Online Media Bias and Political Participation in EU Member States; Cross-National Perspectives

Abstract

This study aims to evaluate the complex relationship between online media consumption, the quality of the digital landscape, and participatory democracy in EU member states. The research is focused on a long-term statistical series from 2000 to 2024. It evaluates the temporal dynamics and structural shifts in media consumption and democratic participation across EU member states. The paper evaluates the influence of social media usage, online media consumption, traditional media, and online media partisanship on different levels of democratic participation based on theoretical frameworks of liberal and deliberative democracy and networked political communication. The results show that the use of social media for offline political networks is positively associated with democratic participation across all quantiles. In contrast, online media consumption has a more pronounced impact among already active citizens. Online media bias is negatively correlated with participatory democracy, especially at high levels, suggesting that media partisanship could inhibit or demotivate civic participation. Traditional media, when consumed critically, remains an important vector of democratic engagement, especially for active citizens. The results exhibit the ambivalent role played by online media, which might stimulate or constrain democratic participation by the level of partisanship.

1. Introduction

The digital environment of the last two decades has become an important element of social, cultural, and political life that has reshaped public and political participation. This evolution of the digital space created the opportunity for individuals to express themselves, to organize in virtual groups or associations, to participate in public or political debates, and to interact with political institutions. This complex digital landscape changed the traditional socialization and dynamics between citizens, media, and political institutions. While traditional media and mechanisms of political participation enhanced freedom of speech and association between individuals, the new environment is characterized by several important challenges related to partisanship, polarization, selective perceptions, and radical attitudes and behaviors. Anyone can produce online content, and the internal mechanisms of social media platforms influence what information reaches the public, how often, and in what kind of social or political message. Although the online environment is associated with an increased level of public participation, in practice, there are several challenges and threats related to misinformation, prevalence of fake news, or polarization of opinions and attitudes. The new online media also had an important impact on the decline of trust in traditional democratic institutions, with differences across EU member states.

An increased level of civic accountability and engagement is observed in the Northern and Western EU states. In these regions, online media is correlated with civic mobilization and consolidation of participatory democracy. In contrast, in Eastern EU countries, online media partisanship contributed to the acceleration of social divisions, the emergence of forms of populism, and the weakening of democratic institutions and mechanisms.

In accordance with these assumptions, this paper aims to create the nexus between online media partisanship and the intensity of democratic participation in EU member states. Moreover, the paper highlights the relevance of the critical role played by traditional media (print/broadcast media) and the importance of online media consumption in the field of civic engagement. Together with traditional and digital factors, the paper examines the complex and intriguing relationship between the quality of the online media environment and the participatory democracy index across EU member states. Using secondary statistical data, the paper captures these interactions and dynamics in a long-term statistical analysis from 2000 to 2024. Research methodology is quantitative, being based on both descriptive and inferential statistics. Therefore, statistical design is based on a quantile regression model to explore the influence of traditional and online media elements in the sphere of civic engagement and political participation in EU member states.

1.1. From Traditional Models to Political Participation in the Digital Age; Academic Literature and Deducing Hypotheses

The freedom of citizens to communicate and associate with one another is a sine qua non principle of democracy (Beetham, 2004). Among its key features are the right of citizens to form and join various associations that express interests and values; free access to alternative sources of information and independent media; the freedom to create, maintain, and communicate beliefs and opinions; and the right to participate in public dialogue (Diamond, 2003, p. 35). In line with this idea, a fundamental characteristic of a democratic political system is the responsiveness of government institutions to citizens’ preferences, according to Robert Dahl (1971, p. 1). This implies that all citizens should be able to articulate and communicate their preferences so that decision-makers can consider them equitably (Dahl, 1971, p. 2). Therefore, freedom of expression emerges as a critical institutional guarantee for democracy (Dahl, 1971, p. 3). Most constitutions of democratic societies enshrine this right, and many international organizations, particularly in Europe, emphasize the defense of freedom of expression, establishing monitoring mechanisms to ensure compliance (Voorhoof & Cannie, 2010; Woods, 2006, 2014).

A corollary of freedom of expression is freedom of the press. A key liberal argument in favor of press freedom is grounded in the understanding of human rights, freedom being one of the individual’s natural rights (Cruft, 2022)—further theoretical justification provided by utilitarian—consequentialist frameworks, which are grounded in the normative tradition of classical liberalism. John Stuart Mill argues that suppressing opposing opinions obstructs the path to truth and affects the process of deliberation either by preventing error correction or by depriving individuals of the ability to strengthen their convictions through rational debate (Mill, 2003, p. 20). For Mill, the path of truth is the path of progress, one that is secured by liberty, by freedom of thought, which is fostered precisely through the possibility of free expression. The maximum type of intellectual development that a community can experience is an expression of the full affirmation of freedom of thought (Mill, 2003, pp. 35–36). Prohibiting the expression of ideas that challenge dominant beliefs denies human fallibility. Such suppression is harmful because opinions rarely encompass the whole truth or complete justice; even flawed views often contain elements of truth (Mill, 2003). Freedom of expression promotes progress and the public interest (Haworth, 2007) and, accordingly, press freedom is integral to both the theory and practice of democracy. A free press is seen not only as an expression of liberty, but also as a contributor to societal well-being. Building on these normative foundations, empirical research has further elaborated the democratic functions of a free press. Scholarly research identifies the press as a key variable in democratization, good governance, and development (Norris, 2007). Norris (2007) identifies three roles of a free press: (a) the press as a watchdog of the public interest; (b) the press as a civic forum; and (c) the press as an agenda-setter. As a watchdog, the press scrutinizes government actions, exposing incompetence or abuse of power. As an independent actor, it promotes accountability and protects citizens’ interests (Bennett & Serrin, 2005; Curran, 2007). Having the role of civic forum, the press offers platforms for communication and debate, enabling citizens to express preferences without prior authorization (Schmuhl & Picard, 2005). The press, with its multiple communication channels and debate frameworks, enables citizens to formulate and make their preferences known (Schmuhl & Picard, 2005). It fosters a lively public dialogue, including achieving a certain level of aggregation of preferences (Anderson, 2014). The press provides means for articulating citizens’ wishes and representing their interests, thus having a significant role in the proper functioning of democracy (Curran, 1993). The press contributes to the formation of the public sphere and thus to influencing political decision-making. A free press ensures public access to information and thus the effectiveness of the public sphere (Gilwald, 1993). As an agenda-setter, the press signals the tribulations that arise in society, forcing the administration and the government to respond, or at least to evaluate the respective situations (McCombs, 2005). Press independence is essential to ensure that no important issues are suppressed due to political sensitivities. Greater press freedom increases the likelihood that all community concerns are openly reported and not censored.

However, these classical liberal arguments for press freedom have lost some persuasive force in recent times. Increasingly, many self-identified liberals stress the risks and limitations of unrestricted expression (Reiff, 2024). Compared to the past, new media tools offer the possibility for a much larger number of people to express themselves publicly, in a much freer manner, less, or even not at all, controlled by academic or editorial norms. Now liberals seem more likely to demand regulation of the channels and content of publicly expressed messages, while conservatives present themselves more as promoters of freedom and unfettered communication (Stoner, 2023). This shift is partly explained by the rise in new media, which allows more individuals to express themselves with minimal oversight. Thus, mechanisms are needed to combat disinformation (Olan et al., 2024), manipulation, or proliferation of untruths through new media (Sievi & Pawelec, 2025).

To uphold liberal values such as social justice, some argue that regulatory mechanisms are necessary to limit harmful or radical expressions (Schejter & Tirosh, 2015). Also, in order to preserve or promote liberal principles of social justice, regulations are seen as necessary to prevent the expression of radical views, opposed to those desiderata (Schejter & Tirosh, 2015; Zhang & Davis, 2022). According to these theoretical foundations, we postulate the following hypotheses:

H₁: 

Traditional media, based on freedom of expression and “watchdog” role, is positively related to an increased level of participatory democracy.

H₂: 

Governmental capacity to regulate online content in ways that protect liberal principles and freedom of expression is positively correlated with an increased level of participatory democracy.

New media offer, under the conditions of freedom of expression, the powerful potential for individuals or groups to promote conservative ideas or doctrines, avoiding the classic central press, gaining widespread support, and mobilizing a large number of supporters (Schradie, 2019). The relationship of new media with democracy is nuanced (Asimakopoulos et al., 2025; Alkiviadou, 2024; Hunter, 2023; Persily & Tucker, 2020). On the one hand, as suggested by classical arguments, the diversity and accessibility of new media, facilitating the formulation and communication of preferences, can lead to a better functioning of democracy (Jha & Kodila-Tedika, 2020; Fischer & Jarren, 2024; Wischmeyer, 2019; Jennings et al., 2020). However, it is crucial to be cautious about the dangers brought by new media—fragmentation of messages, self-referential encapsulation of opinions, lack of epistemic criteria, the possibility of targeting a specific audience with partisan messages, etc.—that can also have an opposite effect (Sunstein, 2018; Rodilosso, 2024; Gradwohl et al., 2025).

