The Scary World Syndrome: News Orientations, Negativity Bias, and the Cultivation of Anxiety

ABSTRACT

Negativity bias is one of the most salient features of news reporting. According to cultivation theory, this bias can foster anxiety about societal issues among news audiences. The relationship is, however, likely to depend on the audience’s news orientations and the issue under consideration. Drawing on a content analysis of mainstream and alternative news media and a three-wave panel survey, both conducted in Sweden, we examine how general and alternative news orientations relate to egotropic anxiety (worry about being personally affected or harmed) about violent crimes and climate change. The results show that while alternative news media portray violent crimes more negatively than mainstream news media, the opposite is true for climate change, which mainstream news media portray more negatively than alternative news media. Consistent with this finding, alternative news orientation is related to higher levels of anxiety about violent crimes, while general news orientation is related to higher levels of anxiety about climate change, illustrating how people seek information that concur with and thereby maintain or reinforce their beliefs. These results have consequences both for cultivation theory and for our understanding of the role played by mainstream and alternative news media in society.

“You cannot use the media if you want to understand the world.”

Those are the words of the now late Hans Rosling (DR, 2015), professor of International Public Health and co-founder of the Gapminder Foundation. Rosling was known for criticizing the news media’s tendency to focus too much on bad things without giving space for all the progress that in general characterizes the world. Negativity bias is indeed one of the most salient features of the news media’s modus operandi (Esser et al., 2016). Societal problems, conflicts, violence, and disasters are staple ingredients in news media coverage all over the world. Newsrooms have good reasons for prioritizing negative news stories, as research shows that people pay more attention to negative information (Soroka & McAdams, 2015; Soroka et al., 2019). Or, as Rosling describes it, by instinct we give more attention to “the overdramatic worldview” (Rosling et al., 2019, p. 13).

At the same time, the negative news bias has potential adverse—and largely unintended—consequences for both audiences and society more broadly. Research shows that news consumption has negative effects on psychological outcomes, such as causing anxiety and lowering mental well-being (Boukes & Vliegenthart, 2017), and that repeated exposure to negative news may generate even more adverse effects (Hopwood & Schutte, 2017). Experimental research on constructive journalism also shows that exposure to negative, problem-focused news engenders negative emotions, leaving the public feeling depressed, disengaged, and disempowered (Hermans & Gyldensted, 2018; McIntyre, 2019; McIntyre & Gyldensted, 2018; Meier, 2018; Parks, 2019). Negativity bias in the news can in turn induce people to tune out from the news altogether (Toff & Nielsen, 2022), causing or exacerbating problems both for the news media (loss of audiences) and for society (less knowledgeable and engaged citizens) (Skovsgaard & Andersen, 2020). These “unpleasant side effects” (Boukes & Vliegenthart, 2017, p. 137) of news consumption warrant further systematic examination. Therefore, this study examines the relationship between citizens’ news media orientations and the development of anxiety about societal risks and threats over time, comparing two emblematic risk-related issues: violent crimes and climate change.

The study builds on the oldest and most well-known research tradition linking long-term media exposure to the development of negative perceptions and emotions about reality: the cultivation theory (Gerbner, 1998; Gerbner & Gross, 1976). Cultivation studies have since the start in the 1960s primarily focused on how representations of crime in media content relate to the development of negative crime perceptions and fear of crime among the audience (Jamieson & Romer, 2014). In doing so, cultivation researchers established the term mean world syndrome to conceptualize how media exposure to crime over time cultivates a notion of the world as a mean and dangerous place (Gerbner, 1998, p. 185). Our study contributes to this area of research by extending the cultivation approach to include societal risks and threats other than crime and violence. By testing the theory on a broader set of societal risks and threats, we ask if the notion of a mean world syndrome should be amended to a scary world syndrome, including a heightened sense of insecurity and danger to societal risks in general.

In relation to news consumption, cultivation research has historically focused almost exclusively on traditional news media, specifically television (Shi et al., 2019). To make cultivation research more relevant in today’s fragmented, high-choice media environment, we use the concept of media orientations (McLeod & McDonald, 1985; Shehata et al., 2022) to analyze the different ways citizens approach news, distinguishing between general news orientation and alternative news orientation. Although negativity bias is considered a universal feature of the news media logic (Esser et al., 2016), research shows that negativity bias is even more pronounced in alternative media (Holt, 2016). Consequently, we examine whether an orientation to alternative news has stronger effects than a general orientation to news on anxiety about violent crimes and climate change.

The study relies on a three-wave panel survey dataset collected in Sweden in 2018 and 2019 and a content analysis of negativity in the coverage of climate change and violent crimes in mainstream and alternative news media during the same period. In the next section, we elaborate on our theoretical argument before we describe our methodological approach in detail and present the results.

