Media criticism as a propaganda strategy in political communication

In contemporary societies, journalism’s ability to scrutinise political power is an essential feature of pluralist democracy. At the same time, media criticism is an important form of political discourse through which journalists and journalism institutions can be held accountable when publishing false information or when they violate ethical standards in their reporting. However, attacks on the news media for political purposes have also become increasingly common over the last decade. Spearheaded by right-wing populist leaders, such as Donald Trump, Jair Bolsonaro, and Viktor Orban, the institutional legitimacy of legacy media has not only been questioned, but also increasingly undermined by political power (Carlson et al., 2021; Parks, 2022). The use of digital media technologies by actors seeking to influence public attitudes toward legacy media raises new questions related to what could be described as a weaponisation of media criticism, through which actors deliberately establish public mistrust in the media for their own political gain.

In order to understand how political actors use media criticism for political purposes, we argue that propaganda studies can vitalise more conventional approaches to political communication (Robinson, 2019). Historically, the concept of propaganda was widely used in the understanding of political communication (Bernays, 1928/2005; Lippmann, 1922), but it has since been separated from the latter, mainly due to its historically negative connotations. Mainstream approaches have primarily had an (instrumental) interest in the effects of political communication, thus reflecting a problem-solving approach, addressing the classic relation of political power and legacy media, and the latter’s role as gatekeeper for political discourse (Robinson, 2019). In a recent authoritative volume on political communication, The Oxford Handbook of Political Communication (Kenski & Jamieson, 2017), the term propaganda is almost absent, and when addressed, it is mainly used as a reference to historical (i.e., pre–World War II) examples of political persuasion. Even in more critical approaches to political communication, for example, Herman and Chomsky’s (2010) “propaganda model”, the focus has been on how large and powerful media corporations’ direct public discourse on political and economic issues in the interest of state and capital. The classical work on propaganda by Ellul (1965) also stressed the mass media’s central position for propaganda. Other critical perspectives have focused on the state as a “propagandist” actor, utilising strategic communication for explicit political purposes, for example, in times of war and conflict (Davis et al., 2020; Snow & Taylor, 2006). However, rather than focusing on the strategic communication of the state, or on legacy media’s ability to shape the mindset and behaviour of the public, this study focuses on media criticism expressed by elected politicians. We take our point of departure in the contention that online communication of individual political actors could be analysed through a revitalised understanding of propaganda in political communication (Benkler et al., 2018).

The purpose of this article is to identify various communication strategies that politicians use to criticise the news media, and thereby reveal how and to which extent they seek to undermine the legitimacy and institutional role of journalism. Moreover, we want to illuminate how the use and digital circulation of these strategies vary between party types. The context is Sweden, where a comparatively strong public trust in legacy media has now started to decrease, particularly among people with critical attitudes towards Swedish immigration policies (Andersson & Weibull, 2017), and where media criticism is a common feature of the political discourse online (Sehl et al., 2022).

In order to explore media criticism as a propaganda strategy, we assessed the Twitter flows of all members of parliament with an active account four weeks prior to the 2018 national election. Drawing on a content analysis of all media-related tweets identified in the sample (N = 2,330), we address the following research questions:

In the subsequent theory section, we first discuss the relationship between political communication and propaganda, then we move on to a discussion of how media criticism can be used as a propaganda strategy by political actors seeking to undermine the legitimacy of legacy media and journalism. We end with a description of the Swedish media system and a review of previous research on election news and media bias in the Swedish context.

The digital transformation towards more open and direct forms of interaction between political actors and the public has challenged the role of legacy media’s centrality for political discourse. Political actors can now steer and interact with public information flows in new and strategic ways. This opens up for revitalised approaches to political communication, where propaganda studies have much to offer. We understand propaganda as strategic communication for political purposes that uses non-consensual persuasion strategies to influence public opinions and attitudes (Robinson, 2019). Propaganda can utilise strategies such as disinformation, lies, distortion, distraction, “fake news”, omission, and misdirection, but also affective manipulation or framing of accurate claims in order to persuade or obtain behaviour outcomes of particular groups in society or the general public (Benkler et al., 2018). Propaganda can also be understood as a weaponisation of information by powerful actors (Miller & Sabir, 2012). One such strategy is disinformation, for example, spreading “explicitly false or misleading information” for political purposes (Benkler et al., 2018: 32). Unlike misinformation, which lacks the intent to deceive and manipulate (Benkler et al., 2018), disinformation manipulates the target group to believe or act in a certain way. Manipulation can come in various forms, for example, by deliberately blending true and false information in statements, or by using strong affective exaggerations on specific matters. The goal is ultimately to mislead segments of the public to achieve political ends. Disorientation is another key feature of propaganda, where the target population “loses the ability to tell truth from falsehood or where to go for help in distinguishing between the two” (Benkler et al., 2018: 24). This has become particularly evident within a digital ecology saturated by information and multiple sources. Propaganda strategies have become more complex with the rise of information networks. Since propaganda circulates between actors in online networks, supporters can add credibility and “recall” the manipulative content to the target population within a network (Benkler et al., 2018: 33). This also lowers the cost, or risk, for actors who disseminate manipulative content or create disorientation among users through information that is inaccurate, vague, or purposively misleading (Benkler et al., 2018).

