Feelings among friends and foes in populist communication: The Finns Party leadership’s emotional appeals on social media during the 2023 Finnish parliamentary election

Populism as a discursive framework

In recent years, scholars have increasingly examined the rhetorical, emotional, and performative dimensions of populist communication (Moffitt, 2016). At the same time, populism remains a contested concept, characterised by a wide range of definitions and analytical approaches, reflecting its context-dependent manifestations across political systems, cultures, and ideological traditions (for an overview, see Herkman, 2022: 11–36). While this plurality has enriched the field, it has also contributed to fragmentation in how populism is conceptualised. Rather than attempting to reconcile these competing perspectives, this study adopts a discursive framework approach, conceptualising populism primarily as a mode of political communication (see Järviniemi, 2025; Reinemann et al., 2017). Within this framework, populist communication is characterised by an antagonistic “us versus them” narrative that constructs boundaries between the virtuous and homogeneous “people”, corrupt “elites”, and perceived “dangerous outgroups” (often referred to as “dangerous others” in the populism literature; Reinemann et al., 2017). From this perspective, populist communication is not confined to a particular ideology and can be adopted by political actors across the ideological spectrum (see Jensen et al., 2026; Järviniemi, 2025). The analytical strength of this framework lies in its flexibility, as “the people”, “elites”, and “dangerous outgroups” vary across political, ideological, and national contexts (de Vreese et al., 2018; Reinemann et al., 2017).

At the core of the populist discursive framework lies people-centrism, in which “the people” are constructed as the rightful locus of political authority (Järviniemi, 2025). However, this notion of “the people” is highly flexible and can be adapted to different ideological projects and contextual realities (de Vreese et al., 2018). Anti-elitist populism combines people-centrism with anti-elite discourse, portraying elites as betraying the “will of the people” (de Vreese et al., 2018; Jagers & Walgrave, 2007). Depending on the context, “elites” can be political, economic, cultural, intellectual, or judicial (de Vreese et al., 2018; Jagers & Walgrave, 2007; Mudde, 2017). Exclusionary populism constructs “dangerous outgroups” that are portrayed as threats to the collective identity of “the people” (Järviniemi, 2025). The identity of these outgroups is similarly fluid and shaped by ideological orientation and political context. In populist radical right communication, immigrants, ethnic minorities, groups portrayed as exploiting the welfare system, and individuals or groups with progressive political agendas (e.g., left-wing activists or advocates of multiculturalism) are frequently constructed as “dangerous outgroups” and portrayed as threats to cultural integrity, social cohesion, or economic stability (Herkman, 2022; Järviniemi, 2025; Reinemann et al., 2017).

Emotional appeals in populist communication

Research in political psychology and political communication has demonstrated that emotions play a fundamental role in shaping how citizens interpret political information, evaluate political actors, and engage in political behaviour (Brader, 2005; Marcus et al., 2000). Much of the literature on emotions in politics has traditionally relied on the two-dimensional valence model of emotions, which conceptualises positive and negative emotions as two independent and orthogonal dimensions (Marcus et al., 2000). However, more recent research has challenged this valence-based approach for overlooking important differences between emotions that share the same positive or negative valence, emphasising that discrete emotions of the same valence can trigger distinct cognitive appraisals and behavioural responses (see, e.g., Rico et al., 2017; Wagner, 2014).

We follow Brader’s (2005: 390) definition of emotional appeals as “communications intended to elicit an emotional response from some or all who receive them”. Building on this, a growing body of research recognises the central role of emotions in populist communication (see, e.g., Bliuc et al., 2026; Caiani & Di Cocco, 2023; Cartes-Barroso et al., 2025; Domínguez-García et al., 2025; Hameleers et al., 2017; Hloušek et al., 2024; Rico et al., 2017; Salmela & von Scheve, 2017; Tietjen, 2023; Widmann, 2021; Wirz, 2018). Populist actors consistently rely on emotional appeals as a central feature of their speeches, language, and overall presentation (Moffitt, 2016). Given the relational logic of the antagonistic populist discursive framework, emotional appeals are central both to the construction and consolidation of collective identities around the ingroup (consisting of “the people” and populist actors) and to the exclusion of perceived outgroups (consisting of “elites” and “dangerous outgroups”). Through these appeals, populist actors mobilise support, foster collective identities through shared grievances, and position themselves as authentic representatives of “the people” while attributing responsibility for societal grievances to “elites” and “dangerous outgroups” (Hameleers et al., 2017).

