A digital principal? Substantive representation in the case of the Italian Five Star Movement Open Access

The emergence and seemingly unstoppable success of populist parties are often connected to the profound crisis of representation that has been ravaging contemporary democracies (Morlino and Raniolo 2017). Populist parties build their raison d'être on the exaltation of the virtues of the people and popular sovereignty (Mény and Surel 2002), in opposition to what they perceive as a corrupted and self-oriented elite, deaf to citizens’ demands. Accordingly, the central populist message is that ‘popular control [over politics] has to be restored’ (Kriesi 2014, p. 363).

This mission implies a different way of doing politics, one that aims at bringing it ‘closer to the people’ (Rovira Kaltwasser et al.2017, p. 4). Populists do so by highlighting issues that are often neglected by mainstream political parties (Hobolt and de Vries 2015; Meguid 2008), but also responding to members’ priorities (Ezrow et al.2011) to enhance responsiveness and reduce intra-party division. Yet, not all populist parties have the same organization and the same way to connect members with the leadership (Vittori 2020). For those parties which embrace post-bureaucratic forms of organization (Bimber 2003), digital technologies play a pivotal role in bridging the gap between the public and politics (Blumler and Kavanagh 1999), seeking to reverse long-term trends of crisis of representation through the direct participation of citizens in decision making (Deseriis and Vittori 2019). This paper seeks to explore whether and how populist parties’ digital platforms contribute to realign élite policymaking with member policy priorities. This topic is still underdeveloped (Kaltenegger et al.2021), even though recent works emphasized that niche parties strategically (de-)emphasize issue in response to party divisions (Ezrow et al.2011; Van de Wardt 2014). As the literature shows, party incongruence between members and the élite has important consequences for the party: members tend to exit the party if their views are not represented among the élite (Barnfield and Bale 2022; Gaasendam et al.2021) or they might even opt for voting another party (Kolln and Polk 2017). Moreover, ideological incongruence constrains the leadership (Kolln and Polk 2017, Greene and Haber 2014) and might push party élite to adapt the party agenda (Van de Wardt 2014). To analyse the party élite-member ideological congruence, we rely on an in-depth analysis of the elected representatives and the membership preferences of a technopopulist party (Bickerton and Invernizzi Accetti 2018),¹ the Italian Five Star Movement (FSM). The FSM is one of the most relevant cases of a populist party that invested heavily in digital infrastructure to put citizens at the centre of democratic politics. In the words of its founder Beppe Grillo, the very aim of the party is to ‘introduce direct democracy tools within institutions […] transforming the relation between voters and elected representatives in an absolutely transparent and continuous manner’ (Grillo et al.2013, p. 191). For the FSM, the use of a digital platform aims at a better registration of the opinion of citizens and a more responsive representation of their views by elected representatives, who are conceived as simple executors/agents of the people's will. We look at the FSM's participative digital law-making features as the place where member priorities and élite policymaking can be matched. The party's digital platform innovatively introduced the possibility for members to propose draft laws and for the Party in Public Office to request members’ feedback on their law proposals discussed in Parliament. Despite the decreasing participation of members in the platform and the top-down non-deliberative approach of the platform affordances (Mosca 2020), FSM organization invested many resources in promoting a different conception of democracy, based on the direct participation of both members and the party élite in the party policymaking. Given these premises, our study differs from previous ones that looked at the FSM's adoption of digital tools as an attempt to transform intra-party democracy (Vittori 2020) and focuses on the policy congruence between the members and elected representatives.

We consider the FSM as a revelatory case study (Yin 2018) allowing us to explore how the use of digital technologies can potentially redesign the substantive representation within a party. The study is based on an original dataset which comprises the law-making activities of FSM members and parliamentarians as presented and discussed on the digital platform of the party, called Rousseau, between 2013 and 2019. The period covers the debut of FSM elected representatives during the 17th legislature of the Italian parliament (15 March 2013–22 March 2018), as well as the first year and a half of the 18th legislature (23 March 2018–December 2019), in which the FSM entered government in coalition first with the League – a populist radical right party – and then with centre-left parties. We find that FSM elected representatives’ agenda, albeit partly diverging from that of members, still changed in the direction of member priorities through time.

