Faces of Europe: Structural Drivers of Visual Personalization in Political Parties' Facebook Campaigns.
1. IntroductionScholars have long examined political personalization, defined as an increasing focus on individual politicians rather than parties or institutions (Rahat & Sheafer, 2007). As party loyalties decline and politics becomes more individualized and lifestyle-oriented (Bennett, 2012), politicians increasingly construct personal images that transcend traditional party boundaries (Metz et al., 2020). The democratic implications of this trend remain contested. Personalization has been linked to rising populism, democratic erosion, and polarization (Frantz et al., 2021). Critics argue that personalization shifts attention from parties and policy issues to individuals. It is also argued that privatization, as a sub-type of personalization, oversimplifies politics, limits rational opinion formation, and reduces citizens' access to substantive information (Adam & Maier, 2010). Such concerns may be intensified in visual communication, where space for political issues is inherently constrained. Personalization may also weaken accountability and fuel cynicism by prioritizing charisma over institutions and blurring public-private boundaries (Pedersen & Rahat, 2021; Van Aelst et al., 2012). These dynamics can benefit populist leaders, who often rely on images, emotional appeal, and anti-establishment rhetoric. Langer (2007) speculates that they may increase politicians' vulnerability to shifts in public sentiment and media framing. Gendered forms of personalization may further reinforce stereotypes and undermine women's political authority (Magin et al., 2024).
Other scholars, by contrast, highlight potential democratic benefits of personalization. Presenting complex issues through politicians' personalities can simplify them and increase comprehensibility, showing that personalization and substantive communication are not inherently contradictory: Citizens' political evaluations often combine judgments of personal traits with assessments of policy positions, suggesting that personalization and issue engagement can be mutually reinforcing. In this perspective, privatization is viewed more negatively than individualization because it shifts attention from political issues to politicians as private individuals (Adam & Maier, 2010). In a context of declining institutional trust, personalized communication may also narrow the gap between elites and citizens by leveraging the interactive affordances of digital media (Pedersen & Rahat, 2021). From this angle, personalization has the potential to enhance political interest and knowledge among less engaged citizens and stimulate participation in distrustful environments (Papacharissi, 2014).
Legacy media has long contributed to political personalization by highlighting politicians' private lives (Van Zoonen, 2005), but social media is widely seen as intensifying this trend (Ekman & Widholm, 2017; Enli & Skogerbo, 2013; Metz et al., 2020; Peeters et al., 2023; Vuckovic, 2023). These platforms enable politicians to foster closeness with voters through personal, unscripted images (McGregor, 2018), which tend to be well received by followers (Farkas & Bene, 2021). The platforms focus strongly on visuals: Images are often better recognized and remembered than verbal content and convey nuanced social cues relevant for political decision making (Grabe & Bucy, 2009). To gain visibility on social media, politicians must meet algorithmic selection criteria, which has made visual personalization central in election campaigns. Much of this visual personalization stems from so-called self-personalization (McGregor, 2018)--strategic displays of personal or family life to align with media logics favoring individual personalities over institutions (Stromback, 2008). Sharing elements of politicians' lifestyles can foster intimacy and political identification by reflecting voters' values and everyday experiences (Mazzoni & Mincigrucci, 2022).
While research on social media campaigning has grown rapidly, it still relies heavily on single-country studies or comparisons of only a few countries (Boulianne & Larsson, 2024). This narrow focus limits understanding of how campaign strategies, while shaped by similar platform logics, remain embedded in national political contexts (Ha[beta]ler et al., 2021). Reflecting this gap, scholars have repeatedly called for more cross-national research on social media campaigning (e.g., Bene et al., 2024; Boulianne & Larsson, 2024), including in visual political communication (de-Lima-Santos et al., 2024) and candidates' visual self-presentation (Steffan, 2020). Comparative research is needed to reveal how structural factors systematically shape the content of social media campaigns.
This study responds to this call by providing the first large-scale, cross-national comparison on how party-and country-level factors systematically affect visual personalization. This is measured through a content analysis of individualization and privatization captured via the public-facing Facebook campaigns of 138 parties and party coalitions across 23 EU member states. As Facebook remains one of the most widely used campaign platforms (Darius et al., 2024; Dommett et al., 2024), it provides a valuable and substantively important setting for examining how structural factors shape visual communication. We investigate how party-level (populist vs. non-populist; government vs. opposition) and country-level factors (electoral system; degree of authoritarianism)--the latter not previously studied in relation to visual personalization--affected campaign strategies in the 2024 European Parliament (EP) elections. The near-simultaneous EP elections across diverse political and media contexts create a unique opportunity to assess how these structural factors influence visual campaign practices. Most previous personalization research has focused on candidates (e.g., McGregor, 2018; Metz et al., 2020), yet parties are the central actors in most European political systems and in EP elections. Because European candidates' campaigns are typically closely coordinated with party strategies, a party-focused approach offers broader insights into campaign dynamics.
Our findings challenge widespread assumptions about populist and authoritarian communication styles and reveal that different dimensions of visual personalization respond differently to party- and country-level structures. This demonstrates the need for a more explicitly multidimensional approach to personalization, and for theories that account for the contextual drivers of each dimension.
2. Visual Personalization
Personalization of politics refers to a contemporary state of affairs, but is also a concept that has been used to discern a long-term globalized trend toward increasing personal focus (Adam & Maier, 2010). Our cross-sectional design captures personalization as a state, though we assume social media has contributed to its rise over time. Scholars differ on whether personalization is episodic and shaped by leadership styles and political contexts (Rahat & Sheafer, 2007) or a lasting feature of modern politics influencing electoral communication and leadership over recent decades (e.g., Grbesa, 2008; Holtz-Bacha et al., 2014). Although widely regarded as multidimensional, the concept has been operationalized in various ways, contributing to mixed empirical findings. Some authors distinguish professional, emotional, and private personalization (Metz et al., 2020; Russmann et al., 2024; Van Santen & Van Zoonen, 2010), while others focus on individualization and privatization (Van Aelst et al., 2012). We adopt this latter, well-established, clear-cut two-dimensional model, which is better suited for cross-country comparisons and applies to both textual and visual content, the latter being our focus.
2.1. Key Dimensions of (Visual) Personalization
Van Aelst et al. (2012) define individualization as a shift in focus from parties or institutions to individual politicians, with two dimensions: general visibility (emphasis on individuals over groups) and concentrated visibility (focus on leaders; see also HaBler et al., 2024). The latter aligns with the concept of a presidentialization of non-presidential political systems, whereby party leaders or top candidates increasingly dominate political communication. Individualization as a rhetorical strategy highlights politicians' competencies and proposals, typically through formal, professional portrayals (Farkas & Bene, 2021; Pedersen & Rahat, 2021).
Van Aelst et al. (2012) define privatization as emphasizing politicians' private lives. It involves politicizing the personal self (Langer, 2007) and intimizing political communication by bringing personal values, relationships, and experiences into the public sphere (Stanyer, 2013; Van Zoonen, 2005). Privatization aims to show the "human face" of politicians (Rahat & Sheafer, 2007) by highlighting personal characteristics and private lives (Ha[beta]ler et al., 2024), often through informal settings such as casual clothing or spontaneous moments (Farkas & Bene, 2021). Holtz-Bacha (2004) argues that privatization can serve various functions such as humanization, simplification--including oversimplifying or diverting attention from substantive issues--emotionalization, and the pursuit of celebrity status (Street, 2004).
2.2. (Visual) Personalization on Social Media
Research on political personalization on social media has focused mainly on either political actors' communication strategies or users' reactions. Regarding the latter, numerous studies show that personalized--especially privatized--content increases user engagement: Visual self-personalization (Metz et al., 2020; Russmann et al., 2024), references to politicians' private lives (Metz et al., 2020; Peeters et al., 2023), emotionally charged posts (Bene, 2017), and emotional or visual appeals in politicians' messaging (Jost, 2023) all enhance online engagement.
Regarding the strategic use of personalization we investigate here, it has been shown that politicians, including members of the EP, are active on social media (e.g., Daniel & Obholzer, 2025), suggesting at least some degree of individualization. However, research also shows that many remain hesitant to adopt highly intimate visual strategies (Filimonov et al., 2016; Ha[beta]ler et al., 2024; Hrbkova & Mackova, 2021; Russmann et al., 2024), despite the engagement potential of personalized content. Even Barack Obama's social media presence used personal cues only sparsely (Ha[beta]ler et al., 2024; Vuckovic, 2023). By contrast, about one-third of posts during the 2016 Swedish election were classified as "everyday private" (Ekman & Widholm, 2017). Such mixed findings may stem from different operationalizations but also align with Rahat and Kenig's (2018) observation that personalization varies across national contexts. With data from 23 countries and consistent measurement instruments, the present study allows us to assess how personalization depends on party and country contexts.
