Media policy attitudes and political attitudes: the politization of media policy and the support for the ‘media welfare state’

ABSTRACT

Previous research has neglected media audiences’ and citizens’ opinions on how the media should be organized, how they should function in society and what individual, corporate and state responsibilities should be in regard to these questions. In an attempt to understand the relationship between citizens’ broader political attitudes and their attitudes on media-related politics and responsibilities, this study uses a survey (n = 2003) of the adult Swedish population to investigate the distribution of a range of media political attitudes in the contemporary space of political positions. The results reveal overlaps between the space of media political attitudes and the broader political space, where support for a Nordic ‘media welfare state’ corresponds to leftist and GAL-oriented values, while TAN-oriented and right-wing attitudes link to scepticism towards state interventionism in the media landscape. A small but highly opinionated right-wing and TAN-oriented segment displays laissez-faire views on media policy that are reflected in current policy propositions from right-wing political parties in parliament.

KEYWORDS:

• Media policy
• media political attitudes
• political attitudes
• multiple correspondence analysis
• media welfare state

Introduction

Media policy has always been an arena in which conflicts between different political interests, ideas and ideologies have shaped the media societies that we inhabit (Freedman 2008). During the last decades, however, it would be safe to say that public debates on how the media are organized, regulated and operating have become more vivid and heated. To some extent, this might be due to the fact that we lead increasingly media-saturated lives (Hepp 2019). Rapid technological development, as well as democratic and economic challenges in the so-called ‘platform society’ (Van Dijck, Poell, and De Waal 2018), have raised new questions and brought new issues to the table of media policy debate. Of equal importance is the changing political climate in many parts of the world, which includes the rise of right-wing populist movements and radical right-wing and neo-fascist parties, for whom the traditional media have become a main target (Carlson 2018). The new right-wing rhetoric comes with a framing of the media – and of public service media and journalism in particular – as being allied with the ‘elites’ of society and standing in the way of the ‘people’ (Sehl, Simon, and Schroeder 2020; Holtz-Bacha 2021). There are thus reasons for the contention that media policy has been re-politicized, and that conflicts and debates on how to organize the media in society have become more intense. Ordinary citizens’ views on how the media should be organized and function, and on individual, corporate and state responsibilities regarding the media, can be broadly labelled as ‘media political attitudes’ and are referred to as such in this paper. Although this concept has recently attracted scholarly interest, media policy research has not fully understood the link between media political attitudes and people’s broader, non-media-related political attitudes. This study sets out to remedy this gap.

Even though public debate on media policy has become more intense in recent years, media policy largely remains an elite discourse in which experts, opinion leaders, politicians and lobbyists play the main roles. While media policy in Europe has changed rapidly – in a neoliberal direction since the 1980s (Jakobsson, Lindell and Stiernstedt 2021, Ala-Fossi 2020) and currently in a more authoritarian direction (Holtz-Bacha 2021; Sehl, Simon, and Schroeder 2020; Surowiec and Štětka 2020) – change has mainly been orchestrated from ‘above’ or from external stakeholders in international flows of policy transfer (Freedman 2003; Sarikakis and Ganter 2014), rather than being the result of public political debates or popular demands (Freedman 2010; Ward 2002). At best, the extent to which media policy shifts are the outcome of democratic processes and popular demand remains unclear. In this context, research has an important role to play in the public debate by highlighting citizens’ ‘normative and value-based expectations concerning media performance’ (Hasebrink 2011, 334). To date, this task has not been taken up by media scholars, which implies that we have little understanding of the extent to which people support overarching media policy paradigms (Van Cuilenburg and McQuail 2003) and how that support (or lack of support) relates to people’s political attitudes. As argued by Flew (1998, 324), media policy processes have often been a ‘translation of normative principles to processes of technical calculation and of routine administration’. Obtaining a better understanding of popular support for various forms of media organization and of how this support connects to ideological sentiments and political orientations is hence a way for policy research to help to reconnect the issues of media policy to a political sphere, in which such questions belong, and to increase opportunities for public participation in policy formation (Hasebrink 2011).

It is against this backdrop that we in this article analyse media policy attitudes in the Swedish population and understand them in relation to general political attitudes. The aim is hence to clarify how the specific subset of media policy issues is related to the overall attitudinal structure among the citizens.

We take as our case Sweden – a Nordic ‘media welfare state’ (Syvertsen et al. 2014) known for well-funded public service media, extensive media subsidies, a consensus-oriented media market and high levels of media access, use and trust. In what follows, we probe citizens’ views on various functions, institutions and responsibilities connected to the Swedish/Nordic media model. Previous research (Lindell, Jakobsson and Stiernstedt 2022) has identified strong citizen support for a welfare-oriented media policy and identified significant gaps between neoliberal media policy measures taken in the political field over the last 30 years on the one hand, and citizens’ views on how the media should be organized on the other (ibid.). While welfare-oriented media policy has been shown to be supported by a majority of Swedish citizens, disagreements exist between citizens with different political views. In this article, our aim is to analyse further the relationship between citizens’ ideological differences and their attitudes towards media politics.

The research question that we set out to answer with the help of a web based survey of Swedish citizens (n = 2003) is: what is the relationship between citizens’ political and ideological orientations, conceptualized here as a multidimensional space of political attitudes, and their attitudes on media politics? This study is relevant particularly because of (1) the technical and administrative character of media policy, which makes the connection to political ideologies and party preferences uncertain; (2) the hidden nature of media policymaking, which has resulted in questions on media policy often having been absent from public debate (Freedman 2010); (3) the recent politicization of media policy, which appears to have moved questions on media policy onto the political agenda; and (4) the fact that such research can assist in facilitating a stronger link between citizens’ opinions and policymaking.