Media criticism is an important and necessary element of a functioning democracy, not least as politics have become increasingly mediatised and dependent on media logic (Esser et al., 2012). The ability to criticise news reporting, including political journalism, is a way to scrutinise the power of legacy media. By holding news journalism responsible for accuracy in the information they produce and disseminate, media criticism can function as a powerful control tool. As with any powerful social institution, holding news journalism accountable is a way of evaluating how it “fulfils its democratic functions” (Egelhofer et al., 2021). If used sensibly and accurately, media criticism can also improve news journalism in various ways (Egelhofer et al., 2021). Democratically relevant media criticism is dependent on rational, truthful, and transparent political discourse. Pinpointing factual errors in the news is a prime example of criticism that – if communicated truthfully – can help citizens make informed decisions, and also improve the practices and routines of journalism. The same could be said about political bias, if the underlying evidence for such a bias is provided in an honest, factual, and transparent way by the communicating actor. Current research shows that politicians’ use of social media has increased the presence of media criticism to new record levels (Carlson, 2017). This criticism is often neither constructive nor democratically useful for citizens but takes increasingly “toxic” forms (Cheruiyot, 2019: 224). This includes media criticism that delegitimises journalism solely for political purposes, altering public perceptions of legacy media in negative ways. This development has been spearheaded by the “fake news” accusations by influential political leaders such as Donald Trump. It has also been the modus operandi of right-wing populist actors across Europe for some time now (Egelhofer et al., 2021; Figenschou & Ihlebæk, 2019). For them, the purpose is not to fact-check, to improve the quality of journalism, or provide evidence for political bias, but to simply influence various target groups’ perceptions of legacy media in a negative way. Increased audience fragmentation has also made these propaganda strategies relatively effective, especially within certain user segments characterised by low trust in legacy media. In these segments, online network logics and architectures (that steer information flows and create feedback, reactions, and circulation) also accelerate propagandist efforts by political actors (Benkler et al., 2018).

Distinguishing media criticism that builds on rational, truthful, and transparent argumentation from criticism that seeks to undermine the legitimacy of journalists or news institutions solely for political purposes is an important research endeavour. In this article, however, we do not analyse rationality (or the lack thereof) in political media criticism per se. Our focus is on the broader systematic aspects of political media criticism. We seek to identify what types of arguments politicians of different parties express, the main targets of their criticism, and how these aspects can be understood in the context of current research about political journalism in Sweden. An important aspect to consider, therefore, is media systemic aspects.

Generally, the Swedish media system is characterised by a high degree of professionalisation of journalism and high circulation of newspapers, but at the same time, a low level of parallelism between political actors and news media companies (Hallin & Mancini, 2004). There is still a rich flora of strong, yet economically pressured, newspapers, private television, and radio channels, but also influential public service media institutions that reach millions of Swedes through television (SVT), radio (SR), and the Internet on a daily basis (Syvertsen et al., 2014). Studies show that that people with right-leaning opinions, and particularly those who sympathise with the Sweden Democrats, trust the news media to a lesser extent than citizens with attitudes positioned at the centre or to the left of the pollical spectrum, and this relationship is particularly strong when it comes to trust in public service media (Andersson & Weibull, 2017). Longitudinal studies of election news in Sweden over several decades show no support for claims that public service media are systematically biased along party lines. On the contrary, public service media stand out as more balanced in their coverage compared with newspapers. The tendency over the last decade is that the right-wing parties have received slightly more favourable media coverage compared with the parties on the left. However, the far-right populist party, Sweden Democrats, stands out in the sense that it has received slightly more critical and unfavourable media coverage compared with other parties. This pattern can be seen in election coverage of newspapers and commercial television, as well as by public service media institutions (Asp & Bjerling, 2014; Johansson, 2022; Johansson & Strömbäck, 2019). On the other hand, studies of non-election periods show that the Sweden Democrats have been increasingly normalised in the journalistic discourse (Ekström et al., 2020).

