Abstract
This study examines how digital parties differ from traditional parties in their external political communication on Facebook during the French 2022 presidential election. Using a supervised machine learning method, we analysed posts (N = 17,060) published on the Facebook pages (N = 544) of French parties during the campaign. Our findings show that digital parties use Facebook intensively to mobilise their activist base, confirming their reliance on online participation. However, contrary to expectations, they do not structure their communication more around their party leader than traditional parties, nor do they systematically generate higher engagement levels. Instead, they grant greater autonomy to their members’ Facebook pages, particularly at the regional level, where activists adapt national campaign strategies to the local context. Finally, while digital parties use Facebook to amplify their candidate's media coverage, they rely less on diverse content formats such as videos and images, challenging previous assumptions about digital campaigning strategies.
Scholars have demonstrated that the resources provided by the internet have ushered in significant changes in party organisation, leading to the conceptualisation of a new subtype of party organisation, the ‘digital party’ (Barberà et al., 2021; Deseriis, 2020a; Gerbaudo, 2019). This notion reflects recent changes in the organisational mode of new political parties, which have used digital technology resources to (re)structure their internal organisation on a more ‘participatory’ basis (Gerbaudo, 2018), allowing party members to express themselves, participate in decision-making and become more directly involved in party activities. Digital parties are often made up of young members who are more open to new communication tools to foster more decentralised and less hierarchical organisations (Magin et al., 2017). These parties are emerging within a broader international political context marked by a crisis of party ideology and the decline of traditional political issues in favour of new social and political concerns that mobilise citizens’ engagement (digital rights, new welfare provision, new forms of democratic participation). Because these issues are not central to traditional party ideologies, they are championed by charismatic figures who, more often than not, enter the political scene via civil society rather than through the ruling parties. To implement their alternative political proposals, these ‘hyperleaders’ then establish new political parties whose organisational structures align with digital parties, utilising ICT resources to unite a ‘superbase’ of activists and engage them in party activities (Gerbaudo, 2018). Scholars have therefore analysed how digital technologies have facilitated the emergence and ‘internal’ organisation of these digital parties, promoting more democratic and decentralised processes. Consequently, they also expect digital parties to use these advantages of ICT, allowing them to build a more effective ‘external’ political communication (Gerbaudo, 2018; Peña and Gold, 2023). For example, scholars suggest that digital parties are expected to manage their political communication on social media during electoral campaigns by posting more content than traditional parties and encouraging their supporters to engage more with their publications to increase their visibility. However, there is a lack of empirical evidence regarding this aspect of digital parties’ external political communication, based on a real data-driven study collected from the social media accounts of digital parties.
To fill this gap, we examine how French parties managed their Facebook pages (N = 544) during the 2022 presidential campaign. We analyse the political issues addressed in their publications (N = 17,060) as well as the intensity of user engagement (in terms of likes, shares and comments). The aim is to compare the ways in which digital parties and traditional parties use Facebook in order to determine whether the external political communication of digital parties relies more on: (a) participationism and supporter engagement; (b) the party leader and media coverage of their campaign; or (c) political issues, particularly divisive ones. The French context is particularly well suited to this analysis, as it allows us to determine whether the ‘external’ communication of the three digital parties varies according to their political ideology (a far-left vs. centrist vs. far-right digital party) and their involvement in government (governing party vs. opposition party).
Theoretical background
Hyperleaders in the electoral communication of digital parties
Paolo Gerbaudo (2018) distinguishes digital parties from traditional political parties by their organisational structure and mode of participation. He describes how new political formations, such as Podemos in Spain or La France Insoumise in France, rely on digital platforms that facilitate member participation, breaking away from the hierarchical structures of traditional parties. They are characterised by technological centralism, in which a charismatic leader or a small ruling group controls decisions through digital tools.
