Undermining the legitimacy of the news media: How Swedish members of parliament use Twitter to criticise the news media

From a democratic perspective, free and independent news media are an essential institution (Cook, 2005), necessary for democracy to function well (Baker, 2002; Cook, 2005; Ferree et al., 2002; Strömbäck, 2005). Among their key democratic functions are providing citizens with verified and truthful information about societal matters, acting as a watchdog against abuses of power, and deterring those in power from illegal or unethical behaviour (Baker, 2002; Kovach, & Rosenstiel, 2021).

For the news media to be able to fulfil their democratic functions, it is, however, important that they are both broadly used and trusted (Fawzi et al., 2021; Strömbäck et al., 2020). Otherwise, it would signal a lack of epistemic authority, and it would not matter how informative the news media's coverage is or how many wrongdoings they expose: People would either not get exposed to or not trust the information they are exposed to, and thereby the news media would not be able to “aid citizens in becoming informed” (Holbert, 2005: 511).

At the same time, recent developments suggest that news media have increasingly become the target of politically motivated attacks seeking to undermine trust in news media. The most famous example is former American president, Donald Trump, who repeatedly accuses the news media of providing “fake news” and being “an enemy of the people” (Carlson et al., 2021; Meeks, 2020), but there are numerous other examples from countries such as Australia (Farhall et al., 2019; Wright, 2021) and Germany (Egelhofer et al., 2021; Koliska & Assmann, 2021), not to mention different semi-democratic and authoritarian countries (Solis & Sagarzazu, 2020; Waisbord & Amado, 2017).

Based on previous theory and research, two reasons for the increasing prevalence of politically motivated attacks on the news media are the increasing prominence of right-wing populism (Esser et al., 2017; Fawzi, 2019; Hameleers, 2020), and that digital and social media have made it much easier for political actors to bypass the news media and express media criticism (Benkler et al., 2018; Engesser et al., 2017; Farhall et al., 2019). Importantly, this also includes elected politicians, who have power not only to influence public opinion but also to shape the rules and regulations that news media must operate within (Waisbord & Amado, 2017). From that perspective, it could be argued that attacks on the news media coming from elected politicians pose a greater danger to free and independent news media than media criticism from other groups in society. This holds particularly true for media criticism that can be characterised as seeking to delegitimise the news media.

Thus far, there is only limited systematic research, however, on the extent to which politicians use social media to attack the news media, not least including a representative sample of politicians rather than just politicians who are considered populist (Egelhofer et al., 2021; Farhall et al., 2019; Waisbord & Amado, 2017; Wright, 2021). The most important exception is Egelhofer and colleagues (2021), who investigated the extent to which the main candidates for all Austrian and German parties engage in delegitimising media criticism on Facebook. Despite this exception, a key research problem is that our knowledge about the prevalence and forms of delegitimising media criticism coming from politicians across ideological affiliations is limited.

The aim of this study is therefore to investigate political elites’ media criticism, focusing on Sweden. In brief, this case was chosen because traditionally, Sweden represents a media welfare state where there has been broad political consensus about the media's main political roles, and that media may be supported by political means (press subsidies, public service) but should be independent and governed by self-regulation (Syvertsen et al., 2014; von Krogh, 2012). While media criticism by political elites is by no means new, it has traditionally aimed at improving the functioning and quality of traditional news media and holding them accountable (von Krogh, 2012). Delegitimising criticism from political elites has been rare, at least in the public debate (von Krogh, 2012). Relatedly, Sweden represents a very different case compared with the US, where most research has been done. This holds true for the media system as well as the political system, the degree of polarisation, and levels of trust in news media (Brüggeman et al., 2014; Humprecht et al., 2022). Thus, focusing on the context of Sweden will contribute to broaden extant research.

