ArticlePDF Available

Abstract and Figures

While the discussion on the individual level variables that affect responses to political scandals has focused mainly on variables such as partisan identity or political cynicism, we suggest that media skepticism could also moderate whether and how individuals respond to political scandals. To test this relationship, we rely on panel data from the United States gathered before and after the Trump–Ukraine scandal occurred (Wave 1 in June 2019, Wave 2 in October 2019). Our results show that individuals who rank higher on media skepticism hold comparatively more positive views of Trump after the scandal, even when previous evaluations and alternative explanations are controlled for. Conversely, we find no effect of media skepticism in trust toward the US political system and government. We believe our findings have significant consequences to understanding the relationship between the governed and those governing in times of widespread media skepticism.
Content may be subject to copyright.
International Political Science Review
1 –16
© The Author(s) 2022
Article reuse guidelines:
DOI: 10.1177/01925121211073005
Media skepticism and reactions to
political scandals: An analysis of the
Trump–Ukraine case
Hugo Marcos-Marne
University of Salamanca, Spain
Pablo González-González
University of Salamanca, Spain
Homero Gil de Zúñiga
University of Salamanca, Spain
Pennsylvania State University, USA
While the discussion on the individual level variables that affect responses to political scandals has focused
mainly on variables such as partisan identity or political cynicism, we suggest that media skepticism could also
moderate whether and how individuals respond to political scandals. To test this relationship, we rely on panel
data from the United States gathered before and after the Trump–Ukraine scandal occurred (Wave 1 in June
2019, Wave 2 in October 2019). Our results show that individuals who rank higher on media skepticism hold
comparatively more positive views of Trump after the scandal, even when previous evaluations and alternative
explanations are controlled for. Conversely, we find no effect of media skepticism in trust toward the US
political system and government. We believe our findings have significant consequences to understanding the
relationship between the governed and those governing in times of widespread media skepticism.
Media skepticism, scandals, political evaluations, selective exposure, United States
The effects of political scandals have long interested social scientists working from very different
perspectives, which is hardly surprising, given their consequences can affect the functioning of
*Homero Gil de Zúñiga is also affiliated to Universidad Diego Portales, Chile
Corresponding author:
Hugo Marcos-Marne, University of Salamanca, Facultad de Derecho, Paseo Francisco Tomás y Valiente S/N, Salamanca,
37007, Spain.
1073005IPS0010.1177/01925121211073005International Political Science ReviewMarcos-Marne et al.
Original Research Article
2 International Political Science Review 00(0)
democracies (von Sikorski, 2018). On the one hand, individuals may react negatively to political
scandals, diminishing their support levels toward political institutions and actors. But, on the other
hand, scandals may end up having positive consequences in terms of support, mainly when known
to the public, and as long the politicians involved are punished for their misbehavior (Maier, 2010;
Praino and Stockemer, 2021). While it seems uncontested to say that there is a high variation in how
individuals react to political scandals, there is still room to understand the micro-foundations of this
variation. This paper seeks to advance this literature by introducing media skepticism as an anteced-
ent of reactions to political scandals. Building upon literature that examines how much information
individuals receive about public affairs and how this information is processed (Garrett et al., 2012;
Gil de Zúñiga et al., 2018), we expect high levels of media skepticism to diminish people’s negative
reactions to political scandals. That is, higher levels of media skepticism will lead to less negative
reactions in terms of political trust (buffering effect), after a scandal takes place.
To test this relationship, we use an online survey panel study conducted in the United States
(US). The first wave of data collection began in June 2019, and the second one in October 2019, a
couple of weeks after the Trump–Ukraine scandal made it to the front pages of influential media
outlets. The scandal started with a whistleblower complaint filed on August 12th, 2019, in which
an anonymous intelligence officer showed his concern because Trump may have used ‘the power
of his office to solicit interference from a foreign country in the 2020 US elections’ (Mettler and
Lieberman, 2020). This hint led to a phone call between Trump and the Ukrainian president,
Volodymyr Zelensky, in which Trump used a rhetoric of ‘quid pro quo,’ alluding implicitly to with-
held funding and inviting Zelensky to the White House. In exchange for that, Zelensky would
announce investigations on the potentially corrupt behavior of the son of Joe Biden, Trump’s com-
petitor for the 2020 elections, and allegations stating that Ukraine interfered in the 2016 US elec-
tions by hacking the Democratic National Committee’s network. Utilizing a quasi-experimental
setting, our panel study allows us to observe changes in attitudes after a scandal while controlling
for previous evaluations of actual political actors (i.e., Donald Trump) and institutions (i.e., US
government and political system), something difficult to obtain when experiments with fictional
candidates are applied (von Sikorski et al., 2020).