In The Platform Party (2019), Gerbaudo further develops this idea by describing how these parties operate like platforms, inspired by digital companies (such as Facebook or X), where activist engagement is often limited to online interactions (votes, surveys, discussions) rather than active participation on the ground. He criticises a form of ‘participationism’ that is more symbolic than genuine: although these parties claim to promote more direct and horizontal democracy, control remains concentrated in the hands of a few leaders who manage the platform. He defines a hyperleader as a plebiscitary-charismatic figure tasked with representing the party in the media and internet spectacle, by attending TV talk shows and intervening obsessively on social media. Through his histrionic performances, the hyperleader makes up for the lack of a strong and dependable organisation. (2019: 21)
This led Deseriis (2020b) to emphasise the need to distinguish between two subtypes of digital parties, as not all digital parties are organised around a charismatic leader. He refers to them as platform parties when this is the case and suggests using the term ‘network party’ when the coordinators of a social movement have exploited the resources of ICTs to federate the partisan base and create a political party. As we shall explain later, in this article, we use the term ‘digital party’ (in the sense of ‘platform parties’) because French digital parties are organised around charismatic leaders rather than social movements. We can expect that the organisation of digital parties around a charismatic hyperleader would have them to amplify the ‘personalisation’ of their political communication on Facebook, by focusing attention on their candidates rather than on their parties (Enli and Skogerbø, 2013). This personalisation is often staged using posts that share press articles highlighting the candidate, in particular their influence on national or international public opinion. This is why the question of the personalisation of electoral communication on social media is intrinsically linked to the media coverage of the campaign.
H1a: To increase their candidate's media visibility, digital parties are more likely to share posts focusing the attention on their leader.
Social media has increasingly influenced media coverage of election campaigns. However, research on intermedia agenda-setting shows that traditional media continue to dominate the news agenda during elections (Harder et al., 2017), particularly on social media. For example, during the 2017 French presidential campaign, the majority of information shared by political parties on their Facebook pages came from mainstream press websites (Figeac et al., 2020). Parties often share these articles to promote their candidates by highlighting the success of their rallies or of their media presence on divisive issues. As digital parties are organised around a ‘hyperleaders’ (Gerbaudo, 2019), we can expect that they use social media to promote their leader's image more than traditional parties do.
H1b: To increase their candidate's media visibility, digital parties are more likely to share posts about their candidate's campaign media coverage.
Participationism in the electoral communication of digital parties
Scholars have found that new parties place a higher value on social media than other parties do, particularly valuing the platform's ability to facilitate voter interaction and mobilisation (Gibson and McAllister, 2015; Lilleker et al., 2015; Larsson, 2016). This trend is expected to be even more pronounced in digital parties, given the centrality of ICT in their internal organisation, which relies on a high level of commitment from the militant base (Barberà et al., 2021; Deseriis, 2020a; Gerbaudo, 2019). This leads us to believe that the mass participation of these parties’ supporters in social media translates into a higher number of comments and a greater tendency to share or ‘like’ content.
H2a: Digital parties are more likely to use social media to promote voter interaction (like, share, comment) and mobilisation.
More specifically, scholars have assessed whether young parties make greater use of social media to mobilise their followers by sharing calls for mobilisation. These calls can be defined as social media posts that ‘provide information on where or how citizens can take political action, including links to petitions or calls to join a political protest’ (Heiss et al., 2019). There is still a lack of evidence on the distinction between offline (e.g. rallies and leaflet distribution) and online (e.g. signing petitions, requests to forward information and participating in live tweets) calls for mobilisation (Segesten and Bossetta, 2017). Here, we analyse both types of calls, as they can support the election campaign by bringing together more supporters for offline events. They might also stimulate user engagement on social media by encouraging parties’ supporters to comment on or share such posts with their network of followers, which in turn encourages them to get involved. However, previous research suggests that this is not the case: such calls may have no effect or even a negative one on user engagement (Bene, 2017; Heiss et al., 2019; Xenos et al., 2017). Despite this result, digital parties may stand out from this trend because they are more dependent on social media resources to run their campaigns, and their support networks are less established outside major cities (Neihouser and Figeac, 2024). These contextual elements lead us to defend the hypothesis that these on- and offline calls for mobilisation are over-represented in the posts of digital parties. In particular, the Facebook pages run by party members outside of the major cities.