Turning to specific emotional appeals, research has particularly highlighted the role of anger in populist communication (e.g., Caiani & Di Cocco, 2023; Rico et al., 2017; Salmela & von Scheve, 2017; Widmann, 2021). Anger can be understood as an actor-oriented emotion, arising when individuals attribute adverse events to the deliberate actions of external actors rather than to their own behaviour or to uncontrollable circumstances (Wagner, 2014; Wirz, 2018). Anger has been conceptualised as a reaction to perceived vulnerabilities among citizens, often channelled in populist communication against corrupt “elites” and stigmatised “outgroups” (Caiani & Di Cocco, 2023; Rico et al., 2017; Salmela & von Scheve, 2017). However, through appeals to anger, populist actors may also engage with narratives that emphasise shame, moral wrongdoing, and social injustice (Caiani & Di Cocco, 2023; Salmela & von Scheve, 2017). In particular, within populist communication, narratives of humiliation and status loss may transform feelings of shame into anger toward those perceived to be responsible (Salmela & von Scheve, 2017).

Fear represents another central emotional appeal in populist communication (Bliuc et al., 2026). Fear is likely to emerge when individuals face threatening circumstances that lack a clearly identifiable source of responsibility, where individuals feel little or no control over the outcome (Wagner, 2014). Feelings of fear and insecurity have been closely tied to perceptions of economic instability, social disintegration, relative deprivation, and a sense of powerlessness (Salmela & von Scheve, 2017). According to Salmela and von Scheve (2017), the perceived “losers” of globalisation are particularly susceptible to fear and feelings of insecurity, because of perceived declining status, identity loss, and the erosion of traditional values. Moreover, fear has been shown to decrease partisan habits and increase individuals’ receptiveness to new information and persuasion (Scheller, 2019). Consequently, appeals to fear may enable populist actors to attract new supporters (Widmann, 2021).

Beyond anger and fear, other negative emotional appeals also reinforce the antagonistic logic of populist communication. For instance, contempt functions as an exclusionary emotion, characterised by a short-term derogation and long-term negative appraisal of another person’s character with the goal of socially excluding this other person (Fischer & Roseman, 2007). In populist communication, contempt is appealed to by signalling the perceived moral or intellectual inferiority of political opponents, for instance, as expressed through sarcasm, ridicule, irony, mockery, or stereotyping. Furthermore, appeals to sadness also feature in populist communication (Widmann, 2021). Sadness is an emotion connected to disappointment, failure, loss, and misfortune (Small & Lerner, 2008). Sad individuals have been shown to engage in more systematic processing of political information, while reducing reliance on prior biases (Small & Lerner, 2008). Consequently, populist actors may employ appeals to sadness to resonate emotionally with disillusioned citizens, cultivate a sense of shared grievance with “the people”, and potentially attract new supporters (Widmann, 2021).

At the same time, comparatively less empirical attention has been devoted to positive emotional appeals in populist communication, particularly within studies of populist radical right communication. Whereas left-wing populism has long drawn on positive emotional appeals to promote grassroots solidarity and engagement, emerging research indicates that populist radical right actors also use positive emotional appeals to mobilise support (see, e.g., Caiani & Di Cocco, 2023). However, positive emotional appeals in populist communication are not inherently inclusive. By strengthening attachment to the ingroup, they may simultaneously reinforce moral boundaries between ingroups and outgroups, thereby contributing to exclusionary forms of collective identity construction (Salmela & von Scheve, 2017).

