link with disintermediation strategies.

In recent years, we have witnessed the promotion of proximity (Le Bart and Lefebvre 2005) between politicians and citizens: the direct contact between them would constitute a possible solution to the crisis of trust that affects the institutions of representative democracy. The creation, through the field activity of volunteers, of what Nielsen (2012) defines "personalized political communication"—personalised in the sense of no longer mediated by communication technologies, but by real people—can be considered an attempt to create an unmediated relationship between the candidate and the citizens, which is different to the one traditionally carried out by party members.

Based on these premises, this chapter will be organised as follows. In the second section, we will discuss the transformations of electoral campaigns in the three phases outlined in the research literature: pre-modern, modern and post-modern or digital. Although apparently paradoxical, it is precisely in the last phase, characterised by the massive use of digital media, that on-the-ground and grassroots campaigns, in which the fundamental factor is the human one, regain their importance. In a context marked by the alleged transformation and crisis of party membership, canvassing and, in general, campaigning carried out with volunteers on the ground appears both as the reinvention and modernisation of ancient practices, and as a substitute for the classical functions of party membership. Finally, I will briefly discuss some of the most significant foreign experiences of campaigns of this type (especially in the US, but also in France) that are implicit or explicit points of reference for the case we have analysed.

The third section will be dedicated to the interpretation and "translation" of this type of campaign in Italy and in particular in Turin. Against the backdrop of the context of the 2016 Turin mayoral elections, we will analyse the setting, the main actors and the objectives of the NST project. In the fourth section, we will deal with the development of the campaign and with its reception by the actors directly involved: the volunteers and the Partito Democratico, analysed both through direct observation and semi-structured interviews.

#### *Canvassing: Electoral Campaigns "Back to the Future" 2.*

Studies on the evolution of political communication in general (Blumler and Kavanagh 1999), and of electoral campaigns in particular (Norris 2000), highlighted the succession of different phases. If, on the one hand, these distinctions are more useful in tracing analytical borders than recording real breaks, it is true that some social, political and technological transformations have profoundly changed the way of conducting and communicating politics over the last few decades.

In this respect, it is possible to identify three main phases (Cepernich 2017; Mazzoleni 2012). The first, which goes from the post-war period to the fifties, is marked by the supremacy and the hegemony of parties. In a context in which political alternatives seem "frozen" (Lipset and Rokkan 1967) and in which citizens' identification with political parties is high, parties are the main actors of political communication and electoral campaigns, also because of the control they can exercise over the media. In this phase, electoral campaigns are managed independently by the party, through direct forms of interaction between candidates and voters (Manin 1995) and propaganda broadcast through partisan media.

The second phase, which can be said to have begun in the sixties, coincides with two trends: the diffusion of television and the loosening of party loyalty. These are two crucial changes that radically transformed political communication and electoral campaigns. The result was the spread of mediatisation processes (Mazzoleni and Schulz 1998): an increasing influence of the so-called media logic (Altheide and Snow 1979) in all fields, including politics, whose institutions, were increasingly dependent and shaped by the media, especially by television, rather than vice versa, as in the previous phase.

A greater independence and commercialisation of the media imply a different way of dealing with politics which is more oriented towards forms of spectacularisation, popularisation and fusion with entertainment (van Zoonen 2005; Mazzoleni and Sfardini 2009). Electoral campaigns are increasingly played out on television, and therefore their organisation is increasingly entrusted to professionals that don't belong to the party. Since a specific expertise is needed to manage this kind of campaigns, the party relies less and less on members for mobilisation and propaganda activities, also because their number has declined (Dalton and Wattemberg 2002; van Biezen, Mair and Poguntke 2012). Therefore, with the passage from the first to the second phase, the personal and direct contact between citizens and candidates was loosened (Lefebvre 2016), and the distance between citizens and politicians grew (Manin 1995).

An opposite tendency seems to have emerged in the current phase. The massive diffusion of digital media, especially those relating to the so-called web 2.0, represent a deep paradigm shift compared to the past, and is defined by Castells (2009) as "mass self-communication". On the one hand, with the internet, citizens have a greater responsibility in the selection of information; on the other, they can also become producers and distributors of information, bypassing journalistic mediation. Information has therefore become more individualised and personalised, and potentially more horizontal and bidirectional, even if asymmetry and hierarchy have not disappeared on the internet. At the same time, political identities and parties' organisational structures have been further transformed into post-bureaucratic forms. Light and multi-speed forms of participation and affiliation have emerged (Scarrow 2014).

Disintermediation strategies, as we have seen, can be considered parties' strategic responses to these changes. In this context, over the past fifteen years, starting from the United States, electoral campaigns that enhanced direct contact with the voters, through telephone and personal contact especially, have appeared. It is what Nielsen (2012) defines as "personalized political communication": premeditated (in the sense of organised) practices that use people as media for political communication. The United States has a vast tradition of canvassing that has never disappeared. But, starting from the nineties, and especially around 2000, we witness a growth in this practice, which we can attribute to the diffusion of the research by two political scientists from Yale, Alan S. Gerber and Donald P. Green (2000; 2004), and in general to the acknowledgement, in communication research, of the limited effects of mass communication (Nielsen 2012). Gerber and Green, through experimental research, demonstrated the electoral effectiveness<sup>32</sup> of canvassing, especially regarding voters that wouldn't have otherwise voted (hence the term "mobilisation campaign", i.e. that does not aim to convince voters already inclined to vote for the other candidate, but to mobilise as much as possible the voters belonging to their own field), because of the social pressure that the campaign would create.

There are two main aspects, related to the use of new technologies, that differentiate this practice from similar ones from the past. On the one hand, new technologies allow more rational, scientific and managerial organisation of activists and volunteers involved in an electoral campaign, for example through applications such as MyBarackObama (Vaccari 2009). The second is the use of data to guide the electoral campaign: the construction of databases, their analysis and segmentation in order to profile

<sup>32</sup> Gerber and Green's experimental research shows that door-to-door activity is more effective than posters, e-mails and phone calls. In fact, canvassing allows parties to "earn" one vote for every 14 conversations, against one for every 38 telephone calls and one for 100,000 flyers.

voters and contact them (the so-called micro-targeting), both indirectly for example with text messages or phone calls—and directly, through canvassing. According to Nielsen (2012, 6) "in political campaigns new technologies have not replaced older forms of communication as much as they have revived them".

Although there are some interesting previous experiences (for example, the Republican election campaign for the US presidential elections in 2004), it is Obama's 2008 campaign which made this type of mobilisation globally famous and became the reference point for all future experiences of this kind, also because of the candidate's victory. As noted by Lefebvre (2016), in fact, Obama's victory helped to legitimise the use of these new practices and to "export" them, even despite the different characteristics of the European and American social and political contexts, and consequently the different structures of their electoral campaigns.

In this respect, an interesting importation of this kind of experience in the European context is the French one, which took place during the 2012 presidential elections<sup>33</sup>. Three young French scholars affiliated to US universities, after having participated in the Obama campaign in 2008, proposed a similar campaign to the French Socialist Party. The three proponents, called *Les Bostoniens*, proposed using the Socialist Party's members in order to organise a massive national canvassing campaign (Liégey, Muller and Pons 2013; Belkacem and Talpin 2014; Lefebvre 2016), reinventing and innovating in the direction of greater managerialisation, not without resistance, the political communication practices of a Socialist Party affected by the decline in its membership.

In conclusion, the fundamental characteristic of this type of campaign is the union of the possibilities offered by new technologies with the search for personal and direct contact with voters, or rather the attempt to organise in a scientific and managerialised way, also through new technologies, direct contact with voters, which was characteristic of the first phase of political communication. It is for this reason that we can define canvassing as a sort of "back to the future" for electoral campaigns, although it would be more accurate to say that this kind of campaign has affinities, but also differences, with the first phase campaigns. An

<sup>33</sup> In the United Kingdom, canvassing is a consolidated practice, also because of the incentives of the uninominal electoral system. Even in this country we have witnessed, especially from 2015 onwards, the profound Americanisation of electoral campaigns, certified by the use of US political consultants that previously worked for Obama by both the Conservatives and Labour.

affinity undoubtedly consists in the valorisation of the direct and personal contact with the electors, whose importance had already been underlined by the studies on the two-step flow of communication and on opinion leaders (Lazarsfeld, Berelson and Gaudet 1948). The main differences lie not only in the managerialisation of this contact, but also in the fact that contemporary electoral campaigns take place in a context in which the role of parties, at least in Europe, has profoundly changed. Campaigns based on the mobilisation of volunteers and on canvassing therefore appear on the one hand as the reinvention and modernisation of ancient practices, and as a substitute for the classical function of party membership on the other.

#### *From the USA to Turin: The Case of NST in the 2016 Electoral Campaign 3.*

The mobilisation campaign that took place in Turin in spring 2016 was the first of this kind organised in a large Italian city. The organisation of NST was autonomous and detached with respect to the PD, but specifically aimed at the re-election of the mayor Fassino. Piero Fassino, the incumbent mayor, is a leading figure of the PD. 67 years old in 2016, he has been active in politics since 1968, holding important positions as an elected representative, party official and member of the government at the local and national level in the PCI, in the DS and finally in the PD, of which he is one of the founders. In 2011, he was elected mayor of Turin in the first round. The main opponent of Fassino and of the centre-left coalition that supported him was Chiara Appendino, the mayoral candidate for the M5S.

The two opponents seemed to represent change and continuity, respectively. While Fassino relied on his strong experience as a public administrator and politician, Appendino, 32 years old at the time, pledged radical discontinuity with respect to the previous administrations: after 23 years of uninterrupted government of the city by a centre-left coalition, Chiara Appendino's victory in the second round testified to a demand for change. This demand was expressed despite the fact that opinion polls showed citizens' overall satisfaction with the previous administration34. But satisfaction did not translate into consensus on election day. The electoral geography of the vote, with the peripheral neighbourhoods voting for

<sup>34</sup> In 2015, a poll showed that Fassino enjoyed the approval of about 60 per cent of the citizens. He was the fourth most highly rated mayor in Italy (Governance Poll 2015, Ipr Marketing).

Appendino and the central ones for Fassino, has led to the hypothesis that the peripheries felt excluded from the narration of the change of Turin from a city characterised by the presence of factories to a city of culture and knowledge.

In fact, in her electoral campaign, Appendino insisted on the existence of "two cities": the city centre and the peripheries, the latter depicted as excluded from the benefits of renewal (Cepernich, Cittadino and Pellegrino 2018). On the contrary, according to Cepernich and Vignati (2016), the electoral campaign for Fassino was constructed on a framework of continuity, based on the results achieved in the previous five years, on the claim of the mayor's competence, and on the validity of the development model of the city that had been promoted in the previous two decades. Fassino stigmatised Appendino as inexperienced and incompetent, and opposed to large infrastructural investments.

It is in this context that Fassino's electoral campaign in general, and the NST project in particular, was organised. NST can thus be seen as a strategy carried out by a mainstream party, particularly by a candidate belonging to the so-called establishment and challenged by an outsider, to deliver an unmediated relationship with citizens and to get closer to them. Nevertheless, I must clarify that Noi Siamo Torino was only a part of Fassino's electoral campaign, certainly a minority part concerning the economic investment by the candidate. Fassino relied largely on the classic electoral campaign tools such as billboards, and on the network of the PD's circles and party militants that organised stands and other campaign events during the electoral campaign. NST was therefore an organisation autonomous but integrated into Fassino's electoral campaign.

How was the idea of a US electoral campaign "imported" into the Turin context? The creators and coordinators of the project were Cristopher Cepernich, a professor of sociology at the University of Turin and an expert in political communication, and Flavio Arzarello, an expert in political communication and a member of the Innovation Department of the PD. Cepernich and Arzarello proposed the project to the candidate Piero Fassino who, in a context in which the party was suffering from a membership crisis, accepted hoping to obtain additional mobilisation. But the two coordinators were driven by partially different motivations. If, for Cepernich, NST was a scientific project, proposed to a candidate whose party had a territorial structure capable of welcoming it, for Arzarello, who was an active member of the PD, the motivation was also political and dictated by his proximity to the party.

In general, the two creators of the project showed a great fascination for American election campaigns. Cepernich was, in fact, present as an observer in Lecco, a city in which a similar campaign for the re-election of the mayor Virginio Brivio was managed by Mike Moffo, a media strategist in Obama's 2012 electoral campaign. Cepernich and Arzarello met on the occasion of a conference by Moffo organised by Arzarello in 2014. Even in this case, as in the French one, there seems to have been a "contagion"<sup>35</sup> from the United States: the reference to the US experience and to Obama is, in fact, explicit<sup>36</sup>. It is also important to consider that university played an important role in this case, as it also did in the US and in France: Cepernich, in fact, played the dual role of scholar and political consultant.

The members of the NST staff (about 15–20 people) were also university students. They were students of public and political communication at the University of Turin, interested in experiencing an electoral campaign. They were very young, all under the age of 30. Their relationship with the electoral campaign was of a formative and professional nature. This is considered valuable experience in the field of political consultancy, to be "added to one's CV", regardless of the candidate's political colour ("if they had called me in Milan to campaign for Berlusconi, I would have gone"—one of the staffers told me).

On paper, the staff's task, under the direction of Cepernich and Arzarello, was to organise and coordinate the mobilisation of volunteers. Indeed, in the minds of the creators, NST was to stand on two legs. As we can read in the slides of the project, presented in PD circles, on the one hand, the goal was the mobilisation of volunteers, with the task of "reactivating the nodes of relationship in the territory", "reactivating participation and increasing turnout" and "promoting the desired voting behaviour" through direct and interpersonal communication with voters. On the other hand, volunteers would have to collect data on voters "for the communication and the strategy of the campaign".

Before I analyse these two goals in detail, the use of the term "volunteers" deserves further attention. In Italy, unlike in other countries such as

<sup>35</sup> As regards the use of the word "contagion", Robert (2007, 16) invites us not to think about the importation of managerial practices in politics in terms of simple diffusion, in which the new practices do not face resistance or undergo transformations. In his opinion, it is more appropriate to investigate the strategic appropriation of these practices by actors and the reasons for their interest in them.

<sup>36</sup> See Giambartolomei, A. *Fassino vuole fare Obama: studenti arruolati a Torino*, in "Il Fatto Quotidiano", June 1st 2016.

the United States, the term volunteer refers to the semantic field of social volunteering. By defining the volunteers as such, it seems that the goal is to untie the volunteers for Fassino from the political dimension, to which they would belong by nature, and to link them to civic activism.

The relationship between volunteering and politics in Italy has been marked by different phases (Biorcio, Caruso and Vitale 2016). Since the fifties, the research by Almond and Verba (1963) showed that in Italy the figures for membership in associations were lower than in other European countries, a fact that helped to explain, according to these scholars, Italy's "particularistic" civic culture. In reality, the peculiarity of the Italian case lay in associations' strong ties with mass parties, in particular with the PCI and the DC. With the decrease in party identification, and with the birth of the new social movements between the end of the sixties and the eighties, and especially after the collapse of the so-called First Republic, we witness the growing autonomy of associations from parties. Given the "crisis of parties", and in a social context characterised by the weakening of the welfare state, it seems in this phase that volunteer associations ended up replacing, albeit in a different way, some of the tasks previously performed by parties and institutions, such as advocacy and service provision (Biorcio, Caruso and Vitale 2016).

From Toqueville onwards, participation in associations has always been considered a "school of democracy", capable of socialising citizens to democratic practices, and increasing participation and citizens' trust in institutions. But although even political activism can be considered a voluntary activity, the two types of participation differ substantially. While social participation is situated within civil society, and its aim is to protect rights and common goods and to support disadvantaged individuals outside the actions of parties (Moro 2013), party membership and activism belong to the political field, and we can say that their aim is to influence the "authoritative allocation of values in a society" (Easton 1965) through the election of candidates to public offices. Moreover, while active citizenship tends to be self-organised, participation in parties takes place within hierarchical structures organised from above (Michels 1911).

We can say then that NST tried to import some elements of social volunteering, or more generally of active citizenship, into the political field. It is therefore possible to frame this campaign within the concept of depoliticisation, in the sense of a transition of electoral campaign practices, caused by anti-politics and mistrust in parties, from the political to the personal-social dimension (Flinders and Woods 2014; Woods and Flinders 2014). As far as NST volunteers are concerned, coherently with this frame, in the presentation of the project they were depicted as separated and disconnected from the Partito Democratico. The goal of volunteers should be to interact with and listen to citizens, demonstrating—through their presence on the ground—the closeness of the candidate to the citizens. Nevertheless, although one of the aims of the campaign was to mobilise lay citizens, not necessarily active in politics and the party, the promoters also seeked to structure and organise their mobilisation with the collaboration of the local sections of the Partito Democratico (*circoli*).

Volunteers could in fact register through a dedicated website, and were coordinated and managed by NST, but the local sections of the Partito Democratico were also supposed to play a part in the campaign. Thanks to their territorial rooting, the local sections were meant to represent a sort of "logistic base" for NST volunteers and contribute to the campaign with activists who knew the neighbourhood and could therefore help volunteers. During the presentations of the project in the local sections, organisers told party members that NST would have done "what parties have always done", but with a strategy and in a systematic way. However, in reality, the organisers' idea was that the party and volunteer campaigns were two electoral campaigns with different objectives and targets. At the meetings that took place in the early phase of the campaign, in various local sections of the party, the organisers presented a clear "division of labour" between the two types of actors involved in the campaign: NST volunteers, unlike PD members, should have established a relationship with people based on their personal and subjective experience of the city, and not on the basis of an electoral programme or party identity.

But the electoral campaign was based not only on the mobilisation of volunteers and on their direct interaction with citizens. The other parallel and specular side of the campaign was the systematic collection by volunteers of voters' data, which were used in order to direct the electoral campaign. In fact, conversing with the citizens shouldn't be intended as an end in itself. The aim was to fill in a form that, in addition to a voter's personal data, contained two pieces of information that will serve to profile the voter and to send him or her targeted communication (micro-targeting): the priorities for Turin and the voting orientation. The collection of data by volunteers could take place in three different ways: first, through the volunteers' private relationships (e.g. relatives, acquaintances); secondly, through the presence in the neighbourhoods' public places; and finally, through canvassing. In the last two cases, the volunteers were equipped with a red bag and a red bib emblazoned with the writing "Piero Fassino candidate mayor of Turin" and were instructed to conduct a "completely natural" conversation based on the following script:

Good morning/good evening, my name is Chiara, I am a volunteer with Noi Siamo Torino for the electoral campaign of Piero Fassino. I would like to talk to you briefly about our city. Are you available?

If so:

In June there will be the elections for the mayor of Turin. Have you already decided who to vote for? (If you notice enthusiasm, ask if the person is willing to dedicate a part of his or her time to help as a volunteer). In your opinion, what is the most urgent problem that the next mayor of Turin will have to solve? Do you agree to give us your contact details to establish regular contacts and receive detailed information about our activities and our programme?

If not (quickly understand why):

If he or she is adamant about voting for another candidate, quickly close the conversation and move on. If he or she fears revealing his or her opinions, be reassuring and underline the VOLUNTARY aspect of your work.

Finally, a further aim of the campaign, in addition to mobilising volunteers and collecting data for micro-targeting, was to "integrate storytelling into field mobilisation". Through social networks and media coverage, the narration of the mobilisation of NST was intended to become a part of Fassino's electoral campaign.

#### *The Volunteers, the Party and the Unfolding of the Campaign 4.*

"Hello, we are volunteers for Piero Fassino; can we ask you some questions about the city?". This was the typical beginning of interaction between NST volunteers and a citizen of Turin. We have seen how, behind this simple question, there was a wider strategy and project, which can be linked, more generally, with the transformations of electoral campaigns and party membership. However, an organisation's strategy, which appears to be highly rationalised, is often different from the actual practices implemented by militants and volunteers (Belkacem and Talpin 2014). This is why it is worth analysing the actual realisation of the campaign.

In order to structure the mobilisation in a capillary way throughout the city, eight small groups were created in correspondence to the eight administrative districts of Turin. Each group was managed by two staff members (called "captains"), who had the task of coordinating the operations of the volunteers, preparing the field operations according to their own and the volunteers' availability. The operations mainly consisted of two types of "outings" (*uscite*): those in public places (e.g. markets) and door-to-door canvassing. The PD's local sections were supposed to serve as a logistic base. A higher level of coordination was represented by the two campaign managers (Cepernich and Arzarello) and by another person in charge of the organisation. She had the task of contacting people who register as volunteers through the website and send them to the captains. All the outings were inserted in a Google calendar, and the staff was coordinated through a WhatsApp chat group. Once a week, the staff met at the headquarters of the electoral committee to discuss the progress of the campaign. It is at this central level that it was possible to monitor the progress of the micro-targeting data collection. The data were processed using software (Target 51) that allowed NST to profile voters and to send them targeted communication.

As regards volunteers, it is not possible to say exactly how many of them actually took part in the campaign, as the purpose of NST was precisely to mobilise a network of both formal and informal supporters. In a leaflet distributed during a meeting between Fassino and the volunteers, also aimed at recruiting new volunteers, we can read that volunteers can have different forms of commitment and involvement, ranging from taking part in the ground operations ("Come with us on the ground") to lighter forms of participation, such as speaking about NST to friends or following NST on social networks. As mentioned, the idea of involving personal networks is greatly emphasised ("introduce us to friends"). Therefore, in a broader sense we can also consider anyone who solicited a vote for Fassino with phone calls a volunteer. Here, however, we will only consider the most active volunteers, those who took part in the activities on the ground.

Thanks to the collaboration of the staff, we asked each captain to indicate the number of volunteers who had taken part in at least one outing on the ground. The number of volunteers turned out to be 27. We then asked the captains to indicate to us the volunteers who had taken part at least three times, and the number turned out to be 12. Of these, we interviewed 10 volunteers (a summary table of the interviewees can be found in the Appendix)<sup>37</sup> .

To start, it is necessary to highlight why their number is very low, in general and in relation to the population of Turin, which is around 900,000. This, against the broad visibility of the project, advertised through all the channels of Fassino's election campaign, indicates the low appeal of such an initiative in the Turinese context and, as an unexpected consequence, led to staff members having to commit themselves to the field activities much more than they initially thought, becoming volunteers instead of volunteer coordinators. This created a partially distorted representation of the campaign. The fact that the young staff members were the most present, on the ground and on social media, led to the fact that NST was identified as "the youth campaign for Fassino". In fact, although nothing was done to make people believe the contrary, it is not openly stated that the staff members were interested in the campaign for almost exclusively educational and professional reasons.

From the point of view of their age and profession, we note that eight out of the ten volunteers belong to two categories: students (4) and retirees (4). The other two, middle-aged men, were already active in associations or trade unions. The volunteers were predominantly men (8/10). In terms of the political identity of the volunteers, seven out of the ten were currently members of the Partito Democratico (the eldest were also members of the PCI or the parties that followed) and regularly participated in the activities of a PD local section. One volunteer was a member in the past, but disagreeing with some of the party's choices, he has not renewed his membership. However, he still voted at the primaries; another is not a member but considered himself close to the party. We also asked the volunteers whether they were active in the field of social volunteering too: six volunteers were, while three said that for them their "volunteering" activity was political ("I am already a volunteer in the party", Int. 2). In this sense, the position of volunteer n. 10, a long-time party member, is interesting as he suggests that party membership is understood by party members themselves as a "voluntary" activity (as unpaid and aimed at a goal such as emancipation) and that, for them, the political activity is

<sup>37</sup> This part of the research was carried out in collaboration with a member of the staff, Jacopo Pelliciari, who was writing his master's dissertation on NST. Together with Pelliciari, I conducted the ten semi-structured interviews, which were also used in his work.

perceived as more important than the social one ("We were more volunteers").

Therefore, despite the attempt to import some elements of volunteering and more generally of active citizenship into the political field, we note that the people interested in this project had, in the vast majority of cases, a very strong political and party background. We can thus compare volunteers' representations of the campaign with their other experiences of political activism. We find both similarities and differences to their participation practices as members of the PD. In the first place, volunteers noticed and appreciated the presence of young people in the project. This can be read in relation to PD membership that is increasingly composed of old people. Moreover, one of the PD's problems is dropping membership and weak participation in the local sections, which leads as a consequence to weak rooting in their territory, which is very different from the past. What emerges from the interviews, especially with regard to older militants, is in fact a contrast between the experience of NST, characterised by the presence of volunteers on the ground, and the experience of the party in recent years and its detachment from its territory. On the other hand, NST represents a sort of continuity with "what we did". For instance, PCI members used to deliver the newspaper *L'Unità* (an organ of the PCI) door-to-door and had a strong rooting in the neighbourhoods.

If this [NST] continues, it will become like what we did: every Sunday we sold *L'Unità* door-to-door. In the neighbourhood they knew you because they always saw you. There was constant activity. That is what is missing now, in fact, because many local sections are always closed (Int. 10).

As regards PD members that decided not to participate in NST, with whom I had the opportunity to speak during the presentation of the project in the local sections, we can find two main reactions. In general, it seems that what was said by the organisers was accepted by members. According to them, NST volunteers "do what we have always done", but with two differences: they do it in a more structured way and are young. The positive reactions to the project show how volunteers, in anti-political times, can benefit from the fact of "not being marked" (not being labelled as belonging to the political field and not being identifiable as PD). The negative reactions can be summarised in the suspicion that the volunteers were "paid"<sup>38</sup>: staff members were people unrelated to local sections, and it was difficult for members to understand why they are committed to Fassino and not to the PD. Generally, the presence of NST was perceived as the consequence of the absence of party members on the ground.