The literature on political parties agrees on the fact that party membership in the last decades has declined markedly (Scarrow and Gezgor 2010; Van Biezen et al.2012; Van Haute 2011). The decrease displays an almost uniform and falling trend through Europe, with few exceptions, i.e. the democratic latecomer such as Spain, Portugal and Greece. More recent research, however, has shown that the decline trend at the party system level hides important differences at the party level (Kölln 2017; Van Haute et al.2018). In particular, more consolidated parties tend to have less members than newer ones, while differences among party families albeit present are not huge (Kölln 2017). The introduction of public funding schemes in Europe and the concomitant technological transformation, i.e. the spread of mass media to all sectors of societies, gradually made members less crucial for the success and survival of the party organizations (Katz and Mair 1995). Furthermore, the personalization and mediatization of politics (Mazzoleni and Schulz 1999) contributed to disintermediate the relationship between party leaders and the electorate, further accelerating the process of transformation of parties into state agencies disconnected from the society (Katz and Mair 2009). Also, many authors have emphasized how factors like the globalization and financialization of the economy, on the one side, and institutional supranational integration on the other, have increasingly compromised the responsiveness of mainstream parties (Boix 2000; Mair 2006 and 2009).

The work of Scarrow (2015) has shed a light on the multi-speed membership that several parties had introduced to cope with the necessity to contrast the membership decline. The creation of party clubs, alongside the classic party cells, or the introduction of new figures such as that of sympathizer were intended to provide lighter and more indirect forms of participation, which nonetheless allowed parties to spread ideas and to prevent the defection of the less involved membership. Similarly, party primaries that allowed members to directly elect leaders and candidates were (also) intended to expand intra-party democracy (Rahat and Hazan 2001) and to give party supporters a voice in the party decisions. Digital media further transformed the role and the function of party membership. First, digitalization has impacted on all organizational aspects of political parties: campaigning, fundraising and even meetings have increasingly transferred on the internet (Gibson et al.2017; Pedersen and Saglie 2005). Party membership has changed accordingly. Technopopulist parties, such as Podemos and the FSM, have entirely digitalized the enrolment procedure since their genesis, so that members acquire the right to vote once they have completed an online form (Vittori 2020). This trend is now common also among mainstream parties (Raniolo et al. 2021). Moreover, technopopulist parties make extensive use of digital platforms to propose new forms of engagement and participation among members. The aim of these platforms, besides providing a managing online tool for electoral campaigns, is to promote direct democracy procedures and redefine the concept of representative democracy within complex organizations. Digital platforms lower the barriers to make all members voting who can be potentially involved in all kinds of decisions without the organizational effort to set up an ‘offline’ vote. This is what happened in the case of Podemos and the FSM and this is also the case of mainstream parties organizing online primaries at both local and national level.

While a sizeable part of the bourgeoning literature has theorized the specific ways in which populist parties represent (among others Canovan 2002; Ochoa Espejo 2011; Werner and Giebler 2019), only few have empirically assessed how they perform their representative function and whether they are indeed more responsive (to members) than mainstream parties (Heinisch 2003; Heinisch and Werner 2019; Plescia et al.2019).