Turning to positive emotional appeals, emotions such as enthusiasm, joy, love, pride, hope, and nostalgia can reinforce collective identities, foster a sense of belonging and unity of “the people”, and strengthen affective attachment to the nation and charismatic populist leaders (Albertazzi & Bonansinga, 2024; Bonansinga, 2025; Tietjen, 2023). Through such appeals, populist actors can simultaneously portray themselves as credible agents capable of restoring the “will of the people” and as protectors of this community. Enthusiasm is a prospective emotion that arises in response to cues that signal goal-congruent and desirable outcomes and reinforces commitment and motivation (Brader, 2005). Joy is a retrospective emotion, experienced when past developments are evaluated as consistent with one’s goals (Widmann, 2021). Love can be conceptualised as an enduring and affectively rich emotion that fosters sustained attachments to valued objects and entails a concern for the well-being and flourishing of that object (Tietjen, 2023). Similar to joy, pride is elicited by previous success and reflects an enhancement of one’s identity through attribution of credit to valued achievements, either one’s own or those of a group with which one identifies (Lazarus, 2001).

Hope represents a prospective and mobilising emotion (Wirz, 2018), elicited when individuals evaluate a future outcome as compatible with their personal goals, meaningful to them, realistically achievable yet uncertain, and expected to bring about an improved future (Chadwick, 2015). By connecting present grievances to expectations of a better future, hope thus serves as a powerful emotional appeal in electoral campaigning. Consequently, in populist communication, hope can connect dissatisfaction with the present to expectations of political renewal by portraying populist actors as capable of restoring the “will of the people” and correcting the failures of the political establishment. Nostalgia, commonly expressed as yearning for lost or idealised “good old days”, is another central emotional appeal in populist communication (Salmela & von Scheve, 2017; Smeekes et al., 2021). It is generally conceptualised as an ambivalent, though predominantly positive, emotion characterised by a sentimental longing for a personally meaningful and warmly remembered past (Sedikides et al., 2008). In populist communication, nostalgia is often used to evoke idealised visions of the past and historical narratives that glorify traditional, ethnically and culturally homogeneous communities and celebrate their past achievements (Salmela & von Scheve, 2017).

Moreover, populist communication is often associated with humorous rhetoric, including satire, sarcasm, irony, mockery, and ridicule, which can make political messages more accessible and engaging (Sakki & Martikainen, 2021). However, humour can serve both inclusive and exclusionary functions. Analysing the Finns Party’s 2019 election campaign video, Sakki and Martikainen (2021) showed how humorous ridicule of political opponents can simultaneously evoke enjoyment from the perceived humiliation, failure, or discomfort of political adversaries (i.e., schadenfreude) and reinforce feelings of contempt and superiority toward political opponents. Humour may thus frame derogatory or norm-violating content as acceptable, thereby masking or legitimising hostility toward political opponents or outgroups (Sakki & Martikainen, 2021). Accordingly, we distinguish between different uses of humour based on the emotional responses they seek to elicit. Humour that primarily entertains, engages, or fosters relatability is coded as an appeal to amusement, whereas derisive humour based on irony, ridicule, sarcasm, or mockery is coded according to the negative emotional response it seeks to evoke, such as contempt or anger (Sakki & Martikainen, 2021).

Populist communication and social media platforms

The affordances of social media have often been argued to align with the people-centrist dimension of populist communication, allowing political actors to cultivate direct, personalised, and seemingly unmediated connections with “the people” (Ernst et al., 2017). Consequently, a growing body of research suggests that social media platforms have facilitated the spread of populist communication by enabling politicians to bypass traditional journalistic gatekeepers and communicate directly with citizens (Ernst et al., 2017). At the same time, candidate-constituent interaction varies across platforms, reflecting differences in their technological affordances, user demographics, and broader sociocultural norms (see Nelimarkka et al., 2020). These differences in social media platforms provide political actors with opportunities for strategic self-presentation (Magin et al., 2024).