This [NST] was a good example of the state in which the party's local sections are. I find it extremely positive, but if the *circoli* worked, the militants of the local sections would be the volunteers for Piero Fassino! There would be no need to adopt another structure dedicated to the electoral campaign; there would be militants for each *circolo*. It is positive that they found them, but in the end, you see, young people were found for the activity, limited to the election of the mayor. So, it means that involving people in something concrete and specific still works, because those are truly volunteers; they are not paid, they are volunteers […]. [This situation] is the result of an absence. The absence of those who were once called party militants. Now we call them volunteers, but they are militants (Int. 10, PD).

Finally, regarding party officials, they seemed to appreciate NST, especially from a narrative point of view, that is, for the storytelling in the campaign. Against a party that lacks young people and rooting in the territories, NST is perceived as a project to be exploited in order to improve the image of the candidate and the party<sup>39</sup>. Storytelling, as previously mentioned, has a strategic importance for NST. NST has a Facebook page where images of volunteers during the outings were posted daily. Staff members were, in fact, always eager to take photographs during these excursions. Therefore, seen in opposition to NST, which created the perception of a group present in the territories, the party is seen as closed, both physically and figuratively. Indeed, through the outings, NST promoted the perception of closeness and the proximity of politics and the mayoral candidate to the citizens.

With regard to the outings, they can be classified into two categories: those in public places and door-to-door canvassing. On these occasions, volunteers were supposed to talk with and listen to citizens, proposing

<sup>38</sup> This is a fundamental difference to the United States, where the presence of paid volunteers is completely normal.

<sup>39</sup> Note that in the flyer for the run-off, it is possible to see a picture of Fassino surrounded by staff members and that on the occasion of a rally with Matteo Renzi at the Teatro Alfieri in the first rows of the theatre there was a sign stating, "reserved for young people and volunteers".

their personal motivations for voting for Fassino and indirectly showing the candidate's closeness to the voters, and finally to collect data for micro-targeting<sup>40</sup>. In short, volunteers became "means of communication" (Nielsen 2012), intermediaries between Fassino and his electorate.

In the face of a highly rationalised strategy, it is interesting to analyse what happened, concretely, during the outings, and how the volunteers interpreted the role they had been assigned. A starting point is the general climate of negativity and of anti-politics among citizens. Faced with this climate of opinion, the strategy of volunteers was to shift the focus from the figure of Fassino (perceived, as we have seen, as part of the establishment, of the "system") to issues related to the city. "Do not immediately say that you are here for Fassino", is the advice given. In short, what appears to have taken place is an attempt to depoliticise the local campaign. "It's like selling a product, but here you sell a political idea," says a volunteer. The consequence of this approach, however—also given the absence of booths and gazebos, typical of a party electoral campaign, and the presence of bibs without party symbols—is that the volunteers were mistaken for ActionAid or Greenpeace volunteers, who ask for donations on the street and therefore tend to be ignored by passers-by.

Secondly, there is a discrepancy between the declared objectives of the project and their concrete realisation. For some staff members, the goal seemed to be essentially to collect data rather than to interact with citizens. If, according to the coordinators, the module is the tool with which to create a relationship and not the goal of the interaction, for some staff members the goal seemed to be simply to fill out the form. This clearly emerges in an exchange between a volunteer and a staff member:

The volunteer (a man in his mid-60s) asks the staff member what the goal is, whether to "collect data" or to "sow". The staff member answers without hesitation that the goal is to collect data. The volunteer tries to understand what these data are for: he is a little doubtful but interested. The staff member is interested in the data, and it is clear his goal is to collect as much data as possible, and in fact he runs swiftly from one interaction to another. Instead, the volunteer starts talking to people calmly. People reject him less because he is an older man, and he seems nice and quiet. He does what volunteers should do on paper: he asks people to talk about the city and listens to them. In fact, I notice that that on some occasions this works: some

<sup>40</sup> The organisers declared that they had collected 11,507 forms.

people who initially do not want to talk eventually give their data. Although he does not talk about his personal experiences, he seems interested in what people are saying and listens to them. However, it does not always work: some people don't want to talk. "It's like selling a product, but here you sell a political idea," he tells me. The volunteer seems especially pleased to see a young man who engages in politics. As soon as the outing ends, the staff member tries to get rid of the red bag, because he is going to the grocery store to do his shopping. In the end, he turns it inside out, so that it is impossible to see the writing (Fieldwork note, 5/4/2016).

The element of rejection and mistrust was also present during the canvassing. On these occasions the volunteers were supposed to go from house to house to remind voters about the elections and to solicit a vote for Fassino (the so-called GOTV—Get Out The Vote). Especially when it was done in the absence of people belonging to the neighbourhood, or party members or candidates, there was a strong reluctance to opening doors to strangers. In the absence of the party's previous presence in the neighbourhoods, the experiment seemed destined to fail.

I'm going to canvass with two staff members. We have to walk the entire street, contacting all the people who vote in an electoral section. We have, in fact, a list of all voters residing in that street who vote in a certain section (the walk-list). One of the staff members tells me that he contacted the members of the Partito Democratico who live in that road to go in their building with them, but no one is available. We are in a street in a residential neighbourhood. When I arrive at the meeting point, I find one of the two staff members sitting on the bench, the Fassino bag upside down, as the other one did. We start. We ring the bell and say, "Hi, we are volunteers for Piero Fassino; can we ask you some questions about the city?" Most people do not respond (also because it's 4 p.m. on a working day). Whoever answers, with rare exceptions, says no. A very strong suspicion withholds them from opening the door. When we find a building door open, we enter and knock on all the doors. We manage to hand out three flyers, including one to a couple from the Partito Democratico. The lady is part of the local section and wants to know who we are. She is part of the local section and does not know us (but why wasn't she contacted?). The two staff members are tired and have no enthusiasm. In fact, it takes enthusiasm and conviction (or a salary) to spend hours ringing doorbells. After about an hour, when we are halfway down the list, the two cannot take it anymore. One of the two jokingly proposes ringing the bells all together (Fieldwork note, 13/5/2016).

The fact that the staff member didn't find anyone available for canvassing shows scarce cooperation with the party. In addition, we see how the lack of involvement in the campaign by the staff members led to an absence of enthusiasm and to consider the time dedicated to the campaign as working time (see the trick of turning the bag inside out once the outing is over). Another problem was the difficulty of collecting data from people who tend to be suspicious probably because of their fear of fraud or illicit use of their data. Faced with these difficulties, some volunteers said that they would stop collecting data, thus violating the rules provided by the staff ("I personally collected two email accounts and a phone number, then I stopped asking", Int. 6). Faced with these difficulties and the low number of volunteers mobilised, as the weeks passed by, and especially during the run-off, NST tended to become only a form of leafleting:

The staff member told me that today there would be two candidates and two volunteers; instead there are only the two candidates. They are two very young women who were elected in the district in the first round. They are happy and relaxed, and they do not even hand out the flyers to all the people they meet. The activity has been reduced to handing out flyers. The staff member says that it is like this now, that people do not stop anymore, and they are tired. In fact, we walk and many people, especially older people, do not really want to talk to us. And they [the candidates] do not do anything to make them change their minds (Fieldwork note, 15/6/2016).

#### *Conclusions 5.*

Electoral campaigns are one of the areas in which parties are said to have "externalised" their functions the most. The voluntary work of members and activists is said to have been replaced by the work of professional campaigners, and the relationship between party members and citizens with an unmediated leader–electorate connection, carried out through television or social network sites. The aim of this chapter was to analyse in-depth a local electoral campaign, in order to observe both if these assumptions hold true and the responses and the adaptations of a mainstream party at the local level.

Electoral campaigns are indeed a relevant moment which analyse political parties and can tell us a lot about their inner workings. Over last few years, electoral campaigns based on direct and personal contact between parties and citizens have emerged. These campaigns, characterised by managerialisation and the use of new technologies, appear both as the reinvention and modernisation of ancient practices, and as a substitute for the classical function of party membership. In this chapter, I observed an electoral campaign based on the mobilisation of volunteers in the city of Turin that was aimed at the re-election of the PD's mayor, Piero Fassino. On paper, the project seems to represent the overcoming of the party's role on the ground using an unmediated relationship between citizens arranged by an organisation separate from the party, but the in-depth analysis of the functioning of the campaign shows that the picture is more nuanced. My results show that we witness both the persistence of traditional forms of intermediation and the emergence of new ones.

As regards the first aspect, the vast majority of volunteers that joined NST, apart from young staffers interested in the campaign for professional reasons, were in reality members of the PD. This result, on the one hand, can be interpreted as a persistence of the role of party members, even in this new kind of electoral campaign. However, through volunteers/members' representations of NST, it is possible to highlight the main weaknesses and shortcomings of the party: the PD lacks the presence of young people, rooting in the territories, and the perception of closeness and proximity with citizens. An electoral campaign such as NST is constructed precisely in order to counterbalance these weaknesses and to replace traditionally intended party membership.

With regard to the second aspect, and in accordance with Nielsen (2012), my results show that these forms of communication between citizens do not constitute a genuinely unmediated relationship, given that the whole interaction is orchestrated and carried out by "wider campaign assemblages", working on behalf of a candidate who is not physically present at the door. In addition, the interactions carried out by NST were not of a spontaneous kind; on the contrary, they were strongly managerial. The NST campaign, even if it tried to represent itself as spontaneous, was carried out by a highly professionalised structure different from that of the party. The electoral campaign was carried out with managerial logic that is *a priori* foreign to the political field (Robert 2007), and so we can say that it constitutes a new managerial form of intermediation.

Finally, the low number of volunteers, the lack of rooting in the territories, also due to difficult collaboration with the PD, together with Fassino's defeat, could lead us to think that NST had no effect. However, the data presented by Cepernich (2017, 106) indicate that, in the electoral sections where campaign mobilisation was higher, there was an increase in voter turnout of 8 percentage points, compared to the average of the city of Turin. We don't know whether and in which way this fact is related to NST's campaign. Nevertheless, in conclusion it seems appropriate to focus on the narrative dimension of this type of experience and on its impacts. Against the backdrop of a strong anti-political climate, and of local politics perceived as closed and detached from citizens, through canvassing its mediatised storytelling, the candidate attempted to show himself as symbolically close to people and staged an unmediated relationship between the candidate and citizens, whose effects can be observed mainly from a rhetorical point of view (Belkacem and Talpin 2014).

## The Organisation of the Movimento 5 Stelle: Disintermediation or New Forms of Intermediation? 6.

#### *Introduction 1.*

Italy is one of the countries in Western Europe characterised by the highest percentage of votes expressed for new political parties (Chiaramonte and Emanuele 2016). The most successful new party<sup>41</sup> that recently appeared in the Italian political system is the Movimento 5 Stelle (M5S). The M5S was founded in October 2009 and at the February 2013 elections obtained 25.5 per cent of the votes in the Chamber of Deputies and 23.8 per cent of the votes in the Senate of the Republic. This was an extraordinary electoral success: the highest figure ever recorded in Western Europe for a new party at its first national electoral test<sup>42</sup>. Five years later, at the 2018 general elections, the M5S achieved another extraordinary result. On that occasion, unlike other new parties during their second national electoral test, it managed to increase its votes by 7.2 per cent, obtaining 32.8 per cent in the Chamber of Deputies (Biancalana and Colloca 2018) and entering government at the national level for the first time.

The M5S presents a unique internal organisation, combining horizontality and verticality: at the time of its foundation, its leader defined himself as a simple "megaphone", its elected representatives presented themselves as mere "spokespersons", allowed to serve as representatives for a maximum of two terms, and its rhetoric was characterised by horizontality, transparency and an emphasis on online direct democracy. But, at the same time, the M5S has always had an opaque internal decision-making structure, in which the leaders played a very relevant role (Passarelli, Tronconi and Tuorto 2013 and 2017; Caruso 2015; Ceccarini and Bordignon 2016; Biancalana and Piccio 2017).

<sup>41</sup> Deschouwer (2008) points out that there are three different dimensions of "newness" to consider: the age of the party, its ideology and the type of party organisation, and the M5S can be considered new in all the three aspects. As is known, the M5S rejects the label of "political party", preferring that of "movement" or "non-party". In this work, however, following the well-known definition by Sartori (1976), we will define it as such.

<sup>42</sup> To have a comparison, Forza Italia obtained 21 per cent in the 1994 elections, the first of the so-called Second Republic.

I defined disintermediation strategies as rhetoric or practices developed by parties in order to stage or deliver an unmediated relationship between leader and followers, which happens through the weakening of the party's intermediate organisation. The aim of this chapter is to examine the evolution of the M5S's organisation in order to analyse whether and how a new party employs disintermediation strategies in its organisation. Moreover, studies on political actors born in opposition to mainstream parties, which are characterised by an organisational model based on grassroots democracy, have demonstrated that these parties' participation in institutional politics led them to centralise their organisation over time (Frankland, Lucardie and Rhioux 2001; Rhioux 2001; Poguntke 2002). It is therefore interesting to examine which dimension of disintermediation prevails among those outlined and if there are changes over time in the M5S's organisation.

My analysis of the M5S's organisational history will be carried out using Pedersen's "lifespan model" (1982), as adapted by Rihoux (2001). The lifespan model is a tool that has often been used to analyse the evolution of new parties, as it allows an examination of the different phases of party development, linking them with the organisational evolution of these actors. According to Pedersen, during their evolution, political parties pass through a number of phases, or thresholds. Not all thresholds are necessary for a political party to exist, and not all thresholds have the same weight, but there is a given sequence in the parties' evolution, and crossing some thresholds is considered to be particularly critical for the parties' development.

Adapting Pedersen's model, Rihoux identifies five thresholds: declaration (the decision to take part in national elections); authorisation (the acquisition of the formal requirements to be allowed to compete in national elections); representation (entering the parliament); institutionalisation (the first national electoral confirmation); and government participation. According to Pedersen, each of these steps will bring about the emergence of new challenges and will determine different organisational changes, in order for parties to cope with the new functions that they will have to enact. Consequently, we can expect that, with the passing of these different thresholds, the internal structure of the M5S would become more complex and develop new forms of intermediation.

Five dimensions will be analysed for each phase: the selection and role of the leader; the role of party members; the selection of candidates; the determination of policies; and the role and characteristics of the party's intermediate bodies. The analysis will be carried out by comparing two types of sources: on the one hand, the party's official documents; on the other hand, interviews with privileged witnesses and party members. As highlighted by Katz and Mair (1992), party statutes tell only one side of the story, which is why it is then useful to compare them with the accounts of party members and party personnel regarding the actual distribution of power within the party. This is particularly true for a party such as the M5S, whose structure is very different compared to traditional parties. Moreover, we have seen that it is crucial to understand how actors perceive the stimuli that lead to change and the reasons for some strategic choices, as well as the discourses that surround party organisation: interviews are the most appropriate tool with which to do this.

#### *The Genetic Phase (2005–2009) 2.*

The M5S was officially born on October 4th, 2009. However, at that time, it already had a history whose prominent aspects are worth mentioning. As highlighted by Panebianco (1988), the genetic phase of organisations marks them and influences their evolution in the following years. For the M5S, this phase corresponds to the years between 2005 and 2009. In 2005, the meeting between Beppe Grillo, a famous Italian showman and comedian, and Gianroberto Casaleggio, an entrepreneur and owner of the company for online strategies Casaleggio Associati, led to the opening of a website (*Beppe Grillo's Blog*), through which Grillo began to carry out an activity of counter-information on issues such as transparency in politics and environmental protection (Vignati 2015a; Ceri and Veltri 2017).

Neither Grillo nor Casaleggio are professional politicians. Grillo is a famous comedian who has always paid close attention to social and political issues in his shows. For instance, in the eighties his shows were even removed from public television because of his satire on government parties. Casaleggio, who started his career as a software designer, was an entrepreneur and a web expert: it was Casaleggio who proposed Grillo should open a blog. Through his website, Grillo also encouraged its readers to mobilise (as, for instance, in the case of the "Clean Parliament" initiative<sup>43</sup>) and to physically meet using the online platform *Meet-up*. This

<sup>43</sup> In 2007, groups related to Grillo's blog collected 350,000 signatures in order to present a popular initiative law to prevent the election of convicted MPs, to set a limit of two mandates for MPs and to reintroduce preference votes in the electoral system.

led to the foundation of the "Friends of Beppe Grillo" groups in many Italian cities, heterogeneous groups of people that coordinated themselves through *Meet-up*, and in many cases overlapped with pre-existing associations and protest committees (for the case of the No TAV movement in the Piedmont region, see Biancalana 2020). In these groups, the interest in the issues raised by Grillo linked to existing territorial mobilisations.

[the first groups] came from a certain type of background that was "movement-like". But it is natural, you first agglomerate those who already have a political and social conscience of a certain type; they are the first ones to commit (Int. 8).

Some events testify to Grillo's attempt to influence politics from the outside in the first period. The failure to achieve results in this way might have led to the decision to create a new party. For instance, in 2006, Grillo met the prime minister Prodi, in order to present him the results of some discussions that took place on his website, but without a result. Moreover, in 2009 Grillo tried to present his candidacy to the Partito Democratico's open primaries, but it wasn't accepted. Faced with the parties' and the institutions' lack of responsiveness to the requests made by Grillo and his audience, the project to move from an activity of "external pressure", as it had been until then, to direct participation in electoral competitions began to take shape, because "being only movement-like was not enough" (Int. 8).

A first step in this direction took place between 2007 and 2009, when Grillo launched the idea of the "Five Stars<sup>44</sup> Civic Lists" "certified" by him. The aim of the civic lists was to bring ordinary citizens into institutions: the candidates that satisfied some requirements, such as having a clean criminal record and not having been previously elected with other parties, could autonomously create a civic list and take part in local elections with a unitary symbol, and benefit from the visibility of Grillo's blog. The permission to use the symbol was given by Grillo, which is the "certification".

[The role of Grillo and Casaleggio was] to give us a chance, that is to say: if you want, we can enter the municipal councils, with civic

<sup>44</sup> The five stars represent the five most important issues carried out first by Grillo and then by the M5S. They are: public water, defence of the environment, free internet connectivity, sustainable development and sustainable public transportation. These issues can be considered to belong to the so-called post-materialistic left. Despite the changes in the M5S's agenda, and the broadening of its scope with respect to its origins, the name of the party didn't change until 2021.

lists and with a well recognisable symbol at a national level, and then we can put into practice what we think, i.e. we make proposals with tangible acts, since the others don't, of what we would like to do. And that was their role in the organisation: the input (Int. 2).

Overall, the genetic phase of the M5S was already characterised by some characteristics that were to mark the organisational evolution of this player. From the very beginning, we see a combination of horizontality and verticality. On the one hand, through his blog, Grillo connected people belonging to civic associations and, in general, the social movements' world, fostering civic participation and the autonomous development of organisations in the territory from below. For this reason, the M5S has been likened to social movements (Mosca 2015)<sup>45</sup>, as its aim was to skip traditional forms of representation and to bring citizens directly into institutions, in a sort of DIY politics. On the other, it is Grillo who guided this development, and decided from above on the most important organisational issues, such as the creation of civic lists, the definition of the requirements for their certification, as well as the drafting of the programmatic points of the civic lists (the so-called Florence Charter, *Carta di Firenze*), which were presented during a meeting in Florence in March 2009. In summary, we can see that Grillo's aim was to bring citizens directly into the control room of politics, without the need for a party organisation, which is replaced by the internet, but under his supervision.

#### *Declaration (2009) 3.*

On October 4th, 2009, at the *Smeraldo* theatre in Milan, Beppe Grillo, alone on the stage, announced the foundation of the Movimento 5 Stelle. "It was a political communication event", "a show", recounted an activist present that day. On that occasion, there was the presentation of the political programme and the statute of the party (significantly called Non-statute). The foundation of the party happened indeed in a decidedly top-down way: the 120 programmatic points that—according to official announcements—came from the proposals made by activists in the comments on

<sup>45</sup> For these reasons, some scholars have defined the M5S as a movement-party (Ceccarini and Bordignon 2016). For a critique of the use of this category in the case of the M5S, see Biancalana and Piccio 2017.

the posts written by Grillo on his website, seem assembled in an arbitrary and non-transparent way.

At the time, Grillo's blog was very lively in the comments section; there was a debate there, there were many *Meet-ups*, and there was turmoil and discussion. And Grillo read, around, on the net; obviously he didn't participate, but he read at that time; we discovered that he had read things that we had written, and then, probably, in that mass of thoughts they decided "let's do it". Of course, Casaleggio, who was the organisational and even theoretical mind, chose the form that he felt was best, and in fact already at the beginning there were controversies because all this stuff that came from above, some people said "we had to be part of it too" (Int. 7).

The same applies to the rules contained in the Non-statute, a document that was never discussed or voted on within the party. In clear contradiction to the rhetoric of horizontality and leaderlessness carried out by Grillo, who used to state that "there is no leader in the M5S, the leader is the Movement […]. I'm not a leader, I control the situation" (Corriere della Sera, October 14th, 2014), according to the statute he is the owner of the "right to use the symbol" (article 3) and his website is considered the "headquarters" of the "non-party" (article 1). The M5S is described not as a party but as a "consultation tool" (article 4), a "centre for candidate collection", a "tool" (article 7) for the selection of candidates for elections. The candidates will be "authorised each time and in writing to use the name and symbol" (article 7). The identity and destiny of the organisation are therefore closely linked to those of its leader, whose role and selection are incontestable. Furthermore, procedures for his recall are not specified. However, this contradiction between rhetoric on horizontality and the presence of a strong leader doesn't seem to bother the interviewees.

[The M5S] is something that they [Grillo and Casaleggio] invented; then one could join or not. When they founded the M5S, they stipulated these four or five rules; all the people who joined and who decided to become a candidate did so in full knowledge that those were the rules (Int. 4).

Grillo is the owner of this brand. For anyone who wants to use it, it is enough to make himself or herself available. There are rules. If they need to be respected, they must be respected (Int. 10).

A second relevant organisational aspect is the absence of statutory provisions regarding the so-called "party in the central office". The M5S is characterised by a strong anti-party stance, which is also mirrored in its internal organisation. Indeed, in contrast with traditional party organisations, no internal bureaucracy or intermediate structures (such as local and regional branches) are formally established. According to official documents, the M5S is made up of three subjects: the leader (the owner of the party symbol, who has the power to place it on the lists of candidates that want to take part in elections with the M5S), the elected representatives, who are authorised each time to use the symbol, for a maximum of two terms, and the members, who are connected through the internet.

The internet, according to article 4 of the Non-statute, has an acknowledged "central role in the phase of registration with the M5S, in consultation, deliberation, decisions and election". "The role of government and direction normally attributed to a few" is acknowledged as being attributed "to the totality of the internet users" (article 4). Indeed, it is only possible to become a member of the M5S online, and this membership "does not require more formalities than registering on a normal website". Registration is free of charge, and the only requirement is not to belong to political parties or associations which have an object or purpose that contrasts with those of the M5S (article 5). The rhetoric of the M5S on the internet has been defined as "cyber-optimist" (Mosca, Vaccari and Valeriani 2015): in the absence of intermediate structures, direct democracy can be achieved thanks to the connection between citizens and power that takes place online.

But if the "role of government" is acknowledged as being attributed to the "totality of internet users", only members, through an online participation platform (which has been called *Rousseau* since 2015), can take part in internal decision-making processes. Indeed, it is only possible to take part in the national decision-making processes of the party online. According to the party's rhetoric, in the absence of a central structure, similar to that of traditional parties, citizens can directly influence politics, either by entering into institutions or giving directions to elected representatives, through the party's online participation platform. It is crucial here to highlight that the management of the platform, as well as of all the sites connected to the M5S, was at that time in the hands of the company Casaleggio Associati (owned by the co-founder of the M5S, Gianroberto Casaleggio), which in the absence of any other intermediary structure acted as a sort of surrogate of the party in central office: it managed online registrations, the participation platform, as well as the certification process of the lists of candidates that wanted to take part in elections under the M5S's symbol.

The "staff" are Casaleggio's employees; it's Casaleggio who pays them; they help him with the management of the website, Twitter, in managing the certified lists, when there are problems and then with the expulsions; that's the staff, but how many of them are there, fifteen? I don't know. But they support us, and they don't take a euro from us. They are paid by Casaleggio [...]. We do not pay a euro to the Milan organisation (Int. 2).

[The role of Casaleggio Associati is a role] of IT support, but it also goes beyond this: Gianroberto was the founder of the M5S together with Beppe Grillo, so this is clear; it must be clear (Int. 9).

At a local level, the situation appeared decidedly less controlled from above. In fact, if at national level it is possible to join the Movement only through the websites managed by Casaleggio Associati, and participation in decision-making processes is only possible online, at a local level it is possible to become an "activist" by taking part in face-to-face activities in the territories<sup>46</sup>. Formal registration with the M5S is only required, according to the rules of the party, to become a candidate. Indeed, in this phase, at the local level, activists organised themselves autonomously; there were no official sections or branches<sup>47</sup>, or specific rules that apply to this level, except for those relating to the process of candidate selection. Since the local groups had no place in the organisational structure, no decision-making power was granted to them. Only members, intended as individuals, could participate in the national decision-making processes that happen online. The only official representatives of the M5S in the territories were candidates and elected representatives, who were also the only ones authorised to use the symbol.

<sup>46</sup> We will call the ones that are officially registered with the M5S through the website and have the right to take part in online decision-making processes "members". In contrast, we will call those who, whether or not officially registered, take part in face-to-face activities in the territories "activists".

<sup>47</sup> The *Meet-up* platform was, at least initially, the main tool through which the M5S's activists organised themselves in the territories. However, these local groups do not represent an official local branch of the M5S. This is certified by a post dated July 19th, 2015, *A letter to the Meet-ups*, in which it is specified that the *Meet-ups* don't have the right to officially represent the Movement (for instance, they can't use the party logo), as the only ones entitled to use the symbol are the candidates and the elected representatives, who have obtained the "certification", i.e. specifically the authorisation to use the party symbol.