Our results show that media skepticism influences reactions to political scandals. Individuals
who ranked higher on media skepticism displayed comparatively more favorable views of Trump
after the Ukraine scandal, even when previous evaluations and relevant competing explanations
were controlled for. The effect of media skepticism was not moderated by a key variable in this
subfield, partisan identity, which speaks of the potential strength of our explanatory variable. We
found no effects of media skepticism in trust towards the US political system or the government.
We attributed this to the lack of short-term spillover effects in the case under examination. The
implications of these results are detailed in the discussion and conclusion section, together with
limitations of our study. An important one is that we cannot guarantee our results are solely
explained by the Trump–Ukraine scandal, because we lack explicit questions about it in our panel
survey. Therefore, while our results provide initial evidence for the buffering effect of media skep-
ticism after scandals occur, more experimental studies will be necessary to confirm it.
Political scandals and media skepticism
While political scandals may occur whenever politicians transgress norms, the use of the term in
social science research often includes the communication and information processes through which
‘alleged transgressions or failures of public figures, groups, organizations, or institutions are
denounced with the aim of eliciting public outrage,’ also known as scandalization (Geiß, 2017: 1).
This perspective emphasizes the joint existence of, first, an illegal/inappropriate behavior by a
Marcos-Marne et al. 3
politician; second, media conglomerates that made it known to the public; and third, a public opin-
ion that might change their opinion and even behavior. We derive from this approach that politi-
cians’ behaviors susceptible to being ‘scandalized’ by media are not uncommon, although only a
portion of them make it to the front page.
Research on the effects of political scandals on public opinion is organized along two main
theoretical approaches: the dysfunctional and the functional one (Maier, 2010). The dysfunctional
approach assumes that scandals are harmful to the image of both political institutions and actors.
Its underlying idea is that individuals will react to the inappropriate or even illegal behavior from
politicians with increased skepticism and distrust. Alternatively, the functional theory posits that
reporting political scandals may have a positive effect on how political institutions are perceived.
Far from suggesting that individuals reward political misconduct, what lies at the core of the func-
tional theory is that citizens may end up being more trustful of the system if they perceive miscon-
duct is identified and, essentially, punished. However, both theories agree in predicting that the
image of the politicians involved in a scandal will be damaged (Maier, 2010; von Sikorski, 2018).
Whether negative reactions spread beyond them, operationalized as spillover effect, is dependent
on two intertwined factors: the extent to which individuals identify politicians involved in scandals
as part of a larger network of political institutions, and the strength of these connections, dubbed as
the accessibility–diagnosticity framework (Roehm and Tybout, 2006).
In line with the dysfunctional theory, empirical research has often found that when political
scandals ocurr, overall political distrust increases and the image of the politicians is largely affected
(Brody and Shapiro, 1989; Praino et al., 2013; Stockemer and Praino, 2019). Nonetheless, there are
nuanced multilevel explanations for this phenomenon.
At the aggregate level, literature shows that scandals are likely to trigger more political distrust
in situations of extreme economic uncertainty (i.e., with high inflation or unemployment issues),
and whenever approval rates are low among supporters of opposition parties (Carlin et al., 2015;
Nyhan, 2015). Attitudinal reactions to scandals are also dependent on how often they are reported
in the media (coverage) and their treatment (framing) (Kepplinger et al., 2012; von Sikorski, 2018).
