The Polarization Paradox: Social Media, Young Voters, and the Challenges to the Open Society

Abstract

The ideal of the open society, grounded in rational deliberation and pluralism, faces growing challenges in the digital age. This study examines how affective polarization—characterized by emotional hostility and antagonism across political identities—and populist sentiments among young Portuguese voters (ages 18–21) reshape democratic discourse, with a focus on the role of social media platforms. Based on a sample of 130 first-time voters in the April 2024 legislative elections, the results show that individuals with populist attitudes are 27% more likely to exhibit affective polarization. Moreover, support for radical right narratives is associated with a 27.6% increase in polarization. Notably, X (formerly Twitter) is associated with higher levels of affective polarization than Instagram, highlighting the differential effects of platform architectures. The study provides empirical evidence for the mediating role of digital platforms in fostering discursive fragmentation and ideological antagonism. It contributes to theories of digital democracy and political communication by clarifying how algorithmic environments intensify group-based hostility and undermine the normative foundations of the open society.

1. Introduction: The Open Society and the New Digital Ecosystem

The open society, as envisioned by Karl Popper, is grounded in transparency, critical thinking, and pluralism. It depends on freedom of expression and rational debate, supported by institutions that enable scrutiny and correction of error (Popper 1945). Historically, this model was sustained by traditional media, which acted as gatekeepers of reliable information (Shoemaker and Vos 2009). However, transformations in the media landscape, from television to the rise of social media, have fragmented public discourse and shifted communication toward entertainment and emotional engagement (Postman 1985; Kubin and von Sikorski 2021).

In this article, we adopt the concept of Open Society 2.0 to refer to the digital transformation of the public sphere and the challenges it poses to deliberative democracy, such as algorithmic mediation, affective polarization, and platform capitalism. According to Mounk (2018, 2022), these dynamics present significant threats to the integrity of democratic systems by reinforcing echo chambers—environments where users are exposed primarily to opinions that reinforce their existing beliefs (Pariser 2011), diminishing opportunities for pluralistic deliberation and deepening social divisions. His analysis emphasizes that the functioning of social media platforms and digital technologies can undermine trust in democratic institutions and hinder civic engagement, thereby generating a legitimacy crisis that threatens the sustainability of substantive democracy. Consequently, the notion of Open Society 2.0 highlights the urgent need to rethink participatory structures and develop strategies that foster inclusion and plurality within the context of rapidly evolving digital environments (Mounk 2018, 2022). In the digital age, social platforms amplify sensational content and reinforce ideological bubbles, closed networks of thought that limit exposure to diverse perspectives (Pariser 2011), challenging the foundations of democratic deliberation (Allcott et al. 2019; Boulianne and Larsson 2023). The weakening of traditional verification filters and the prioritization of engagement over accuracy contribute to affective polarization—a process that erodes trust and reduces shared spaces for dialog (Levy 2021; Pinkus 2018). As Iyengar et al. (2012) and Allcott et al. (2020) show, this dynamic fosters deep divisions across political identities, with consequences for cohesion and democratic resilience. Young voters, especially digital natives, are among the most exposed to algorithmically curated content and emotionally charged discourse. Their centrality in the evolving public sphere has been highlighted in studies on digital socialization and civic engagement (Loader et al. 2014; Boulianne and Theocharis 2020).

This study contributes to ongoing debates by analyzing how radical right populist parties (RDPs) foster affective polarization among young voters (Gidron et al. 2023). Although theoretical accounts have highlighted the role of social media in fragmenting the public sphere and undermining deliberative norms, empirical research on these dynamics remains limited in the Portuguese context, especially regarding first-time voters in the aftermath of the 2024 legislative elections. Addressing this gap, the study focuses on Portuguese youth aged 18 to 21, a demographic that relies heavily on social media for political information and is particularly susceptible to digitally mediated populist discourse. Grounded in the conceptual framework of Open Society, it examines whether populist attitudes and support for radical right narratives are associated with higher levels of affective polarization and how different platform architectures, particularly X and Instagram, shape these dynamics. The analysis is based on a sample of 130 respondents who participated in the April 2024 elections, bridging theoretical insights with empirical evidence on the digital mediation of political antagonism.

By examining the intersection of populism, polarization, and platform architecture, it contributes to broader discussions on how to safeguard democratic values in a fragmented and algorithmically mediated public sphere. Based on the literature, we hypothesize that (i) supporters of radical right parties are more likely to express populist sentiments and hostility toward perceived out-groups (Gidron et al. 2023); (ii) young people who support radical right narratives, such as anti-”gender ideology” discourse, display greater affective polarization (Iyengar et al. 2024); and (iii) different platforms foster different levels of polarization depending on their affordances and algorithmic logics (Allcott et al. 2020; Kubin and von Sikorski 2021).