If you come to our meetings, I won't ask you if you have already registered. So in reality the meetings are always open to everyone and we do not have a "register" of the people who reside in the city and are officially members of the M5S. There are some people who maybe participated for months and then they tell us; maybe it happened when we drew up the lists for the elections, and to become a candidate it is necessary to be registered: "But I'm not registered yet" (Int. 4).

To sum up, the Non-statute ratified what was already implicit in the genetic phase of the M5S. According to the 2009 statute, Grillo is the leader and owner of the M5S, with significant power in deciding on national policies, and members don't have the chance to remove him. Moreover, there is no structure comparable to that of traditional parties. All the key national decision-making process of the party (starting, as we have seen, from party enrolment) take place online and are managed by a private company owned by the co-founder of the M5S, Gianroberto Casaleggio. On the other hand, according to the statute, party members have significant rights. In particular, there is the idea that citizens can directly influence politics through digital tools.

At the local level, despite the fact that local groups do not represent official branches of the M5S, activists can organise themselves autonomously, hence deciding on local candidates and policies. We can find, then, the two dimensions of what I called disintermediation strategies: great powers are given to the party leader (disintermediation from above) and to party members (disintermediation from below). In the case of the M5S, the intermediary role of the party is not simply weakened: according to party documents, in this political actor there is no party structure at all. In the next paragraphs, we will see whether and how this changed with the passing of some important thresholds.

#### *Authorisation (2010–2012) 4.*

In this second phase, which ended in 2012 with the M5S's participation in the 2013 general elections, we find the first signs of the emergence of an (albeit informal) internal organisational structure, also driven by the growing number of local elections in which the M5S took part, often with considerable successes. For instance, the M5S elected representatives in the regional councils of the Piedmont and Emilia-Romagna regions in 2010, and in Sicily in autumn 2012; in the administrative elections of spring 2012, Federico Pizzarotti became mayor of Parma. We see that, after passing the declaration threshold, the procedures for the selection of candidates for elective offices (local, regional and then national) became defined more clearly. Nevertheless, in this phase, the procedures were often specified near elections, through a post on Grillo's blog; only later were they included in the party's rules.

Again, in this phase, at the local level, in order to take part in elections under the symbol of the M5S, certification (authorisation to use the symbol) is required. To obtain it, it is necessary to send the necessary documentation to the Casaleggio Associati offices<sup>48</sup>, which are responsible for the process. At this level, in most cases activists formed electoral lists and defined the programmes in a completely autonomous way, with internal voting taking place in offline assemblies. However, when more than one list of candidates wanted to run for elections under the M5S symbol in the same city or in large cities, where it was difficult to organise offline consultations and where the centre wanted to control local decision-making processes, candidates at the municipal level were selected through online consultations. At this stage, and at least until 2014, even because of the small number of activists, the selection of candidates for the regional elections also took place mainly through consultations in offline assemblies.

At the national level, however, the internet replaced the assembly. For instance, the procedure for the presentation and selection of candidates for the 2013 general elections, which took place from December 3rd to 6 th,, 2012 (the so-called *Parlamentarie*), was completely online. Among the requirements to become a candidate, there was: to have previously taken part in an electoral consultation under the symbol of the M5S or the civic lists and not to have held a position as an elected representative on the date of October 29th, 2012; in order to take part in the selection of candidates it was instead necessary to be enrolled in the M5S by September 30th, 2012, just two months before the consultation.

As the selectorate was not extended to the M5S's "voters", as in the Partito Democratico, this cannot be considered an open primary; however, members of the M5S had a considerable decision-making power in the choice of their own candidates for parliament. Nevertheless, if the inclusiveness of the selectorate was guaranteed, the "rules of the game" were

<sup>48</sup> To become an M5S candidate, it is necessary: to not be a member of other parties, to not have completed more than one electoral mandate and to not have been convicted. The fact that it was Casaleggio's company that managed the process was not codified officially, but it was clear to activists: when they talk about the certification process, they say they have to send the documents "to Milan".

neither shared nor previously discussed with the activists, but simply published in a post by Grillo. Moreover, we must say that online consultations are a problematic tool, given the private ownership of the IT system that manages the operations and the absence, with only a few exceptions, of an external agency that controls the regularity of the vote. This task was entrusted to Casaleggio Associati's staff, who may engage in manipulation (see Mosca, Valeriani and Vaccari 2015; Deseriis 2017a).

On the other hand, the procedure for the definition of the programme at the national level was decidedly not inclusive: the twenty programmatic points for the 2013 elections were simply communicated by the leaders, without an actual programmatic discussion. "They come out of the hat", as an MP, who was later expelled, remembers (Int. 6). We have to note that for a party such as the M5S the procedure that defines the programmatic agenda is of particular importance. As we have said, elected representatives are considered mere "spokespersons" of citizens' requests and they are bound to respect the established programme. Especially in the presence of such a "binding mandate", the fact that programmatic decisions are in the hands of the leaders reveals a highly centralised structure.

Lastly, with regard to the emergence and characteristics of intermediate bodies, we can note the first signs of the formation of a sort of "coordination network" within the party, albeit a completely informal one. In the absence of a formalised structure, the elected representatives at the local level became organisational reference points and carried out coordination activities (this has also been highlighted by Vignati 2015b). These transformations were not codified in official documents but emerged from interviews with privileged witnesses and activists.

In 2010, the first regional councillors were elected in Piedmont and Emilia Romagna and therefore they became a reference point certainly at the regional level, but the two presidential candidates, Bono and Favia, also became a national reference point at the time, until 2013 of course, until the general elections. [...] In 2013, 160 MPs were elected and therefore the focus shifted a little. In the years between 2010 and 2012, there were very few of us, Bono and Favia were a reference point for everyone, and it was really difficult for them to combine their activity in the regional council, their activity at the municipal level, their coordination activity at national level and so on. [What is this "national coordination"?] Basically, it does not exist [...] the fact is that if there are problems, if there are issues, or simply journalists, or anything else, in 2012 they called Bono or Favia and now they call Di Maio or Di Battista (Int. 9).

In this phase, the contours of the organisation are defined more clearly. In the first place, we see that members have relevant decision-making powers regarding candidate selection: all the M5S's candidates are chosen by members through online or offline consultations. On the other hand, this this does not mean less power held by the leaders, who have not only the power to decide on the definition of the national electoral programme, but also to control above all the decision-making processes that take place online, and to give local groups permission to use the party logo. As regards the growth of an internal organisation, faced with the absence of an official intermediate structure, the first locally elected representatives (Davide Bono and Giovanni Favia, elected to the regional councils in 2010) began to play an informal coordination role. This—as we shall see will be amplified in 2013 with the passing of the representation threshold by the M5S and the election of more than 150 MPs at the national level.

#### *Representation (2013) 5.*

According to Pedersen, the passing of the threshold of representation constitutes one of the most significant moments in the evolution of a political party. This threshold was crossed by the M5S during the parliamentary elections of February 24th and 25th, 2013 (ITANES 2013). This was an extremely significant step forward for this political player, which partially unexpectedly gained more than 150 elected representatives in the two branches of the Italian parliament. Given the way in which the party was founded, the elected representatives did not share a common political culture. This was their first political experience, and they had arrived in parliament unexpectedly in many cases: "with a handful of clicks [...] we were put on the list". They were "neophytes", people who "cross a *stargate* and find themselves in another universe" (Int. 8).

A few days after the elections, Grillo published the *Code of Conduct* for the newly elected MPs on his website. In addition to the prohibition of associating with other parties or groups if not for votes on individual shared points, to the obligation of resigning after a first-degree conviction and to the commitment to reduce their salary, some specifically organisational rules were defined in this document, such as the principle of rotating the figure of the party whip and spokesperson in the Chamber of Deputies and Senate. Moreover, it established that decisions on institutional offices will be taken by MPs by majority.

It is difficult to give an unambiguous interpretation of these choices. On the one hand, similar decisions, especially the one on the rotation of offices, recall some practices that were common in Green parties (including the Italian Greens). These were aimed at avoiding hierarchies, centralisation and the emergence of strong personalities within the party. On the other hand, the horizontality that such practices would like to preserve in a context of authentically bottom-up parties is likely to turn into forms of uncontrolled and uncontrollable centralisation if introduced into an organisation, such as the M5S, in which from the beginning the leader was of great relevance. Preventing the formation of internal "intermediate bodies", in fact, would prevent the formation of groups or factions within the party, and hence the risk of the leadership being criticised or deposed. It was not by chance, in fact, that the great electoral success and the considerable media exposure of the new party led to the expression of internal dissent by MPs critical of the party line being discouraged. In the M5S, dissent was punished with expulsion, which was proposed by the leader and ratified by members with an online vote. As noted in the interview below:

[2013] was also an excuse used by the leadership of the Movement to eliminate any kind of internal discussion and thus to create a party made of people who obey, so there has been a cultural transformation. To begin with, we preached active citizenship, the fact that citizens must know, must be distrustful of power, understand if what the rulers say is true and have their own opinion. Suddenly it has become: "we must all be aligned, we must not have different ideas, we must repeat what the leaders say and never make any criticisms; if not, we help newspapers that hate us" and so on. And so, there is precisely an anthropological change (Int. 7).

Among the provisions present in the *Code of Conduct*, we can also mention MPs being prohibited from participating in talk shows or the decision, taken by Grillo, to form two "communication groups", whose composition was also defined by the leader, one for the Chamber and the other for the Senate, to manage the party's external communication. The same provision was inserted into the *Code of Conduct* for EMPs in 2014. The 2014 *Code* also stipulated a fine of 250,000 euros if the norms are not respected.

In this phase, we do not find any further formal changes with respect to the dimensions examined, even though, as it has been noted with respect to the local level in 2010, with the entry of more than 150 MPs into parliament we see, despite rhetoric contrary to personalisation, the strengthening of some personalities that, thanks to their visibility, became prominent figures in the M5S. Therefore, there started to be a sort of "intermediate structure" within the parliament, which will later be called the "coordination structure" of the M5S.

Di Maio, Di Battista and Fico are the three who have taken on importance... Especially Di Maio and Di Battista, and so, by now, they are quite on their own, they have their networks in the parliamentary groups... (Int. 7).

#### *Institutionalisation (2014–2017) 6.*

According to the lifespan model, the passing of the institutionalisation threshold corresponds to the confirmation of the party's parliamentary presence. In the case of the M5S this took place on the occasion of the elections of March 4th, 2018 (Valbruzzi and Vignati 2018). However, since 2013 there have been various local and regional electoral competitions: the M5S took part in them, obtaining fluctuating results (Colloca and Marangoni 2017). On the one hand, the European elections of May 2014, in which the party achieved a good result, were a confirmation of the electoral strength of the M5S at the national level<sup>49</sup>. On the other hand, at the local level, over the years the M5S has had numerous municipal and regional councillors elected as well as a number of mayors, including Turin's (Biancalana 2019) and Rome's (De Rosa and Quattromani 2019). We can thus consider 2014 as the starting point of the institutionalization phase. This is a phase full of organisational changes for the M5S, which are even present for the first time in official party documents.

The party rules of the M5S were indeed published in December 2014. For the first time, the rights of members (article 1) were codified. The M5S's members have the right to: discuss and determine the political direction of elected representatives through the party's online participation tools; take part in online consultations; become candidates. Neither the powers granted to the leadership, nor the mode of its selection changed; on the contrary, Grillo's figure as the incontestable "political leader" (*capo politico*) of the M5S was established formally, including his relevant powers

<sup>49</sup> 21.2 per cent of the votes; the outcome was, however, perceived as a defeat because of the high expectations that were placed on these elections.

over the organisation such as the power to authorise the use of the symbol, the power to call an online consultation and the power to start an expulsion procedure.

We can also observe a succession of the "dynastic" type following the death of Gianroberto Casaleggio, which occurred in April 2016. Although, at first, his role had not been formalised, over time his son Davide—also a member of the family business: he had been involved from the beginning in the management of the blog and in the development of the online participation platform—also started to represent the M5S, taking part in official events and press conferences. The role of Davide Casaleggio in the organisational structure of the M5S was formalised in 2016. In that year, Associazione Rousseau (*Rousseau Association*) was created: this is a private association that had the task of managing the online participation platform of the M5S. Davide Casaleggio is the president both of the association and of the company Casaleggio Associati.

At the national level, the candidates' selection process remained substantially unchanged and happened through online consultations. At the 2014 European elections, those who had been members of the M5S since at least December 31st, 2012, and did not hold any elective office could become candidates and be voted by other members. Those who had been members since at least June 30th, 2013, could vote. Furthermore, the party rules published in December 2014 officialised the procedure of online voting for the selection of regional councillors.

With regard to the procedures used to define the party's programme and its policy agenda, they seemed to move in a more inclusive direction. We have seen that the electoral programme presented by the M5S for the 2013 general elections had been defined "from above". However, starting from 2014, the M5S began to use online consultations more systematically, giving members the opportunity to decide on topics such as the electoral law, the abolition of the crime of illegal immigration and civil unions. In addition to the definition of electoral programmes and policies, online consultations were used to select candidates for elective offices, to ratify the expulsions proposed by the leader, to decide on matters of alliances and strategies as well as, as we will see, for the election of the members of the new-born intermediate organs.

Such practices of inclusion of members in the decision-making processes of a political party are unprecedented in Italian political history and constitute a significant innovation introduced by the M5S. At the same time, this opening of the decision-making processes is controversial because the content of the consultations, as well as the options on which it is possible to vote, are decided "from above", "from Milan", as highlighted in this excerpt.

The rule was that everything that goes beyond our five stars, our issues, our lines, it's to be voted for on the blog so the request started from us, or we said, "we would do this", and from Milan they would say to us "yes, but it is better to vote for it" (Int. 8).

As observed by Mosca (2015a), in fact, online voting is a problematic consultation tool from many points of view, including transparency, timing, choice of topics and incomplete and asymmetric information regarding the issues to be voted on, as the leader's ideas are always highlighted more than those of the opposing positions.

But in the end this stuff [online votes] is now only a façade; when you want or need legitimation, maybe to overcome an internal obstacle or to make a good impression with the newspapers then we vote; if you are afraid of the result, consultations are not organised (Int. 7).

Last but not least, in this phase the control of the regular unfolding of online votes was entrusted to Casaleggio Associati staff, which could allow for manipulation of the results. In addition to the increase in the number of online consultations, in this phase the entire online participation platform of the Movement (since July 2015 renamed *Rousseau*) was extended with new functions and tools. For example, *Lex* allowed members to express their opinions on the laws that national elected representatives or regional councillors intend to present; *Lex iscritti* enabled members to propose draft laws; in *E-learning*, online courses for newly elected representatives were offered; *Sharing* allowed representatives to share their activities and the proposals that they presented in the assemblies.

The most important organisational change, however, relates to the emergence in this phase of official intermediate bodies for the first time. In November 2014, the so-called *direttorio* (directorate) was created. It was defined in a post on Grillo's blog (post *Political statement no. 55*) as the "representative structure" of the M5S. It was composed by five people who were to become a "wider reference for the M5S", meeting regularly with Grillo "to examine the general situation, to share the most urgent decisions". A sort of intermediate structure besides the leader and the Casaleggio Associati, even if not formalised in the Non-statute, was thus created, and was formed by MPs, formalising the role that some figures (Di Maio, Di Battista, Fico) assumed in the M5S. However, the methods of selection are also interesting: it was a closed list of five people<sup>50</sup> defined by the leader, who the members could only accept or reject with an online vote. Grillo announced the dissolution of the *direttorio* in an interview in autumn 2016.

At national level, there is no organised structure, hierarchical structure; we have chosen to do so, and this has meant that then, at some point, some sort of leadership is necessary, and, in the end, you see that the *direttorio* is born (Int. 5).

Casaleggio decided, he saw the chaos that was there; the idea of rotating the whip didn't work and so he thought: if we have to have leaders, it's better that I choose them and that I trust them (Int. 7).

Who chose those people? They chose them because they are trusted people... They needed someone to call and say, "take care of this" (Int. 8).

Only a month later, in December 2014, the appeal committee (*comitato d'appello*) was created. The new intermediate body, established in the party rules of the M5S (a document never voted on or discussed by members), had the task of examining the contested expulsions, contrary to what happened previously, when the expulsions decided by the leader were unquestionable, and of acting as a guarantor in the case of modifications to the rules. The appeal committee was composed of three MPs, two of whom were chosen by members among a list of names proposed by the leader, while the third was nominated directly by the leader.

A third intermediate body was established as a result of an update of the party rules and of the statute, which took place in September 2016. On this occasion, members had the opportunity to approve or reject the new version of the statute and to choose between two different versions of the party rules with an online vote. Following this update, a new intermediate body was created: the board of advisors (*collegio dei probiviri*), composed of three MPs chosen with an online consultation on the proposal of the political leader of the M5S. The board of advisors had the task of managing disciplinary sanctions, including expulsions (whose re-examination, in the event of a dispute, was the task of the appeal committee).

From 2016, therefore, there are three intermediate bodies of the M5S: the assembly of members, that is, all the members of the M5S who had

<sup>50</sup> The members of the *direttorio* were: Luigi Di Maio, Alessandro Di Battista, Roberto Fico, Carla Ruocco and Carlo Sibilia.

the right to vote in online consultations; the board of advisors; and the appeal committee. These are intermediate bodies "without any directive or representative function" (Party rules, article 1bis). At this stage, there was also the creation of an informal network of elected representatives that act as internal reference points of the M5S. This is what is called, informally, the "coordination structure" of the M5S (post *The future of the M5S*, 12/9/2016). It is a sort of structure composed by some elected representatives that, as in the case of the *direttorio*, perform some important "functions". In this case, no online consultation was organised. The elected representatives seem to have been simply chosen by the leader, using non-transparent methods.

They are functions, we have solved the problem in that way, precisely because we want to be a movement in theory without leaders, then television creates the leaders anyway, but we tried to propose a series of functions that we must necessarily have […] and therefore, in this sense, it is a response by the M5S to the need, on the one hand, to organise and, on the other, not to be structured in a party-like way, because we realise very well that there are difficulties, there are some shortcomings on the part of the M5S, when you choose not to have a party structure (Int. 9).

In this phase, members' rights were clearly defined for the first time and, as we have seen, they included a vast array of powers. However, rights were granted to members as individuals and could be exercised only though an online platform controlled by the leaders. Moreover, the statutory changes that occurred in this phase could be seen as the first steps towards the formation of an intermediary structure in the M5S, albeit different with respect to those of traditional parties. But we have seen that the members of these intermediate bodies, despite being elected with online consultations, were in reality substantially chosen by the leader. Furthermore, it is necessary to remember that the M5S's intermediate bodies were composed of MPs who, according to the rules of the party, can complete a maximum of two mandates.

Therefore, the emergence of such an internal structure didn't essentially change the M5S's disintermediation strategies. As we have seen, these strategies consist in replacing the traditional party structure with an internet-based one, since members can have a say on political and party decisions through the platform *Rousseau*. This can be seen as a prevalence of the dimension of disintermediation from below. However, we have also seen that the leaders strictly control the platform, and the decision-making processes that happen within it are guided from above.

#### *Government Participation (2018) 7.*

At the 2018 general elections, the M5S obtained 32.8 per cent of the valid votes (more than ten million) and had more than 300 MPs elected (Biancalana and Colloca 2018). Contrary to its precedent provisions and principles on alliances with traditional parties, in June 2018 it formed a coalition government with Salvini's Lega (League), and in September 2019 with the Partito Democratico. In both cases, the prime minister was expressed by the M5S: he was Giuseppe Conte, a law professor near to the M5S but without any previous political experience, has become one of the most important characters in Italian politics over time, also because of his role of PM during the first outbreak of the pandemic<sup>51</sup> .

After one year in government, in May 2019, the M5S participated in the European elections, losing 15.3 per cent of the votes, and achieving its worst result in a national electoral competition (17.1 per cent). We know that in the life cycle of all new parties, crossing the threshold of government marks the start of a crucial and completely new phase. It can be suggested that the first experience of government is especially problematic: it represents a completely novel set of circumstances—circumstances of which the parties concerned have no experience and for which they have no established routines. It can affect the parties' internal organisation and their relations with other parties, especially when, as in the present case, the other parties' members in the coalition have more experience.

The phase that preceded the 2018 elections was full of organisational changes. At the end of 2017 a new statute was released, and this brought many innovations to the internal structure of the M5S. In reality, a major change had already occurred in September 2017, when Luigi Di Maio, a young MP elected in 2013, without any previous political experience, was elected as the M5S's candidate prime minister in an online vote. This seemed to herald a more relevant role for Di Maio within the organisation, and that is precisely what happened three months later, with the publishing of the new statute.

<sup>51</sup> His increased political experience, media exposure and approval rating among citizens and members led him to try to take over the party in early 2021.

The first thing to note is that, just like all other M5S's documents, the content of the 2017 statute (the name Non-statute was abandoned) was not subject to a public discussion. It was simply published, and even the positions that, according to it, were supposed to be elective, were already set. The two most important organisational novelties recorded by the statute were the growth and institutionalisation of intermediate bodies and the formalisation of two different roles of leadership, Grillo's and Di Maio's. On the one hand, we witness a detachment of Grillo from the M5S, also indicated by the fact that, in January 2018, he separated his personal blog from that of the Movement; on the other, this did not result in less centralisation of the party.

As regards intermediate bodies, six internal organs of the M5S are listed in the new statute: the guarantor; the political leader; the treasurer; the members' assembly; the guarantee committee (*comitato di garanzia*); and the board of advisors. The guarantor, Beppe Grillo, has the task of protecting the fundamental values of the M5S and has the power to interpret the rules of the statute. He remains in office indefinitely and can be removed by the proposal of the guarantee committee by an absolute majority of its members; the decision should be ratified by an online consultation of the members. The political leader, Di Maio, is the legal and institutional representative of the M5S, in charge of ensuring the unity of the political address of the M5S. The political leader is elected by the members, holds office for 5 years and can be re-elected once. He has many powers including deciding to hold online consultations and a sort of veto power over candidacies. He can be removed following a decision by the guarantee committee or the guarantor, and the decision must be ratified by an online consultation.

Moreover, the guarantee committee supervises the correct application of the rules of the statute, decides on the existence and loss of the requirements to become a member or a candidate, and approves the regulations and codes proposed by the political leader. The three members of the guarantee committee are elected by the members: they can choose them from a list of six names, decided by the guarantor. It lasts 4 years and is not renewable; from the next election, its members can't be elected representatives. Finally, the board of advisors imposes disciplinary sanctions. It is composed of three members elected through an online consultation; members can choose them from a list of at least five names proposed by the guarantor.

As regards the procedure of candidate selection, even in the case of the 2018 elections it happened completely online. Compared to the 2013 and 2014 ones, we witness further opening of the decision-making processes, but also closure of them: *all* members, without restrictions, could vote and be voted for, but the political leader had a sort of veto power over candidacies, and can directly decide on the candidates of the single-member constituencies (37 per cent of the seats according to the current electoral law). Even in the case of the selection of candidates for the 2019 European elections all members could be voted for and vote, but the political leader had the power to decide the names of the first candidate in the list of each constituency autonomously.

In contrast to 2013, the programme for the 2018 general elections was the subject of a discussion within the party, since a series of online consultations were held on *Rousseau* during the year preceding the elections. Starting from December 2016, the members of the M5S had the opportunity to vote on sixteen points of the programme. It should be noted, though, that these were votes on predetermined options, the answers to which could often only be "yes" or "no". Despite these limitations, we can say that the M5S notably opened its internal decision-making processes, regarding the definition of the programme. In the case of the European elections of 2019, where the programme was less significant, members had the opportunity to choose the "priority actions" that the future European MPs would carry out in Europe.

In the *Code of Conduct* for MPs, renamed *Code of Ethics*, the provision of the rotation of the assignments within parliamentary groups was maintained. Nevertheless, the names of the representatives holding that position had to be chosen from a list prepared by the political leader, who also had the power to decide on the first party whip and spokesperson, who will last 18 months. Also, communication groups, whose composition is again decided on by the political leader, were maintained. Moreover, MPs were required to finance the platform *Rousseau* monthly, which is intended as the official means of communication for MPs.

As regards participation in TV programmes, it is allowed, but MPs are obliged to coordinate their participation with the national communication managers of the party, who are in turn designated by the party leader. Finally, in this new version of the statute the role of the Associazione Rousseau, whose president, as we have seen, is Davide Casaleggio, was officialised as the manager of the IT system (the platform *Rousseau*), through which all the decision-making processes of the M5S pass (article 1c). This is relevant because Associazione Rousseau is an entity detached from the M5S, which basically manages the entire decision-making process of the party, and over which members have no form of control.

To sum up, with the publication of the new statute, the organisational structure of the M5S became more complex. The 2017 statute gave the M5S an almost party-like shape, at least on paper. Although Grillo's position remained unquestioned, we witness a veritable change of leadership as more and more and more power is held by Di Maio, also thanks to his prominent role both in the governments with the Lega and the Partito Democratico. But the change of leadership and the enlargement of intermediate bodies did not change the essentially centralised nature of the party. It is true, on the one hand, that individual members had a vast array of rights, but when the stakes were high (for instance, decisions on key positions within the party, or in the case of the most secure candidacies), the leadership retained its power, also with respect to the elected representatives.

#### *The Reform and Refoundation of the Party (2019–2020) 8.*

After the disappointing results of the 2019 European elections, Di Maio started a process of reform of the organisation that was to shake the very foundations of the party and led, in the first half of 2021, to the veritable refoundation of the M5S. In summer 2019, not only was an exception to the limit of the two terms for municipal councillors introduced (the so-called *mandato zero* or "zero term"), but the role of the "facilitator" was also created. The facilitator is a sort of coordinator of the party, and a transmission belt between the centre and the territories. According to the party's rhetoric, his task is to "support" the political leader and to "listen to the requests from below […] and transform them into activities such as organising events, training moments, communications, etc. to be carried out in the territories". Indeed, facilitators can operate at the regional or national level. While regional facilitators have the task of reinforcing the linkage between the centre and the territories, national level facilitators are responsible for the organisational aspects of the party and for the articulation of policies.

