The Complicated Relation Between News Frames and Political Trust


The Complicated Relation Between News Frames and Political Trust: A Case Study of Romania


Madalina Botan, Nicoleta Corbu, Dani Sandu[1]



Abstract: This study tests through an experiment the hypothesis that heavy emphasis on conflicts in the news undermines political trust and has an impact on citizens’ cynicism and political participation. Findings do not support a uniform negative impact of the conflicts covered in the news but demonstrate a cumulative effect of their levels of intrusiveness and incivility. In addition, we found consistent evidence that personal characteristics influence people’s reactions to conflicts. For extrovert individuals, who are typically more assertive in public matters, have higher levels of political knowledge and feel more politically efficacious, exposure to conflict frames does not necessarily result in lower political trust.

Keywords: Conflict Frames; Political Trust; Issue Obtrusiveness; Civility; Extroversion, Political Knowledge; Political Efficacy


1. Introduction[2]

Media coverage of politics in Romania is increasingly conflict-oriented, reflecting the growing internal political turmoil. Politics in Romania over the past years have been marked by domestic conflict and allegations of abuse of power. Romanian politicians have been widely accused of power abuses and ignoring the nation’s needs while focusing on their conflicts with each other.[3] The internal political conflicts and the way they have been framed by the Romanian media have been reported by various journalistic institutions and international NGOs.[4]

Building on this, we test through an experiment whether the exposure to such high levels of political conflict covered in the news necessarily leads to public mistrust and apathy towards political participation or if there are other variables, which can help us understand how the political controversies are linked to the public's representations of political leaders and institutions.

In addition, we wanted to add a more fine-grained understanding of the impact of the political conflicts routinely covered by the media; therefore we have tested two variables which have not been previously analysed in direct relation to the impact of journalistic coverage of politics on political trust:[5] issue obtrusiveness and the valence of political conflicts over contentious issues (civil vs. uncivil debates and controversies). We have anticipated that these two variables might influence the way people react to political disputes. Thus we have included measures of ‘issue obtrusiveness’ and ‘civility of debate’ in our experimental conditions in order to find an accurate scale for measuring the effects of political conflicts, as framed in the news. The main goal of this study has been to provide a theoretical understanding of news frames revolving around political conflicts and to advance an experimental design that could potentially isolate their differential impacts on political trust.

Lastly, we believe that personality traits can help researchers understand the differential effects that media have on individuals. In our experiment, we have tested a set of personality characteristics, such as introversion - which have been tested in correlation with public behaviour of social interaction but never in relationship with political trust – and political knowledge and efficacy: traditionally linked to political participation (Cappella and Jamieson 1997; Mutz 2002; Avery 2009) in order to better disentangle their impact on how people process and use information. Testing for psychological and personality characteristics is still very new in the field of media effects research. However, we consider that this is a very fertile research ground which will receive increasing scholarly attention in the near future. Our findings are modest, but indicate a very rich and diverse palette of cognitive and emotional reactions towards political information.

The trend of increasing coverage of political conflicts in the Romanian media deserves to be better scrutinised, especially since in other European countries an opposing trend seems to be prevalent. Various authors have noticed a decline of conflict-oriented news in Germany and in the Northern countries, such as the Netherlands and Denmark, where the presence of political conflicts in the news peaked in the 1990s or the early 2000s, and has dramatically declined ever since (Takens 2013; Zeh and Hopmann 2013). News focused on conflicts between journalists and politicians or on conflicts among politicians seem to be a distinctive trait of newscasts in certain countries, such as Spain and France, while in other European countries (the Northern countries and Germany, for example) there are very few political conflicts and controversies covered in the media (Maurer and Pfetsch 2014). Trying to understand if the predominance of political conflicts in the news is a country characteristic and a metaphor for the local political culture, we have started by having a closer look at journalistic practices and the prevalent frames used when covering politics. Our main goal – as previously stated – was to identify if the particularities of covering politics in the media influence the overall political trust as well as to see if there is a differential impact of the political conflicts prioritised by the media and if certain people (based on their personality characteristics) respond differently when exposed to political conflicts, such as when they are commonly filtered and framed by the journalistic practices.[6]

Framing of information in the news was demonstrated to consistently affect the direction of public opinion (Chong and Druckman 2007; Druckman 2001; Iyengar and Kinder 1987/2010; Schuck and de Vreese 2006; Sniderman and Theriault 2004). A news frame points out ‘the essence of an issue’ and suggests ‘what the controversy is about’ (Gamson and Modigliani 1987: 143). When citizens are repeatedly exposed to a certain frame, they internalize it and use it to understand the issue at hand (Entman 1993).

The fact that one of the most common frames within contemporary political journalism revolves around conflict has consistent consequences on citizens’ political support (Cappella and Jamieson 1997; Iyengar 1991; Schuck and de Vreese 2012). Conflict frames are organized around disagreements over issues of public interest or political clashes and manoeuvres (Gross and Brewer 2007) and are traditionally more unsubstantiated than frames that focus on the stakes of the discussed matters (Bennett 2006).

