Faculty of Social and Behavioural Sciences
Graduate School of Social Sciences
Department of Political Science Supervisor: Dr. Armèn Hakhverdian
21st June 2019
Master’s Thesis
The Impact of Populist Leadership on Media Freedom
Viola Koegst viola.koegst@alice.de
Student Number 12231010
M.Sc. Political Science - International Relations Research project: A global crisis of democracies? Change and continuity in 21st-century politics
Acknowledgement
Danke an meine Eltern, die mich seit fast 25 Jahren in allem was ich tue in der bestmöglichen Weise unterstützen und mich meinen Weg gehen lassen. Ohne euch hätte ich heute buchstäblich keine Masterarbeit abgeben können, ich hätte mir den Spaß nämlich gar nicht leisten können.
Abstract
In the past years, many incidents of populist leaders attacking the free media in their country verbally, structurally or directly have been reported. Although the topic is widely covered by the news, it has hardly been researched yet. In this paper, a thorough empirical quantitative analysis is used to shed light on the question of how strong the impact of populists in power on media freedom actually is. This is accompanied by a literature review on the possible incentives of populists to attack the media. The findings show that the outcomes of populism studies are often hard to compare as they are based on different definitions which can be misleading. However, drawing on the popular definition of populism by Cas Mudde (2004), I find that populist leadership is significantly negatively associated with media freedom with an average loss of 1.5 points on a 0-100 scale in the first year of populist rule and cumulative long-term effects of -9.8 points on average.
Table of contents
1. Introduction………...………..………….p. 2 2. Literature review and theoretical framework………....………..p. 5 2.1 Defining populism………....……..………..p. 5 2.2 Theoretical framework and hypothesis:
Why is populism a threat for press freedom?……….………..p. 7 3. Data and methods…………...………..……...p. 12
3.1 Dependent variables………....…….p. 12 3.2 Independent variables………..p. 14 3.3 Methods………...p. 18 4. Results and analysis………...p. 20 5. Conclusion………...……..p. 25 References…….……….……..……….………..……….…………....…...p. 27 Declaration of authorship……….…..………..p. 30 Appendix………...……..………….p. I
Overview of tables and figures
No. Title of table or figure Page No.
Figure 1 Comparison of the press freedom indices 14
Figure 2 Percentages of populist rule in the definition of Kyle/Gultchin and Kenny 16
Table 1 Descriptive statistics for all variables 19
Figure 3 Standard error bar charts for press freedom under populist and
non-populist rule 20
Table 2 Main results of the regression models 23
Index of abbreviations
FPÖ: Freiheitliche Partei Österreichs (Freedom Party of Austria) ORF: Österreichischer Rundfunk (Austrian broadcasting service) TV: Television
V-Dem: Varieties of Democracy, Institute at the University of Gothenburg CSO: Civil society organisation
GDP: Gross domestic product
2 1. Introduction
Austria has just recently, in May 2019, witnessed a government breakdown as former chancellor Sebastian Kurz had to end his coalition with the right-wing populist party FPÖ after a secretly recorded video showing the former FPÖ vice chancellor Heinz-Christian Strache during a meeting with an ostensible Russian oligarch was published. In the video which was recorded two years ago, shortly before the Austrian parliamentary elections, the decoy offers to support Strache’s party financially in return for construction contracts with the state. The video reveals him making plans with the “oligarch” to invest in the Austrian newspaper Kronen Zeitung and make it FPÖ-friendly by replacing critical journalists with pro-government ones. The same should happen with the national TV station ORF so that the FPÖ could get a “push” before the elections and gain higher turnouts. When the video was published a few weeks ago, Strache’s political career came to a sudden end as Kurz had to distance himself and his party from the scandal and dissolve the coalition.1 Although those plans made in 2017 have not been realized, in retrospect one can notice
a certain dangerous development for press freedom in Austria over the past two years after the populists came into power. There have been several scandals about verbal attacks of journalists by FPÖ-members. This is reflected in the study by Reporters without Borders, where Austria fell by five places in the world-wide ranking of press freedom since last year.2
The case of Austria is not a unique example of populists in power attacking the media. There are currently 18 countries in the world that are governed by populists3, some of which belong to the
most populous in the world. Donald Trump, populist president of the United States, is well-known for his anti-press statements. Trump has almost 60 million followers on Twitter who read his regularverbal attacks on the press on this platform, for example:
“The Press has never been more dishonest than it is today. Stories are written that have absolutely no basis in fact. The writers don’t even call asking for verification. They are totally out of control. Sadly, I kept many of them in business. In six years, they all go BUST!” 4
04:20 - 20. Feb. 2019
Statements like this have become typical for the US-president and are not even surprising anymore. He has repeatedly called the press the “enemy of the people” and the “real opposition party” who
1 http://time.com/5592347/austria-government-video-scandal/, accessed 30.05.2019;
https://www.washingtonpost.com/world/2019/05/20/how-boozy-video-confirmed-critics-views-austrian-far-right-triggered-government-break-up/, accessed 30.05.2019.
2 https://rsf.org/en/austria, accessed 30.05.2019.
3 Kyle/Gultchin (2018): Populists in power around the world.
https://institute.global/sites/default/files/articles/Populists-in-Power-Around-the-World-.pdf. p. 28-31. 4 https://twitter.com/realDonaldTrump/status/1098195686540169223, accessed 26.03.2019.
3
deliberately spread “fake news” among the people and act dishonest and misleading.5 Getting used
to populist anti-media rhetoric, however, does not solve the question of how dangerous populists can really become for media freedom.
In addition, this is not only a Western phenomenon. Populists all over the world put free media at risk in their countries. The president of the Philippines, Rodrigo Duterte, has been threatening journalists since he came into office and regularly calls them things like “garbage”. The media have increasingly become government-controlled and the Philippines are currently one of the countries with the highest rates of unnatural deaths of journalists in the world.6 Nicolás Maduro, until recently
president of Venezuela, stated last year that because of his policies, “only the debris of the bourgeois media is left.” Indeed, in 2017 the regime shut down 40 radio stations and the number of newspapers in the country has decreased by three quarters since 2013.7 Turkey has imprisoned
more journalists over the past three years than any other country.8 President Recep Tayyip Erdogan
has justified those human right violations with statements like “journalists are the gardeners of terrorists” 9 and that they would nurture terrorism through their writing until at some point they
would turn into terrorists themselves.10
These are only a few out of numerous examples that show hostile behaviour of populists against the media. But are these attacks only bully-boy tactics or do they really put the free work of the media into danger? In how far do populists in power really threaten the free press in their country? This is the subject of this paper which I base on the research question:
To what extent does populist leadership have an impact on media freedom?