The creation of regional facilitators relates to the relevant aspect of the relationship between the local and the national levels in the party. We have seen that the local level is crucial for the M5S in many respects: it was at the local level that the M5S originated, through the autonomous formation and self-organisation of groups of active citizens, who later participated in local elections. At this level, the party has always worked essentially as a "brand-giver": candidates that wanted to stand in local elections had to request a certification in order to use the party logo. Besides this, local groups, although they do not represented official branches of the party, organised themselves autonomously, and there was not a structure, except for the elected representatives. Apparently, the position of the facilitator was conceived to fill this gap.

There are two types of national level facilitators: six are responsible for the organisational aspects of the party (the six areas of action are: communication; training and staff; local activism; election campaigns; support for local authorities administered by the M5S; and coordination and internal affairs<sup>52</sup>), while twelve are responsible for the articulation of policies. Thematic facilitators<sup>53</sup> are supported by an 8-people group of experts. Together, the facilitators compose what has been called the *Team del futuro* (Future Team), which can be considered a new organisational structure of the M5S. Nevertheless, these changes were not to be codified in the party statute. Facilitators are essentially coordinators with no formal rights or duties, but they testify to the growth of an internal structure in the party and the attempt to create a more widespread structure of power, albeit strictly controlled from above.

As regards the control from above, it is worth mentioning the selection procedure of the members of the Future Team. In the case of the organisational facilitators, the choice left to the members was only to accept or refuse them in full, while in the case of the thematic ones, members had the opportunity to propose themselves and to form a group composed of other members. Different groups and projects were subsequently voted for online. In the case of regional facilitators, members had the opportunity to propose themselves and vote, but in the end it was the leader that had the power to choose the facilitators from the three voted for most by members. Needless to say, the leadership could screen candidacies and thus exclude candidates at any time.

Di Maio resigned as political leader in January 2020, on the occasion of the public presentation of the regional facilitators. The reasons for his resignation were the Movement's electoral defeats and the loss of members in the parliamentary groups. According to the statute, his role was taken

<sup>52</sup> The six facilitators are all elected representatives except Elena Sabadini, the facilitator in the area dedicated to coordination and internal affairs. She is a close collaborator of Davide Casaleggio.

<sup>53</sup> The twelve thematic areas are: agriculture and fishing; the environment; the economy; foreign affairs and the European Union; justice and institutional affairs; business, innovation; education, research and culture; work and family; health; security and defence; transport and infrastructure.

on by Vito Crimi, as a senior member of the guarantor committee. Consequently, a meeting of the base was called for the spring, the so-called *Stati generali*, a kind of party congress. This was a very important step for the party, which had always kept his distance with respect to traditional political actors. For this reason, the Movimento 5 Stelle had never had a real congress. It had only held some annual meetings, called *Italia 5 stelle*, that were more a matter of communication than of decision. The congress was called for the spring, but the outbreak of the pandemic stopped the process and, in the end, the *Stati generali* were held online between October and December 2020.

With regard to the process, the congress had three main steps. In the first place, in October 2020 regional facilitators called for regional online meetings, which were open to registered members resident in the region. The task of the meetings was to prepare a regional document, based on three areas: rules and principles, organisation and structure, and political agenda. In the same meetings, 305 delegates were elected by the regional assemblies to represent the requests of the region in the next phase of the congress. According to the party, 8,000 members participated in this phase. There were three types of delegates: members, elected representatives at the local level, and elected representatives at the regional, national and European levels; each type of actor could elect only his type of delegates.

Second, in two days in November 2020, the regional documents were discussed by delegates in order to create a single synthetic document. The phase was managed by Avventura Urbana, a company specialising in participatory processes. Finally, the third step, in December, was the members' online vote. 23 proposals were prepared by the political leader on the basis of the document, which were consequently voted on via *Rousseau*. In February 2021, after an online vote, six new points were added to the statute. In the month of December, the party also opened a second phase of the *Stati generali*, in order to create the new political agenda of the party from below. This phase was also open to non-members.

The *Stati generali* testify to a deep change for the party. In the first place, in terms of procedures, for the first time the focus was on deliberation rather than on decision, and the existence of forms of delegation was acknowledged. If it is true, on the one hand, that in the end the points were assembled by the political leader and voted on on the platform with Yes/No votes, the points of the proposals stem from a deliberative process that involved registered members first and then delegates. The "rules of the game" came from above, and the structure remained centralised, but it was a rather inclusive process from below, which took into consideration the requests of the base within a codified process.

Even with regard to content, the *Stati generali* represented a real Copernican revolution for the M5S, although it occurred in the wake of the process of reform started in 2019. In the final document, we can read that members asked for the principle of collegiality to inform all the decision-making and governing organs of the party, including the party leadership. This was an important novelty for a political actor that had always placed great importance on the figure of the leader. As regards intermediate organs, members asked for their number to be enlarged, to also structure them also at the regional level, and for the opportunity to recall them. Moreover, they asked for the formalisation and strengthening of the party's territorial structure, for instance with the acknowledgement of the role of local groups, even when they don't have an elected representative. Against the backdrop of a party born in opposition to party structures, this testifies to a clear need for intermediation expressed by party members.

Finally, a last point made by members during the *Stati generali* regarded the relationship of the party with Davide Casaleggio and Associazione Rousseau. In the 2017 statute, it is written that all the decision-making processes of the party have to pass through *Rousseau*, which is managed by Associazione Rousseau. One of the outcomes of the congress was the request to regulate this relationship with a formal contract: "Relations with the platform manager must be governed by a specific service contract or partnership agreement that defines the services, roles and reciprocal duties". Indeed, the relationship between the M5S and Associazione Rousseau is a thorny one, as many MPs over time criticised the role of Casaleggio and his association with the party. As early as autumn 2020, MPs in the Chamber of Deputies changed the statute of the parliamentary group, stating that the group's means of communication shouldn't be only those managed by Casaleggio, such as the site and the platform.

For his part, on October 4th—the anniversary of the birth of the M5S— Davide Casaleggio wrote a post on the blog recalling that the Movement should not become a party, and that "if the transformation into a party starts" Associazione Rousseau will no longer support such a political actor. Coherently with these positions, he decided to leave the party in June 2021, also ending the collaboration between Associazione Rousseau and the M5S.

The M5S was born with some promises […]. The first of these is that we would never become a party. The party has a group of a few people who decide everything for everyone. Electoral lists, appointments, programmes, electoral alliances in the various cities. In the Movement, on the other hand, power is exercised from below and there are ways to ensure transparency and the sharing of choices among members. The party believes in delegation to a representative, the Movement in the active involvement of individual participants. The party takes public funding, creates salaried structures for former elected officials, does not believe that there are limits to parliamentary mandates and creates decision-making structures that expropriate citizens from their role of direction and choice. We will guarantee the activities that will be requested by the political leader of the M5S […] for the realisation of the path that the Movement wants to take, but if, for some reason, the transformation into a party starts, our support will no longer be guaranteed (Post *Noi siamo Movimento*, October 4th, 2020).

To conclude, we see that the passing of the last threshold and the first years in government contributed to completing a course already undertaken by the party. With the 2017 statute, the creation of the facilitators, and the outcome of the *Stati generali* we see clear enlargement and strengthening of the party's intermediate structure. As we could expect on the basis of the literature on new parties' evolution, this was an inevitable step, since the Movement increased its electoral strength and acquired government responsibilities. Despite Casaleggio's views on the future of the M5S, even members' requests testify to a need for organisation and intermediation. However, at least until the end of 2020, this did not change the fundamentally centralised nature of the party, in which the leader retained great shares of power.

With respect to what we could define as the party in central office, that is, members of the intermediate structure, the political leader has the power to propose and, in some cases, choose them. With respect to elected representatives, he has relevant decision-making powers regarding their communication. With respect to members, he has the power to call online consultations through which they could express themselves on various party issues. Finally, we should not forget that, although some procedures for his removal formally exist, in the 2017 statute Beppe Grillo was appointed indefinitely as guarantor of the party with relevant decision-making powers. So, if centralisation was one of the main characteristics of the M5S at the end of 2020, we can say that the clashes between Conte and Grillo, as well as the refoundation of the party on the figure of Conte in the summer 2021, confirm once again this trait of the party<sup>54</sup> .

Studying Italian politics is like chasing the future; therefore, I cannot analyse in detail here the most recent evolution of the M5S or the new statute released in the summer 2021. However, summing up the transformations until the end of 2020, we can say that in this phase members had the opportunity to vote extensively on candidates and policy selection, and on important party decisions. For the first time, they also had the opportunity to express themselves on the party's structure, through a codified and structurated process that led to important (albeit short) statutory changes, such as the proposal of collective leadership. Nevertheless, we can conclude that even in this last phase the dimension of disintermediation from above prevailed. Members' power, that is, disintermediation from below, was limited by the fact that the "rules of the game", and the infrastructure in which their decision-making occurs, are strictly controlled by the leadership. It remains to be seen, and it will be the task of future research, whether and in which ways the new structure of the Movimento 5 Stelle led by Conte will resemble its predecessor.

#### *Conclusions 9.*

The aim of this chapter was to analyse the organisation of the M5S and its evolution in order to assess whether or not it makes use of disintermediation strategies, how they are interpreted by this political actor, and which dimension prevails among those outlined. From the very beginning, the M5S made use of disintermediation strategies. The will to create an unmediated link between citizens and power is, according to the official

<sup>54</sup> In July 2021, after a series of clashes between Beppe Grillo and Giuseppe Conte, the former PM and new leading figure of the Movement, a new statute was presented by Conte. Beppe Grillo will remain guarantor but, in line with the evolution underlined in this chapter, the new statute paves the way for an internal organisation of the M5S more similar to that of a traditional party. For instance, territorial groups are acknowledged for the first time, as well as a national council of the party. In the month of August, the statute was approved by members through an online vote on SkyVote, a voting platform managed by a private company that, differently from Associazione Rousseau, is completely detached from the M5S. A few days after, again through an online vote on SkyVote, Giuseppe Conte was elected as president of the party with a 92,8 per cent approval rate. This step, which cannot be analysed here, represents the last step in the institutionalisation of the party.

rhetoric, the main aim of this political actor and is mirrored in its organisation, which, at least initially, is made up only of the leader, members and elected representatives, who are connected by the internet. Disintermediation strategies can be defined precisely as rhetoric or practices developed by parties in order to stage or deliver an unmediated relationship with citizens. In particular, we have seen that disintermediation in party organisations involves the weakening of the party's intermediate organisation, through greater powers given to the party leader (disintermediation from above) and party members (disintermediation from below).

In order to understand the prevailing dimension of disintermediation, we can go back to the five dimensions taken into consideration within the course of this analysis. As regards the candidates' selection process, we have seen that the M5S's members have always had a lot of power at the local<sup>55</sup> and national levels. This is true, especially in the last period, regarding the definition of programmes and policies too. On the other hand, the prerogatives of the leadership remained relevant and substantially unchanged until the end of 2020. Despite the formal leadership change from Grillo to Di Maio, the leadership retained significant powers and control over the organisation. Notwithstanding the discussions on the creation of collective leadership during the first congress of the Movement, the *Stati generali*, the recent clashes between Beppe Grillo and Giuseppe Conte over the future of the M5S testify once again to the importance of personal leadership for this political actor.

Although a new statute was presented by Conte only a few months after the statutory changes had stemmed from the congress, the *Stati generali* represented an important moment of change for the Movement. If until the end of 2020, the internal rules of the M5S, starting from its founding documents, weren't the object of a real discussion, but were simply presented to members, this started to change with the *Stati generali*, both in method (deliberation instead of decision) and in substance (requests for intermediation instead of disintermediation). Moreover, an important step towards the institutionalisation of the party emerged with respect to the IT

<sup>55</sup> We must specify that, in some cases, the influence of the centre on the territories was relevant. For instance, in 2016, Grillo upset the result of the online consultation for the candidate mayor of the city of Genoa. In 2018, but this had also happened various times in the past, he decided not to give "certification" to two local groups (Siena and Vicenza) that wanted to participate in local elections. This decision was neither shared with nor explained to the local groups. They, as we have seen, are for the moment not formalised in the party model and have no codified way to make claims to the centre.

infrastructure in charge of managing the decision-making processes of the Movement. Following the exit of Davide Casaleggio from the M5S, the online decision-making processes of the Movement are no longer controlled by Associazione Rousseau: the Movement will collaborate with private companies for the managing of online votes. This can be seen as a sign of enfranchisement of the Movement not only from one of its founders, and thus a decrease in the dimension of disintermediation from above, but also from the most radical promises of online direct democracy.

We have seen that with the evolution of the organisation, we witness a growing presence of intermediate bodies in this political actor. As expected, with the passing of the various thresholds identified by Pedersen, and in particular with the passing of the last one, that is, government, the organisation of the M5S has gradually become more complex, both through an informal "coordination structure", not formalised in official documents and with the codification of more formalised internal organs. With respect to its origins, the M5S now has an almost party-like structure. But, in terms of the characteristics of that structure, we have also seen that it is substantially different from a traditional one, as until the end of 2020 it was the political leader who had significant powers in appointing its members. Moreover, the M5S continued to be different with respect to traditional party actors in many respects, first of all with respect to the two-term limit for elected representatives. It remains to be verified if the forthcoming formalisation of the new party structure in 2021—with territorial groups and a national council—could imply a sort of return to "old" forms of intermediation.

To sum up, we can affirm that in the M5S's disintermediation, strategies from below and from above coexist, but that until now disintermediation from above has prevailed. On the one hand, this is not a "memberless party" (Mazzoleni and Voerman 2017). Through the internet, members have significant decision-making powers, compared to those traditionally granted to party members. However, on the other hand, the structure within which they can exercise their power is highly centralised. This party did not present a structure similar to that of traditional parties: the M5S, at least at the beginning, was composed only of the leader, the elected representatives and members, who were connected with the internet. It is through the internet that an unmediated connection between citizens and power, but also between leader, elected representatives and members, was created.

In the M5S, we are not faced with a simple attempt to *weaken* the party's internal organisation. From my analysis of party documents, we can see that this political actor, at least at the beginning, did not present an intermediate structure or a middle-level elite at all, which were replaced by an internet-based organisation. The request for collective leadership, an outcome of the *Stati generali*, could be seen as a sign of the weakening of the dimension of disintermediation from above that has long been present in this political actor, not through more disintermediation from below, as promised by the party rhetoric, but with more intermediation. Future analyses of the new party documents presented by Giuseppe Conte will tell us if this is the case.

Finally, in addition to the intermediate bodies created after 2017, we can say that over time new and old forms of intermediation emerged. In the first place, if the internet can be seen as a tool of disintermediation, it can also represent intermediation of a different kind. I will analyse this point in detail in the next chapter, which is dedicated to the M5S's online participation platform *Rousseau*, but we have already seen that Casaleggio Associati and Associazione Rousseau acted as powerful gatekeepers in the party's decision-making processes. In the second place, we have seen that even in an organisation that presents strong rhetoric on leaderlessness and that affirms it is contrary to personalisation and opposed to traditional party organisations, forms of formal and informal leadership and an internal organisation have emerged over time. This is true not only with respect to Grillo and Casaleggio, who had a recognised and codified role in terms of leadership from the beginning. Over time, and we will see this in the chapter dedicated to the M5S in Turin too, elected representatives start playing coordination roles and later have relevant formal positions in the organisation, such as in the case of Luigi Di Maio or, more recently, Giuseppe Conte.

## Between Members' Empowerment and Pseudo-Participation: The Movimento 5 Stelle's Online Participation Platform *Rousseau* 7.

#### *Introduction 1.*

The Movimento 5 Stelle represents one of the most advanced experiences of use of the internet for participatory and organisational purposes in a political party. The internet has always been fundamental to the party's narrative and organisation: we can say that it was through the web that, in the founders' view, an unmediated relationship between citizens and power, but also between leader, elected representatives and members, would be created. The M5S considers its elected representatives to be citizens' "spokespersons" and the internet a substitute for traditional party structures: in the M5S's rhetoric, the party on the ground has the right to "direct" the other two "faces" of the party's organisation, and until 2021 the digital platform *Rousseau* was the place in which members decided on parliamentary and party activities. The platform *Rousseau* thus falls into the direct democracy category of the use of the internet by parties; the tools belonging to this category could potentially empower party members, giving them a say in party decisions and then altering the distribution of the internal decision-making power in their favour.

In this chapter, after having analysed the references to the internet in the party statute and rules, and their evolution over time, I will analyse the functioning of the platform *Rousseau*. Although in early 2021 *Rousseau* stopped being the official participatory and decision-making tool of the party, the experience of this platform is one that is worth examining. *Rousseau* can be considered the core of the M5S's disintermediation strategies, as it gave members the opportunity to decide directly on many relevant issues, replacing to some extent the party's internal organisation. However, it is worth asking whether there was actual empowerment of members through this tool or whether, according to the "pseudo-participation" hypothesis (Pateman 1970. For an application of this concept to cyber-parties see Biancalana and Vittori 2021a), its function was mainly symbolic.

According to this assumption, digital tools wouldn't be used to grant members more power, but to give them the *impression* of influencing the decision-making processes of the party: they would be used more from a narrative than from a practical point of view. In these situations, no participation in decision-making actually takes place: for decision-makers the concern is to create a *feeling* of participation and directness, while retaining power in their own hands. In other words, in terms of practices, disintermediation from above would prevail over disintermediation from below.

Based on this premise, *Rousseau* will be analysed from two points of view: on the one hand, a description of its affordances (Dahlberg 2011) and functioning; on the other, an assessment of its impact, especially on party organisation. Three dimensions will be examined: the architecture of the platform and its affordances, that is, the features present in the digital tool and the activities that users are encouraged to perform; the transfer of power from the top to the bottom, that is, members' rights and powers and their ability to influence the "rules of the game"; the consequences on party organisation, in particular on the internal distribution of power.

The chapter will be structured as follows. In section 2, I will analyse in detail the role of the internet in the M5S's organisational structure, especially the references to the internet in the party's statute and rules and their evolution over time. In section 3, I will describe the available tools and functions that were present on the platform *Rousseau*, focusing on the kind of participation encouraged by them. In section 4, I will examine a specific function, online votes, through which M5S's members decided directly on some party decisions, such as the selection of candidates and the definition of electoral programmes. For this purpose, I have collected, classified and analysed all the online votes held by the M5S in the time span 2012–2020. Finally, in section 5, I will summarise the main findings of my analysis and assess to what extent *Rousseau* allowed the actual empowerment of the M5S's members.

### *The Role of the Internet in the M5S's Narrative and Organisational Structure 2.*

The internet has always been regarded by Grillo and Casaleggio both as an alternative tool for communication and counter-information, as well as a means of political organisation. The M5S's rhetoric on the internet has always set it apart from other parties and has been defined as "cyberutopian" (Mosca 2020): in the party's rhetoric the internet is pictured as "an inherently transparent, democratic, and accountability-enhancing technology" (Mosca, Vaccari and Valeriani 2015, 127. See also Biancalana 2014; 2017; Natale and Ballatore 2014). The vision of the internet as an ontologically positive technology, a natural creator of horizontal, transparent, participatory and non-hierarchical processes, is one of the main elements of the democratic vision of the M5S that emerges from the analysis of the leaders' discursive production and which makes it remarkably distinct from traditional parties (Mosca 2020).

In a relatively short time, everything will change. Traditional media will disappear, together with the majority of the hierarchical structures that govern the various aspects of society and economics. Among those are parties, which will be replaced by movements (Grillo, Fo and Casaleggio 2013, 7).

Representative democracy was probably the best model that we could have until a few years ago. But with the use of the internet and the set of tools that can be used through the internet, today participation is probably the best democracy that we can have (Post *Presentazione del #NuovoRousseau: Participate! Don't delegate*, 2/7/2017).

Online direct democracy has always been fundamental in shaping the M5S's identity. Although it has been demonstrated that its rhetoric and practices differ, as the rhetoric of horizontality, lack of leadership and spontaneity of the party have been used to mask, facilitate and eventually legitimise centralised and authoritarian practices (Treré and Barassi 2015), and that over time the most radical promises of members' empowerment through digital tools have been curtailed, it is true that the party managed to create an innovative organisation: the M5S represents a web-based organisational model very different from that of traditional parties in terms of membership, structure and forms of participation.

Although over time some intermediate bodies and an almost party-like structure emerged, until the end of 2020 the M5S did not have an official party in central office, as all the main party activities (for instance, enrolment) passed through the party's websites. We can say that Casaleggio Associati—the company owned by one of the founders of the M5S (Gianroberto Casaleggio) and, after his death, by his son Davide—and later Associazione Rousseau can be considered unofficial parties in central office. From the beginning, Casaleggio Associati managed the M5S's websites, members' enrolment and the certification of the local electoral lists, that is, the permission to use the party logo. Casaleggio Associati created the platform *Rousseau* and still have links with it: since April 2016 *Rousseau* has been managed by an association called Associazione Rousseau, whose president is Davide Casaleggio. Associazione Rousseau is an association detached from the M5S, over which members had no control, that until 2021 had, consistently with the provisions of the 2017 statute, the right to manage all the online decision-making processes of the party.

With regard to the party in public office, since 2012, the M5S has had local and regional councillors, MPs, members of the European Parliament, as well as mayors of various important towns such as Rome, Turin and Livorno elected. According to the rules of the party, elected representatives can complete a maximum of two mandates<sup>56</sup>. So, the people elected with the M5S are essentially citizens that become politicians for a short period of time, since it is, at least at the time of writing, impossible to complete a third mandate under the M5S's symbol. According to the party's narrative, elected representatives are considered spokespersons, who have to bring the requests of citizens, not considered by professional politicians, into the assemblies. At the regional and local levels, until the end of 2020, there weren't official regional and local branches of the party<sup>57</sup>, so that the elected representative "represented" the party both in the assembly and in the territory, as he or she was the only authorised subject to speak in its name, having obtained its certification and the right to use its symbol.

The M5S's party on the ground is composed of all the people registered on the party website. To the end of 2020, the declared number of members was around 170,000. Indeed, party enrolment was equated with website registration: the only way to officially<sup>58</sup> join the party and participate in its national activities was online. Every Italian citizen that is not yet enrolled

<sup>56</sup> Since 2019, there has been an exception for municipal councillors: the so-called *mandato zero.* See chapter 6. The derogation from this rule at the national level currently discussed within the framework of the refoundation of the party by Giuseppe Conte—would mean the complete institutionalisation of this actor.

<sup>57</sup> The creation, in 2019, of the position of the regional facilitator, can be considered a first attempt to create a regional structure. However, these changes have not been fomalised in the party statute. The new 2021 statute mentions "territorial groups" for the first time.

<sup>58</sup> It is true, though, that it is possible for everyone to participate at the local level. Indeed, while participation at the national level is online only, at the local level face-to-face participation is fundamental and highly developed. At the local level, the party works essentially as a "brand-giver". Candidates that want to run for local elections have to request "certification" in order to use the party logo. Besides this, in general, local groups, although they do not represent official branches of the party, organise themselves autonomously. Indeed, due to these characteristics Tronconi (2018) stated that the party structure of the M5S reflected Carty's "franchise model" (2004).

in another party could join the M5S. The registration is free of charge. Control over registrations—that is, control over party membership—is a key responsibility, and Associazione Rousseau was in charge of it.

Before we analyse members' digital rights and the affordances of the party's online participation platform, it is interesting to analyse the references to the internet in the party's statute. The first version of the statute of the M5S, the one significantly called Non-statute, was soaked in what has been defined as "cyber-optimist" rhetoric. According to the Non-statute, the party's "headquarters" were Beppe Grillo's website (article 1), and the internet was considered central for the party's "enrolment, consultation, deliberation, decision and election" (article 4). Moreover, the M5S granted "all internet users" the power to steer it, which is usually retained by a few (article 4).

The M5S is not a political party, nor will it become one in the future. The M5S wants to achieve an efficient and effective exchange of opinions and a democratic debate outside associations and parties and without the mediation of governing or representative bodies, *acknowledging to the totality of the users of the internet the role of government and direction* normally detained by a few (Article 4, Non-statute, 2009, emphasis added).

The point contained in this article is key to understanding the use of the internet by this party: in a techno-populist fashion (Deseriis 2017b; De Blasio and Sorice 2018), the internet is seen as the tool that directly links people and power. This provision seemed to soften in the 2017 statute.

The M5S aims to achieve an efficient exchange of opinions and a democratic debate, *acknowledging to all members*, in accordance with the provisions of this statute and especially through digital tools, *an effective role in the direction and determination of the fundamental choices for the association's political activity* (Article 2a, Statute, 2017, emphasis added).

In the first place, it is no longer "all internet users" but "all members" that have the right to exercise the role of government and direction. In the second place, the more general "role of government and direction" becomes the "determination of the fundamental choices of the association", that is, the M5S. This can be seen as an indicator of the normalisation of the cyber-optimist rhetoric that characterised the M5S in previous years<sup>59</sup> .

According to the party statute, the members of the M5S have a number of rights (article 3b, 2017 statute), all of which can be exercised online: the right to contribute to the definition of the political direction of the elected representatives; the right to participate in online consultations called in order to determine the fundamental choices for the political action of the M5S; the right to become a candidate at the national, European, regional and local levels; and the right to formulate law proposals that, if approved by members through an online consultation, can be taken into consideration by elected representatives. As regards online consultations, they can be called by the political leader or, in his absence, by the guarantor. According to the 2017 statute (article 4a), they can be used to: elect the political leader, the guarantor, the guarantee committee and the board of advisors; choose candidates and approve the political programme; approve members' law proposals; and impeach the political leader and the guarantor.