Political attitudes and attitudes towards the media

Media policy scholars have generally not prioritized attitudes among citizens as an area of research. As argued above, however, we find that media research has an important role to play in analysing and highlighting citizens’ ideas, wishes and attitudes in regards to media policy and thereby creating a link between citizens and the policy field. Verboord and Nørgaard Kristensen (2021) present a recent exception to this rule – albeit one that concerns cultural policy more broadly – in their work on ‘cultural value orientations’ in relation to EU cultural policy and different media systems in Europe.

Comparative media systems researchers have included citizens in their research to some extent, but this has mainly been done by conceptualizing citizens as media audiences. Research interest within this field has thus been directed towards differences in media use in and between different countries – for instance, in terms of levels of news consumption (Hallin and Mancini 2004; Syvertsen et al. 2014). The media user as a political subject and her role as a citizen and participant in processes of media governance have thereby been side-lined.

There is some research on popular movements for media reform (e.g. Pickard 2014; Segura and Waisbord 2016), as well as research on forms of media activism (e.g. Freedman 2019; Milan 2017); however, such practices only include very small fractions of the general population. The question of the broader citizenry and their attitudes on how to organize the media in society and what the functions and responsibilities of the media are, as well as the relation between such attitudes and general political sentiments, have been overlooked. Consequently, there is little previous research to build on in this study. Nevertheless, it is reasonable to believe that general differences in political attitudes, such as an acceptance of state interventions versus more laissez-faire contentions, play into citizens’ views on media political issues. In order to contextualize this study, we rely on research on the connection between political attitudes, ideologies and party preferences on the one hand, and media preferences, media use, perceptions of the media and media trust on the other. Here, research has consistently shown that the political orientation of citizens affects how they view the media. This body of research provides clues on how the political orientation of citizens might overlap with their attitudes on media policy.

Firstly, an important part of the Nordic media model – the media system taken as our case – is its strong public service institutions, which are upheld by generous funding, institutionalized autonomy from the state and a broad mission provided in the broadcasting license. Even though the last 30 years have seen an increasing marketization of broadcasting in the Nordic countries (e.g. Sjøvaag 2012), previous research has suggested that there is a relative consensus around this policy design, at least among the elites, and that ‘diverging political perspectives on public service media should not be exaggerated as they have so far more focused on principles than on practices’ (Arriaza Ibarra and Nord 2014, 82). In a study from the Netherlands, which has a similar media system and political system to that of Sweden, Bos et al. (2016) found that public service media act as a bridge between different segments of the media audience and that people consume public service media regardless of political affiliation. In a similar vein, Dahlgren (2019) found that both the political left and the political right in Sweden turn to public service media and that this has not changed during the years from 1986 to 2015, despite an increased availability of commercial alternatives in the Swedish media market. Rather than left-wing or right-wing viewpoints, substituting public service media with other media outlets can be explained by political disinterest and an attraction to political parties outside parliament (Dahlgren 2019).

Consequently, it seems reasonable to expect that policy attitudes towards public service media institutions in Sweden should be more or less positive across the political spectrum. However, previous research has also revealed differences in the use of and trust in public service media among Swedish citizens (Bergström and Wadbring 2012; Jakobsson, Lindell and Stiernstedt 2021) that can be explained by age, generation and political orientations. Historically in Sweden, there has been a divide between left-wing and right-wing parties in their relation to public service media, with the former generally having a more positive opinion to a ‘broad’ public service media (Arriaza Ibarra and Nord 2014). In recent years, the polarization between right-wing and left-wing parties on the issue of public service media has intensified, with proposals on re-fashioning and defunding the public service media coming from the political right (Jakobsson, Lindell and Stiernstedt 2021). If this intensification is reflected in the popular attitudes of right-wing voters, we should expect to find a polarized view on public service media within the citizenry.

Secondly, a key dimension of Nordic media policy concerns the broader role of the press and journalism. The Nordic countries are not only characterized by a high degree of press freedom (Press Freedom Index 2021), but also by a relatively high news readership among all strata of the population, even though such readership is decreasing, especially among younger people who tend to turn to social media for news (Bergström and Jervelycke Belfrage 2018). The Nordic press has a strong tradition of self-regulation (i.e. ethical codes and an ombudsman) and has been supported by the state through uniquely generous direct subsidies, which have increased in recent years (Nord and von Krogh 2021). Do previous studies provide any clues on whether this policy framework is an area of political conflict? For one thing, previous research has shown that people with populist attitudes tend to prefer ‘tabloid’ media over ‘quality’ media (Hameleers, Bos, and de Vreese 2017). Since the tabloid press is more often accused of transgressing the ethical boundaries of the press (Biressi and Nunn 2008), this leads us to expect that people who support the Swedish right-wing populist party might be more negatively attuned to the notion that media companies should be obliged to take social responsibility and to press ethics and other forms of self-regulatory frameworks for media publishing. This expectation is also supported by the fact that the most visible ‘alternative’ media actors in the Nordic countries, which operate outside of the established system of professional journalism, are mainly right-wing populist media outlets.

In the last few years, press subsidies in Sweden have been transformed into platform-neutral media subsidies; they have also increased. However, there has not been much public debate on this issue, and no research has been conducted on whether there is popular support for this policy measure. Since the subsidies mainly go to established and traditional news outlets, it might be expected that people who do not consume news (or are hostile towards mainstream news) would be less in favour of media subsidies. This assumption relates to findings on political selective exposure, which indicate that people with extreme political views are more likely than others to turn to partisan news outlets that support their political preferences and to avoid content that contradicts their political position (Rodriguez et al. 2017). Previous research also indicates that distrusting traditional media outlets and political selective exposure is more common among people on the right-wing side of the political spectrum (ibid.). The consumption of alternative news sites operating outside of the political support systems for news and journalism should thus be more prevalent among people who agree with conservative or right-wing ideologies. This expectation is supported by the fact that most of the high-profile alternative news outlets in the Nordic countries brand themselves as right-wing outlets (Ihlebæk and Nygaard 2021) and that the consumers of these media are more critical towards traditional media (especially public service media) and more sceptical of news quality in general (Schulze 2020). These segments of the citizenry display what Holt (2018) refers to as a general ‘anti-systemness’. It can thus be expected that people with (extreme) right-wing views are less positive towards media regulations that support news and journalism in various ways.