As we have seen, this is a very current topic of high societal relevance that we often get confronted with in the daily news. However, the link between media freedom and populist rule has not been made very often in academic literature yet. But some very recent research shows that populist rule is associated with a decline in press freedom. Kenny (2019) conducted a study on this topic using his own dataset from 2017 where he coded all populists in power from 1980-2010 and that he extended until 2014 for this purpose. Kenny finds that populist rule has a negative impact on all
5 https://twitter.com/realDonaldTrump/status/1082268365081767936, accessed 08.06.2019. 6 https://thediplomat.com/2018/09/dutertes-media-war-in-the-philippines/, accessed 08.06.2019. 7 https://www.cjr.org/the_media_today/venezuela_crisis_maduro_trump.php, accessed 08.06.2019. 8 https://abcnews.go.com/International/3rd-straight-year-turkey-jailed-journalists-country-report/story?id=59791362, accessed 08.06.2019. 9 https://www.dw.com/en/erdogan-in-paris-journalists-are-gardeners-of-terrorism/a-42037145, accessed 08.06.2019. 10 Ibid.: accessed 08.06.2019.
4
the examined indicators of press freedom and freedom of expression.11 Although this study is
important as one of the first ones in this field, it is characterized by Kenny’s rather unconventional approach to populism and is therefore not directly comparable with other studies. He uses his organizational definition of populism. Apart from this, one could also critique the restrained use of control variables on the effect of populism on media freedom.12 This may leave important
factors out that have an impact on media freedom which could distort the analysis.
Ruth-Lovell et al. recently published a broader study on the implications of populist rule where they also looked at the press freedom under populist governments. They coded populism based on the average degree of populist discourse of political leaders over their term. Their findings show a strong negative effect of populist rule on media freedom.13 According to their research, at the end
of a term under populist leadership, the average press freedom index is almost two points lower (on a 0-100 low-to-high scale) than the average press freedom index at the beginning of a term.14
The limits of their research lie in the relatively small dataset of only 40 countries and their focus on the 21st century.15 The scope of the analysis of press freedom is even smaller. It is only based on
159 cases up until the year 2016.16
Kyle/Mounck (2018)17 have examined the influence of populist rule on media freedom based on
the list of cases of populists in power around the world by Kyle/Gultchin (2018)18 that I too use
for part of my analyses in this paper. They looked at cases between the years 1990 and 2014 and their findings show a decline of press freedom by 7% when a country is ruled by populists.19
These three studies are to my knowledge the first ones that specifically examine the impact of populists in power on media freedom. They have contributed to the overall understanding of populism and opened a new field of research. However, they have also left some questions open and the topic still needs more elaboration and requires more detailed research. Also, it is interesting that their results differ quite strongly in their strength which is presumably due to the different methodologies that were used. For these reasons, this thesis takes their research further and
11 Kenny, P. (2019): “The Enemy of the People”: Populists and Press Freedom. Political Research Quarterly 1/25/2019. p. 1–15, here: p. 9.
12 Ibid.: p. 7-8.
13 Ruth-Lovell et al. (2019): Consequences of populism memo for the Guardian’s The New Populism project. http://populism.byu.edu/App_Data/Publications/TP_Consequences_Memo.pdf, p. 16-17.
14 Ibid.: p. 21. 15 Ibid.: p. 2. 16 Ibid.: p. 16.
17 Kyle/Mounck (2018): The populist harm to democracy: An empirical assessment. http://institute.global/insight/ renewing-centre/populist-harm-democracy. p. 22, 36-37.
18 Kyle/Gultchin (2018): p. 28-31. 19 Kyle/Mounck (2018): p. 22, 36-37.
5
expands it by using two indicators for press freedom as well as two different definitions of populists in power to test for robustness. Kyle/Gultchins’ (2018) conceptualization of populism is close to the one by Mudde (2004). To make the project more comprehensive, I test my assumption based on Kenny’s definition of populism, too. In this paper, a quantitative empirical analysis is used to determine the influence of populists in power on media freedom. All cases of populists in power from 1990-2018 that fall into one of these definitions are taken into account. In total, I use 1127 cases from 39 countries in my analyses. This makes the study more extensive than the previous ones and provides deeper and more detailed insight into a very current yet hardly researched topic and is therefore of value for both academia and society.
This paper proceeds as follows. I first elaborate on the different possible definitions of populism which is essential for the rest of this work. The second part of the thesis deals with possible reasons why populists act hostile against the media. I draw an extensive theoretical framework and deduct the hypothesis to be tested from it. The next part is on the empirical analysis in which I research how strong the impact of populists in power on media freedom actually is. I explain the data and methods I use before I go over to the detailed discussion and analysis of the results. The paper ends with a conclusion.
2. Literature review and theoretical framework
This section contains a literature review on the implications of populists in government and the theoretical framework for the conducted analysis. The first part deals with the definition of populism. I then elaborate on why populist leadership can be hostile to press freedom and deduct the hypothesis to be tested from this.
2.1 Defining populism
Populism itself is a highly contested phenomenon. It seems to be hard to grasp and conceptualize which results in numerous competing definitions in academic literature. Probably prevailing among scholars at the moment is the definition that was introduced by Cas Mudde in 2004. He sees populism as:
“[…] an ideology that considers society to be ultimately separated into two homogeneous and antagonistic groups, ‘the pure people’ versus ‘the corrupt elite’, and which argues that politics should be an expression of the volonté générale (general will) of the people.” 20
6
Populism in his understanding is moreover a “thin-centred ideology” which means it does not provide a fully developed model to live after but can be applied to many different aspects.21 This
definition has found many followers among scholars who either applied it directly to their studies [e.g. Rooduijn (2013)22, Batory (2016)23, Huber/Schimpf (2016)24] or used it in a slightly modified
way such as in Kyle/Gultchin (2018)25, Engesser et. al. (2017)26 and Haller/Holdt (2018)27.
Although this ideational approach has been widely accepted, there is still a number of other interesting conceptualizations in the academic discourse.
Another concept is based on the assumption that populism is a political movement that is characterized by a strong, charismatic leader and supported by a heterogeneous group of people.28
Roberts (2007) uses one form of this organizational definition. For him, populism is:
“[…] the top-down political mobilization of mass constituencies by personalistic leaders who challenge elite groups on behalf of an ill-defined pueblo, or ‘the people’.” 29
Kenny (2019) also follows this definition of populism in his study on which this thesis is partly based on.30 His work will be discussed further in the following sections.
Some scholars have come up with their own interpretations of populism that do not fit into the major conceptualizations. De Vreese et al. (2018) define populism as a communication phenomenon. They combine the ideological factor with a distinctly populist communication style. For them, populism arises in the interaction and communication of politicians, the media and citizens.31 Interesting about this approach is that it takes the reciprocal relationship between
populism and the media into account. Inglehart/Norris (2016) add a cultural dimension to Mudde’s
21 Ibid.: p. 544.
22 Rooduijn, M. (2013): A populist Zeitgeist? The impact of populism on parties, media and the public in Western Europe. University of Amsterdam. p. 5-6.