Cultivation research and the mean world syndrome

In media effects research, cultivation studies provide the most longstanding and prolific analyses of the relationship between negative representations of reality in the media and negative beliefs about reality among the public. Questions about fear and anxiety were at the heart of cultivation research from the outset, and Gerbner and colleagues were early to warn about the consequences of the heightened sense of risk and insecurity resulting from the media’s ubiquitous focus on crime and violence: “The consequences of living in a symbolic world ruled largely by violence may be much more far-reaching. Preparation for large-scale organized violence requires the cultivation of fear and acquiescence to power” (Gerbner & Gross, 1976, p. 178).

Television fiction was for long seen as the main source of cultivation messages, but alongside the transition to a more scattered and heterogeneous high-choice media environment, cultivation research began to turn the attention to cultivation effects from a variety of genres, such as news (Hermann et al., 2021; Morgan & Shanahan, 2010). Numerous studies showed significant relationships between news consumption—television news in particular—and negative crime perceptions, as well as fear of being personally victimized (Morgan & Shanahan, 2010). Cultivation research related to other issues than crime or violence was, and still is, much less prevalent. Studies of the cultivation of environmental beliefs and fear of environmental risks, for example, have been scant and focused on the effects of television viewing—with mixed results (Dahlstrom & Scheufele, 2010; Good, 2009; Shanahan et al., 1997).

The cultivation research that was founded by George Gerbner and colleagues in the 1960s focused on aggregate and system-level effects where the “notion is that living in a symbolic environment in which certain types of institutions with certain types of objectives create certain types of messages, tends to cultivate (support, sustain, and nourish) certain types of collective consciousness” (Morgan & Shanahan, 2010, p. 339). Gerbner and colleagues perceived television as the chief source of “repetitive and ritualized symbol systems cultivating the common consciousness of … heterogenous mass publics” (Gerbner & Gross, 1976, p. 174), and they claimed that the worldview of the public was homogenized to align with the televised reality through an ongoing process of mainstreaming (Gerbner, 1998; Gerbner & Gross, 1976). This macrosystems approach was categorically uninterested in individual level selectivity and message processing (Potter, 2014, p. 1016) and critical of methods that are currently in fashion when analyzing media effects, such as experimental designs and panel designs aimed at estimating intra-individual effects.

Our study aligns with the more recent reformist agenda for cultivation research (Hermann et al., 2021), both methodologically by employing a longitudinal research design to measure over time effects and substantially by moving away from the ubiquitous focus on crime and violence. This approach departs from the original macro level focus of earlier cultivation research, making a “movement into micro” (Potter, 2014, p. 1021) where individual level effects are the primary target. Considering how far contemporary cultivation research, such as this study, has come from cultivation theory’s original macro-oriented approach, Potter (2014) rightly questions if today’s research that claims to use cultivation theory has deviated so far from Gerbner’s initial ideas to even be called cultivation studies. Still, we argue, there are specific tenets that set the cultivation framework apart from other research on media effects—tenets, as we describe below, that are also sustaining the present study.

First, the idea that media effects are cumulative and builds up over time in response to exposure to dominant and consistent media messages—in our case, negative news. This foregrounds the presence of maintenance and reinforcement effects on beliefs, rather than conversion effects (cf. Shehata et al., 2021), and emphasizes that self-selection to confirmatory news sources is likely to play a key role in the underlying dynamics (van der Meer et al., 2020). As such, the direction of the relationship between news consumption and beliefs becomes less evident. This idea therefore also advises the use of longitudinal research designs to properly delineate the relationship between media usage and the formation of beliefs over time. Cultivation studies are usually based on cross-sectional methods that provide limited cues about causality, and the lack of longitudinal designs is indeed a recurrent criticism voiced against cultivation studies (Hirsch et al., 1981; Potter, 2014; Shi et al., 2019). Using a longitudinal panel design that appropriately estimates within-person effects of news media usage over time is important, not the least as a recent cultivation study that used a longitudinal panel design showed little support for a causal relationship between media exposure and crime perceptions (Shi et al., 2019).

Second, the notion that the magnitude of effects is related to the amount of accumulated exposure to entire categories of media, rather than examining exposure to specific messages or single news outlets. The cultivation approach emphasizes patterns of media usage or orientations to media categories that offer message systems with similar content. The original cultivation studies made a distinction between light and heavy users of television and measured the cultivation effect as the difference—the cultivation differential—between these groups of users. We adopt a similar line of reasoning and assume the media effect on anxiety about be contingent on the extent people orientate themselves toward news in general and alternative news more specifically. We use the term news orientation to conceptualize to what extent people turn to news as a message system within today’s hybridized and fragmented high-choice media environment. Recent research suggest that broad patterns of news consumption are very stable (Andersen et al., 2022). Thus, rather than inventing increasingly more fine-grained measures of individuals’ exposure to specific news outlets (e.g., Andersen et al., 2016), we devise generalized measures of an individual’s orientation toward news. News orientation considers both the level of usage and the motivation and interest in news in a single measure (cf. McLeod & McDonald, 1985; Shehata et al., 2022). People with a high level of news orientation are, hence, avid consumers of news as well as highly reliant on and interested in news.