H2b: Digital parties are even more likely to use on- and offline calls for mobilisation, particularly members of parties running Facebook pages outside major cities.
Political issues in the electoral communication of digital parties
The concept of a digital party does not merely refer to technical issues or the way ICTs are used by these political movements to engage their supporters (H2) in backing their leader's candidacy (H1). This term refers to a crisis in the representation of traditional parties, which are unable to meet citizens’ expectations regarding participation. Digital parties respond more effectively to these expectations through their use of participatory digital platforms. Scholars believe they also succeed because they champion new ideas related to campaign themes that matter to some citizens (Gerbaudo, 2018, 2019), such as new forms of democratic participation and new welfare provision. Therefore, it appears that digital parties are more inclined to utilise social media to discuss traditional political issues than traditional parties (Gibson and McAllister, 2015; Larsson, 2016). However, there still exists a lack of empirical evidence on this trend.
H3a: Compared to other parties, digital parties are more likely to use social media to inform about traditional political issues.
However, it also appears that parties tend to emphasise in their communication those issues that are central to their political agenda and that differentiate them from their opponents (Walgrave et al., 2015). For example, it's well known that populist parties have been successful in attracting supporters and establishing themselves in the political landscape by focusing on controversial issues such as identity and immigration policies (Kriesi et al., 2008). This pattern is also observed in social media, where parties treat political issues differently depending on their party ideology. Most research on Western countries (Bene, 2022) sets apart the information-sharing behaviour of extreme right-wing parties because of the over-representation of divisive issues, particularly immigration. In contrast, while scientists also measure political issues that are over-represented in far-left publications, international comparisons do not identify any ones in particular (Bene, 2022).
H3b: The over-representation of information on political issues in the online communication of digital parties depends on their party ideology.
Scholars have suggested that digital parties don’t just cover more political topics on social media (Gerbaudo, 2018). Because they are internally networked, they find it easier to use these platforms to engage in narrative battles over divisive issues. These polarising communication strategies are more characteristic of a subtype of digital parties, the ‘digital populist parties’ (Peña and Gold, 2023), that position their alternative political offer on divisive issues. There remains a lack of evidence to determine whether digital parties, particularly populist ones, are more inclined to address divisive issues on social media.
H3c: Divisive issues are over-represented in the publications of digital parties, especially populist ones.
France: a political context transformed by the emergence of digital parties
The political context of the 2022 French presidential election provides an ideal scenario for analysing how digital parties specifically manage ICT resources compared to other parties. Indeed, this campaign involved six main candidates, three of which emanated from digital parties: LFI, REM and R! (cf. Table 1).
Party organisation of the most important French political party.
LFI (La France Insoumise, Far-Left) and REM (République En Marche, Centre) emerged a year prior to the 2017 presidential election. In 2021, another digital party was formed, the far-right party Reconquête (R!), relying heavily on ICT resources. These parties retain several defining features of digital parties. Firstly, we consider these political movements to be digital parties because they rely on a superbase of activists who organise their internal structure using ICTs. Secondly, they are organised around ‘hyperleaders’ (Gerbaudo, 2019; Deseriis, 2020b) with the aim of winning the presidential election for them: Jean-Luc Mélenchon at LFI, Emmanuel Macron at REM (the current president of the republic) and Éric Zemmour at R! Thirdly, these leaders have positioned their political agendas and proposals in the media landscape as alternatives to traditional mainstream parties. REM challenges the left-right divide by calling for a renewed approach to social rights; LFI brings together various radical left parties by prioritising the fight against tax evasion and the defence of minorities; and R! positions itself against the established far-right party (Rassemblement National) by promoting the ‘Great Replacement Theory’ and advocating for stricter migration policies.