In terms of platforms where political elites may express their media criticism, this study focuses on Twitter. While it is only one among several platforms where political elites can bypass and criticise mainstream news media, previous research suggests that media criticism may be more common on Twitter than other platforms (Farhall et al., 2019). One reason is that Twitter is popular among journalists (Molyneux & McGregor, 2022). By levelling their media criticism on Twitter, political elites can directly reach the targets of the attacks. More generally, Twitter is often used by politicians to reach and influence both journalists and politically interested groups and to amplify messages also distributed on other platforms (Jungherr, 2016; Silva & Proksch, 2022). Hence, media criticism on Twitter is important in itself, and communication on Twitter may spill over to other contexts (Åkerlund, 2020). How members of parliament (MPs) tweet may thus serve as a proxy for how political elites express media criticism on other platforms and in other contexts.

More specifically, then, the purpose of this study is to investigate the extent to which Swedish MPs on Twitter engage in delegitimising media criticism targeting mainstream news media, the party distribution of those engaging in such media criticism, and the targets and expressions of such critique. By mainstream news media, we refer to news media such as broadsheet and tabloid newspapers, radio, and television that are guided by journalistic norms emphasising that the news should be verified, factual, impartial, and serve to inform the public (Kovach & Rosenstiel, 2021).

In modern and highly complex societies, no individual can ever acquire all the information they need to function by themselves. Instead, we are dependent on others, ranging from peers to experts within knowledge-producing and knowledge-disseminating institutions such as schools, universities, and the news media. Thinking and knowing can thus be described as “collective action”, where “our intelligence resides not in individual brains but in the collective mind” (Sloman & Fernbach, 2017: 5). This reflects that contemporary societies “cannot function without a social division of labor and a reliance on experts, professionals, and intellectuals” (Nichols, 2017: 14).

In this context, the news media are essential. To begin with, for most people, the news media constitute the most important source of information about matters beyond their personal experiences (Mitchell et al., 2016; Mutz, 1998; Shehata & Strömbäck, 2014). Even though people increasingly get news from social media and political alternative media, most news are produced and disseminated by traditional news media. The news media also serve as crucial intermediaries of information coming from other knowledge-producing institutions, such as universities and governmental authorities. Equally important, key functions of the news media are to be independent of those they cover, to verify information before it is published, and to provide people with truthful information about current affairs and societal matters (Ferree et al., 2002; Kovach, & Rosenstiel, 2021; Strömbäck, 2005). In contrast to political parties and political alternative media, which are guided by what Benkler and colleagues (2018) labelled “propaganda feedback loops”, the news media are largely guided by “reality-check dynamics” (see also Strömbäck et al., 2022).

For the news media to fulfil their democratic functions, it is thus essential that they are widely used, trusted, and recognised as epistemic authorities (Ekström & Westlund, 2019; Wikforss, 2021). Following Gieryn (1999: 1), an epistemic authority is recognised as having “legitimate power to define, describe, and explain bounded domains of reality”, thereby leading to “public acceptance of knowledge claims” (Ekström, 2002: 261).

In this perspective, it is important to distinguish between media criticism that accepts news media as an epistemic authority and that seeks to hold the news media accountable and improve journalistic practices and quality (Carlson, 2009), and media criticism that is used strategically to challenge their epistemic authority and influence how others perceive the news media (Farhall et al., 2019; Solis & Sagarzazu, 2020; Waisbord & Amado, 2017). Following Egelhofer and colleagues, (2021), the latter can be characterised as delegitimising media criticism. Typically, delegitimising media criticism can be distinguished from other forms of media criticism by being characterised by a lack of substance, lack of civility, and lack of reasoning or specificity (Egelhofer et al., 2021). Although delegitimising media criticism can take many different expressions, some examples are claims that news media are “fake news”, an “enemy of the people”, and that they are hiding the truth from the public or are lying (Cheruiyot, 2018; Egelhofer et al., 2021; Farhall et al., 2019; Figenschou & Ihlebæk, 2019; Koliska & Assmann, 2021; Tandoc, 2019; Wright, 2021). Another type of delegitimising criticism is sweeping and unsubstantiated claims of political bias, which may be used in politically motivated attacks on the news media (Domke et al., 1999; Eberl, 2019; Watts et al., 1999). Delegitimising media criticism thus has all the trappings of an intention to undermine trust in, and the legitimacy of, the news media.