2. Literature Review

2.1. Affective Polarization and the Erosion of Democratic Debate

Popper’s notion of the open society stresses rational deliberation, institutional transparency, and a commitment to pluralism (Popper 1945), offering a normative framework that remains highly relevant for analyzing contemporary democratic challenges. Building on this legacy, Weinert (2025) argues that the preservation of open societies depends on more than formal institutional design; it requires the continuous reinforcement of critical scrutiny and reflexivity within democratic structures. Such mechanisms of accountability, institutionalized and sustained over time, are indispensable for fostering contestation and self-correction, thereby ensuring the adaptability of democratic systems.

At the same time, Weinert underscores the civic dimension of openness, emphasizing that democratic vitality cannot be secured without trust in institutions, civic responsibility, and active participation. Yet, these virtues are increasingly eroded by the combined pressures of populism and disinformation, which corrode the social foundations upon which open societies rest. Consequently, open societies must be conceived not as fixed achievements but as dynamic and inherently incomplete projects, whose endurance depends on the constant renewal of democratic practices and the cultivation of civic engagement.

In contrast, the rise of affective polarization—emotional hostility across political identities—threatens these ideals by undermining democratic discourse (Iyengar et al. 2012; Gidron et al. 2023). Affective polarization not only deepens societal divisions but also erodes trust in institutions and fosters intolerance toward opposing viewpoints (Kiousis 2001; Lan and Tung 2024). Populism, especially in its radical-right variants, further intensifies this threat by promoting antagonistic worldviews and reducing the political field to a binary of “the people” vs. “the elite” (Mudde and Kaltwasser 2017). Such discourse encourages simplification and scapegoating, which diminishes the space for deliberative engagement and fosters distrust (Laclau 2005). This logic gains strength in the context of algorithmically driven platforms that reward divisive and emotive content, thereby amplifying polarization and populist rhetoric (Kubin and von Sikorski 2021; Bennett and Segerberg 2013). Moreover, the increasing reliance on emotionally charged content exacerbates the challenge to uphold the norms of rational debate and pluralism essential to the open society (Tucker et al. 2018). As a result, the intersection of affective polarization, populism, and algorithmic influence presents a formidable obstacle to the realization of Popperian ideals in the digital age.

The intersection of these dynamics—affective polarization, populism, and platform logic—forms the conceptual core of this study, which examines how these forces manifest themselves among young voters in a digital public sphere increasingly disconnected from traditional forms of deliberation (Habermas 2006; Mounk 2022). Affective polarization, a dimension of political polarization, reflects how individuals emotionally identify with their group while rejecting ideological opponents (Sharma and Sood 2020; Gidron et al. 2023). It goes beyond policy disagreement, fostering negative perceptions based on identity markers such as preferences, religion, or sexuality, reinforcing group-belonging through contrast.

In advanced democracies, political hostility fosters social distancing and weakens the capacity for dialog and compromise (Iyengar et al. 2012). As Popper (1945) emphasized, the viability of an open society depends on pluralism and critical scrutiny, which promote contestability and reflexivity. Yet, these foundations are undermined when partisanship fuses with core social identities. Under such conditions, ideological disagreement is no longer perceived as legitimate dissent but as a personal or group-level threat, eroding the critical distance necessary for deliberation.

This identification diminishes openness to alternative viewpoints and narrows the space for societal differentiation, as individuals retreat into in-group narratives. As a result, the balance that sustains an open society, marked by institutional independence and the capacity for critique, is destabilized. Polarization grows, while the reflexive mechanisms essential for democratic resilience and evolution are weakened.

Digital media intensifies this fragmentation. Platforms like Facebook and X (formerly Twitter) privilege engagement over deliberation, while YouTube and TikTok reinforce content homogeneity through algorithmic repetition (Allcott et al. 2019; Cinelli et al. 2021). These dynamics nurture ideological echo chambers and emotional polarization, conditions conducive to the spread of populist narratives.

The paradox of Open Society 2.0 lies precisely in this tension: while digital media expand access and participation, they simultaneously erode the discursive foundations of democratic life. Highly polarized users often reject information from rival groups, deepening institutional distrust and dismantling shared interpretative frameworks (Levy 2021; Iyengar et al. 2024). As Popper warned, democracy demands rational debate—yet the algorithmic structures of today’s platforms often constrain rather than enable such exchange.