Third, the interest in how media exposure affects the audiences’ conceptions of social reality, the very fabric of peoples’ worldviews, rather than examining very specific effects on attitudes (Hermann et al., 2021). The mean world syndrome (Gerbner, 1998, p. 185) is indeed broad in scope, referring to a view of the world as a grim and dangerous place where other people generally cannot be trusted. Our study is motivated by a similar interest in how long-term exposure to risks and threats in the media world sensitizes people to societal risks in general, engendering an apprehensive and pessimistic outlook on society. This focus also motivates the comparison between the two issues under consideration. Although very different in the risks they pose to individuals and society, both violent crimes and climate change are problem-focused and conflict-laden news topics and as such they epitomize many of the tenets of negative news reporting. Both constitute political issues that are very salient on the media agenda. More importantly, both issues (but climate change in particular) saw an increase in news media salience over the course of our data collection. Although politically debated along the lines of current ideological cleavages in Swedish politics—left-leaning partisans emphasize climate change risks, right-leaning partisans emphasize crime risks—we expect that news orientation will have effects on anxiety about both societal issues due to the shared negativity bias in the news.

Cultivating fear and anxiety: Emotional effects of the negative news bias

The main outcome we are interested in is the effects of news orientations on egotropic anxiety about violent crimes and climate change. Egotropic refers to feelings of anxiety or worry about being personally affected or harmed by negative events related to societal risks. It is different than being concerned or worried about the world at large, humanity, or society in general; the term sociotropic anxiety is used to conceptualize the latter (Djerf-Pierre & Wängnerud, 2016; Mutz, 1994). Egotropic anxiety about violent crimes thus refers to an individual’s fear of being victimized or threatened by violence (e.g., robbery, abuse, assault, or rape). Egotropic anxiety about climate change entails an individual’s fear of being directly harmed by a changing climate, such as experiencing disastrous weather events (e.g., wildfires, flooding, or droughts), or dangers related to health and living conditions (e.g., heat stress, spread of diseases, or loss of access to food and water) (IPCC, 2022).

Fear and anxiety are both part of our “defensive motivational system” (Barlow, 2000, p. 1249) but while fear is an immediate emergency response reaction to a concrete threat, triggering a fight-flight response, anxiety is a future-directed response “in which one is ready or prepared to attempt to cope with upcoming negative events” (Barlow, 2000, p. 1249; Clayton, 2020). Being worried about societal threats—such as being the victim of robbery or assault or that flooding or forest fires will destroy your house—is, therefore, often a matter of anxiety. Consequently, many communication studies that set out to study fear are rather studying anxiety (Wirth & Schramm, 2005, p. 12).

The focus of the present study is on the fear-anxiety spectrum of negative emotions (Barlow, 2000), specifically how anxiety is linked to the negativity bias in news reporting. Negativity bias has been identified as a core psychological mechanism when individuals process information such as news: “There is evidence of this negativity bias—or, more broadly, the relative strength of negative over positive—throughout psychology” (Soroka & McAdams, 2015, p. 2). Negativity increases physiological arousal, perceptions, attention, and learning (van der Meer et al., 2020, p. 943), and audiences react emotionally more strongly to negative news when tested in a laboratory setting with techniques that measure actual psychophysiological reactions to news (Soroka & McAdams, 2015).

As a basic psychological mechanism, negativity bias is clearly not just relevant for studies about crime and violence. It has a much wider scope and implications. Doherty and Clayton (2011) argue that there are indirect and vicarious effects on emotions including anxiety, depression, grief, numbness, and apathy associated with observing the impacts of climate change worldwide through media images. Experimental studies on constructive journalism comparing audiences’ responses to traditional, negative problem-focused and solutions-focused news about societal issues found that problem-focused news generally produces more negative emotions (McIntyre, 2019; McIntyre & Gyldensted, 2018; Meier, 2018). Boukes and Vliegenthart (2017) go as far as to claim that the consumption of hard news in general has a negative effect on the development of mental well-being.

Based on the basic premises of cultivation theory and other research connecting negativity bias in the news to emotional outcomes, the first two hypotheses for the study are as follows:

H1:

Individuals with higher levels of general news orientation display higher levels of egotropic anxiety about (a) violent crimes and (b) climate change.

H2:

Higher levels of general news orientation increase egotropic anxiety about (a) violent crimes and (b) climate change over time.

Negativity bias in mainstream and alternative news media

Negativity in the news pertains to both news selection and presentation (Esser et al., 2016; Soroka & McAdams, 2015). Negative news thus involves the inclusion of negative or problem-focused news topics as well as presenting reality in negative terms and/or with the use of alarmistic language. Although negativity bias is often described as an endemic feature of news, there are differences in how mainstream news media and alternative news media portray reality that should significantly impact anxiety.