This is reflected in studies of populist communication on different social media platforms. For instance, research has shown that populist communication is more widespread on Facebook compared with other social media platforms (Ernst et al., 2017; Magin et al., 2024). This suggests that populist actors may perceive Facebook as a useful venue for populist communication, given its extensive reach, opportunities for direct interaction with citizens, and capacity for longer-form messages (Ernst et al., 2017). By contrast, X (Twitter) has more frequently been associated with elite discourse, and research suggests that right-wing populist actors have used the platform as a space for confrontation with political, journalistic, and institutional elites rather than for direct voter engagement (Schwartz et al., 2023).

Turning to Instagram and TikTok, recent research suggests that these platforms enable populist actors to cultivate a more personal, playful, and sympathetic image through self-personalisation and lifestyle-oriented content (Albertazzi & Bonansinga, 2024; Bonansinga, 2025; Schwartz et al., 2023). For instance, Bonansinga (2025) has shown how Marine Le Pen uses Instagram for “dedemonisation”, sharing curated glimpses of her private life to cultivate a more relatable and ordinary public image. At the same time, TikTok is underpinned by an engagement-driven algorithm that prioritises high-arousal content and has been linked to the amplification of far-right radicalisation (Shin & Jitkajornwanich, 2024). Moreover, emotionally intense content tends to achieve greater visibility on the platform, and recent studies find both positive and negative emotional appeals to be prominent in populist radical right communication on TikTok (Cartes-Barroso et al., 2025; Domínguez-García et al., 2025).

Against this backdrop, while previous research suggests that social media platforms may be used in distinct ways and that populist communication varies across them, less is known about how different emotional appeals are distributed across these platforms. Differences in emotional appeals across social media platforms may reflect a range of factors, including individual actor preferences and audience compositions.

The Finnish context: The political system and the Finns Party

Finnish politics has traditionally emphasised consensus and cooperation, with broad coalitions and relatively low levels of political conflict (von Schoultz & Strandberg, 2024). At the same time, Finland’s candidate-centred electoral system, where voters select individual candidates within multimember districts, encourages personalised campaigning and competition among candidates of the same party (von Schoultz & Strandberg, 2024). As a result, individual visibility and personalised communication play a central role in electoral competition (Van Aelst et al., 2012). Thus, Finland’s electoral environment provides a well-suited context for studying social media communication in electoral campaigns, a dynamic further reinforced by the country’s high Internet usage and candidates’ early adoption of digital platforms (Carlson & Strandberg, 2024; Newman et al., 2024). The Finns Party constitutes a particularly relevant case in this context, having positioned itself as a digital challenger to established parties by actively and innovatively using digital media platforms early on to bypass traditional media gatekeepers and mobilise supporters directly (Hatakka, 2018).

Founded as a challenger to the established party system in Finland, the Finns Party’s electoral breakthrough came in the 2011 parliamentary election, when it secured 19 per cent of the vote. In 2015, the party received nearly 18 per cent and entered government as the second-largest party in a centre–right coalition led by the Centre Party. Government participation proved to be a turning point, and in 2017, a leadership change brought Jussi Halla-aho to power, representing the party’s more radical right-wing faction. This shift triggered an internal split, as supporters of former leader Timo Soini formed Blue Reform, which remained in government until shortly before the 2019 election. Under Halla-aho, the Finns Party consolidated a more explicitly radical right, nativist, and fiscally conservative economic political agenda. In the 2019 election, the party received 17 per cent of the vote and again became the second-largest party in Finland. In 2021, Riikka Purra was elected party leader, under whose leadership the party further strengthened its radical right profile. This ideological consolidation has brought the Finns Party into closer alignment with other populist radical right parties across Europe, sharing core features such as nativism, authoritarianism, and anti-establishment populist communication (see Arter, 2025). In the 2023 election, the Finns Party increased its vote share to 20 per cent and joined a right-wing coalition government led by the National Coalition Party, again as the second-largest party.