2.3. Structural Factors Influencing Parties' Strategic Use of Visual Personalization
What parties publish on social media is part of their broader strategic campaigning. Research on strategic party behavior (e.g., Muller & Strom, 1999) shows that parties for example emphasize issues depending on whether they are vote-, office-, or policy-seeking (Strom, 1990). We view visual personalization as a strategic campaigning tool. While we cannot infer parties' motivations directly from social media content, we can examine how structural contexts shape their likelihood of using visual personalization, offering insights into factors influencing strategic decisions.
As this indicates, election campaigning does not take place in a vacuum. Parties make strategic decisions within opportunity structures at both the meso level (party characteristics) and macro level (country contexts; Esser & Stromback, 2012), which can encourage or discourage individualized or privatized visuals. While many contextual factors could influence the strategic use of visual personalization, we focus on four variables--two party-level and two country-level--that are either grounded in prior research or derive from clear conceptual expectations about how opportunity structures shape campaign strategies. At the party level, we examine populist orientation and government status, as these shape parties' strategic incentives and align with our party-centered perspective, which treats visual personalization on party pages as coordinated organizational choices rather than individual candidates' self-presentation. At the country level, we assess the electoral system and degree of authoritarianism, two structural factors that predate the campaign, shape the broader incentive environment in which parties operate, and vary significantly across EU member states.
All four factors lie outside campaign communication itself, helping us avoid circular explanations driven by campaign-specific dynamics. Moreover, their cross-national data availability makes them well-suited for a comparative design, enabling us to examine how broader institutional contexts influence parties' incentives to use visual personalization.
2.3.1. Party-Level Factors
2.3.1.1. Populist Parties
Research on personalization has focused heavily on right-wing populists (e.g., Bast, 2024; Gonzalez-Aguilar et al., 2023), but populism extends beyond the political right (de Vreese et al., 2018). Mudde (2004, p. 543) defines populism as "a thin ideology" dividing society into two homogeneous, antagonistic groups--"the pure people" and "the corrupt elite"--which "argues that politics should be an expression of the volonte generale of the people." The definitions of both groups vary according to the broader ideology (e.g., liberalism, nationalism) to which populism is linked (Bast, 2024; de Vreese et al., 2018). We follow this understanding and include the full range of populist parties in our analysis of visual personalization. The past decade has seen growing success for populist parties in Europe and beyond, with the rise of social media being considered a key factor (Gerbaudo, 2018).
A defining feature of many populist parties is their reliance on a strong, charismatic leader whose personal appeal is central to the party's success (Pappas, 2016). This tendency aligns closely with personalization (Rahat & Sheafer, 2007; Van Aelst et al., 2012). Populist leaders often construct a persona embodying the "common man," presenting themselves as "just like the people" (Bast, 2024). Social media posts showing leaders with ordinary citizens or engaging in everyday activities can reinforce this relatable image (Lalancette & Raynauld, 2019). This makes visual personalization a strategic tool through which populist parties can seek legitimacy, combining an authentic-seeming connection with an unfiltered projection of power. By using positive, highly personalized content, populist leaders can strategically manufacture legitimacy, normalize their authority, and secure public support (Szebeni & Salojarvi, 2022). Thus, we hypothesize:
H1: The visual elements in the Facebook campaigns of populist parties are more likely to show individualization (H1a) and privatization (H1b) than those of non-populist parties.
2.3.1.2. Governing vs. Opposition Parties
In party-centered systems, such as in most European countries, personalization tends to take a professionalized form, emphasizing leadership visibility and credibility rather than intimacy or emotional self-disclosure (Enli & Skogerbo, 2013). Moreover, EP elections are typically seen as second-order contests in which voters express dissatisfaction with national governments rather than decide on EU issues (Ehin & Talving, 2021). Research suggests that domestic dynamics continue to shape the behavior of parties, media, and voters in EP campaigns (van der Brug et al., 2021; van der Brug & de Vreese, 2016). As a result, EP elections are often framed as referenda on national government performance (Maier et al., 2024), placing incumbents in a vulnerable position and giving challengers incentives to target the government. On social media, governing and opposition parties therefore face distinct communicative challenges, which we assume lead to different visual personalization strategies.
Incumbents must sustain attention and trust in competitive, fast-moving environments (van der Brug & de Vreese, 2016). Individualization can serve as a tool of strategic control: By highlighting well-known office holders through professionalized, leader-centered imagery, they give institutions a human face while keeping communication firmly under organizational authority. Such individualized visuals help reinforce leadership recognition, continuity, and competence (Hopmann et al., 2011). Unlike these incentives for individualization, however, incumbents have weaker motivations to share privatized or informal imagery that might blur the boundary between personal and institutional roles.
Opposition parties, by contrast, lack an incumbent bonus and often have less well-known leaders, giving them fewer incentives than governing parties to use individualized visuals. For them, however, privatization can be a useful strategy to gain attention in the continuous flow of social media content and to differentiate themselves from governing actors. As Kollberg (2024) shows, challenger parties in EP elections compensate for their weaker institutional position through communicative differentiation, emotional and symbolic appeals, and greater willingness to experiment. With fewer reputational constraints on communication strategy, opposition parties may also use personalization more opportunistically to attract attention, signal authenticity, and mobilize emotion. These conditions favor more expressive and affective campaign styles, in which privatized visuals function particularly well as an attention-maximizing strategy on social media. We therefore hypothesize:
H2a: Government parties are more likely to use individualized visuals in their Facebook campaigns than opposition parties. H2b: Opposition parties are more likely to use privatized visuals in their Facebook campaigns than government parties.
2.3.2. Country-Level Factors
2.3.2.1. Electoral System
Visual personalization is most effective when it features candidates familiar to voters. In EP elections, however, candidates are often not well-known in their home countries (Gattermann, 2022). Electoral systems that allow voters to vote for individual candidates create incentives for parties and candidates to actively build candidate visibility, making visual personalization more beneficial. We therefore expect parties to use visual personalization strategically, particularly when voters can cast votes for individual candidates. Such opportunities depend on a country's electoral system.
Although EP elections follow uniform procedures and proportional representation, national electoral systems vary in how candidate-centered or party-centered they are (Hoyland et al., 2019). Candidate-centered systems encourage personalized campaigning (Bowler & Farrell, 2011) because they incentivize the personal vote (Carey & Shugart, 1995), leading individual candidates to compete directly with one another, including within the same party. This is typical of systems with open ballots. These systems also tend to have smaller electoral districts--i.e., fewer seats per district (Walter & Emmenegger, 2023)--which further encourages personalized campaigning since it is easier for candidates to reach voters personally and build individual visibility.
In party-centered systems, by contrast, voters either cannot vote for individual candidates when parties determine electoral lists (closed ballots) or are constrained by the pre-determined order of the party lists (ordered ballots). Electoral districts are usually larger, and constituencies may even be absent, as in the Netherlands. As a result, parties have weaker incentives to highlight individual politicians and, if anything, campaign more with lead candidates than with ordinary ones. We therefore hypothesize:
H3: In party-centered electoral systems, political parties are less likely to campaign with Facebook posts that contain individualized (H3a) or privatized (H3b) visual elements than in candidate-centered electoral systems.
2.3.2.2. Degree of Authoritarianism
Over the past decade, scholars have highlighted authoritarian and illiberal tendencies in several EU member states (Bakke & Sitter, 2022; Stetka & Mihelj, 2024), noting democratic regression (Markowski & Kotnarowski, 2025) and political actors adopting strategies that undermine liberal institutions, political norms, and the principles of pluralism and accountability (Enyedi, 2024). These developments are often associated with authoritarian figures in power and manifest in institutional and normative erosion (Szebeni & Salojarvi, 2022), normalizing previously radical ideas and moving them into the political mainstream (Berman, 2017).
To our knowledge, authoritarianism and visual personalization have not yet been directly linked in the literature. Logically, if authoritarian actors employ communication strategies similar to those of populist parties (see above), they are likewise incentivized to use visual personalization. The number of such actors in a country--and whether they are part of government--is therefore likely to influence the overall use of visual personalization, as other parties may react and (partially) adjust to these patterns. We thus expect the degree of authoritarianism in a country to shape how strongly political communication generally focuses on individual politicians and hypothesize:
H4: The higher the degree of authoritarianism in a country, the more likely parties' Facebook posts are to contain individualized (H4a) and privatized (H4b) visual elements.