1.2. Theoretical Perspectives on the Double-Edged Sword: Online Media and Its Effects on Democratic Participation

Starting with the Arab Spring protest movement, continuing with the current role of different platforms in electoral campaigns, such as TikTok, Twitter, YouTube, or Facebook, we may easily see that social media evolved as a mass communication tool becoming a trigger for political activism and social engagement: ‘No politician, no party, no NGO, and no social movement can do without profiles on social media today’ (Fuchs, 2023, p. 17). Social media platforms are, at the same time, instruments for political information and disinformation. This dual characteristic is an important topic of scholarly debate. Therefore, the relationship between social media and democracy is important and yet contested by different stakeholders due to its embodying risks: misinformation, polarization, surveillance, or algorithm bias. Important scholars, such as Habermas, Papacharissi, Fraser, Lipschultz, Segerberg, or Bennett, underline the ambivalent face—promises and pitfalls—of social media in democratic societies.

A broader perspective presents social media as a virtual space represented by digital platforms that create and offer networked connections on a global scale. It refers to “tools that promote interaction” (Lipschultz, 2018) between multiple actors from the private or public sphere, often delimited by a thin line. Social media impacts both private and public behavior, influencing civic and political participation and democratic processes. Social media term is grounded in a large-scale conceptual framework, and some of the main theories are related to personal needs and the dynamics of change within society.

Agenda-Setting and Personal Influence theory, developed in 1955 by Katz and Lazarsfeld, is an important starting point for social media studies. Everett Rogers in 1962 (Rogers, 2003) analyzed the increasing role of new communication instruments as a way to explore the diffusion of innovations theory based on the S-curve adoption model. Katz et al. (1974) proposed the Uses and Gratifications Theory that explains how people need to join social networks for information and social integration. Manuel Castells (1996) states that social media determines the configuration of cultural, political, and social systems, and this is a result of Network Society Theory. More recently, Computer-Mediated Communication theory, initiated by Baym (2010), explains the dominant role of digital connectivity in interpersonal interactions. Cass Sunstein is a pioneer voice for the echo chambers concept that expresses the role of digital technologies in creating group polarization on a large scale based on a shared narrative (Sunstein, 2001, 2002, 2009). This is a result of social fragmentation and intellectual isolation, explained by Eli Pariser’s book, The Filter Bubble (2011) (Pariser, 2011). The echo chamber effect is even more present in the current social media context due to algorithmic filtering options. Recent research (Cinelli et al., 2021), based on more than 100 million topics about abortion, gun control, and vaccination, from Facebook, Twitter, Gab, and Reddit, confirms this idea. Social network analysis (SNA), rooted in structuralist theory, evolved from early sociometry into a paradigm based on graph theory, network dynamics, and relational sociology. It illustrates how digital platforms using digital traces determine ‘common affiliation, communities of practice, or collective action’ (Scott & Carrington, 2011, pp. 168–169).

Jürgen Habermas’s (1989) approach to the public sphere, which is considered the basis of the deliberative democracy model, highlights the rational–critical thinking of individuals, free from distortion and domination. Stevenson (2002) and other scholars have stated that this perpetual challenge to adapt to a digital landscape of communication is characterized by algorithmic mediation, emotional appeals, and fragmented attention. In such an environment, democratic communication is biased by access inequalities and structural exclusions, which nowadays are digitally reproduced (Fraser, 2007). Additionally, social media offer access to both conventional and alternative news outlets, making the potential effects of perceived bias difficult to predict (Ardèvol-Abreu & de Zúñiga, 2017).

In the current context of research, there is a shift from deliberation to participation. The role of emotional activism is dominant, and it offers a new perspective based on affective publics. This concept describes the perception rooted in common feelings rather than rational discourse, as Papacharissi explains: ‘I argue that networked digital structures of expression and connection are overwhelmingly characterized by affect’ (Papacharissi, 2015, p. 8). In this regard, social media is also an important instrument for mass mobilization by connective action (Bennett & Segerberg, 2012) using horizontal communication of digital networks that offer a dynamic and flexible way of organizing protest movements, for example. In accordance with these theoretical aspects, we postulate the following research hypothesis:

H₃: 

Higher levels of social media usage for political actions and online media consumption are positively associated with increased democratic participation.

Feenberg (2002) argues that online platforms should be designed in line with democratic values. The spread of digital capitalism with a mercantile focus and increased surveillance mechanisms (Fuchs, 2023) is one of the dark sides of social media. As a response, Fuchs proposes a public service Internet, which is a manifestation of the digital public sphere and digital democracy (Fuchs, 2023, p. 17) in reaction to the public sphere concept argued by Habermas. The new digital era of fringe democracy, which might be seen as a “heterodox democracy”, is a result of contemporary change within the society that includes actors and narratives opposite to liberal democracy and democratic values. By using algorithm-driven platforms like Twitter, Facebook, and YouTube, populist figures and extremist groups transform a fringe discourse into a prominent one (Boccia Artieri et al., 2025, pp. 4–6). Different studies suggest that an important aspect of participation, specifically being politically informed, is related to the algorithmic functions of social media platforms, such as Facebook (Thorson & Wells, 2016; Yoo & Gil de Zúñiga, 2019), which act like a business model (Papa & Photiadis, 2021, p. 2).

In their speech, online media communication platforms declare their neutrality, but in reality, they are the big players of the moment in the power equation. In this regard, the book Custodians of the Internet explains the “myth of the neutral platform” based on the idea that online platforms not only host the content, but they influence it by using algorithmic choices and content moderation that make them powerful actors (Gillespie, 2018, pp. 21–44). Therefore, Lipschultz (2018) describes this as an era of networked political communication, where users are both producers and consumers of political content, creating hybrid models of participation. He underlines the increasing role of social media in all sectors of the public sphere and the impact it has in activating civic engagement, or different kinds of social and political movements (Lipschultz, 2018, p. 13).

An important aspect is revealed by Jacobs and Spierings (2016, p. 19) about the impact that social media may have in increasing the visibility of the political parties that implicitly contribute to inequality in power balances. Huckfeldt and Sprague argue that political information is related to personal contacts (Huckfeldt & Sprague, 1995, pp. 125–126). Therefore, based on these theoretical assumptions, we formulate the following research hypothesis:

H₄: 

Increased online media bias erodes participatory democracy by reducing citizens’ engagement and trust in the democratic process.

Nevertheless, digital platforms are the new players in the communication arena, and they significantly impact individual public participation and the decision-making process. In this complex landscape, there are some key factors that influence democratic participation, such as the rate of engagement, platform governance, and social and political context. Digital media may be both enablers and inhibitors of democratic practices. Social media platforms, though primarily driven by commercial goals (Thorson & Wells, 2016), unintentionally expose users to political content, often through incidental or peer-shared encounters, that might influence political perceptions and offline participation (Bode, 2016; Kümpel, 2020; Boulianne, 2009). In the same vein, recent studies on the impact of social media highlight the need for further research exploring its role in development communication and social change (Ihsaniyati et al., 2023, p. 30). However, the extensive access to social-media platforms for all individuals requires abilities, skills, and specific competencies (Polanco-Levicán & Salvo-Garrido, 2022).

In the context of the digitalization of political communication, political polarization impacts political division (López-López et al., 2023, p. 2). This polarization is increased by micro-segmentation algorithms, which reinforce selective exposure and weaken the deliberative potential of the digital public space. Online media acts as a medium for the crystallization of political beliefs and attitudes; consequently, political polarization is intensified. Selective exposure to partisan content interferes with individuals’ cognitions, generating social and political polarization. This dynamic is influenced by factors such as personality, socio-economic status, and political ideology (Rowden et al., 2014; Hoewe & Peacock, 2020; Yarchi et al., 2020; Hout & Maggio, 2021). Scholars observed negative correlations between media bias, polarization, and participatory democracy (Johnson et al., 2017; Levendusky & Malhotra, 2015; Mellon & Prosser, 2017). In line with these theoretical perspectives, we aim to test the following research hypothesis:

H₅: 

An increased level of media bias is positively correlated with an increased level of political polarization.

The impact of digital platforms on polarization is contingent—influenced by contextual factors such as algorithmic design and user preferences (Barberá, 2020, pp. 38–46). Most research explains the polarizing effects of the media, but Kubin and von Sikorski highlight in their investigation that the media can also contribute to reducing political polarization. Deliberate exposure to opposing opinions and the promotion of cognitive openness can reduce affective polarization. This depolarizing potential remains poorly explored, and there is a need for experiments aimed at stimulating media contact with different points of view and recalibrating recommendation algorithms to favor informational diversity (Kubin & von Sikorski, 2021, pp. 196–199). Thus, the media could become not only a generator of division, but a tool for the reconstruction of democratic dialogue.