Even though party representatives are believed to be detached from members’ priorities (Katz and Mair 2009), there are empirical assessments (Kaltenegger et al.2021) on whether they substantively represent party members. Substantive type of representation, along with the descriptive one, is one of the four ways political actors have to represent voters (Pitkin 1967). While descriptive representation has already been analysed in the literature (Tarditi and Vittori 2021), substantive representation is far less investigated. The assumption that members are by definition more radical than the élite – the law of curvilinear disparity elaborated by May (1973) – have been substantiated by the literature, even though some recent contributions (Norris 1995, Van Haute and Carty 2012) put into question its validity across party families: in a nutshell, it should not be presupposed that members are more radical than the élite and the leadership nor that the members by definition diverge in their priorities compared to other internal party actors. Substantive representation is crucial when analysing the relationship between party representatives and the membership: depending on the relevance of the membership within a political party, members can shape party representatives’ composition (via the direct election of the represented) and, more important, their priorities. While selecting party representative is much common among parties, since intra-party democracy is conceived mainly as including members in the election of the party leadership or the candidates (Von Dem Berge and Poguntke 2017), party members are only rarely analysed in the literature (Kaltenegger et al.2021). Substantive representation is even more relevant for technopopulist parties because of the main constitutive elements of their ‘thin’ ideology: a vision of society as neatly divided into two groups, the people and the elites; the idea of an antagonistic relationship between these two groups; and the idea that the people are the only legitimate source of the ‘general will’ and, as such, should be considered the ultimate source of authority in the state (Mudde 2004; Ochoa Espejo 2011). Rejecting a political system in which representatives are part of and beholden to the elites and detached from the preferences of ordinary people, populist representation vindicates a congruence between representatives’ actions and claims of the represented (Mudde 2004). However, there are also other strategic reasons for parties to enhance congruence between the élite and the members. Firstly, while one would expect that niche parties for their limited membership are more responsive than mainstream parties (Ezrow et al.2011), Van de Wardt (2014) shows that both challenger and mainstream (de-)emphasize issues following internal party divisions. The reason why parties consider party congruence between members and the élite as crucial for the party might be different: members with a high ideological misfit vis-à-vis the party might abandon the party, threat the internal party cohesion, thus constraining the leadership (Kolln and Polk 2017; Van Haute and Carty 2012) or even vote for another party (Gaasendam et al.2021). Furthermore, in a compelling contribution about radical right parties Art (2011; see also de Lange and Art 2011) suggested that contextual factors – such as the historical legacy of right-wing extremism and the in-tolerance toward the existence of those parties – influence radical right quantity and quality of membership and this aspect, in turn, affect parties’ electoral performances. Thus, political parties and especially non-mainstream parties have strong incentive to care about the party congruence between the élite and the membership.

Technopopulism label has been used to describe a new party family (Deseriis 2017), in which digital affordances are a key organizational aspect. Technopopulists emphasize their distance from mainstream models of organization and instead favour a model of disintermediated representation through instruments of direct democracy like digital platforms. In this regard, technopopulist parties are part of the non-mainstream ‘family’ in which organizational aspects play a crucial role in structuring party preferences. Such instruments lead to further differentiation of party membership as interested citizens participating in platform-based discussions become quasi-party activists without necessarily being members of the party apparatus in any other way (Scarrow 2015). More importantly, digital platforms add digitally involved citizens as a new and dynamic principal whose mandate is not fixed only at the time of elections but constantly accompanies the parties’ representative activities. In the course of a legislature unpredictable issues, new understanding of policy problems, emergence of new public debates – factors usually identified as influencing both parties’ and public priorities (Froio et al.2017) – can move this principal's attention and energies to new issues. Under conditions of a dynamic organization of the party mandate, made possible by the employment of digital platforms, responsive populist representatives would then be expected to dynamically adjust to members’ desiderata. Yet, as the law of curvilinear disparity suggests (May 1973), addressing members’ priorities might affect party responsiveness at the electoral level making the party less responsive to the (median) voter. This might be detrimental for the electoral results and that is why mainstream parties usually adapt their positions to those of their constituencies (Ibenskas and Polk 2021). Thus, being responsive to members’ priorities is not without problems: yet, as specified above, there are also good reasons for party to care about what members prioritize. The literature suggests that parties and their representatives tend to stick to the manifesto at the beginning of the parliamentary term, with the issue content of manifestos being correlated to the topics addressed through the introduction of bills (Bräuninger and Debus 2009). As time goes by, however, while party pledges to voters remain static, member inputs (collected by digital platforms) may reorient the attention of party representatives to make them responsive to their digital principal. Even though we acknowledge that member priorities might diverge from that of the leadership and that the leadership might not be incentivized to adapt to stick to the law of curvilinear disparity, we believe there is also a good reason for party élite to re-orient their priorities, as explained above. Furthermore, having technopopulist parties the digital affordances to involve members and to listen to their desiderata, we believe that members’ priorities from a theoretical standpoint should be a core concern for these parties. For that reason, we elaborate a first set of ‘congruence-based’ hypotheses.