Prior to the 2023 election, Finland was governed by a “red–green” coalition including the Social Democratic Party, the Centre Party, the Green League, the Left Alliance, and the Swedish People’s Party. The government was led by Prime Minister Sanna Marin and was notable for being headed by female leaders from all coalition parties. Traditionally, Finnish parties campaign without ruling out potential coalition partners. The 2023 election marked a departure from this norm (Kestilä-Kekkonen et al., 2024). In January 2023, the National Coalition Party announced its willingness to cooperate with any party, including the Finns Party. In response, the Greens, Social Democrats, and Left Alliance ruled out collaboration with the Finns Party, thereby contributing to a campaign environment in which competition increasingly resembled a contest between a right-leaning alternative centred on the National Coalition Party and the Finns Party, and a red–green coalition led by the Social Democrats (Wass et al., 2024). Public finances, national debt, healthcare reform, national security, and labour-based immigration were among the central issues of the 2023 election campaign (Kestilä-Kekkonen et al., 2024; Ylisalo, 2024). In 2023, the Finns Party focused on immigration, a longstanding issue for the party, linking it to debates on gang crime and school segregation, while simultaneously criticising the government’s handling of public finances and national debt (Kestilä-Kekkonen et al., 2024).

Data

Given Finland’s candidate-centred electoral system, the empirical analysis focuses on individual-level campaigning, where we examine the social media communication published on the public personal social media accounts of six members of the Finns Party leadership on X (Twitter), TikTok, Facebook, and Instagram. The selected politicians occupied key leadership positions within the party during or immediately following the study period, including party leader Riikka Purra; former party leader Jussi Halla-aho; vice party leaders Sebastian Tynkkynen and Mauri Peltokangas; and parliamentary group leaders Ville Tavio (until the 2023 parliamentary election) and Jani Mäkelä (after the 2023 parliamentary election).

The data were collected manually in April 2025 and consist of content published on these accounts during the 2023 Finnish parliamentary election campaign. The data collection focused on a three-week period leading up to election day, representing a high-attention phase of the electoral cycle, marked by an escalation in political messaging, increased media coverage, and intensified voter engagement. Accordingly, the data collection period began on 12 March 2023, included the advance voting period (22–28 March), and concluded on election day (2 April 2023).

The distribution of social media posts varied considerably across the selected politicians. Jani Mäkelä accounted for the largest share of posts (44%, n = 231) but generated comparatively low engagement. By contrast, Riikka Purra (23%, n = 122) and Sebastian Tynkkynen (17%, n = 91) attracted the highest levels of engagement. Ville Tavio (10%, n = 54), Mauri Peltokangas (3%, n = 18), and Jussi Halla-aho (2%, n = 8) contributed smaller shares of posting activity and received smaller shares of engagement. These differences should be considered when interpreting actor-level patterns. Posting activity also varied across platforms. X (Twitter) accounted for the majority of posts (60%, n = 317), followed by Facebook (26%, n = 135), Instagram (9%, n = 46), and TikTok (5%, n = 26). Consequently, platform-level differences should be interpreted with caution. The final dataset comprised 524 social media posts. A detailed overview of posting activity and engagement across actors and platforms is provided in the Supplementary Material file (available to download in the “Additional files”; see Appendix B, Figure B1 and Table B1).

Methodological approach

The analysis is based on a human-coded quantitative content analysis (for a full overview of the codebook and coding scheme, see Appendix A in the Supplementary Material). Manual content analysis is considered to provide greater interpretive validity than automated approaches by allowing coders to account for contextual, cultural, and linguistic nuances in communication (Aslanidis, 2018). This is particularly important when analysing emotional appeals. While automated and dictionary-based approaches have been employed in studies of populist communication (e.g., Caiani & Di Cocco, 2023; Widmann, 2021), they may be less well suited to identifying emotional appeals that rely on context, cultural references, or indirect forms of expression, such as contempt conveyed through irony, sarcasm, mockery, or ridicule.