From a journalistic perspective, issues that involve conflicts or disputes are generally considered newsworthy, and therefore they are more likely to get media attention (Tuchman 1978). Beyond newsworthiness, conflict-driven information is more appealing to the public, and thus it has even more chances to be picked up by the media (Graber 2010). In the process of covering political conflicts or disputes, journalists should normally try to seek a balance between competing positions. However, journalists are rarely neutral transmitters of information; therefore, even if presumably unintended, the use of conflict frames is ‘politically charged in practice’ (Gross and Brewer 2007: 130). In other words, journalists suggest particular themes and interpretations by which conflicts tend to be understood through their choice of conflict framing. Another problem with this approach to covering politics is that the substantial issues under debate are often relegated to the periphery.

Existing research dedicated to political conflict looks almost exclusively at the impact on voter turnout (Brooks and Geer 2007; Kahn and Kenny 1999; Mutz 2002; Min 2004). Correlated to this, heavy emphasis on conflict also seems to be responsible for the erosion of public trust in politicians and political institutions (Hibbing and Theiss-Morse 1995), as well as for citizens’ cynicism and disengagement with politics (Avery 2009; Cappella and Jamieson 1997). Even if the majority of the empirical research dedicated to media effects was undertaken in the United States, recent European research dedicated to this topic also validates the correlation between negative media coverage of politics and institutional mistrust (Hakhverdian and Mayne 2012). An interesting finding from such cross-national research is that education seems to be a key characteristic influencing political trust, but with a different valence depending on the country: while in Eastern-European countries the greatest political mistrust is exhibited by the most educated, in various Western countries, the less educated have less trust in the government and political institutions (Hakhverdian and Mayne 2012). A possible explanation might again be related to local journalistic practices. Eastern European media is more focused on political conflict and disparity, therefore citizens who have higher incomes are more media savvy and politically knowledgeable, understand in more substantial ways what the controversies are about and, consequently, become more sceptical towards politics in general, in comparison with their poor, less educated compatriots, who do not understand the stakes of the political games of power and manoeuvres and are, therefore, more insensitive to the political information covered in the media. In Western countries, where the media coverage of politics seems to be more consensus oriented and less prone to underlining the disparities among various political entities, the better educated seem to understand the necessity of opposing positions as part of a healthy, democratic debate, while the lower-income, less educated citizens are more prone to distrusting politicians and political institutions, misinterpreting the debates as signs of an inefficient political system.


2. Conflict frames and their differential impact on political trust: a moderated relation

Coming back to the local context in Romania, we wanted to test if the relevance of conflicts between politicians over contentious issues in Romanian media can alter citizens’ trust in political institutions and political parties, both reunited in an index of ‘political trust’. Scholars have only recently begun to examine the possibility that different conflicts portrayed in the media may affect citizens’ political attitudes and behaviours in different ways. There is consistent evidence that people distinguish between useful negative information presented in an appropriate manner and irrelevant or harsh mudslinging (Kahn and Kenny 1999).

It is difficult to find encompassing taxonomies and typologies of political conflicts. Some authors consider the ‘harshness’ of the tone a reliable criterion (Kahn and Kenny 1999); others have suggested the ‘fairness’ versus ‘unfairness’ of attacks, or the evaluation of the employed argument or idea (Jamieson 1992). Different concepts are deployed in these studies, usually for measuring the same features, and the majority of authors, with few exceptions (Geer 2006), do not make a clear distinction between ‘conflicts’, ‘attacks’, ‘contrast’ or ‘negativity’. We argue that such distinctions are crucial in understanding the differential role conflict plays in shaping political trust. Conflicts portrayed in the news vary according to their civility and debated topics, and audience members presumably respond to this variation. However, no research has tested this hypothesis in order to examine if and how different types of conflicts play a role in shaping political support.

In this study, we explore the obtrusiveness of the political topics and the civil vs. uncivil way in which such topics are debated in the news, in an effort to identify the differential impact of the conflicts portrayed in the news on political trust. We argue that political conflicts have different identifiable effects on the audience’s level of political trust depending on how they are framed in the news. We maintain that the civility of the political conflicts presented in the news and their level of obtrusiveness influence political trust amongst media audiences. In trying to draw boundaries between particular aspects of conflicts between politicians over contentious issues we differentiate between: A) conflict style, which pits civil conflicts (legitimate remarks expressed in an appropriate manner) against uncivil conflicts (accusatory remarks presented in a strident and pejorative manner); and B) conflict topic, which covers obtrusive conflicts (focused on issues of personal relevance, important to audience members) and unobtrusive conflicts (focused on aspects of low-relevance for a certain audience). This taxonomy of political conflicts – as portrayed in the media – has allowed us to isolate the differential impact each of these conflict dimensions has on political trust.

In this context, we wanted to further explore what some of the individual factors susceptible to moderating the impact of conflict frames on political trust were. In this attempt we have tested three variables: political knowledge, political efficacy and introversion.


a. Moderation variables for media effects on political trust

Negative information seems to command more attention, and attacks are considered to have the highest level of negativity, mainly because attacks tend to be personal, and more unfair than substantive criticism (Geer 2006: 64). Other scholars (Geer 2006; Kahn and Kenny 1999; Min 2004) have distinguished between campaign ads that focus on personal traits versus those that focus on policies. Even though they have reached different conclusions, such authors seem to agree that personal attacks are less relevant to political life and have more potentially detrimental effects than policy attacks. Using the policy versus personality dichotomy, some authors have contended that the former is more effective since it is perceived as more legitimate (Gross and Brewer 2007).