2. Materials and Methods

2.1. Research Directions, Questions, and Objectives

In order to analyze the complex interactions between democratic participation and independent factors, this research is based on several research objectives such as O1: to evaluate the relationship between traditional media, particularly its “watchdog” role, and the level of participatory democracy across EU member states; O2: to assess the impact of the governmental capacity to regulate online content, in ways that enhance liberal values, on the field of participatory democracy index; O3: to estimate the impact of online media consumption in the field of civic engagement and participatory democracy across EU member states; O4: to evaluate the relationship between online media bias and participatory democracy index by diminishing citizens’ engagement and political trust in democratic institutions; O5: to examine the complex interplay between rising levels of online media bias and political polarization across EU member states.

2.2. Data Collection and Statistical Design

This section shows the variables included in the analysis and the criteria for sample selection. The process is focused on ensuring cross-national comparability and methodological validity within the European context. Understanding the complex relation between online media and participatory behavior is complex and requires a multidimensional analytical framework that accounts for both media and behavioral variables. This includes not only the nature and diversity of online media content, but also the mechanisms through which individuals engage in civic and political life. As part of the data collection process, the research focused on a set of key analytical dimensions for a better understanding of the patterns and nuances of the participatory behavior across EU member states. In this respect, our data are relevant for media trust and polarization, governmental capacity to regulate digital content, biased traditional vs. digital media influence, and mobilization for democracy and civic engagement. Exploring democratic participatory behavior requires more than simply measuring voter turnout or civic engagement. The channels through which individuals access information and crystallize political opinions and cognitions have become increasingly fragmented and ideologically polarized. This affects how people think about public life and the political process. The cognitive process of forming opinions becomes more reactive and less reflexive as long as individuals are exposed to a constant stream of emotionally charged or contradictory political messages. Political distrust is connected to an increased level of media fragmentation by exposing individuals to inconsistent and contradictory information. This is often correlated with difficulties in building stable political opinions and feeling confident about engaging in the democratic process. Thus, following the academic literature, we aim to investigate the dynamics of participatory behavior across European member states using both political and mass-media statistical indicators.

This study is based on secondary statistical data provided by the V-Dem dataset (Varieties of Democracy), one of the most comprehensive data sources of democratic indicators. V-Dem Institute is an international think tank developed by the University of Gothenburg, which brings together scholars and researchers in the field of social and political sciences. This research project provides statistical data related to social transformations, traditional and digital media, and democracy across more than 200 countries on long-term statistical series. Statistical indicators selected for this study reflect key dimensions of democratic participatory behavior in a digital context: social media usage, online media consumption, online media perspectives, traditional media critical perspectives, online media bias and fractionalization, governmental regulatory policies in the field of digital media, and the participatory democracy index. The V-Dem dataset does not rely on population-based surveys but rather on expert-coded assessments. Each country and year observations are evaluated by five independent experts with a comprehensive knowledge of the political systems. This methodology ensures both conceptual and comparative design. In contrast, this methodology does not guarantee statistical and socio-demographic representativeness by age, gender, income, or education.

(a)

Dependent Variable

As we pointed out in the first part of this section, the study examines the dynamics of civic engagement and participatory behavior across EU member states. An important indicator for capturing the evolution of civic engagement and democratic participation is represented by the participatory democracy index. According to V-Democracy’s perspectives and conceptualization, participatory democracy refers to a model of democracy based on citizens’ involvement in both electoral and non-electoral processes. Together with electoral, liberal, egalitarian, and deliberative components, participatory democracy highlights the crucial role of civil society and other social and political organizations in fostering direct civic engagement and collective civic action.

(b)

Independent Variables

The independent factors that could explain the variation in participatory democracy across EU member states are represented by media consumption and social media usage, the quality of the digital media environment, and governmental capacity to regulate, by the national legal framework, the online content. In the field of social media usage, we used several statistical indicators such as online media consumption, social media usage for offline political actions, and online mobilization for democracy. Therefore, in the field of quality of digital media, we integrated several variables such as online media perspectives, traditional media measured by the print/broadcast media, online media bias, online media fractionalization, and political polarization. The last dimension is characterized by the government’s capacity to regulate online content in accordance with national and European legal framework the national and European legal frameworks.

In the analytical framework developed by V-Democracy, online media consumption is captured through the indicator known as “online media existence”. This variable suggests the extent to which individuals use domestic online media as a source of information. The variable is measured on an ordinal scale from 0 to 3, where 0 reflects the lack of online media consumption, and 3 represents widespread use, with citizens extensively engaging with online news content. The variable social media usage for offline political usage represents the extent to which average citizens use online platforms to coordinate different forms of political participation. However, the variable highlights the frequency with which social media platforms are used for organizing collective actions, such as protests, meetings, demonstrations, and other forms of civic participation. Using the scale from 0 to 4, V-Dem researchers capture the variation in the extent to which citizens use social media platforms to organize political actions, from a complete absence of such mobilization to frequent and diverse forms of civic engagement. Therefore, social media usage refers to citizens’ engagement in political actions using platforms such as Facebook, Twitter, and Instagram. Moreover, online media consumption expresses the use of digital news outlets and web pages for political information. In terms of the online mobilization for democracy, we stress that this variable measures the frequency and magnitude of collective public actions to support free elections, civil rights, the rule of law, and a democratic society. The scale used by V-Democracy is from 0 to 4, reflecting the variations between a lack of pro-democratic online mobilization and widespread events, both large and small-scale.

Concerning the quality of the media environment, an important variable is represented by online media perspectives. This indicator, measured from 0 (a heavily restricted media environment) to 4 (an inclusive media sphere), evaluates the extent to which the domestic digital media landscape presents a large spectrum of political viewpoints. In correlation to the academic literature, we used data regarding traditional media for understanding the “watchdog role” played by print or broadcast media in new societies. In this respect, traditional media play an important role in preserving liberal dimensions of democratic societies, enhancing transparency, freedom of speech, and diversity. Traditional media, measured through print/broadcast media, critically expresses the presence and prominence of critical coverage within national newspapers, radio, and television. The data were measured from 0 (fully controlled and censored media landscape) to 3 (media pluralism where major outlets occasionally criticize the government). Political polarization is a key variable in assessing the quality of democratic life. This variable illustrates the extent to which political fragmentation influences social relationships beyond formal discourse. The value 0 is related to the absence of polarization and 4 to a high level of political polarization, where citizens interact with political opponents in a hostile manner.

Among the indicators included in the analysis, online media bias and online media fractionalization stand out as important variables for understanding and assessing the quality of the digital media environment. These variables reflect both the ideological balance and structural cohesion of online public discourse. Online media bias presents the extent to which political parties or candidates are disadvantaged in online public discourse. In line with this conceptual perspective, this variable refers to the editorial stance and ideological orientation of online media outlets. It does not encompass broader phenomena such as fake news or public disinformation. However, online media bias reflects whether the digital media environment presents political actors in an impartial or balanced manner, favoring the government party. We used this variable due to its consistent operationalization in the V-Democracy database. The metrics provided by V-Democracy allowed us to make country comparisons and robust longitudinal analyses. To ensure methodological consistency across the set of indicators, the original coding of the variable, where 0 represents higher levels of bias, was reversed. In this version, higher values indicated greater media bias, aligning with the directionality of all other variables used in our analysis. Therefore, the value 0 reflects proportional and balanced media coverage, and the value 4 expresses an increased level of media bias. In correlation with online media bias, online media fractionalization measures the degree of divergence in how online media presents significant political events. Initially, the indicator ranged from 0 (high level of fragmentation and divergence in news presentation) to 4 (narrative coherence in presentation). For consistency across the dataset, the scale was reversed. In the reversed scale, higher scores (3, 4) expressed a high level of online media fractionalization, and lower values (0, 1) correspond to a coherent journalistic perspective on different political events.

In the field of governmental control of online content, we used the variable known as government capacity to regulate online content as an indicator. Thus, this variable measures the state’s capacity to implement a legal framework and to regulate online content. The variable is measured from 0 to 4, where higher values are related to an increased technical and legal capacity of the governments to regulate online content. In

Table 1. Research variables: symbols, questions in the V-Democracy dataset, and scales of measurement.