In this study, we primarily focus on the linguistic dimension of populist communication, prioritising the textual and verbal content of social media communication. This approach reflects the understanding of populism as a discursive framework in which the moralised division between “the people”, “elites”, and “dangerous outgroups” is constructed (Reinemann et al., 2017). At the same time, we acknowledge that the audiovisual context in which linguistic elements appear shapes message interpretation (see Bonansinga, 2025), and we therefore treat audiovisual elements as a contextual layer to the discursive content. In the case of audiovisual content, all videos were viewed in full, and spoken content was transcribed during coding. Accordingly, linguistic elements, including post captions, spoken language, and on-screen text, served as the primary basis for coding emotional appeals. Audiovisual elements such as images, music, editing style, and visual symbolism were used as contextual cues that could inform interpretation when the textual content alone was insufficient to determine the emotional appeal or its target. This was particularly relevant on TikTok and Instagram, where audiovisual elements frequently complemented or reinforced the communicative message, but it was also considered when relevant for content published on X (Twitter) and Facebook.

Given the study’s focus on the linguistic dimension of communication and its inclusion of four social media platforms with distinct communicative formats, particular attention was devoted to the unit of analysis. Treating the entire social media post as the unit of analysis risks obscuring within-post variation, as social media posts often contain multiple argumentative statements that differ in both their targets and emotional appeals. This understanding also reflects how social media content is received, as different components of the same social media post may evoke distinct emotional responses among audiences (see Lindholm et al., 2026). A standardised unit of analysis is therefore necessary both to ensure comparability across social media platforms and to capture variation in emotional appeals within individual social media posts.

Research on populism has employed a variety of approaches to defining, analysing, and segmenting populist discourse (see Aslanidis, 2018). Building on this work, we adopt a logic of segmentation by dividing each social media post into Core Argumentative Units, defined as distinct argumentative expressions encompassing claims, ideas, or statements. The segmentation follows the principle that each unit should represent one internally coherent argumentative expression. Accordingly, sentences were grouped into the same unit when they conveyed a common argument, conceptual meaning, and emotional orientation. Segmentation was conducted collaboratively by both authors, with ambiguities resolved through discussion and consensus until a shared understanding of the segmentation principles was established. Ambiguities occurred most frequently in video-based content, particularly on TikTok, where longer videos and layered audiovisual elements occasionally made the coherence of arguments less apparent. Accordingly, the length of individual units of analysis varied according to the coherence of the argument being expressed. Consequently, longer and more argumentatively complex posts generated a larger number of units. This is considered a substantive feature of the data rather than a methodological artefact, as different components of the same message may contain distinct emotional appeals and targets. Applying this segmentation procedure resulted in a final dataset comprising 1,259 units of analysis derived from 524 social media posts.

Prior to coding, both authors completed a training phase using the codebook, during which coding instructions were reviewed, examples were discussed, and ambiguities were resolved through consensus. Intercoder reliability was subsequently assessed using Krippendorff’s alpha (Krippendorff, 2019). The initial reliability assessment yielded coefficients below the threshold for robust agreement, prompting refinement of the codebook and additional coder training. Reliability was then reassessed using a stratified subsample comprising approximately 10 per cent of the dataset (126 units of analysis derived from 42 social media posts), proportionally distributed across platforms and political actors. The second reliability assessment produced coefficients of α = .83 for emotional sentiment, α = .69 for emotional appeal, and α = .82 for target. Following Krippendorff (2019), coefficients above .80 indicate good reliability, while values above .667 are considered acceptable for drawing tentative conclusions. Although reliability was lower for the emotional appeal variable than for emotional sentiment and target, the coefficient remained above the accepted minimum threshold for drawing tentative conclusions (Krippendorff, 2019). The lower level of agreement reflects the challenge of distinguishing between conceptually related emotional appeals within the same emotional valence, rather than disagreement about whether a message contained an emotional appeal or whether that appeal was positive or negative. This is supported by the higher reliability obtained for emotional sentiment, indicating strong agreement regarding the overall emotional orientation of the coded units.