Conflict frames are likely to decrease political trust because they emphasize the disagreements between political parties and, in a broader way, underline the perceived inefficiency of political institutions (Hibbing and Theiss-Morse 1995). Since a focus on conflict in the political news is not likely to consolidate the image of a cohesive and efficient political institution, we expect conflict frames to have a strong negative effect on political trust. Therefore, we expect that:


H1: Conflict frames have a negative impact on political trust.


Furthermore, we argue that the differences go beyond the distinction between personal and issue-based attacks, and that the manner in which conflicts are presented can make a substantial difference in how political conflicts are understood by the audience. There is consistent evidence that dishonest, unprincipled criticism makes citizens uncomfortable with the presented information (Hibbing and Theiss-Morse 1995), and therefore leads to increased cynicism and lower political interest (Cappella and Jamieson 1997).

In addition, political conflicts portrayed in the news are likely to decrease political trust because they emphasize the disagreements between political parties and, in a broader way, underline the perceived inefficiency of political institutions (Hibbing and Theiss-Morse 1995). Since a focus on conflict in the political news is not likely to consolidate the image of a cohesive and efficient political institution, it is not counter-intuitive to expect conflict frames to have a strong negative effect on political trust.

The public’s cognitive and affective reactions depend also on the topic of the news stories, therefore the conflict topic and, implicitly, its capacity to engage the public, is crucial when trying to evaluate the impact of conflict-driven information. Obtrusive topics might be more likely to draw attention and to trigger a reaction than unobtrusive topics. An issue is obtrusive if the public has direct experience with it, or unobtrusive if the public has no direct contact with it. Domestic economic issues such as inflation and unemployment are often cited as examples of obtrusive issues, whereas foreign affairs are considered to be an unobtrusive issue. Framing effects are presumably stronger for unobtrusive issues, because individuals rely more on media information about those topics.


Combining both conflict style and topic, we hypothesize that:


H2: Conflicts presented in an uncivil manner about obtrusive topics will depress political trust to a larger extent than conflicts phrased in a civil manner about unobtrusive topics.


3. Individual characteristics moderate the impact of the frames present in the news

Recently, more fine-grained distinctions among the effects of news frames have moved away from content features toward variations of personal characteristics, since individuals are not homogenous in their knowledge of or motivation to think about an issue, and, therefore, they are not ‘equally receptive or responsive to the same messages’ (Chong and Druckman 2007: 7).

As shown by past research, news frames resonate with pre-existing political knowledge and reinforce partisan attitudes (Valentino et al. 2011). Identifying prior knowledge is thus an important step in exploring the impact of conflict framing on political trust. Extensive literature indicates that people filter media messages through their own political views – if any – (Zaller 1992); as a result, political knowledge together with other individual characteristics shape how citizens respond to news frames (Iyengar 1991). Existing knowledge is a key variable, especially since individuals, when trying to make a judgment about certain issues, rely the most on their prior knowledge regardless of the nature of the analysed issue (Nelson et al. 1997). When audience members are exposed to new information for which they do not have a knowledge base, they are more likely to be influenced by news frames, since they do not have an existing opinion on the presented matter. In line with these arguments, we assume that when citizens have greater political knowledge, conflict framing will not necessarily result in lower political support. In contrast, less knowledgeable individuals will be more stongly influenced by conflict-oriented political news.

Next, the second moderating variable we have tested is extroversion, a trait associated with assertiveness and enthusiasm, which has been positively correlated to political participation (Gerber et al. 2011). Conflict-driven information seems to trigger different reactions depending on people’s general reluctance to publicly express their opinions. Recent research suggests that people with high levels of extroversion are more likely to become politically involved (ibid.). Since there seems to be a significant positive relationship between extroversion and various forms of political participation (ibid.; Vecchione and Caprara 2009), we believe that extroverted individuals, who are typically more assertive to public matters, will not be negatively influenced by conflict frames to the same extent as introverted individuals. 

The political impact of extroversion has been studied in conjunction with another classical moderator: political efficacy – our third moderating variable. In fact, there is consistent proof that the link between extroversion and participation goes through political efficacy (Gallego and Oberski 2012; Vecchione and Caprara 2009).

Political efficacy refers to the feeling that one’s participation can actually make a difference and that elected politicians serve the public interest (Delli Carpini 2000: 396). As demonstrated by previous research, exposure to negative or cynical portrayals of politics reduces citizens’ sense of efficacy (Ansolobahere et al. 1999; Valentino et al. 2011). However, this seems to be less the case with individuals high in political efficacy, because they feel they understand how to take part in politics, and are not intimidated by the challenges, conflicts or disagreements that occur in that arena (Valentino et al. 2011: 308). Along this line, we expect conflict frames to have a differential impact on individuals according to their political efficacy. For less politically efficacious individuals, the negative impact of conflict frames might be higher, while for politically efficacious individuals it might have a more modest effect.

Building on such findings, we argue that the main variables which moderate the effect of conflict frames are prior political knowledge, political efficacy and extroversion. In order to test this assumption, we have formulated the following hypothesis:


H3. Political knowledge, political efficacy and extroversion will influence the level of political trust, based on the type of conflict frame people are exposed to.