Variables	Symbol	Questions in V-Democracy Database	Scale of Measurement
Online media consumption	OMC	Do people consume domestic online media?	0–3
Social media usage for offline political actions	SMU	How often do average people use social media to organize offline political action of any kind?	0–4
Online mobilization for democracy	OMD	In this year, how frequent and large have events of mass mobilization for pro-democratic aims been?	0–4
Online media perspectives	OMP	Do the major domestic online media outlets represent a wide range of political perspectives?	0–4
Print/broadcast media critical	PBMC	Of the major print and broadcast outlets, how many routinely criticize the government?	0–3
Online media bias	OMB	Is there media bias against opposition parties or candidates?	0–4
Online media fractionalization	OMF	Do the major domestic online media outlets give a similar presentation of major (political) news?	0–4
Political polarization	PP	Is society polarized into antagonistic, political camps?	0–4
Government capacity to regulate online content	GRO	Does the government have sufficient staff and resources to regulate Internet content in accordance with existing law?	0–4
Participatory democracy index	PD	To what extent is the participatory principle achieved?	0–1

, we present the research variables, symbols, research questions, and units of measurement:

To test our research hypotheses, we operated a sample of EU member states in a long-term statistical series from 2000 to 2024. We used the year 2000 as a starting point in our temporal series, as it marked an increased growth of online media and platforms. We included 28 EU member states in our analysis as of 2020, when the UK left the European Union as a consequence of BREXIT. Our dataset comprises 695 observations that include both dependent and independent variables. Data regarding the current year (2025) is not available on the public V-Democracy databases.

The research methodology is quite quantitative, being based on both descriptive and inferential statistics. Moreover, at the descriptive level, we analyzed the statistical indicators of central tendency, dispersion, and statistical distribution for all research variables. Statistical data were analyzed using several data analysis software, such as JASP 0.19.3.0 and MS-Excel Data Analysis for descriptive statistics and IBM SPSS 29 for inferential statistics. Both Shapiro–Wilk and Kolmogorov–Smirnov were applied to test the normality of the statistical distribution. Due to the fact that p-values in both cases are significant (p < 0.001), we used nonparametric tests to describe and explain the interactions between dependent and independent variables. Therefore, we used Kruskal–Wallis (H) to compare differences between research variables and effects. In accordance with these perspectives, at the inferential level, we used quantile regression to estimate the significant predictors of the participatory democracy index in EU member states. By these premises, we defined our statistical model in the following terms:

Let Y be the dependent variable and X the independent factor, the quantile regression between X and Y is as follows:

$$Q_Y\left(\tau |X\right)={\beta }^{\left(\tau \right)}·X$$

where $Q_Y\left(\tau /X\right)$ is the conditional quantile function of Y if X, $\tau ϵ\left(0;1\right),$ and $\tau$ is the quantile level. ${\beta }^{\left(\tau \right)}$ is the specific vector of quantile coefficients, and X is the independent variable/predictor of the model.

The regression coefficients ${\beta }^{\left(\tau \right)}$ are estimated using the following formula:

$${min}_{\beta }={\displaystyle \sum _{i=1}^n}\rho \tau (y_i-x_i^T\beta )$$

Let $X_1,X_2\dots \mathrm{a}\mathrm{n}\mathrm{d} X_n$ be a set of predictors, then the quantile regression has the following formula:

$$Q_Y\left(\tau |X_1,X_2..X_n\right)={\beta }_0^{\left(\tau \right)}+{\beta }_1^{\left(\tau \right)}·X_1+{\beta }_2^{\left(\tau \right)}·X_2+\cdots +{\beta }_n^{\left(\tau \right)}·X_n$$

where ${\beta }_0^{\left(\tau \right)}$ is constant and ${\beta }_{1,n}^{\left(\tau \right)}$ are regression coefficients of the independent factors. In order to evaluate the impact of the independent variables in the field of participatory democracy, we used several quantiles: $\tau =0.25;\tau =0.5;\tau =0.75$. Based on these assumptions, we used three statistical models:

Model I ($\tau =0.25)$:

$$Q_Y\left(0.25|X_1,X_2..X_n\right)={\beta }_0^{\left(0.25\right)}+{\beta }_1^{\left(0.25\right)}·X_1+{\beta }_2^{\left(0.25\right)}·X_2+\cdots +{\beta }_n^{\left(0.25\right)}·X_n$$

Model II ($\tau =0.5)$:

$$Q_Y\left(0.5|X_1,X_2..X_n\right)={\beta }_0^{\left(0.5\right)}+{\beta }_1^{\left(0.5\right)}·X_1+{\beta }_2^{\left(0.5\right)}·X_2+\cdots +{\beta }_n^{\left(0.5\right)}·X_n$$

Model III ($\tau =0.75)$:

$$Q_Y\left(0.75|X_1,X_2..X_n\right)={\beta }_0^{\left(0.75\right)}+{\beta }_1^{\left(0.75\right)}·X_1+{\beta }_2^{\left(0.75\right)}·X_2+\cdots +{\beta }_n^{\left(0.75\right)}·X_n$$

Using these statistical methods, we aim to estimate the significant predictors that could influence the dynamics of the participatory democracy within EU member states over the past 25 years. The following section of the paper highlights the main findings and relevant digital factors for explaining civic engagement in the countries included in our sample. Statistical results are considered statistically significant when $p\le$0.05.

3. Results

This paper highlights the importance of social media usage and media bias in the sphere of participative political behavior across EU member states. If social media usage is related to an increased level of political participation and mobilization for democracy, media bias creates premises for civic demotivation and political mistrust. Biased media distort public opinion, generating misinformed voting and civic engagement. Also, media bias is related to political polarization, contributing to the echo chamber effect and ideological divides. An increased level of media bias erodes public trust in journalism and political institutions. Therefore, there are significant adverse effects generated by an increased level of media bias in the field of democratic practices and civic participation. This section presents the differences between EU member states regarding online media consumption, media bias, and participatory democracy. Statistical results indicate the importance and significance of the mass media predictors (both online and traditional media) in creating models of participatory behavior in EU member states.

3.1. Digital Environment and Participatory Democracy Across EU Member States

This section presents the evolving dynamics of civic engagement and participatory democracy in EU member states in a context characterized by an important digital transformation. By examining the descriptive statistics of the key variable, we aim to create an important insight regarding the relevance of the digital environment and quality in the field of democratic practices. This analysis provides a comprehensive perspective of the complex relationship between political behavior and digital transformation, stressing both opportunities and challenges in fostering an informed and inclusive citizen involvement.

Participatory democracy (PD) captures the level of participation in public affairs beyond the electoral process. Using the scale from 0 to 1, the median value in EU member states is 0.62 and the std. deviation is 0.06. These values reflect a moderate level of civic participation in public affairs across the states included in the sample. The political environment could provide good channels for civic participation in public life outside of the basic electoral process. Civic organizations, public consultations, and involvement in decision-making are several mechanisms for civic participation in public life. The score reflects the average citizens’ involvement in the public sphere, with significant disparities generated by the geographical position, regime type, and historical legacy. Regarding the particularities of social media usage, we stress the fact that online media consumption (OMC) has a median value of 2.53 and std. deviation of 0.34. These values indicate a consistent rate of online media consumption with low variation rates, expressing a common habit of the individuals included in our statistical sample. The quantitative results align with existing research, which suggests that an increased proportion of the world population uses digital media daily. Therefore, 92% of the internet users consume social media, gaming, and news (General Trends in Online Media Consumption). In accordance with these results, social media usage for offline political actions (SMU) has a median value of 2.55 and std. deviation of 0.62. SMU is an important predictor of participative behavior, with increasing access to political news and civic engagement. This factor could boost political interest and mobilization. Platforms such as Facebook, TikTok, Instagram, or X are utilized for political mobilization, petition dissemination, and the organization of various social movements that can disrupt the political order. In this respect, these platforms could have an ambivalent role in the field of political behavior: they can either build social and political trust or erode it by using a large spectrum of fake news. Online mobilization for democracy (OMD) has a median value of 0.96 and std. deviation of 0.69. Although statistical results indicate increased median values of online media consumption and social media usage for offline political actions, the score of mobilization for democracy demonstrates a decreased level of political engagement in offline movements that support democratic regimes. The high coefficient of variation reflects an increased level of dispersion in the field of mobilization for democracy. The mobilization for democracy is not uniformly distributed, with several countries having a minority of actively engaged citizens. Taking into account all these descriptive results, we underline the fact that EU member states are characterized by an increased level of media consumption and social media usage, with low rates of political mobilization for democracy. The high level of dispersion is explained by both historical and cultural factors that shaped different ways of democracy in Western and Eastern countries of the European Union.