In the 2022 election, three traditional parties also received a significant number of votes: EELV, LR and the RN.
Les Républicains (LR) is the historical party of the right. It was created in 2015 when the Union pour un Mouvement Populaire (UMP) changed its name and statutes. The UMP itself was founded in 2002 as a merger of the then-main right-wing party, Rassemblement pour la République (RPR, founded in 1976), and its key allies.
Europe Écologie Les Verts (EELV) is a left-wing ecological party founded in 2010 through the merger of Les Verts (established in 1984) and the Europe Écologie coalition. In 2023, the party rebranded itself as Les Écologistes to reflect a broader ecological movement.
Finally, the Rassemblement National (RN) is a far-right party. It was created in 2018 when the Front National (FN), founded in 1972 by Jean-Marie Le Pen, changed its name.
Data and methods
Data
To test our hypotheses, we conducted a content analysis of party Facebook posts during the 2022 French presidential election campaign. France has a mixed electoral system, and we considered candidates who received at least 4% of the vote in the first round of the election: Emmanuel Macron (centrist, REM, 27.84%), Marine Le Pen (far-right, RN, 23.15%), Jean-Luc Mélenchon (far-left, LFI, 21.95%), Eric Zemmour (far-right, R!, 7.07%), Valérie Pécresse (right, LR, 4.78%), Yannick Jadot (left, EELV, 4.63%). We collected 17,060 posts from the Facebook pages (N = 544) managed by these candidates and their supporter networks from 1 January 2022 to 24 April 2022 using the Facebook API. In defining the perimeter of this data set, we then identified three main types of Facebook posts: those shared by the candidates (or their community managers) on their personal Facebook accounts (N = 2941), those published on the Facebook pages of political parties (N = 8889) and those relayed by the Facebook pages of the supporters and the party members (N = 5230) most often at a regional level.
Statistical models
The unit of analysis in this study was Facebook posts, and the dependent variable was Facebook posts published by digital parties. This allows us to determine other parties. To obtain valid results, we control for a wide range of variables. Thus, our dependent variables are directly derived from the characteristics of Facebook profiles (FPC) or the characteristics of posts (FpC). We used multilevel modelling to test our hypotheses by controlling for the effects of the ‘Facebook Profile Characteristics’ (FPC) on model estimates at the level of the ‘Facebook posts Characteristics’ (FpC) (Bene et al., 2022; Heiss et al., 2019). This approach allowed us to account for the influence of the characteristics of Facebook profiles (FPC) on the model estimates in relation to the Facebook posts published by digital parties.
Variables based on the Facebook profiles characteristics (FPC)
First, we will examine whether the use of Facebook by digital parties varies according to their ‘party ideology’. We distinguish between a far-left party (LFI), centrist party (REM) and far-right party (R!). Second, we compare the parties’ digital campaigns on the basis of ‘digital campaign funding’ (ranging from €158,600 to €654,117; M = €318,917) and ‘number of votes’ in the 2022 French presidential election campaign (ranging from 1,627,853 to 9,783,058; M = 4,852,228). Third, we identified the ‘types of Facebook profiles’: the ‘official pages’ managed by party community managers, the ‘personal pages’ of candidates and the ‘members pages’ created by party members to support their candidates. Fourth, we distinguish according to the ‘scope of action of the Facebook pages’, distinguishing between ‘national pages’ and ‘regional pages’.
Variables based on the Facebook posts characteristics (FpC)
First, ‘traditional political issues’: Based on a qualitative analysis of a randomly selected sample of posts (N = 1500), two annotators identified eight central issues raised by the parties during the 2022 French presidential campaign: economic policy, environmental policy, social policy, education policy, health policy, security policy and foreign policy.
Second, ‘divisive issues’: We differentiate between information on ‘migration policy’ and ‘disruptive events’ introduced into the campaign by the media in accordance with current affairs, such as the COVID-19 pandemic and the war in Ukraine.