So, despite the softening of the cyber-utopian narrative that characterised the M5S in its early years, we see that, according to the 2017 party statute, the M5S's members can deeply influence both parliamentary and party activities through digital tools. In the next section, the architecture and affordances of the online participation platform *Rousseau* will be described.

#### Rousseau*: Design, Architecture and Affordances 3.*

The place in which online consultations and the other online decisionmaking processes of the M5S took place was *Rousseau*. Significantly named after the philosopher Jean Jacques Rousseau, it has been online in different versions since the end of 2012, and its available features have increased over time. The reasons why the M5S chose to call its platform "Rousseau" must be looked for, clearly enough, in the fact that the author of the *Social*

<sup>59</sup> In the 2021 statute, we witness further normalisation of the cyber-optimistic rhetoric: "The Movimento 5 Stelle promotes, *through internet platforms or other methods, possibly including non-telematic ones*, the involvement of its members in the process of identifying those who will disseminate and implement the Association's ideas, projects and social, cultural and political awareness campaigns" (Article 2d, 2021 statute, emphasis added).

*Contract* is one of the most notorious proponents of what is known today as "direct democracy." The quality of being "direct" refers, in Rousseau's Republican theory, to the refusal of a representative form of the Hobbesian sort: the united will of all citizens, the Genevan philosopher claims, cannot be represented, for it is either itself or something else. This explains why he argued that the popular will ought not to be articulated by elected deputies in a representative assembly but directly by all citizens in a popular assembly (Rustighi 2021).

In the views of the M5S, the platform was constructed precisely to enable private citizens to constantly participate in the party's decision-making processes instead of limiting themselves to just authorising its representatives through elections. Any mediation between the citizens and the government was supposed to be eliminated in this manner. However, theory and practice do not always align. Moreover, the ways in which members' participation was organised concretely matters. Technology is not neutral: a tool can have certain features that can enable particular uses and outcomes: the term affordances refers precisely to how a technology encourages users to perform certain activities (Dahlberg 2011). Analysing the affordances of *Rousseau* can help us to understand the conception of internal democracy and participation in the M5S.

In summer 2020, there were twelve main tools embedded in *Rousseau*. In the first place, members could search for other members (*Cerca altri iscritti*) and for spokespersons (*Cerca i portavoce*). Members and elected representatives could create a public profile and complete it with their CV and their "merits", that is to say, their experiences of participation and professional or education abilities. Merits could be used, on the occasion of online consultations, to screen candidates for public offices and for internal positions (on the "merit system", see Biancalana and Vittori 2021b). Another important tool, which I will analyse in detail in the next section, is *Vota* (Vote). On certain occasions, M5S's members were asked to have a say on some topics, ranging from the selection of candidates to the definition of electoral programmes, through online votes.

*Lex* was a part of Rousseau in which members could comment on the law proposals of M5S's elected representatives. *Lex* was launched in late 2013 and had a regional, national and European section, for regional, national and European law proposals respectively. Elected representatives that wanted to present a law in their assembly had to post it on *Lex* beforehand, together with a brief explanation and a video. Then, for a fixed period of time activists could comment on the law; afterwards, the elected representative had to integrate the comments into the law and present a written report in which he or she explained the added modifications and also why some comments hadn't been accepted. The elected representative was in charge of the whole procedure and was able to decide whether or not to accept comments. Moreover, comments are not "structured" (that is, in *Lex* there is not a discussion among members, but only untied comments temporally ordered), and in most cases the comments are incoherent with respect to the topic.

It has been noted (Deseriis 2017a) that the relationship created on *Lex* was an "asymmetrical relationship" between representatives and members, and that this function is designed to avoid the possibility of horizontal interaction among users, allowing only interactions between individuals and MPs. Studies on the evolution of participation using *Lex* (Mosca 2020) show that participation in the debate on law proposals decreased markedly over time: the average number of comments per law amounted to 446 in 2014, 184 in 2015, 144 in 2016, and 63 in 2017. Moreover, only 39 per cent of the law proposals received at least one answer by the MP in charge of it, and only in 15 per cent of the cases did they provide information on the outcome of the discussion (in 22 cases the proposals were modified, in 14 there was no change at all, in 13 the result is unclear).

Starting from May 2016, members could also propose a law with the function *Lex iscritti* (Lex Members). With the slogan *Con Lex le leggi le fai tu!*, that is, "With Lex you make your own laws!", this tool was advertised as a sort of DIY law-making. In reality, there was a set of steps that the law proposal had to undergo before being actually proposed. First, the proposed law had to fit certain requirements and it was evaluated by the "Rousseau staff". The drafts that passed this first evaluation<sup>60</sup> could be later voted on by members in occasional online consultations. The two most voted for draft laws in each consultation were assigned to an elected representative, who would then post the law on *Lex* and consequently bring it to parliament. Deseriis (2017a) defined *Lex* as a form of "direct parliamentarism", that is, as an extension of parliamentary processes to the web, in which the MPs retain the freedom to determine "which bills should be introduced into Parliament, which should be given priority, and how they should be initially drafted".

*E-Learning* was a section dedicated to online courses for elected representatives and lay citizens, for instance on the internal workings of Parliament

<sup>60</sup> Deseriis (2017a) reports that during the first four rounds of voting, M5S MPs received a total of 3,280 proposals, but only 523 of them, roughly 15 per cent of the total, were considered eligible to be voted for by members.

or on European issues. The M5S's elected representatives were, in the vast majority of cases, common citizens without political experience: in the absence of a party structure, this section was intended to serve as a tool of political socialisation. In the same way, in the tool named *Sharing* it was possible for elected representatives to share the activities in the councils or in Parliament, so that other councillors, MPs or citizens could know about them and take inspiration from the best proposals.

*Rousseau* also allowed citizens to create and find events or local mobilisations organised by elected representatives, other activists or the party. The function *Crea o partecipa a eventi* (Create or Participate in Events) consisted of three sub-sections: *Activism*, *Eventi dei portavoce* (Spokesperson's Events) and *Campagne elettorali* (Electoral Campaigns). In *Activism* it was possible for members to create a local version of some mobilisation campaigns proposed by the party. This function is interesting because it can be considered an evolution of the *Meet-up* experience. The platform *Meet-up* was used in the first period by M5S activists to organise themselves at the local level<sup>61</sup>, together with, in some cases, independent experiments such as Airesis, an open-source digital platform created spontaneously by some M5S activists in order to favour participatory but also deliberative experiences (Ceri and Veltri 2017, 176).

Through *Meet-up*, members had the opportunity to organise meeting and events, which nevertheless got out of the centre's control. In a process of increasing centralisation, these independent and open-source tools have been marginalised over time, and only participation through the official "operating system" was encouraged and recognised, even though—as we will see in the case of Turin—local groups also continued to organise themselves through infomal channels. *Eventi dei portavoce*, on the contrary, was a section in which the official events, created by elected representatives

<sup>61</sup> It is interesting to note that Meet-up in Italy is "monopolised" by the M5S. In 2014, Meet-up in Italy had 250,000 members (Italy was the second most represented European country after the UK), but 170,000 of these users were registered to groups that refer to Grillo (see Costa, F. *Il problema italiano di Meetup*, https:/ /www.ilpost.it/2014/10/16/problema-italiano-meetup/, October 16th, 2014). It is worth noting that in July 2015 a post appeared on Grillo's blog (*Meet-ups alone aren't the M5S*) stating that members of the *Meet-ups* weren't entitled to speak on behalf of the M5S. Only those that obtained the permission to use the party logo, that is, its certification, were able to do so. Moreover, since most groups changed the *Meet-up* name from "Friends of Beppe Grillo" to "Movimento 5 Stelle" the post urged *Meet-up* members not to use the party name. Even in the *Call to Action* section, it is underlined that the events collected on that page are not official M5S's initiatives. The official ones can be found in the section *Activism*.

of the M5S, were listed. In the section called *Campagne elettorali*, materials for organising and promoting the M5S's events and initiatives were made available to members.

Finally, *Open candidature* was the tool through which members could become candidates for public offices or apply for party positions, such as facilitators. *Segnala un iscritto* (Report a Member), was a section through which members were encouraged to report rule violations committed by other members to the party. In *Scudo della rete* (Net Shield), there was a list of lawyers that could help M5S elected representatives and activists for felonies related to the freedom of expression on the internet. Finally, like on most party websites, there was the facility to donate to the party.

The tools contained in *Rousseau* covered a vast array of functions. Online votes, as we will see in the next section, represented the direct link between members and the other two faces of the party's organisation. With *Lex* party members, as well as elected representatives and party leaders, could easily control the activities of their representatives, and potentially take part in the drafting of laws, while with *Lex iscritti*, members could make their own law proposals. *E-Learning* and *Sharing* were tools specifically conceived for representatives. The M5S's elected representatives are people without previous political experience: these tools were able to help them carry out their activities within assemblies. Finally, the tools included in the section *Crea o partecipa a eventi* were aimed at organising online and offline mobilisation events. *Rousseau* thus represents a very innovative experience of internal participation for a political party. Not only could members register online, but they could also comment on and propose laws, take online courses, monitor the activity of representatives and become aware of party mobilisation events, as well as have a say on some relevant issues, such as the selection of candidates and the determination of the party's policies.

However, from my analysis of the platform's affordances, we can note that members' participation through *Rousseau* is characterised by three main features: it is asymmetrical, individual and centralised. It is asymmetrical because it is designed to avoid the possibility of horizontal interaction among users, allowing only communications between individuals and MPs, like in the case of the function *Lex*. It is also this way because there is an imbalance of power in favour of the political leader, who has the power to call an online consultation, and in favour of the "staff" that managed the platform, who oversaw screening Lex members' proposals. Moreover, it is an individual kind of participation because the platform is designed for citizens to decide, vote and participate at the individual level: the absence of tools for discussion is an indicator of the atomised nature of this kind of participation. Finally, we can define *Rousseau* as centralised because, as we have seen in the case of the *Meet-up* experience, the leadership tried (and managed to) centralise all the party's various digital tools and online functions in one virtual place, which was strictly controlled from above.

#### *The Quantity and Quality of Participation Through Online Votes 4.*

One of the most relevant tools on *Rousseau* is *Vota*, through which M5S's members could have a say on some party decisions, such as the selection of candidates and the definition of electoral programmes. For the purposes of this analysis, all the online votes held from December 2012 to December 2020 have been collected, creating a complete database. Although the section *Vota* was only accessible by members, online votes were publicised on the party's website and, in most cases, turnout was also reported. Usually, online consultations were advertised on the party's website and announced at short notice with an e-mail sent to members, and last between eight and twenty-four hours.

With regard to the quantity of participation, 209 online consultations were held in the time span 2012–2020 (108 at the national level, 78 at the regional level and 23 at the local level)<sup>62</sup>. Overall, we see that the number of online consultations increased over time, reaching a peak in the year preceding the party's participation in government (2017), during which members participated in the drafting of the electoral programme (Figure 7.1). Due to the fact that, as we will see, online votes were also used to select candidates, the number of votes, especially at the regional level, also depended on the electoral competitions planned for each year.

As regards participation rate, if we compare the evolution of the M5S's membership with the evolution of the turnout of national online consultations, we see that, while members grew in number over time, the number of people that took part in online consultations remained more or less steady, with a decrease in the last period (Figure 7.2). So, online participa-

<sup>62</sup> The tally refers to voting sessions. In some cases (e.g. definition of the electoral programme), the M5S opened a single voting session in which more than one question was asked to members. In these cases, I counted the ballot as one. In the case of local and regional consultations, usually only the members that live in the city or region were allowed to vote.

tion did not seem to grow with the growth in party members. As we will see in the rest of the chapter, this is an interesting datum, and we can hypothesise that the fall in online participation is linked to members' dissatisfaction with the outcomes of the participatory process (see also Mosca 2020).

What are online consultations concretely used for? As the rules of the M5S have been incrementally defined over time, due to the need to regulate contingent issues or to adapt to specific normative requirements (Mosca 2020), in the past online votes have also been used for different purposes with respect to the rules on consultations written in the statute (the election and recall of political personnel; selection of candidates; approval of the political programme and of members' law proposals).

At the local level, in the vast majority of cases online consultations have been used to select candidates (mayors and/or the city councillors). Since 2018, no local online consultation has been held, as the process is managed through the tool *Open candidature*. However, even in the past, in most cases there were no online consultations to select candidates at the local level. Online consultations at this level have been used in two cases: if there were conflicts in a local group and therefore there were two different lists that wanted to run under the M5S's symbol and for big cities (e.g. Rome and Naples in 2016), where it was impossible or very difficult to leave the whole process to a local group<sup>63</sup>. In the first case, the vote was on the entire list; in the second, members could select single candidates. Where a local group was able to fill a list, no online consultation was needed; paperwork was sent to Casaleggio Associati's offices, where it was checked, and if all the documents were in order, the so-called certification was granted.

Even at the regional level, in the vast majority of cases, online consultations have been used to select regional councillors and regional candidate presidents. The regional level is interesting because it shows the institutionalisation of the method of online consultations to choose regional candidates. Since there are no official regional branches of the party, but only the elected representatives, it was difficult for members to meet and decide on candidates, like in the cities.

<sup>63</sup> One case that it is worth mentioning is that of Genoa. In 2017, Grillo cancelled the result of the online consultation regarding the choice of the mayor, because he did not trust the winning candidate. It is also worth noting that in some cases certification wasn't given to any list, without explanations, thus preventing local groups from participating in the elections.

*Evolution of the number of online consultations over time by territorial level Figure 7.1.*

*Evolution of the M5S's membership figures and average turnout in national online consultations Figure 7.2.*

Note: The mean for the year 2013 also includes the only consultation held in December 2012.

Nevertheless, until 2012, when there were fewer members and the organisation was easier, no regional online consultations were organised for the seven regional elections held in that period. In 2013, two out of seven regions held an online consultation to choose candidates, whereas in 2014 and 2015 all regions did it. Moreover, unlike in 2014, in 2015, all the procedure, starting from the proposal of the candidate, was online. Like in local consultations, Casaleggio Associati (and later Associazione Rousseau) was in charge of the whole process. So, at the local and regional level, we can see that online votes are mainly used in order to select candidates. As regards the national level, we can divide online consultations into ten categories (Table 7.1).

Even at the national level, the most important type of online consultation can be considered that regarding the selection of candidates. The first online vote ever was held in December 2012, in order to select candidates for the 2013 general elections (the so-called *Parlamentarie*), then a similar procedure was repeated in 2014 to select candidates for the European Parliament and on the occasion of the 2018 general elections and 2019 European elections. Essentially, these were online closed primaries: candidates for Parliament and the European Parliament were chosen by members in a completely online process. Both in 2012 and in 2014, there were limitations both to becoming a candidate (only members that had already tried to be elected with the M5S but weren't could run as MPs in 2012; in contrast, only seniority was required in 2014) and to voting—a requirement was to have been a member for a given period of time. In 2018 and 2019, as opposed to 2012 and 2014, *all* members could vote and be voted for. But in the case of the 2018 elections, the political leader of the M5S had a sort of "veto power" on candidacies and could directly decide on the candidates in the single-member constituencies (37 per cent of the seats according to the current electoral law); in the case of the 2019 European elections, he had the power to choose the first candidate on the electoral list in each constituency. In all cases, the rules of the competition were decided on by the leadership and released in a top-down way, and the process was managed first by Casaleggio Associati and later by Associazione Rousseau.


*National online consultations by category (2012–2020) Table 7.1.*

It is worth reiterating that in the M5S, at least until 2021, there has never been a consultation to decide on the party's leadership<sup>64</sup>. Beppe Grillo, as

<sup>64</sup> In August 2021, after the presentation of a new statute, Giuseppe Conte was elected as president of the M5S. The question: "Are you in favor of the election of Prof. Giuseppe Conte as President of the Movimento 5 Stelle?" was asked to members through an online vote on SkyVote (a voting platform managed by a private company that, differently from Associazione Rousseau, is completely detached from the M5S). Conte was elected as president of the party with a 92.8 per cent approval rate.

the founder of the party, has never been selected, and currently holds the position of guarantor indefinitely. In contrast, Luigi Di Maio was elected as the party's candidate MP for the 2018 general elections in September 2017. In this case, all elected representatives of the M5S were able to propose their candidacy. Nevertheless, the only prominent figure of the M5S that decided to run was Luigi Di Maio: he was, in fact, considered the M5S's "natural candidate", and 30,937 out of 37,442 participants voted for him. Following the 2017 statutory changes, Di Maio became political leader without a formal election: in the 2017 statute, which was never voted on or ratified by members, the positions that were supposed to be elective were already set. After the disappointing result of the 2019 European elections, Di Maio called for a recall vote. The result was a plebiscite: out of 56,127 voters, 44,849 voted "yes" to the question: "Do you *confirm* Di Maio as political leader?" (emphasis added).

In terms of electoral programmes, it is worth noting that on the occasion of the 2018 general elections the M5S organised online consultations on electoral programmes for the first time, despite rhetoric on the "programme written by citizens" that was already present in the 2013 general elections. In that case, no consultation on the electoral program occurred: the programme was released in a top-down way, assembling the themes and topics addressed by the blog in past years.

In contrast, when approaching the 2018 general elections, the M5S carried out a series of online consultations in order to define the electoral programme of the party on different topics, such as energy, work, public transportation, defence, tourism, school, health, banks and the environment. Each topic was broken up into different sub-themes; for each one, a post with an in-depth analysis was published on the blog. After the presentation of the topic had been completed, members had the opportunity to vote on each single sub-theme, on predetermined and closed options: the M5S's declared that the electoral programme would have been constructed according to the results of these consultations<sup>65</sup>. Questions included:

Choose the three priorities for a new model of economic development (members could select predetermined choices).

Do you think that the constraints on *our economic sovereignty* contained in the European treaties must be radically renegotiated, and that if

<sup>65</sup> An enquiry by the newspaper *Il Foglio* showed that after the elections the text of the electoral programme changed with respect to the first drafts, which were devised starting from members' answers in online votes.

negotiations lead to downward compromises, they must be rejected *for the good of the Italian people*? (Yes/No answer, emphasis added).

In the first case, members did not have the opportunity to add a new priority; in the second example, the question is framed in a way that makes the outcome obvious. In the same vein, in the case of the European elections of 2019, members had the opportunity to choose the "priority actions" that the future European MPs would carry out in Europe from a predetermined electoral programme.

With regard to political strategy, several votes fit into this category: whether or not to meet Matteo Renzi (February 2014); which European group to associate with (June 2014 and January 2017); approval of the coalition government (twice, in 2018 and 2019); the vote on an electoral alliance in Umbria (September 2019) and on local alliances with traditional parties (August 2020); and whether or not to participate in some regional elections in Emilia Romagna and Calabria (November 2019).

In the first case, Matteo Renzi, leader of the Partito Democratico and appointed by the President of the Republic to form a new government, proposed a meeting to the M5S. The thorny decision of whether to accept this request was left to the members, and Grillo and Casaleggio wrote that they didn't agree to meet Renzi; in contrast, members voted in favour of it. Grillo then agreed to go to the meeting but didn't take it seriously: because of Grillo's attitude, the meeting lasted a few minutes and there was no real discussion, so that the result was the same as not accepting the meeting.

In the second case, the result was practically predetermined. Indeed, the choice was between EFL, ECR and non-attached members, while some members and MEPs wanted to enter the Green group. This option wasn't pursued, and it is known that Grillo wanted an alliance with UKIP's leader Nigel Farage (EFL). In January 2017, with a new, unexpected online consultation, Grillo asked members to leave the current European group and to adhere to the ALDE group, a group that is very different compared to the M5S. The vast majority of members accepted the change, but after the failure of the negotiations the M5S remained in the EFDD group.

As regards alliances, the decision to approve the formation of the two national coalition governments in which the M5S participated (first with the Lega and then with the Partito Democratico) was also left to members<sup>66</sup>. Here it is interesting to compare the two questions presented to

<sup>66</sup> An online vote was organised also in the case of the participation of the M5S to the national unity government led by Draghi, in February 2021. The question

members on the occasion of the two consultations, which were held in June 2018 and September 2019. The first one refers to the alliance with the Lega. The name of the party, which was unpopular among a vast segment of the M5S's electorate, was not mentioned. Instead, the pretentious name given to the alliance ("government of change") was highlighted.

Do you approve of the "government of change"?

Do you agree that the M5S will start a government, together with the Partito Democratico, chaired by Giuseppe Conte?

In contrast, in the case of the alliance with the Partito Democratico, the name of the allied party was presented in the question, even though it is widely known that M5S electors have no sympathy for it. These different formulations reflect internal clashes within the party and the will of different factions to influence members' votes, that are essentially used as a tool of legitimation. In the first case, 94 per cent of members approved the alliance, while in the second case "only" 80 per cent did.

Finally, members were asked to vote on an electoral alliance at the local level (in the Umbria region in September 2019) to form local alliances with traditional parties (August 2020) and to decide, in November 2019, whether or not to participate in two regional elections, in order to prepare the party's congress and because of the weakness of the Movement and the fear of an electoral loss. In both cases, the leadership called for a consultation to legitimate their choices (for instance, the consultation on the alliance read "*Do you agree* with the proposal made by the political leader…?"), but in the second one, members decided to vote against the will of the leadership. Consequently, the M5S participated in regional elections in the two regions, suffering a heavy defeat.

Another category is the election of party personnel, and this happened several times: the election of the so-called *direttorio*, the election of the appeal committee and of the board of advisors, and the selection of national and regional facilitators and of the party treasurer. The first vote took place in November 2014. Grillo proposed 5 MPs in order to "help him" in managing the party, and members were asked to accept or not to accept

read as follows: "Do you agree that the M5S supports a technical-political government: that will provide for a super-Ministry of Ecological Transition and that will defend the main results achieved by the Movement, together with the other political forces indicated by the appointed president Mario Draghi?". 59.3 per cent of voters approved the decision. The observations made with respect to the other two consultations are valid also with respect to this one.

the 5-person closed list. Eventually, about 90 per cent of members voted in favour of it. The second consultation happened in December: members were asked to choose two out of three members of the appeal committee from a five-person list prepared by the leader, while one member was chosen directly by the leader. Finally, in November 2016, an online consultation was held to elect the board of advisors. Even in this case, there was a list composed of three people, to be approved or not approved. In 2018 and 2019, other votes to select single members of the board (for instance, after resignations) were held. In those cases, the leader proposed a closed list of names from which members could select the missing member.

In 2019 the M5S began an internal reorganisation, albeit one not codified in the party's statute, in which the positions of regional and national facilitators were created. A facilitator is a sort of coordinator of the party, and a transmission belt between the centre and the territories. Indeed, facilitators can operate at the regional or national level. While regional facilitators have the task of favouring the linkage between the centre and the territories, national level facilitators are responsible for the organisational aspects of the party (organisational facilitators) and for the articulation of policies (thematic facilitators). Facilitators were, of course, chosen in online votes.

In the case of the organisational facilitators, the choice left to the members was only to accept or refuse a closed list of six people decided on by the leader. In contrast, in the case of the thematic ones, members had the opportunity to propose themselves and to form a group, composed of other members (the opportunity to create a profile on *Rousseau* and to search for other members on the platform was created precisely on this occasion). Different groups and projects were subsequently voted on online. In the case of regional facilitators, members had the opportunity to propose themselves and vote, but in the end it was the leader that had the power to choose the facilitators from the three most voted for by members. Needless to say, the leadership was able to screen candidacies, and thus exclude candidates, at any time. Finally, in December 2020, the political leader proposed a name for the role of party treasurer, and members had the opportunity to accept or reject it. Eventually, it was accepted by a narrow margin (52.6 per cent).

As regards the definition of party documents, four votes have been held, including the change of the party symbol (November 2015), the change of the party statute and rules (2016), the votes on the new organisation of the Movement (July 2019), and the votes concluding the *Stati generali* (December 2020). In September 2016, a vote on the change of the party statute and rules was held. Indeed, it was the first time ever in which the M5S's members actually voted on party rules. Members were asked to approve or not approve a new version of the party statute; moreover, two versions of the party rules to choose from were proposed to them. We can see that it wasn't a bottom-up process, as the changes were decided on top-down: members had only to ratify them. And indeed, the vast majority (about 90 per cent in the case of the statute, 70 per cent in the case of the party rules) ratified the decision. Also, it was the first time that a consultation lasted for more than a few hours, as was usually the case: it lasted a whole month, in order to get as many members as possible to vote and to obtain legitimacy.

In July 2019, members had the opportunity to vote on the "new organisation" of the M5S. Five questions, ranging from the opportunity for municipal councillors to run for an additional mandate to the regional and national reorganisation that was to lead to the creation of the Future Team were proposed to members, to either accept or reject<sup>67</sup>. All five questions were accepted in the end, with different percentages of approval.

Finally, in October and November 2020, the members had the opportunity to discuss the organisation of the party, its principles and its political agenda in the first party congress, the so-called *Stati generali*, for the first time. Through a series of regional and national online meetings<sup>68</sup> , a document was written, and eventually 23 proposals were prepared by the political leader on the basis of that document. The 23 points—which included the request for the principle of collegiality to inform all the decision-making and governing organs of the party and the strengthening of the intermediate and territorial structure—were consequently approved by members with a vote on *Rousseau* and added to the statute. The statute was changed in February 2021 with an online vote on six points.

Additionally, between 2013 and 2015 online votes were used by the M5S for the expulsions of members, especially elected representatives. Since 2012, five expulsion votes have been held, involving nine MPs. In all these cases, members confirmed the expulsion that was proposed by the leader. Expulsion, besides signifying the withdrawal of the Movement's

<sup>67</sup> In the preceding months a specific area of Rousseau called *Area di ascolto* (Listening Area) was created to collect input from members regarding the new organisation. However, there was not a codified procedure through which this input could affect the leadership's decisions.