Holt et al. (2019) describe alternative news media as media actors that position themselves as correctives to the dominant mainstream news media by providing a platform for marginalized and underrepresented voices and trying to influence public opinion according to their agenda. Historically, alternative news media was at first seen as representing a radical, leftist political agenda with close ties to social movements. More recently, many countries have witnessed a rise of alternative news media that promote nationalist and populist discourses, describing reality from an extremist right-wing ideological stance. While left-leaning alternative news media often see themselves in opposition to a commercialized and profit-seeking mainstream media, right-leaning outlets are often criticizing the perceived ideological bias within professional journalism (Ihlebæk et al., 2022).

Of importance to the present study is how mainstream news media compare with alternative media regarding negativity. Holt (2016) compared the content of right-wing alternative news media with mainstream news media in Sweden, examining both topic selection and forms of news presentation. Interestingly, he found larger differences in news presentation than in topic selection. Both alternative news media and mainstream news media favored hard news, which, according Boukes and Vliegenthart (2017), can be regarded as predominantly negative. However, that emphasis was even stronger in alternative news media, specifically for the topics of politics, war, and crime. The amount of crime news in alternative news media was almost double that of mainstream news media, whereas environmental issues such as climate change were largely invisible in the alternative news sources. The most substantial differences were still found regarding style and tone. News in alternative media was to a much larger extent negative in tone and more interpretive, as opposed to descriptive, and personal in style. Altogether, 44% of the news in alternative news media displayed a negative tone compared with 17% in mainstream news media. There is, therefore, empirical evidence for assuming that alternative news media have a stronger negativity bias than mainstream news media, at least in Sweden.

Based both on the pervasiveness of crime news in alternative media and the observed differences in negativity between alternative and mainstream news media, we have the two following hypotheses:

H3:

Individuals with higher levels of alternative news orientation display higher levels of egotropic anxiety about violent crimes.

H4:

Higher levels of alternative news orientation increases egotropic anxiety about violent crimes over time.

We still know very little about how alternative news media report on climate change and whether the negative tonality is the same across different news topics. Thus, we lastly ask:

RQ1:

Does the negative tonality in news about climate change and violent crimes differ between mainstream news media and alternative news media?

RQ2:

How does alternative news media orientation affect anxiety about climate change?

Methods

To examine how news orientations are related to and affect anxiety about violent crimes and climate change over time, individual-level longitudinal data is needed. For this purpose, we utilize a three-wave panel survey conducted in Sweden in 2018 and 2019. To examine how mainstream and alternative news media cover violent crimes and climate change, we further draw on a content analysis of Swedish news media in the same period. Belonging to the democratic corporatist media system (Hallin & Mancini, 2004; Humprecht et al., 2022), Sweden has a high degree of journalistic professionalism, a high demand for news, strong state support for news media, and a low degree of political parallelism. At the same time, however, the demand for alternative news is also comparatively high (Heft et al., 2020). Most of the alternative news media are right leaning, occupied by topics like immigration, crime, and the political correctness of mainstream media (Holt, 2018; Holt et al., 2019; Ihlebæk & Nygaard, 2021). Well-known examples are Fria Tider, Nyheter Idag, and Samhällsnytt, reaching approximately 6% to 8% of the population on a weekly basis (Westlund, 2022).

Panel survey

The panel survey includes three waves and was conducted in collaboration with the Laboratory of Opinion Research (LORE) at the University of Gothenburg.¹ A sample of 3,397 respondents, stratified on gender, age, education, and political interest, was drawn from LORE’s pool of participants. LORE’s sampling procedure is based on probability sampling using both telephone and regular mail during the initial recruitment phase. Wave 1 was fielded from March 22 to April 16, 2018, Wave 2 from December 10 to January 8, 2018, and Wave 3 October 7 to October 28, 2019. Wave 1 included 2,291 respondents (AAPOR RR5: 67%), 1,880 in Wave 2 (59%), and 1,819 in Wave 3 (63%).

As we conceptualize news orientations as the extent people turn to news as a message system within today’s hybridized and fragmented high-choice media environment, we focus on the habitual usage patterns and motivations characterizing individuals’ news consumption. General news orientation was measured in the first wave using the three survey items “I follow news about politics and society every day,” “I usually try to avoid news about politics and society” (reversed), and “I try to follow news about politics and society as much as I can” with a five-point response scale from 1 = Completely disagree to 5 = Completely agree. Likewise, alternative news orientation was measured in the first wave with three survey items, asking the respondents “How often do you use online news websites or social media to follow … ” followed by (1) “News about societal issues not reported by the traditional media,” (2) “News that gives a different picture of social issues than traditional media,” and (3) “News that provides new perspectives on important social issues.” Here, the response categories ranged from 1 = Daily to 6 = Never (reversed). A PCA revealed that the six news orientation items loaded on two different factors, and the items were therefore averaged into two indices ranging from 0 to 1, one for general news orientation (α = .81) and one for alternative news orientation (α = .86).