Thirdly, much current debate on media policy has concerned ‘platformization’, which refers to the rise of large tech companies and their role in the Swedish media landscape, along with the related issues of personal integrity, safeguarding of personal information and control over personal data. Another issue concerns freedom of speech and the occurrence of hate speech and defamation on these platforms, which has pushed the issue of media education and media and information literacy to the forefront. These are issues in which the state can play a key role through the educational system. Wagner (2019, 87) argues, however, that media and information literacy does ‘not have a central foundation in Swedish politics, nor is there an overall political framework for such matters’. There is no previous research on how such policy issues have been interpreted and embraced (or not) by the public; however, considering the absence of media and information literacy questions in the Swedish political system, a low political polarization regarding this issue might be expected. Cocq et al. (2020) showed that Swedes are generally aware and critical of online surveillance, as they tend to value their online privacy and integrity. However, rather than seeking technical or political solutions to this problem, they tend to handle online surveillance by adapting their individual behaviour. Therefore, a low level of political polarization on the issue of the importance of control over data might be expected.

The debate on hate speech and freedom of expression on digital platforms is arguably a more polarized discussion, in which suggestions for regulation have been manifold (e.g. Banks 2010). The fact that this discussion taps in to the broader political issues of interventionism versus laissez-faire, which have a long history in the area of freedom of expression (Peters 2005), makes it reasonable to expect that political attitudes will be a predictor for where an individual stands on both the regulation of content on digital platforms and the support for media education and media literacy programmes. Furthermore, the fact that parties and groups belonging to the radical right have been shown to be active in creating and spreading hate speech and using social media and digital platforms for distributing such content, as well as in attacking political opponents (Edström 2016; Pettersson 2020; Wahlström, Törnberg, and Ekbrand 2021), suggests that a far-right political affiliation might affect an individual’s standpoint on this media political issue.

The space of political attitudes

In this study, our interest lies in mapping the relationship between citizens’ political attitudes and their attitudes to media policy. Political attitudes can be approached in many different ways. Following Harrits et al. (2010), we conceptualize citizens’ political attitudes in an open-ended and empirical way as a multidimensional space of position-takings. The ‘materialist’ or ‘economic’ left-right dimension has dominated political analysis for a long time and is arguably still the most important political dividing line in today’s politics. However, in order to nuance contemporary political dividing lines, political scientists have added the so-called post-materialist or GAL (green, alternative, libertarian)/TAN (traditionalist, authoritarian, nationalist) dimension (Hooghe, Marks, and Wilson 2002). Political disputes along this dimension revolve around ‘questions relating to the environment and climate change, (nuclear) war, rights for homosexuals, racism and xenophobia’ (Flemmen 2014, 545). It can be questioned, however, whether ‘post-materialism’ is a good label for these issues, given that all have a clear economic dimension – such as the intersection of race and class, the uneven economic consequences of climate change and how migration is believed to impact the economy (Evans and Tilley 2017). Nevertheless, the two-dimensional approach used in many contemporary analyses has been shown to be useful to study overlaps, or homologies, between, for instance, social class positions and political attitudes (Harrits et al. 2010; Enelo 2012; Harrits 2013; Flemmen and Haakestad 2018).

As mentioned above, previous research shows that the right-/left-wing dimension influences how people consume and relate to the media. It is thus reasonable to expect that such ideological oppositions affect people’s attitudes towards media policy issues. The inclusion of post-material attitudes (e.g. GAL/TAN values), however, is not particularly common in media studies. Considering that media policy is a form of cultural policy (as well as economic policy), and since previous research has revealed connections between GAL-TAN orientations and support for welfare-oriented media politics – with people holding GAL-views being more positively attuned (Lindell, J., P. Jakobsson, and F. Stiernstedt 2022) – it is necessary to include the post-material dimension in the present analysis.

The drawback of the ‘material’ and ‘post-material’ approach to contemporary ideological conflicts is that it does not fully capture the rise of right-wing populism. Populism is often described as a ‘thin ideology’ (Mudde and Rovira Kaltwasser 2013), as it is defined with just a few key characteristics and does not hold a fixed place along the two dimensions of the political spectrum described above. Still, as previously discussed, populist attitudes seem to correlate with attitudes towards news and journalism. It is thus relevant to include populist and non-populist orientations in an analysis. One way to address this issue is to include party preferences in the analysis. The current Swedish parliament hosts one party – the Swedish democrats (with 17.5 percent of the votes in the 2018 general election) – that shares some characteristics with European right-wing populist parties (Hellström and Kiiskinen 2013). Moreover, including current party preference in the analysis provides additional details to the overall analysis, since ideological opinions do not always overlap with party preferences (Dahlgren 2019).

Data and method

The aim of this study was to study the relationship between citizens’ media political attitudes – that is, their views on how the media should be organized, how they should function, and the responsibilities held by individuals, corporations and the state with regard to these questions – on the one hand, and peoples’ general political attitudes on the other hand. This endeavour required fairly specific data, not only on people’s standpoints on various media policy measures, but also on their political opinions. To obtain the necessary data, we developed a web-based survey that was administered by the research institute Kantar-Sifo. The survey was distributed to 10,395 Swedes between the ages of 18 and 99 in November 2020. We received 2003 responses, for an answering rate of 19.3 percent. While the survey captured well citizens of different ages (M = 49, SD = 18) and genders (men/women = 50/50 percent) it did not adequately represent people with no or very low educational qualifications, which should be taken into consideration when interpreting the results.