In contrast to accountability media criticism, delegitimising media criticism can be theorised to have at least three forms of effects. First and most importantly, it might lower the trust in and legitimacy of journalism and news media (Figenschou & Ihlebæk, 2019; Hanitzsch et al., 2018; Ladd, 2012; Strömbäck et al., 2020; Tsfati et al., 2022). Second, it might shield individuals and organisations that are subjects of critical media coverage from criticism and deflect public attention (Carlson, 2009; Carlson et al., 2021; Neo, 2022; Ross & Rivers, 2018). Third, it might increase the demand for, and use of, partisan or political alternative media, that is, media that are guided by political motivations and propaganda feedback loops (Benkler et al., 2018; Figenschou & Ihlebæk, 2019; Holt et al., 2019; Müller & Schulz, 2021).

Importantly, each of these potential effects may contribute to undermining the news media as a democratic institution, including their ability to provide people with the information they need rather than just the kind of information that confirms people's worldviews and attitudes, their ability to hold those in power accountable, and their ability to deter those in power from wrongdoings. All these potential effects may, in fact, be the very purpose of delegitimising media criticism.

Despite the potentially negative effects that delegitimising media criticism has on the public trust in news media as a credible democratic institution, our knowledge of the prevalence and forms of delegitimising media criticism with politicians as attackers is quite limited. That holds particularly true beyond the US, where there is extensive research on, among other things, former American president Trump's attacks against the media and the Republicans’ recurring allegations of a liberal news bias (D’Alessio, 2012; Domke et al., 1999; Hassell et al., 2020; Watts et al., 1999).

One exception, though, is a study on the extent to which a sample of Australian federal politicians in the aftermath of the 2016 American presidential election used the fake news discourse on different platforms (Farhall et al., 2019). Among other things, the results show that this discourse was quite rare and that it was mainly used by conservative politicians. Moreover, it was found that such media criticism was more common on Twitter than on other social media. In another Australian study, Wright (2021) investigated attacks on the news media by a sample of 26 federal politicians on Twitter during 2011–2018. The results show an increase in attacks on the media over time, with a substantial upswing in 2016 as Trump took office. In this study as well, most of the attacks came from conservative and right-wing populists, and the most frequent target was public service news media. Among the accusations were fake news, bias, and criticism toward specific news stories for being untrue or dishonest.

A more prominent exception is the already mentioned study by Egelhofer and colleagues (2021), in which they investigated the prevalence of delegitimising criticism against the media in Facebook posts from all major parties and their main candidates in Austria and Germany. Their study covered about a year, and it explicitly focused on delegitimising media criticism. Dwelling on a deliberative approach for what constitutes healthy democratic conversation (and criticism), delegitimising media criticism was defined as being characterised by 1) a lack of reasoning, or 2) presence of incivility. Lack of reasoning means the absence of arguments for why a journalistic practice or coverage is being criticised, whereas presence of incivility refers to the use of, for example, insulting language, name-calling, and character assassination. Altogether, Egelhofer and colleagues found about 3,200 references to media or journalism, but only about 6 per cent of these included any media criticism. Out of these, about 40 per cent were classified as including delegitimising media criticism. Similar to the aforementioned studies, they also found that right-wing and populist actors used more delegitimising media criticism than others. The most common target was media in general, but when specific media were targeted, the most common targets were public service media and quality papers.

Finally, there is one Swedish study investigating how Swedish MPs tweet about the news media (Widholm & Mårtenson, 2018). The study covered four weeks and found that 7 per cent of all media-related tweets included some kind of media criticism. The extent to which that media criticism was delegitimising is, however, not clear.