In psychological research, anxiety has been conceptualized both as an emotional trait and state (Zsido et al., 2020). Anxiety as trait refers to generalized anxiety that is part of an individuals’ personality, but it could influence state anxiety about societal problems across the board. State anxiety can, in turn, be studied as affect, mood, or feeling. Anxiety as feeling refers to a first-person experience associated with emotional processing (Spezio & Adolphs, 2007), and, according to some theories, thus involves a certain amount of cognitive appraisal (Wirth & Schramm, 2005, p. 4). Anxiety as affect is often unconscious and requires measurements of palpable effects on the body, whereas anxiety as feeling requires conscious retrieval and is thus possible to report verbally in an interview or questionnaire. As such, this study examines anxiety as feelings, measured in each survey wave by asking “Looking at your own situation, how worried are you about the following:” followed by (1) “That I will be a subject to violent crime” and (2) “That the effects of climate change will affect me.”

In addition, measures of gender (binary with 1 = female), age, and education were also included in the analysis, as these variables represents fundamental antecedents of both news consumption and perceptions of societal issues. Age was measured with six categories (below 30, 30–39, 40–49, 50–59, 60–69, and 70 or above) and rescaled from 0 to 1. Education was measured with a scale of different educational levels in Sweden, which was recoded to two categories, separating respondents with 1 = a university degree from those with 0 = no university degree. Descriptive statistics for all variables can be seen in the Table A1 in the online supplementary materials.

Content analysis

We know from previous research conducted in Sweden that the level of negativity is most likely higher in alternative news media than in mainstream news media (Holt, 2016). There is, however, a lack of research on the differences in negativity for different news topics. Our content analysis compared two indicators of negativity—the tonality and use of alarmistic language—in the news coverage of violent crimes and climate change in mainstream and alternative news media. The analysis was conducted over a two-year period that largely corresponded with the panel survey and was based on a representative sample of news stories from five mainstream news media outlets and three alternative outlets: Dagens Nyheter (broadsheet), Svenska Dagbladet (broadsheet), Expressen (tabloid), Aftonbladet (tabloid), SVT News (public service television), Fria Tider (right-wing alternative), Nyheter Idag (right-wing alternative), and Samhällsnytt (right-wing alternative). Additional details on the analysis procedures and timeline are included in the online supplementary materials. A proportional sample was drawn for each outlet and issue for each month. The number of articles in the sample was: Violent crimes in mainstream media = 858; Violent crimes in alternative media = 221; Climate change in mainstream media = 900; Climate change in alternative media = 58.

The coding of tonality was based on indications of negative environmental and criminality development, respectively, in the articles. Examples of negative tonality were articles saying that environmental conditions are deteriorating, crime rates are increasing, or policy measures are failing. The coding of alarmistic wording measured if there were multiple usages of strong adjectives and/or emotionally charged words in the news article. Examples of alarmistic language were the usage of emotionally charged words such as “system collapse,” “alarming,” “crisis,” “catastrophe,” “crash,” “disaster,” “fiasco,” “emergency,” or “out of control” when describing the developments.

The content analysis was conducted using a single coder, but an additional person completed the intercoder reliability test for news about violent crimes. The coding process followed Neuendorf’s (2017) suggestions on code training and codebook revision. The codebook was devised to provide the coder with very detailed coding instructions, and it was continuously revised throughout the code training period to ensure that the instructions were straightforward for all coding circumstances. Thereafter, a pilot coding was carried out where the same units were coded twice. This was done to ensure that the instructions were detailed enough to eliminate discrepancies and to make sure that the codebook was reliable and repeatable. The intercoder reliability test for crime news was conducted on 54 articles, yielding a Krippendorff’s alpha of .68 for alarmism (alarmistic wording) and .83 for tonality. For climate change, an intra-coder reliability test was conducted on 50 articles, resulting in an alpha of .69 for alarmism and .69 for tonality. The content analysis is described in more detail in the supplementary online materials.

Analytic approach

Analytically, we present our results in three steps. First, we provide descriptive statistics of news orientations and the patterns identified in the content analysis to identify the message systems that individuals are exposed to. Second, we use multiple linear regression to examine how general and alternative news orientation relates to egotropic anxiety about violent crimes and climate change. As we examine the relationships between each news orientation and anxiety holding the other news orientation constant, we here identify the unique contribution of general and alternative news orientation, respectively. Lastly, we present a parallel growth curve model that estimates two sets of intercepts and slopes based on our repeated measures of egotropic anxiety about violent crimes and climate change (Acock, 2013; Preacher et al., 2008). While the latent intercepts represent initial starting levels of each factor, the latent slopes capture individual linear change trajectories over time. This procedure allows us to estimate how news orientations affect the developments in anxiety over time. To test the sensitivity of our findings, we report analyses based both on the sample participating in all three panel waves as well as models using full information maximum likelihood (FIML) to address missing data and thereby make full use of all information available.