3. Methods
3.1. Sample and Data Collection
To test our hypotheses, we conducted a manual standardized content analysis of posts published on the official Facebook pages of national parties during the 2024 EP election campaign across 23 of the 27 EU member states: Austria, Belgium, Croatia, Cyprus, Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Malta, Netherlands, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, and Sweden. Our aim was to include all 27 countries, but because the coding required knowledge of both national language and political context, we could only include countries where we could recruit collaborators with this expertise. For Bulgaria, Estonia, Lithuania, and Luxembourg, this proved not to be feasible. Despite this limitation, we contend that our sample provides sufficient variation and coverage to draw generalizable conclusions about parties' Facebook campaigning in the 2024 EP elections. It covers 97.5% of the EU population (439 million inhabitants; Eurostat, 2025) and spans diverse party systems, media systems, and electoral systems. This breadth enables a meaningful comparison of structural drivers of visual personalization across different political contexts.
Our sample includes posts from the official Facebook pages of all political parties that won at least one seat in the newly constituted EP and received at least 5% of the national vote, except for Ethniko Laiko Metopo (Cyprus), Kommounistiko Komma Elladas (Greece), and TISZA (Hungary), which surpassed 5% but did not maintain official Facebook accounts throughout the entire study period. In the case of coalitions, we included at least one party in the coalition and/or the coalition itself. In total, the sample comprises 149 parties and coalitions (referred to hereafter as parties; see Table A1 in the Supplementary File).
Although Facebook use is declining, it remained the most widely used social media platform in most EU countries in 2024--except Austria, Germany, and Spain (Newman et al., 2024; no data for Cyprus, Latvia, Malta, and Slovenia)--and continues to be one of the most important campaign platforms (Darius et al., 2024; Dommett et al., 2024).
We analyzed posts from official party pages rather than candidate pages to complement previous research which has largely focused on individual candidates (e.g., McGregor, 2018; Metz et al., 2020). This approach also reflects the fact that individual candidates in Europe generally operate within party-defined frameworks, which is particularly relevant in the party-centered EP elections. Crucially, analyzing party pages enables cross-country comparability: The presence and visibility of individual EP candidates vary widely across member states, while party pages exist in all countries and mirror the legally party-centered nature of EP elections, in which voters choose parties rather than individual candidates.
We used CrowdTangle to collect all posts published during the four-week "hot phase," ending one day after election day in each country. For countries where voting took place over several days, we used the last voting day. Due to differing election days, the investigation period differs slightly across countries (Netherlands: June 6; Ireland: June 7; Czech Republic, Latvia, Malta, Slovakia: June 8; all others: June 9). Full samples were coded for most countries. Due to research constraints, however, in Belgium, Finland, and Hungary we coded random samples of 50% of each party's posts, with a minimum of 50 posts per party unless a party had published fewer. This yielded different country-level sample shares (Belgium 63%, Finland 59%, Hungary 50%). In total, the dataset includes 15,447 coded posts (from 16,961 collected) from 149 parties. Because our focus is on visual communication, only posts containing images and/or videos were analyzed.
3.2. Measurements
3.2.1. Dependent Variables
Since we focus on visual personalization, we coded only images and videos contained in Facebook posts, but not the accompanying text. We coded the first image and the first video (up to one minute), reflecting the fleeting way many users consume social media content. For testing hypotheses H1-H4, we used the two dimensions of visual personalization--private personalization and individualization--as dependent variables, measured as follows (for examples of posts coded as individualized or privatized, see the Supplementary File).
3.2.1.1. Individualization
We measured individualization by combining two indicators: (a) the number of people visible in an image/video, and (b) the type of actors shown (e.g., politicians, ordinary citizens). Based on these measures, we created a dummy variable coded as individualized (= 1) when a post displayed one or two individuals and at the same time included at least one politician from the posting party (national politicians or European top candidates). Posts showing no people, more than two people, and/or no politician from the posting party were coded as not individualized (= 0). We focus on settings with one or two persons because we assume that individualization requires a limited number of visible individuals; in larger groups, the focal figure is typically unclear. While many national-level studies operationalize individualization via concentrated visibility (i.e., highlighting party leaders), this approach is difficult to apply consistently across EU member states, since the prominence of national and European lead candidates varies substantially. We therefore rely on the general visibility of party politicians as a more comparable cross-national indicator of individualization.
3.2.1.2. Privatization
We coded a binary variable indicating whether an image/video contained a private background story (= 1) or not (= 0), defined as showing politicians in a private context (e.g., with partners/family), presenting their interests or hobbies, or depicting their childhood or personal development through "throwback" images. Following Van Aelst et al.'s (2012) conceptualization of privatization--which distinguishes between personal characteristics and the presentation of private life--our indicator captures the latter sub-dimension. In the limited research on visual personalization on social media, privatization is commonly operationalized through such background stories (e.g., Farkas & Bene, 2021; Ha[beta]ler et al., 2024).
3.2.2. Independent Variables
3.2.2.1. Populist vs. Non-Populist Parties (H1)
We created a dummy variable for populist parties using the 2024 European edition of the Chapel Hill Expert Survey (CHES; Rovny et al., 2025). We identified populist parties through the variable "anti_elite_salience," which measures the extent to which a party employed anti-elite, anti-establishment rhetoric in 2024 (0 = not salient; 10 = extremely salient), which is a central element of populist political communication (de Vreese et al., 2018). Parties scoring above 6.67--the top third of the distribution--were classified as populist. Parties not matched to CHES were left uncategorized. Parties with separate Facebook accounts but forming a joint electoral alliance in CHES (e.g., Fidesz and KDNP in Hungary) were matched to their alliance. Electoral coalitions in our dataset were matched to the CHES party that won the most EP seats. Accordingly, the French coalition Besoin d'Europe was matched to Renaissance; the Czech coalition Spolu21 to Obcanska demokraticka strana; the Czech coalition Stacilo! to Komunisticka strana Cech a Moravy; and the Portuguese Alianca Democratica to the Partido Social Democrata. The Dutch coalition GroenLinks/Partijvan de Arbeid was not assigned because both parties won an equal number of EP seats and could not be matched unambiguously. Eleven parties in our sample were excluded from analyses because they could not be matched to any party in the 2024 CHES dataset. This procedure resulted in 36 populist parties, 102 non-populist parties, and 11 uncategorized cases. The final dataset for analysis therefore included 14,553 posts from the 138 categorized parties (see Tables A1 and A2 in the Supplementary File for the distribution across countries and parties).
3.2.2.2. Government vs. Opposition Parties (H2)
We coded whether parties were members of the domestic government or the opposition at the time of the election. Since the Portuguese sample did not include a government party, we excluded Portugal from testing H2.
3.2.2.3. Electoral System (H3)
We included a dummy variable indicating whether the electoral system is candidate-centered or not, following the classification of Hoyland et al. (2019) and adaptations by Gattermann (2022).
3.2.2.4. Authoritarianism (H4)
We accessed the V-Dem dataset (Coppedge et al., 2025) via https://demscore.se for the Liberal Democracy Index (Coppedge et al., 2015, 2025). The index ranges from 0 to 1, with higher values representing stronger liberal democracy. To align with our theoretical expectations, we reversed the scale so that higher values indicate higher levels of authoritarianism.
3.2.3. Controls
3.2.3.1. Political Ideology
We control for party ideology with a variable from CHES which measures general political ideology on the left-right scale and ranges from 0 (extreme left) to 10 (extreme right). We include this variable because progressive parties are often more decentralized in their organization than conservative parties, which may influence the extent to which they employ visual personalization strategies.
3.2.3.2. District Size
In addition to the candidate-centered electoral system measure, we control for the average district size used in the 2024 EP elections. In most countries, this equals the size of the national delegation, as the entire country serves as a single constituency. For Ireland, Italy, and Poland, we calculated district size by dividing the number of elected MEPs by the number of constituencies. For Belgium, we distinguished between Flanders (13 seats) and Wallonia (8 seats), excluding parties from the German-speaking electoral college.
3.2.3.3. Education Level
We control for the share of people with lower educational attainment in each country in 2024, defined by Eurostat as those with less than primary, primary, or lower-secondary education. This ranges from 11.8% (Poland) to 38.7% (Portugal). Education is linked to media literacy and political engagement, which may shape how advantageous visually personalized content is for parties.