Our first step was to delineate the structure of the Swedish space of political attitudes. In following previous studies on the space of political attitudes (e.g. Harrits et al. 2010; Enelo 2012; Harrits 2013; Flemmen and Haakestad 2018), and in order to plot media political attitudes in the space of political attitudes, to visualize the results, and to avoid assumptions on linearity in the model, we relied on multiple correspondence analysis (MCA).

To construct a statistical representation of the contemporary Swedish space of political attitudes 16 measurements of various political standpoints, including both traditional left-wing versus right-wing attitudes (eight variables) and post-material (GAL/TAN) attitudes (eight variables), were used. While the process of operationalization drew inspiration from previous attempts to study the space of political attitudes (e.g. Harrits et al. 2010; Harrits 2013; Flemmen and Haakestad 2018; Lindell and Ibrahim 2021), it also included specific items related to the political landscape in Sweden (e.g. attitudes on the Employment Protection Act [LAS] and tax subsidies on household services and renovations [ROT/RUT]). Table 1 lists the active variables that were used in MCA to construct a statistical representation of the Swedish space of political attitudes by extracting and visualizing the principal factors, or dimensions, among the 16 variables. The variable ‘political party preference’ was used as a supplementary variable to study where prospective voters according to their political attitudes were positioned in the political space. In order to avoid a skewed model, variables were recorded so that no individual value held less than 5 percent of the observations (Hjellbrekke 2018).

Table 1. Active variables creating the space of political attitudes in Sweden.

Heading	Variable	Categories	N	%
Left-wing vs. right-wing	Politicians should reduce economic gaps	-/- -	268	13.7
		+	783	40.1
		+ +	903	46.2
	Taxes are too high	- -	398	20.4
		-	504	25.9
		+	592	30.4
		+ +	452	23.2
	Remove ‘karensdag’	- -	295	15.5
		-	395	20.8
		+	655	34.6
		+ +	552	29.1
	Welfare should be non-commercial	- -	240	12.6
		-	295	15.5
		+	646	33.9
		+ +	724	38
	ROT/RUT tax cuts should increase	- -	385	22.8
		-	459	27.2
		+	546	32.3
		+ +	300	17.7
	Reintroduce wealth tax	- -	781	42.8
		-	309	16.9
		+	473	25.9
		+ +	264	14.5
	Make ‘LAS’ more flexible	- -	650	33.9
		-	450	23.5
		+	631	32.9
		+ +	187	9.7
GAL vs. TAN	It is good that homosexual couples can get married	-/- -	116	6.3
		+	119	6.5
		+ +	1608	87.2
	Give the police force more authority	-/- -	269	14.2
		+	613	32.4
		+ +	1011	53.4
	Increase penalties for crime	-/- -	228	11.7
		+	604	31
		+ +	1120	57.4
	Reduce immigration	- -	243	12.6
		-	321	16.7
		+	615	32
		+ +	746	38.7
	Increase taxes on fossil fuels	- -	310	16.3
		-	377	19.8
		+	727	38.1
		+ +	492	25.8
	Stay in the EU	- -	153	8.2
		-	178	9.2
		+	586	31.4
		+ +	948	50.8
	Give more resources to the armed forces	- -	119	6.5
		-	272	14.7
		+	745	40.3
		+ +	711	38.5
	Liberalize wolf hunting	- -	411	27.8
		-	399	27
		+	392	26.6
		+ +	274	18.6
	Increase gender equality	- -	100	5.2
		-	134	6.9
		+	608	31.3
		+ +	1099	56.6

- - means ‘I don’t agree at all’; - means ‘I don’t agree’; + means ‘I agree’; and + + means ‘I fully agree’. ROT/RUT are tax cuts on household services, LAS is the Employment Protection Act, and ‘karensdag’ is an initial day of sick leave without pay.

To study citizens’ opinions on how the media should be organized, how they should function, and the responsibilities of individuals, corporations and the state with regard to these questions, we probed peoples’ attitudes on various functions and institutions connected to the media system they inhabit – that is, the Swedish, or Nordic, media system (Syvertsen et al. 2014). Among these 16 variables were statements saying, for instance, that communication services should be organized in a way that ensures their character as ‘public assets’, that extensive subsidies exist and that requirements should be set for universal access (e.g. broadband expansion, public service media). Other statements captured whether or not the respondents agreed that regulation should exist to ensure the freedom of the press, editorial freedom and professional autonomy in the media industry and in the journalist corps. Table 2 shows the supplementary variables and their distribution. In the subsequent analyses, these variables were recoded so that no value held less than 5 percent of the observations.

Table 2. Citizen’s media policy attitudes (percentages).