Understanding how social media amplifies affective polarization is thus vital to preserving the conditions for pluralistic democracy. The notion of Open Society 2.0 calls for a rethinking of participatory structures and the development of strategies that sustain inclusion, contestability, and plurality within rapidly evolving digital environments (Mounk 2018, 2022).

2.2. Digital Populism and the Crisis of Truth

Reflecting on Popper’s open society is particularly relevant considering the rise of digital populism and the erosion of truth. While the open society promotes rational debate and the free exchange of ideas (Popper 1945), the architecture of social media favors emotional, controversial content, challenging these ideals. Historically, populists gained visibility through mass media but were constrained by journalistic filters—the editorial norms, fact-checking processes, and professional standards that traditionally governed the production and dissemination of news (Reese and Shoemaker 2016). With the rise of the Internet, these constraints diminished, leading to a new “one-step flow” of unmediated communication (Bennett and Manheim 2006).

This shift has fueled anti-media populism, marked by growing distrust in traditional journalism and the perception of the media as aligned with political elites (Fawzi and Krämer 2021; Krämer 2014; Stroud 2008). Social networks now function as alternative arenas for political discourse, where emotional engagement and algorithmic dynamics amplify polarizing narratives (Allcott et al. 2019). Studies show that populist messages, particularly those blaming elites or minorities, intensify negative public attitudes (Hameleers and Schmuck 2017; Harteveld et al. 2022).

In such an environment, disinformation spreads easily, creating fragmented realities where truth becomes secondary to ideological resonance (Guess et al. 2019). Events like the Capitol riot (2021) and the Brasília attacks (2023) illustrate how populist actors exploit these dynamics to mobilize distrust and undermine democratic institutions. Ultimately, digital populism challenges the foundations of the open society: instead of fostering rational dialog, social platforms propagate division and fuel a “crisis of truth” that threatens democratic resilience.

2.3. The Paradox of Digital Participation

Affective polarization has gained attention with the rise of social media, which plays a key role in shaping political identity. Platforms like Facebook allow users to signal preferences—intentionally or not—through content sharing, reinforcing group belonging and ideological segmentation (Settle 2018). This dynamic contributes to polarization by making political orientations more visible and socially salient.

Empirical studies are consistent with these effects: algorithmic recommendations (Cho et al. 2020) and exposure to hostile comments (Suhay et al. 2017) increase polarization, while deactivating Facebook can reduce it (Allcott et al. 2020). Platform architecture and interaction design are central to these outcomes (Fenoll et al. 2024). Facebook, with its emphasis on personal ties, encourages echo chambers; X (formerly Twitter), due to its rapid and public format, amplifies confrontational exchanges. Visual platforms like Instagram and TikTok, although less overtly political, still foster homogeneity and identity-based segmentation by filtering content through aesthetic and lifestyle cues.

These effects are not uniform: networks that promote emotional involvement and personalization tend to increase polarization, while more diverse interaction environments have a weaker impact. The “engagement-first” business model of social platforms favors divisive content, often isolating users in fragmented digital communities. According to the Pew Research Center (2022), 62% of 18–24-year-olds trust social media more than traditional sources, increasing their exposure to manipulation and polarization.

Understanding the impact of these platforms on young voters is therefore critical. While networks like Instagram and TikTok enhance access to information and facilitate mobilization, they also foster radicalization and cognitive distortion (Allcott et al. 2019). These dynamics raise concerns about the political, emotional, and mental vulnerability of young people in increasingly polarized environments (Anderson et al. 2024).

2.4. The Populist Radical Right and the Enemies of the Open Society

Radical right-wing parties have gained prominence in contemporary politics due to their role in fostering affective polarization (Vanagt et al. 2024). Although these movements operate within democratic systems, they often undermine liberal democratic norms by exploiting the tension between “the people” and “the elites,” advancing exclusionary, nativist agendas that target minorities and challenge pluralism (Moffitt 2018; Pappas 2019). Their discourse frequently includes anti-intellectualism and opposition to cultural elites, deepening ideological divisions.

Across Europe, parties like Rassemblement National, AfD, and Vox, and in Portugal, Chega, represent a response to what Popper called the crisis of the open society (Harteveld et al. 2022; Marchi and Zúquete 2024). Rather than fostering inclusive dialog, these actors weaponize cultural conflict to define a “true society” based on homogeneity and exclusion (Gidron et al. 2023). Their political style emphasizes emotion, fear, and resentment, mobilizing voters through narratives that cast external groups as existential threats (Engesser et al. 2017).