3.3. Reliability
A total of 64 trained coders--each familiar with the language and political context of the country they coded--coded the Facebook posts using a joint English-language codebook developed by the international research network Digital Election Campaigning Worldwide (DigiWorld). To ensure reliability, two tests were conducted per country. In the intra-country test, coders coded about 100 posts from their national parties in their native language. In the inter-country test, coders in all countries coded 94 English-language posts from European political parties such as the European People's Party or the Party of European Socialists (six of the intended 100 had been deleted on Facebook), which required no specialized national political knowledge.
We used Brennan and Prediger's Kappa (B&P Kappa), a robust chance-corrected coefficient suitable for binary, often zero-inflated data like ours, and additionally report Holsti's composite reliability (CR). The results indicate a consistent understanding of the categories (individualization: B&P Kappa = .65, CR = .94; privatization: B&P Kappa = .83, CR = .98). Table A3 in the Supplementary File reports reliability scores per country.
3.4. Data Analysis
We employed multilevel logistic regression models with binomial outcomes to test our hypotheses, as both dependent variables--individualization and privatization--are binary. We estimated two sets of five models: one using individualization, the other privatization as the dependent variable. For each hypothesis, we ran a separate model to isolate the effect of the corresponding independent variable while controlling for the same set of potential confounding factors. We then estimated a full model including all independent variables to assess the robustness of the results. Lastly, we included country-level random intercepts to account for baseline differences in individualized and privatized posts across countries. We did not include random slopes, assuming that the effects of the independent variables are consistent across countries.
See Table A4 in the Supplementary File for descriptive statistics of all coded variables and Table A5 for the scores on all independent variables and controls at the country-level.
4. Findings
As shown by the descriptive statistics (Figures A1-A3 in the Supplementary File), individualization was by far more commonly used than privatization. Below, we test H1-H4 on how structural factors affected visual personalization on parties' Facebook pages during the 2024 EP elections. Tables 1 and 2 report odds ratios (OR) and standard errors from logistic regressions for individualization and privatization, where OR = 1 represents no change in odds, OR > 1 indicates higher odds, and OR < 1 lower odds of individualization/privatization. For instance, an OR of 1.50 reflects a 50% increase in odds, whereas an OR of 0.50 reflects a 50% decrease.
Table 1 reports how party- and country-level factors relate to the odds of parties using individualized posts. In the full Model 5, populist parties have 39.5% higher odds of using individualized posts (OR = 1.395, p < .001) than non-populist parties, confirming H1a, and governing parties have 45.2% higher odds (OR = 1.452, p < .001) than opposition parties, confirming H2a. Our other independent variables--candidate-centered systems and degree of authoritarianism--are not significantly associated with individualization, falsifying H3a and H4a. Regarding controls, a one-unit increase in right-wing positioning raises the odds by 1.9% (OR = 1.019, p < .05), while lower levels of education reduce the odds by 2.3% (OR = 0.977, p < .05). It is clear that district size does not affect individualization in these models.
Regarding privatization (Table 2), Model 5 finds that being a populist party reduces the probability of using privatized posts by 50.2% in the odds (OR = 0.498, p < .001), falsifying H1b. In contrast to individualization, government or opposition status does not significantly affect privatization, falsifying H2b. While candidate-centered systems do not significantly affect privatization, falsifying H3b, we find that a one-unit increase in authoritarianism reduces the odds of using private personalization by 98.3% (OR = 0.017, p < .01), falsifying H4b. Regarding controls, lower levels of education decrease the odds of privatization (OR = 0.933, p < .01), while stronger right-wing ideology increases them (OR = 1.075, p < .01). Again, we show that district size does not influence the use of privatized posts.
5. Conclusion
This study investigated how political parties across 23 EU countries strategically employed two dimensions of visual personalization--individualization and privatization--on Facebook during the 2024 EP election campaign. It explored how this strategic use was shaped by four structural factors: populist vs. non-populist orientation, government vs. opposition status at the party level, as well as electoral systems, and the degree of authoritarianism at the country level.
Individualization emerged as the dominant form of visual personalization, whereas privatization remained comparatively rare. This is in line with previous research on European parties' use of personalization strategies (e.g., Ha[beta]ler et al., 2024; Magin et al., 2024) and might be explained by the party-centered nature of political systems across Europe. The predominance of individualization further suggests that parties leveraged social media's visual affordances to humanize political actors and establish personal connections with voters, yet they maintained professional boundaries by avoiding excessive private disclosure. This strategic choice aligns with theoretical expectations that parties balance the need for authenticity and relatability with concerns about maintaining credibility and avoiding the risks associated with overly intimate political communication (Stanyer, 2013).
Starting with the party-level factors we investigated, we found differential effects of being a populist party on the use of visual personalization: Populist parties were more likely to use individualization, but less likely to use privatization than non-populist parties. These findings challenge simplistic assumptions about populist communication styles and refine our theoretical understanding of how populist actors navigate social media platforms. The emphasis on individualization without corresponding privatization suggests that populist parties strategically deploy leader visibility to reinforce their anti-establishment narrative and cultivate direct connections with citizens, yet they avoid personal vulnerability that might undermine their projected strength and decisiveness (Gerbaudo, 2018). This pattern reflects a calculated communication strategy wherein populist parties harness the visual potential of social media to amplify individual political leaders while maintaining a carefully curated public persona. The finding extends previous research on populist visual communication (Farkas et al., 2022) by revealing that populism shapes not only the extent, but also the specific dimensions of personalization, highlighting the need to conceptualize personalization as a multidimensional phenomenon rather than a unitary construct (Van Aelst et al., 2012).
Regarding the second party-level factor, we found that governing parties were more likely than opposition parties to use individualized visuals. This pattern indicates both a professionalized form of personalization designed to manage visibility and reinforce leadership recognition (Enli & Skogerbo, 2013) and a strategic use of individualization as message control in environments where political actors directly curate their own visibility (van der Brug & de Vreese, 2016). At the same time, government or opposition status did not affect the use of privatized visuals. Regarding governing parties, this supports the view that institutional responsibility constrains informal, emotional self-presentation (Hopmann et al., 2011). Interestingly, opposition parties were not shown to have made greater use of privatization than governing parties, even though they face fewer reputational risks and can thus afford to experiment with expressive or affective content, which opens up the use of personalization as a differentiation strategy to attract attention (Kollberg, 2024). These findings may simply indicate that privatization does not really fit in with European party systems and EP elections, both of which focus strongly on parties rather than individual politicians. Given the concerns that excessive privatization of politics could undermine the substance of political discourse and thus democracy (Adam & Maier, 2010), this result can be viewed as positive.
The country-level factors we investigated had overall a smaller effect on visual personalization strategies than the party-level factors, which suggests that parties' strategic decisions are more strongly influenced by party structures than by the surrounding conditions of the respective country, in line with Gibson and Rommele (2001). Whether an electoral system is candidate- or party-centered did not at all affect the use of individualized or privatized visuals, suggesting that the closeness of the candidates to their electorates was less influential than expected. Contrary to our expectations, we found that more liberal (rather than more authoritarian) national contexts promote greater privatization in parties' visual communication, while individualization appeared unaffected by this factor. This resonates with Grbesa (2008), who raised questions regarding the relevance of the political history of the country regarding both personalization and privatization of party communication, especially in post-socialist countries (several of which are represented in our sample) with strong resentment towards strong leadership and developing democratic experience. Our study thus underlines the need for a deeper, more complex understanding of national and historical factors affecting the use of parties' visual personalization strategies.
As with any research, our study has several limitations. While the four structural factors we examined offer new insights into how contextual conditions influence visual personalization in social media campaigning, future studies should complement them by additional characteristics--such as political culture, party-system polarization, or levels of social media use--to further clarify when political actors consider visual personalization a strategically advantageous approach. While our focus on party-level communication extends previous candidate-focused research on personalization on social media (e.g., McGregor, 2018; Metz et al., 2020), our operationalization does not allow for measuring presidentialization since it does not distinguish whether individualization reflects general visibility of politicians or concentrated visibility focused on leading figures. Future research should systematically compare the social media accounts of parties, political leaders, and other politicians (where all exist) to assess whether the structural factors identified here also shape candidates' visual personalization.