	“I do not agree at all”	“I do not agree”	“I somewhat agree”	“I fully agree”	“I do not have an opinion”	Total
It is every citizen’s duty to gather information and stay up to date on important societal matters	1,4	3,3	39,9	53,2	2,2	100%
The state and the educational system should ensure that all citizens have the right skills in order to use important internet services	4,4	14,0	47,1	29,7	4,7	100%
It is important that news media report on issues related to minority groups	3,5	7,7	40,2	44,6	4,0	100%
Public funds should be used to ensure diversity on the media market, for example by supporting smaller actors and news outlets	8,5	15,4	44,4	21,6	10,0	100%
The financial support to public service media should be decreased	39,7	21,9	17,2	13,2	8,0	100%
Public service media are important in order to guarantee access to independent journalism	7,0	5,4	20,9	64,0	2,7	100%
Public service media should not produce broad formats, but solely focus on news and societal information	49,9	24,7	15,8	6,7	3,2	100%
Global internet companies should be taxed by the Swedish state	8,3	8,2	24,7	38,4	20,5	100%
It is important that I have control over the information that is collected about me when I use social media	0,8	2,2	21,6	73,0	2,5	100%
Commercial media companies (including social media companies) should act against false information	2,0	1,6	12,7	80,9	2,8	100%
Commercial media companies (including social media companies) should take democratic responsibility by supporting local journalism	5,4	11,2	41,9	24,9	16,5	100%
Public service media should expand digitally and become an alternative to commercial actors	21,2	20,9	26,7	11,8	19,4	100%
Internet access is a citizen right and as such internet access should be publicly funded via taxes	16,9	19,3	36,7	20,0	7,0	100%
Citizen journalism is a viable alternative to established news media in the parts of the country lacking professional journalism	34,0	27,2	16,4	2,9	19,5	100%
Journalism is not important in a society where everyone can share and consume information on the internet and social media	71,6	15,5	7,3	3,4	2,2	100%
Knowledge on how various media work should become a more important element in primary school	0,6	1,8	20,9	74,1	2,7	100%

We were only concerned with respondents who displayed their political standpoints. In the following analyses, all ‘I don’t have an opinion’ answers in the active variables were coded as missing.

Results and analysis

In order to study the relationship between attitudes on how the media should be organized and function in society and the wider space of political attitudes, we began by constructing a statistical representation of the contemporary Swedish space of political attitudes. We then plotted our 16 variables measuring media political attitudes in the space of political attitudes as supplementary, passive variables.

The space of political attitudes

Figure 1 shows a statistical representation of the contemporary Swedish space of political attitudes. More specifically, it displays the result of an exploratory statistical method (i.e. MCA) that extracted dimensions from the 16 variables used to measure the respondents’ political attitudes. The first dimension (explaining 64.9 percent of the variance) and second dimension (explaining 14.5 percent) account for a total of 79.4 percent of the variance; we have chosen to focus only on these two main dimensions (see Table A1).

Figure 1. The Swedish space of political attitudes. MCA, axis 1 and 2. Missing values (including “I don’t have an opinion”) have been removed (n = 874). - - = “I don’t agree at all”, - = “I don’t agree”, + = “I agree”, + + = “I fully agree”.

The first dimension (horizontal axis in Figure 1) captures differences (in the form of distances in the space) between left-wing and right-wing views, as well as between GAL and TAN views (see Table A2 for the active variable contribution to dimensions 1 and 2). To the right in Figure 1 are individuals who favour both right-wing and TAN-oriented policy. Individuals located towards this pole in the space tend to prefer economic liberalism (e.g. reducing taxes, creating a more flexible labour market and keeping a commodified welfare system) over material redistribution. At the same time, they favour social conservatism or TAN values (e.g. opposing the legalization of homosexual marriages, wanting a stricter migration policy and opposing the allocation of energy and resources into creating more equality between genders) over GAL values. In opposition, to the left in the space are individuals with left-wing and GAL-oriented attitudes.

The second (vertical) dimension differentiates individuals according to the intensity of their political views. Individuals located at the top of the space tend to strongly agree or disagree with the policy statements provided. In contrast, the majority of the respondents, which are positioned in the bottom half of the space, hold more modest political views.

These results are in agreement with previous studies on the space of political attitudes in that we identify an ‘intensity of opinion’ dimension (e.g. Flemmen and Haakestad 2018; Harrits 2013; Harrits et al. 2010; Lindell and Ibrahim 2021). The results, however, go against the body of previous research arguing that the GAL/TAN and left-wing versus right-wing dimensions make up two separated realms of contemporary politics (e.g. Hooghe, Marks, and Wilson 2002). While separating the GAL/TAN dimension from the left-wing versus right-wing dimension makes conceptual sense and might reflect the policy profiles of political parties quite well (Hooghe, Marks, and Wilson 2002), our analysis aligns with those of Evans and Tilley (2017) and Lindell and Ibrahim (2021) in that attitudes connected to ‘material’ matters and redistribution (left-wing versus right-wing) and ‘post-material’ values (e.g. GAL or TAN orientations) seem to be highly interlinked, at least in terms of the attitudinal constellations of ordinary citizens.

Projecting the variable measuring respondents’ current favourite party preference as a supplementary, passive variable reveals that the Left Party is preferred by people located towards the top left in the space – that is, where people with intense GAL/leftist views are located. In contrast, the Sweden Democrats are the preferred party for individuals in the middle of the top-right quadrant of the space, where relatively intense TAN/right-wing attitudes prevail. Citizens that favour the Centre Party and the Liberals are found in the middle on the horizontal (GAL/left vs. TAN/right) axis and 0.5 deviations south of the centre on the vertical axis – a finding that implies that people who prefer these parties do not hold intense opinions on the policy issues probed here. People favouring the Social Democrats and the Green Party are unsurprisingly drawn to the left-wing/GAL pole, but they do not hold as intense views as the Left Party voters. Finally, respondents preferring either the Moderate Party (conservative) or the Christian Democrats are found towards the TAN/right-wing pole in the space, but they do not display as intense opinions as the Sweden Democrats.

This initial step, which yielded a statistical representation of the space of ideological conflicts between contemporary Swedish citizens, provides the basis for understanding the relationship between political ideology and polarization and views on how the media should be organized and function in society.