<sup>68</sup> The process, also due to the coronavirus pandemic, happened completely online. An online vote also concerned the choice of the 30 delegates that participated in the national meeting of the *Stati generali*.

certification and the permission to use its logo and implying the obligation to leave the parliamentary group, equates to the deletion of a member's account and consequently to the impossibility of them accessing *Rousseau*, voting in online consultations and becoming a candidate. Following the publication of the new statute (2017), it is now the board of advisors that decides on expulsions, without the need of an online vote. Several votes related to *Lex iscritti* (Lex Members). Occasionally, members' law proposals proposed through Lex Members were voted on by other members: the two most voted for proposals were brought in parliament to become M5S bills. To date, no law proposal has been presented in parliament. Finally, some votes also related to the allocation of party funds, derived by the cutting of M5S MPs' salaries.

Online votes have also been used in the past to directly influence the behaviour of M5S MPs in the assemblies. In recent years, this type of consultation has been used less and less. One explanation for this could be that control over the party in public office by the party on the ground has been more and more exercised through *Lex*. On the other hand, especially with the passing of the threshold of government, we have seen that the M5S is normalising and moving away from its original and radical ideas on digital direct democracy.

As regards this kind of votes, in two cases (2013 and 2015), members had the opportunity to select the M5S's candidate for the presidency of the Republic, who MPs had to vote for in parliament. Nine votes (held in 2014) related to the electoral law: an official M5S electoral law was constructed through nine consecutive votes. Nine questions were posed to members, starting from general questions (proportional or majoritarian system) to specific ones (electoral threshold, preference voting). It is relevant to say that all questions were "closed": that is, members could vote either for one or for the other solution, without proposing one themselves. Another consultation on the electoral law happened in June 2017. In this case, the question was also framed in a way that made the result predictable:

Are you in favour of the approval of a German electoral system which is respectful of the Constitution, possibly with the introduction of constitutionally legitimate governability corrections?

The other consultations related to specific issues, such as: the abrogation of the crime of clandestine immigration, metropolitan cities, the selection of the supreme council of the judiciary's judges, civil unions, the selection of a member of the constitutional court, the anticorruption law, living will and euthanasia, the selection of the members of the board of RAI (Italian public television), and the parliamentary vote on the incrimination of Salvini.

In January 2014 members were asked to decide whether or not MPs should vote on the abrogation of the crime of clandestine immigration. In fact, in October 2013 two MPs had proposed the abrogation, which was eventually voted on in a parliamentary committee. The leaders then wrote on the blog that the issue "wasn't in the electoral programme" and that nobody was consulted. But the problem, as outlined by them on the party's website, was more in the substance of the proposal than in the method used. Basically, they were openly against the abrogation of the clandestine immigration crime. In January, when the law had to be voted on in the Senate, an online consultation was held, and members voted in favour of abrogation, and thus against the leaders. Consequently, MPs voted in accordance with the online vote. This consultation is relevant since we can see that while practically it is the leader and not members that decides whether or not to hold a consultation, in this case the results were respected, even against the leader's will.

In October 2014, members were asked to vote on civil unions (vote in a parliamentary committee), and the vast majority voted in favour of them. In February 2016, the law had to be voted on in parliament, but Grillo and Casaleggio left the MPs free to vote, because the law now included the so-called stepchild adoption, which wasn't included in the past consultation. They chose not to repeat the online vote because "ethical issues" were at stake. This consultation is relevant because it shows clearly that the relevant choice of whether or not hold a consultation is taken by the leaders.

Finally, in February 2019, M5S MPs had to vote on Salvini's right to benefit from parliamentary immunity in a case regarding the disembarkation of some immigrants in an Italian harbour. The thorny decision of whether to concede defeat in the M5S's fundamental battle against politicians' privileges, or to help the government's ally, because following Salvini's incrimination the coalition government might fall, was left to members. Nevertheless, we can note that the formulation of the question (the reference to the "state interest") and the fact that a "yes" vote would counterintuitively mean denying the authorisation to proceed against Salvini made the question lean unambiguously towards "yes", which was eventually voted for by 59 per cent of the participants in the vote.

Did the delay in disembarking the *Diciotti* ship, to redistribute migrants to various European countries, happened to protect a state interest? (Yes: deny authorisation to proceed).

To sum up, online consultations for "directing" MPs have rarely been used. It seems that online consultations were used (or not used) to solve internal conflicts and to legitimise and ratify choices: to "pass the buck" to members. It is significant, then, that online consultations have been (or not been) held on divisive issues, clearly identified as left-wing or right-wing, such as immigration and civil rights.

#### *Conclusions 5.*

A cyber-utopian narrative has always been fundamental in shaping the M5S's rhetoric and practice: according to the founders, it is through the internet, and in particular through the party's websites and platform, that an unmediated relationship between the leader, elected representatives and members can be created. Therefore, in the M5S, at least at the national level, online participation replaced traditional forms of intra-party participation69. The most relevant example of the power held by the M5S's members is the candidate selection process: candidates at the local, regional and national levels are decided on by members, in most cases through online votes. Other than that, online consultations have been used to define the party's policies, strategies and internal organisation, as well as to select party personnel.

If it is true that this represents one of the most advanced experiences regarding the use of the internet by a political party, there are also some weaknesses that can be outlined. The first is related to accountability and transparency issues. *Rousseau* was created by the private company Casaleggio Associati, and it was managed through Associazione Rousseau by the son of one of the two founders, Davide Casaleggio; Associazione Rousseau is an association detached from the M5S that had, in line with the provisions of the 2017 statute, the right to manage all the online

<sup>69</sup> Except for the annual meeting *Italia a 5 stelle*, which cannot be considered a congress, in the M5S, there aren't national or regional congresses or other official occasions at or on which it is possible to take part in the life of the party at the national level. An exception could have been the *Stati generali*, the party congress that was scheduled for spring 2020 but that was held online in autumn because of the pandemic.

decision-making processes of the party, but members had no control over it, also because *Rousseau* doesn't employ open-source software (Deseriis 2017a). The end of the collaboration between Associazione Rousseau and Movimento 5 Stelle, in early 2021, can be considered both a further step towards the party's institutionalisation and a dismissal of the most radical promises of online direct democracy.

The second is related to the definition of the "rules of the game", which are always released in a top-down way and are not negotiable by members. This applies not only, for instance, to the rules for becoming a candidate and to the right to vote in consultations, but also to who decides to hold a consultation, its timing, and what the available options are. It is the leader who holds the power to call for a consultation, and it is not possible for members to request an initiative. Moreover, the way in which a consultation question is framed and the choices available to vote on are also relevant issues: all the M5S's consultations were held on predetermined options, and members could not add or propose new options.

Finally, in *Rousseau* there was no space within which members were able to discuss. On the M5S's platform, participation was intended as a means to decide, not to discuss or deliberate. The only place in which they could write comments was *Lex*, but the comments were untied and there was no real discussion among members, as the comments were addressed to the elected representative. To sum up, the M5S's individual members were given a lot of opportunities to have a say on the internal life of the organisation, but the M5S's party on the ground was an "atomised" one, which took shape during online consultations, always called for from above<sup>70</sup> .

Nevertheless, the M5S represents one of the most advanced experiences regarding the use of the internet by a political party, especially considering its electoral successes. Other experiences comparable to that of the M5S are the Spanish party Podemos and its online participation platform *Participa.podemos* (Pizarro and Labuske 2015) and Pirate Parties. Comparative studies of these parties' use of the internet (Deseriis 2020; Biancalana and Vittori 2021a) show similarities and differences between the three cases. For instance, in Podemos online participation does not substitute for offline participation, as in this party traditional congresses, albeit held with

<sup>70</sup> This is partially different at the local level, where face-to-face participation is more important. This is also indicated in our data from the fact that most online consultations happen at the national level, and that at the local and regional level they are used mostly to select candidates.

the help of digital tools, are still held. More generally, the organisational structure of Podemos is more party-like with respect to the M5S, and this is also mirrored in the management of the platform, which is run in a more transparent way through a team linked to the party and composed of party personnel. However, also in this case the platform serves only as a space of consultation, and the issues to be discussed are defined in advance by the party leadership. The case of the Pirate Parties, and especially the German one, seems to be more promising in terms of horizontality and members' empowerment. In this case, digital tools, and in particular the software LiquidFeedback, could potentially allow complex decision-making processes. However, internal conflicts and the lack of leadership have made the actual impact of online deliberation limited in this case too.

What was, then, the impact of *Rousseau* on the M5S's organisation? On the one hand M5S's members had relevant decision-making powers within the platform, and could directly decide, without formal intermediation, on many relevant issues. On the other, new forms of hierarchy and concentrations of power arose: the absence of an official party structure led, in the end, to a greater concentration of power in the hands of the leaders, or to anyone who holds the keys to the operating system. Members *formally* had a lot of power, but in the end the predominance of the leadership was never questioned. Due to the fact that the "rules of the game" were not shared and that members had neither control over the management of the platform, nor the power to call initiatives, and that the questions and available options were often framed to make members ratify decisions taken elsewhere, the outcomes of the consultations rarely contested the leadership. For these reasons, despite the deep innovation brought about by these tools with respect to the decision-making processes of traditional parties, we can say that members' power was mainly symbolic.

Members' participation resulted, in the end, in an *impression* of participating, which can have its most significant effects, for instance, in members' identification with the party or in its positioning *vis-à-vis* other parties in the public sphere, rather than in concrete decision-making processes (Biancalana and Vittori 2021a). This can be also considered one of the reasons behind the, on average, constantly low participation rate in online consultations. The role of the internet and of the platform *Rousseau* in the party's disintermediation strategies was thus mainly linked to a rhetoric or a narrative on overcoming traditional party structures and on citizens' empowerment. When it comes to practices, it seems that new and less accountable forms of intermediation were created. In particular, the role of Casaleggio Associati and Associazione Rousseau can be seen as a new powerful and unaccountable form of intermediation.

#### Not Just a Cyber-Party: The Movimento 5 Stelle in Turin 8.

#### *Introduction 1.*

Online participation has always been very important for the M5S. Although this partially changed over time, not only is the organisational structure of the party based on the internet, but also registration and, in general, most of the national decision-making processes happened online. Although we have seen that members' power exercised through digital tools is mainly symbolic, and new forms of hierarchy and concentrations of power arose, in the case of the M5S the internet allows and favours disintermediation strategies, that is, the creation of an unmediated relationship between leader and followers. We could expect, then, that in such a party a web-based unmediated relationship between leader and followers would replace traditionally defined party membership, even at the local level. In reality, this is only one side of the coin.

If, on the one hand, it is true that in the M5S the official party membership is only online, and that the official decision-making processes happen exclusively online, on the other hand participation in local physical groups of the M5S appeared to be highly developed (Biorcio 2015; Biorcio and Sampugnaro 2019). The local level is important for the M5S in many respects. In the first place, it is precisely at the local level that the M5S had its origins, through the autonomous formation and self-organisation of groups of active citizens, who later took part in local elections. In the second place, the local level is potentially characterised by a greater margin of independence for those groups. Until 2019, when the role of the regional facilitator was created<sup>71</sup>, according to the official discourse, the influence of the centre was limited to the management of the certification of the lists, without which it was impossible to take part in the elections under the M5S's symbol. This is the reason why the M5S is substantially different in every city where it is present.

For this reason, it is important to consider and analyse context. Turin and Piedmont have always been among the most important places for

<sup>71</sup> These changes haven't been codified in the party statute. In 2021, the new statute of the party mentioned and acknowledged the existence and role of "territorial groups" for the first time.

the M5S, and the party has been present in the region since the very beginning, both in connection with local mobilisation events and through an early presence in its institutions. For instance, in 2008, the second V-Day, a big gathering of the M5S's supporters, was held in Turin. The local group took part in the provincial elections in 2009, even before the official founding of the M5S. In 2010, two regional councillors were elected in the regional council, the first ones, together with those of the Emilia-Romagna region. In 2011, two city councillors were elected to the city council and finally in 2016 Chiara Appendino, a former city councillor, became mayor of the city. Moreover, in recent years, in Turin and Piedmont, and especially in Valle di Susa, a strong local conflict against the high-speed train line Torino-Lione (TAV) emerged. The M5S has always supported the movement against the construction of the line, and this has had consequences both on the profile of the first activists and on the results of local electoral competitions (Mosca 2015; Biancalana 2020).

The aim of this chapter is to analyse in-depth the history, the structure of the organisation and the forms of participation of the M5S in the city of Turin. Through this case study, the goal is to analyse how activist participation is structured in a party that, at the national level, employs internet-based disintermediation strategies. On the one hand, in such a national organisation, the role of organised membership could appear as pointless. On the other, it is true that parties' strategies can be different at the national and at the local level: I will thus observe how the M5S's members organised themselves and participated in a local context, and if there are differences with respect to the national one. Accordingly, in the first part of the chapter, I will reconstruct the history and evolution of the M5S in the city of Turin72. In the second part, I will focus on the structuring of the organisation and the forms of participation of the Turinese M5S, especially during the 2016 local electoral campaign.

<sup>72</sup> The main focus of the analysis is the city of Turin. Nevertheless, especially in the first part, we will also refer to the Piedmont region for two reasons. In the first place, Turin is the regional capital, where the Movement's regional offices are located; secondly, the election of two regional councillors in the regional council of Piedmont in 2010 had a great impact on both local and national dynamics.

#### *From the First Meet-Ups to the Election of the Mayor 2.*

Following the opening of Grillo's blog and its call to readers to meet using the platform *Meet-up*, groups of citizens in Turin sensitive to the issues handled by Grillo—called *Amici di Beppe Grillo*—began to gather. In Turin, some of these people were already active in existing social movements (e.g. the No TAV movement73, the movement against the privatisation of the public water system, the groups against an incinerator) and parties located on the left side of the political spectrum.

The first groups came from the movements. But it's normal to attract those people who already have a social and political awareness first they are the first to commit (Int. 8).

So, if at first the M5S in Turin attracted people belonging to existing social movements (Mosca 2015) not interested in taking part in elections and seizing power, this component (the "movement-like" one) seems to decline with the passing of the years and the growth of the organisation. According to interviewees, in 2016 for most activists the M5S was their first political experience and "very few people have previous political experiences" (Int. 1). Initially, the activities of the group consisted in making citizens aware of some issues, mainly related to the environment, and proposing alternatives to the institutions. However, at the local level and at the national one, the attempt to exert "external pressure", i.e. bringing the M5S's stances into the assemblies without entering them directly, seemed to fail.

<sup>73</sup> There is a strong proximity between the M5S in Turin and Piedmont and the No TAV movement, both from a programmatic point of view and regarding the biographies of the activists and electoral successes (Mosca 2015). At the 2013 general elections, the first seven municipalities in which the M5S obtained its best results were located in Piedmont. The first one is Venaus, a small town in the Val di Susa. This result can be attributed to the proximity of the M5S to the No TAV movement. It should also be noted that during all the public events of Turin's M5S (for example, the celebrations for the victory of Chiara Appendino in June 2016 and the closing of the electoral campaign for the "no" to the constitutional referendum, December 2016), numerous No TAV flags were waved, together with the flags of the Movement. After the passing of the threshold of government at the national level, the M5S abandoned some of its historical battles, including the No TAV one. Indeed, having been the first party to "represent" the No TAV movement in Parliament, it ended up as the party whose government gave the final go-ahead to the project in 2019 (Biancalana 2020).

At that time, as *Amici di Beppe Grillo* we made the request of an alternative plan to manage waste in the province of Turin to the provincial council. Obviously, our plan was rejected, the provincial council didn't even listen to it. Meanwhile, everyone followed their own issues: someone dealt with the Tav, someone with public water, someone with the incinerator… (Int. 9).

Between 2008 and 2009, a significant rift occurred in the group. At that time, Grillo and Casaleggio gave local groups the opportunity to compete in local elections with "certified" civic lists (*Liste civiche a 5 stelle*). In order to obtain "certification", that is, the permission to use the Movement's symbol, it was necessary to prove one possessed some requisites—for example, having no convictions and not being registered with other parties. The certified local lists could therefore compete with a unitary symbol and benefit from the visibility of Grillo's blog, but other than that they could organise themselves autonomously and horizontally. This opportunity created a conflict within the group between those who wanted to compete in elections and those interested only in "civic activism".

We realised that the movement *Amici di Beppe Grillo* was turning from a movement that brought proposals to other political parties into a political movement that competes in elections, so the first civic lists are born and in Turin the group splits into two parts: a part that wants to remain "movement-like" (*movimentista*), and a part that wants to compete in elections instead […] [so, two different entities are created:] a "movement-like" reality that faded away […] and the part that was more involved in the creation of the list (Int. 9).

In the end, the "faction" that wanted to compete in elections prevailed: the goals of the organisation were clearly defined, even though the ties with local social movements and mobilisation groups were not severed.

It was spring 2008. Actually, the M5S in Turin was already divided into two or three groups. The group I entered was the one that wanted to create a civic list for the local elections, ten people more or less. There was another group, maybe more numerous, that wanted to engage in civic activism only, even because they were part of other parties, at that time there were a lot of people from leftist parties. Then, obviously, those that weren't interested in competing in elections left the group two or three years later (Int. 7).

The first elections the group took part in were the provincial ones in spring 2009. The candidate mayor, chosen by other activists, was Vittorio Bertola, and the list got 7,423 votes: a modest result that testified to the early presence of a nucleus of activists and supporters. Following the principles of the Movement ("everyone counts as one"), the electoral campaign was carried out without any form of personalisation.

Bertola was disputed when he was the candidate for the provincial elections because "La Stampa" conducted a three-column interview, the only one in the entire electoral campaign about our list, and then put a stamp-sized photo of him in the article, and the critics went "That is personalisation, because you have the photo in the newspaper; this is a form of leadership, you had to put on the symbol of the M5S and say that you are only one among many, and one counts as one" (Int. 7).

This first phase testifies to the early presence of the M5S in Piedmont and Turin: as early as in 2008 there was a nucleus of activists that were to form the core of the M5S in the years to come. Initially, the M5S attracted people from social movements not necessarily interested in competing in elections. The decision, taken at a national level, to allow local groups to compete in elections created a rift in the group, and only those interested in electoral politics remained. This therefore led to significant consequences regarding the definition of the group's aims, which were clearly defined ("change politics, from the inside", Int. 5). The "movement-like" component therefore left the group, even if this didn't mean a clear break with the world of social movements.

The real turning point for the M5S in the region were the regional elections of 2010. In that year, just one year after the official foundation of the party, the M5S ran for elections for the first time in five regions, gaining four elected councillors in two regions: Emilia-Romagna and Piedmont. In Piedmont, the Movement got 3.6 per cent of the votes (17,217 votes in Turin and 90,086 votes in Piedmont) and had two councillors elected<sup>74</sup>. On the one hand, as regards organisation, in the absence of a formalised structure, we observe that the regional councillors became a reference point both at the national and at the local level. But this

<sup>74</sup> The presence of the M5S on the ballot probably influenced the electoral competition in the region. A large transfer of votes from the left-wing parties to the M5S was recorded in Valle di Susa, the region more affected by the high-speed line. The little margin of victory of the centre-right coalition (fewer than 10,000 votes) has led scholars to hypothesise that the M5S altered the outcome of those elections (Bobba and Seddone 2010).

didn't imply neglect of the M5S's original principles of horizontality. At that time, the M5S, at this territorial level, was open and horizontal, and candidates were chosen by other activists and requested activists' help to carry out their administrative activities. On the other hand, the election of two representatives is also important for the history of the group because the M5S started dealing with "real", "institutional" politics, as opposed to the previous "movement-like" politics. The funds given by the regional council allowed the creation of an office with some employees; some of them also became MPs in 2013.

At the regional elections, two regional councillors were elected, and they hired four or five aides to work with them, who were activists too. A first office was created and it was the occasion to begin to work more, to be more active also from an institutional and political viewpoint, and not only from a "movement" [*movimentista*] point of view, like it was with the participation in the movement for the referendum on public water, in the No TAV group, the No Incinerator group and all the other groups that we used to meet at the demonstrations… And we began to conduct activities, real politics, even of the institutional kind. We had a little office and then we could produce flyers, have a logistic base (Int. 8).

One year later, in 2011, at the elections for the mayor of the city of Turin, the M5S got 5 per cent of the votes (22,403 votes). In 2011, the M5S was then a stable political actor in the region (Bobba and Cilluffo 2015). The candidate mayor, chosen by the activists, was again Vittorio Bertola. And again, the electoral campaign was carried out without any form of personalisation.

When Bertola was candidate mayor, his name wasn't on the flyers; we put it in in the last two or three weeks because people asked us, "You're with Fassino?" [the candidate mayor of the centre-left, then elected]; they believed that we were part of some coalition. Now you've seen the electoral campaign of Chiara Appendino, with the big billboards with her face on them! (Int. 7).

In 2011, Vittorio Bertola and Chiara Appendino were elected to the city council. Moreover, one M5S district councillor was elected in each of the ten district councils of Turin. This is important because it allowed the strengthening of the organisation and the diffusion of the Movement in all the neighbourhoods of the city.

The group grows, in 2011 we manage to put, and that's a very important thing, a person, I believe this is fundamental for the M5S, a councillor in each district council and each one of these councillors agglomerates… begins to organise activities with stands [*banchetti*], which are our preferred means of meeting and discussion with citizens, and so every Saturday or Sunday we organise these stands, but every time there is something to do: the referendum on public water, a campaign for some other reason, the No TAV battle… (Int. 8).

So, we can say that after 2010 the group began to structure itself and to engage in "institutional" politics, without losing its characteristics of horizontality and contact with the territory and the social movements. In a horizontal organisation such as the M5S, which does not have a traditional and structured party organisation, an important role, albeit an informal one, was that of the elected representatives, especially those at the higher levels than the local one. As mentioned, elected representatives are the only ones who have the right to use the party symbol<sup>75</sup> and, although they placed themselves in an equal relationship with activists, they became coordinators and points of reference both for the local level and for the national level, creating forms of informal leadership.

It is especially in the run-up to the 2016 mayoral electoral campaign that we witness the creation and strengthening of informal leadership configurations. Indeed, in June 2016, after having been chosen by other activists as candidate mayor in an offline meeting in which she was the only candidate, Chiara Appendino beat Piero Fassino, the incumbent mayor and leading member of the Partito Democratico, and became mayor of Turin (Biancalana 2019). In a situation in which there is not a formalised party in central office, or official local and regional branches of the party, the elected representatives (who are selected directly by other activists) have a relevant role, not only in terms of their political activity but also, for example, in the management of disputes and, more generally, in terms of the control of the organisation. With regard to the organisation of the 2016 electoral campaign, interviewees recall that in the absence of a party structure that manages it, the candidate mayor and her trusted people had an important role both in organisational and in policy decisions.

<sup>75</sup> Until 2021, local groups didn't represent official branches of the party and had no place in the party model. In 2021, the new statute of the party mentions and acknowledges the existence and role of "territorial groups" for the first time.

During the first meeting after Chiara Appendino was chosen as candidate mayor, she told the assembly, "The people that will collaborate with me are X and Y, we're taking decisions because there is no time to involve everyone; it's too complicated, so we're deciding". Someone told her, "But what's the meaning of this, we have to share, all together!" and she goes, "No, I'm the candidate mayor and these are my choices" (Int. 7).

The workgroup on transportation decided, years ago, to be against the second line of the subway […]. They wrote so in the draft of the electoral programme and they sent it to Appendino. Then they came to the public presentation of the electoral programme and they discovered from the slides that the M5S is in favour of the second line of the subway! They asked, "What happened?" and Appendino told them that "politically, we can't be against the second line of the subway because the people want it and so we have to be in favour for reasons of political communication, and then we will see". And they said, "But what about us, do we count?" (Int. 7).

Furthermore, as far as the communicative aspects of the electoral campaign are concerned, we notice an important shift towards personalisation, as Appendino's 2016 electoral campaign was strongly related to her name (*L'alternativa è Chiara,* "The Alternative is Clear" is the electoral campaign's slogan) and to her face, which was pictured in posters.

Beppe Grillo has always said "No to big faces" [on posters], and in fact the 2010 campaign had been conducted entirely without faces, but even in the 2011 one, there was no person on the posters. This has gradually changed; in fact, now Chiara Appendino is conducting an image campaign, her face is everywhere (Int. 1).

Finally, after the electoral campaign and the election of Appendino as mayor, we witness a shift from the importance of activists' participation to accountability. The activists' assembly, which as we will see in the next section was the decision-making body of the local M5S, became "of accountability" because "the decisions are taken by the city council now" (Int. 9). And while for some "it would be unimaginable if the activists' assembly gave directions to the city council assembly", for others this posed a problem with respect to the M5S's principles.

Under this new management, activists suddenly disappear. From the day after the elections, yes, some meeting was held, but Chiara Appendino takes all the political decisions, at most with the council, at most with the city councillors, but the city councillors don't listen to the activists and so all the others say, "We worked for five years, we devised propositions, ideas, etc., we wanted to fulfil them and now we can't even speak internally" (Int. 7).

Indeed, from interviews and later accounts (Biancalana 2019) it emerges that, after the Movement crossed the threshold of city government, elected representatives became increasingly detached from activists. During Appendino's mandate, activists had greater opportunities to influence the M5S at lower levels than at the level of the city, that is, in the city's districts, where the M5S was in opposition and "there are five councillors and 20 activists". (Int. 9)

#### *Organisation and Participation in Turin's M5S (2011–2016) 3.*

Between 2011 and 2016 the M5S thus structured and organised itself, entering institutions without losing its characteristics of horizontality and openness, or contact with citizens and the territories. The M5S does not display a structure similar to that of traditional parties. Its only official relationship with the centre is the certification of the lists, that is, the authorisation to the use of the symbol, given to the candidates who meet certain requirements<sup>76</sup>. So, local groups organise themselves autonomously<sup>77</sup> , and elected representatives (chosen by local activists) act independently on issues related to their territorial level. However, since local groups are not considered official branches of the party, they have no influence on the national level, as participation in the national decision-making processes is individual and took place through the online participation platform *Rousseau*.