Our study covers only a single election period in a specific electoral setting (EP elections), which limits the generalizability of the findings. Nonetheless, because parties remain the central actors in European democracies and EP elections continue to function as second-order contests (Ehin & Talving, 2021) strongly shaped by national contexts (Ha[beta]ler et al., 2021), our results should remain relevant for party-centered environments beyond this particular election, at least to a certain extent.
Since platforms such as Instagram and TikTok have distinct visual conventions and user demographics that likely shape parties' visual personalization strategies, future studies should compare our Facebook-based results with comparable data from other platforms. Finally, longitudinal analyses covering both election and non-election periods would clarify whether the patterns identified here reflect only strategic campaign choices or rather more stable communication styles.
Summing up, our findings indicate that visual personalization in parties' Facebook communication reflects a strategic adaptation to platform logics shaped by party characteristics and national contexts. The dominance of individualization over privatization indicates that parties have adapted long-standing personalization strategies to social media's visual affordances rather than fundamentally transforming their communication approach. Political actors have long highlighted individual politicians to humanize and simplify their messages and establish voter connections; social media simply offers new channels for these established strategies. However, the empirical variation we document shows that this adaptation is neither uniform nor automatic. The differing use of individualization and privatization by populist vs. non-populist parties and by government vs. opposition actors indicates that institutional and ideological factors mediate how parties translate traditional communication strategies into digital contexts.
What appears new, therefore, is not visual personalization itself but the systematic, strategic calibration of its dimensions in response to platform affordances, competitive pressures, and institutional constraints. Our contribution lies in providing the first large-scale, cross-national evidence of how parties strategically navigate this adaptation, both revealing and contradicting patterns previously assumed but not empirically demonstrated. As platforms continue to evolve and political actors refine their digital strategies, sustained investigation of the multidimensional nature of visual personalization and its contextual determinants remains essential for understanding contemporary political communication in (European) democracies.
Acknowledgments
This publication is part of the project Campaigning for Strasbourg (CamforS). The authors would like to thank all collaboration partners who contributed to the infrastructure of the project, the coding scheme, and the creation of the dataset used in this publication. A list of all collaborators can be found on the project website: https://digidemo.ifkw.lmu.de/camfors.
Funding
The International Visegrad Fund supported this article under project no. 22410053: EU Campaign Race in V4--Political Communication in Social Media before the 2024 EP Elections. Krisztina Burai's research was supported by the Momentum program of the Hungarian Academy of Sciences (MTA-ELTE TK Lendulet "Momentum" PRISMA research project [LP2024-2/2024]). Anna-Katharina Wurst contributed as part of the junior research group DigiDeMo, funded by the Bavarian State Ministry of Science and the Arts and coordinated by the Bavarian Research Institute for Digital Transformation (bidt). Publication of this article in open access was made possible through the institutional membership agreement between the University of Innsbruck and Cogitatio Press.
Conflict of Interests
The authors declare no conflict of interests. In this article, editorial decisions were undertaken by Viorela Dan (University of Vienna), Philipp Muller (University of Mannheim), and Anne Schulz (University of Zurich).
Data Availability
Interested researchers are encouraged to contact the authors to request access to the data, which will be provided whenever feasible. For copyright and privacy reasons, the data cannot be published freely.
LLMs Disclosure
Several co-authors used LLM-based tools when working on this article. ChatGPT and DeepL were used for translating self-written text. ChatGPT, Claude, and Copilot were used to shorten self-written text. ChatGPT, Claude, Copilot, DeepL, Grammarly, Instatext.io, and Perplexity were used for language editing of self-written text. ChatGPT was used during the analyses to help debug code when it was not running correctly. The initial texts and code were always written by the authors. All results were carefully reviewed and edited. The following author states not to have used LLMs when working on the article: Katjana Gattermann.
Supplementary Material
Supplementary material for this article is available online in the format provided by the author (unedited).
References
Adam, S., & Maier, M. (2010). Personalization of politics: A critical review and agenda for research. Annals of the International Communication Association, 34(1), 213-257. https://doi.org/10.1080/23808985.2010.11679101
Bakke, E., & Sitter, N. (2022). The EU's enfants terribles: Democratic backsliding in Central Europe since 2010. Perspectives on Politics, 20(1), 22-37. https://doi.org/10.1017/S1537592720001292
Bast, J. (2024). Managing the image. The visual communication strategy of European right-wing populist politicians on Instagram. Journal of Political Marketing, 23(1), 1-25. https://doi.org/10.1080/15377857.2021.1892901
Bene, M. (2017). Go viral on the Facebook! Interactions between candidates and followers on Facebook during the Hungarian general election campaign of 2014. Information, Communication & Society, 20(4), 513-529. https://doi.org/10.1080/1369118X.2016.1198411
Bene, M., Magin, M., & Ha[beta]ler, J. (2024). Political issues in social media campaigns for national elections: A plea for comparative research. Politics and Governance, 12, Article 8727. https://doi.org/10.17645/pag.8727
Bennett, W. L. (2012). The personalization of politics: Political identity, social media, and changing patterns of participation. The Annals of the American Academy of Political and Social Science, 644(1), 20-39. https://doi.org/10.1177/0002716212451428
Berman, S. (2017). The pipe dream of undemocratic liberalism. Journal of Democracy, 28(3), 29-38. https://doi.org/10.1353/jod.2017.0041
Boulianne, S., & Larsson, A. O. (2024). Comparative digital political communication: Comparisons across countries, platforms, and time. Social Science Computer Review, 42(5), 1092-1100. https://doi.org/10.1177/08944393241258763
Bowler, S., & Farrell, D. M. (2011). Electoral institutions and campaigning in comparative perspective: Electioneering in European Parliament elections. European Journal of Political Research, 50(5), 668-688. https://doi.org/10.1111/j.1475-6765.2011.01988.x
Carey, J. M., & Shugart, M. S. (1995). Incentives to cultivate a personal vote: A rank ordering of electoral formulas. Electoral Studies, 14(4), 417-439. https://doi.org/10.1016/0261-3794(94)00035-2
Coppedge, M., Gerring, J., Knutsen, C. H., Lindberg, S. I., Teorell, J., Altman, D., Angiolillo, F., Bernhard, M., Cornell, A., Fish, M. S., Fox, L., Gastaldi, L., Gjerlow, H., Glynn, A., God, A. G., Grahn, S., Hicken, A., Kinzelbach, K., Marquardt, K. L Ziblatt, D. (2025). V-Dem: Codebook (v15). V-Dem Institute. https://www.v-dem.net/documents/55/codebook.pdf
Coppedge, M., Lindberg, S., Skaaning, S.-E., & Teorell, J. (2015). Measuring high level democratic principles using the V-Dem data (V-Dem Working Paper Series 2015:6). V-Dem Institute. https://www.v-dem.net/media/publications/v-dem_working_paper_2015_6.pdf
Daniel, W. T., & Obholzer, L. (2025). Social media campaigning across multiple platforms: Evidence from the 2024 European elections. West European Politics, 49(1), 293-319. https://doi.org/10.1080/01402382.2025.2501464
Darius, P., Drews, W., Neumeier, A., & Riedl, J. (2024). The EUDigiParty data set: The digital campaigning presence of 401 political parties during the European Parliament election 2024 including websites and social media handles on Facebook, Instagram, TikTok, X/Twitter, and YouTube [Data set]. Harvard Dataverse. https://doi.org/10.7910/DVN/U6UWPN
de-Lima-Santos, M.-F., Goncalves, I., Quiles, M. G., Mesquita, L., Ceron, W., & Lorena, M. C. C. (2024). Visual political communication on Instagram: A comparative study of Brazilian presidential elections. EPJ Data Science, 13, Article 72. https://doi.org/10.1140/epjds/s13688-024-00502-0
de Vreese, C. H., Esser, F., Aalberg, T., Reinemann, C., & Stanyer, J. (2018). Populism as an expression of political communication content and style: A new perspective. The International Journal of Press/Politics, 23(4), 423-438. https://doi.org/10.1177/1940161218790035
Dommett, K., Kefford, G., & Kruschinski, S. (2024). Data-driven campaigning and political parties: Five advanced democracies compared. Oxford University Press. https://doi.org/10.1093/oso/9780197570227.001.0001
Ehin, P., & Talving, L. (2021). Still second-order? European elections in the era of populism, extremism, and Euroscepticism. Politics, 41 (4), 467-485. https://doi.org/10.1177/0263395720986026
Ekman, M., & Widholm, A. (2017). Political communication in an age of visual connectivity: Exploring Instagram practices among Swedish politicians. Northern Lights: Film & Media Studies Yearbook, 15(1), 15-32. https://doi.org/10.1386/nl.15.1.15_1
Enli, G. S., & Skogerbo, E. (2013). Personalized campaigns in party-centered politics. Information, Communication & Society, 16(5), 757-774. https://doi.org/10.1080/1369118X.2013.782330
Enyedi, F. (2024). Concept and varieties of illiberalism. Politics and Governance, 12, Article 8521. https://doi.org/10.17645/pag.8521
Esser, F., & Stromback, J. (2012). Comparing election campaign communication. In F. Esser & T. Hanitzsch (Eds.), The handbook of comparative communication research (pp. 289-307). Routledge.