In Portugal, Chega exemplifies this dynamic, rejecting pluralism in favor of nationalist, anti-elite rhetoric. While not directly analyzed by Mudde, Chega aligns with far-right parties elsewhere in Europe through its use of affective polarization and cultural antagonism. This strategy not only fractures public debate but also corrodes democratic cohesion by promoting hate speech and identity-based division (Lapa and Di Fátima 2023).

Ultimately, these movements pose a direct challenge to the open society, advancing a politics that weakens democratic foundations and empowers its ideological adversaries.

2.5. Gender Ideology as a Theme

Contemporary populist dynamics intensify the divide between “ordinary people” and a “corrupt elite,” posing a direct challenge to Popper’s vision of the open society, which depends on pluralism and tolerance (Popper 1945). Social media has amplified these polarizing narratives, enabling the spread of exclusionary discourses that target groups embodying inclusive values, such as human rights advocates, feminists, racial minorities, and LGBTQI communities, often portrayed as threats to a “natural” social order (Mudde and Kaltwasser 2015; Johnston and Wronski 2015).

These dynamics are exacerbated by the structure of digital platforms, which simultaneously enable open debate and facilitate the viral spread of misinformation and polarization. The debate on gender offers a clear example: the opposition between gender as a social construct and, as a biological fact, has become a symbolic battleground between progressive and conservative worldviews. Social media functions as a space where these disputes are amplified, reinforcing either equality or traditional hierarchies (Dietze and Roth 2020).

A central feature of this conflict is the discourse around “gender ideology”—a term employed by conservative populist actors to discredit LGBTQ+ rights and equality policies. On social media, this rhetoric thrives through misinformation campaigns aimed at mobilizing conservative voters against a perceived progressive “agenda.” Framed as a struggle between “true patriots” and liberal elites, this discourse erodes pluralistic values and weakens the foundations of the open society (Dietze and Roth 2020; Mudde and Kaltwasser 2015).

These developments raise a critical question: how can an open society, built on freedom of expression, respond to the digital amplification of narratives that threaten its pluralist core? Balancing open debate with the need to protect democratic values is a fundamental challenge in the age of affective polarization (Iyengar et al. 2012).

2.6. Comparison with Other Studies

Affective polarization and the influence of social networks on political opinion formation are not unique to Portugal. In the United States, studies such as Iyengar et al. (2012) highlight how interpersonal animosity between partisans has deepened, amplified by partisan media that foster ideological echo chambers (Stroud 2008). In Portugal, although polarization is increasing, it is primarily driven by social networks rather than traditional media.

Across Western Europe, the rise of radical right parties, such as Rassemblement National (France), AfD (Germany), and Vox (Spain), has been linked to growing polarization, often through populist discourses that frame politics as a conflict between “the people” and “the elites” (Mudde and Kaltwasser 2015; Harteveld et al. 2022). Portugal follows this trend more recently, with social media playing a key role in amplifying these dynamics.

Brazil offers a contrasting case where private messaging platforms like WhatsApp have had a central role in spreading disinformation and fuelling polarization, especially during the 2018 and 2022 presidential elections (Guess et al. 2019). While X and Instagram dominate political information consumption among Portuguese youth, Brazil’s experience underlines how different platform ecosystems shape distinct patterns of polarization.

3. Hypotheses, Methods, and Results

Although specific hypotheses are formulated, this study is exploratory in nature due to its small, non-probabilistic sample. No generalizations are claimed; rather, the aim is to identify patterns and raise theoretically informed questions for future research. Young first-time voters, often relying on social media as their primary source of political information (Anderson et al. 2024), provide a relevant case for examining these dynamics. Instagram and X offer contrasting platform architectures—one visual and identity-oriented, the other centered on text and public argument—allowing for an analysis of how design logics shape political emotions and polarization (Cinelli et al. 2021). The hypotheses thus function as analytical lenses, not as instruments for confirmatory testing.

3.1. Hypotheses

Based on the literature review developed on the previous pages, the following hypotheses can be formulated:

Hypothesis 1. 

Young people with populist sentiments show higher levels of polarization in social media interactions.

This hypothesis is in line with the open society framework because it reflects how populism can foster polarization and restrict rational debate, creating ideological bubbles that hinder pluralism.

Hypothesis 2. 

Young people who support causes defended by right-wing populist parties have higher levels of affective polarization.

This hypothesis is related to the challenge of an open society to maintain a space for balanced debate, as affective polarization can lead to intolerance and the perception of political opponents as enemies, threatening the principle of pluralism.

Hypothesis 3. 

Users of the X network show higher levels of polarization compared to users who favor the Instagram network.