Third, ‘on and offline mobilisation’: During the political campaign, parties share messages to coordinate their offline (meetings, door-to-door campaigning, etc.) and online campaigns (e-petitions, etc.) (Heiss et al., 2019) on Facebook.
Fourth, ‘self-centred promotion of the candidate’: Political communication often aims to promote the candidate by highlighting his or her personality, skills and ability to win and defend the party's ideas.
Fifth, ‘media coverage of the campaign’: It is worth noting whether digital parties provide more information about the media coverage of their candidate's campaign.
Sixth, ‘campaign phase’: As media coverage of the campaign is more prominent a few weeks before the election, we distinguished two phases of the campaign: the pre-campaign (1 January 2022–28 February 2022) and the campaign phase, when media coverage intensifies (1 March 2022–24 April 2022).
Seventh, ‘visual content’: We distinguished between visual (video and photo) and non-visual content in parties’ political communication.
Eighth, ‘post length’: We also measured the effect of text length (between 1 and 4999; M = 188; SD = 408).
Finally, ‘user engagement’: The Facebook data specify user engagement with politicians’ Facebook posts by referring to the number of likes (ranging between 0 and 91,371; M = 1021), comments (ranging between 0 and 57,587; M = 226) and shares (ranging between 0 and 32,743; M = 231).
Supervised machine learning method
Variables based on the characteristics of Facebook posts were created using a supervised content analysis method based on machine learning, which involves a two-stage process (Bene, 2021). The first stage was performed by two annotators to create a gold standard annotated set on a randomly selected sample of 1500 posts (8.8% of the posts in the main dataset). All of these posts were labelled by referencing them under 13 categories (cf. Table 2). Despite the diversity of topics, a significant degree of agreement was found between the two annotators, as shown by a Cohen's kappa index of 0.94 across all labels and a Krippendorff's alpha coefficient of 0.955. In the second stage, this gold standard set was divided into a training set (80%) and a validation set (20%). The training set was used to train an algorithm to identify the patterns liking the messages to a unique label based on the data from the training set. Once trained and given its good performance, the algorithm was used to extend the labelling process to the entire dataset.
Performance metrics for semi-supervised detection of each topic.
The algorithm used was a pre-trained French language model: ‘CamemBert’. The model was then fine-tuned on the training set to learn to label each post, with one of the 13 categories represented as numbers ranging from 1 to 13. Several trainings were launched in parallel with different hyperparameters (number of epochs, batch size and learning rate), and the model with the best accuracy on the validation set was selected (Table 2). The fined-tuned model exhibits a consistent precision of 76% on the validation set and an F1-score of 0.745.
Results
Our findings compare the use of Facebook by digital parties with that of traditional parties to determine whether their external political communication on social media is primarily structured around: (H1) hyperleaders and media coverage, (H2) participationism and activist engagement and (H3) thematic differentiation, particularly regarding divisive issues. These results contribute to a broader understanding of how digital parties position themselves within contemporary democratic systems and how their communication strategies may or may not signify a fundamental shift from traditional party practices.
Digital parties and political communication based on hyperleaders
Contrary to our expectations (H1a), the self-centred communication of the candidate (SCCC) is one of the two most characteristic features of the political communication of traditional parties rather than digital parties (Figure 1). Digital parties do not necessarily personalise their messages more, despite the fact that they are organised with the aim of bringing a charismatic leader to power. This trend can be explained by the fact that traditional party leaders are often perceived as uncharismatic and as veteran strategists who have waited many years for a successful appointment as party leaders. As a result, traditional parties’ political communication on social media is primarily characterised by a strategy of emphasising the merits of their candidates, especially in the first phase of the campaign. Conversely, the SCCC seems to be less important in the publications of the three French digital parties because their leaders have high levels of media visibility.