Beyond the fact that we will expect populist parties to, overall, take into account both voters and members while performing their representative function, we can also expect some issues to be more carefully addressed than others. Parties stress specific issues to signal voters their intention to give priority to certain topics (van der Broug 2017). Populist parties in general, and technopopulist in particular, centre their actions on specific topics, neglected by mainstream parties, which usually vary according to the ideological mix that accompanies the thin core of the populist worldview. Accordingly, populist radical right parties may appeal more to issues such as immigration, public order and European integration, while left-wing populist parties will be more likely to focus on issues such as unemployment and social policies, and redistribution. In this respect, the case of the FSM defies straightforward classifications, as the ‘historical’ core of the party message combines a mix of substantive issues that cut across different, and non-coherent, ideologies (Mosca and Tronconi 2019). As a technopopulist party, we expect the FSM to give particular priority to the following issues: criticism of the establishment, moralization of politics (understood as fight against corruption and cuts to the waste of politics) and call to popular sovereignty (Mudde 2004). However, it is also important to consider that the FSM is a party that, particularly in the founding phase, was characterized by a programmatic proposal strongly focused on neo-environmental issues (Tronconi 2015). These issues, alongside the theme of digitalization, are included in the 5 stars that give the party its name: public water, sustainable mobility, alternative energy, sustainable development and internet connectivity. Therefore, emphasis on environmentalism and technological innovation as well as ‘genuinely’ populist issues should be particularly present in the legislative activity of its representatives. Given this premise, we expect more congruence over core ideological issues compared to other issues:

The FSM rests upon a set of ideas that have been considered typical of technopopulist parties (Deseriis 2017). Such parties bridge populist and technolibertarian discourses: anti-establishment claims, reproach of representative democracy and its actors, and stress on the general will of the people are connected to ‘a blind faith in technological progress and free-market economics with a deep distrust in statist, bureaucratic, and hierarchical forms of authority’ (Deseriis 2017, p. 442). Nonetheless, the FSM has a specific peculiarity when it comes to direct democracy and membership involvement in the decision-making: it has been depicted as a plebiscitarian party (Vittori 2022), which involves members mainly to ratify party leadership decisions. However, the FSM is the only party in Italy to have developed a digital platform where members were allowed to directly connect representatives and the membership (yet, without success in terms of élite responsiveness) (Mosca 2020), while at the same time granting members the opportunity to draft law proposals and to see them endorsed by MPs in Parliament (see below). Another rather unique characteristic of the party is that the leadership of the party is an organizational leadership, more than anything else: the leader of the first generation (Grillo and G. Casaleggio) and the leader of the second generation (Grillo again, Luigi Di Maio and D. Casaleggio) did not have a say in the party ideological agenda, as the elaboration of the party manifesto testify (Vittori 2022).²

The identity of the FSM is strongly based on the idea of direct democracy made possible by members’ participation through an online platform. Throughout its history, the internet has been used by the party as an instrument for involving members in internal decision-making through online voting, for organizing activists (through the Meetup platform), for communicating with voters and sympathizers, and for creating and diffusing a cyberutopian narrative presenting the web as the very solution for contemporary ‘corrupted’ democracy (Mosca et al.2015).

Over time, the FSM employed a digital platform called ‘Rousseau’ – named after the Genevan philosopher to recall the centrality of direct democracy in the party identity – that allowed party members a high degree of participation between its launch in 2014 and 2021 when a change in leadership also entailed a change in the participatory platform, now called SkyVote. Through Rousseau, FSM members took part in several decision-making activities, including voting in online consultations, selecting candidates, and participating in discussions over draft laws (Mosca 2020). However, the platform was not only a decision-making instrument involving members but also an organizational hub aimed at collecting calls for action and mapping events organized by local groups, providing training opportunities to newly elected representatives, archiving all acts passed by Movement's representatives at different territorial levels, and offering legal defence to its members.