Specifically, we expect political efficacy to influence the level of trust in politics and politicians in general; at the same time, we expect political trust to be higher for extroverted than for introverted people, but we also believe there are differences in intensity and even in the level of significance of the correlations between variables, based on the type of conflict frame in the news item a person reads. To better understand their influence, we build two regression models predicting variations in political trust (dependent variable), based on the type of conflict exposure (uncivil obtrusive vs. civil unobtrusive conflicts).

In summary, we maintain that political trust fluctuates as a consequence of exposure to conflict-driven information (H1). We expect conflict frames to negatively impact political trust especially when they are formulated in an uncivil manner around obtrusive topics (H2). In the second part of this paper, we advance the prediction that conflict-driven information has a differential effect on political trust depending on the political knowledge, political efficacy and extroversion level of the audience members (H3).


4. Method

a. Participants and Design

The experimental design allowed for the testing of hypotheses regarding the impact of the conflict frames. In order to test the combined effect of civil vs. uncivil and obtrusive vs. unobtrusive types of frames, we built a 2x3 experiment, plus a control group. There is constant proof among researchers that student-based sampling is a fairly good sampling choice, with a higher potential for generalizing results (Nelson et al. 1997). However, the high homogeneity of the sample raises problems of representativeness, and does not offer much variation in socio-demographics. We will address these limitations in the discussion section. A sample of 2nd and 3rd year undergraduate students (N=617) was used in the experimental design. The sample was composed of 517 women (83.8%) and 100 men (16.2%), aged between 19 and 25 years old. In terms of political preference, more than 50% (53.8%) did not mention a political preference (don’t know/no answer or undecided or would not vote). Six hundred and seventeen undergraduate students in a medium size university in Romania were randomly assigned to the 7 stimulus conditions: Obtrusive and unobtrusive topics treated in a civil or uncivil manner, a non-conflict (neutral) presentation of both obtrusive and unobtrusive topics, and finally a control group which was not exposed to any kind of stimulus.[7]

Before participating in the experiment, students signed a consent form informing them that they would each read a newspaper article and fill out a questionnaire asking how they feel about the article and about politics in general.

For the selected population (students), the obtrusive topic was related to an increase in tuition fees, while the unobtrusive topic focused on agricultural subsidies granted by the Government. The experiment was conducted in November 2012 during the election campaign in Romania, a period of time in which the level of interest in news in general, and particularly in political news, was presumably higher than in non-election periods.


b. Experimental Stimuli

The stimuli for the six conditions were newspaper-like articles, featuring fictitious political characters (the same in every condition). In every treatment message, a fictitious opposition leader was presented as attacking the Government members’ positions on either tuition fees or agricultural subsidies. Manipulations varied in terms of the tone of criticism (i.e., civil versus uncivil) and its topic (i.e., obtrusive versus unobtrusive).[8] We therefore systematically manipulated the topic and extent of civility, without altering the political content of the exchanges. We expected that when obtrusive topics were presented in an uncivil fashion, they would encourage more negative, distrustful attitudes toward politics and politicians.

Consistent effort was put into ensuring that the reporting style of the experimental stimuli is as close as possible to the style routinely used in day-to-day political news in Romania. The stimuli were pretested on a sample of 36 students, and minimally adapted, based on some minor comments.


c. Procedures

The pre-test/post-test questionnaire and stimuli were pre-tested on a small sample of 36 students, who were not included in the final sample. The pre-test questionnaire included the measurements of the moderators considered for this study: political knowledge, introversion, and efficacy. The subjects first completed the pre-test part of the questionnaire, then read the ‘newspaper article’ containing one of the framing conditions, and then completed the post-test part of the questionnaire. The post-test questionnaire included the dependent variable, political trust, as well as other variables considered for further testing.

Three scales were slightly adapted using classical wording to measure the moderators considered in this experiment: political knowledge, introversion and political efficacy. Another scale was used to measure political trust. We will further detail the final items included in each of the four scales.

Political knowledge was measured using 7 multiple-choice questions testing for basic political information about the Romanian political life, from very simple questions, such as the name of the prime minister, to more complex ones, such as the procedure in case of a president’s death, etc. The answers were binary coded as dichotomous variables and the composite variable measuring the level of political knowledge resulted as the mean of the 0 or 1 point ‘obtained’ for each wrong or right answer, with ‘0’ representing the lack of political knowledge and ‘1’ a very strong political knowledge. The internal consistency of the scale (Cronbach’s alpha = .523) was considered high enough for a binary variables scale, even though below the .600 limit usually used for dichotomous variables scales (see Semetko and Valkenbourg, 2000).

The level of introversion was measured using the classic Goldberg scale of 10 items (4 coded in reverse order), measured on a 5-step Likert scale (Goldberg 1999; Goldberg et al. 2006). Cronbach’s alpha of .842 argued for a reliable scale. For this scale, ‘1’ represents the introvert extreme of the scale and ‘5’ the extrovert extreme.

Political efficacy was measured using 3 questions usually used to test for people’s opinions about how much they can influence politics in general. Subjects were asked to estimate (on a 7 point scale) how much they agree with the following statements: ‘People like me have no influence on the decisions politicians make’; ‘I don’t believe the government cares too much about what people like me care about’; ‘The government doesn’t give too much attention to problems people like me have’. The internal consistency of the scale was measured using Cronbach’s alpha (alpha = .761). The scale was adapted from classical scales used to measure political efficacy (Gerber et al. 2011; Gallego and Olberski 2011; Schuck and de Vreese 2011). For the efficacy scale, ‘1’ represents people’s trust that they can change or have an influence on or at least matter to the government, while ‘7’ stands for a general sense of powerlessness people associate with politicians and politics in general.