Altogether then, it is unclear how common delegitimising criticism against the news media is, although it seems to constitute a rather small share of all social media posts related to the news media. To further explore this, we build on the framework of delegitimising criticism by Egelhofer and colleagues (2021) and ask the following research question:

RQ1. How prevalent is delegitimising criticism toward the news media among Swedish MPs?

A recurring finding in previous research seems to be that delegitimising media criticism in Europe and the US is more common among right-wing populist politicians than among other politicians (Egelhofer et al., 2021; Farhall et al., 2019; Wright, 2021). In the case of Sweden, that would mean that MPs from the right-wing, authoritarian populist party, the Sweden Democrats, are likely to be more hostile to the media than MPs from other parties (Norris & Inglehart, 2019; Strömbäck et al., 2017b). However, the study that investigated Swedish MPs tweeting about the media found that media criticism was equally likely to come from MPs belonging to the Sweden Democrats as from the conservative party, the Moderates (Widholm & Mårtenson, 2018). During the last years, the Moderates and the Sweden Democrats have furthermore begun to collaborate more extensively. This suggests that the distinction between populist and non-populist parties and politicians is not clear-cut.

At the same time, Swedish research shows that trust in news media is lower and that hostile media perceptions are more common among Sweden Democrats than among other political affiliations (Andersson, 2021; Strömbäck & Karlsson, 2017; Oleskog Tryggvason & Andersson, 2022). Moreover, several of Sweden's right-wing and media-critical political alternative media are, directly or indirectly, linked to the Sweden Democrats (Ekman & Widholm, 2022; Vergara, 2016, 2021). Combined with the notion that right-wing populists tend to perceive the news media as part of the elite that they are against (Fawzi & Krämer, 2021; Hameleers, 2020; Holt, 2020), this creates incentives for MPs belonging to the Sweden Democrats to use delegitimising media criticism. Hence, our first hypothesis, followed by our second research question, is:

H1. Delegitimising media criticism is predominantly expressed by MPs from the Sweden Democrats.

RQ2. What is the party distribution of those MPs that express delegitimising media criticism?

In terms of the targets of delegitimising media criticism, we expect the main target to be news media in general, in line with previous findings from other countries (Egelhofer et al., 2021). With news media in general, we refer to traditional mainstream news media such as radio, television, and newspapers in their traditional or digital formats. Beyond that, we expect the main target to be public service media. In Sweden, public television (SVT) and public radio (SR) are the most widely used media (Martinsson & Andersson, 2022; Nord & Grusell, 2021), and several parties towards the right are more critical toward and want to decrease the budget and commission of public service media (Bengtsson, 2021; Nord, 2021; Strömbäck, 2022). The earlier study by Widholm and Mårtenson (2018) also found public service media to be the main target of media criticism. Hence, our next hypotheses are:

H2. Delegitimising media criticism is mainly targeting news media in general.

H3. When delegitimising media criticism is targeting specific media, the main target is public service media.

In terms of how delegitimising media criticism is expressed, previous research suggests that recurring themes are accusation of political bias, but also that the news media spread false or fake news, mislead the public, and hide important truths (Cheruiyot, 2018; Egelhofer et al., 2021; Farhall et al., 2019; Figenschou & Ihlebæk, 2019; Holt, 2020: Koliska & Assmann, 2021; Tandoc, 2019; Wright, 2021). Sometimes the criticism is quite general, as when “the media” is accused of being biased. Other times it is targeted at the news coverage of specific topics. In Sweden, that holds particularly true for the news coverage of immigration (Strömbäck et al., 2017a; Truedson, 2016), where research also shows that people have lower trust in the news coverage of immigration than in the coverage of other issues (Andersson, 2017; Tsfati et al., 2022). The prevalence of different expressions of delegitimising media criticism is, however, unclear. Hence, our last research question is:

RQ3. How prevalent are different expressions of delegitimising media criticism?