Concerning participatory features, Rousseau allowed FSM members to directly intervene into the law-making process through a feature tellingly called ‘Lex’. To make the law-making process more bottom-up and more participatory, this function allowed for interactions between representatives and platform users at different territorial levels (European parliament, national parliament and regional representative assemblies). Through a function called Lex Parliament, available between 2013³ and 2021, elected representatives published draft laws on the platform and members could provide opinions, as well as amendments, within 60 days from their publication. After that, the proponent (MP) evaluated the amendments, chose whether to modify the text and published the final version of the bill on the platform, presented in parliament afterwards (Mosca 2020). In July 2016, a new section was added on Rousseau, called Lex Members. Through this function, members could upload their law proposals, which were then put to an online vote in subsequent sessions, after fulfilling a formal examination (see Appendix Table A1). The two most voted proposals of each session were then transformed into bills and presented in parliament.⁴

Our study is based on an original dataset which comprises the law-making activities of FSM members and parliamentarians as presented and discussed by Rousseau (Feo et al.2022). We use platform users’ proposals as recorded on Lex Members to determine their issue preferences. Likewise, we employ proposals from Lex Parliament to infer FSM representatives’ issue preferences. Thus, we treat Lex Members as a proxy for the ‘digital membership’ principal. Being the various proposals advanced by members gathered in different sessions, we considered the first eleven ones, which took place from 2016 to 2017. We opted to restrict our focus to the first eleven sessions because the law proposals voted from session 12 onwards changed in a way that would have biased our analysis. While from session 1–11 law proposals were pre-selected and included into chronologically-ordered sections, the only requirement being the respect of some formal criteria by users,⁵ from session 12 onwards law proposals were incorporated into thematic session (e.g. public management of water service, environment, digital citizenship, corruption and anti-mafia), thus according to their content and not the timing of their presentation. Including these last sessions in our dataset would have skewed our results, as only the first 11 ones had no issue-related pre-selection. In total, members presented 1182 law proposals in 11 sessions (with a mean of 107 proposals per session); we were able to code 1175 proposals, as for few of them some basic information were either missing or not provided (see Appendix, Table A1).

We treat Lex Parliament as a proxy for the ‘elected representatives’ agent. National MPs presented FSM-sponsored bills on this section of the platform and explained in a short video the main features of the proposal. The proposal was then discussed among members. However, MPs were not required to take into consideration the discussion before presenting the final draft in parliament (Mosca 2020). As anticipated, we coded all law proposals presented by MPs on Lex Parliament in the previous parliamentary term (2013–2018), but we also considered the proposals presented in the first year and a half of the last legislature⁶ (2018–December 2019). We opt to keep the two terms apart to preserve the internal coherence of the comparison with Lex Members. We have coded 335 law proposals in the first term (2013–2018) and 100 in the first part of the second one (2018–2019). To make the comparison between Lex Members and Lex Parliament meaningful, we used as a cut-off point for Lex Parliament proposals the year in which Lex Members was introduced for the first time (2016).