Logistic regression with random intercepts estimates the ‘Facebook posts Characteristics’ (FpC) of the digital parties (vs. other parties). #p < .1, *p < .05, **p < .01, ***p < .001. Standard errors are in parentheses. If no reference category is mentioned, variables were coded as present or not present. All numeric independent variables including the log-transformed variables (‘log’ in parentheses) were mean-centred for convergence reasons. See Table 1 in the supplementary materials.
Digital parties also tend to use Facebook more to increase the media coverage of their candidate's campaign (H1b; Figure 1). However, it appears that this is less the case for the ruling party (REM; see Figure 2), as political journalists often comment on the president's agenda. This topic generated the most comments, particularly on party members’ pages, where they gave their views on their candidates’ television appearances.
Logistic regression with random intercepts estimates the main topics of the posts published by the French political parties on Facebook. #p < .1, *p < .05, **p < .01, ***p < .001. Standard errors are in parentheses. If no reference category is mentioned, variables were coded as present or not present. All numeric independent variables including the log-transformed variables (‘log’ in parentheses) were mean-centred for convergence reasons. See Table 2 in the supplementary materials.
Lastly, as with the SCCC, the messages about the media coverage of the campaign generate the most user engagement in terms of likes and comments (Figure 2). The SCCC also characterises the contributions of local support pages with the least number of followers. This finding suggests that the political communication of regional support networks is often rooted in the personal characteristics of their candidate, his image and the elements that contribute to his charisma.
Digital parties and the user engagement of the superbase of activists
First, our results confirm our hypothesis that digital parties are more likely to use Facebook to coordinate both on- and offline mobilisation (H2b; see Figure 1). Their campaign communication on Facebook aims to stimulate the participation of their super base of activists, confirming the findings already suggested by studies of the internal communication of digital parties (Gerbaudo, 2018). However, they do not rely on diverse types of content, such as videos and photos, any more than traditional parties do to structure their external political communication. This challenges the assumption that digital parties strategically leverage a wider variety of content formats to enhance engagement, as suggested by previous research on online party strategies (Staender et al., 2019; Casas and Williams, 2019; Koc-Michalska et al., 2021).
However, our findings show that the communication of digital parties varies according to their position in the government. Only the governing party (REM) is over-represented in offline coordination compared with the other parties (Figure 2). This trend can be explained by the fact that the exercise of power at the national and regional levels has allowed this digital party to build a denser network of activists across the country. During electoral periods, REM's communication managers use Facebook to publicise the activities organised in their region. This use of social media differs from that of LFI, which shares more coordination posts online to encourage its supporters to organise election activities locally.
Our results also validate our expectations by demonstrating that the strong ‘participationism’ of the superbase of activists most distinctly characterises the electoral communication of digital parties on social media (Gerbaudo, 2018). Compared to traditional parties, the most active Facebook pages of digital parties are those run by supporters, not the official or candidate pages (Figure 1). This trend is strongly corroborated by the number of likes, which is higher on these pages. Despite this, our findings do not confirm the hypothesis (H2a; see Figure 1) that the Facebook pages of digital parties would generate more user engagement. This may be because supporters of these parties share fewer publications with their network of Facebook friends and tend to comment on them in the same way as supporters of traditional parties do. This leads us to consider that the mass participation of supporters in the internal organisation of these political parties (Gerbaudo, 2018) may not necessarily be observable on social media.
Digital parties and political issues differentiation
First, our results confirm one of our expectations (H3a) by showing a slight over-representation of traditional political issues in the Facebook posts of digital parties (see Figure 1, Model 1). Digital parties address foreign policy more often, and they are less likely to address environmental and security policy.