The complex relationship between the Sweden Democrats and the news media has been an important driver behind the development of a new hyper-partisan and alternative-media ecology in Sweden with close ties to the party (Ihlebaek & Nygaard, 2021; Ekman & Widholm, 2022). Criticism of legacy media is an important ingredient of these alternative media outlets, but the prime target has been the Swedish public service media institutions. This reflects a broader trend among right-wing populist parties in Europe, who not only criticise but also seek control over public service media institutions (Sehl et al., 2022). An important contextual factor behind this development is that cultural diversity and gender equality have become increasingly central for the public service media remit in many countries (Horsti & Hultén, 2011). Such ideals clash in many respects with the conservative or nationalist ideology of far-right populist parties. As we shall see, all these aspects are important to have in mind when analysing media criticism from a propaganda perspective.

This study builds on a content analysis (Krippendorff, 2013) of tweets published by politicians on the national level in Sweden during the 2018 election campaign. We draw on a census sample including all on-duty members of the Swedish parliament with an active Twitter account. In addition, we included all party leaders and ministers of the Swedish government, in case they did not have a seat in the parliament. The collected accounts (155 in total) were selected based on information gathered through the Swedish parliament’s open data (Riksdagen, 2018), from which we obtained information about the politicians’ party affiliation and whether they were in office during the study period. We used a data collection platform at Stockholm University which utilises automated searches through Twitter’s Graph API and captures tweets and engagement data from selected accounts according to specified time intervals. The timeframe for the study was set to four weeks (28 days) prior to the election day. Data were exported from the platform one week after the election day to ensure that tweets published at the end of the period had enough time to generate user engagement.

Publications on Twitter take three main forms: tweets (including quote tweets), retweets, and replies (Molyneux & Mourão, 2019). Tweets and quote tweets can be considered original messages, while retweets are uncommented redistribution of another actor’s message. Replies consist of direct responses to another actor’s tweet. Since we were primarily interested in the politicians’ own communication, we chose to limit the analysis to tweets and quote tweets. In order to find a reliable sample of media-critical tweets, we started by identifying publications with explicit relationships with news media. Media-related tweets were defined as tweets containing links to news media or mentions of news media organisations, specific newsrooms, or individual journalists (N = 3,179). To further limit the data, we excluded uncommented shares of Internet links, since such tweets do not express explicit opinions or commentary. The total number of tweets corresponding to this definition was 2,330 (n).

The Moderate Party had the largest number of tweeters, followed by the Social Democrats, mirroring the fact that they were the biggest parties in the Swedish parliament. A comparison of the power balance in the parliament with the power balance on Twitter (see Table 1) shows a noticeable overrepresentation of parliamentarians from the Moderate Party on Twitter, while the Social Democrats and the Sweden Democrats were underrepresented. Differences for the other parties were less distinctive. These relationships are important to have in mind, especially regarding questions relating to user engagement in the analysis.

We started the coding process by conducting a pilot study in order to inductively operationalise variable values suitable for our study objectives. The pilot study resulted in a codebook comprising a broad set of variables of which we use seven in this study. The codebook consists of manually as well as automatically coded variables. Name and party affiliation were detected automatically through the data collection platform. Party affiliation was manually recoded into three party types corresponding to the ideological landscape of Swedish politics at the time (Demker & Odmalm, 2021): liberal/conservative parties (comprising the Moderate Party, the Centre Party, the Christian Democrats, and the Liberal Party); right-wing populist parties (represented by the Sweden Democrats); and left-wing/green parties (comprising the Social Democrats, the Left Party, and the Green Party). It should be mentioned that these party constellations were reformed some months after the election, when the Centre Party, formerly aligned towards the right, voted in such a way that the Social Democrats and the Green Party could form a government. The Moderate Party and the Christian Democrats formed at the same time a new block with support from the Sweden Democrats, who previously had been excluded from all cross-party alliances in the parliament (Öhberg, 2020).