With regard to the M5S in Turin, in the period 2011–2016 there were two levels of organisation: the city level and the district level. At the city

<sup>76</sup> The figure of the regional facilitator, created in 2019, was conceived to create a tighter link between the centre and the territories. Nevertheless, these changes have not been codified in official party documents.

<sup>77</sup> The level of autonomy varies from case to case. Turin seems a case in which control from the centre is lower. For instance, in Turin the candidate mayor and the city councillors didn't have to sign a *Code of Conduct* that obliges them to pay 150,000 euros in case of "image damage" to the Movement (like in the case of Rome); the candidate mayor and the councillors were chosen in offline meetings, over which the "centre" has no control.

level, we can find the activists' assembly and the workgroups. The activists' assembly gathered all the "activists" in the city and was the decision-making body of the local M5S. At the local level, in contrast to the national one, there isn't an online enrolment procedure. As we have seen, it is only possible to become an official member of the M5S online. According to the 2017 statute (article 3b), "members" have the right to: contribute to the definition of the political direction of the elected representatives; take part in online consultations; become a candidate; and make legislative proposals.

The local M5S didn't know the identities of the "official members" of the party, that is, those registered online, so at this level those who take part in the M5S's face-to-face activities were considered "activists", and then had the right to take decisions at the local level. In Turin, activists were counted and coordinated through a closed and secret Facebook group that, in March 2016, had approximately 300 members.

We have a closed (hidden) Facebook group of activists from Turin; we drafted a regulation and if you want to take part in our activities—we always start from the territory—and you want to participate in the M5S, you have to write to us. "In which district do you live?" […] "Well, you're in district X, so contact this person who is the district councillor" and you start to take part in district X. So, if a person is really interested, he/she will contact the district councillor, who will tell him/her: "We have a stand this Saturday; we have to do this and that," even simple stuff like going and getting the materials, being there and so on. Or "we need you to take part in this committee in the district council". You come, you do it, after two or three times of showing up, you can be part of the closed Facebook group, you can come to our city meetings, you can vote and so on (Int. 1).

This marks a huge difference between participation at the national and at the local level. If at the national level *members'* participation is online only, at the local level *activists'* face-to-face participation is considered fundamental and allows members to take part in the local decision-making processes.

For us, activism is something very tangible, very material; I am sure that there are more people registered with the M5S than activists (Int. 1).

Due to the limited number of people participating at the local level and their digital literacy ("We are not digital champions," says an activist, Int. 1), the preferred means of interaction and participation was the offline one. Against the cyber-optimistic rhetoric of Grillo and Casaleggio, it seems that there is, at the local level, a very realistic vision of digital democracy.

When you are in the Municipality, or in Parliament, it is impossible to consult the internet for everything. At some point, you will have to make decisions and you have to be accountable for them; that's absolutely legitimate, there needn't be separation between the base, the territory and elected representatives. However, we cannot even think of asking for opinions on everything. Actually, this issue, we got over it (Int. 1).

Nevertheless, in the absence of other forms of organisation, digital tools were widely used by the groups as coordination tools. In addition to the closed and secret Facebook group, which acts as a replacement for the members' register, and to the various mailing lists and Facebook pages of each district group, WhatsApp chats were highly developed and, in some cases, they constituted a sort of permanent assembly: a continuous form of participation, parallel to the face-to-face one<sup>78</sup> .

Among the rights given to activists, there was the selection of local candidates. In Turin, at the city and district levels, this has always been done with offline consultations. All activists could vote and be voted for. To become a candidate, it was necessary to be an "activist", that is to say, to participate in the face-to-face activities and to be officially enrolled in the M5S. This is the only occasion, at this level, on which it was necessary to be an official member of the M5S. In fact, the two forms of belonging, that is, membership and activism, overlapped only with regard to the candidates' selection process79. To use the words of an activist, as regards the candidates' selection process: "without registration it is like not being an activist".

<sup>78</sup> For example, at the end of an open meeting, it was necessary to decide on some issues, such as the date of the next meeting or where to organise a stand. Activists, instead of discussing the topic in the assembly, decided to "talk about it in chat", thus excluding those who were not part of that WhatsApp group chat. On another occasion, during a meeting of the district council and faced with a thorny decision, the councillors wrote in the chat to get directions and advice from other activists.

<sup>79</sup> Although becoming an activist requires a greater effort than joining the M5S online, we can say that the boundaries of the organisation are quite open. In contrast, to become a candidate, a concrete and continuous commitment to the organisation, witnessed by other activists, is required.

The candidate has two levels of requests. The national request: you have to be registered for the blog, you mustn't have a criminal record and so on […]. On the other hand, you have to be an activist, a person that did something […]. your group must agree, they know you and vote for you. You have to come forward when candidacies open; we draft a list of people that want to become a candidate and then we vote (Int. 3).

The relationship with the elected representatives was equal and horizontal. Indeed, the second layer of the organisation at the city level were the workgroups. The workgroups were groups of activists gathered according to areas of interest that helped and supported the elected representative in his or her administrative activity and were in charge of writing the electoral programme. Furthermore, as I said, in each district there was an M5S district group. It is at this level that the relationship with citizens was stronger. The groups at this level carried out mainly three activities: weekly stands, meetings and support for the elected representative. Stands (*banchetti*) were the trademark of the M5S. Stands were organised on a weekly basis, both during electoral campaigns<sup>80</sup> and during other periods, and embodied the link of the M5S with citizens and its presence in the territory. District group meetings could be open to citizens or restricted to activists and were one of the ways to recruit new activists.

If you come to our meetings, I won't ask you if you have already registered. So, in reality the meetings are always open to everyone, and we do not have a "register" of the people who reside in the city and are officially members of the M5S. There are some people who have maybe taken part for months and then they tell us; maybe it happened when we drew up the lists for the elections; and to become a candidate, it is necessary to be registered: "But I'm not registered yet" (Int 4).

Even at the district level, the activity of supporting the elected representatives was perceived as very important. In the eyes of the M5S, an elected representative is seen only as a spokesperson, a lay citizen sent into institutions to bring citizens' demands to them.

<sup>80</sup> In the period 2010–2016, the M5S took part in five elections: the 2010 regional elections, 2011 local elections, 2013 general elections, and the 2014 regional and European elections. In 2014 the M5S got 93,806 votes in Turin. The candidate president was again Davide Bono, and the M5S had eight regional councillors elected.

What I always say is that I am alone, but the branches, the roots around me are all the people around me, so I cannot be in the whole district at the same time, but all the activists live here, and then they can report, collect and write documents (Int. 4).

Finally, stands represented the fundamental activity of the district groups. They were held every week, whether or not during an election campaign, and represented the way in which the Movement kept contact with citizens, as well as a way to recruit new activists. Through the booths, in fact, it was possible to listen to citizens' demands and to communicate the activities carried out by the M5S at the various territorial levels. This activity was considered to be the most important by activists, a real trademark to be proud of ("Where are the others?" activists note during non-electoral periods when they are the only ones in the street).

#### *Activists' Mobilisation During the 2016 Electoral Campaign 4.*

In June 2016, in the second round of the local elections, Chiara Appendino beat Piero Fassino, the incumbent mayor and leading member of the Partito Democratico, by a margin of 9 percentage points (54.6 per cent vs. 45.4 per cent<sup>81</sup>). Appendino's victory interrupted a long government cycle: in the previous twenty-five years, centre-left governments had guided the transition of Turin from an industrial and Fordist city (Bagnasco 1990) to a new model of development based on culture, tourism and knowledge. Landmarks in this respect are the 2006 Winter Olympic Games, promoted by mayor Chiamparino. In the public narrative, the 2006 Winter Olympics symbolised this change in the city, while for the oppositions (and, in particular, for the M5S) the Games represented the cause of the city's huge debt, and then one of the reasons for the inequalities that affected the city. As noted by Berta (2016), perhaps in 2016 a part of the city that, like other Italian cities, had been hit by the economic crisis and is characterised by a high rate of youth unemployment, failed to recognise itself in this narrative of change and decided to vote for renewal.

Indeed, the two main opponents, Fassino and Appendino, plastically symbolised continuity and change, respectively. While Fassino relied

<sup>81</sup> In the first round, Fassino got 41.8 per cent of the votes and Appendino 30.9 per cent. From the first to the second round, Fassino gained only 8,000 votes, while Appendino gained more than 84,000. Analysis of electoral flows (Cepernich and Vignati 2016) showed that most of these votes came from the centre-right.

on his experience as a public administrator and politician, Appendino pledged radical discontinuity with respect to the previous administrations. Moreover, Appendino focused on the inequalities that affected the city, and in particular on the existence of "two cities"<sup>82</sup>: the centre, the main character of the narrative of change and the setting of the cultural events that symbolised it, and the disadvantaged peripheries, characterised by poverty and negligence (Cepernich and Vignati 2016). Her slogan "The Alternative Is Clear" (*L'alternativa è Chiara*) summarised her promises. And, at first sight, the electoral results seemed to confirm this centre–periphery cleavage, with the central neighbourhoods voting for Fassino and the peripheral ones, once fortresses of the left, for Appendino<sup>83</sup> .

But, although she was the "outsider candidate", we must say that Appendino was not a complete outsider. The former city councillor chosen by other activists as candidate mayor had a background partially different to that of most M5S activists, or rather different from the stereotype of the M5S activist. Aged 31 years at the time of the election, having graduated in economics from a prestigious university in Milan, she is a young woman from the upper middle class of Turin. Her father is close to the industrial *milieu* and her husband is a local entrepreneur. Moreover, Appendino surrounded herself with collaborators who were not foreign to local politics. To sum up, her profile doesn't seem to be grounded in a "movement-like" background.

How did activists' mobilisation unfold during the electoral campaign? The M5S is an organisation born in a top-down way, in which the leaders

<sup>82</sup> In a video posted on Facebook during the electoral campaign, Appendino stated that "I live in a city divided in two: one is that of the queues in front of the museums, the other is that of the queues in front of the soup kitchens". Besides this, the electoral programme written by the workgroups' activists included several different themes. As the M5S has always paid attention to matters related to the environment and sustainable development, the first is the environment. As regards city planning, the M5S committed itself to stopping the consumption of soil, and in particular the construction of new shopping centres, to improve the separate waste collection, to ensure the rights of animals and to defend public water. The second relates to the organisation of the municipality; here, the M5S committed itself to reorganising and cutting the costs of municipal administration, but also to improving the transparency of administration and empowering citizens' participation.

<sup>83</sup> In reality, it has been shown that the picture is more nuanced, as in reality peripheries are multiple and diversified and the support for the outsider candidate gradually increased in the passage from the centre to the peripheries (Cepernich, Cittadino and Pellegrino 2018).

hold considerable decision-making power, and in which the rules and the most important decisions are not shared. We can thus expect that members' participation on the ground is not very developed and is controlled from above. However, we have also seen that, at the local level, groups potentially organise themselves autonomously and that, at least in the case of Turin, are not strictly influenced by the centre. Despite the presence of a "restricted group", guided by the candidate mayor, that has the task of managing the campaign, reports of the campaign tell that, at a lower level, activists' mobilisation was spontaneous and that the organisation benefitted from their constant presence in the streets and in the neighbourhoods, which we have seen is a sort of trademark of the M5S.

I participated in a meeting that was called to "organise the participation" of the people who have made themselves available to help Chiara Appendino in the campaign. The room is not very large and when I arrive it is overcrowded. The meeting seems spontaneous and unorganised. This is, I think, one of the contradictions of the M5S: at the local level it is really "bottom-up", and yet it was created and is organised "from above". It is a sort of franchise of the auto-organisation. When activists start to talk, at the beginning of the meeting, they say that five years ago they, those who now speak, were "on the other side", and that everyone in the M5S is a volunteer. They say that it is necessary to talk to people, starting with acquaintances, but in reality, there is no discontinuity with what they do in other periods, because they are on the streets in other periods too. An activist carries a yellow bag with some stickers of the M5S and of Chiara Appendino attached to it. Speaking of the sticker, he says that before, five years ago, when wearing the sticker there was the risk of being insulted, but now, on the contrary, people approach and congratulate them when they see it. To carry the M5S sticker is a matter of pride (Fieldwork note, 9/4/2016).

The M5S is used to being present in the neighbourhoods, for instance with weekly stands. Stands were organised on a weekly basis, and there is not a huge difference between the electoral and non-electoral period, except for the higher activists' participation. Participation is characterised by enthusiasm on the part of M5S's activists in conducting campaign activities.

I am at a stand. The atmosphere is pleasant. I can see that they know each other, that they are happy to stay at the stand; they do it with passion, they do not consider it as burden. They tell me that they have been doing this for three years, every week (Fieldwork note, 23/4/2016).

At the city level, the electoral campaign also consisted of big rallies with national MPs, but at the district level, district groups organised small events autonomously. On these occasions, wearing the symbol of the M5S was considered a source of pride for activists.

There is a lot of enthusiasm among them, and I see M5S logos everywhere: candidates have t-shirts emblazoned with the symbol of the M5S and their name on it (as if they were a team) and stickers; they stick stickers everywhere. I also notice a flag on a stroller, and a lighter and a smartphone cover with the logo of the campaign (*L'alternativa è Chiara*) on it. Supporters bring pins or t-shirts... They are very happy to bring the M5S's symbol (or Chiara's), so much as to cover anything with it (Fieldwork note, 15/5/2016).

To sum up, whether in electoral campaigns or not, the image of the local M5S as a "cyber-party" appears to be inaccurate. In the absence of a formalised party structure, the internet was used for coordination purposes, and offline participation and activists' mobilisation were relevant. Moreover, the strong identification with the party could seem unexpected for an organisation with such a short history and no roots with the past. However, it seems that the M5S managed to recreate some characteristics of mass parties, despite being born to contrast with traditional party organisations.

#### *Conclusions 5.*

The aim of this chapter was to analyse how the M5S organised itself at the local level and how activists' participation unfolded in a party that, at the national level, employs internet-based disintermediation strategies. Indeed, at the national level the most relevant strategy of disintermediation involves the use of the internet to substitute for the party's organisation: through *Rousseau*, members could decide directly on political and party issues, even though the IT infrastructure, as well as the national organisation of the party, are strictly controlled from above.

In such a national organisation, the role of organised membership on the ground could appear pointless. However, in the case and in the period considered, the M5S was very grounded in the territory, and offline participation was strongly developed. Here, the M5S was not just a cyber-party. At this level, the cyber-optimist rhetoric that characterises the national level was absent: the internet was used only as a coordination tool; for instance, in the absence of a formalised party structure, it replaced some functions, such as the members' register. In contrast, face-to-face participation and commitment to the organisation were perceived by activists as very important. This is also confirmed by other research: as regards participation practices, it has been noted that "the consensus-oriented deliberative practices that drive activism at a local level seem almost at odds with the direct democracy model based on preference aggregation engendered by *Rousseau*" (Deseriis 2020, 1782).

In the period considered, the M5S was indeed grounded in the territories with a constant presence in the streets. This together with the strong identification with the party symbol are characteristics that, unexpectedly, recall some features of mass parties. Writing about "digital parties", a category that would also include the M5S, Gerbaudo wrote that the digital party "comes close to the mass party in its ambition of reconstructing a culture of mass political participation by adapting it to the individualized experience of the digital society" (Gerbaudo 2019, 178). But a relevant difference between the M5S and mass parties is that local groups, however fundamental at the local level, were not—at least until 2021—officially recognised by the national one<sup>84</sup>, and that participation at the national level is individual and takes place through online consultations. Moreover, in contrast to traditional or mass parties, there is not a local party structure at this level either: activist self-organisation replaces it.

At the local level, the M5S thus responded to the "crisis of parties" creating an open and horizontal organisation, in which there is no official party structure. We have seen that activists organized themselves autonomously and besides the relevant question of the certification of the list, without which it is impossible to participate in elections, created bottom-up local groups and elected representatives. At least initially, the M5S was an informal and grassroots organisation, in which there were no formal hierarchies and everyone was considered equally, thus recalling some traits of social movements. If then, at first, the dimension that seemed to prevail in the case analysed is the empowerment of activists, we see that, especially with the passing of the government threshold, informal leadership emerged.

<sup>84</sup> In the 2021 statute, territorial groups are officially recognised for the first time (article 6). It is also worth mentioning article 25c, according to which "Local groups and territorial formations self-constituted over time or, in any case, already in effect are dissolved, starting from the approval of this statute".

This trend has been confirmed by later research: with a dynamic quite similar to that of the national level, and coherently with expectations on the institutionalisation of new parties, during her mandate the mayor gained power and became more independent with respect to councillors, who, in turn, failed to act as transmission belts with activists, who consequently failed to intervene in decision-making at the city level (Biancalana 2019). Thus, although no official party structure was created, just like at the national level elected representatives played an increasingly important role in the organisation to the detriment of activists. A question that is worth asking, and that will be subject of future research, is whether activists' demobilization played a role, among other factors, in the electoral defeat of the Turinese M5S at the 2021 municipal elections<sup>85</sup> .

<sup>85</sup> At the 2021 municipal elections, held in October, the local M5S obtained 9.1 per cent of the votes. The party's candidate mayor was Valentina Sganga. The centreleft coalition, led by Stefano Lorusso (PD), won the elections in the second round agains the centre-right one.

#### Party Change in the National, Virtual and Local Arenas 9.

The aim of this study was to answer three main questions. The first one concerned how parties adapt and change in response to certain social and political changes, and in particular to a context characterised by the perception of a refusal of intermediate bodies, especially with regard to their internal organisation. I hypothesised that disintermediation strategies, defined as *rhetoric or practices developed by parties in order to stage or deliver an unmediated relationship between leader and followers, which happens with the weakening of the party's intermediate organisation* are parties' answers to such changes.

The second question involved the emergence of new forms of intermediation—either in the form of the rise of new intermediaries or in the form of old intermediaries' attempts to avoid disintermediation—following parties' disintermediation strategies. Starting from the idea that disintermediation implies a *transformation* of intermediaries rather than their removal, I expected to find both new forms of intermediation and the persistence of old ones.

Finally, my aim was to assess whether different parties, and in particular "old" (traditional or mainstream) parties and "new" parties, adapt in different ways. In a context in which mainstream parties are accused of being disconnected from society, I expected them to open up their decision-making processes in order to gain legitimation. In particular, I assumed that they used rhetoric characterised by disintermediation from below and practices marked by disintermediation from above, and that they could try to avoid disintermediation through reintermediation strategies. Starting from different assumptions, also with regard to new parties I expected that, in their rhetoric, they enhanced disintermediation from below, but that, over time, they could increasingly put into practice disintermediation practices from above. In this chapter, I will summarise my findings and outline similarities and differences between the two parties in the three "arenas" considered in this work: the national, the virtual and the local one.

#### *The National Arena 1.*

The first arena that I considered is the national organisation. In chapter 3, we saw that the Partito Democratico presents a party model based on the promise of opening decision-making processes not only to members but also to voters. In particular, voters can participate in the direct election of the party secretary, which occurs through an open primary. This is an almost unique innovation in Europe. Indeed, Italy is one of the two European countries that use open primaries for the selection of the party leader, the other being Greece, where Pasok uses such a selection method. In terms of elective offices, it is worth remembering that, since 2012, the French Socialist Party has used open primaries for the selection of the candidate for President of the Republic (Lefebvre and Treille 2016). This decision ignited a debate similar, in part, to the Italian one (for a comparison between Italy and France, see De Luca and Venturino 2015; Giannetti and Lefebvre 2015).

On paper, the PD party statute offers voters many ways to influence the internal decision-making of the party, but in practice the opening of decision-making processes is limited to the direct election of the party secretary. We have also seen that, through the mechanism of primaries, the party leader would gain strength both outside (for instance, in terms of personalisation of the leadership) and inside the party, given that the intermediate bodies of the party are composed starting from the voters' vote in the primaries, and that members are marginalised in favour of voters.

Therefore, since the decision-making power of voters is limited to voting in primaries and there are no accountability mechanisms from one election to another, from the analysis of party documents we can conclude that in the case of the PD disintermediation results in an increase in power for the leader who, through primaries, can increase the scope of their power and autonomy *vis-à-vis* the party's organisation (see also Seddone, Sandri and Sozzi 2020). But, although it has unusual and innovative aspects, such as the opening of decision-making processes to voters and the composition of its internal organs starting from the votes of supporters, we have seen that the organisational form of the PD can be equated with that of traditional parties. For instance, we don't witness a decrease in the number of party layers between the highest executive body and the party congress, which, according to Pizzimenti, Calossi and Cicchi (2020), is an indicator of disintermediation.

In the case of the PD, more than the emergence of new forms of intermediation, it is possible to find the permanence of those typical of traditional parties. For instance, the analysis of the PD's organisational history tells us that primaries are managed and controlled by the dominant coalition, which uses them to gain legitimacy outside or to move challenges within the party. It is perhaps for this reason that, in the end, the Partito Democratico failed to deliver on its promise of strong leadership. Except for Renzi's experience, which ended up being absorbed and normalised, we see that—and the resignation of Zingaretti at the beginning of 2021 testifies to this—what the PD lacks is precisely strong leadership.

In the Movimento 5 Stelle, the innovation with respect to traditional party models is even more radical, as the web replaces the classic party structure: the party, at least initially, was composed only of the leader, members and elected representatives—who present themselves as ordinary citizens that become politicians for a limited time span—connected through the internet. In the case of the Movimento 5 Stelle, ordinary members (and not voters) had the opportunity to directly choose their candidates and, potentially, also to direct their activities, through the online participation platform *Rousseau*.

However, we have seen that this is not a genuinely horizontal organisation: the structure in which members can exercise their powers is highly centralised, with the leader maintaining important prerogatives in the organisation. Moreover, over time elected representatives and other personalities took on relevant formal and informal roles in the organisation, and an intermediate structure started to arise. In the case of the M5S, and we will see this clearly in the part dedicated to the virtual arena, we can state that disintermediation strategies led to the emergence of new forms of intermediation, mainly linked to the use of the internet and to the strengthening of the role of personal leadership in the organisation, but also to the unfolding of the institutionalisation process.

The two parties present both similarities and differences. In terms of the from-above dimension, this prevails in the practices of both parties. In both cases it is the leadership that gains power when the intermediate organisation weakens. From-below disintermediation, i.e. the greater decision-making power promised to the base, is instead predominantly a rhetorical or a symbolic aspect. But, on the other hand, it is true that the two parties arrived at the same result through different paths. We can outline two main differences between the two cases. The first one concerns the intermediaries involved in disintermediation strategies. In the PD, disintermediation results more in the bypassing of *party members*, who are marginalised in favour of voters, while in the case of the M5S it results more in the weakening of the *middle-level elite*, who initially is practically non-existent and takes shape over time with modalities and forms partially different with respect to those of traditional parties.

The second one relates to the diachronic evolution of the two parties: in the PD, despite the modification of the statute, we can't find formal changes to the party's organisation over time. In a context of accelerating technological change and of a deep crisis of representation, Renzi exploited the opportunities that were already given by the open party model more than previous party leaders, increasing the personalisation of its leadership, creating a direct link with supporters and thus developing disintermediation strategies more than his predecessors. However, his experience has been normalised and absorbed, and the essence of the party model—a party model in which the leader is potentially strengthened from the outside through primaries, which are the main tool of citizens' participation, despite rhetoric on the open party—didn't change with the election of Zingaretti, even though he presented himself as the promoter of a party in which "sovereignty must move towards the base of the pyramid".

In the M5S, on the other hand—although there is not a substantial decrease in the prerogatives of the leadership—we witness a progressive increase in intermediation, the growth of formal and informal intermediate bodies, which is testified to by the profound changes in party documents and organisation with the development of the institutionalisation process. This increase in intermediate bodies can be seen as a consequence, as stated in Pedersen's model, of the internal complexification brought about by the passing of the different thresholds.

#### *The Virtual Arena 2.*

The second arena considered in this study is the virtual one. The use of the internet of the two parties was analysed, especially the tools potentially capable of influencing party decisions and the internal distribution of power. Although the use of the internet by the M5S is much more advanced than the PD's, we can say that in both cases there is no real empowerment of the base through digital tools: the new technologies are used by the two parties, according to the pseudo-participation hypothesis (Pateman 1970), to *give the impression* to members or supporters that they can influence the decision-making processes of the party.

The digital tools of the PD are completely detached from the decisionmaking processes of the party, while the online consultations of the M5S, despite the fact that they touch important areas of decision-making, are piloted from above and take place within an infrastructure that is centralised and managed in a poorly transparent way. Even in this case, the from-above dimension prevails in the practices of both parties. In both cases, digital tools don't lead to the empowerment of the base, but are used mainly with rhetoric or symbolic purposes, leading in the end to a greater concentration of power in the hands of the leadership.

But the two cases present also relevant differences. In the case of the M5S, the role of the internet as a new form of intermediation emerges rather clearly. In chapter 6 and 7 we have seen that the lack of a party structure does not lead to a genuine bottom-up organisation, leading instead to new hierarchies and concentrations of power. In particular, Associazione Rousseau, which managed all the decision-making processes of the M5S through the control of the online participation platform *Rousseau*, acted as a new gatekeeper of the digital age, and acquired increasing power over time, until it was discharged in early 2021.

In contrast, in the case of the PD, digital tools don't represent a new form of intermediation: they are managed within the traditional party frame. The potential innovation brought about by digital tools, of which *Rousseau* is a perfect example, is completely normalised and kept under control in the case of the PD, in the case of *Bob* even more than in that of the previous experience of online circles.