Results

In today’s fragmented high-choice media environment people to varying degrees orientate themselves toward news in general or alternative news more specifically. Figure 1 illustrates the distributions of general and alternative news orientation. As seen from the two plots, the respondents have a high orientation to news in general, whereas the orientation to alternative news is less pronounced and more varied. The two indices are only weakly related (r = .19, p < .001). These results square with available data on news consumption patterns in Sweden, showing that mainstream news media have a very broad reach, while alternative news media reach a much smaller audience (Westlund, 2022).

Figure 1. Distributions of general and alternative news orientation (percent).

Notes. General news orientation is an index based on the following items: 1) “I follow news about politics and society every day,” 2) “I usually try to avoid news about politics and society” (reversed), and 3) “I try to follow news about politics and society as much as I can.” Responses were measured with a five-point response scale from 1 = Completely disagree to 5 = Completely agree. Alternative news orientation is an index based on three items, asking the respondents “How often do you use online news websites or social media to follow … ” followed by (1) “News about societal issues not reported by the traditional media,” (2) “News that gives a different picture of social issues than traditional media,” and (3) “News that provides new perspectives on important social issues.” Here, the response scales ranged from 1 = Daily to 6 = Never (reversed). Both scales were transformed to range from 0 to 1. N = 1,454.

What are the worldviews that people meet when orienting themselves toward news in general and alternative news more specifically? As expected, the content analysis showed that the prioritization of news about violent crimes was immensely higher in alternative news media than mainstream news media. For each published news story about climate change, alternative news media published 3.8 stories about violent crimes. For mainstream news media, the proportion was only 0.9, meaning that there were slightly more articles about climate change than violent crimes.

The results of the content analysis are illustrated in Figure 2. Regarding negativity, the results show that alternative news media were significantly more negative in tonality and more alarmistic in wording when covering violent crimes compared to mainstream news media. Importantly, climate change reporting displayed the opposite pattern. Here, the negative tonality was much more pronounced in mainstream news media. These findings answer RQ1, showing that the negative tonality in the news about climate change and violent crimes does indeed differ between mainstream and alternative news media. For violent crimes, the level of negativity is much higher in alternative news media, whereas climate change is described in much more negative terms in mainstream news media.

Figure 2. Negative and alarmistic news about violent crimes and climate change in mainstream and alternative news media (percent).

Notes. Alarmistic wording entails multiple usages of strong adjectives and/or emotionally charged words in an article, such as “system collapse,” “alarming,” “crisis,” “catastrophe,” and “disaster” etc. Negative tonality refers to when an article describes the issue/problem in negative terms, such as failures to fight crime, rising crime rates, increasing levels of greenhouse gases emissions, environmental destruction etc. Other options for the coding of tonality were “positive,” “neutral,” and “balanced” (both negative and positive). The sample included articles from five mainstream news media outlets and three alternative outlets: Dagens Nyheter (broadsheet), Svenska Dagbladet (broadsheet), Expressen (tabloid), Aftonbladet (tabloid), SVT News (public service television), Fria Tider (right-wing), Nyheter Idag (right-wing), Samhällsnytt (right-wing). The content analysis covered the period March 1, 2018 - June 25, 2020. Number of articles in the sample: Violent crimes in mainstream media = 858; Violent crimes in alternative media = 221; Climate change in mainstream media = 900; Climate change in alternative media = 58. For more information about the sample and the content analysis, see the online supplementary materials.

Next, we examine how general and alternative news orientations relate to egotropic anxiety about violent crime and climate change. We start by examining these relationships in multiple linear regression models focusing on respondents participating in all three waves of the panel survey, as illustrated in Figure 3. These results show that general news orientation is positively related to anxiety about climate change (b = .38, SE = .14, p = .006) but unrelated to anxiety about violent crimes (b = −.18, SE = .13, p = .172), while alternative news orientation is positively related anxiety about violent crimes (b = .71, SE = .11, p < .001) and (marginally significant) negatively related to anxiety about climate change (b = −.20, SE = .11, p = .081). General news orientation is thus related to higher egotropic anxiety about climate change, while alternative news orientation is related to higher egotropic anxiety about violent crimes.

Figure 3. Estimates from multiple linear regressions.

Notes. Illustrates unstandardized b-values from multiple linear regressions with each independent variable ranging between 0-1. Egotropic anxiety about violent crime and climate change are measured on 1-5 scales. N = 1,454.

To get a more nuanced understanding of these relationships, and in particular how news orientations are related to the development of anxiety over time, we lastly present the parallel growth curve model displayed in Table 1. The latent intercepts represent the initial starting value for egotropic anxiety about violent crimes and climate change, respectively, while the latent slopes reflect the linear growth trajectory in anxiety over time. Thus, the model estimates how our independent variables, most importantly news orientations, relate to between-person differences in both levels and changes over time. The findings in Table 1 are based on respondents participating in all three waves of the panel. Overall, this model with linear anxiety slopes has a decent model fit (RMSEA = 0.066; CFI = 0.974; SRMR = 0.02).

Table 1. Estimates from the parallel growth curve model.