To address the question of which parties and party types expressed the most media criticism (RQ1), we used a dichotomous variable, distinguishing tweets where media criticism was present from tweets lacking such criticism. Media criticism was defined as negative and critical comments about the content, practices, or values of journalism, specific news media companies, specialty desks (sport, cultural, editorial, etc.), or individual journalists. To address the question of what types of media criticism politicians express (RQ2), we coded dichotomously for seven potentially overlapping dimensions: 1) inaccuracy, for example, tweets pointing to concrete factual errors or journalistic misconceptions; 2) political bias, where the focus is on alleged partiality and lack of journalistic objectivity; 3) name calling, for example, tweets that seek to establish distrust in specific news media institutions or in journalism more generally; 4) deceptive framing, focusing on how journalists obscure certain aspects of reality or mislead their audiences; 5) news values and ethics, criticising unethical journalistic practices and banal news angles on allegedly irrelevant societal phenomena; 6) systemic criticism, focusing on private media ownership structures as well as financial models for public service media; and 7) ideological criticism, focusing on interpretations and arguments expressed by opinion journalists and subjective columnists.

The seven dimensions connect in different ways to our definition of propaganda. As previously noted, transparency is key in constructive forms of media criticism – that is, to provide evidence for the argument stated (e.g., pointing to concrete factual errors or examples of how a story is deceptive), but equally important is to avoid over-simplification and misleading generalisations. Name-calling is the most apparent form of generalisation that only serves to delegitimise news institutions or journalists. Likewise, if politicians accuse journalism of biased reporting repeatedly and systematically without providing evidence, it too can be seen as a deliberative delegitimisation campaign (Benkler et al., 2018). The various targets of media criticism (RQ3) were coded both conceptually and specifically. On the conceptual level, we coded nominally for whether the criticism was directed against specific media companies or individuals, or whether the focus was on journalism in general. On the specific level, the collection platform enabled us to automatically categorise which media companies were mentioned or referred to in the tweets. These were then manually re-coded into the following categories: public service media, commercial television and radio, press (i.e., newspapers), and journalism in general. Tweets criticising several media companies at the same time (e.g., commercial television and public service television) were coded as general. Finally, to address the question of user engagement (RQ4), we drew on a numerical engagement variable, based on the number of retweets, likes, and replies each tweet received. The sum of these different types of interactions is what we refer to as engagement in the analysis. The manual coding was conducted by a research assistant. An inter-coder agreement test of a random sample of media-critical tweets (n = 50) was made to ensure a high level of reliability. We used Holsti’s method for determining reliability. The test showed coefficients over 0.8 for all variables, indicating that the coding is methodologically reliable. Since the study is based on a census sample, statistical significance tests are not relevant (Krippendorff, 2013).

Over the last decade, media criticism in social media has become an increasingly prevalent feature of election campaigns, especially among right-wing populist parties. This tendency was also reflected in Sweden during the 2018 election.

As seen in Table 2, the vast majority of all media-critical tweets (82%) were posted by politicians of either the Moderate Party or the Sweden Democrats. It is clear that the further a party is positioned to the right on the political spectrum, the more likely it is that the party utilises media criticism in communication on Twitter. The Sweden Democrats had the highest share of media-critical tweets of all parties, followed by the Moderate Party and the Christian Democrats. The Left and Green Parties were the ones with the lowest level of media criticism, although the differences were relatively small when compared with the Liberal Party and the Centre Party. Hence, while there is a historical tradition of leftwing media criticism in Sweden, based on criticism of private ownership structures, dominance of bourgeoise editorial pages, and commercialisation debates (Svensson, 2015), as well as a strong political criticism of the content of public service broadcasters, particularly in the early 1970s (Djerf Pierre & Weibull, 2001), it is obvious that few contemporary left-wing politicians consider media criticism to be a fruitful instrument of influence on Twitter.

Previous research suggests that the constant struggle for attention among political actors on social media contributes to increased political personalisation (Ekman & Widholm, 2017; Metz et al., 2020). Through social media, individual politicians can run their own personal campaigns and position themselves in relation to competitors of other parties, as well as internally by emphasising certain political niches, even in otherwise party-centred political systems (Enli & Skogerbø, 2013). Subject matters can also be divided strategically between actors of a party to ensure influence across a broad range of political areas. Media criticism can be seen as such a communicative niche, particularly within the Moderate Party, where one parliamentarian, Lars Beckman, published the majority of the party’s media-critical tweets, and almost half of all media-critical tweets in the sample (see Table 3). Martin Kinnunen of the Sweden Democrats had a similar, yet not as distinctive, position in his party. It should be emphasised though, that media criticism is also established on a broader level within these parties – 7 out of 14 representatives from the Sweden Democrats, and 18 out of 47 representatives of the Moderate Party expressed on Twitter a critical stance against the news media at least once during the election campaign. It is moreover noteworthy that the top media critic of the Centre Party, Staffan Danielsson, changed party affiliation after the election, and since 2021, he is a member of the Christian Democrats. While there are indeed clear ideological connections between right-wing party positions and dissemination of media criticism on Twitter, our results show that the phenomenon must also be understood in terms of personalisation.