Together with social media usage, the quality of the digital media environment is an important set of predictors that could explain the evolution of participative democracy in EU member states. With regard to the online media perspectives (OMP), statistical values indicate a median value of 3.49 and a std. deviation of 0.28. Online media perspectives are still an important source of civic information. The accessibility and editorial rigor are relevant features of traditional media, fostering an informed and engaged citizen in public affairs. With regard to this variable, we used the level of print/broadcast media critical (PBMC) to estimate media and speech freedom. Statistical results indicate a median value of 2.81 and a std. deviation of 0.26. Statistical values show that traditional media remains a vector of democratic participation. Despite traditional media continuing to provide a stable and credible information framework, results indicate that the digital landscape is characterized by ideological polarization and information fragmentation. In this respect, online media bias (OMB) is a key variable for stressing the impact of ideological polarization in the field of political participative behavior. Therefore, descriptive statistics indicate a median value of 0.41 with a std. deviation of 0.25. The values associated with media bias are mostly moderate, and statistical distribution highlights the existence of different groups that perceive strong political and ideological influences in media content. In connection with media bias, online media fractionalization (OMF) is seen as a major challenge for public communication, journalism, and social cohesion. Descriptive statistical analysis indicates a median value of 1.13 and a std. deviation of 0.68. Central tendency values might indicate the division of the audience into ever smaller groups, each consuming personalized content, often from ideologically distinct sources. This fact is evident in the fields of social cohesion, political polarization, and public opinion aggregation. An important marginal effect of online media bias and online media fractionalization is represented by political polarization. Academic literature on political polarization highlights a series of negative implications in the field of governance, public choice, political participation, and democratic order. However, an increased level of political polarization (PP) is related to legislative gridlock, erosion in institutional democratic trust, information fragmentation, radical groups, and social violence. The median value of political polarization across EU member states is 1.41 with a std. deviation of 0.98. Statistical results suggest that political polarization is a common phenomenon in the countries included in our sample. Therefore, there is a significant, but not uniform, ideological divide, generated by demographic, cultural, and media differences. The broad distribution of the statistical values might suggest that certain groups could experience political polarization more intensely. Media fractionalization and ideological bias could explain this important social effect in the field of democratic societies. Concerning the governmental capacity to regulate online content (GRO), we estimated a median value of 2.37 and a std. deviation of 0.68. These values show that the government’s ability to regulate online content is perceived as moderate. Regional and cultural differences in the perception of the political authority, personal experiences, and social trust in political actors and democratic institutions could explain these significant variations. Based on these quantitative results,

Table 2. Descriptive statistics for research variables.

Descriptive Statistics	PD	SMU	OMC	OMD	OMP	PBMC	OMB	OMF	PP	GRO
Median	0.62	2.55	2.53	0.96	3.49	2.81	0.41	1.13	1.41	2.44
Std. Deviation	0.06	0.62	0.34	0.69	0.28	0.26	0.25	0.68	0.98	0.68
Skewness	−1.02	−0.40	−0.89	1.15	−0.49	−1.33	1.56	0.61	0.36	−0.62
Kurtosis	0.85	−0.14	0.50	1.42	−0.64	1.12	3.79	−0.14	−1.11	−0.29
Range	0.26	2.48	1.34	2.88	1.01	0.98	1.21	2.34	3.28	2.56
Minimum	0.45	1.00	1.62	0.31	2.84	2.00	0.09	0.24	0.17	0.83
Maximum	0.71	3.48	2.95	3.18	3.85	2.98	1.30	2.58	3.45	3.38
25th percentile	0.58	2.13	2.38	0.57	3.22	2.60	0.32	0.90	0.85	2.06
50th percentile	0.62	2.55	2.53	0.96	3.49	2.81	0.41	1.13	1.41	2.44
75th percentile	0.64	2.97	2.67	1.55	3.60	2.91	0.59	1.44	2.53	2.86

shows the descriptive statistical values for all the variables included in our analysis:

An important issue in the field of participatory democracy in EU member states is represented by a decline in political participation and active citizen involvement over the last twenty years. The longitudinal data show a downward trend in the participatory democracy index, with maximum values in early 2010 (PD = 0.63) and minimum values in 2024 (PD = 0.56). This decline in participatory democracy could indicate the erosion of democratic engagement, civic activism, and disillusionment with the political process. Generational studies and democratic theoretical perspectives underline that new generations are less interested in public affairs, and a decreased level of social trust is strongly related to flawed democracies. Together with these aspects, the complexity of the political processes and new forms of political participation are relevant for explaining this trend in the field of participatory democracy. Figure 1 presents the long-term dynamics of the participatory democracy index from 2010 to 2024:

Taking into consideration the geographical distribution, we underline the fact that the highest scores of PD are found in Northern and Western EU member states. In this respect, Denmark (PD = 0.713, $\sigma =0.008$), Sweden (PD = 0.667, $\sigma =0.01)$, Belgium (PD = 0.644, $\sigma =0.008$), and Germany (PD = 0.643, $\sigma =0.01)$ are relevant examples of an increased level of participatory democracy index. In contrast, Southern and Eastern EU member states are characterized by the lowest scores of democratic participation: Croatia (PD = 0.522, $\sigma =0.04)$, Hungary (PD = 0.488, $\sigma =0.14$), Romania (PD = 0.457, $\sigma =0.06$), and Bulgaria (PD = 0.453, $\sigma =0.01)$. Figure 2 presents the average values of the PD in EU member states.

The rise in digital platforms has a significant impact on the sphere of civic participation and engagement in correlation with the dynamics of the participatory democracy index. With regard to the dynamics of social media usage for offline political actions, data reflect a positive trend in using these platforms between 2010 and 2024. Moreover, a similar statistical trend is relevant for online media consumption. Thus, online media consumption is a key indicator of the transformations of the digital environment in a new framework for political information and engagement. Consequently, social media usage and consumption reveal new forms of political participation through users’ interactions with online content, political campaigns, political discourses, and prime-time news. The year 2010 could be considered a kind of “turning point”, marking the moment of competition between traditional media and new online media perspectives. Around this time, most internet platforms, such as Twitter, Facebook, and YouTube, became mainstream and were used on a large scale by internet users. This year marked the transition from traditional political behavior to digital civic engagement, as people used all these digital platforms to participate in political debates, forums, online discussions, and deliberations. These aspects are relevant for understanding generational differences and significant shifts in social and civic engagement. Figure 3 presents the evolution of social media usage for offline political actions and online media consumption as average values for EU member states from 2000 to 2024.

An important effect of digital media transformation consists of how people think, understand, and interpret political context and political information. A prominent impact on the sphere of social cognitions and political behaviors is the diversity of information sources, ideological polarization, and the increased amount of information to which individuals are exposed. This fact led to information polarization and audience fragmentation for various media channels. In this respect, alongside the “mere exposure effect” (Zajonc, 2001), a new social and psychological phenomenon has appeared, based on both “selective perception” and “echo chambers” (Garrett, 2009). Selective exposure, information repetition, and algorithmic filtering are the central coordinates of the cognitive bias and the “echo chamber” effect to which online media users are exposed. The main consequence of this type of digital interaction is represented by ideological polarization of some groups or individuals, who become closed and radicalized. Alongside this effect, we can observe that the lack of informational diversity and media bias is strongly related to the prevalence of fake news. These aspects contribute to the accentuation of the cognitive bias and a high degree of either social polarization or isolation. An important nexus between online media bias and participatory democratic behavior lies in the manipulation of information and public opinion. Biased online media often cover narrative frames, shaping citizens’ perceptions and political behavior. A first aspect of this complex interaction is represented by the fact that information selection creates an optimal framework for filtered social reality and manipulated public opinion. A second aspect derived from this complex interaction between online media bias and civic behavior is given by the narrative framing. In this context, the presentations of political events can lead to a single set of emotions or values, excluding other ways of understanding, interpreting, or explaining political reality. A third important aspect of this correlation consists of the diminishing of critical thinking and objective evaluation of the political reality through the emphasis of a single version, which is perceived by most of the internet users as the truth. The prevalence of the cognitive bias represents a final relevant aspect of this association. This bias is correlated with an increased level of political polarization, with individuals eliminating alternative sources of information that contradict their opinions or expectations. Overall, informational bias is more likely to lead to subjective thinking, informational confusion, unnecessary information accumulation, and the development of various stereotypes.

With regard to the statistical data related to online media fractionalization and online media bias, we observed significant differences in the dynamics of these indicators between 2000 and 2024. Online media fractionalization remains relatively constant over 25 years of analysis. In contrast, online media bias has experienced a positive trend since 2015, with a significant acceleration trend after 2021. The current evolution of the online media bias produces social and political polarization, emotional and partisan perspectives, as well as an increase in the intensity of the virulence of the negative news and content. An important consequence of this phenomenon is the lack of trust in traditional sources, leading to a migration to social platforms where the degree of information bias is very high. Marginal effects of this phenomenon are met in the sphere of democratic societies, where an increased level of information bias is more likely to be correlated with distorted perceptions and radical political behaviors. Media bias remains a challenge for journalists who want to maintain ethical standards in their approach, in a digital landscape characterized by informational and commercial pressure. Figure 4 presents the dynamics of online media fractionalization and bias as average values over 25 years of data analysis.