To investigate the hypotheses and address the research questions above, we draw on a content analysis of all media-related tweets by Swedish MPs over one full year: 31 October 2020–31 October 2021. This period was selected because we wanted to capture a routine political period when there was no election.

The data collection proceeded in several steps. First, all MPs with an identifiable Twitter account were identified manually. The Swedish parliament has 349 seats and eight parties in parliament. The parties currently belonging to the centre-left are the Centre Party, the Social Democrats, the Green Party, and the Left Party, while the parties belonging to the centre-right are the Liberal Party, the Christian Democrats, the Moderate Party, and the Sweden Democrats. The latter is considered an authoritarian, right-wing populist party (Norris & Inglehart, 2019; Strömbäck et al., 2017b). At the time of data collection, 219 Twitter accounts were manually identified and included in the study (see Table 1).

Second, Twitter data from all analysed MPs within the one-year study frame was collected and downloaded via Twitter's REST-API using Chorus. Altogether, this resulted in a sample of 117,039 tweets.

The unit of analysis in this study is single tweets including keywords related to media. To identify these, our third step was to create a search string consisting of terms related to media and journalism, including both media in general and prominent specific media (the search terms were Aftonbladet, Dagens Nyheter, Expressen, Svenska Dagbladet, SVT, SR*, public service, radio*, DN*, SVD, blask*, journali*, medi*, presse*, svenskan*, tidning*, televis*, TV4). This generated a sample of 5,620 units (single tweets, replies, and re-tweets). Fourth, we excluded replies and re-tweets and focused only on original tweets, the reason being that these represent the most proactive form of tweeting about, in this case, criticism of news media. We also checked that the remaining tweets actually referred to news media and journalism. This resulted in a final sample of 2,244 tweets, which constitute the units of analysis in this study.

To investigate delegitimising media criticism, we developed a codebook that included formal variables such as party affiliation and date. But more importantly, the codebook included the following variables intended to capture the presence of delegitimising media criticism:

presence of media criticism

Moreover, the codebook included six variables intended to target different types of media criticism:

presence of left-wing bias accusations toward the media

All these variables were coded on a binary yes–no basis. In addition, we also coded the target of media criticism inductively in free text, as it was difficult to deduce all the possible targets a priori. For the codebook, see the online Supplement.

Following Egelhofer and colleagues (2021), we define delegitimising media criticism as media criticism that includes at least one element of either incivility or absence of reasoning. Based on previous research (Sobieraj & Berry, 2011; Egelhofer et al., 2021), incivility was operationalised using indicators such as rude and unnecessarily disrespectful language, name-calling, attempted character assassination, ideologically extremising language, and caricaturing or sarcastic language. Similarly, absence of reasoning was defined as tweets that accuse the media without giving specific examples or an argumentation (Egelhofer et al., 2021. Fake news attributions are an obvious example, but other examples are broad and unspecified accusations of bias, lies, falsehoods, and deceit. Table 2 provides five examples of tweets including delegitimising media criticism.

To check for intercoder reliability, one external coder coded 7 per cent of the sample (n = 153), resulting in an intercoder reliability (Krippendorffs’ alpha) of .75. Although this is somewhat lower than the customary .80, it is above the .667 limit (Krippendorff, 2013: 325) and may depend on the skewness of the values on some of the variables. Hence, we consider it satisfactory.

Turning to the results, let us begin by providing an overview of the number and party distribution of all tweets from Swedish MPs that contained any reference to media and journalism. First of all, it is worth noting that tweets making any reference to media and journalism (n = 2,244) make up less than 2 per cent of all tweets by MPs (n = 117,039) during the one-year time period. Beyond that, Table 3 displays the distribution of tweets that contained any reference to media and journalism across all party affiliations. Based on these results, a first observation is that MPs from some parties are clearly overrepresented when tweets refer to media and journalism, and that the distribution of tweets in this respect does not reflect the size of the party in terms of seats in parliament. Perhaps the most significant example is the Social Democrats which, despite its role as the largest and governing party at the time of investigation, stood for a marginal share of the sample.