To identify and code issues, we created a coding scheme borrowing from the approach employed by the Comparative Agenda Project (CAP) (Jones and Baumgartner 2005), which capture the issue content. Our original coding scheme comprises 17 categories: economy and trade, civic rights (materialism), civil rights (post-materialism), social policy, agriculture, labour (employment and unemployment), culture and education, environment, immigration, infrastructure and public transportation, law and order, defence and foreign policy, technological innovation, moral politics, corruption (and transparency), democratic institutions, EU politics and policies. While many of the categories are directly borrowed from the 20-odd categories employed by the CAP project, we modified the coding scheme to capture the specificity of technopopulist parties’ substantive representation, in line with our research questions (for a description of the codebook see Appendix, Table A2). The categories we added are ‘moral politics’, ‘corruption and transparency’ and ‘technological innovation’. These new categories are aimed at capturing the policy proposals advanced by MPs and Members for issues that are highly sensitive for the parties: here, we do not refer to generic claims to fight the corruption or to bring ‘morality back’ into politics, but to specific policies addressing these issues. For example, all proposals related to reduce the ‘privileges’ for politicians, to make parties accountable for their misconduct in the electoral campaign, to make institutions accountable to citizens fall under the ‘moral politics’ label. Few examples are (a) the proposal to create an agent provocateur figure, whose mandate is to check only whether parties misbehave during the electoral campaign, (b) forcing elected representative to remain in the parliamentary group they were elected and impeding switching from one group to another, (c) to force the local authority to stream the local council sessions online. Accordingly, we include ‘corruption and transparency’ and ‘technological innovation’ categories to check how many policies were proposed to fight corruption and enhance transparency and how salient was the issue of facilitating the introduction of new technologies (especially those related to digitalization) in both the public and the private sphere. An example of the latter is a proposal in the first session of Lex Members, in which the proponent proposed to create a web portal to evaluate the satisfaction of local health services.⁷

Since law proposals in both Lex Members (henceforth Members) and Lex Parliament (henceforth MPs) cannot be always straightforwardly assigned to a specific issue, we include the possibility of double-coding. In this case, we opt to consider both issues and to count them separately, rather than imposing which of the two issues have a priority in the law proposal. This procedure has the advantage of sorting out dubious code attributions and being as faithful as possible to the proponent's intentions. Accordingly, the number of observations was increased: our final N (issues coded) are 1479 for Members and 543 for MPs.

We present a couple of examples that illustrate the coding strategy. First, a law proposal published in Members (Session 1) concerning the introduction of the binding mandate in the Italian constitution. The law proposal falls both: (a) in the ‘democratic institutions’ and (b) ‘moral politics’ categories. As for the first category, this law foresees a change of the Constitution; as for the second, looking at the description made by the proponent, the content of the proposal refers to the enforcement of the binding mandate in representative institutions, preventing MPs from party switching. For that reason, it was assigned to the ‘moral politics’ category as well. The second example comes from MPs and concerns a draft bill titled ‘Prevention and contrast to mafia infiltration in agriculture and sheep farming sectors in state-owned areas’. Again, we coded the proposal with a double code: (a) ‘agriculture’, as the content refers to both agriculture and sheep farming, and (b) ‘law and order’ as the law proposal refers to quelling mafia infiltration.

As for intercoder reliability, we run three separate tests on Members’ first session, which counts a high number of proposal (129). The Krippendorff's alpha, calculated with kripp.alpha function in R was .75 for each variable.

To explore the dynamics of representation of the FSM and test our two sets of hypotheses we rely on one measure, called Delta, which we employ to measure issues’ congruence between MPs and Members.

Delta is defined as the mean of the absolute discrepancy in issues’ emphasis between the digital principal (members) and its agent (elected representatives).

To control for within discrepancies, that is the absolute differences between each of the 17 issues for Members and MPs, that might be overlooked by the overall result, we control the discrepancy value for each of the 17 issues coded. Overall, the results do not differ markedly.⁹

The descriptive statistics provide some preliminary interesting findings. In Figure 1, we report only those issues reporting a salience>5% in Members and MPs. Slightly in contrast with our expectations, we notice that issues such as (a) Economy and Trade, (b) Social policies and (c) Labour – which we can be considered ‘materialist’ issues – have the highest salience in both Members and MPs. Despite its post-ideological stance, the FSM do emphasize economy, contrary to other right-wing challenger families, for which economic positioning is less relevant (Rovny and Polk 2020).

The core issues, in which the FSM is expected to have clear-cut entrepreneurship, such as Democratic innovations, Moral politics or Corruption lag behind the previous three categories. Parliamentarians are the ones who place more attention to issues of Moral politics and Corruption, compared to platform's members. However, there are other relevant differences between Members and MPs: the former place much more emphasis on the above-mentioned categories in 2016 and in 2017 compared to their agent, while the latter are comparatively more focused on the environment, another entrepreneurial issue for the FSM.

Moving on to other two issues at the core of FSM ideology, Technological Innovation and Infrastructure and Transportation, these are not salient in both Members and MPs.