Table 5 displays the types of news media that politicians criticise the most, and it is clear that the prime targets are the two public service media organisations SVT and SR. This holds true for all party types, although the liberal/conservatives stand out as most critical against public service media. The scarce criticism from the left/green parties is more varied; for example, it is to a greater extent directed towards the press and commercial broadcasters.

In terms of foci (see Table 6), these parties also tended to be more specific in their criticism; for example, they did not use generalised language (e.g., “mainstream media”, “PC media”, “fake news”, etc.). The more individualised criticism from the left-wing parties can be related to their tendency to more often criticise the press, where subjective columnists, essayists, and pundits have a more central role.

There are four key factors that explain why public service media receives much more criticism compared with other media. First, their size an impact are an important factor. SVT and SR have both local and regional audiences as well as a large nationwide audience, and their influence is unparalleled in the Swedish media system. A second factor is related to content: European public service media organisations focus more on hard news and policy issues rather than soft news and “horce-race journalism” related to the political game (Esser et al., 2012). SVT and SR have considerable resources for investigative reporting, they arrange the most-watched debates, and they hold politicians accountable for their actions through critical interviews more often than commercial media companies do. For a politician who wants to “correct” alleged journalistic misrepresentations of reality (Ekman & Widholm, 2015), the content of public service media seems to be unavoidable. A third factor is the widespread notion by the far- and populist-right that public service media organisations are biased in favour of the political left. Studies show that people with right-leaning opinions, and particularly those who sympathise with the Sweden Democrats, trust the media to a lesser extent than citizens with attitudes positioned at the centre or to the left of the political spectrum, and this relationship is particularly strong when it comes to trust in public service media (Andersson & Weibull, 2017). Hence, media criticism in social media resonates with, but also fuels, established attitudes in the Swedish citizenry. Moreover, research shows that far-right populist actors see public service media as part of a broader “left-liberal” establishment and elite. Therefore, strategic attacks on public service media institutions and actors are often carried out solely for political purposes (Sehl et al., 2022; Holtz-Bacha, 2021). A fourth factor is related to political influence. Political actors have the power to change the mission and broader orientation of public service media institutions, not least through changes in cultural policy and budget regulations (although such changes take time). This is also why questions relating to public service media are more disputed. The fact that public service media companies are funded through taxation motivates high demands in terms of quality of the output, but it also creates political expectations of certain outcomes for the resources invested (e.g., more of a particular type of content orientation, the establishment of certain values in the organisations, etc.). In addition, right-wing parties have traditionally been more critical of the public service media institutions in general, reflecting their more market-friendly policies. Politicians cannot exert formal influence over the practices and values of professional journalism in Sweden; their main impact lies instead on the broader systemic level. However, media criticism may be influential on a more informal level, for example, if it leads to self-censorship or avoidance of certain political topics, sources, or angles in the news (Cheruiyot, 2022; Löfgren Nilsson & Örnebring, 2016).

The uneven distribution of media-critical tweets between party types is also reflected in the figures for digital user engagement, but the differences between the left/green parties and the others are in fact even more accentuated. Figure 1 shows the distribution of user engagement across party types (%, bars), and mean (yellow line) and median (green dotted line) values for all media-critical tweets in our sample. 64.9 per cent of all engagements can be attributed to the liberal/conservative parties, 33.6 per cent to the right-wing populist parties, and only 1.5 per cent to the left/green constellation. Moreover, when left/green parties eventually published media-critical tweets, they created few reactions among users, reflected in their significantly lower mean and median values. While this is unsurprising, given the right-wing dominance on social platforms in Sweden (Larsson, 2020), our study underscores that political media criticism has become a predominantly right-wing phenomenon on Twitter.