23 Batory, A. (2016): Populists in government? Hungary's “system of national cooperation”. Democratization, 23(2). p. 283-303, here: p. 284.
24 Huber/Schimpf (2016): Friend or Foe? Testing the Influence of Populism on Democratic Quality in Latin America. Political Studies Vol. 64(4), p. 872-889, here: p. 873.
25 Kyle/Gultchin (2018): p. 11-12.
26 Engesser et al. (2017): Populism and social media: how politicians spread a fragmented ideology. Information, communication and society 20(8), p. 1109-1126, here: p. 1111-1113.
27 Haller/Holt (2018): Paradoxical populism: how PEGIDA relates to mainstream and alternative media, Information, Communication & Society, DOI: 10.1080/1369118X.2018.1449882. p. 3.
28 Rooduijn (2013): p. 5.
29 Roberts, K. (2007): Latin America's Populist Revival. SAIS Review 27(1). p. 3-15, here: p. 5. 30 Kenny (2019): p. 9.
31 De Vreese et al. (2018): Populism as an Expression of Political Communication Content and Style: A New Perspective. The International Journal of Press/Politics 2018, Vol. 23(4), p. 423 –438, here: p. 425-427.
7
definition. Populist nativist values are seen as the opposite of cosmopolitan liberal values in their work.32
The various conceptualizations of populism contribute to a vital and active discourse on the topic. While this is interesting and fruitful, it also makes it hard to generalize the outcomes of a particular study. Populism remains a highly contested phenomenon that is hard to measure. Because scholars have not found a universal definition over decades of research, it makes it particularly important to use the term carefully and make clear which definition is used. Depending on the definition of populism, studies sometimes vary strongly in their choice of cases as well as results. This is an intrinsic difficulty of the study of populism. This thesis tries to bridge this issue by using both two very different definitions of populism and two indices for media freedom to test the hypothesis on. The focus lies, however, on the main definition by Mudde and his followers.
2.2 Theoretical framework and hypothesis: Why is populism a threat for press freedom?
In this section, possible reasons for the presumed negative effect of populist leadership on press freedom as well as other implications of populist rule are discussed, based on existing literature. First, it is agreed upon among scholars that populists in power have different kinds of effects on the democratic quality of a country. Populists could represent a threat for the separation of powers and the system of checks and balances as they are more likely to try to undermine it than other politicians.33 These institutional changes are often associated with democratic erosion as they
influence other factors like election quality or liberal democratic values.34 Kyle/Mounck (2018) find
that 23 percent of populist governments have caused significant democratic backsliding and more than half of populist leaders make changes or amendments to their constitutions, often leading to less checks on the executive or extensions of leadership terms.35 In the sense that populists typically
would put aside legal limitations for their own political gain in the name of the “will of the people”, populism can therefore even be seen as an “antithesis of constitutionalism”36 which makes the
anti-democratic tendency of populists clear. Though, when looking at this phenomenon, one has to bear in mind that the dynamic is often more complex. In many cases, the liberal democratic
32 Inglehart/Norris (2016): Trump, Brexit, and the Rise of Populism: Economic Have-Nots and Cultural Backlash. Harvard Kennedy School Faculty Research Working Paper Series. Available at: http://ssrn.com/abstract=2818659. p. 6-7. 33 Huber/Schimpf (2016): p. 875-876; Kenny (2019): p.1. 34 Ruth-Lovell et al. (2019): p.1. 35 Kyle/Mounck (2018): p. 4. 36 Batory (2016): p. 284.
8
institutions of the regarding state have not been fully functioning for a while by the time populists came into power.37
In general, democracies score higher in media freedom than autocracies.38 At the same time,
authoritarian tendencies are a core feature of the populist ideology. Populists justify their own will to power with the idea of representing the “will of the people” which is thought to give them the right to lead the people in a strong and charismatic manner. Also, populists often show a preference for forms of direct democracy like polls and referenda rather than a representative democracy with strong institutions.39 This authoritarian mindset that is often displayed by populists enhances the
risk of democratic erosion when populists come into power.40
Democratic backsliding and restrictions of media freedom go hand in hand. The press and democracy are intertwined and cannot be completely separated from one another.41 Citizens are
dependent on press freedom and sufficient media reach in order to be able to make informed decisions, express their will in elections and hold their leaders accountable for their actions. A free press is therefore directly associated with the citizens’ right of political participation.42 Democracy
and media freedom hence constitute and influence each other. Free media can be seen as a part of the link between political institutions and citizens and is in this way a defining determinant for democracy.43 So if populists in power threaten democratic quality, they put media freedom at risk,
too.
But a rise of populism does not only have these broad, multi-layered implications for democracy. Populists also often attack the media directly and use hostile rhetoric against them. So what about the media scares them? Why do they work against them and why do they try to build up an enemy stereotype of the free media?
To understand this, one has to make the link between the intrinsic anti-pluralist and anti-elitist attitudes of populists and the media. For Mudde (2004), elitism and pluralism are the two exact
37 Kyle/Gultchin (2018): p. 5.
38 Stier, S. (2015): Democracy, autocracy and the news: the impact of regime type on media freedom, Democratization, 22(7), p. 1273-1295, here: p. 1273.
39 Inglehart/Norris (2016): p.7.
40 Wilson, G. (2017): Brexit, Trump and the special relationship. The British Journal of Politics and International Relations 19(3), p. 543 –557, here: p. 547-548.
41 Waisbord, S. (2012): Democracy, journalism, and Latin American populism. Journalism 14(4), p. 504-521, here: p. 505.
42 Dutta/Roy (2016): The interactive impact of press freedom and media reach on corruption. Economic Modelling 58(2016), p. 227-236, here: p. 228;
Kyle/Mounck (2018): p. 21.
43 Nisbet, Erik/Stoycheff, Elizabeth (2011): Let the People Speak: A Multilevel Model of Supply and Demand for Press Freedom. Communication Research 40(5), p. 720 –741, here: p. 733, 736.