This trend varies according to political ideology, as traditional political issues are under-represented in the posts of the far-left digital party and centrist digital party (see Figure 2), which validates H3b. Our findings also confirm another of our expectations by showing an over-representation of divisive issues in the publications of the far-right digital parties (H3c, cf. Figure 1, Model 2). It appears that the issue of immigration is strongly over-represented only in the publications of this party (R!), founded by Éric Zemmour in response to the so-called laxity of the traditional far-right party (RN). Zemmour proposed a tightening of immigration policy in order to stop what he calls the ‘great replacement’ of the French population by immigrants. Conversely, this divisive issue is under-represented in the political communication of the other digital parties (LFI and REM). The same trend holds regarding the divisive issues introduced into the campaign by disruptive events. Political parties in France shared conflicting messages around the war in Ukraine and the COVID-19 pandemic. While the president's party praised the government's handling of these two events, the opposition parties criticised the president's handling of the situation. Our results show significant differences between digital parties on these divisive issues, as they are only under-represented in the communication of the far-right party (Figure 2). This trend confirms the value of differentiating between various subtypes of digital parties, such as the ‘populist digital parties’ (Peña and Gold, 2023), which are more likely to address divisive social issues on social media in order to present themselves as alternatives to traditional (populist) parties.
Discussion
The findings of this study offer substantial insights into the external political communication strategies of digital parties, particularly regarding hyperleaders, participationism and the differentiation of political issues. These results refine the existing theories on digital parties and challenge some common assumptions about their campaign strategies. They contribute particularly to a broader understanding of the ongoing crisis of political representation and of how digital parties – not just populist ones – are reshaping the landscape of party politics.
Hyperleaders and the personalisation of electoral communication
While digital parties are often associated with highly centralised leadership structures, our findings suggest that they do not necessarily emphasise their leaders more than traditional parties do in their electoral communication (H1a). Instead, traditional parties may use candidate-centred messaging as a compensatory mechanism for their less media-visible leaders. This highlights the need to reconsider the extent to which digital parties rely more on leader-focused messaging in their electoral campaigns than their traditional counterparts (Gerbaudo, 2019; Deseriis, 2020b). Rather than systematically relying on hyperleadership, digital parties appear to leverage digital platforms primarily to amplify or counterbalance mainstream media narratives (H1b). This trend is particularly evident among opposition parties, such as La France Insoumise (LFI) and Reconquête (R!), reinforcing the idea that digital platforms serve as a strategic tool for outsider parties (Deseriis, 2020b).
Participationism and activist engagement
One of the key assumptions regarding digital parties is that they foster higher levels of engagement through participatory online infrastructures (Gerbaudo, 2018). While our findings confirm that digital parties are more likely to orchestrate both online and offline mobilisation (H2b), they challenge the notion that these parties necessarily generate stronger user engagement than traditional parties do (H2a). One possible explanation for this paradox lies in the nature of digital participation itself. While digital parties claim to empower their supporters, scholars argue that the latter's participation often remains highly structured and centralised (Gerbaudo, 2019), while digital elites maintain significant control over engagement rather than fostering horizontal interactions. Moreover, Heiss et al. (2019) highlight that liking a post constitutes a low-effort form of engagement compared to commenting or sharing. This suggests that while digital party supporters may be highly present on social media, their participation remains largely passive, reinforcing the idea that digital activism within these parties does not always translate into higher levels of visible engagement.
Beyond this, our findings also highlight the key role played by local support networks in the online communication strategies of digital parties. Compared to traditional parties, digital parties’ presence on Facebook is more decentralised, with significant activity occurring on pages run by party members rather than on official party accounts. This pattern is particularly evident on regional Facebook pages, where party activists in less populated areas adapt national digital strategies to local political contexts (Neihouser and Figeac, 2024) in the vein of citizen-led campaigns (Gibson, 2015). This finding challenges the dominant view of digital parties as strictly hierarchical structures centred around hyperleaders (Gerbaudo, 2019; Deseriis, 2020b). Rather than enforcing a fully centralised model of digital communication, these parties appear to allow a degree of grassroots mobilisation, raising questions regarding the degree to which hyperleaders maintain control over their activist base in electoral contexts.