FIGURE 1

A notable difference can be seen between the right-wing populist and the liberal/conservative parties: An average media-critical tweet published by the former generated a higher level of interaction than an average tweet published by the latter. This explains why the right-wing populist actors’ share of the total engagement (34%) was higher than their share of the total number of published tweets (21%). While these differences are clear, it is important to note that media criticism must also be understood on the party level, and not least in relation to individual politicians’ strategic decisions. The dominance of the liberal/conservative parties is first and foremost an effect of how members of the Moderate Party communicate on Twitter – their tweets generated as much as 96 per cent of all engagements within the liberal/conservative party group (see Table 7). In addition, it is clear that individual actors within the Moderate Party were able to create a much wider circulation of certain tweets. The most engaging tweet was published on Twitter by one of Sweden’s most influential politicians, Hanif Bali, and alone generated 8,287 interactions. In comparison, that is nearly six times more than Sweden Democrats’, and nearly 23 times more than the Social Democrats’ most engaging tweet during the period.

This study found that political media criticism in Sweden was primarily expressed by a small group of influential politicians connected to the Moderate Party and the Sweden Democrats. Notably, one member of parliament from the Moderate Party accounted for nearly half of all media-critical tweets in the sample. A clear majority of the criticism was aimed at public service media. As previously noted, studies of election news in Sweden show no support for claims that public service media are systematically biased in favour of the political left (Asp & Bjerling, 2014; Johansson & Strömbäck, 2019). An important conclusion that can be drawn in light of our data is that media-critical actors do not primarily react to known factual circumstances of systematic bias, but as a result of strategic decisions. The size and reach of public service media, the type of news they produce, existing media-critical opinions among the citizenry, and political actors’ ability and willingness to apply pressure to the media in terms of structural influence are key to understanding these prioritisations. The specificity of our case – an election campaign – also involves situational factors that likely contributed to the strong criticism against public service media. One such factor was the Swedish public service television broadcaster SVT’s decision to officially denounce a statement made by the leader of the Sweden Democrats during the final television debate. The decision instigated many reactions on social media, and leading politicians of the Sweden Democrats refused interviews for SVT in the aftermath. The Swedish Press and Broadcasting Authority later concluded that SVT violated the rules of impartiality in content (SVT, 2019). However, the final debate took place at the very end of our study period, and most of the criticism of public service media was expressed before this event.

Criticism of news media’s political bias and name-calling were the dominating dimensions among liberal/conservative and right-wing populist politicians, but not among left/green politicians. Criticism pertaining to factual errors and to news values and ethics were more common in the small number of media-critical tweets published by the left/green representatives. This points to the fact that a weaponised criticism directed towards mainly public service television and radio is common among party representatives from the right-wing of Swedish politics. We have also shown that a more general media criticism is relatively common, for example, accusing “the media”, “journalists”, or multiple companies at the same time for being “fake” or “biased”. These forms of propaganda strategies could be understood as a way of undermining the epistemological authority and legitimacy of news institutions.

The patterns in the Swedish election campaign on Twitter reflect a general development internationally, where legacy media have become a common target of right-wing populist actors, and where more traditional conservative political actors also use media criticism aimed at legacy media (Cimaglio, 2016). The criticism articulated by the member from the conservative right-wing party, the Moderate Party, towards mainly public service media, could also be understood from the perspective of personalisation, as the vast majority of the tweets were posted by one particular MP. The fact that the media criticism of right-wing populist politicians creates higher levels of circulation and reactions on Twitter (compared with the other party groups) points to the effects of feedback loops of political propaganda (Benkler et al., 2018). The distrust towards legacy media in general, and public service media in particular, among supporters of right-wing populist politics, makes media criticism an effective tool for affective mobilisation (Ekman, 2019) and increases the risk of disorientation on factual matters. Distrust in legacy media has shown to be highly asymmetric (Benkler et al., 2018), and since supporters of right-wing populist parties stand out in this respect, a provocative style, including weaponised criticism in online communication, does not only confirm already prevailing distrust, but also risks increasing it. There is also a risk that it propels threats against journalists and news professionals working within legacy media (Löfgren Nilsson & Örnebring, 2016). Because contemporary political communication takes place in an increasingly fragmented and complex media ecology, we argue that propaganda studies could shed new light on the political communication of elected representatives online. By highlighting propaganda strategies, including micro targeting, future research should distinguish between constructive forms of political communication that can strengthen democratic deliberation, and more destructive ones that undermine a much-needed transparent critical discussion about the societal role of the news media.

To conclude, political media criticism can be a constructive way of pinpointing problems in news reporting such as factual errors or commercialised news angles that push sensationalism before accuracy (Cheruiyot, 2018). But as this article has shown, a vast majority of the media criticism seems to be targeted campaigns where a weaponised criticism of public service media stands out, and where the goal is to undermine the credibility and legitimacy of these institutions for temporary feedback goals. Certain target groups have shown to be particularly susceptible for such propagandistic media criticism.