Eurostat. (2025). Population on 1 January [Data set]. https://ec.europa.eu/eurostat/databrowser/view/tps00001/default/table
Farkas, X., & Bene, M. (2021). Images, politicians, and social media: Patterns and effects of politicians' image-based political communication strategies on social media. The International Journal of Press/Politics, 26(1), 119-142. https://doi.org/10.1177/1940161220959553
Farkas, X., Jackson, D., Baranowski, P., Bene, M., Russmann, U., & Veneti, A. (2022). Strikingly similar: Comparing visual political communication of populist and non-populist parties across 28 countries. European Journal of Communication, 37(5), 545-562. https://doi.org/10.1177/02673231221082238
Filimonov, K., Russmann, U., & Svensson, J. (2016). Picturing the party: Instagram and party campaigning in the 2014 Swedish elections. Social Media + Society, 2(3). https://doi.org/10.1177/2056305116662179
Frantz, E., Kendall-Taylor, A., Nietsche, C., & Wright, J. (2021). How personalist politics is changing democracies. Journal of Democracy, 32(3), 94-108. https://dx.doi.org/10.1353/jod.2021.0036
Gattermann, K. (2022). The personalization of politics in the European Union. Oxford University Press. https://doi.org/10.1093/oso/9780198798712.001.0001
Gerbaudo, P. (2018). Social media and populism: An elective affinity? Media, Culture & Society, 40(5), 745-753. https://doi.org/10.1177/0163443718772192
Gibson, R. K., & Rommele, A. (2001). Changing campaign communication: A party-centered theory of professionalized campaigning. The International Journal of Press/Politics, 6(4), 31-43. https://doi.org/10.1177/108118001129172323
Gonzalez-Aguilar, J. M., Segado-Boj, F., & Makhortykh, M. (2023). Populist right parties on TikTok: Spectacularization, personalization, and hate speech. Media and Communication, 11(2), 232-240. https://doi.org/10.17645/mac.v11i2.6358
Grabe, M. E., & Bucy, E. P. (2009). Image bite politics: News and the visual framing of elections. Oxford University Press. https://doi.org/10.1093/acprof:oso/9780195372076.001.0001
Grbesa, M. (2008). Personality politics in Croatia [Unpublished doctoral dissertation]. University of Ljubljana. HaBler, J., Magin, M., Russmann, U., & Fenoll, V. (Eds.). (2021). Campaigning on Facebookin the 2019 European Parliament election: Informing, interacting with, and mobilising voters. Palgrave Macmillan. https://doi.org/10.1007/978-3-030-73851-8
HaBler, J., Wurst, A.-K., & Pohl, K. (2024). Politicians over issues? Visual personalization in three Instagram election campaigns. Information, Communication & Society, 27(5), 815-835. https://doi.org/10.1080/1369118X.2023.2227684
Holtz-Bacha, C. (2004). Germany: How the private life of politicians got into the media. Parliamentary Affairs, 57(1), 41-52. https://doi.org/10.1093/pa/gsh004
Holtz-Bacha, C., Langer, A. I., & Merkle, S. (2014). The personalization of politics in comparative perspective: Campaign coverage in Germany and the United Kingdom. European Journal of Communication, 29(2), 153-170. https://doi.org/10.1177/0267323113516727
Hopmann, D. N., de Vreese, C. H., & Albaek, E. (2011). Incumbency bonus in election news coverage explained: The logics of political power and the media market. Journal of Communication, 61(2), 264-282. https://doi.org/10.1111/j.1460-2466.2011.01540.x
Hoyland, B., Hobolt, S. B., & Hix, S. (2019). Career ambitions and legislative participation: The moderating effect of electoral institutions. British Journal of Political Science, 49(2), 491-512. https://doi.org/10.1017/S0007123416000697
Hrbkova, L., & Mackova, A. (2021). Campaign like a girl? Gender and communication on social networking sites in the Czech Parliamentary election. Information, Communication & Society, 24(11), 1622-1639. https://doi.org/10.1080/1369118X.2020.1716040
Jost, P. (2023). How politicians adapt to new media logic. A longitudinal perspective on accommodation to user-engagement on Facebook. Journal of Information Technology & Politics, 20(2), 184-197. https://doi.org/10.1080/19331681.2022.2076271
Kollberg, M. (2024). The challenger advantage--How challenger parties disrupt mainstream party dominance in the European Parliament. Journal of European Public Policy, 32(9), 2252-2281. https://doi.org/10.1080/13501763.2024.2391510
Lalancette, M., & Raynauld, V. (2019). The power of political image: Justin Trudeau, Instagram, and celebrity politics. American Behavioral Scientist, 63(7), 888-924. https://doi.org/10.1177/0002764217744838
Langer, A. I. (2007). A historical exploration of the personalisation of politics in the print media: The British prime ministers (1945-1999). Parliamentary Affairs, 60(3), 371-387. https://doi.org/10.1093/pa/gsm028
Magin, M., Ha[beta]ler, J., Larsson, A. O., & Skogerbo, E. (2024). Walking the line of the double bind: A cross-country comparison on women and men politicians' self-presentations on social media. Nordic Journal of Media Studies, 6(1), 157-184. https://doi.org/10.2478/njms-2024-0008
Maier, J., Nai, A., & Verhaar, N. (2024). More negative when it matters less? Comparing party campaign behaviour in European and national elections. Journal of European Public Policy, 32(9), 2307-2329. https://doi.org/10.1080/13501763.2024.2404154
Markowski, R., & Kotnarowski, M. (2025). Alternatives to liberal democracies and their consolidation in Central and Eastern Europe. Politics and Governance, 13, Article 9545. https://doi.org/10.17645/pag.9545
Mazzoni, M., & Mincigrucci, R. (2022). Storytelling through images: How leaders managed their visual communication on Facebook during the 2019 European election campaign. Journal for Cultural Research, 26(3/4), 221-243. https://doi.org/10.1080/14797585.2022.2106444
McGregor, S. C. (2018). Personalization, social media, and voting: Effects of candidate self-personalization on vote intention. New Media & Society, 20(3), 1139-1160. https://doi.org/10.1177/1461444816686103
Metz, M., Kruikemeier, S., & Lecheler, S. (2020). Personalization of politics on Facebook: Examining the content and effects of professional, emotional and private self-personalization. Information, Communication & Society, 23(10), 1481-1498. https://doi.org/10.1080/1369118X.2019.1581244
Mudde, C. (2004). The populist zeitgeist. Government and Opposition, 39(4), 541-563. https://doi.org/10.1111/J.1477-7053.2004.00135.X
Muller, W. C., & Strom, K. (Eds.). (1999). Policy, office, or votes? How political parties in Western Europe make hard decisions. Cambridge.
Newman, N., Fletcher, R., Robertson, C. T., Ross Arguedas, A., & Nielsen, R. K. (2024). Reuters Institute digital news report 2024. Reuters Institute for the Study of Journalism.
Papacharissi, Z. (2014). Affective publics: Sentiment, technology, and politics. Oxford University Press.
Pappas, T. S. (2016). Are populist leaders "charismatic"? The evidence from Europe. Constellations: An International Journal of Critical & Democratic Theory, 23(3), 378-390. https://doi.org/10.1111/1467-8675.12233
Pedersen, H. H., & Rahat, G. (2021). Political personalization and personalized politics within and beyond the behavioural arena. Party Politics, 27(2), 211-219. https://doi.org/10.1177/1354068819855712
Peeters, J., Opgenhaffen, M., Kreutz, T., & Van Aelst, P. (2023). Understanding the online relationship between politicians and citizens. A study on the user engagement of politicians' Facebook posts in election and routine periods. Journal of Information Technology & Politics, 20(1), 44-59. http://doi.org/10.1080/19331681.2022.2029791
Rahat, G., & Kenig, O. (2018). From party politics to personalized politics? Party change and political personalization in democracies. Oxford University Press.