9
opposites of populism.44 Therefore, they are always going to be opposed to them. The media, on
the other hand, are an incorporation of pluralism. For populists, there is one clear answer to who should be representing the people in politics and what policies are the best for them. The media, in contrast, encourage the discourse between political actors, civil society groups and others. 45 They
hence widen the room to debate about policies and politicians which is typically perceived by populists as a threat to their power and the “will of the people”. When citizens are free to choose to inform themselves from a large number of sources representing different political orientations, this will result in a more pluralist society with a wide variety of political ideas. This stands in contrast to the populist assumption that there is one universal “will of the people”. In populist ideology, there is no room for minority views.46
Also, the media can be seen as an elite actor and it fits into the populist ideology to interpret them as such. The populist understanding of the media is that they are instruments of the elites and can never be autonomous. Populists see journalism and press ownership as the same thing, reporters are only figures in the overall legal, political and economic environment of the news organizations for them. The media are seen as a way of expression of the establishment. Consequently, in this understanding, the media can never really be free.47 This view is certainly deeply influenced by
ideology. However, there is actual evidence that the established mainstream news media have an elitist bias in most countries. They have a tendency to cover mostly topics of interest for the ruling classes of politics, culture and the economy, even when they criticize them. Usually, they downplay or combat populism in their publications which adds to the populist anti-media sentiment.48
Another typical populist perception is that mainstream media as elite actors have no interest in publishing information that does not fit into an established “politically correct” agenda. They, therefore, accuse the media to withhold information or distort the news in order to serve the interests of the elites.49 Populists on the other hand often stand for the unconventional and the
“truth” that nobody else dares to address. So another point of critique of the media by populists is often that due to their bias, they don’t sufficiently cover the failures and delinquencies of the elites.50
44 Mudde (2004): p. 543. 45 Kyle/Gultchin (2018): p. 18.
46 Goldstein, A. (2018): The Press and Classical Populism in Argentina and Brazil. Latin American Perspectives 220 (45) No. 3, p. 109-123, here: p. 111.
47 Waisbord (2012): p. 510.
48 Mazzoleni, G. (2008): Populism and the Media. In: Albertazzi/McDonnell [ed.]: Twenty-First Century Populism. The Spectre of Western European Democracy, p. 49-64, here: p. 51.
49 Haller/Holdt (2018): p. 1.
50 Krämer, B. (2018): How Journalism Responds to Right-Wing Populist Criticism. The “Lying Press” Attack and the “No Censorship” or “No Ammunition” Defense. In: Otto/Köhler [ed.]: Trust in Media and Journalism: Empirical Perspectives on Ethics Norms, Impacts and Populism in Europe, p. 137-154, here: p. 142.
10
The media naturally act to some degree as a mediator between institutions and the public. Even if produced with the highest objectivity and accuracy, a medium remains a secondary source of information. Populists make use of this fact for their ideology by claiming that the media cannot depict the “real truth” that is known among the people and that the media could always systematically distort information or directly lie to the people.51 Another factor that plays into the
notion of the media as mediators is that the political field is actually to some extent defined by the media. By showing who or what the main actors, policies, politicians, ideas, and cleavages are, the media do have an impact on public opinion and discourse which reflects on political outcomes.52
What the media publish has its roots in structural factors of society. They take part in social, cultural and political processes and can to some degree be seen as “mirrors of society”53.
When we study the connection between populism and free media, we have to be aware that the two stand in a complex reciprocal relationship with each other.54 Populists are dependent on the
publicity they receive through news coverage and mass media. At the same time, the way populists are portrayed in the media is usually not of affirmative nature.55 Independent media sometimes
help populists to gain popularity by reporting critically about past governments. Once they are in power, the media reports typically get more opposed to them.56 Populists make use of the media
as a mobilization tool. Populism therefore also works through the media as the press creates the conditions for mass mobilization. It is key to understand that the media can be turned into allies of populist leaders, sometimes without them realizing.57 This can happen to all kinds of press.
Against a common notion, popular newspapers and quality newspapers do not differ regarding their level of sympathy towards populist parties.58 The complex relation between populism and the
media is rooted in their ideology. Because they stand for popular empowerment, they often criticize the concentration of media corporation ownership. On the other hand, they also see themselves as leaders of the “true people” which is connected to the notion of top-down centralized power and communication.59 51 Kyle/Gultchin (2018): p. 18. 52 Krämer (2018): p. 149. 53 Mazzoleni (2008): p. 64. 54 Ibid.: p. 62. 55 De Vreese et al. (2018): p. 428. 56 Ruth-Lovell et al. (2019): p. 16. 57 Mazzoleni (2008): p. 50; De Vreese et al. (2018): p. 428-429.
58 Akkerman, T. (2011): Friend or foe? Right-wing populism and the popular press in Britain and the Netherlands. Journalism 12(8), p. 931-945, here: p. 942.
11
Populist parties and politicians have developed numerous strategies to deal with and make use of the media. The press is a central communication tool for populists. Populists believe that only state-influenced journalism under populist rule can be truly democratic because populists stand for the people.60 This explains their tendency to try to influence journalists and pull media
organizations over to their side. This often happens through funding or even corruption.61
Social media have become a big part of most politicians’ communication but especially of the populist media strategy. They provide a much more direct connection to the citizens and are hence a big chance for them. Because social media represent a communication space without any mediation by other actors like the press, the Internet is even thought to have the potential to change the political power structure in a more populist direction. The logic of professional mass media outlets and the network-structured social media differ substantially, which is why populists typically use a hybrid communication strategy. While official statements that address a larger audience are usually made on the classic media, social media are used for more polarizing or radical topics.62 As
“unmediated media”, social media give politicians the opportunity to portray themselves exactly the way they want to be viewed.63 Also, the preference for social media can be explained by the
suspiciousness against mainstream media outlets of populists and large parts of their (potential) voters. It is part of their strategy to claim that the mainstream media hide information from the public or intentionally lie to the people because this affirms the voters in their media distrust and at the same time makes the populists appear as the only ones who tell “the truth”. By speaking directly to the people they may appear more honest and reliable.64 The term “lying press” has
become prevalent in populist rhetoric. Claiming that the traditional media lie to the people and that they are the only ones who can be trusted is a populist strategy to convince voters of their ideas. This can substantially hurt the established media as they rely on the people’s trust in them as news providers.65 On top of that, social media have become very important and influential in
agenda-setting. If a topic becomes viral on the Internet, it is very likely that it will be picked up by both the
60 Waisbord (2012): p. 516.
61 Samet, R. (2017): The Denouncers: Populism and the Press in Venezuela. Journal of Latin American Studies 49, p. 1-27, here: p. 18.
62 Engesser et al. (2017): p. 1113.
63 Groshek/Engelbert (2012): Double differentiation in a cross-national comparison of populist political movements and online media uses in the United States and the Netherlands. New Media & Society 15(2), p. 183-202, here: p. 184.
64 Haller/Holdt (2018): p. 1, 5.
65 Beiler/Kiesler (2018): “Lügenpresse! Lying press!“ A content analysis study of the bias of journalistic coverage about ‘Pegida’, the movement behind this accusation. In: Otto/Köhler [ed.]: Trust in Media and Journalism: Empirical Perspectives on Ethics Norms, Impacts and Populism in Europe, p. 155-179, here: p. 159-160.