#### *The Local Arena 3.*

The third arena considered is the local one, in particular members' and activists' participation during an electoral campaign. In chapter 5 we saw how the PD, through the organisation Noi Siamo Torino, staged a direct relationship (or rather, one mediated by volunteers) between citizens. This way of conducting a campaign represents, on the one hand, the reinvention of old practices such as canvassing, and on the other is an attempt to offset the absence of party militants, who have a marginal role in the party model of the PD. On paper, non-partisan volunteers are supposed to replace party militants, creating a direct and unmediated relationship with citizens. Nevertheless, in the analysis of this experiment of externalisation of the campaign, we saw the emergence of managerial organisation of participation, as well as—partially unexpectedly—the relevance that party militants maintain in this new kind of electoral campaign.

On the other hand, unlike what we might assume, given the importance of the internet in the national organisational structure, at the local level the participation practices of the Movimento 5 Stelle are mainly offline: the internet is used by activists only as a tool of coordination. In chapter 8 we saw that, at a local level, the M5S forms horizontal and autonomous groups, which have some traits of social movements. In the absence of a formalised party structure, activists organise themselves autonomously (but they depend on the centre in order to obtain the authorisation to use the symbol), creating horizontal groups rooted in the territories. However, besides the fact that they act in a very centralised national organisation, even at the local level hierarchies and informal forms of leadership emerge. The absence of formal party structures leads to the strengthening within the organisation of the role of the elected representatives.

At the local level, the analysis aimed to explore the two parties with ethnographic methods, and therefore the dimensions used for the examination of the parties at the national level (disintermediation from below and from above) are limiting. However, we can state that the rhetorical or symbolic dimension of disintermediation can also be found in the case of Noi Siamo Torino. Moreover, we see that a common trait of the two examples of local mobilisation is the attempt to substitute party-mediated relationships with people-mediated ones. They are "staged" in the case of NST and more "genuine" in the case of the M5S, but the increasing relevance of direct and personal relationships established between common citizens can be considered both a sign of parties' weakness and delegitimisation and a new form of intermediation, albeit one typical of the past (Manin 1995).

On the one hand, the NST campaign was organised by a managerial structure external to the party, which can therefore be seen as a new form of intermediation. On the other, the main players of the campaign were not "real" volunteers, but party militants; therefore we can see the permanence of the old forms of intermediation. As regards the case of the M5S, we have seen that the local party is grounded in the territories with a constant presence on the streets. This together with the strong identification with the party symbol are characteristics that, unexpectedly, recall some features of mass parties, even though in chapter 8 we saw that the M5S differs from traditional parties in various, relevant ways. For instance, local groups are not officially recognised as local branches by the national party in central office: official participation at the national level is individual and takes place through the online platform. Moreover, unlike traditional parties, there is no local party structure at this level either: activist self-organisation replaces it. To sum up, at the local level, in both cases, we see a combination of disintermediation and old and new forms of intermediation.

#### *Mainstream and New Parties in Comparison 4.*

The two parties considered in this study are very different: one is a party that can be defined as mainstream; the other is a new and anti-establishment party. Their responses to the changed context are in turn different, but we can state that in both, although through partially different paths, we find the will to weaken the intermediate party structure and to create a direct link between leader and supporters, which—despite rhetoric pertaining to members' and supporters' empowerment and direct participation—has resulted in an increase in the power of the leadership or the persistence of old forms of intermediation. Open primaries and online participation platforms are two different tools used by the two parties in a similar fashion to achieve this result.

In the case of the PD, open primaries are presented as a way to increase citizens' direct participation, but, in reality, they are used to keep tight control over elite recruitment and to increase the scope of the party leader's power and its autonomy *vis-à-vis* party organisation. According to Sandri, Seddone and Sozzi (2020), they function "as a trojan horse fostering party organisational weakening", and represent an innovative tool used to carry out traditional political activities through the centralisation of power in the hands of the leader. In a similar vein, behind the façade of disintermediation we also find the persistence of the leadership in the case of the M5S.

Dealing with digital parties, a category in which the M5S would also fall, Gerbaudo (2019) defined the way in which—in these kinds of actors opening the party's lower levels through online participation is accompanied by an increasing concentration of power in the hands of the party leader as "distributed centralisation". In digital parties there is a contradiction between a narrative of radical disintermediation and leaderlessness, and a reality in which leadership and hierarchy are very far disappearing. As it is in the software of the platform and in the process of management that takes place in its back end, power relations simply become more concealed (Gerbaudo 2019, 184).

These dynamics seem to take place following the same general logic, regardless of the type of party. On the one hand, we expected mainstream parties to open up their decision-making processes in order to gain legitimation, but since they are conservative organisations, we expected them not to change, or give away their power, simply for the sake of change. On the other hand, the goal of new parties is often to empower members' participation. However, even in this case we expected them to tend, over time, towards institutionalisation and centralisation. Thus, as expected, in both cases we can say that the different devices promoted to foster members' empowerment in reality hide plebiscitarian dynamics and the centralisation of power in the hands of the leadership.

Contrary to rhetoric characterised by disintermediation from below, when it comes to practices, it is disintermediation from above that prevails: it is the leadership that becomes strengthened when the intermediate organisation—be it party members or the middle-level elite—weakens. Even though in the last few years parties have tried to convince us of the opposite, favoured by social and political trends such as the decline of deference and mistrust in politics, party members and supporters have not acquired much power. Disintermediation strategies, when a greater decision-making power is promised to the base, are primarily of a rhetorical or symbolic type.

But disintermediation strategies, especially if they are of a symbolic or rhetorical type, do not leave a vacuum. Studies conducted in the field of commerce show how old intermediaries can re-establish their power in the face of disintermediation, and how new intermediaries can appear; so, we have considered disintermediation as a process that involves both the persistence of old forms of intermediation and the emergence of new ones. On this topic, results are not univocal, as old and new forms of intermediation coexist in the two parties.

On the one hand, it is true that the PD, although presenting major innovations in its party model, remained connected with old party forms more than the M5S. But, on the other hand, at the local level the M5S, which at the national level provided a radical innovation in its party organisation with respect to traditional parties, presents participation practices that are more "old-fashioned" than the PD, the heir of the two most important mass parties of the so-called First Republic. In the case of the M5S, at the national level disintermediation strategies led to the emergence of new forms of intermediation, mainly linked to the use of the internet. However, over time we also witness a strengthening of the role of elected representatives in the organisation at the national and at the local level, and structuring of the organisation in a "party like" shape. The demand for collective leadership as an outcome of the *Stati generali* at the end of 2020 testifies to a clear need for intermediation expressed by party members, and it is worth noting that in its new statute, the PD uses the word "congress", which was absent before. Finally, while at the local level the campaign carried out by NST can be considered a new form of intermediation based on people-mediated relationships carried out by volunteers, in terms of the use of the internet at the national level, the PD fails to break away from traditional party forms.

# Conclusions

The aim of this work was to understand how parties adapt to a context characterised by the refusal of intermediate bodies, or rather how they adapt to their perception of such a context. The refusal of intermediate bodies is said to be part of a larger process of transformation, for some of a crisis, of representation. In this context, parties, perceived as more and more disconnected from society, devise solutions to get closer to citizens and obtain legitimacy. Among the various fields of enquiry related to party change, this work focused on organisation and, in particular, on what we defined as strategies of disintermediation. Although the study of party organisational change is a highly developed field of research, disintermediation in politics, and specifically in parties—unlike in other areas, such as communication—is a field still partially unexplored.

Disintermediation strategies in party organisations have been described as rhetoric or practices developed by parties in order to stage or deliver an unmediated relationship between leader and followers, which happens through the weakening of a party's intermediate organisation*.* Therefore, we have identified two dimensions through which to study disintermediation in parties. The first opposes rhetoric and practices: do disintermediation strategies imply a real change in the internal distribution of power or are they constituted only by a discourse or narrative of change? And, in the event in which disintermediation strategies are put into practice, how are they interpreted? The second opposes disintermediation from above and disintermediation from below, i.e. the two directions of the weakening of the intermediate party structure: this can happen either by giving greater power to the leader or by granting it to its members or supporters. Finally, this work questioned the emergence of new forms of intermediation and the possible persistence of the previous ones, considering disintermediation also as a process.

In order to verify whether disintermediation strategies are present in parties, what the prevailing dimensions are and, more generally, how those strategies are interpreted and applied, two Italian parties were studied from three different points of view. There are three arenas considered in this work: the national, the virtual and the local one. The results of this study show that these are three partially different ways of carrying out and interpreting disintermediation, which nevertheless correspond to the same logic: the will to create a direct link between leader and supporters through the weakening of the intermediate party structure.

While at the national level—despite rhetoric relating to members' and supporters' empowerment and direct participation—in both the cases considered this creation of an immediate relationship is only apparent as it has as an outcome, although with different results, an increase in the prerogatives of the leadership and the persistence of old intermediaries, at the local level a common trait of the two cases is the attempt to substitute party-mediated relationships with direct and personal relationships established between common citizens. And, in both cases, digital participatory tools are used more to *give the impression* to members of participating than to grant them effective decision-making power.

My analysis of the two parties' disintermediation strategies conducted in three different arenas gives us a composite, though not exhaustive, picture of the possible responses and transformations of the parties in the current context. The results of my research are obviously limited to the two cases taken into consideration, which can be considered heuristic case studies aimed at generating hypotheses for future research and at stimulating new theories in the still largely unexplored field of disintermediation in politics.

In this concluding chapter, starting from the results achieved in this work, I will trace some paths for future research, highlight the contribution that my research can make to the conceptualisation of disintermediation, and finally try to understand what the implications and consequences of the phenomena under investigation are on the future of political parties and, more generally, representative democracy. With regard to future research, it would be interesting to test whether the concept of disintermediation can "travel", taking into consideration different parties in different countries. There are more than a few cases that, at first glance, would fit into the framework presented in this work.

In the first place, the case of Podemos in Spain seems an interesting one to take into consideration. Podemos has been compared to the Movimento 5 Stelle in various respects (Bickerton and Invernizzi Accetti 2017; Montesanti and Tarditi 2017; Vittori 2017; Deseriis and Vittori 2019), not only because they are both new parties that emerged in the same years in a context of a deep crisis of representation. With respect to the focus of this work, we can say that in the case of Podemos we also find the attempt to weaken the party's intermediate organisation and to favour citizens' direct participation. Moreover, even in this case there is a strong leader (disintermediation from above), and the internet is widely used to empower members and influence the party's internal decision-making process (disintermediation from below) (Pizarro and Labuske 2015; Biancalana and Vittori 2021a). But it is also true that, compared to the M5S, Podemos has an internal organisation more similar to that of traditional parties and that its leadership retains less autonomy (Chironi and Fittipaldi 2017).

The Pirate Parties can be considered another case of parties in which we find weakening, or better the disappearance, of the internal organisation. Pirate Parties have also been likened to the Movimento 5 Stelle because of their use of the internet and their faith in its democratic potential (Deseriis 2020), even though—in contrast to the M5S—Pirates value decentralisation and individual autonomy and entrust individual users with several deliberative capacities. With regard to organisational aspects, in the case of Pirate Parties, in contrast to that of the M5S and Podemos, we find the suspicion towards any form of centralisation and authority, and mostly decentralised forms of engagement (also testified to by the use of the software LiquidFeedback). Here, the only dimension of disintermediation present is the one from below, and it would be interesting to investigate if there is a link between the absence of a leader and the scant electoral successes that these parties achieved, with respect to the M5S and Podemos.

Finally, En Marche! (Evans 2017; Dolez, Fretel and Lefebvre 2019) could be another relevant case to analyse, as this is a new party created by a (mainstream) leader, Emmanuel Macron, in which there is no intermediate structure, but only local electoral committees directly linked with the leader (En Marche! is, according to Lefebvre "un club de mobilisation électorale", cfr. Lefebvre 2018, 29). Finally, it would also be interesting to continue investigating the Italian case, taking into consideration the parties that—mainly on the right of the political spectrum—put the figure of the leader before the organisation.

It was not the aim of this study to generate a theory of disintermediation. However, it is true that my research has contributed to the conceptualisation and operationalisation of this concept, which was lacking in the field of politics. In this respect, recently Pizzimenti, Calossi and Cicchi (2020) further developed the concept of disintermediation applied to party organisations. According to these scholars, the concept is a heuristic tool capable of identifying parsimoniously a number of organisational changes, thus addressing "crucial aspects of the actual functioning of political parties, by enriching a common descriptive vocabulary of party change" (*ibidem*). The concept of "internal disintermediation" would entail cutting off the intermediate articulations of the party and the traditional organs of political mediation, while making the relationships between party leadership and party members/supporters more immediate.

More in detail, patterns of disintermediation are associated with an organisational profile in which the party leader is provided with extended prerogatives; party executive organs are dominated by the representatives of the party in public office, while the complexity of party structural articulation decreases; and the procedures to join the party are simplified, while the number of members sharply declines in parallel to the dismissal of collateral organisations. According to their empirical research based on a set of indicators<sup>86</sup>, most of the parties considered in the countries analysed (Austria, Belgium, Denmark, Germany, Ireland, the Netherlands, Norway, Sweden, the United Kingdom) have adopted more disintermediated organisational profiles, and new parties tend to be more disintermediated than the old ones.

This concept thus captures a relevant trend in contemporary politics. In this respect, the results of this research show that it is possible to grasp not only contradictory vertical and horizontal trends, but also the emergence of new forms of intermediation. Indeed, in this work, I considered disintermediation in two different but complementary ways. The first one is disintermediation as a *strategy*, which we have defined as rhetoric or practices developed by parties in order to stage or deliver an unmediated relationship between leader and followers, which happens through the weakening of the party's intermediate organisation, and of which we identified two dimensions, from below and from above. The second one is disintermediation as a *process*. Seen in this way, the concept of disintermediation also takes into account the diachronic dimension and can be linked to the "IDR cycle" developed by Chircu and Kauffman.

If it is conceived in this way, it could be possible to define disintermediation as a process that, starting from a situation in which intermediate bodies (such as the middle-level elite or party members) operate in a party (I), leads to disintermediation—that is, a weakening of the party's intermediate organisation (D) and, from there, possibly, to new forms of intermediation (R). However, it is not necessarily a unidirectional process.

<sup>86</sup> The indicators are: a) the opening of boundaries of party organisations (for instance, opening the organisation to "friends" and/or "sympathisers"); b) the dismissal of the party's collateral organisations; c) the decrease in the number of party layers between the highest executive body and the party congress; d) a greater presence of representatives of the party in public office in the party's executive organs; e) the expansion of the rights and functions of the party leader.

We can picture it as standing over a continuous line, on which *it is possible to travel in two directions*, that is—like in the case of the parable of the Movimento 5 Stelle or, to some extent, the transition between Renzi and Zingaretti—also from disintermediation to intermediation. So, we can define disintermediation processes as *the transformations of the party's intermediate organisation: its weakening, the creation of new or the recreation of traditional forms of intermediation*.

To conclude, we can ask ourselves: what are the implications and the consequences on parties and, more generally, on the future of representative democracy of the phenomena under investigation? In a recent article focused on the causes and consequences of the increasing democratisation and inclusiveness of mainstream parties, Ignazi (2020, 5) claimed that the creation of "un-intermediated relations between leaders and followers", favoured by the opening of the decision-making process to individual members:

has had an unexpected, but unsurprising, outcome: the rise of rightwing populist parties which play the plebiscitary card unscrupulously. The naive hyper-democratic drive towards members' and citizens' empowerment in the party decision-making process has unwittingly promoted leaders and parties that legitimize their voice playing on a direct appeal to the undifferentiated, homogeneous, people, with a call which recasts the holistic imprint of politics (Rosenblum 2008). The attempt to counteract dissatisfaction and mistrust towards parties by opening them up and relying on the "resource leadership" has proved unsuccessful and counterproductive. Membership shrinks and demobilizes, leaders go their own way and confidence plummets—and antidemocratic forces gain momentum. Democracy itself is under threat from a populist surge, because developments in intra-party dynamics have an effect on the external environment too.

As far as the two cases considered in this work are concerned, it is obviously impossible to tell whether the opening of the decision-making processes operated by the PD has led to the rise of the M5S: there are too many variables to consider, and to isolate a single causal dynamic would be simplistic. What it is possible to say, however, is that that it is true that both the PD and the M5S throve, as we have seen, in a context of a deep crisis of confidence in parties that has its roots in the passage from the First to the Second Republic and that it is highly likely that the two parties influenced themselves. For instance, Beppe Grillo opened the M5S's website in 2005, the same year as the first primaries of the PD, regarded as its "founding myth"; the M5S was founded in 2009, two years after the foundation of the PD and the first primaries for the election of the party leader. So, we can see that the idea of involving voters/members in the internal decisions of parties was already circulating in the Italian political system—it is to be noted that at the 2006 elections Beppe Grillo backed Prodi, and in 2009 he tried to participate in the primaries of the PD, but his application was rejected. In the same way, the innovative use of the internet by the M5S had an impact on the PD, which, as we saw in chapter 4, explicitly developed *Bob* to counter the challenge of *Rousseau*. Also, the use of the so-called *Parlamentarie* by the PD on the occasion of the 2013 general elections can be seen as a further opening in the process of candidate selection, fostered by the use of a similar method by the M5S.

But it is also true that, in the long run, the balance of power between the two parties was reversed, as at the last general elections the PD obtained its worst result ever, and the M5S its best (Valbruzzi and Vignati 2018). Can this be seen as the success of a party that "plays the plebiscitary card unscrupulously" over a party that tries to "counteract dissatisfaction and mistrust towards parties by opening them up and relying on the resource leadership"? It would be an exaggeration to affirm that the fact that the two parties governed together for one year represented the union of two interpretations of disintermediation, but—as we have seen in this work—in this respect the two parties show some interesting similarities: rhetoric relating to members' and supporters' empowerment that ends up in increasing the autonomy of the leadership; the use of participatory tools in a plebiscitary way; the persistence or recreation of an intermediate structure.

More generally, we can ask ourselves whether disintermediation strategies, conceived in order to give parties more legitimacy in a context of refusal of intermediate bodies, do in reality weaken them, thus contributing to the circulation of rhetoric about the uselessness of the parties, and also having an effect on the "external environment", that is, democracy. Both mainstream parties and new parties seem entangled in this problem. Disintermediation appears then as a double-edged sword, both because, in the end, contributing to the delegitimisation of the intermediate bodies at the intra-party level, it fosters parties' de-legitimation as transmission belts at the systemic level, but also because fully keeping the promises implied by the rhetoric on disintermediation (especially disintermediation from below) is, in the end, impossible.

So, once again, whoever says organisation says oligarchy? In 2013, Carty wrote that party types that have emerged with the transformation of societies have not been able to overturn Michels' "sociological law", and my research appears to confirm once again the validity of that statement. And if, on the one hand, we could say that yet another confirmation of what was said by Michels more than one hundred years ago is pointless, on the other I believe that it is interesting to have proved it in the current context, in an environment that, more than in the past, enhances the absence of mediation and hierarchies and the apparently horizontal nature of political relationships.

But what Michels also stated is that if democracy is an ideal model that cannot become reality, the very fact that we are trying to reach it is relevant and can have positive outcomes. It is the metaphor, written in the conclusions of *Political Parties: A Sociological Study of the Oligarchical Tendencies of Modern Democracy*, of the farmer that is ploughing a field looking for treasure that is not there and that will not be found, but in doing so he improves the quality of the soil. In Michels' perspective, the farmer was the labour movement and the field democracy, but history tells us that there will always be newcomers, parties and political actors calling for *more* democracy, empowerment or disintermediation from below, and that this could be seen not only as a negative circumstance.

Indeed, according to Canovan, one of the greatest challenges facing contemporary democracies is that of the discrepancy between the democratic ideal and the reality of political practice (Canovan 1999). We can say, then, that disintermediation is one of the answers to this perception of disequilibrium, a response that concerns the desire for a democracy in which, literally, the power is truly "of the people" and exercised directly by them, without parties, or through organisational forms that favour unmediated relations between citizens and power. Disintermediation, therefore, can be conceived as a different way of imagining and practising representation within democracies.

It is an issue that has been raised recently (Mastropaolo 2018). If, on the one hand, representation has become much more unstable than it was when it was structured by parties, it is inappropriate to consider this a "crisis of representation". According to Mastropaolo, we are rather faced with a new adaptation in the use of an ancient institution. If anything, some "entrepreneurs" and some representative "companies" are in crisis, while other entrepreneurs and other companies, such as the so-called populist parties, have a lot of success. Disintermediation can thus be considered rhetoric through which parties promise to citizens—who feel they have been betrayed by those who represented them—a radical alternative to the forms of conventional political representation.

https://doi.org/10.5771/9783748912644, am 28.09.2022, 11:13:25 Open Access – - http://www.nomos-elibrary.de/agb

# References


Bordignon, F. and Ceccarini, L. (2019) *Five Stars, Five Years, Five (Broken) Taboos*, in Ceccarini, L. and Newell, J. L. (eds.) *The Italian General Election of 2018*, London, Palgrave Macmillan, pp. 139–163.

Bourdieu, P. (2000) *Propos sur le champ politique*, Lyon, Presses Universitaires Lyon. Calise, M. (2010) *Il partito personale*, Roma-Bari, Laterza.


212


214


Morlino, L. and Tarchi, M. (eds.) (2006) *Partiti e caso italiano,* Bologna, Il Mulino.


Woods, M. and Flinders, M. (2014) *Rethinking Depoliticisation: Beyond the Governmental,* in "Policy & Politics", vol. 42, no. 2, pp. 151–170.

https://doi.org/10.5771/9783748912644, am 28.09.2022, 11:13:25 Open Access – - http://www.nomos-elibrary.de/agb

# Appendix


# *Number of interviews and occasions of participant observation*

## *Appendix*


*List of interviewees, Partito Democratico*


*List of interviewees, Movimento 5 Stelle*


*Noi Siamo Torino, summary of the profile of the interviewees*

# Acknowledgements

It is hard to summarise an experience of almost seven years in just a few lines, and to try to remember all those who have accompanied and helped me along this path. It is difficult, of course, because there is always the risk of forgetting someone, but also touching, because in the "individual" work of academic research sometimes we feel lonelier than we really are.

First of all, I would like to thank my three (!) supervisors. In order of appearance: Alfio Mastropaolo, who followed me throughout this long journey, encouraging me to to strive for continuous improvement, leading me to think outside the box and giving me great support, not only in intellectual but especially human terms. Oscar Mazzoleni agreed to co-tutor this thesis when the project was already underway, and in these years he was always present, providing me with fundamental comments for the success of this work, pushing me to never give up, and finally offering me the opportunity to continue my academic career. Franca Roncarolo "adopted" this thesis when it was already grown up and helped me, with an attentive and external gaze, to see its weaknesses but also to appreciate it more.

The members of the commission Emilie van Haute, Michele Sorice and André Mach read the first version of the manuscript carefully. The discussion that followed, during the *colloque privé*, was truly inspiring: I consider it a privilege that three scholars like them devoted time and energy to my work.

But they are not the only ones to have influenced this work. Roberto Biorcio pushed me to investigate the case of the M5S in Turin, which is – at the end – the part of this research I'm most proud of. Martina Avanza read and sharply commented the chapter on Noi Siamo Torino, which could not have existed without the help (but also the inventiveness) of Cristopher Cepernich. Marco Deseriis' research on participation platforms was a point of reference for the part of this research dedicated to the use of the internet by parties. On this piece of research, I also owe much to my collaboration with Davide Vittori and the cyber-parties group. Not a few ideas included in this thesis came from the workshops of the Giangiacomo Feltrinelli foundation, organised by Spartaco Puttini. Chiara Geloni, met by chance on the train, put me in touch with the executives of the PD and, also, with a piece of the "political history" of my family.

As for the friends I met along the path, I would like to thank my colleagues of the XXX cycle of the doctorate in Social and political change: Valeria Bianchi, Marco Piasentier, Antonello Picucci and especially Elena Ciccarello, with whom I shared not only a desk but also the joys and sorrows of our professional and personal lives for two years. It is difficult to find four real friends all at once: I was very lucky.

At the Department of Culture, Politics and Society I met Maria Trapani, Antonio Vesco, Vittorio Martone, Luca Bossi, Chiara Maritato, Stella Pinna Pintor, Giulia Marroccoli and Tommaso Frangioni, with whom I shared lunches, coffees, cigarettes, drinks and paranoia. Before the doctorate, Turin was a place like any other; if I now feel it is my home, it is also because of them.

Precious, and not only academic, support came from Irene Bono, Massimo Cuono, Gianfranco Ragona, Giuliano Bobba, Antonella Seddone, Moreno Mancosu, Marinella Belluati and Roberta Ricucci. And in particular from Daniela R. Piccio, who starting from the writing of an article on the organisational evolution of the M5S, became a great friend. Stefania Profeti and Guido Legnante, with whom (unfortunately) I do not share a workplace, allowed me to call them whenever I needed.

The colleagues and friends of the Institut d'études politiques of the University of Lausanne (Sabina, Yitang, Iris, Andrea, Roberto, Karim, Ludovic, Steven and Steven, but especially Maxime) accompanied me in my first experience of life abroad in 2018. Eventually, and luckily, my time in Lausanne didn't end with the completion of my PhD. This allowed me to meet Grégoire Yerly, who warmly supported me in the last period.

Massimo Bavastro read some first drafts of this work for reasons that, I believe, can be understood only through the mystery of friendship. Giorgio Malet would probably have read something beforehand, if only I had sent it to him. To his severe judgement I submit this research study, now that it is published.

Finally, I must thank the people who have been close to me over the past years.

My flatmates from via Nizza 32: Enrica, Pierangelo, Virginia and Giovanni.

My family, without whom this work would not have been possible.