	Anxiety about violent crime (AVC)		Anxiety about climate change (ACC)
	Latent intercept	Latentslope	Latent intercept	Latent slope
General news orientation	-.05 (.03)	.00 (.02)	.10** (.03)	.00 (.02)
Alternative news orientation	.14*** (.02)	.01 (.01)	-.05* (.02)	-.03** (.01)
Female	.37*** (.06)	.01 (.03)	.53*** (.06)	.03 (.03)
Age	-.23** (.06)	-.03 (.03)	-.14*** (.02)	.02 (.01)
University	-.04* (.02)	.01 (.01)	.01 (.07)	.01 (.03)
R-squared	.08	.01	.12	.04
Covariances
Intercept AVC / Intercept ACC		.18*** (.03)
Intercept AVC / Slope AVC		-.03 (.03)
Slope AVC / Slope ACC		.03*** (.01)
Intercept ACC / Slope ACC		-.06 (.03)
Model fit
Chi-square (17)		125.941, p < .001
RMSEA		.066
CFI		.974
CFI		.02
N		1,488

Notes. Entries are unstandardized coefficients with standard errors in parentheses.

*p < .05, **p < .01, ***p < .001.

First, we see how the relationships between news orientations and anxiety largely reflect the patterns found in the multiple linear regression models. General news orientation is positively related to initial anxiety about climate change (b = .10, SE = .03 p = .003) but not to initial anxiety toward violent crimes (b = −.05, SE = .03, p = .099). Individuals with higher levels of general news orientation consequently display higher levels of egotropic anxiety about climate change (supporting H1b) but not to violent crimes (rejecting H1a). Importantly, however, the effect of general news orientation on climate anxiety appears sensitive to sample composition as the coefficient is non-significant when full information maximum likelihood (FIML) is used (b = .05, p = .089). Alternative news orientation, on the other hand, is positively related to initial anxiety about violent crimes (b = .14, SE = .02, p < .001) but negatively related to initial anxiety about climate change (b = −.05, SE = .02, p = .019). Thereby, individuals with higher levels of alternative news orientation display higher levels of egotropic anxiety about violent crimes (supporting H3) but lower levels of egotropic anxiety about climate change (answering RQ2). These findings are consistent also using the FIML specification. Full results are available in Table A2 in the online supplementary materials.

In terms of growth over time, we only find one significant relationship with news orientations: alternative news orientation is negatively related to growth in egotropic anxiety about climate change (b = −.03, SE = .01, p = .002). In other words, over time alternative news orientation is related to less anxiety about climate change (again answering RQ2). In contrast, as general news is neither related to increases in anxiety about violent crimes (b = .00, SE = .02, p = .791) nor climate change (b = .00, SE = .02, p = .928) we did not find support for H2a and H2b. Likewise, as alternative news orientation is not related to increases in egotropic anxiety about violent crimes (b = .01, SE = .01, p = .524), we did neither find support for H4. These findings are also consistent using the FIML option and are available in Table A2 in the online supplementary materials.

Finally, the covariances between the latent factors indicate that both the intercepts (COV = .18, p < .001) and the slopes (COV = .03, p < .001) of anxiety about climate change and violent crimes are positive related: people who are anxious about violent crimes are generally also more anxious about climate change—and over time these trends are related.

Discussion and conclusion

Although the negative bias pertaining much contemporary news reporting can draw audience attention (Soroka & McAdams, 2015; Soroka et al., 2019), it also comes with the risk of several undesirable consequences, such as biased perceptions of reality (Morgan & Shanahan, 2010), lower mental well-being (Boukes & Vliegenthart, 2017), and news avoidance (Toff & Nielsen, 2022). For these reasons, audiences’ emotional responses to news have gained increasing attention in media and communication research in recent years. Fear and anxiety about violence, terrorism, war, and crime are a common theme in this recent research, drawing attention to the vast socio-political consequences “a discourse of fear” may have for individuals and society (e.g., Altheide & Michalowski, 1999). Departing from cultivation theory (Gerbner, 1998; Gerbner & Gross, 1976), this study examined how violent crimes and climate change are covered in mainstream and alternative news media along with how general and alternative news orientations are related to egotropic anxiety about these societal issues.

Our content analysis illustrated how violent crimes are covered more negatively in alternative news media, while climate change is covered more negatively in mainstream news media. Additionally, in comparison to violent crimes, climate change is covered less in alternative news media while mainstream news media gives equal amounts of attention to these two issues. As such, news audiences will be exposed to two different media worlds depending on how they orientate themselves in today’s high-choice media environment. Consequently, results from the survey analyses demonstrated how alternative news orientation is related to higher anxiety about violent crimes, while general news orientation is related to higher anxiety about climate change, although not significant in all model specifications. Those differences reflect cultivation differentials between people with lower and higher orientations toward news in general and alternative news, respectively. Thereby, our results support the idea of expanding the concept of a mean world syndrome (Gerbner, 1998, p. 185) to a scary world syndrome. Importantly, however, the magnitude of this syndrome depends on the amount of negativity bias pertaining to people’s news orientation.