Not all tweets referring to media and journalism are critical, however. In most cases, MPs made neutral references to media and journalism or mentioned the news media in passing, for example, when sharing news articles. In the end, the results show that 1,181 of the Swedish MPs’ tweets during the period contained some kind of valuation – positive or negative – of news media. Out of these, 73 per cent (n = 857) included some kind of criticism and 35 per cent (n = 413) some kind of support. In some cases (8%; n = 89), the tweets were both critical and supportive.

Turning to our research questions and hypotheses, RQ1 asked how prevalent delegitimising criticism toward the media is among MPs, while H1 hypothesised that such criticism would predominantly be expressed by MPs from the Sweden Democrats. Beginning with RQ1, the results first show that media criticism in general is not very prevalent, as less than 2 per cent of all tweets made any reference to media and journalism. Second, of all tweets containing any type of negative criticism toward the media, 75 per cent (n = 642) included some form of delegitimising media criticism. Thus, while tweets critical of the news media are quite rare overall, when MPs make some kind of critical evaluation of the news media, they are far more likely to use delegitimising than constructive media criticism. In fact, delegitimising media criticism is more common than both supportive tweets and constructive media criticism.

Concerning H1, Table 4 displays the party distribution of all tweets that include delegitimising media criticism. The results show significant differences across parties (Cramer's V = .334, p <.001). In contrast to H1, the party standing out is, however, not the Sweden Democrats, but the Moderate Party. More specifically, the results show that the Moderate Party stand for a clear majority, and more precisely 57 per cent, of all tweets including delegitimising media criticism. Table 4 also addresses RQ2, concerning the party distribution of those MPs that express delegitimising media criticism. As can be seen, the Sweden Democrats stand for the second largest share. In fact, together, the Sweden Democrats and the Moderates stand for 89 per cent of all tweets that include delegitimising media criticism, while the other six parties together stand for only about 10 per cent. Taken together, these results show that most delegitimising media criticism comes from the political right, broadly speaking. Additional analyses show, however, that MPs from the Sweden Democrats stand out in the sense that a larger share (45%) of all their media-related tweets included delegitimising media criticism compared with MPs from the other parties. In second place come MPs from the Moderate Party (36%), while MPs from the Green Party (32%) come in third place (26%). By this measure, MPs from the Left Party (6%), the Christian Democrats (7%), the Liberal Party (9%), and the Centre Party (9%) are much less prone to include delegitimising media criticism in their tweets.

Going beyond the party affiliation, further analyses show that some MPs are particularly engaged in tweeting delegitimising media criticism. More specifically, the results show that just one MP stands for fully 47 percent of the tweets that include delegitimising media criticism: Lars Beckman (Moderate Party). The findings also show that the five MPs who most often express delegitimising media criticism stand for about seven out of ten (71%) of the tweets that include delegitimising media criticism: Aside from Lars Beckman (Moderate Party) with 299 tweets, these are Robert Stenkvist (Sweden Democrats) with 64 tweets; Björn Söder (Sweden Democrats) with 46 tweets; Josef Fransson (Sweden Democrats) with 26 tweets, and Hanif Bali (Moderate Party) with 24 tweets. This suggests that some MPs, in particular Lars Beckman, have made delegitimising media criticism an important part of their political personae (see also Widholm & Mårtensson, 2018).