This hypothesis highlights the role of digital platforms in shaping public opinion, suggesting that certain digital environments can amplify polarization. This reinforces the challenge for the open society because if certain social media favor the fragmentation of discourse, they become obstacles to an informed and diverse debate.

3.2. Methodology and Sample

Given the exploratory nature of this study, the sample was non-probabilistic and obtained through convenience sampling, using mailing lists and dissemination via personal contacts and digital communication networks, including email and social media. While this approach limits the generalizability of the findings, it was considered appropriate for conducting an initial survey shortly after the 2024 legislative elections. The timing of data collection enabled the capture of impressions, attitudes, and patterns of affective polarization during a period of heightened political engagement, offering valuable insights into the role of social media and populist discourse in shaping youth political opinions.

To assess the adequacy of the sample size (n = 130), a post-hoc power analysis was conducted using G*Power software (Version 3.1.9.7). The results indicate that the sample provides sufficient statistical power (≥0.80) to detect medium effect sizes (e.g., d = 0.50; r = 0.30) in bivariate analyses such as chi-square tests and correlation coefficients, using a significance level of α = 0.05. This suggests that the sample is adequately powered for the study’s analytical scope. Nonetheless, the use of convenience sampling introduces potential biases, as participants were not randomly selected but recruited through digital and interpersonal networks. This may limit external validity, as the sample could overrepresent politically engaged or digitally active individuals, potentially inflating levels of polarization or populist sentiment. Future research should prioritize more diverse and representative samples to validate and extend these findings.

Two inclusion criteria were applied: respondents had to be users of at least one social network and eligible to vote for the first time in national elections—in this case, the Portuguese legislative elections held on 10 March 2024. An online questionnaire was administered between 26 February and 5 March 2024, yielding 130 valid responses. The sample showed key characteristics relevant to the study’s aims, such as high media consumption, attentiveness to civic and social issues, and a balanced gender distribution.

Data were analyzed using descriptive statistics and bivariate contingency tables. Inferential analyses were limited to basic measures, such as correlation coefficients and effect size estimates (e.g., Cohen’s d), in line with the exploratory scope of the research. The findings provide a preliminary basis for understanding how different social networks influence polarization and highlight directions for future research with broader and more representative samples.

3.3. Instruments and Operationalization

3.3.1. Demographic Control Variables

Two demographic control variables were included, gender and age, which are also considered to play a role in the political participation process. In terms of age, all the individuals were between 18 and 21 years old. It was found that 43.1 per cent of respondents were male and 56.9 per cent female. No additional demographic variables were collected, as the study’s focus was on attitudes and platform preferences. Future research could integrate socio-economic or educational variables to expand the analytical scope.

3.3.2. Populist Sentiments

To assess the existence of populist sentiments, the questionnaire included some instruments to measure core components of populism. Although more comprehensive instruments have been developed to capture dimensions such as anti-elitism, popular sovereignty, and homogeneity of ‘the people’ (Akkerman et al. 2014), the present measure focuses on the first two axes, adhering to protocols for exploratory studies with sampling constraints (Newman et al. 2019; Schulz et al. 2018): Q1: “I think that most politicians don’t care what people like me think” and Q2: “I think that ordinary people should be consulted whenever important decisions have to be made, namely through popular referendums”. Both measures were intended to capture the central ideas associated with the populist ideology, namely those that reflect the antagonism between the people and the elites, the dissatisfaction with the actions of these same elites, and the importance attributed to the perspective of popular sovereignty. Each question had a 5-point response scale, the first two of which were contrary to statements Q1 and Q2 (“totally disagree” and “partially disagree”), a neutral center point (“neither agree nor disagree”), and two points of agreement (“partially agree” and “totally agree” and “totally agree”). Following the methodology applied by previous studies, these two questions were combined into a single variable with two categories. Individuals who answered that they agreed that most politicians do not care what people think and that ordinary people should be consulted whenever important decisions must be made, namely through popular referendums, were categorized as having populist attitudes; all the rest were categorized as having mainstream attitudes.

The results show the following distribution: 52 individuals (40 per cent) with populist attitudes and 72 individuals (60 per cent) with mainstream attitudes, with a similar distribution in terms of gender (

Table 1. Distribution of feelings by gender.

			Feelings		Total
			Mainstream	Populist
Gender	Female	Frequency	44	30	74
		% Gender	59.5%	40.5%	100.0%
	Male	Frequency	34	22	56
		% Gender	60.7%	39.3%	100.0%
Total		Frequency	78	52	130
		% Total	60.0%	40.0%	100.0%

).