The dependent variable, political trust, was measured using a 4-question scale, adapted from the scale used by Schuck and de Vreese (2012). Thus, the wording of the four questions, measured on a 7-point scale of level of agreement, were: ‘Politicians generally have good intentions’; ‘When politicians make public statements, they are usually telling the truth’; ‘Most politicians can be trusted to do what is right’; ‘Despite what some people say, most politicians try to keep their promises, and work for the benefit of the people’, with ‘1’ meaning a lack of trust in politicians and politics in general and ‘7’ a high level of trust. Cronbach’s alpha (alpha = .817) showed an internally consistent scale.


5. Results

When testing the impact of conflict news frames on political trust, the first hypothesis was not validated. Differences between conflict frames vs. non-conflict frames are not statistically significant (independent T tests not significant). However, when comparing means of political trust across experimental subgroups exposed to conflict news frames vs. non-conflict frames and no treatment, data show a trend of increasing/decreasing (variations in the level of) political trust. These levels are not statistically significant, but the consistency of the identified trends give us reasons to believe that, with a stronger stimulus, such a hypothesis could be validated.

Table 1. Variations of political trust among experimental subgroups




Std. deviation

Control Group




Non-conflict frame




Conflict frame









The hypothesis has been invalidated, but there is a consistent basis for arguing that frames revolving around conflict in political news lead to a decrease in political trust and vice-versa (H2). The level of political trust reported by the groups subjected to conflict news frames is smaller than the one reported by the control group which, in turn, is smaller than the political trust reported by the non-conflict frame group. In short, this proves that people subjected to non-conflict news frames tend to report a higher level of political trust than those who are subjected to conflict frames.

When comparing responses to the conflict portrayed in the news, clear differences can be observed in the ‘extreme’ treatment conditions (unobtrusive civil vs. intrusive uncivil conflicts). Exposure to these treatments shows a significant difference in levels of political trust (t[174] = -2.286, p<.01). The mean of the value of political trust in the uncivil intrusive treatment (M = 2.12, SD = .98) is significantly lower than the mean of the value of political trust in civil unobtrusive treatment (M = 2.47, SD = 1.02).

Trying to explain such variations, we can argue that people exposed to a treatment of uncivil interaction about an obtrusive topic display a lower level of political trust than people exposed to a treatment of civil interaction about an unobtrusive topic. Therefore, we can clearly state that the nature of a political conflict covered in the media (civil or uncivil) has a more consistent impact on influencing political trust than the bare existence of a conflict. People seem to relate very differently to conflicts depending on their civility valence. Conflicts that are debated in a civil manner (regarding unobtrusive topics) elicit the public to score higher on the political trust scale. In contrast, uncivil conflicts (regarding obtrusive topics) make the public less trusting of politics in general.

This finding casts a new light on the debate about the influence of conflict in the news in general. Our data suggests that it is not only the conflict that matters, but more specifically the type of conflict to which people are exposed that has a strong influence on their level of political trust. Without a doubt, political campaigns are dominated by conflicts, by their very nature as ‘horse races’, therefore political campaign staff and politicians should pay more attention to the tone of the conflicts and the topics around which they are built.

Moving on to the third hypothesis (H3), we attempted to isolate the impact of personal characteristics on how people respond to media frames focused on conflicts; findings validated the existence of a mix of variables with potential impact on political trust. In order to understand the impact of our selected moderators on political trust we ran two regression models within each of the ‘extreme’ subgroups.


Table 2. Regression model predicting political trust (unobtrusive civil conflict)



Std. error






Political knowledge








Introvert Extrovert Scale




Note: Dependent Variable = Political Trust


The advanced model has enough explanatory power (Adjusted R square = .481) in order to argue for the choice of the most important moderators. A limit related to the homogeneity of the sample used prevented taking into consideration the socio-demographic characteristics. We acknowledge that, in further research, a good variability in the socio-demographic attributes of the subjects used in the experiment would allow for a more robust predictive model.

Based on our data, we have identified two significant predictors of political trust: political efficacy and extroversion, the former being the strongest. The more politically efficient a person is, the more he/ she trusts the political scene, politicians and their decisions. In other words, the more people believe they matter to politicians and that their voice can be heard, the more they trust politicians and politics in general. Moreover, the more extroverted a person is, the more he/she shows political trust. (Table 2)


Table 3. Regression model predicting political trust (obtrusive uncivil conflict)



Std. error






Political knowledge








Introvert Extrovert Scale




The second regression model identifies three significant predictors: efficacy, extroversion, and political knowledge (in this order). The strongest one is once again political efficacy. The explanatory power of the model is high enough (Adjusted R square = .455) to argue for the choice of the most important moderators. This second model confirms the three moderators’ (political efficacy, introversion, and political knowledge) impact on political trust. We emphasize that the political knowledge only becomes a significant predictor for political trust for people exposed to an obtrusive, uncivil conflict frame.