Media political attitudes in the space of political attitudes

Figure 2 retains the structure explored above, in which the contemporary space of political attitudes in Sweden seems to be structured around two main principles of division: an opposition between GAL/left-wing views on the one hand and TAN/right-wing views on the other hand (horizontal axis), and a differentiation between those with ‘intense’ views on political matters and those with less animated views (vertical axis). In this figure, the active variables have been made invisible; instead, Figure 2 shows the distribution of our supplementary variables: the respondents’ media political attitudes. As a first observation, it is noticeable that many variable categories are located relatively far (>.4 deviations) from the centre of the space, which suggests that oppositions in political orientations explain differences in views on how to organize the media rather well. In most cases, the relationships between attitudes on how to organize the media in society and views on responsibility in media politics and the axes described above are statistically significant (Table A3). However, a couple of media political attitudes tend to be distributed very close to the centre of the space, meaning that variations in these variables are not well explained by peoples’ political views and the intensity of those views. These media political attitudes include the view that citizens have a duty to keep up to date with current affairs and that it is important to have control over personal data collected on social media. As discussed above, the latter result was expected, due to the fact that Swedish citizens regard online privacy issues as an individual responsibility rather than a political one. The notion that citizens have a civic duty to keep up with news and current affairs is shared by virtually all citizens (Lindell 2018), and is thus not a politically divisive opinion.

Figure 2. Projecting media policy attitudes into the Swedish space of political attitudes. MCA, axis 1 and 2. Missing values (including “I don’t have an opinion”) have been removed (n = 874). - - = “I don’t agree at all”, - = “I don’t agree”, + = “I agree”, + + = “I fully agree”.

In regard to most of the variables measuring citizens’ attitudes on media politics, there are clear political divisions within the citizenry. The role, function and funding of public service media is one of the most divisive issues, where individuals drawn to the top-right corner of the space (right-wing/TAN attitudes) are of the opinion that the public service media should be defunded, that these organizations should not expand their activities in the digital realm and that being part of public service media is not a guarantee of independent journalism. This result only partly aligns with what we had expected based on previous research on how audiences value the public service media. Studies have shown that there are differences in whether media audiences trust public service media or not that are dependent on political ideology (Bergström and Wadbring 2012); however, previous studies also suggest that public service media in the Nordic countries (Dahlgren 2019) and in other democratic-corporativist countries (Hallin and Mancini 2004) function as a bridge between people with widely differing political views among the audience (Arriaza Ibarra and Nord 2014). Judging from our results, it seems that the increased politization and polarization of the public service media has begun to be reflected in the attitudes of citizens. The differences within the citizenry analysed here, in contrast to the results of Arriaza Ibarra and Nord (2014), are not only a matter of consumption habits but of principles, since they involve attitudes on the defunding of the public service media and the belief that being a part of the public service media does not guarantee independent journalism. Furthermore, this stand against public service media is associated with preferring the right-wing populist party (i.e. the Sweden Democrats) in Sweden, as shown in Figure 2. To some extent, this is an expected result, since the Sweden Democrats are the political party with the most outspoken and aggressive rhetoric towards the public service media in their communications and in the policy suggestions they have put forward, both in parliament and in the public debate (Bengtsson 2021).

Overall, the segment of the population holding relatively intense TAN and right-wing views, which overlaps with the segment voting for a right-wing populist party, holds a more laissez-faire view on how to organize the media, with resentment towards state intervention in the media market. These segments of the population do not support the system of press subsidies and oppose a publicly funded strategy for broadband expansion. This lack of support for press subsidies was expected, based on the observation that right-wing attitudes are associated with a distrust of the mainstream news media (Schulze 2020) and on the establishment of right-wing populist alternative media outlets, some of which are not included in the Swedish system for press subsidies. We found no previous research on the support of media policy measures for Internet and broadband access; however, the results show that such support is distributed in the same way as support for press subsidies. The economic liberal standpoint is also reflected in the opposition to the taxation of global tech companies. Even though this segment has attitudes that can be interpreted as nationalistic – pro-military, against the EU and with migration scepticism – the economic liberalist position held by these individuals seems to override their national interest in curbing the power of the global tech giants, most of which are based in the United States or China.

The same segment of the population (right-wing, TAN-oriented) has strong views against the responsibility of the commercial media to act against false information. This finding aligns with the laissez-faire sentiments explored above. However, the survey question did not mention regulation by law but referred to moral obligations, so a negative response is compatible with a libertarian political standpoint. The response to this question that stands out in Figure 2 might therefore primarily link to how the issue of ‘fake news’ and ‘misinformation’ has become politicized and discursively configured in the contemporary moment. While the notion of ‘fake news’ has been appropriated by right-wing populist leaders (Farkas and Schou 2018), the fight against ‘misinformation’ is culturally coded as an elite project and as having alleged attempts to produce news that is in line with current ideas about ‘political correctness’. It is clear, for example, that negative opinions about the EU cluster together with strong opinions against the need for the media to take democratic responsibility and combat false information. The EU has been outspoken about its intention to combat disinformation (Kuczerawy and Kloza 2019), which might explain why these opinions are found close together. A similar line of reasoning might explain the opposition towards the inclusion of media literacy as a responsibility of the state and the education system. Media literacy, it might be speculated, is in some people’s view associated with a state-sanctioned ‘correct’ way of critiquing and interpreting media content, rather than being perceived as a politically neutral skill of processing information.

Taken together, these findings suggest that there are clear overlaps between the space of political attitudes and the space of media political attitudes. People holding GAL and leftist views have a more positive attitude towards the values, functions and institutions connected to the Nordic media model, whereas individuals with TAN and right-wing views have a more negative attitude. In addition, intense political views reflect intense media political views.

Conclusions and discussion

Media policy is sometimes seen as a technical, administrative affair (Freedman 2008). As such, there is a risk of it becoming disconnected from public debates and decided upon by experts, lobbyists and politicians, who are not held accountable for their decisions. Media research has a role to play here, by highlighting citizens’ views on crucial issues within media policy debates (Hasebrink 2011). This is one of the contributions of this study: connecting citizens’ attitudes on media political issues to their political ideologies and standpoints, and identifying clear affinities between the two realms. Although Sweden was used as the empirical case in this article, we hope that our open-ended approach in studying the relationship between ordinary citizens’ political standpoints and their views on how to organize the media can be used and developed in other contexts as well, not least in other media systems. Given the relatively low response rate on our survey, and the fact that people with no or very low educational qualifications were underrepresented we also encourage more studies in Sweden.