Rahat, G., & Sheafer, T. (2007). The personalization(s) of politics: Israel, 1949-2003. Political Communication, 24(1), 65-80. https://doi.org/10.1080/10584600601128739
Rovny, J., Polk, J., Bakker, R., Hooghe, L., Jolly, S., Marks, G., Steenbergen, M., & Vachudova, M. (2025). The 2024 Chapel Hill Expert Survey on political party positioning in Europe: Twenty-five years of party positional data. Electoral Studies, 97, Article 102981. https://doi.org/10.1016/j.electstud.2025.102981
Russmann, U., Klinger, U., & Koc-Michalska, K. (2024). Personal, private, emotional? How political parties use personalization strategies on Facebook in the 2014 and 2019 EP Election campaigns. Social Science Computer Review, 42(5), 1204-1222. https://doi.org/10.1177/08944393241254807
Stanyer, J. (2013). Intimate politics: Publicity, privacy and the personal lives of politicians in media-saturated democracies. Polity Press.
Steffan, D. (2020). Visual self-presentation strategies of political candidates on social media platforms: A comparative study. International Journal of Communication, 14, 3096-3118.
Stetka, V., & Mihelj, S. (2024). The illiberal public sphere: Media in polarized societies. Palgrave Macmillan.
Street, J. (2004). Celebrity politicians: Popular culture and political representation. The British Journal of Politics and International Relations, 6(4), 435-452. https://doi.org/10.1111/j.1467-856X.2004.00149.x
Strom, K. (1990). A behavioral theory of competitive political parties. American Journal of Political Science, 34(2), 565-598. https://doi.org/10.2307/2111461
Stromback, J. (2008). Four phases of mediatization: An analysis of the mediatization of politics. The International Journal of Press/Politics, 13(3), 228-246. https://doi.org/10.1177/1940161208319097
Szebeni, Z., & Salojarvi, V. (2022). "Authentically" maintaining populism in Hungary--Visual analysis of prime minister Viktor Orban's Instagram. Mass Communication and Society, 25(6), 812-837. http://doi.org/10.1080/15205436.2022.2111265
Van Aelst, P., Sheafer, T., & Stanyer, J. (2012). The personalization of mediated political communication: A review of concepts, operationalizations and key findings. Journalism, 13(2), 203-220. https://doi.org/10.1177/1464884911427802
van der Brug, W., & de Vreese, C. H. (2016). (Un)intended consequences of European Parliamentary elections. Oxford University Press.
van der Brug, W., Gattermann, K., & de Vreese, C. H. (2021). Electoral responses to the increased contestation over European integration: The European elections of 2019 and beyond. European Union Politics, 23(1), 3-20. https://doi.org/10.1177/14651165211036263
Van Santen, R., & Van Zoonen, L. (2010). The personal in political television biographies. Biography: An Interdisciplinary Quarterly, 33(1), 46-67. http://www.jstor.org/stable/23541048
Van Zoonen, L. (2005). Entertaining the citizen: When politics and popular culture converge. Rowman & Littlefield.
Vuckovic, M. (2023). Politicizing, personalizing, and mobilizing in online political communication: Drivers and killers of users' engagement. International Journal of Communication, 17, 3033-3053.
Walter, A., & Emmenegger, P. (2023). Designing electoral districts for proportional representation systems: How electoral geography and partisan politics constrain proportionality and create bias. The Journal of Politics, 85(3), 1123-1138. https://doi.org/10.1086/723975
About the Authors
Melanie Magin is a professor of media sociology at the Norwegian University of Science and Technology (NTNU), Trondheim. She works at the intersection of political communication, social media, and comparative research, focusing on traditional and new media's societal role as well as the chances and risks associated with them.
Uta Russmann is a full professor of media and communication studies with a focus on democracy research at the University of Innsbruck, Austria. Her research focuses on political communication, media and election campaigns, digital communication, (visual) social media, public relations, and strategic communication.
Rossella Vulcano is a PhD candidate in a joint program between Grenoble University, France, and Groningen University, the Netherlands, in political science and social and behavioral sciences, respectively. Her research interests focus on civic political participation and the application of quantitative analyses to study social and political behavior.
Felix-Christopher von Nostitz is an associate professor in political science at ESPOL, Universite Catholique de Lille. His research interests are political participation, parties, and elections in the digital age. He is involved in various research projects including ActEU and INES. He is the co-coordinator of the Jean Monnet CoE TransformEU.
Anna-Katharina Wurst is a research associate at the Department of Media and Communication at LMU Munich. Her research interests include political communication, social media, research data management, and the application of computational methods.
Katjana Gattermann is an associate professor at the Amsterdam School of Communication Research, Department of Communication Science, University of Amsterdam, the Netherlands. Her research interests include elections and campaigning, political contestation, personality politics, misinformation, media-politics relations, and the media's role in democratic attitude formation within and across the European Union.
Laura Alonso-Munoz is a senior lecturer in journalism at the Department of Communication Sciences of the Universitat Jaume I, Spain. Her research focuses on political communication, disinformation, and populism on social media.
Delia Cristina Balaban (PhD, Europa University Viadrina Frankfurt/Oder, Germany) is a communication science professor at the Department for Communication, Public Relations, and Advertising, Faculty for Political, Administrative, and Communication Sciences, Babes-Bolyai University, Cluj-Napoca, Romania. Her research focuses on political communication on social media and social media influencers.
Pawet Baranowski is an assistant professor and deputy director at the Institute of Journalism and Social Communication, University of Wroclaw. His research intersects political communication and journalism studies, with a particular focus on social media, advertising strategies, and the impact of emerging communication technologies.
Krisztina Burai is a junior research fellow at the ELTE Centre for Social Sciences and a PhD candidate at Eotvos Lorand University. Her research focuses on political communication, social media, and news consumption.
Jean Claude Cachia is a senior lecturer at the University of Malta's Institute for European Studies. He joined the Institute in 2015 after graduating with a PhD in politics from the University of Lincoln.
Tomaz Dezelan is a professor of political science at the University of Ljubljana and coordinator of the American studies PhD programme. He has led numerous scientific projects, has published in prestigious scientific outlets, and has advised different international actors (e.g., OSCE, International IDEA, COE, EC, EP) in the fields of participation, representation, and elections.
Michal Garajis an assistant professor and deputy director at the Institute of Political Science and Public Administration, Faculty of Social Science, University of Ss. Cyril and Methodius in Trnava. His research focuses on European political parties, political communication on social media, and political participation.
Babette Hermans is a PhD researcher at the Media, Information & Persuasion Lab at KU Leuven. Her research lies at the intersection between journalism studies and political communication, with a focus on political knowledge of young adults and how fact-checks can help increase that knowledge.
Konstantinos Kallinikos is a postgraduate student in the joint master's program in public discourse and digital media at the University of Western Macedonia, Greece, where he also earned his bachelor's degree in communication and digital media.
Elisa Kannasto is a head of the Master School at Seinajoki University of Applied Science. Her research focuses on political communication on social media, self-branding, election campaigns, and strategic communication.
Simon Kruschinski is a senior researcher of platform data and computational social science at GESIS Cologne (Germany) and the principal investigator of a DFG-funded research project. His research focuses on how digital platforms, algorithms, and artificial intelligence are shaping the communication strategies of political actors and how this affects citizens' attitudes and behaviors.
Georgios Lappas is a professor of social and political informatics at the Department of Communication and Digital Media, University of Western Macedonia, Greece, and a visiting scholar at the Department of Media and Communication, LMU Munich, Germany. His research interests are in social informatics, artificial intelligence, digital democracy, and digital campaigning.
Sara Machado (PhD) is a researcher at the Transdisciplinary Research Center for Culture, Space and Memory (CITCEM) at the University of Porto (Portugal), where she has been teaching communication modules as an invited assistant since 2021. Her research focuses on political communication and participation, youth, and social media.
Alena Pospfsil Mackova is an assistant professor at the Department of Media Studies and Journalism, Faculty of Social Studies, Masaryk University. Her research focuses on political communication, social media, news consumption, and political attitudes.
Anamaria Dutceac Segesten is an associate professor in European studies at Lund University, Sweden. Her research interests include pan-European political elections and campaigns, digital politics, and political polarization.
Ilva Skulte is an associate professor at the Faculty of Social Sciences and a researcher at the Social Sciences Research Center, Riga Stradi[eta]s University. Her research focuses on journalism, media and information literacy, and tendencies in political discourse.