3.3.3. Support for the Radical Right

Based on the above characterization, in this paper, we have considered one of the most prominent, fractious and polarizing electoral debate themes of the Chega party as the object of analysis, which we have adopted in this study as representative of the radical right populist parties (RDP) category, i.e., the argument that “Gender ideology is killing schools and putting our children at risk. We must stop it!”, uttered by the party leader and expanded into a campaign theme and slogan (Figure 1). Respondents were asked to express their agreement with this statement, which was transcribed in full, and the answers were grouped into two categories: adherence or non-adherence to this perspective.

The data obtained show that 56.2 per cent of respondents reject this argument, which was accepted by 43.8 per cent of the individuals studied. If we cross-reference this data with the feelings associated with populism, mentioned above, we see that the respondents who reject the need to combat gender ideology mostly have mainstream feelings (71.2% vs. 28.8%); on the other hand, the majority of young people who consider it necessary to combat gender ideology are classified as having populist feelings (54.4%) (see

Table 2. Gender Ideology by Populism.

			Feelings		Total
			Mainstream	Populist
gender ideology	No	Frequency	52	21	73
		%	71.2%	28.8%	100%
	Yes	Frequency	26	31	57
		%	45.6%	54.4%	100%

).

3.3.4. Affective Polarization

In this study, affective polarization focuses on individuals’ attitudes toward fellow citizens, rather than toward political elites, migrants, or other “outsiders.” Unlike some research that uses sentiment thermometers to gauge feelings toward political parties or leaders (Iyengar et al. 2012), this study examines adolescents’ social distancing attitudes. Affective polarization is known to hinder democratic dialog and compromise by fostering antagonism between community members, especially when political affiliation becomes part of one’s social identity (Iyengar et al. 2024; Turner-Zwinkels et al. 2023). This approach helps to identify whether hostile attitudes toward those with different ideological views are formed as early as adolescence. To measure affective polarization, participants were asked to rate their agreement with the statement: “I think it’s okay to be friends with people who have different political beliefs” on a scale from one to eight (Oden and Porter 2023). While limited, this proxy measure has been used in other exploratory studies (Iyengar et al. 2012), particularly when multi-item scales are impractical. Nonetheless, it does not capture the full spectrum of affective polarization, such as emotional intensity, trait attribution, or perceived social distance. Responses (see

Table 3. Affective Polarization by Gender Ideology.

			Affective Polarization		Total
			No	Yes
gender ideology	No	frequency	60	13	73
		%	82.2%	17.8%	100%
	Yes	frequency	43	14	57
		%	75.4%	24.6%	100%

) were grouped into two categories: scores of one to four were classified as “not polarized,” while scores of five to eight were considered “polarized.” The results showed that 103 respondents (79.2%) were “not polarized,” while 27 respondents (20.8%) were “polarized”.

As a hypothesis to be verified (H2), the fact that levels of affective polarization tend to be higher among young people who accept causes defended by the radical right-wing populist parties was presented. In fact, the data obtained allow us to observe that those who adhere to the cause of the radical populist right are more polarized (24.6% versus 17.8% of those who do not adhere, which means a difference in magnitude of approximately 27.64%); to this extent, these data are consistent with Hypothesis 2.

Considering the data collected and Hypothesis 1, which suggests that levels of polarization tend to be higher among young people who have populist sentiments, a contingency table (

Table 4. Affective polarization by Populist sentiments.

			Feelings		Total
			Mainstream	Populist
Affective polarization	No	frequency	64	39	103
		%	62.1%	37.9%	100.0%
	Yes	Frequency	14	13	27
		%	51.9%	48.1%	100.0%

) was drawn up to compare the percentage of individuals classified as “polarized” in the “mainstream” and “populist” categories. While in the category of non-polarized individuals, the percentage of populists is 37.9, in the polarized category, it is 48.1, i.e., 10.2 percentage points higher, with a difference in magnitude of 26.91%. These data allow us to validate Hypothesis 1.

The findings indicate that adolescents with populist inclinations exhibit greater affective polarization, aligning with theoretical expectations, but warrant a nuanced interpretation. While populist sentiments often amplify in-group and out-group dynamics, as suggested by Mudde (2004), it is equally important to consider the role of negative perceptions directed toward populist right parties themselves. Research, such as that by Harteveld et al. (2022), has demonstrated that these parties are often more intensely disliked than other political groups, evoking strong negative emotions across the political spectrum. This dual dynamic—where populist inclinations foster hostility toward opposing groups while simultaneously attracting animosity from the broader electorate—may intensify polarization among adolescents who align with these ideologies.