12
mainstream media and policy-makers. This way, populists can bypass the established media and potentially influence their agenda at the same time.66
The media function as a watchdog for the government. They help to provide transparency and accountability.67 Within the populist ideology, there is no space for the idea that the media should
be sceptical of governments.68 If one criticizes a populist government, one is automatically wrong
and against the true will of the people in their view. It is therefore typical for populists in power to try to erode this watchdog function of the media by restricting press freedom.
All these factors explain why and how populist leadership can be dangerous for media freedom. In the following, this theoretical assumption will be tested empirically. So the hypothesis to be tested in this paper is:
Populist leadership has a negative impact on media freedom.
3. Data and methods
Based on the assumptions of section 2, this section deals with the data and methods used to conduct the quantitative empirical analysis for this paper.
3.1 Dependent variables
In order to gain results of higher validity for the analysis conducted for this research, I use two indices for press freedom as dependent variables. They differ slightly in their focus and included variables. I do two sets of analyses using the two indicators. This makes it more likely to depict the real relationship accurately in the model. One of the dependent variables used is the “freedom of the press index” provided by Freedom House in the version of 2017 (freedomhouse_freedompress) that covers the years up until 2016.69 Freedom House evaluates the
degree of broadcast, print, and digital media freedom in a country and takes the legal, economic and political environment of the media into account. Political pressure may influence reporting and economic factors can have an effect on the access to information.70 Each of the 199 included
countries and territories receives a score between 0-100 with 0-30 meaning “free”, 31-60 meaning “partly free” and 61-100 meaning “not free”. This study covers the years 1990-2018. Because Freedom House only evaluated the press freedom of countries in the three categories “free”,
66 De Vreese et al. (2018): p. 429.
67 Majhosev et al. (2018): European model of regulating the media. Balkan Social Science Review 12, p. 127-141, here: p. 129.
68 Waisbord (2012): p. 511.
69 https://freedomhouse.org/report/freedom-press/freedom-press-2017, accessed 16.05.2019. 70 https://freedomhouse.org/report-types/freedom-press, accessed 16.05.2019.
13
“partly free” and “not free” until the year 1993, the years 1990-1992 are not included in the analysis with this index as the dependent variable. So, only the years that received a numeric score, that is 1993-2016, are part of the analysis.
The second index for press freedom that I base my analyses on is the V-Dem index for “freedom of expression and alternative sources of information” (freexp_altinf) as in the “Country-Year: V-Dem” dataset in the version of 2019 which covers the years 1789-2018.71 This index includes the
variables for media censorship effort, media self-censorship, media bias, harassment of journalists, commonness of government-critical print and broadcast media, representation of a wide range of political perspectives in the major print and broadcast media as well as freedom of discussion for men and women and freedom of academic and cultural expression.72 Countries receive a score
between 0-1 with 0 meaning “no press freedom” and 1 meaning the “highest possible press freedom”. In this case, data is available for all years that are examined in this study. Using both indices, therefore, allows me to gain results for the most recent years and the early 1990s, with the largest amount of years of the analysis still covered by two indices for better validity. Also, as the included variables of the two indices indicate, the projects have chosen a slightly different focus for their definition of press freedom. Because the two indicators contain different variables, the use of both makes sure that the results are not sensitive to one index. The scatterplot below shows that the two indicators do correlate but are far from giving the same score per country-year case. The correlation between the two indices is -0.792 and therefore only on a medium level, taking into account that they are meant to measure the same. The negative correlation is a result of the converse scales.
71 Varieties of Democracy (2019): V-Dem [Country-Year] Dataset v9.
https://www.v-dem.net/en/data/data-version-9/.
14
Figure 1: Comparison of the press freedom indices 3.2 Independent variables
The main independent variable in this study is populist leadership. Again, to increase validity and to be able to conduct a broader and more complex analysis, I base my study on two different conceptualizations of populism. Kyle/Gultchin (2018) have listed all cases of populists in power around the world from 1990 onwards that are covered by their definition (see: appendix I a).73
They define populism as follows:
“[…] populism is the combination of two claims: the people are locked into conflict with outsiders; and nothing should constrain the will of the true people. Populism can be identified according to the prevalence of these two claims.” 74
This definition is certainly similar to the one by Mudde (2004)75 but leaves out the ideological factor.
With this minimal definition, they were able to apply the concept of populism on a wide range of different contexts and countries.76 The list contains all populist leaders and parties that reached an
executive office - that is the prime ministership, presidency or the equivalent - in the years
73 Kyle/Gultchin (2018): p. 28-31. 74 Ibid.: p. 19. 75 Mudde (2004): p. 544. 76 Kyle/Gultchin (2018): p. 19. 30 100 0 80 60 40 20 1 0.8 0.6 0.4 0.2
Press freedom – Freedom House
Pr es s fr eed om – V -De m
15
1990-2018. The fact that both parties and individual politicians are included helps to balance out a possible regional bias as in parliamentary systems typically parties have higher chances while presidential systems favour individuals. Populists who did not reach the highest executive office but were in a government office as minority coalition partners are not included in the database. Also, only cases in at least minimally democratic countries were taken into account which leaves out all cases of populism in power in autocracies or semi-democracies. There is no populism score used in the list which makes “populists in power” a dichotomous dummy variable in my analysis (Populism_KyleGultchin). However, the degree of populism may have changed over time in some cases. This global database on populist leadership is the first of its kind. It represents a helpful tool for my analysis despite its limitations.77 In total, the list contains 45 cases of populists in power in
33 countries.78
The second list of populists in power that I use is by Kenny (2019).79 It differs substantially from
the list by Kyle/Gultchin (2018) as it is based on the organizational definition of populism. Populism in this study is coded by the following criteria: first, whether a party leader used the party as a vehicle for his or her own power or substantially removed constraints to his or her power once in lead and second, whether the party primarily relied on the charismatic appeal of its leader to mobilize the masses and especially swing voters.80 Also, here the campaign prior to the party or
leader taking office is used as a basis for the coding as populists. This is done to avoid the problem of endogeneity or possible changes in the degree of populism over time. Hence, populists in power in the definition by Kenny is a dummy variable, too (Populism_Kenny). In this list, junior coalition members are included but this applies to only three cases (Poland 2005-2007, Austria 2000-2002, Romania 1993-1994). In total, 36 cases in 26 countries are coded as populists in power by Kenny (see: appendix I b).81 Because their understanding of populism and coding is so different, I expect
different results based on the two populism variables.
The inherent conceptual difference between the two variables for populism reflects on their influence on the empirical data. First, it is interesting to have a look at the degree of correlation between the two populism variables which indicates a medium level of correlation between the two variables with a value of 0.421. Keeping in mind that they are meant to measure the same
77 Ibid.: p. 26-27. 78 Ibid.: p. 19.
79 Kenny, P. (2019): “The Enemy of the People”: Populists and Press Freedom. Political Research Quarterly 1/25/2019: Online Appendix, Table A1.