Finally, at the individual level, the relationship seems mostly conditioned by partisan identity
(Dancey, 2012; Wagner et al., 2012), but also by political cynicism and distrust (Dancey, 2012). In
this regard, three main mechanisms have been identified that condition the outcomes of political
scandals at the individual level (von Sikorski et al., 2020). First, scholars have consistently shown
that individuals, by means of media selective exposure, look for information that is coherent with
their beliefs and attitudes (Garrett, 2009). Second, individuals tend to practice selective perception
and are likely to interpret that when exposed to media news, the information received is in agree-
ment with their beliefs (Tsfati and Cappella, 2003). Third, motivated skepticism also drives this
relationship, as individuals more easily assume arguments that are in line with their way of think-
ing while downplaying information that challenges it (Ditto and Lopez, 1992). The combination of
these effects explains why partisan identity strongly moderates the outcomes of political scandals.
Via selective exposure, individuals with a strong partisan identity are more likely to avoid media
outlets and specific news that criticize the politicians/parties to which they feel attached (Stroud,
2010). Due to selective perception and motivated skepticism, these same individuals are more
likely to maintain their perception of the politicians/parties involved in a scandal. This happens
because even if they consume news about the scandal, they will be more prone to downplay atti-
tude-challenging arguments or even to process these arguments in a way that makes them look in
agreement with their beliefs (von Sikorski et al., 2020). This paper suggests similar underlining
mechanisms explain how media skepticism shapes attitude change after a political scandal. To the
best of our knowledge, this connection has not been considered yet, despite its potentially impor-
tant implications in a context increasingly shaped by hostile perceptions of media.
4 International Political Science Review 00(0)
Media skepticism reflects feelings of alienation and distrust towards mainstream media, which
affects perceptions of credibility and reliability, and translates into lower levels of exposure (Tsfati,
2003). In the context of political scandals, we expect this variable to matter because of the crucial
importance of media in making political scandals known to the public, which is well reflected in
the ideas of ‘scandalization’ (Geiß, 2017) or ‘mediated scandals’ (Thompson, 2000). While politi-
cal cynicism is likely to foster negative attitudinal reactions to political scandals (Dancey, 2012),
we expect media skepticism to reduce them, for two main reasons. First, we expect that media
skepticism, via selective exposure, will affect the overall amount of news to which individuals are
exposed. Given that more exposure is associated with a higher likelihood of attitude change (Ernst
et al., 2017), we should see that media skepticism buffers attitude change after a political scandal
takes place, provided that mainstream actors mainly provide the information. Second, even if
media skeptics are exposed to news about a scandal (Tsfati and Cappella, 2003), which is not
unlikely given the complex informational environments citizens live in nowadays, the lack of cred-
ibility assigned to the source will increase the likelihood of selective perception and motivated
skepticism, limiting too attitude change.
These two arguments lead to a common expectation: individuals with higher levels of media
skepticism will change their political attitudes less after a political scandal occurs because they will
be less likely to accept the ‘climate of opinion presented by the media’ (Tsfati, 2003: 69).
Building upon the results of the dysfunctional theory, we assume that political scandals will
harm the evaluation of politicians involved and, to a lesser extent, in the assessment of political
institutions (in this paper, US government and political system). Considering the effects of the
theoretical framework exposed above, we expect the effect to vary due to media skepticism. Hence,
we hypothesize that, after a scandal, individuals who rank higher on media skepticism will evaluate
the politicians involved less negatively (H1a). Also, if attitudinal reactions spread beyond the poli-
ticians involved in the scandal (i.e., if spillover effects exist), we will see that media skepticism
buffers negative evaluations towards political institutions (H1b). Furthermore, and given prior
findings on the crucial association between political scandals and individuals’ partisanship, we also
consider that the effect of media skepticism may be dependent on the latter. The combination of
these two elements may trigger selective exposure, motivated skepticism, and selective perception
in a way that makes the result greater than the sum of its parts. Therefore, we suggest that the buff-
ering effect of media skepticism will be stronger for individuals that identify with the party involved
in the scandal (H2a). If attitudinal reactions spread beyond the politicians involved in the scandal
(i.e., if spillover effects exist), we will see a stronger buffering effect in the evaluations of political
institutions among individuals who identify with the party involved in the scandal (H2b).
Our case study: The Trump–Ukraine scandal
The ‘Trump–Ukraine scandal’ started with a whistleblower complaint filed on August 12th, 2019.