To contextualize the results, it is possible that the heightened polarization observed among adolescents with populist sentiments reflects a compounded effect: on one hand, the antagonistic rhetoric inherent to populism emphasizes divisions, and on the other, the social and political stigma associated with populist parties may reinforce feelings of hostility both within and outside these groups. This interplay suggests a bidirectional process, where the polarization attributed to populist inclinations is not only a product of ideological alignment but also of the societal reactions these alignments provoke. Future research could explore this complexity further, potentially disentangling the contributions of internal group dynamics and external perceptions in shaping affective polarization.

3.3.5. Social Network Usage

The relevance of using social networks as a source of information as a factor with an impact on polarization dynamics was described above. However, as shown above, social networks are not homogeneous in terms of their impact on affective polarization. Several factors, related to the platforms themselves and the uses made of them, play a crucial role in how they affect affective polarization. We have seen, however, that networks that emphasize emotional involvement and content personalization tend to amplify these effects, and that network X tends to generate and amplify this polarization, while Instagram, by valuing the visual dimension and being composed of lighter content, generates a less polarizing environment. We therefore consider hypothesis (H3) that polarization levels tend to be higher among users of network X than among those of Instagram.

Respondents were asked which social network they used to obtain information on election-related matters. The following results were obtained (Figure 2):

4. Results and Discussion

This section presents and discusses the findings considering the three hypotheses defined earlier.

4.1. Populist Sentiments and Affective Polarization (H1)

The first hypothesis proposed that young voters with populist sentiments would display higher levels of affective polarization. The data suggest this trend: while 37.9% of non-polarized individuals exhibit populist attitudes, the proportion rises to 48.1% among those classified as polarized—a relative increase of nearly 27%. These findings align with the previous literature highlighting the antagonistic structure of populist discourse and its tendency to reinforce in-group/out-group distinctions (Mudde 2004; Zehnter 2025).

Interestingly, this polarization may not solely derive from internal group dynamics. Research suggests that populist parties and their supporters often elicit strong negative reactions from broader society, intensifying their sense of isolation and opposition (Harteveld et al. 2022). Thus, the polarization observed may reflect a dual process: both the active construction of hostile narratives and the social stigmatization of populist positions, especially among politically engaged youth.

4.2. Support for Radical Right Narratives and Affective Polarization (H2)

The second hypothesis concerned the relationship between support for radical right populist narratives and affective polarization. The data show that 24.6% of those who agreed with the campaign statement opposing so-called “gender ideology”—a hallmark theme of the radical right—were classified as polarized, compared to only 17.8% among those who disagreed. While this difference may seem moderate, it supports the idea that alignment with exclusionary discourses is associated with heightened antagonism toward ideological others.

This result is particularly significant in the context of digital political communication, where identity-based issues like gender, immigration, or nationalism are instrumentalized to generate emotional responses (Engesser et al. 2017; Pappas 2019). The affective structure of such discourses undermines deliberative norms, eroding the foundational principles of the open society as defined by Popper (1945).

4.3. Platform Effects: Comparing Instagram and X (H3)

Finally, the third hypothesis posited that users of X (formerly Twitter) would display higher levels of affective polarization than Instagram users, due to the former’s structure and emphasis on public, emotionally charged debate. The results support this hypothesis: while 20% of Instagram users were classified as polarized, the figure rises to 27.6% among X users—a 38% relative increase. As shown in

Table 5. Polarization Among Social Media Users.

Affective Polarization	Instagram	X
Not polarized	80.0%	72.4%
Polarized	20.0%	27.6%

, 27.6% of X users were classified as polarized, compared to only 20.0% of Instagram users.

This difference is further supported by a moderate correlation between platform choice and polarization (r = 0.45, p < 0.01) and a Cohen’s d of 0.62, indicating a moderate effect size. The confidence interval for the mean difference (0.32 to 0.92) suggests the finding is robust, even within the limitations of the sample. While the percentage-point differences are modest, they nonetheless point toward a possible trend that warrants further exploration in larger and more representative samples.

These results are consistent with the literature on the differential affordances of social media platforms: while visual-based apps like Instagram promote fewer polarizing interactions, text-driven platforms like X tend to amplify confrontation and ideological signaling (Cinelli et al. 2021; Cho et al. 2020).

5. Limitations

This study is exploratory in nature and based on a non-probabilistic convenience sample of 130 Portuguese first-time voters. As such, the findings cannot be generalized to the broader population of young voters in Portugal or elsewhere. The limited sample size, combined with its specific demographic and contextual focus—voters aged 18 to 21 who participated in the 2024 legislative elections—means that the conclusions drawn should be interpreted as indicative rather than definitive.