One conclusion that can be drawn from this analysis is that attitudes among Swedish citizens on how the media should be organized, how they should function, and the responsibilities of individuals, corporations and the state with regard to these questions tend to cluster together in the middle of the political space, as we conceived of it here. In the discussion of the results above, we mainly focused on the sentiments regarding media policies that were clearly connected to ideological polarization; however, the results also show that the disagreements on media policy are somewhat limited. Indeed, most Swedes have positive responses towards the values, functions and institutions that are connected to the Nordic way of organizing media in society (Lindell, Jakobsson and Stiernstedt 2022). Nevertheless, the differences that we have uncovered shows an overlap between media policy attitudes and general political attitudes, including both the material left- and right-wing dimension and the ‘post-material’ GAL-TAN dimension. We thus nuance previous research that has primarily focused on differences along left-wing versus right-wing oppositions.

A key finding of this study is that one particular segment of the citizenry is positioned outside of the central cluster, far away from the mainstream political and media political consensus. This is a highly opinionated segment with strong views against immigration, the EU, taxes on fossil fuels and same-sex marriages – that is, questions associated with the TAN position in the political space. This segment also contains those who express sympathy for the Swedish right-wing populist party. This part of the citizenry is highly sceptical of the media policy regime connected to the Nordic media model (Syvertsen et al. 2014). At the same time, this relatively small minority of the Swedish population displays media policy attitudes that are aligned with how the mainstream and public service media has been criticized by right-wing populist parties across Europe (Holtz-Bacha 2021). This finding also supports the observation of clear overlaps between citizens’ media political attitudes and their political attitudes.

These results relate to the wider field of media policy debates. The limited disagreements on media policy and the fact that most citizens are supportive of the general features of the Nordic media system create a ‘silent majority’ in media policy issues. When media policy is discussed in the public realm, the more ‘extreme’ or critical positions generally get the most attention (cf. Nord 2020); examples include the idea from the conservative party (Moderaterna) to close public service broadcasting all together (Folkö 2019), or the parliamentary motions from the right-wing parties to, for example, cut press subsidies, defund or refashion the public service broadcasters and increase political control over public service media (Bengtsson 2021). Also ideas from the left, such as harder regulations on platform companies or a ‘public service internet’ (Jonasson2015) represent such ideas, which as we have seen find little support in the attitudes held by the vast majority of the population.

The mis-match between citizens’ views on media policy and the policy debate and suggestions that is put forward in the public debate is most clearly pronounced for a large group of right-wing voters who hold relatively moderate views on media policy and therefore find themselves without political representation in large parts of the current media policy debates. Furthermore, the current, more general, policy discussion on the road ahead for the Swedish media system seems to be strangely disconnected from the standpoints and media policy ideals held by the great majority of the population.

Disclosure statement

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

Additional information

Notes on contributors

Peter Jakobsson

Peter Jakobsson is an Associate Professor in the Department of Informatics and Media, Uppsala University, Sweden.

Johan Lindell

Johan Lindell is an Associate Professor in the Department of Informatics and Media, Uppsala University.

Fredrik Stiernstedt

Fredrik Stiernstedt is an Associate Professor in the Department of Culture and Education, Södertörn University, Sweden.

Appendix

Table A1. Benzécri-adjusted eigenvalues of the top 10 dimensions among the active variables.

Dimension	Adjusted eigenvalue (%)	Cumulative
1	64.9	64.9
2	14.5	79.4
3	11	90.4
4	3.4	93.8
5	2.7	96.5
6	1.2	97.7
7	0.5	98.2
8	0.4	98.6
9	0.3	98.9
10	0.2	99.1

Table A2. Relations of active variable modalities to dimensions 1 and 2 (eta squared).

	Dimension 1	Dimension 2
Reduce economic gaps	0.478	0.158
Legalized homosexual marriage is good	0.074	0.062
Strengthen police	0.375	0.092
Strengthen penalties	0.414	0.137
Taxes are too high	0.583	0.280
Remove ‘karensdag’	0.120	0.141
No profits in welfare	0.333	0.207
Increase ROT/RUT	0.271	0.198
Bring back wealth tax	0.426	0.241
Make ‘LAS’ more flexible	0.321	0.263
Reduce immigration	0.532	0.228
Increase tax on fossil fuels	0.317	0.205
Stay in EU	0.043	0.175
Strengthen armed forces	0.356	0.206
More wolf hunting	0.143	0.206
More gender equality	0.222	0.131

Table A3. V-test for modalities in supplementary variables on the two main dimensions.