80 Kenny (2019): p. 6.
16
phenomenon, this correlation is actually very low, meaning that they are not really comparable as one would assume at first sight.
Because they follow two very different definitions, the cases that are coded as populists in power according to the two lists vary strongly in numbers and cases. As visualized by the following bar charts, the percentage of country-year cases defined as populist rule by Kyle/Gultchin is almost twice as high as the one by Kenny, although populist minority coalition partners are included in Kenny’s list.
Figure 2: Percentages of populist rule in the definition of Kyle/Gultchin and Kenny
I include a range of control variables in my analyses. These variables are as well taken from the V-Dem Dataset of 2019. This provides good comparability as most of them are coded in the same way as their “freedom of expression and alternative sources of information index” which is in an interval from low to high (0-1). Among these are the following: the “presidentialism index” measures to what extent the regime is characterized by “systemic concentration of political power in the hands of one individual who resists delegating all but the most trivial decision-making tasks”82. I include this variable based on the assumption that the system of government has an
impact on both the likelihood of populists coming into power and the risk of an erosion of the freedom of the press. Parliamentary systems are in general more stable and resilient to crises than presidential systems.83 Also, it usually takes populists much longer to gain popularity and come into
power in parliamentary systems. The example of Donald Trump shows us that a populist can reach the presidency in his first run for office. In contrast, in parliamentary systems, populists are often dependent on another party that is willing to form a coalition with them, even if they win the largest amount of seats in the parliament.84 Because there is more executive power in only one person’s
82 Coppedge et al. (2019): p. 262.
83 Stepan/Skach (1993): Constitutional frameworks and democratic consolidation. Parliamentarianism versus Presidentialism. World Politics 46, p. 1-22, here: p. 19.
84 Kyle/Gultchin (2018): p. 38. 100 100 60 0 0 80 20 20 80 40 0 80 40 0 80 60 67.4% 32.6% 82.2% 17.8%
17
hands in a presidential system, populist presidents have typically been implementing anti-free press strategies and policies very quickly and on their own behalf.85
Furthermore, I expect an interdependence of populism, political corruption, and press freedom. I, therefore, control for “regime corruption” which indicates to what extent political actors make use of political office for private or political gain.86 Ruth-Lovell et al. (2019) find that populism is
negatively associated with the control of corruption.87 In some cases, populists in power use
corruption to gain control over the media.88 On the flipside, press freedom is supposed to help to
curb government corruption because it promotes transparency.89
Moreover, it is a common notion that populist leadership is a regional phenomenon and appears more frequently in some areas of the world, especially in Latin America. Although this might be true, populism can definitely not only be found in Latin America.90 One reason why it may be easier
for populists to come into office in Latin America could be the prevalence of presidential systems there compared to other parts of the world such as Western Europe.91 In Eastern Europe and
Post-Soviet Eurasia, populism has steadily been strong, while Asia is experiencing a rise of populism at the moment with currently about 40 percent of its population under populist rule.92 The control
variable “region (geographic)” assigns each country to one out of 19 geographic regions of the world.93
The variable “political civil liberties” is an indicator of freedom of expression and freedom of association. It is formed out of the indices for harassment of journalists, government censorship effort of the media, media self-censorship, freedom of academic and cultural expression, party ban, autonomy of opposition parties, barriers to parties, freedom of discussion for men and women, CSO entry and exit and CSO repression.94 The variable “institutionalized democracy” evaluates
states on the basis of the “presence of institutions and procedures through which citizens can express effective preferences about alternative policies and leaders, […] the existence of institutionalized constraints on the exercise of power by the executive” and “[…] the guarantee of
85 Waisbord (2012): p. 515;
Andersen Jones/West (2017): The fragility of the free American press. Northwestern University Law Review 112(3): p. 567-596, here: p. 569.
86 Coppedge et al. (2019): p. 262. 87 Ruth-Lovell et al. (2019): p. 10. 88 Samet (2017): p. 17.
89 Dutta/Roy (2016): p. 228.
90 Rode/Revuelta (2015): The Wild Bunch! An empirical note on populism and economic institutions. Econ Gov 16, p. 73-96, here: p. 79.
91 Kyle/Gultchin (2018): p. 38. 92 Ibid.: p. 37-43.
93 Coppedge et al. (2019): p. 324. 94 Ibid.: p. 263-264.
18
civil liberties to all citizens in their daily lives and in acts of political participation.”95 The “clean
elections index” is aggregated from indicators that measure registration fraud, election violence, vote buying, systematic irregularities and government intimidation of the opposition.96
The remaining control variables like GDP per capita and the annual inflation rate97 are measured
in absolute numbers. The average life expectancy 98 and average education of citizens over the age
of 1599 are measured in years. The infant mortality rate stands for “the number of deaths prior to
age 1 per 1000 live births in a year.”100 These indicators were included to gain a broader picture of
wellbeing and prosperity in the regarding country. The variable “population” represents the total population in numbers.101 I include it as an additional indicator for the general circumstances of a
country. Because the original distributions of the variables “population” and “inflation” are strongly skewed, I use their logarithms in my regressions (logpop, loginf). The variables show no multicollinearity.
3.3 Methods
I conduct my quantitative empirical analysis via a set of regressions with added layers of complexity. I run the model four times, for all combinations of press freedom variables and populism variables to gain results of high validity. I begin with a comparison of the mean values of the main variables and a simple bivariate regression between the regarding populism and media freedom indicator to gain a first overview. A multivariate regression including the control variables then delivers more nuanced results. However, this pooled model ignores the time-series dynamic of the panel dataset as it only analyses the effect of populists in power on media freedom across countries, not within countries. Therefore, I further refine the model by doing a time-series regression. The use of one-year lagged variables allows us to draw conclusions about the causality of the effect as it takes the change of the coefficients over time into account. I then add country-fixed effects to be able to identify the change of media freedom when populists come into power not only between but also within a specific country and year. In the next step, region-fixed effects are introduced. This gives information about whether the developments are specific for a region or not. The last specification I add are clustered standard errors which are taken into account due to the fact that the observations are not independent of each other but show the development in a country over time.
95 Ibid.: p. 319-320. 96 Ibid.: p. 44. 97 Ibid.: p. 326. 98 Ibid.: p. 329. 99 Ibid.: p. 323. 100 Ibid.: p. 329. 101 Ibid.: p. 329.
19
I also calculate the long-term effects of populist rule on media freedom since the year populists come into power also has an impact on the lagged media freedom variable for the following years. Making use of all these model specifications in combination with the double media freedom and populism variables delivers highly reliable and valid results.
I rely on the existing lists of cases of populism in power and the databases for press freedom and the control variables provided by V-Dem and Freedom House to be able to limit this research project to a feasible scope. This research, although it has its limitations, builds up on existing studies and provides a more refined and complete model and unit of analysis than the former research projects. The following table shows the descriptive statistics of all variables that are used in my regression models.