Furthermore, the demographic data collected was restricted to age and gender, limiting the capacity to account for other potentially relevant variables such as socioeconomic status, educational attainment, regional background, or political engagement history. These factors could significantly shape political attitudes and susceptibility to affective polarization, and their exclusion represents a constraint on the analytical depth of the study. The modest percentage differences observed (e.g., 7 points in H2 and H3) should not be read as strong validation but as preliminary indications, whose substantive significance remains open to further empirical testing. At the same time, the operationalization of populism via only two items further constrains the analysis, as it limits the multidimensional capture of the construct. Future studies should employ validated scales (e.g., (Akkerman et al. 2014) for populism; Iyengar et al. (2024) sentiment thermometer for polarization).

In terms of methodological scope, no multivariate or inferential statistical analyses were conducted beyond basic correlations and effect size estimates. This decision reflects both the exploratory purpose of the research and the constraints imposed by the available dataset. Although simple inferential tests (e.g., t-tests) could be conducted, the decision was made to privilege descriptive patterns due to the exploratory scope and sampling constraints. Consequently, differences between X and Instagram are interpreted as preliminary and indicative, rather than definitive. Factors such as ideological intensity, personality traits, or offline political engagement—none of which were controlled for—may also correlate with affective polarization and media use patterns. As a result, potential confounding effects and interactions between variables remain unexamined, which may limit the explanatory power of the observed patterns.

It is also important to note that the questionnaire relied on self-reported data regarding political preferences, social media use, and emotional attitudes. Such data are inherently subject to social desirability bias, memory inaccuracies, and individual interpretation of survey items, particularly when assessing affective responses or ideological alignment.

Despite these limitations, the study provides meaningful initial insights into the relationship between populist sentiment, affective polarization, and social media use among young voters. It highlights emerging trends and raises questions that warrant further investigation with more robust and representative methodologies. Future research should aim to incorporate larger and more diverse samples, longitudinal designs, and more detailed measures to better capture the complexity of political behavior in the digital age.

6. Conclusions

The study indicates that affective polarization among young voters is shaped by both populist sentiments and the social media platforms they use. The findings illustrate the tension inherent in Open Society 2.0: platforms like X, by prioritizing emotional engagement, invert the Popperian logic of rational deliberation, replacing it with algorithmically mediated identity-based antagonism (Fuchs 2022). While the non-representative sample prevents generalizations, consistent patterns emerge: individuals with populist attitudes tend to exhibit higher levels of polarization, and users of X (formerly Twitter) display more polarized views than those who prefer Instagram. These findings align with previous research suggesting that algorithmic curation and platform design reinforce ideological segmentation and weaken deliberative engagement (Cinelli et al. 2021).

Despite its exploratory nature, the study contributes to ongoing debates about digital populism, youth political behavior, and the erosion of the open society. By focusing on first-time voters—a group particularly shaped by digital environments—it highlights how platform architecture and affective dynamics interact to deepen political antagonism. In line with Popper’s (1945) view, the weakening of mechanisms for critical correction and the intensification of group-based hostility pose significant risks to democratic resilience.

The international context reinforces these concerns. While polarization trends vary, Western democracies show a broader pattern of increased hostility between political camps, often linked to the rise of radical-right movements and declining trust in institutions (Engesser et al. 2017; Grande and Saldivia Gonzatti 2025). In Portugal, the growing influence of the Chega Party among young voters exemplifies how digital populism and dissatisfaction with the political system converge, fostering emotional segmentation and undermining pluralism.

Ultimately, the findings underline the need to rethink strategies that support critical engagement in the digital public sphere. Affective polarization and digital populism challenge the foundations of the open society by prioritizing emotional identification over rational debate. As Popper emphasized, democracy requires institutional and civic commitment to continuous scrutiny and correction. Addressing these threats involves promoting media literacy, regulating algorithmic amplification, and restoring common spaces for deliberation grounded in mutual respect and factual reasoning.

Platforms such as X, by privileging emotional engagement and rapid interaction, undermine the conditions for rational deliberation envisioned by Popper. In the framework of Open Society 2.0, this platform logic amplifies fragmentation and identity-based antagonism, posing structural obstacles to pluralistic debate. Considering these findings, two practical responses emerge: first, the need for regulatory interventions that promote algorithmic transparency and reduce the creation of echo chambers; second, the importance of investing in media literacy programs that help young citizens critically navigate the affective and ideological pressures of digital environments. These initiatives are essential to restoring democratic resilience in a context increasingly shaped by polarization and digital populism.