Finally, our findings show that the role of participationism in digital parties is shaped by their status as governing or opposition parties. REM, as the ruling digital party, prioritises offline coordination, whereas LFI relies more on online coordination, reflecting differences in the ways in which these parties use digital platforms to sustain engagement. This distinction aligns with research suggesting that digital parties are not a homogenous category but instead develop distinct strategies based on their institutional positioning and available resources (Peña and Gold, 2023).
Political issues and ideological differentiation
Our results partially confirm the expectation (H3a) that digital parties use social media more frequently to discuss political issues than traditional parties do. However, this over-representation is relatively modest and varies according to the party's ideological orientation (H3b). In particular, the far-right digital party (R!) prioritises divisive issues in its online communication (H3c). The over-representation of immigration in R!'s posts aligns with existing literature on digital populist parties (Peña and Gold, 2023), which suggests that these parties leverage social media to polarise political debates and set themselves apart from traditional parties.
In contrast, the governing digital party (REM) engages more with policy-related controversies, such as COVID-19 and the war in Ukraine. This finding enriches broader debates on the evolution of digital campaigning by demonstrating that digital parties tailor their social media engagement to reflect their institutional role, ideological positioning and media presence
Limitations and future research
Despite these insights, this study has some limitations. First, our analysis focuses solely on Facebook, which, while being the most widely used platform for political content consumption in France (Neihouser et al., 2023), does not cover the full spectrum of online campaign strategies. While Facebook plays a central role in structuring political discourse and mobilisation, digital parties also leverage other platforms, such as X, Instagram and TikTok, each with distinct affordances and audience demographics. For instance, Instagram and TikTok, which have younger user bases, may encourage more visual and interactive forms of engagement (Larsson et al., 2024), potentially leading to different communication styles and strategies. Future research should explore these differences across various social media platforms to assess whether the patterns observed on Facebook are replicated or not elsewhere.
Second, our study is limited to the French context, where digital parties with varying ideological orientations and political statuses (government vs. opposition) are well-established. While the French semi-presidential system distinctly shapes electoral competition, making the presidential election highly personalised and media-driven, it remains a key moment for the analysis of party communication. This electoral context allows us to assess how digital parties structure their messaging in a high-stakes campaign, where the interaction between party dynamics and leadership-centred communication is most pronounced. However, future research could explore whether these findings hold in other electoral settings, such as legislative elections, where parties may adopt different communication strategies less centred on their candidates.
Conclusion
The transformation of digital parties is part of a broader context of a crisis of political representation characterised by a growing distrust of traditional party structures and an increasing demand for direct citizen participation in decision-making (Dalton and Weldon, 2005; Gerbaudo, 2018). Unlike traditional parties, which rely on hierarchical structures and conventional intermediary mechanisms (party sections, local activists, partisan elites), digital parties seek to address this crisis by adopting more flexible forms of organisation, leveraging digital tools to facilitate direct engagement from members and supporters. However, our study adds nuance to this perspective by showing that, while digital parties heavily mobilise social media, their communication strategies on Facebook remain partially dependent on their political positioning, electoral role, ideological framework and the logic of traditional media. Furthermore, they rely more on text-based content than traditional parties, using fewer videos and images. Instead of indicating a complete departure, digital parties seem to integrate traditional political communication methods with digital possibilities. This implies that their evolution is more of a partial adjustment than a total overhaul of partisan involvement, questioning the belief that digital parties significantly alter the electoral framework (Deseriis, 2020; Peña and Gold, 2023).
Supplemental Material
sj-docx-1-ejc-10.1177_02673231251340803 - Supplemental material for The digital party paradox: Behind participationism, the persistence of traditional political repertoires on Facebook
Supplemental material, sj-docx-1-ejc-10.1177_02673231251340803 for The digital party paradox: Behind participationism, the persistence of traditional political repertoires on Facebook by Julien Figeac, Marie Neihouser and Ferdinand Le Coz in European Journal of Communication
Footnotes
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Agence Nationale de la Recherche (grant number ANR-11-LABX-0066).
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