Dimension 1 (GAL/left vs. TAN/right)		Dimension 2 (Intensity of opinion)
Modality	V-test	Modality	V-test
Civic duty to be informed: - -‎/- (incl. don’t know)	1.277	Civic duty to be informed: - -‎/- (incl. don’t know)	0.306
Civic duty to be informed: +	−1.500	Civic duty to be informed: +	−0.805
Civic duty to be informed: + +	0.909	Civic duty to be informed: + +	0.653
State and edu. system ensure media literacy: - -‎/- (incl. don’t know)	9.115	State and edu. system ensure media literacy: - -‎/- (incl. don’t know)	2.435
State and edu. system ensure media literacy: +	−3.076	State and edu. system ensure media literacy: +	−2.954
State and edu. system ensure media literacy: + +	−4.707	State and edu. system ensure media literacy: + +	1.033
Important self control data: - -‎/- (incl. don’t know)	1.694	Important self control data: - -‎/- (incl. don’t know)	0.054
Important self control data: +	0.041	Important self control data: +	−2.373
Important self control data: + +	−0.799	Important self control data: + +	2.213
More media literacy in school: - -‎/- (incl. don’t know)	3.006	More media literacy in school: - -‎/- (incl. don’t know)	2.414
More media literacy in school: +	4.505	More media literacy in school: +	0.549
More media literacy in school: + +	−5.439	More media literacy in school: + +	−1.424
Important news on minorities: - -‎/- (incl. don’t know)	11.367	Important news on minorities: - -‎/- (incl. don’t know)	5.707
Important news on minorities: +	5.343	Important news on minorities: +	−4.907
Important news on minorities: + +	−13.194	Important news on minorities: + +	0.900
Public funds to ensure media diversity: - -	6.474	Public funds to ensure media diversity: - -	6.628
Public funds to ensure media diversity: -	4.179	Public funds to ensure media diversity: -	−1.028
Public funds to ensure media diversity: +	−1.396	Public funds to ensure media diversity: +	−4.426
Public funds to ensure media diversity: + +	−5.984	Public funds to ensure media diversity: + +	2.037
Public funds to ensure media diversity: Don’t know	−0.340	Public funds to ensure media diversity: Don’t know	−0.713
Cut PSB funds: - -	−13.931	Cut PSB funds: - -	2.243
Cut PSB funds: -	−0.344	Cut PSB funds: -	−6.760
Cut PSB funds: +	7.541	Cut PSB funds: +	−0.458
Cut PSB funds: + +	12.561	Cut PSB funds: + +	5.780
Cut PSB funds: Don’t know	−0.289	Cut PSB funds: Don’t know	0.093
State should tax global tech: - -	7.354	State should tax global tech: - -	5.406
State should tax global tech: -	1.994	State should tax global tech: -	−2.137
State should tax global tech: +	0.306	State should tax global tech: +	−7.569
State should tax global tech: + +	−5.669	State should tax global tech: + +	5.391
State should tax global tech: Don’t know	0.124	State should tax global tech: Don’t know	−0.530
PSB should be stronger on web: - -	9.613	PSB should be stronger on web: - -	5.262
PSB should be stronger on web: -	−0.872	PSB should be stronger on web: -	−3.433
PSB should be stronger on web: +	−2.559	PSB should be stronger on web: +	−4.624
PSB should be stronger on web: + +	−5.005	PSB should be stronger on web: + +	3.481
PSB should be stronger on web: Don’t know	−2.089	PSB should be stronger on web: Don’t know	0.594
Internet access human right should be pub. funded: - -	8.966	Internet access human right should be pub. funded: - -	3.491
Internet access human right should be pub. funded: -	2.293	Internet access human right should be pub. funded: -	−3.271
Internet access human right should be pub. funded: +	−5.454	Internet access human right should be pub. funded: +	−3.404
Internet access human right should be pub. funded: + +	−4.197	Internet access human right should be pub. funded: + +	4.243
Internet access human right should be pub. funded: Don’t know	0.592	Internet access human right should be pub. funded: Don’t know	−0.553
PSB important independent journalism: - - (incl. don’t know)	10.461	PSB important independent journalism: - - (incl. don’t know)	7.850
PSB important independent journalism: -	5.206	PSB important independent journalism: -	−0.071
PSB important independent journalism: +	4.657	PSB important independent journalism: +	−2.685
PSB important independent journalism: + +	−12.490	PSB important independent journalism: + +	−2.291
PSB should not be broad: - - (incl. don’t know)	−3.393	PSB should not be broad: - - (incl. don’t know)	2.128
PSB should not be broad: -	−3.934	PSB should not be broad: -	−4.069
PSB should not be broad: +	4.905	PSB should not be broad: +	−0.542
PSB should not be broad: + +	6.223	PSB should not be broad: + +	3.488
Commercial media should work against false info: - -‎/- ‎(don’t know)	5.226	Commercial media should work against false info: - -‎/- ‎(don’t know)	5.598
Commercial media should work against false info: +	2.750	Commercial media should work against false info: +	−1.332
Commercial media should work against false info: + +	−5.403	Commercial media should work against false info: + +	−1.832
Commercial media should take demo. resp: - -	7.262	Commercial media should take demo. resp: - -	5.885
Commercial media should take demo. resp: -	3.829	Commercial media should take demo. resp: -	−0.836
Commercial media should take demo. resp: +	−1.657	Commercial media should take demo. resp: +	−5.106
Commercial media should take demo. resp: + +	−5.187	Commercial media should take demo. resp: + +	2.879
Commercial media should take demo. resp: Don’t know	1.069	Commercial media should take demo. resp: Don’t know	0.930
Citizen journalism is sufficient: - -	−6.595	Citizen journalism is sufficient: - -	3.114
Citizen journalism is sufficient: -	0.330	Citizen journalism is sufficient: -	−3.499
Citizen journalism is sufficient: +‎/+ +	6.459	Citizen journalism is sufficient: +‎/+ +	0.316
Citizen journalism is sufficient: Don’t know	1.115	Citizen journalism is sufficient: Don’t know	−0.087
Journalism not important today: - -	−8.643	Journalism not important today: - -	1.412
Journalism not important today: -	5.665	Journalism not important today: -	−1.539
Journalism not important today: +‎/+ +	5.568	Journalism not important today: +‎/+ +	−0.253
Moderates, Christian Democrats	14.376	Moderates, Christian Democrats	−2.205
Left Party	−13.154	Left Party	8.483
Sweden Democrats	9.714	Sweden Democrats	5.011
Centre Party, Liberals	−0.487	Centre Party, Liberals	−7.143
Social Democrats, Green Party	−12.000	Social Democrats, Green Party	−2.376