6. Discussion and Conclusion

Various studies have indicated that arguments and conflicts between politicians offer a distorted image of unreliable political institutions and political leaders, careless with public interest (Cappella and Jamieson 1997). As an extension of such studies, our research indicates that the way media frame political conflicts over contentious issues can have a serious negative impact on political trust, especially if the political discussions and debates lack civility and revolve around topics of personal relevance to the audience members. Even though our data are not robust enough to support with statistically significant results the impact that conflict framing (as opposed to non-conflict framing) has on political trust, we do have reasons to argue that future research should confirm the trends we have identified in the matter, namely that conflict frames undermine political trust in general.

Moreover, our data add another nuance to the discussion related to the negative impact of political news frames: if the political debates over contentious topics of public interest are undertaken in a civilized manner, the effect of such news is not necessarily negative. If the audience members have a reasonable amount of political knowledge and also have confidence in their own ability to engage in politics and understand the political game (political efficacy), then they gain even more political trust when exposed to news about political disagreements and controversies. These findings indicate that when citizens have a knowledge base and are more trustful in their own capacity to influence the political game, conflicts among political leaders or parties do not demobilize them or make them less trustful towards politicians. A reasonable explanation is that such individuals understand that there is something at stake at the centre of a political conflict and that political debates are inherent to the political process.

The effects that we could isolate in the experiment are somewhat surprising, if we think of the ephemeral nature of the stimuli (modified short newspaper articles) and, secondly, due to the stability of characteristics such as political trust, which does not traditionally have large fluctuations over time. Sudden changes occur only in exceptional circumstances, for example when a large-scale political scandal takes place. This study thus indicates a surprising malleability of political trust. If simply reading a piece of news can produce a change in terms of confidence in the political class, it is clear that this feature is much more volatile than previously thought. However, this is also a limitation of this study: due to the weak stimuli (a piece of written news, allegedly one of the ‘news of the day’), we could not empirically support the impact of conflict framing on political trust in general. Nonetheless, we argue that the lack of statistical significance is most probably due to the stimuli, given the general direction of the effects that we tested. Moreover, real news and real political controversies – much more virulent and conflict loaded than the articles used as stimuli in this experiment – presumably have an even stronger impact on political trust.

A limitation of the present study is related to the printed articles used in the experimental conditions. The hostility of a written text describing a political conflict is undoubtedly much reduced compared to a video message on the same topic. In a written text, nonverbal and paralinguistic cues, which normally amplify the conflict dimension of a described situation, disappear. A newspaper article cannot therefore trigger the same level of physiological attention as a TV message. The impact on the audience is thereby radically different. In such circumstances, the magnitude of the observed effects is all the more surprising.

At the same time, we have to acknowledge another limitation of the study. This is due to the choice of subjects in the experiment. Even though many exploratory studies use students as participants in experiments, this choice always limits the results to young, educated people’s behaviour. At the same time, the homogeneity of the sample prevented controlling for other personal characteristics, such as age, education, income etc. These first results provide solid grounds for further research, using both stronger experimental treatments and a probabilistic sample of subjects, which will allow for a more thorough examination of the impact of types of conflict on political trust. Even though a moderation analysis in the classic sense of the term was not conducted, we believe political knowledge, political efficacy, and extroversion do moderate the level of political trust in conflict framed media news.

The external validity of the experiment and the generalisation of results beyond the controlled universe of an academic experiment were also addressed. The fairly large number of participants (N = 617) allows us to extrapolate, to some extent, the results to the entire Romanian young population with higher education.

Summing up, we believe that this study has made a modest but valuable contribution to answering a key question: ‘How do political conflicts portrayed in the news affect political trust?’ Our data show that the combined effect of conflict topic (obtrusive vs. unobtrusive) and style (civil vs. uncivil) has a considerable impact on political trust. There are many ways in which we can interpret such results and speculate about the consequences of political conflicts. What is certain is that a high level of political conflicts entails an increased level of mistrust in politicians and political institutions.

However, the present study adds a nuance to the discussion: Not every type of conflict has a negative impact, only the exposure to uncivil conflicts has a negative impact on political trust. In addition, such findings must be correlated with the individual characteristics of audience members. This research tested three variables that moderate the impact of news frames focused on political conflicts: a) political efficacy (confidence that citizens can have an influence on political decisions and that the government cares about the interests of ordinary people); b) political knowledge and c) extroversion. In the case of certain categories of citizens – people who have high political efficacy, are extroverted and more politically knowledgeable – uncivil conflicts do not have a negative effect on political trust.

Returning to the main hypothesis of the research, we found that the way a political conflict is portrayed in the media makes a significant difference in terms of effects. The underlying assumption of this study confirmed by the empirical data suggests the existence of a cumulative effect of conflict-oriented news frames – a combined result of the topic and the civility of the conflicts portrayed in the news. Conflicts in the news may even have a positive effect on political trust if citizens understand what the conflicts are about, what it is at stake. An explanation of this phenomenon - already implied by other studies - relates to the fact that people who have knowledge of the debated topics are who are more extroverted and trust their ability to be part of the public debate ‘understand how and are willing to get involved in politics, without feeling intimidated by the challenges, conflicts or disagreements that may arise on the political scene’ (Valentino et al. 2011: 308).

We believe that our study adds some useful insights into the process of moderated political trust, particularly since such studies still need to gain more popularity beyond the US and Anglophone countries.