Coded variables

In the analysis, we use three coded variables: emotional sentiment, emotional appeals, and target (for an overview of the coding scheme and examples, see the Supplementary Material, Appendix A, Tables A1 and A2, and Coding scheme for targets).

Emotional sentiment is operationalised by coding each unit of analysis as neutral (absence of emotional sentiment), negative (the overall sentiment of the unit is negative), or positive (the overall sentiment of the unit is positive). This serves as a guiding variable: If a unit is coded as positive or negative, it is subsequently classified according to the second variable, emotional appeal.

Emotional appeals are operationalised based on prior research on populist communication, identifying several emotional appeals relevant to populist communication (see, e.g., Caiani & Di Cocco, 2023; Cartes-Barroso et al., 2025; Domínguez-García et al., 2025; Hameleers et al., 2017; Hloušek et al., 2024; Rico et al., 2017; Salmela & von Scheve, 2017; Widmann, 2021; Wirz, 2018). Negative emotional appeals include anger, contempt, fear and insecurity, and sadness, whereas positive emotional appeals encompass enthusiasm, joy, love, or pride, hope, nostalgia, and amusement. While enthusiasm, joy, love, and pride are distinct emotional appeals, they were combined into a single category because they share a common positive orientation toward desirable outcomes, achievements, and valued individuals, groups, or ideals, and because of the difficulty of distinguishing between them with sufficient intercoder reliability.

Target identifies the object toward which the actor seeks to direct the audience’s emotional response. Accordingly, the target refers to the specific individual, group, collective, actor, or institution to which the emotional appeal is directed, rather than the audience exposed to the message. Given that a single emotional appeal may be directed towards multiple targets, the coding scheme permitted multiple target classifications within the same unit of analysis, provided that these did not simultaneously include both ingroup and outgroup targets. The target is operationalised in line with previous research on populist communication (de Vreese et al., 2018; Jagers & Walgrave, 2007).

Ingroup targets consist of two categories. First, the people capture references to a collectively constructed ingroup presented as the homogeneous, legitimate, virtuous, or authentic people. Following Cranmer’s (2011) definition of people-centrism, this includes messages in which the political actor explicitly claims to defend the will of the people (advocacy), asserts accountability to them (accountability), legitimises their position by invoking the people (legitimacy), or refers directly or indirectly to a collective constructed as the ingroup (e.g., “the people”, “Finns”, “citizens”). Second, actors capture references to the party, its members, ideologically aligned actors, or the political actor themselves when these are presented as part of a shared ingroup. Such references often employ collective identifiers (e.g., “we”, “us”, or “our”) or self-referential language to construct a shared collective identity and to position political actors as authentic representatives and defenders of “the people”.

In contrast, outgroup targets include different elites or dangerous outgroups. Anti-elitism can target a broad range of actors and institutions, including political actors and parties (anti-politics), public institutions and bureaucracy (anti-state), international institutions (anti-international), journalists and media organisations (anti-media), research and cultural institutions or actors (anti-intellectual and anti-cultural), and economic actors and institutions such as corporations, unions, banks, or industries (anti-economic). These elites are depicted as self-serving and as opposing the interests of the ingroup (de Vreese et al., 2018; Jagers & Walgrave, 2007; Mudde, 2017). Dangerous outgroups include groups such as immigrants, ethnic minorities, groups portrayed as exploiting the welfare system, and individuals or groups with progressive political agendas (e.g., left-wing activists). These groups are typically constructed as threats to national identity, security, or economic well-being and are portrayed as undermining the unity and purity of the ingroup (de Vreese et al., 2018).

Finally, the category other captures emotional appeals directed toward identifiable targets that cannot be classified as either ingroup or outgroup actors within the populist antagonistic framework.