Important transformations in the field of political socialization and participation characterize the last decades. These transformations created an apparent democratization of access to information, but also the emergence of systemic risks such as political polarization and information bias. Although the online media fragmentation has remained relatively constant, its effects in the field of political behavior are increasingly visible. Overall, internet users and online media consumers are divided into distinct information communities, creating opportunities for the emergence of “echo chambers” and limited social or political dialogue. The digital environment provides rapid access to information. However, it could create cognitive vulnerabilities and bias through the constant exposure to partisan content. The quality and informational coherence have started to suffer significant changes that might lead to distorted perceptions of the social and political reality. The rise in partisanship and the presence of “echo chambers” for most of the platform’s users influences the way citizens vote, discuss, and interact with democratic institutions.

3.2. Traditional vs. Digital Media; Regional Differences

This section underlines the significant statistical differences between the study variables depending on the geographical regions of the EU member states. This research design presents the regional differences in the field of participatory democracy in correlation with social media usage and the quality of the digital landscape. In the cases where online media bias influences participatory behavior, strategic measures (governmental policies, educational initiatives) could be designed to mitigate its adverse effects. In this respect, some regions could be exposed to partisan media sources. Identifying these patterns helps to understand opinion formation and information fragmentation in the territory. In order to assess these regional differences between research variables, we performed the Kruskal–Wallis H-test. The level of significance is p < 0.05.

Table 3. Kruskal–Wallist H-Test.

Variables	Test Statistic	p (Sig.)
Online media consumption	7.016	0.071
Social media usage for offline political actions	5.830	0.120
Online mobilization for democracy	8.505	0.037 *
Online media perspectives	7.223	0.065
Print/broadcast media critical	16.517	0.001 *
Online media bias	14.359	0.002 *
Online media fractionalization	11.740	0.008 *
Political polarization	12.096	0.007 *
Government capacity to regulate online content	6.988	0.072
Participatory democracy index	6.399	0.094

* Statistical results are significant with p < 0.05.

shows the significance values (p) for the research variables, taking into consideration the geographical position of the EU member states.

Quantitative results confirm significant differences between geographical regions for several research variables, such as online mobilization for democracy, print/broadcast critical media, online media bias, online media fractionalization, and political polarization. Regarding regional differences in democracy mobilization, statistical results indicate that countries in the Southern part of the EU have high median values of online mobilization (OMD = 1.70, $\sigma =0.96)$. Another important value is specific to Central and Eastern European countries (OMD = 1.19, $\sigma =0.46)$. Statistical values from the Central-Eastern and Southern parts of the EU could signify the presence of civic activism in the face of political and economic dysfunctions and a collective reaction to social and political polarization. An important issue with significant differences between geographical regions is that traditional media, measured by print and broadcast media, is critical. This variable encompasses both the “watchdog” and “freedom of speech” functions of traditional media in democratic societies. Therefore, statistical results confirmed that Northern (PBMC = 2.91, $\sigma =0.14)$ and Western (PBMC = 2.90, $\sigma =0.05$) countries have increased values in the field of the critical role played by traditional media (both print and broadcast). In Northern and Western EU member states, traditional media tends to fulfill a more pronounced critical function. In these regions, both print and broadcast media act as “watchdogs” and foster informed public discourse. Both “agenda-setting” and “watchdog” roles ensure an increased level of political participation and democratic accountability in these societies. In contrast, in Central and Eastern European countries, we observed the lowest scores related to the critical role played by traditional media (PBMC = 2.53, $\sigma =0.21$). Figure 5 highlights the regional differences between EU member states in the field of traditional media.

Relevant examples of the critical role played by traditional media in Northern and Western countries are represented by Denmark (PBMC = 2.94), Sweden (PBMC = 2.94), Estonia (PBMC = 2.92), Finland (PBMC = 2.91), Germany (PBMC = 2.97), France (PBMC = 2.91), and Austria (PBMC = 2.9). In contrast to these countries, traditional media in Southern and Eastern Europe are weak or moderate, playing roles such as “watchdog” and “agenda-setting”. Therefore, illustrative examples for these regions are Croatia (PBMC = 2.66), Greece (PBMC = 2.6), Bulgaria (PBMC = 2.26), Romania (PBMC = 2.26), and Hungary (PBMC = 2.19).

Based on these assumptions, we observed that online media fragmentation and online media bias are relevant variables with significant differences between geographical regions. In the field of online media fractionalization, we estimated that Northern countries registered the lowest values (OMF = 0.372, $\sigma = 0.378)$. In contrast, Southern (OMF = 1.46, $\sigma = 0.67)$ and Eastern European countries (OMF = 1.24, $\sigma = 0.57)$ have registered the highest scores in the field of online media fractionalization. An important result is estimated in the field of online media bias. Northern (OMB = 0.35, $\sigma = 0.12)$ and Western (OMB = 0.27, $\sigma = 0.10)$ countries are more likely to be correlated with decreased levels of informational bias. In this respect, fully democratic countries are based on civic culture, participatory democracy, and informed citizens. At the opposite pole, Southern (OMB = 0.49, $\sigma = 0.11)$ and Eastern (OMB = 0.712, $\sigma = 0.308)$ countries are characterized by an increased level of media bias. Figure 6 presents the distribution of median values related to online media bias by geographical region.

Relevant examples of countries with a low level of online media bias are Denmark (OMB = 0.09), Sweden (OMB = 0.281), Belgium (OMB = 0.292), France (OMB = 0.225), and Luxembourg (OMB = 0.197). In Southern and Eastern countries, the statistical results indicate an increased level of online media bias in countries such as Croatia (OMB = 0.626), Greece (OMB = 0.607), Spain (OMB = 0.494), Italy (OMB = 0.430), Poland (OMB = 0.713), Bulgaria (OMB = 0.786), Romania (OMB = 0.867), and Hungary (OMB = 1.303). In connection with these results, political polarization has significant differences between geographical regions. In line with the previous results, political polarization registered low values in Northern (PP = 0.635, $\sigma = 0.365)$ and Western countries (PP = 1.12, $\sigma = 0.775)$. In contrast to these results, Southern (PP = 2.50, $\sigma = 0.887)$ and Eastern countries (PP = 2.60, $\sigma = 0.921$) are more likely to be correlated with a high degree of political polarization. To assess the correlation between political polarization and online media bias, we counted the direct effect of media bias on social and political polarization. Therefore, we estimated a moderate positive correlation with R² = 0.47, p < 0.01. An increased level of online media bias is strongly related to political polarization in countries such as Hungary, Romania, Slovakia, and Poland. Low online media scores are correlated with low levels of political polarization in Western and Nordic democracies. Figure 7 captures the moderate positive correlation between online media bias and political polarization within countries included in the sample.

3.3. Online Media Bias, Social Media Usage, and Participatory Democracy: A Quantile Regression Model

This section extends the previous findings by investigating the complex interaction between participatory behavior and factors linked to online media usage within the context of today’s evolving digital landscape. In order to identify relevant and significant predictors of participatory democracy in EU member states, we performed a quantile regression. As we pointed out in the methodological section, we used this regression technique in accordance with our research data and normality tests of distribution. The data on participatory democracy were categorized into three quantiles, 0.25, 0.5, and 0.75, allowing for a differentiated analysis of engagement levels and their correlations with variables related to the digital environment. The first quantile (Q1 = 0.25) consists of seven states with values of participatory democracy from 0.453 to 0.569. In this category, we included countries like Romania, Poland, Hungary, Bulgaria, Malta, Cyprus, and Croatia. The second quantile (Q2 = 0.5) consists of seven countries with values of participatory democracy from 0.57 to 0.62. In this category, we included countries like the Czech Republic, Greece, the UK, the Netherlands, Luxembourg, Lithuania, and Latvia. The third quantile (Q3 = 0.75) has 14 countries with the highest values in the field of participatory democracy (0.63 $\le PD\le 0.713)$. In this category, we included the following EU member states: Slovenia, Slovakia, Spain, Portugal, Italy, Ireland, Germany, France, Belgium, Austria, Sweden, Finland, Estonia, and Denmark.

Table 4. Quantile regression models. Dependent variable: Participatory democracy.