Last but not least, our research can also be viewed as radiography of the local journalistic practices which influence to a large extent they way people (who do not have direct contact with the political process) interpret the political world and make evaluative judgments. Building on this, we can argue that media, traditionally inclined to selectively frame extraordinary, dramatic and controversial issues, is not to be blamed for political mistrust and apathy. When investigating the causes of this so-called 'video-malaise', the personal characteristics of the public is a much more fertile direction of research. The inclusion of personality traits, such as extroversion, also represents a step forward in understanding how political information is consumed and used when making political evaluations.

In conclusion, sentiments towards political leaders are increasingly influenced by the media coverage of politics. In Romania, there is a strong bias towards negativity in the media. However, the fact that citizens are disheartened and feel that the domestic conflict between politicians is overshadowing actual policymaking in the country is not only a consequence of the biased media portrayals of political leaders but it is due to a larger extent on the lack of a strong knowledge base and efficacy related to the political process. Healthy democracies rely on such characteristics; they allow citizens to be an active part in the political game. We will not, therefore, join the choir of media critics, who deplore contemporary journalistic practices and see in them the source of all public 'malaise', but argue instead for the necessity to consolidate the democratic processes in the emergent democracies and test new variables in order to see how the academic community can contribute to this process.



Ansolabehere, Stephen, Shanto Iyengar and Adam Simon (1999). “Replicating experiments using aggregate and survey data: The case of negative advertising and turnout”. American Political Science Review 93(4): 901–909. DOI: 10.2307/2586120

Avery, James (2009). “Videomalaise or Virtuous Circle? The Influence of the News Media on Political Trust.” The International Journal of Press/Politics 14(4): 410–433. 10.1177/1940161209336224

Bennett, Lance W. (2006). News: The Politics of Illusion, 7th edition. New York: Longman.

Brooks, Deborah J. and John G. Geer (2007). “Beyond Negativity: The Effects of Incivility on the Electorate.” American Journal of Political Science 51(1): 1–16. DOI: 10.1111/j.1540-5907.2007.00233.x

Cappella, Joseph N. and Kathleen H. Jamieson (1997). Spiral of Cynicism: The Press and the Public Good. New York: Oxford University Press.

Chong, Dennis and James N. Druckman (2007). “A theory of framing and opinion formation in competitive elite environments.” Journal of Communication 57(1): 99–118. DOI: 10.1111/j.1460-2466.2006.00331.x

Corbu, Nicoleta, Boţan, Mădălina (2011). Telepresidents. An Inquiry into a Presidential Campaign. Bucharest:

Cristea, Dumitru, Corbu, Nicoleta. 2014. “Media Framing of European News in Times of Crisis: the Romanian Case.” In: Constantin Brătianu et al. (eds.), Strategic Thinking in a Changing World. Bucharest:, 334–346

Delli Carpini, Michael X. (2000). “Gen.Com: Youth, civic engagement and the new information environment.” Political Communication 17(4): 341–349. DOI: 10.1080/10584600050178942

Druckman, James (2001). “On the limits of framing effects: Who can frame?” Journal of Politics 63(4): 1041–1066. DOI: 10.1111/0022-3816.00100

Entman, Robert M. (1993). “Framing: Toward clarification of a fractured paradigm.” Journal of Communication 43(4): 51–58. DOI: 10.1111/j.1460-2466.1993.tb01304.x

Gallego, Arina, and Daniel Oberski (2012). “Personality and Political Participation. The Meditation Hypothesis.” Political Behavior 34(3): 425–51. DOI: 10.1007/s11109-011-91687

Gamson, William A. and Andre Modigliani (1989). “Media Discourse and Public Opinion on Nuclear Power: A Constructionist Approach.” American Journal of Sociology 95(1): 1–37. DOI: 10.1086/229213

Geer, John G. (2006). In Defense of Negativity: Attack Ads in Presidential Campaigns. Chicago: University of Chicago Press.

Gerber, Alan S. et al. (2011). “Personality traits and participation in political processes.” The Journal of Politics 39(1): 265–287. DOI: 10.1017/S0022381611000399

Goldberg, Lewis R. (1999). “A Broad-Bandwidth, Public-Domain, Personality Inventory Measuring the Lower-Level Facets of Several Five-Factor Models.” In: Ivan Mervielde et al. (eds.), Personality Psychology in Europe. Vol. 7. Tilburg: Tilburg University Press, 7–28

Goldberg, Lewis R. et al. (2006). “The International Personality Item Pool and the Future of Public-Domain Personality Measures.” Journal of Research in Personality 40: 84–96. DOI: 10.1016/j.jrp.2005.08.007

Graber, Doris (2010). Mass media and American Politics. Washington, D.C: Congressional Quarter.

Gross, Kimberly and Paul R. Brewer. 2007. “Sore Losers: News Frames, Policy Debates, and Emotions.” The Harvard International Journal of Press/Politics 12(1): 122–33. DOI: 10.1177/1081180X06297231

Hakhverdian, Armen and Quinton Mayne (2012). “Institutional trust, education, and corruption: a micro-macro interactive approach.” The Journal of Politics, 1(1), 1–12. DOI: 10.1017/S0022381612000412

Hibbing, John R. and Elizabeth Theiss-Morse (1995). Congress as Public Enemy. New York: Cambridge University Press.

Iyengar, Shanto and Donald Kinder (2010 [1987]). News that matters: Television and American Opinion. Chicago: University of Chicago Press.