Milica Vuckovic is an assistant professor and vice dean at the Faculty of Political Science, University of Zagreb. Her research includes political communication, strategic communication, social media, election campaigns, and information disorder.
Matt Wall is a professor of politics at Swansea University's Department of Politics, Philosophy and International Relations. His research focuses on the interplay of election campaigns, public opinion, and digital technology.
Melanie Magin (1) (iD), Uta Russmann (2) (iD), Rossella Vulcano (3) (iD), Felix-Christopher von Nostitz (3) (iD), Anna-Katharina Wurst (4) (iD), Katjana Gattermann (5) (iD), Laura Alonso-Munoz (6) (iD), Delia Cristina Balaban (7) (iD), Pawet Baranowski (8) (iD), Krisztina Burai (9,10) (iD), Jean Claude Cachia (11) (iD), Tomaz Dezelan (12) (iD), Michal Garaj (13) (iD), Babette Hermans (14) (iD), Konstantinos Kallinikos (15) (iD), Elisa Kannasto (16) (iD), Simon Kruschinski (17,18) (iD), Georgios Lappas (4,15) (iD), Sara Machado (19) (iD), Alena Pospisil Mackova (20) (iD), Anamaria Dutceac Segesten (21) (iD), llva Skulte (22) (iD), Milica Vuckovic (23) (iD), and Matt Wall (24) (iD)
(1) Department of Sociology and Political Science, Norwegian University of Science and Technology (NTNU), Norway
(2) Department of Media, Society and Communication, University of Innsbruck, Austria
(3) ESPOL-Lab, Universite Catholique de Lille, France
(4) Department of Media and Communication, LMU Munich, Germany
(5) Amsterdam School of Communication Research, University of Amsterdam, The Netherlands
(6) Department of Communication Sciences, Universitat Jaume I, Spain
(7) Department of Communication, Public Relations, and Advertising, Babes-Bolyai University, Romania
(8) Institute of Journalism and Social Communication, University of Wroctaw, Poland
(9) Institute for Political Science, ELTE Centre for Social Sciences, Hungary
(10) Eotvos Lorand University, Hungary
(11) Institute for European Studies, University of Malta, Malta
(12) Faculty of Social Sciences, University of Ljubljana, Slovenia
(13) Institute of Political Science and Public Administration, University of Ss. Cyril and Methodius in Trnava, Slovakia
(14) Department of Communication Science, KU Leuven, Belgium
(15) Department of Communication and Digital Media, University of Western Macedonia, Greece
(16) Seinajoki University of Applied Sciences, Finland
(17) Department of Computational Social Science, GESIS--Leibniz Institute for the Social Sciences, Germany
(18) Department of Data Services for the Social Sciences, GESIS--Leibniz Institute for the Social Sciences, Germany
(19) Transdisciplinary Research Center for Culture, Space and Memory (CITCEM), University of Porto, Portugal
(20) Faculty of Social Studies, Masaryk University, Czech Republic
(21) Centre for Languages and Literature, Lund University, Sweden
(22) Social Sciences Research Centre, Riga Stradi[eta]s University, Latvia
(23) Faculty of Political Science, University of Zagreb, Croatia
(24) Department of Politics, Philosophy and International Relations, Swansea University, UK
Correspondence: Uta Russmann (uta.russmann@uibk.ac.at)
Issue: This article is part of the issue "Communication in Election Campaigns: Staggering Changes or Same Old, Same Old?" edited by Viorela Dan (University of Vienna), Uta Russmann (University of Innsbruck), Philipp Muller (University of Mannheim), and Anne Schulz (University of Zurich), fully open access at https://doi.org/10.17645/mac.i508
Submitted: 6 November 2025 Accepted: 12 February 2026 Published: 14 May 2026
https://doi.org/10.17645/mac.11739
Table 1. Individualization: Multilevel logistic regression (country
random effects with standard error).
Model 1 Model 2 Party in
Populist Government (Without
Party Portugal)
Populist Party 1.147 (**) (0.050)
Party in Government 1.277 (***) (0.047)
Candidate-Centered
Electoral System
Authoritarianism
District Size 0.998 (0.003) 0.999 (0.003)
Education 0.977 (*) (0.010) 0.979 (+) (0.011)
Political Ideology 1.035 (***) (0.008) 1.043 (***) (0.007)
Intercept 0.854 (0.201) 0.701 (0.171)
Num. Obs. 14,553 14,553
[R.sup.2] Marg. 0.014 0.017
[R.sup.2] Cond. 0.049 0.054
Model 3 Model 4 Authoritarianism
Candidate-Centered
Electoral System
Populist Party
Party in Government
Candidate-Centered 1.192 (0.196)
Electoral System
Authoritarianism 0.707 (0.417)
District Size 0.999 (0.004) 0.998 (0.003)
Education 0.976 (*) (0.010) 0.977 (*) (0.011)
Political Ideology 1.045 (***) (0.007) 1.045 (***) (0.007)
Intercept 0.776 (0.185) 0.934 (0.300)
Num. Obs. 14,553 14,553
[R.sup.2] Marg. 0.013 0.014
[R.sup.2] Cond. 0.047 0.050
Model 5 All Variables (With Portugal)
Populist Party 1.395 (***) (0.068)
Party in Government 1.452 (***) (0.061)
Candidate-Centered
Electoral System
Authoritarianism 0.760 (0.424)
District Size 0.999 (0.003)
Education 0.977 (*) (0.010)
Political Ideology 1.019 (*) (0.008)
Intercept 0.781 (0.240)
Num. Obs. 14,553
[R.sup.2] Marg. 0.022
[R.sup.2] Cond. 0.055
Note: (+) p < 0.1, (*) p < 0.05, (**) p < 0.01, (***) p < 0.001.
Table 2. Privatization: Multilevel logistic regression (country random
effects with standard error).
Model 1 Model 2 Party in
Populist Government (Without
Party Portugal)
Populist Party 0.547 (***) (0.080)
Party in Government 1.191 (0.139)
Candidate-Centered
Electoral System
Authoritarianism
District Size 1.003 (0.008) 1.003 (0.009)
Education 0.937 (*) (0.027) 0.933 (*) (0.027)
Political Ideology 1.075 (**) (0.029) 1.057 (*) (0.026)
Intercept 0.040 (***) (0.025) 0.037 (***) (0.024)
Num. Obs. 14,553 14,553
[R.sup.2] Marg. 0.087 0.073
[R.sup.2] Cond. 0.246 0.240
Model 3 Model 4 Authoritarianism
Candidate-Centered
Electoral System
Populist Party
Party in Government
Candidate-Centered 1.140 (0.497)
Electoral System
Authoritarianism 0.022 (*) (0.033)
District Size 1.003 (0.009) 0.998 (0.008)
Education 0.931 (*) (0.027) 0.926 (**) (0.024)
Political Ideology 1.054 (*) (0.026) 1.054 (*) (0.026)
Intercept 0.041 (***) (0.027) 0.169 (*) (0.130)
Num. Obs. 14,553 14,553
[R.sup.2] Marg. 0.070 0.126
[R.sup.2] Cond. 0.237 0.252
Model 5 All Variables (With Portugal)
Populist Party 0.498 (***) (0.085)
Party in Government 0.866 (0.120)
Candidate-Centered 0.805 (0.316)
Electoral System
Authoritarianism 0.017 (**) (0.024)
District Size 0.997 (0.008)
Education 0.933 (**) (0.024)
Political Ideology 1.075 (**) (0.029)
Intercept 0.207 (*) (0.159)
Num. Obs. 14,553
[R.sup.2] Marg. 0.149
[R.sup.2] Cond. 0.261
Note: + p < 0.1, (*) p < 0.05, (**) p < 0.01, (***) p < 0.001.
 Printer friendly
 Cite/link
 Email
 Feedback
| |
| Title Annotation: | ARTICLE |
|---|---|
| Author: | Magin, Melanie; Russmann, Uta; Vulcano, Rossella; Nostitz, Felix-Christopher von; Wurst, Anna-Kathar |
| Publication: | Media and Communication |
| Article Type: | Report |
| Geographic Code: | 4EXHU |
| Date: | Jun 1, 2026 |
| Words: | 10534 |
| Previous Article: | "Vote for Me, I Am Authentic": Performed Political Authenticity on Social Media. |
| Next Article: | Party Equalization or Normalization Through Visual Generative AI in the 2025 German Federal Election. |
| Topics: | |