The whistleblower, an anonymous intelligence officer, showed his concern because Trump may
have used ‘the power of his office to solicit interference from a foreign country in the 2020 U.S.
election’ (Mettler and Lieberman, 2020). Subsequent information pointed to a phone conversation
between Trump and the Ukrainian president, Volodymyr Zelensky. Shortly before this call, Trump
had ordered the withholding of US$391 million in military and security aid to Ukraine. In his con-
versation with Zelensky, Trump used a ‘quid pro quo’ rhetoric, alluding implicitly to the withheld
funding and inviting his equivalent to the White House. In exchange for that, Zelensky would
announce investigations on two fronts: first, on the potentially corrupt behavior of the son of Joe
Biden, Trump’s potential competitor for the 2020 elections; second, on allegations stating that
Ukraine interfered in the 2016 US elections by hacking the Democratic National Committee’s
network, opening the frame for Russia’s interference.
Marcos-Marne et al. 5
The Washington Post, a leading outlet in covering political scandals (Nyhan, 2015), played a
key role in making this situation known to public opinion. Its editorial board first reported the situ-
ation on September 5th, 2019, and by September 19th the diary published the whistleblower com-
plaint on Trump’s communications with a (still unknown) foreign leader (Nakashima, 2019). This
event led to President Trump posting several statements on Twitter accusing the whistleblower of
being ‘partisan’ and calling the Washington Post ‘Fake News Media.’ The Washington Post kept
reporting about the scandal on its pages, which subsequently influenced the agenda of other media
outlets (Pramuk, 2019; The New York Times, 2019; The Washington Post, 2019; Wolf, 2019).
Although different framings appeared that put the emphasis on distinct actors involved in the pro-
cess (Blitzer, 2021; Dilanian and Winter, 2020), the importance of the scandal and the attention it
received is beyond doubt. It was covered by virtually all relevant media in the country, even more
since Trump kept accusing Biden of corruption after the revealing of the phone call, which led to
an impeachment inquiry opened by House Speaker Nancy Pelosi (Trump was finally impeached by
the House of Representatives on December 18th, under accusations of abuse of power and the
obstruction of Congress), and it affected public opinion of Trump about this impeachment process,
especially among Democrats and independents (McCarthy 2019). The question is, how did media
skepticism, alone and in combination with partisan identity, influence individuals’ responses to the
Data for this research comes from an online panel survey conducted in the United States in two
waves. The first wave (W1) of the collection started in June 2019 (N = 1,338), and the second one
(W2) in October 2019 (N = 511), a few weeks after the Trump–Ukraine scandal was made known
to the public (a comparison of aggregated socio-demographic features in the two waves is in the
online appendix, Table OA.1). The principal investigator at the MiLab Research Unit hired IPSOS
Austria to provide respondents for the survey, with the questionnaires administered via Qualtrics at
the University of Vienna. IPSOS maintains and curates a massive panel of hundreds of thousands
of subjects from where respondents were selected from a stratified and randomized subsample of
3000 individuals, so they reflect key demographic elements from the US census (i.e., education,
gender, and income) (Gil de Zúñiga et al., 2021). Although acknowledging the limitations of non-
probabilistic sampling, this technique allows us to obtain a diverse and ‘census-balanced sample’
appropriate for our research interest. This is particularly so as we do not focus on estimating popu-
lation values, but we rather develop multivariate models to assess the initial relationship between
two main survey variables: media skepticism and political evaluations (Baker et al., 2010). The
questionnaire we used contained an extensive battery of items designed to measure different types
of political and media-related attitudes and patterns of media consumption. All indexes, otherwise
stated, were measured on a 1–10 Likert type scale.
Dependent variables (DVs)
To see the effects of the scandal, we separately used three main questions that tap into evaluations
of, first, the political system of the United States: ‘Please rate your feelings of trust towards the
political system of the U.S.’ (W1 M = 3.96, SD = 2.52; W2 M = 3.77, SD = 2.41); second, the gov-
ernment: ‘Please rate your feelings of trust towards the government’ (W1 M = 4.39, SD = 2.47; W2
M = 4.37, SD = 2.46); and third, US president, Donald Trump: ‘We would like to know your feel-
ings toward Donald Trump,’ on a scale from 0 to 10, in which 0 means very unfavorable, and 10
very favorable (W1 M = 5.06, SD = 3.9; W2 M = 4.89, SD = 4.06).