Variable N Minimum Maximum Mean Deviation Std.
Populists in Power -Kyle/Gultchin 1127 0 1 .33 .469 Populists in Power - Kenny 971 0 1 .21 .405 Press freedom – Freedom House 925 12 93 42.32 17.571
Press freedom – V-Dem 1127 .20 .98 .7951 .17607
Presidentialism index 1127 .03 .97 .2871 .24089
Clean elections index 1127 .00 .98 .7131 .22277
Political civil liberties index 1127 .12 .99 .8149 .16516
Regime corruption 1127 .03 .96 .4818 .24814 Institutionalized democracy 1083 -88.00 10.00 7.2493 5.61394 Region (geographic) 1127 1.00 18.00 10.8199 5.86470 GDP per capita 1022 961.00 53015.00 14069.3875 11241.39922 Inflation 748 -9.50 7481.69 58.8596 363.01599 Inflation (log) 727 -3.218876 8.920214 2.202456 1.45376 Education 15+ 976 3.25 13.61 8.6992 2.32552 Life expectancy 1098 43.90 84.20 72.5049 6.44662
Infant mortality rate 984 2.00 113.00 23.8716 21.60269
Population 1060 407472.00 1339180127.00 77994501.5000 184580706.65055
Population (log) 1060 12.91773 21.01532 16.98008 1.512328
Valid N (listwise) 587
20 4. Results and analysis
This section contains the results of the quantitative empirical analyses explained in the previous chapter as well as the univariate analysis of all key variables, the multivariate analysis for the full model and the results of significance tests. I evaluate the hypothesis of section 2.2 based on these results.
The standard error bar charts shown below give us an impression about the relationship of the populism- and press freedom variables.
Figure 3: Standard error bar charts for press freedom under populist and non-populist rule
The mean values indicated by the dots show us the following: taking Kyle/Gultchins’ definition, the average score for press freedom in times without populist rule is 39.2 (Freedom House) respectively 0.82 (V-Dem). When populists are in power, Freedom House gives an average score of 48.5; V-Dem of 0.73. This shows a difference of ca. 9 (respectively 0.9) points for both indices. Therefore, media freedom is on average about 9 points lower under populist rule if we follow the definition by Kyle/Gultchin. This sounds alarming, however, taking Kenny’s definition, we come to a different result. On average, Freedom House gives a score of 41.5 when no populists in the
21
definition of Kenny are in government, V-Dem gives a score of 0.79. When populists are in government, the V-Dem rating is 0.82 on average while the Freedom House rating even goes up to an average of 42.9. This means that the mean press freedom index by V-Dem is slightly lower (3 points) when there are no populists in government. But taking the Freedom House index, we see that on average, the press freedom in a country gets rated about 1.5 points better under populists in government, if Kenny’s definition is used. The overlapping confidence intervals in the charts with Kenny’s definition, however, show that the result is insignificant while both graphs for Kyle/Gultchin’s populism variable show significance. Yet, this comparison is only meant to give a short impression of a possible outcome and needs to be tested in a model including other factors which will be done in the following.
First, I run my series of analyses with the variables by Kyle/Gultchin and V-Dem. The bivariate regression supports the result of the comparison of means with a significant negative impact of populist leadership on media freedom. It is remarkable that although the coefficient for populism decreases when more model specifications are added, it stays significant on the 0.01 level. Only in the complete model, the level of significance goes down to the 0.05 level. The value of the coefficient stagnates at -0.0146 after country-fixed effects are introduced. Therefore, the media freedom can be expected to drop by 1.5 points on the 0-100 low-to-high scale in the first year of populist rule. The cumulative long-term effect of populists in power is -9.8 points on the scale because part of the shock is carried over to the following years. So, the actual expected impact of a year of populist rule on press freedom is even bigger with a loss of almost ten points. This is a very strong result and supports the assumption of a negative influence of populist leadership on press freedom.
When testing Kyle/Gultchin’s populism variable on the Freedom House index, the results are slightly different. In the bi- and multivariate regression, we find a negative significant result for populist rule of about the same strength as in the previous model. But as soon as we change the model to a time-series analysis, the coefficients become insignificant and cannot be generalized anymore. Despite their insignificance, the results indicate a negative relationship between populism and press freedom.
With populism defined by Kenny and the V-Dem media freedom index, we only get one significant result which is in the multivariate regression that leaves out further specifications. This coefficient, as well as the insignificant others, is, however, very low which makes the model not suitable to draw conclusions about the relationship of populism and press freedom.
22
In the case of populists in power defined by Kenny and media freedom defined by Freedom House, we get two significant negative results for the multivariate regression and the added time-series. When country-fixed effects are introduced, the results become insignificant. It is interesting though that in the two models using Freedom House’s index, the results become quite similar after country-fixed effects are added.
23
Media Freedom
Index V-Dem Freedom House
Populists in Power – Kyle/Gultchin -0.0912** -0.0211** -0.0111** -0.0146** -0.0146** -0.0146* 9.3398** 2.9190** 0.7333 0.3535 0.3535 0.3535 Control Variables Time-series Country-fixed effects Region-fixed effects Clustered standard errors N 1127 675 675 675 675 675 925 587 587 587 587 587 R² 0.0591 0.9392 0.9534 0.9586 0.9586 0.9586 0.0634 0.7827 0.9541 0.9632 0.9632 0.9632 Populists in Power – Kenny 0.0239 -0.0010* -0.0044 -0.0063 -0.0063 -0.0063 1.4656 2.2925** 1.2118** 0.3588 0.3588 0.3588 Control Variables Time-series Country-fixed effects Region-fixed effects Clustered standard errors N 971 675 675 675 675 675 846 587 587 587 587 587 R² 0.0033 0.9362 0.9528 0.9579 0.9579 0.9579 0.0012 0.7811 0.9546 0.9632 0.9632 0.9632 *p<0.05
24
Comparing the results of the four regression models, it is remarkable that Kenny’s populism variable does not hold the assumption of influencing press freedom negatively under this hard test. Although there seems to be some relationship, we do not reach a significant result with any of the two definitions of press freedom and the coefficients for populists in government are very low when conducting the full analysis. This is contradictory to Kenny’s own study. This may be the case because of his restrained use of control variables. Also, we could make theoretical assumptions. If populism in the main sense of theory does show an impact and populism in Kenny’s organizational definition doesn’t, we can conclude that there must be more to populism than charismatic leadership and party mobilization. The threat of populism for media freedom seems to be rooted in another characteristic of populism that is not covered by Kenny’s definition. While he may measure a certain dynamic that is typical for populists, he does not depict the phenomenon of populism in a holistic way. If we take the more prevalent definition of populist leadership as by Kyle/Gultchin, we gain very different results. Populist rule then has a significant negative impact on “freedom of expression and alternative sources of information” under all specifications of the model. We do not get the same strong result for the Freedom House index which does not change anymore and loses significance after time-series respectively country-fixed effects are added.