Iyengar, Shanto. 1991. Is Anyone Responsible?: How Television Frames Political Issues. Chicago: University Of Chicago Press.

Jamieson, Kathleen H. (1992). Dirty Politics. Oxford: Oxford University Press.

Kahn, Kim F. and Patrick J. Kenney (1999). “Do Negative Campaigns Mobilize or Suppress Turnout? Clarifying the Relationship between Negativity and Participation.” American Political Science Review 93(4): 877–89. DOI: 10.2307/2586118

Krosnick, Jon A. and Donald R. Kinder (1990). “Altering the Foundations of Support for the President through Priming.” The American Political Science Review 84(2): 497–512. DOI: 10.2307/1963531

Maurer, Peter and Barbara Pfetsch (2014). “News Coverage of Politics and Conflict Levels: A cross-national study of journalists' and politicians' perceptions of two elements of mediatization.” Journalism Studies 15(3), 339–355. DOI: 10.1080/1461670X.2014.889477

Min, Young (2004). “News Coverage of Negative Political Campaigns. An Experiment of Negative Campaign Effects on Turnout and Candidate Preference.” Harvard International Journal of Press/Politics 9(4): 95–111. DOI: 10.1177/1081180X04271861

Mutz, Diana C. (2002). “The Consequences of Cross-Cutting Networks for Political Participation.” American Journal of Political Science 46(4): 838–55. DOI: 10.2307/3088437

Nelson, Thomas E., Rosalee A. Clawson and Zoe M. Oxley (1997). “Media Framing of a Civil Liberties Conflict and its Effect on Tolerance.” American Political Science Review 91(3): 1055–78. DOI: 10.2307/2952075

Schuck, Andreas R. T. and Claes H. de Vreese (2011). “Finding Europe: Mapping and explaining antecedents of ‘Europeanness’ in news about the 2009 European parliamentary elections.” Studies in Communication / Media (SCM) 1(2): 265–294. DOI: 10.5771/2192-4007-2011-2-265

Schuck, Andreas R. T. and Claes H. de Vreese (2006). “Between risk and opportunity: news framing and its effects on public support for EU enlargement.” European Journal of Communication 21(1): 5–32. DOI: 10.1177/0267323106060987

Schuck, Andreas R. T. and Claes H. de Vreese (2012). “When Good News Is Bad News: Explicating the Moderated Mediation Dynamics behind the Reversed Mobilization Effect.” Journal of Communication 62(1): 57–77. DOI: 10.1111/j.1460-2466.2011.01624.x

Sniderman, Paul M. and Sean M. Theriault (2004). “The structure of political argument and the logic of issue framing.” In: Willem E. Saris and Paul M. Sniderman (eds.), Studies in Public Opinion: Attitudes, Nonattitudes, Measurement Error, and Change. Princeton: Princeton University Press.

Takens, Janet et al. (2013). “Media logic in election campaign coverage.” European Journal of Communication 28(3), 277–293. DOI: 10.1177/0267323113478522

TI (2015). Transparency International Corruption Index – 2015, available at https://

Tuchman, Gaye (1978). Making News. New York: Free Press.

Valentino, Nicholas A., Ted Brader and Eric W. Groenendyk (2011). “Election Night’s Alright for Fighting: The Role of Emotions in Political Participation.” Journal of Politics 73(1): 156–70. DOI: 10.1017/S0022381610000939

Vecchione, Michele and Gianvittorio V. Caprara (2009). “Personality determinants of political participation: The contribution of traits and self-efficacy beliefs.” Personality and Individual Differences 46(4): 487–492. DOI: 10.1016/j.paid.2008.11.021

Zaller, John (1992). The Nature and Origins of Mass Opinion. New York: Cambridge University Press.

Zeh, Reimar and David Nicolas Hopmann (2013). “Indicating mediatization? Two decades of election campaign television coverage.” European Journal of Communication 28(3): 225–240. DOI: 10.1177/0267323113475409

[1] National University of Political Studies and Public Administration, 30 Expozitiei Street, 012103, Bucharest, Romania. Contact e-mail for all three authors:

[2] We thank an anonymous reviewer for his/her careful reading of our manuscript and many insightful comments and suggestions.

[3] There are numerous international organisations which issue annual reports on media coverage of politics in Romania (for instance The Foreign Policy Research Institute, democracy-watch/romania/), freedom of speech in the media or the abuses of the political elite and institutions over the media (Transparency International or Freedom House, to cite the most notable).

[4] For example in the Transparency International Corruption Index; see TI (2015).

[5] Some authors (Brooks and Geer 2007; Kahn and Kenny 1999; Mutz 2002) have already argued for the necessity to find a scale for the debates and political controversies; Civility and obtrusiveness haven't been, however, tested as stand-alone categories in relation to political trust.

[6] We have replicated some of most prevalent journalistic frames in the Romanian media in our experimental stimuli, as identified in our previous research (Corbu and Botan 2011; Cristea and Corbu, 2014).

[7] The treatment message combinations were as follows: 90 subjects in the control group, 87 for the ‘obtrusive civil’ condition, 87 for the ‘neutral civil’, 87 for the ‘obtrusive uncivil’, 89 for ‘unobtrusive civil’, 89 for ‘unobtrusive uncivil’, and 88 for ‘neutral unobtrusive’.

[8] The other two stimuli were neutrally framed, while there was no stimulus used for the seventh condition, the control group.