Inspired by cultivation theory, our panel analysis focused both on levels of and growth in anxiety over time. While the relationships between news orientations and levels of anxiety reflect a long-term buildup of accumulative media effects, the relationships with growth in anxiety reveal if news orientations are related to any additional outcome changes over time. Taken together, our results support the central idea of cultivation theory that media effects should be seen as slow and gradual, with the news media supporting and sustaining a specific worldview. Maintenance of a certain level of egotropic anxiety is likely to be an outcome of the negative news bias. If people are not reminded about the negative aspects of violent crimes and climate change, they will likely begin to forget and, consequently, become less anxious (Shehata et al., 2023). Importantly, as today’s high-choice media environment offers people ample opportunities to self-select into confirmatory information sources, the result can also be seen as a product of selective exposure to belief consistent news sources.

The general lack of significant relationships between news orientations and over-time increases in anxiety suggest that we are either witnessing a ceiling effect—where pre-treatment hampers additional increases in anxiety—or that the study’s relatively short timeframe of approximately 1.5 years was not long enough to capture potential small and gradual effects of cultivation. Further, it again underlines the relevance of people self-selecting into confirmatory news sources, likely leading to a reciprocal relationship between news consumption and beliefs. Our analysis did show, however, that alternative news orientation is related to an over-time decrease in egotropic anxiety about climate change. This result suggests that while the news media can cultivate a picture of a scary world causing anxiety, it can also work in the opposite direction, easing people’s anxiety through a less negative representation of reality. Important to keep in mind here, however, is that alternative news media in general downplay the threat of climate change and delegitimize climate change action by giving voice to climate deniers and by questioning research.

The findings also raise the question of what the appropriate level of anxiety is. Although fear and anxiety are categorized as negative emotions, they are certainly not altogether bad. To the contrary, they are part of evolution and necessary for survival. As humans, we are predisposed to be anxious to potential risks and threats to avoid harm. Psychologists therefore speak of maladaptive and adaptive levels of anxiety (Clayton, 2020). Anxiety produces a heighten sensitivity and attention to threat, which leads to more deep and effortful information processing (Huddy et al., 2007, p. 206). However, since anxious individuals exhibit a strong “threat bias” (Norris, 2021, p. 72), negativity bias in the news may result in an exaggeration and overestimation of actual risks. High levels of anxiety can result in a person being overwhelmed or paralyzed from worry, which induces inaction as a defensive response rather than productive engagement in concrete actions. Indeed, research shows that there is no significant correlation between climate change anxiety and behavioral engagement (Clayton & Karazsia, 2020). Intense anxiety caused by negative news can thus lead to inaction rather than the opposite. As such, it seems fair to question if the hunt for audience attention has resulted in today’s news being too negative and overdramatic (Rosling et al., 2019).

Our study of course comes with some limitations. Most importantly, although we address a central criticism of cultivation research by employing a longitudinal research design, our study period was relatively short, and as we in general do not find media orientation to be related to increases in anxiety over time, we cannot be sure about the causal direction underlying the identified relationships. Future research is therefore encouraged to follow changes in people’s worldview and their relationships with news orientations over several years, or maybe even decades. Relatedly, although we have expanded beyond violent crimes as a case for studying cultivation effects, we are still only examining two issues. Future research is therefore also encouraged to test the idea of the scary world syndrome on other societal topics, such as, for example, wars, pandemics, and terrorism.

Despite these limitations, the study has demonstrated how people’s orientation toward news in general and alternative news media specifically, and the corresponding amount of the of negative coverage in these two media worlds, are related to egotropic anxiety about violent crimes and climate change. Therefore, the study has illustrated how one of the most fundamental theories of media effects, the cultivation theory, is still relevant in today’s heterogeneous high-choice media environment.

Supplemental data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/15205436.2023.2297829

Disclosure statement

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

Additional information

Funding

The work was supported by the Data collection for this study was funded by the Swedish Research Council (VR 2016– 02262).

Notes on contributors

Kim Andersen

Kim Andersen is Associate Professor at the Centre for Journalism, Department of Political Science, University of Southern Denmark, and Affiliated Researcher at the Department of Journalism, Media and Communication, University of Gothenburg. He does research on people’s news consumption and its consequences for their political knowledge, beliefs, and engagement.

Monika Djerf-Pierre

Monika Djerf-Pierre is Professor at the Department of Journalism, Media and Communication, University of Gothenburg. Her main research fields are journalism, political communication, risk- and crisis communication, and gender and media.

Adam Shehata

Adam Shehata is Associate Professor at the Department of Journalism, Media and Communication, University of Gothenburg. His research focuses on media use as well as media effects on public perceptions, opinion, and political engagement.

Notes

¹ This study was subject of an ethical application and was deemed exempt from full review by the Regional Ethics Board, Gothenburg. The reviewing body gave an advisory statement declaring no objections to the study (Dnr. 2017/1005–17).