Because this is the case for both populism variables, this indicates that the difference stems from the way of measurement of the Freedom House variable. A Bartlett’s test for homogeneity of variances shows that the variance in and between countries differs between the two indices as the Freedom House index contains less variance of press freedom within countries. This is the reason why it does not show as much change over time. It is, therefore, harder to identify the changes within countries over time using the Freedom House index. In comparison, the V-Dem index changes more annually which also reflects in the smaller coefficient. So, the V-Dem measure of media freedom is more sensitive to overtime changes which allows us to find significant results more easily than with the more stable Freedom House index. Although the cross-sectional analysis with the Freedom House index does support the hypothesis, the more specified model does not give full support to the hypothesis which is due to the focus and measurement of the variable. Although this does not allow us to accept the hypothesis in the case of Freedom House, we can still assume some relation between the two variables. I, therefore, conclude that we can accept the hypothesis that populist leadership has a negative impact on media freedom on a 95% confidence level under the condition that media freedom is measured with the V-Dem index and populism is defined by Kyle/Gultchin. For all other models, we cannot accept the hypothesis as we have not gained significant results. Populism in the definition by Kyle/Gultchin has a by far higher
25
explanatory quality for the level of media freedom than the definition by Kenny which leaves us unable to draw any conclusions about an impact on media freedom as we assumed. Kenny’s populism variable measures a different dynamic than the one described by Kyle/Gultchin. This, however, does not influence the level of media freedom in a country as shown in this study. Using the more prevalent definition of populism, we do gain results that show a negative impact of populists in power on media freedom.
5. Conclusion
This paper contributes in several ways to the study of the influence of populist leadership on media freedom and the research of populism in a broader sense. We have seen that populist leadership in the most common definition is indeed harmful to media freedom. We can observe this dynamic both across and within countries and the effect remains statistically significant under the inclusion of all model specifications. Media freedom can be expected to deteriorate by 1.5 points on the 0-100 scale in the first year when populists come into power in a country which has a long-term cumulative effect of -9.8 points. This is a remarkable and alarming result which proves the necessity of taking the threat of populism to free media seriously.
Based on existing theory, we have learned that populists have various reasons to attack the free press. Some are rooted in their ideology, like their anti-elitist and anti-pluralist sentiments that they project on the media. The press is seen as an elite actor, mediator, watchdog for the government and provider of space for the exchange of ideas which results in the populist aversion against them as they see themselves as the incorporation of the true “will of the people”. So, they often try to restrict the free work of independent media and tend to prefer the use of social media as a communication tool.
Moreover, it became clear how essential an unambiguous definition of populism is as my contradictory results based on the two populism variables illustrate once again how difficult and misleading it can be to compare studies on populism that are based on different definitions of it. We have as well seen that the organizational approach to populism does not hold the same explanatory quality as the ideational one. It explains a typical behaviour of populists without capturing the core of the phenomenon and is therefore only conditionally applicable to the study of the negative effects of populist leadership. As also supported by other research, the real threat of populists in power does not stem from their organization or charisma, which as such is not dangerous, but rather their mindset.
This thesis has empirically shown to what extent populists in power threaten the free media. The results are in line with the studies by Ruth-Lovell (2019) and Kyle/Mounck (2018) but do not
26
support the research by Kenny (2019) as I do not find significant results in the models using his definition of populism. The study was conducted in a more thorough way than previous works on the topic as it included two variables for populism and two for media freedom as well as more control variables. This allows us to take a broader picture of the issue. It has therefore taken existing research further and refined it.
This project is, however, limited by its timespan from 1990 to 2018 and the use of cases in democracies only. It would be interesting to see if the strong difference between the two variables holds in another context. Also, I have used a dichotomous variable for populism although the degree of populism of a leader or party may have changed over time in some cases. A more extensive study could include cases of mainstream parties that have at some point made use of populist rhetoric or strategies. We could moreover expect populists to have some impact on the level of media freedom even when they are not in government but in opposition which could as well be a topic for future studies. This study has found empirical proof for the theoretical assumption that populists in power decrease the level of media freedom in a country. Future research could focus on the empirical study of the mechanisms behind this. For example: Why do populists threaten the free media and under which conditions do they do more or less harm to them?
This study allows us to understand populism better and especially the distinct populist use of the media as well as their often hostile rhetoric against them. This is of great academic and societal relevance as it illustrates the real and inherent threat of populism for the media. With this knowledge, we are better prepared to face this problem.
27 References
- Akkerman, Tijtske (2011): Friend or foe? Right-wing populism and the popular press in Britain and the Netherlands. Journalism 12(8), p. 931-945.
- Andersen Jones, Ronnell/West, Sonja (2017): The fragility of the free American press. North Western University Law Review 112(3): p. 567-596.
- Batory, Agnes (2016): Populists in government? Hungary's “system of national cooperation”. Democratization, 23(2). p. 283-303.
- Beiler, Markus/Kiesler, Johanna (2018): “Lügenpresse! Lying press!“ A content analysis study of the bias of journalistic coverage about ‘Pegida’, the movement behind this accusation. In: Otto/Köhler [ed.]: Trust in Media and Journalism: Empirical Perspectives on Ethics Norms, Impacts and Populism in Europe, p. 155-179.
- Coppedge et al. (2019): V-Dem Codebook v9. Varieties of Democracy (V-Dem) Project. Varieties of Democracy (2019): V-Dem [Country-Year/Country-Date] Dataset v9. https://www.v-dem.net/en/data/data-version-9/.
- De Vreese et al. (2018): Populism as an Expression of Political Communication Content and Style: A New Perspective. The International Journal of Press/Politics 2018, Vol. 23(4), p. 423 –438.
- Dutta, Nabamita/Roy, Sanjukta (2016): The interactive impact of press freedom and media reach on corruption. Economic Modelling 58(2016), p. 227-236.
- Engesser et al. (2017): Populism and social media: how politicians spread a fragmented ideology. Information, communication and society 20(8), p. 1109-1126.
- Goldstein, Ariel (2018): The Press and Classical Populism in Argentina and Brazil. Latin American Perspectives 220 (45) No. 3, p. 109-123.
- Groshek, Jacob/Engelbert, Jiska (2012): Double differentiation in a cross-national comparison of populist political movements and online media uses in the United States and the Netherlands. New Media & Society 15(2), p. 183-202.
- Haller, André/Holt, Kristoffer (2018): Paradoxical populism: how PEGIDA relates to mainstream and alternative media, Information, Communication & Society, DOI: 10.1080/1369118X.2018.1449882.
- http://time.com/5592347/austria-government-video-scandal/, accessed 30.05.2019. -