	Model I		Model II		Model III
	q = 0.25		q = 0.5		q = 0.75
Parameter	$\beta$ $\mathrm{S}\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{d}\mathrm{a}\mathrm{r}\mathrm{d}\mathrm{i}\mathrm{z}\mathrm{e}\mathrm{d}$	p (sig.)	$\beta$ $\mathrm{S}\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{d}\mathrm{a}\mathrm{r}\mathrm{d}\mathrm{i}\mathrm{z}\mathrm{e}\mathrm{d}$	p (sig.)	$\beta$ $\mathrm{S}\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{d}\mathrm{a}\mathrm{r}\mathrm{d}\mathrm{i}\mathrm{z}\mathrm{e}\mathrm{d}$	p (sig.)
(Intercept)	0.1		0.166		0.139
Social media usage for offline political actions	0.280	<0.01	0.360	0.013	0.384	<0.01
Government capacity to regulate online content	0.129	<0.01	0.307	0.016	0.104	0.201
Online mobilization for democracy	−0.215	<0.01	0.205	0.136	−0.107	0.245
Online media consumption	0.082	0.09	0.198	0.210	0.365	0.02
Online media perspectives	0.142	0.01	0.01	0.944	−0.328	0.03
Political polarization	−0.076	0.016	0.206	0.201	0.802	<0.01
Print/Broadcast media critical	0.420	<0.01	0.299	0.112	0.703	<0.01
Online media bias	−0.206	<0.01	−0.439	0.033	−0.582	<0.01
Online media fractionalization	0.119	<0.01	−0.063	0.684	−0.066	0.529
R2	0.670	<0.05	0.522	<0.05	0.418	<0.05

Dependent variable: Participatory democracy, p < 0.05 (two-tailed).

captures the significant predictors and their standardized values in correlation with participatory democracy levels. Table 4 presents three models of quantile regression, the standardized coefficients of regression, and their statistical significances.

For data specific to the first quantile, the model has R² = 0.670, p < 0.05. The model captures relevant predictors of participatory democracy in the field of social media usage for offline political actions ($\beta = 0.280, p<0.01)$, online mobilization for democracy ($\beta = -0.215, p<0.01)$, traditional media represented by print/broadcast media critical ($\beta = 0.420, p<0.01),$ and online media bias ($\beta = -0.206, p<0.01)$. Although statistically significant values were obtained for other predictors, the regression coefficients indicate either a lack of meaningful correlation or only minimal associations with the participatory dimension of political behavior. An important influence in the field of participatory democracy is represented by print/broadcast media critical ($\beta = 0.420, p<0.01)$. Individuals who consume traditional media are more likely to develop participatory behavior. Also, social media usage has a positive impact in the field of participatory democracy ($\beta = 0.280, p<0.01)$. This factor suggests that the digital environment not only facilitates the expression of opinions but also influences political actions. This result supports the idea that online activism could have tangible effects on real political participation. Our findings highlight the negative impact of online media bias in the field of participatory democracy ($\beta = -0.206, p<0.01)$. In this respect, the exposure to partisan content could reduce motivation for political engagement. In practice, this factor is more likely to produce demotivated citizens and enhance the “echo chamber” effect of the digital environment.

In the second quantile, the regression model has a moderate value of R² = 0.522, p < 0.05. Relevant predictors related to participatory democracy are social media usage for offline political actions ($\beta = 0.360, p = 0.013)$ and government capacity to regulate online content ($\beta = 0.307, p = 0.016)$. Thus, social media is a bridge between offline actions and the digital environment. Moreover, the government’s capacity to regulate online content is important for reducing the spread of misinformation and fake news. This kind of political control might help to reduce political polarization and foster a democratic digital environment for debates and decision-making. The second regression model captures the same adverse effect generated by the online media bias in the field of participatory behavior. The regression coefficient between online media bias and participatory democracy ($\beta = 0.439, p = 0.033)$ suggests the negative moderate impact of media partisanship in the field of democratic participation. The perception that media outlets are biased could erode credibility and trust in democracy and institutions, amplifying both cognitive bias and political demotivation.

With regard to the third model, quantitative results pointed out a weak to moderate relationship between independent factors and the participatory democracy index, with R² = 0.418, p < 0.05. The model highlights the importance of social media usage in the field of participatory behavior. Therefore, we noticed significant positive correlations between social media usage for offline political actions ($\beta = 0.384, p<0.01)$, online media consumption ($\beta = 0.365, p = 0.02),$ and participatory democracy index within EU member states. Although the online environment is positively associated with participatory behavior, the quality of the digital environment predicts negative trends in the field of civic engagement and accountability. Thus, we estimated negative associations between online media perspectives ($\beta = -0.328, p = 0.03)$, online media bias ($\beta = -0.582, p<0.01),$ and the participatory democracy index. This model captures a moderate negative impact of the online media bias in the field of democratic participation across EU member states.

Empirical findings capture the negative marginal effects generated by online media bias in the field of participatory democracy. This variable is a key factor for explaining the dynamics of participatory behavior in countries characterized by moderate to high levels of political participation. In these situations, an increased level of media partisanship negatively affects civic trust, civic interest in public affairs, and the quality of the national democracy. With regard to the quantile regression models, the prediction lines (q = 0.25, 0.5, 0.75) indicate a decrease in the participatory democracy index. Regression lines for q = 0.5 and 0.75 show that even in countries with moderate and high levels of participation, the increase in online media bias leads to a significant decrease in the quality of democratic participation. The regression line for q = 0.25 confirms that this effect is present even in contexts with low participation, indicating a systemic impact of media partisanship. Our quantitative model offers an important analytical advantage, underlying that the negative relation is not uniform, but persists across the entire distribution of the participatory democracy index.

In this regard, another important result highlights significant statistical associations between traditional media variables and the level of political participation across EU member states. Thus, in both q = 0.25 and q = 0.75, print/broadcast media is positively and significantly associated with an increased level of political participation. For the countries included in the first quantile, we estimated a moderate positive correlation between the participatory democracy index and the usage of print/broadcast media ($\beta = 0.420, p<0.01)$. In line with this result, we estimated a strong positive correlation between these variables in the countries included in the third quantile ($\beta = 0.703, p<0.01)$. The effect of traditional media is more pronounced in countries with a high level of participation, suggesting that a critical attitude towards traditional media is associated with a more robust participatory democracy. Traditional media is more likely to stimulate democratic participation. Figure 8 presents traditional media and online media bias in correlation with participatory democracy.

In conclusion, these results support the idea of a demotivating effect generated by online media bias, which affects civic trust, interest in public affairs and decision-making, and the quality of democratic order. These results underscore the important role played by the balanced and credible media landscape in sustaining civic participation and fostering participative political culture.

4. Discussion

This section discusses the main research findings, integrating descriptive analysis, regional differences, and regression to highlight the relationship between the quality of the digital media environment and participatory democracy in EU member states. The study aimed to highlight to what extent the transformation of mass media in the digital space still preserves its “watchdog” or “agenda-setting” functions, mobilizing citizens for civic activism and democratic participation. Rivalry between media institutions, although associated with informational pluralism, can paradoxically weaken electoral accountability by increasing the risk of the media being captured by political interests (Trombetta & Rossignoli, 2021). The shift towards digital journalism, driven by artificial intelligence and disinformation, underscores the need to adapt European codes of ethics, promote transparency, and ensure rigorous source verification, alongside media education, to maintain information integrity and resilience (Forja-Pena et al., 2024). There is a consistent debate regarding a regulatory agenda for prohibiting disinformation and large technology platforms in the European Union, underlining the importance of protecting the public sphere and democratic values (García & Oleart, 2023).

In the digital era, online media creates the link between the public and private sectors. An important aspect of online media is its influence on civic and political behavior, as well as the democratic process. Scholars have formulated a significant theoretical corpus to highlight the ambivalence of online media: on the one hand, online media facilitate democratic participation; on the other hand, they can amplify social and political polarization, misinformation, and inequality (Dahl, 1971; Schradie, 2019; Jha & Kodila-Tedika, 2020; Fischer & Jarren, 2024; Wischmeyer, 2019). Online media created new forms of political participation, more spontaneous and less hierarchical. Equally, online media has significantly shaped new forms of political behavior across both digital and offline spheres. Based on these normative perspectives, scholars have underlined the fact that social networks and social media created a space for public debate and the expression of citizens’ political preferences (Lipschultz, 2018; Ardèvol-Abreu & de Zúñiga, 2017; Bennett & Segerberg, 2012). However, it creates digital communities that mobilize based on shared emotions, not necessarily rational arguments. An important adverse effect underlined in the academic literature is that online media could amplify social and political polarization and intellectual isolation through algorithms that filter the content (Thorson & Wells, 2016; Yoo & Gil de Zúñiga, 2019; Papa & Photiadis, 2021; Huckfeldt & Sprague, 1995). Online media platforms are not neutral; they shape content through algorithms and internal policies. Algorithms amplify similar content, promoting polarization and excluding divergent points of view (Rowden et al., 2014; Hoewe & Peacock, 2020; Yarchi et al., 2020; Hout & Maggio, 2021).