In terms of the target of delegitimising media criticism, H2 predicted that such criticism would mainly be aimed at news media in general, while H3 predicted that the main target in terms of specific media would be public service media. Here, the results show support for H3, but not for H2 (when addressing these hypotheses, we used the inductive coding of targets of media criticism, and all media could thus be a target and part of this ranking). Of all tweets including delegitimising media criticism, news media in general was the target of criticism in 24 per cent (n = 152) of the cases. This includes terms such as mainstream media, mainstream news media, international media, journalism, and media. In contrast, public service media was the target in 47 per cent (n = 300) of all tweets including delegitimising media criticism (see Table 5). Only 11 tweets included delegitimising media criticism targeting political alternative media (here, we include tweets referring to alternative media, partisan media, political alternative media, Bulletin, ETC, Riks, and Nya Dagbladet).

Finally, RQ3 asked how prevalent different expressions of delegitimising media criticism are. To address this, we investigated the prevalence of five types of criticism: poor quality (including poor standard, sensationalism, and selective coverage), left- or right-wing bias accusations, falsehood accusations, corrective information, and fake news attributions. The results are presented in Table 5.

Perhaps somewhat expected – given that most delegitimising media criticism comes from MPs belonging to the Moderate Party and the Sweden Democrats – the most common expression of delegitimising media criticism is bias accusations. Such accusations were present in 58 per cent of all tweets including delegitimising media criticism. Not surprisingly, such accusations are consequently almost exclusively about alleged left-wing bias. More precisely, 83 per cent of these tweets are about alleged left-wing bias, while 4 per cent are about alleged right-wing bias. The pattern is the same if we look at tweets that do not contain delegitimising media criticism but still accusations of bias (then, 81% include accusations of left-wing bias while 4% include accusations of right-wing bias; the rest refers to other types of biases).

The second most common expression of delegitimising media criticism targeted the quality of the news coverage (30%), including claims that media or journalistic practice hold a poor standard, is sensationalistic, or is sloppy. Thereafter follow accusations about falsehoods (8%), for example, that something in a news story is incorrect or false, or the use of more sweeping terms such as propaganda or manipulations. Importantly, though, this category does not include accusations that media are fake news. Given the great attention paid to this term over the last few years (Egelhofer & Lecheler, 2019; Tandoc, 2019), we coded this separately. However, only 1 per cent of all tweets including delegitimising media criticism used the term fake news. Similar to what was found in the case of Australia (Farhall et al., 2019; Wright, 2021), the use of the term fake news thus seems quite rare among Swedish politicians. When Swedish politicians attack the news media – at least on Twitter – they rather focus on alleged bias or low quality, in general.

From a democratic perspective, free and independent news media are an essential institution, necessary for democracy to function well. Not least important is their function of providing people with the information they need to be free and self-governing and holding those in power accountable (Kovach & Rosenstiel, 2021; Strömbäck, 2005). For the news media to be able to do this, it is key that they are both widely used and trusted.

From that perspective, attacks on the news media that might serve to undermine trust in, as well as the epistemic authority of, news media must be taken seriously (Waisbord & Amado, 2017). This is particularly true when such criticism comes from elected politicians. Thus far, however, there is only limited research on the extent to which politicians engage in criticism or attacks on the news media. To remedy this, the purpose of this study was to investigate the extent to which Swedish MPs engage in delegitimising media criticism on Twitter, the party distribution of the MPs engaging in such media criticism, and the targets and expressions of such media criticism.

While the results show, on the one hand, that it is quite rare for Swedish MPs to tweet about the news media, and that they in most cases only make neutral references, on the other, this study offers five key findings. First, the results show that negative media criticism is far more common than support for the news media. Second, a clear majority of all media-critical tweets can be characterised as delegitimising media criticism, that is, media criticism that may serve to undermine trust in, and the epistemic authority of, the news media. Third, such media criticism is most likely to come from the political right, and the Moderate Party and the Sweden Democrats in particular. Striking is that almost nine out of ten of all tweets including delegitimising media criticism come from just these two parties. These parties also have some MPs that are very active in directing delegitimising criticism toward the news media, where Lars Beckman, MP for the Moderate Party, is in a category for himself with his recurrent delegitimising media criticism. Fourth, the most common type of delegitimising media criticism is bias accusations, followed by criticism of the quality of the news media and their coverage; outright accusations of fake news are, however, quite rare. Fifth, and finally, the most common target is public service media, followed by news media in general.

