The hostile media: politicians’ perceptions of coverage bias

Abstract

Politicians seem to be increasingly criticising the traditional news media for being biased. While scholars usually argue that politicians make such claims out of strategic concerns – they try to undermine the credibility of the potentially harmful media – it might as well be that they actually believe there is a bias in traditional news coverage. Though this so-called hostile media effect – the idea that news content is biased against one’s own ideas or party – is often studied with citizens, it has rarely been examined among politicians. However, in this paper it is studied, drawing on a unique survey in which 183 Belgian politicians were asked to what extent they perceived different media outlets to produce (un)favourable coverage about their party. The exploration shows that politicians, in general, have the tendency to perceive the news media as slightly biased against their party. Importantly, media hostility perceptions are more outspoken among politicians from right-wing parties and among politicians in high-level functions. Interestingly, politicians’ perceptions of partisan bias differ across outlets; especially the outlets that are used by non-party voters are considered to be biased.

Keywords:

• Media bias
• elite perceptions
• elite survey
• political communication
• political parallelism

Acknowledgements

We are grateful for the helpful comments and suggestions from the anonymous reviewers.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Note that another possible cause of elites’ media critique may be the fact that coverage is indeed biased. Politicians would then just be explicating the truly biased nature of the media. Empirical evidence with regard to media bias is inconclusive (see De Swert 2011). However, regardless of this objective reality, it is politicians’ perceptions that drive their behaviour.

2 Interviews were conducted within the framework of the POLPOP-project in Flanders, led by Stefaan Walgrave from the University of Antwerp (Belgium), with funding from the national science foundation FWO (grant number G012517N). Together with our colleagues Pauline Ketelaars, Kirsten Van Camp, and Julie Sevenans, we interviewed 183 politicians.

3 Proportionally as many politicians from each party were willing to participate. There is no self-selection bias.

4 This average perception of bias conceals differences in politicians’ ratings across outlets. Politicians do not uniformly evaluate all outlets as being biased against them or not; they rate some outlets significantly higher and lower than others. Most politicians (60%) rate some outlets as disadvantaging them (score > 5) and other outlets as advantaging them (score < 5). On average, the difference between the highest and lowest rating each politician gave, is 3 on an 11-point (un)favourability scale.

5 As a robustness check, we ran both a crossed-effects regression with random effects on the level of politicians and news outlets, and a regression with eight outlet dummies. Both models yield the same findings.

6 Eén (public broadcaster), De Morgen (newspaper), De Standaard (newspaper), are considered as the ‘quality media’; Het Nieuwsblad (newspaper), Gazet van Antwerpen (newspaper), Het Laatste Nieuws (newspaper), Belang van Limburg (newspaper), and VTM (commercial broadcaster) were the ‘popular media’ (see De Bens and Raeymaeckers 2007).

7 For this dummy variable, we rely on two questions asked in a citizen survey (N = 1,190) fielded in July 2019. Citizens were asked to place themselves on a left–right scale and to indicate the news outlet they primarily consume to inform themselves about public affairs. The differences in the ideological positioning of audiences are small but significant: De Morgen: left-wing audience (3.5 on a 10-point scale); De Standaard, Eén and Belang Van Limburg: center audience (5.2); Gazet Van Antwerpen, Het Laatste Nieuws, VTM and Nieuwsblad: right-wing audience (5.8). The variable political parallelism receives the value of 1 in the stacked dataset if leftist politicians rate De Morgen, if center politicians rate one of the outlets that are identified as ‘center’, and so on.

Additional information

Funding

This work was supported by the Flemish Fund for Scientific Research (FWO) [grant number G012517N].

Notes on contributors

Karolin Soontjens

Karolin Soontjens is a Doctoral Candidate financed by the FWO (grant number 11G8819N). She works at the research unit Media, Movements & Politics (M2P) in the Department of Political Science at the University of Antwerp. [[email protected]]

Annelien Van Remoortere

Annelien Van Remoortere is a Doctoral Candidate. She works at the research unit Media, Movements & Politics (M2P) in the Department of Political Science at the University of Antwerp. [[email protected]]

Stefaan Walgrave

Stefaan Walgrave is a full Professor at the Department of Political Science, University of Antwerp. He works at the research unit Media, Movements & Politics (M²P). [[email protected]]