Using secondary statistical data, we aimed to observe differences between countries and variables in shaping different patterns of democratic participation. An important empirical finding consists of the fact that democratic participation has been increasing, but in recent years, a downward trend has been observed. After 2015, a significant decrease in the participatory democracy index was registered in most of the states included in our analysis. Northern and Western EU countries registered increased values in the field of participatory democracy. These countries have stable institutions, an active civic culture, and respect for fundamental freedoms. In parallel with the negative trends in the field of participatory democracy, statistical data indicate the rise in social media as a political tool. While online media consumption is higher, offline political engagement through social media has increased more steeply, suggesting that the digital platforms have become catalysts for social mobilization. The rise in online engagement through social media shows the democratic potential of the digital environment. However, passive consumption of online media does not guarantee active participation, amplifying isolation and social polarization via “filter bubbles” and “echo chamber” effects. Statistical indicators regarding online media fractionalization suggest a constant evolution over 25 years. The diversity of information sources has stabilized, but not necessarily in a beneficial way: online media users might remain in “information bubbles” without exposure to different opinions. In correlation with this result, online media bias is an important factor in understanding participatory democracy in the digital age. Therefore, the increased online media bias indicates a tendency towards partisanship, where platforms favor content that confirms users’ opinions. The evolution trend of the online media bias reflects the polarization of online discourse, influenced by algorithms, social media, and partisanship. In a democratic context, these trends undermine informed participation, favoring polarization and radicalization.

Another important statistical result is represented by regional differences in the field of participatory democracy, online media usage, and the quality of the digital environment. Comparing traditional media consumption by print/broadcast media with online media evolutions, we observed that traditional media varies significantly across the EU member states, with a clear advantage for Northern countries. This difference influences the level of political participation, trust in institutions, and the quality of democracy. Regions with less critical traditional media risk being more vulnerable to manipulation, fake news, and political polarization. In the field of online media consumption and partisanship, we observed that Northern countries are characterized by low media bias, suggesting more balanced and less partisan online media. In contrast, Eastern EU countries have the highest values of online media bias, indicating an increased level of polarization and possible ideological or political influences in the field of online content. In this case, an increased level of online media bias may affect the quality of public information, political trust in democratic institutions, and the participatory democracy index. In accordance with these results, we estimated a positive moderate correlation between the level of online media bias and political polarization (R² = 0.47, p < 0.01).

Using the quantile regression models, we observed an important effect generated by online media bias in the field of the participatory democracy index. Online media bias has a significant and consistent negative effect on the field of political participation in EU member states. Online media partisanship reduces democratic participation and civic trust in all the states included in the sample. In this context, online media bias is related to the selective and partisan presentation of information. Citizens constantly exposed to biased content may develop a distorted view of political reality, leading to demotivation or radicalization. Both social and traditional media are significant predictors of the participatory democracy index. In practice, the quality of the digital landscape influences the level of participation, trust, and civic accountability.

With regard to the research objective O_1, we tested the research hypothesis H₁ that assumes a positive correlation between traditional media and the participatory democracy index. Table 4 and Figure 8 indicate that print/broadcast media play a vital role in democratic engagement. It aligns with “watchdog” and civic forums’ roles of the press, as discussed in democratic theory (q_0.25: $\beta = 0.420, p<0.01;$ q_0.75: $\beta = 0.703, p<0.01$) (Norris, 2007; Bennett & Serrin, 2005; Anderson, 2014; Reiff, 2024). To achieve the research objective O₂, we tested the research hypothesis H₂. This research hypothesis assumes that governmental capacity to regulate online content in ways that protect liberal principles and freedom of expression is positively correlated with an increased level of participatory democracy. Our empirical findings are significant only for states included in the first two quantiles. Table 4 highlights a very weak positive correlation in states characterized by a low level of participatory democracy ($\beta = 0.129, p<0.01$). A weak to moderate impact of the governmental capacity to regulate online content was observed in the states characterized by a median value of the participatory democracy index ($\beta = 0.307, p = 0.016$) (Gilwald, 1993; Schejter & Tirosh, 2015; Sievi & Pawelec, 2025).

Hypothesis 3 examines the correlation between higher levels of online media consumption and the dynamics of the participatory democracy index, in accordance with the research objective O₃. Statistical results confirm a weak to moderate statistical correlation between social media usage and participatory democracy level. Table 4 presents the statistical results regarding these variables in the field of democratic participation. Online media consumption might reinforce political engagement rather than initiate it. At a higher level of participation (q = 0.75), the effect is weak to moderate and significant ($\beta = 0.365, p = 0.02$), indicating that online media consumption is more influential among highly engaged individuals (Fuchs, 2023; Castells, 1996; Lipschultz, 2018).

In line with the research objective O₄, the findings confirm H₄, which assumes that increased online media bias erodes participatory democracy by reducing citizens’ engagement and trust in the democratic process. The results highlight moderate negative effects of online media bias in the field of participatory democracy (q_0.25: β = −0.206, p < 0.01; q_0.5: β = −0.439, p = 0.033; q_0.75: β = −0.582, p < 0.01). This aligns with recent academic studies showing that online media bias can lead to disengagement, civic demotivation, and civic mistrust (Gillespie, 2018; Lipschultz, 2018; Jacobs & Spierings, 2016; Huckfeldt & Sprague, 1995). To achieve research objective O₅, we tested H₅ that assumes that an increased level of media bias is positively correlated with an increased level of political polarization (R² = 0.47, p < 0.01) (Rowden et al., 2014; Hoewe & Peacock, 2020; Hout & Maggio, 2021).

Regarding the limitations of the study, we stress that this research is conducted in a digital context, which is constantly changing, making it challenging to capture the complex and stable relationship between online media and participatory democracy (Giugni & Grasso, 2022; Pasquino, 2023). Platform dynamics, algorithms, structures, public trust, and civic engagement are evolving rapidly, affecting the universal validity of the research conclusions. There are limitations related to the data of the study. Data are cross-sectional, limiting the possibility of formulating causal relationships between the perception of the online media bias and participatory behavior in EU member states. Further research will take into consideration primary data for testing this complex and intricate relationship between political behavior, trust, traditional, and online media.

Although the study provides a comparative and longitudinal perspective on the relationship between online media and participatory democracy, several methodological and conceptual limits can be acknowledged. The V-Democracy dataset is subjective, being based on expert-coded evaluations rather than population-based surveys. The quantile regression design might be sensitive to data distribution and might not fully capture the complex relationship between online media bias and the participatory democracy index. Additionally, the theory of democratic participation encompasses a reflective dimension that is difficult to measure. Critical thinking and contemplative positioning could influence political behavior. These aspects are not captured by statistical indicators in the current statistical analysis. Other important limits are related to the lack of conceptual equivalence between traditional media and online media bias. While online media bias reflects the editorial and ideological partisanship in online contents, government criticism expressed by print/broadcast media serves as a proxy variable for media independence. These conceptual limitations highlight the challenges of comparing traditional and digital media environments.

Further research will take into consideration primary data for testing this complex and intricate relationship between political behavior, trust, traditional, and online media, with a more extended focus on the impact of new technologies on public sphere participation in different societies.

Nevertheless, we may extend the research based on recent studies that advocate for a more effective cooperation between policymakers and educators in the current era of rapid change in digital communication, with significant impact on the mental health of the young generation that will affect the entire public sphere (Chang et al., 2025). For example, when assessing the public’s perception of AI-generated journalism, a dual perception is found: potentially objective and impartial, but susceptible to errors and biases. Skepticism increases with the level of AI literacy of users (Yeste-Piquer et al., 2025). In the same vein, other studies analyze the complex challenges of disinformation and misinformation in the digital environment, insisting on the need to improve media literacy skills in adults. For example, the analysis of Greek-speaking societies, Cyprus and Greece, where the level of digital literacy is below the EU average, shows the need for sustainable and adapted educational programs, gamification, interactive learning methodes, institutionally supported continuous training, media literacy campaigns that stimulate critical thinking, and multiple verification of sources (Taxitari et al., 2025, pp. 12–18).

5. Conclusions

This study highlights the complex relationship between the quality of media in the digital landscape and participatory democracy across EU member states. Using a long-term statistical design, the study captures the complex interplay between online media usage and consumption, online media bias and fractionalization, and the intensity of political participation in EU member states. This research offers several original contributions: (a) methodologically, it applies a quantile regression design to capture nuanced trends in online media influence and democratic participation; (b) it confirms the direct effect of online media bias on political polarization and the major impact in the field of participatory political behavior. These models allowed for a nuanced analysis, revealing that the effects are not uniform but variable depending on the intensity of participation. In a context marked by the fragmentation of public discourse, online media facilitates the expression of political preferences and civic mobilization. On the other hand, online media bias can distort democratic deliberation through echo chambers and partisan information. Using a quantitative research design, the study presents the negative trend in participatory democracy over 25 years in EU member states, in correlation with increasing online media consumption and partisanship. Statistical differences are relevant for understanding the important role of “watchdog” played by print/broadcast media in Western and Northern states of the EU. In contrast, Southern and Eastern EU countries are characterized by an increased level of online media bias and fractionalization. This research contributes to the understanding of how online media bias affects political participation in EU member states, but also highlights the need for longitudinal, multidimensional, and interdisciplinary studies to capture this evolving and complex correlation fully.