Looking at Table 1, the overall saliency of populist issues (Moral politics, Corruption and Democratic institutions categories) varies between 13.4% and 25.3%. If we sum together other-than-populist issues for which FSM has an entrepreneurship – Technological Innovations (almost ignored by both MPs and Members) and Environment plus Transports and Infrastructures – we get a range from 13.5% to 23.1%. The total of the core issues ranges from 33.7% to 43.7% of the total mention in MPs and from 32.8% to 35.1% in Members.

This evidence drives to some first, general reflections: rather than downsizing the emphasis on populist-related issues and attempting to redirect and diversify their agenda to other issues, Members and MPs tend to focus their attention on issues at the core of the party message. While emphasis on technological innovation and green issues tend to be rather similar across the two, MPs devote more attention to populist issues, compared to Members.

We move on now to our first set of hypotheses. We run a chi-squared test for all 11 Member sessions (see Appendix, Table A3), which were previously codified following the 17 categories identified above, to check for their independence to one another and every group results independent. Having discarded the possibility that sessions have some sort of dependence, we followed up with our statistical analysis.

Our descriptive statistics of our dataset comprise a contingency table with 17 columns (one for each coded issue). In each row, we include the levels of analysis.¹⁰ Following our formula, we obtain a Delta for each pair we analyse: one pair is constituted by MP proposals in one year and by Member proposals in 1 year.

Thus, it seems that in articulating their agenda, FSM parliamentarians do take into serious account the dynamic mandate they received from their members, rather than sticking to the more general mandate agreed on at the moment of the elections. We, thus, find confirmation for H1a and H1b. In line with the party congruence literature, thus, the élite cares and adapts to (digital) members. It would be unreasonable, though, to expect a perfect match between members’ and MPs’ priorities (Van Haute and Carty 2012): the FSM is not an exception to this trend. Our data do not allow to inquire why did they adapt, but in the following section, we provide some potential explanations.

To test our second set of hypotheses, we calculate the Delta for the populist issues (H2a), for the green issues (H2b) and for the technological innovation issues (H2c) and we compare these three categories with other three categories, economic issues, civil right issues and policing issues, which are the most recurrent issues in our dataset.¹² We did not set expectation throughout time in this case, but we hypothesize that the congruence between the first three is higher than the second group of issues.

Populists’ claim to fill the representation gap which democracies are effectively experiencing went under scrutiny in the literature: their responsiveness to the members should be in principle higher than that of other parties, not only because niche parties in general are responsive to members ideological shift (Ezrow et al.2011), but because their core message is to bring people closer to politics.

Yet, so far, the congruence between populist élite and their members has not been properly scrutinized. In this article, we test the ability of populist parties to deliver this promise by focusing on the revelatory case of the FSM, a prototypical populist party in several respect (Mosca and Tronconi 2019).

We focused in particular on what we consider a new principal of technopopulist parties, e.g. digital members, to whom they claim to be responsive. We then hypothesize that populist representatives are overall congruent with their members and that due to the digital interaction between members and representatives their congruence increases over time, granting members the possibility to influence representatives during their mandate, resulting in a dynamic mandate for populist representatives. Furthermore, we proposed another set of hypotheses according to which representatives converge with digital members on specific issues, in which the party has a clear-cut entrepreneurship.