In other words, what unfolds from this study is a clear pattern in Sweden where predominantly MPs from political right – and the Moderate Party Sweden Democrats in particular – engage in media criticism seeking to delegitimise public service media and traditional news media in general. Importantly, this pattern reflects findings from other countries (Egelhofer et al., 2021; Farhall et al., 2019; Wright, 2021) and suggests that delegitimising media criticism has become an important part of right-wing and populist political communication (Carlson et al., 2021; Hameleers, 2021). Also important is that these parties also want to cut down on funding and limit the mission of public service media, while the Sweden Democrats in addition have strong links to right-wing alternative media with their recurrent attacks on traditional news media (Figenschou & Ihlebaek, 2019, 2021). Hence, delegitimising media criticism from these parties’ MPs might be interpreted as a means towards influencing public opinion to accept more radical changes to public service and, in the case of the Sweden Democrats, to create greater space for and legitimise right-wing alternative media. As noted by the party leader, Jimmie Åkesson, the Sweden Democrats “have much to thank alternative media for [translation]” (tweet on 29 December 2020).

While these are important findings in themselves, they may also help to explain why research has found that trust in news media in general, and trust in public service media in particular, is lower among supporters of the Sweden Democrats and those leaning to the right ideologically, and that there is a trust gap between people with different political belongings. One theoretical explanation for this has been the hostile media phenomenon, according to which people tend to perceive the media as biased against their own party or side in a political conflict (Hansen & Kim, 2011; Perloff, 2015; Strömbäck & Johansson, 2017). However, this phenomenon does not fully explain why it is more common to perceive the news media as biased among supporters of some parties than among others or among right-leaning citizens. Another theoretically important and potential explanation might be that right-wing politicians have been more active and successful in shaping people's perceptions of and trust in the news media – not just on Twitter, as investigated here, but also on other platforms and in other contexts. From that perspective, lower trust in media among Sweden Democrats and people leaning to the right ideologically can be interpreted as an outcome of successful opinion formation on Twitter and other platforms. What also adds to this interpretation are findings that right-wing and populist political alternative media recurringly attack and seek to delegitimise traditional news media (Benkler et al., 2018; Ekman & Widholm, 2022; Figenschou & Ihlebaek, 2019, 2021; Holt, 2018).

As with any other study, some limitations should be noted. One such limitation is that the study only covers one year. Hence, it is unclear whether delegitimising media criticism has become more common or whether the prevalence of such media criticism differs between parliamentary terms. Future research is thus encouraged to follow up this study with a longitudinal design covering several terms. Another limitation is that we did not include the specific names of political alternative media in the search string, even though such tweets were captured by some of the other search terms. A third limitation is that this study only focused on delegitimising media criticism on Twitter. Hence, another avenue for future research would be to include more social media, such as Facebook, and compare the prevalence, targets, and expressions of media criticism across platforms. Future research should also go further in comparing delegitimising media criticism from politicians across countries.

Further research should also deepen the study of the importance of individual MPs in the context of delegitimising media criticism. As shown in this and earlier studies (Widholm & Mårtensson, 2018), some MPs are particularly active in delivering delegitimising media criticism, which raises important questions about their intentions and motivations, strategic reasoning, and, in the end, what they seek to accomplish.

These limitations notwithstanding, we believe this study represents an important step forward in understanding patterns of delegitimising media criticism and, in extension, how to understand the trust gap between supporters of different parties and people leaning to the ideological left versus the ideological right. These issues will likely not become less important in the future, as democracy and the legitimacy of the news media are increasingly challenged around the world.