We show that the overall congruence between elected representatives and digital members rather than diverging over time is converging, thus showing that MPs are adjusting their priorities to members’ desiderata, in what we have called a process of lagged adaptation. In this paper, admittedly we do not go back to look at membership motivations: in a nutshell, we do not know why members prioritize one issue over the other and which are the mechanisms that lead members to draft a proposal on a specific issue. It might well be that they are influenced by external factors: however, what we show is that given members’ issue priorities, MPs adapt to them. Further research in this regard is needed to check who are members who propose policies, what drives them and which are the mechanisms that push members to engage in digital platforms and, specifically, in online law-making. In this regard, there might emerge patterns of structural inequalities, e.g. the degree of offline activism linked to socioeconomic centrality, that might transform the opinion of most active and motivated members in the majoritarian opinion of the party. Be as it may, lacking comparable cases, we could not speculate further about the role of the digital platform in facilitating élites’ adjustment to member desiderata. Nonetheless, we can speculate about the possibility that the digital affordance might have helped MPs knowing members’ priorities, as the organizational structure of the party up until its last reform in 2021, substantially impeded a connection between members and MPs (Vittori 2022). In line with the literature on party congruence, we show that the élites care about members’ priorities (Barnfield and Bale 2022; Kolln and Polk 2017; Gaasendam et al.2021; Van de Wardt 2014; Van Haute and Carty 2012): there might be several reasons behind this convergence. First, MPs might be genuinely attentive to member priorities, as they want to put into practice the techno-populist ideal type for which members might become the most relevant agent in the party. Second, as the literature suggested (Barnfield and Bale 2022), they might be concerned about membership loss and about limiting internal ideological divisions, especially within a party that openly discouraged divergent opinions with respect to leadership positions. Third, it might be also the case that responsiveness from MPs is a strategic move to show members, who elect them in closed primaries, that they are attentive to their priorities.

Furthermore, representatives give priorities to many issues that mirror the issues at stake for the members. However, our second set of hypotheses is only partially confirmed. We found a convergence between digital members and their representatives on technological issues. Our findings are counter-intuitive as while there is congruence between the two, the topic is substantially neglected or ignored: while the leadership of the party frame this topic as a defining ideological element (Mosca 2020), the issue is not incorporated by its representatives and its members. As for populist issues, we noticed that the Delta between the digital principal and its agent increased over time. This seems to show that the legislative activity of MPs could have been influenced by the approximation of general election of 2018, when a growing emphasis on populist issues could have been considered as particularly rewarding vis-à-vis the input of digital members. Lastly, the increasing convergence between MPs and digital members concerning green issues seems to show that there could be a return of focus on a topic which defined the party identity in its origin and that was partly diluted in the parliamentary daily routine.

In conclusion, elected representatives appear significantly influenced by the issue preferences of their digital members. However, the promise to fill the gap of democratic representation is only partially addressed by technopopulist parties, especially if we take into account the low number of party members in comparison with party voters (Vittori 2022).

This paper has some limitations, though: firstly, it focuses on a peculiar technopopulist case, whose digital platform is unique and not easy to replicate. While we deem our methodology as a useful tool to be replicated for other parties, potentially belonging to the same party family, we are also aware that the dataset we compiled represents an exception, rather than a norm. Furthermore, the requirements needed to propose a law in the platform might have created a self-selection bias: we acknowledge this limitation, but we are also confident that our large N in a way helped reduce the bias for a party whose membership has been always quite limited. However, we hope that this paper will be the first step to further inquire the relevance of the congruence between members and the party elite and, equally important, the consequence of the dealignment between two agents.

We would like to thank Francesca Feo for her relevant contribution to the development of the codebook, the collection and coding of law proposals, and the systematization of the dataset we use. Lorenzo Mosca has received funding for this project from the Department of Social and Political Sciences of the University of Milan.

No potential conflict of interest was reported by the author(s).

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Davide Vittori is a Post-Doc fellow at Université Libre de Bruxelles and adjunct professor at the University of Antwerp. He was previously post-doc at LUISS University (Rome). He holds a PhD in ‘Politics: theory, history and science’ at LUISS University. His first monography is titled ‘Il Valore di Uno’ (2020, LUISS University Press). He works mainly on political parties, European political system and on voting behaviour. His works were published by Comparative European Politics, Government & Opposition, European Journal of Political Research, South European Politics & Society, West European Politics and others.

Lorenzo Mosca is full professor of Sociology of Culture and Communication at the Department of Social and Political Sciences, Milan University. He is president of the MA in Public and Corporate Communication and member of the national commission for the Abilitazione Scientifica Nazionale (ASN) in his area of expertise. Among his research interests lie sociology of digital media and political communication, sociology of participatory processes, civil society and social movements, populism.