6 International Political Science Review 00(0)
Independent variables (IVs)
To measure media skepticism, we used the averaged index of three questions: ‘I think the news
media prioritize being first to report a story’; ‘I think the news media get in the way of society solv-
ing its problems’; ‘Overall, I am skeptical about the news media’ (W1 Cronbach’s α = .80, M = 6.58,
SD = 2.25) (see Tsfati, 2003).1 Regarding partisan identity, we utilized a question that asked
respondents where they place themselves on a 0–10 scale where the extremes represent strong
Democrat and strong Republican, respectively, and middle positions self-location as independent.
From it, we created three categories: Democrat/Democrat-leaning (0–4), independent (5), and
Republican/Republican leaning (6–10) (W1 = 478 Republican/Republican leaning, 443 Democrat/
Democrat leaning, 363 independent). An alternative operationalization that considers a broader
category of independents (4–6) does not affect our results.
Rest of covariates and controls
To isolate the effect of the scandal on attitudes as much as possible, we incorporated an exhaustive
battery of controls that allowed for considering alternative explanations. We controlled for levels
of generalized trust using the average of two questions: ‘Generally speaking, would you say that
most people can be trusted, or that you need to be very careful?’; ‘Could you tell me how much you
trust people that you meet for the first time?’ (Spearman-Brown ρ = .84, M = 4.18, SD = 2.35); sat-
isfaction with life: ‘In most ways, my life is close to my ideal’; ‘All things considered, I am satis-
fied with my life as a whole’ (Spearman-Brown ρ = .87, M = 6.04, SD = 2.38); dissatisfaction with
the government in fields not related to the scandal, such as ‘crime, unemployment, difference
between rich and poor, and the cost of living’ (Cronbach’s α = .78, M = 6.82, SD = 2.04), and media
consumption, a composite index of respondents’ usage of newspapers, radio, TV, social media, and
online platforms to get news.2 This index relates to 41 questions in which respondents answered
questions about the frequency with which they get news from different media outlets (Cronbach’s
α = .95, M = 3.92, SD = 1.76).3 We further controlled for age, gender (female as reference), educa-
tion (Bachelor’s degree or more as reference), race (white as reference), and income levels (dichot-
omous variable considering the median family income in the United States).4
Admittedly, the availability of specific questions related to the scandal would have been ideal to
pinpoint its effects on political attitudes, but, as it is often the case in quantitative studies of politi-
cal scandals, data were not originally collected for these purposes (Geiß, 2017). In this sense, we
were lucky to capture the occurrence of an exogenous shock (the Trump–Ukraine scandal) between
the first and second wave of our longitudinal study, which situates our research within the frame of
‘natural experiments’ combined with survey research (Barabas and Jerit, 2010; Minkus et al.,
2019). However, one must consider the two main strengths of our design. First, our study evaluates
actual candidates and political institutions, which reduces possible biases associated with moti-
vated reasoning in experimental designs with fictional candidates (von Sikorski et al., 2020).
Second, due to its panel nature, we have data on the same individuals before and after the external
natural ‘stimuli’ (i.e., the Trump–Ukraine scandal) affected the political scene.
Analytical approach
To analyze the relationship between media skepticism, alone and in combination with partisan
identity, and political evaluations after a scandal, we use Ordinary Least Square Regression (OLS).
We conduct three different models: a cross-sectional one where all variables used to belong to our
baseline survey in the first wave (DV and IVs from W1); a lagged one, which addresses some
Marcos-Marne et al. 7
temporality relationships between wave 1 and wave 2 (DV from W2, all IVs from W1); and finally,
an autoregressive model, which allows for more stringent causal order effects in time, as it also
controls for prior levels of the dependent variable in time 1 (lagged model plus a control of evalu-
ations of political institutions and Trump from W1). While the first two models allow for a control
of the consistency and coherence of our results, we focus our interpretation on the third and last
model, the autoregressive. This model allows us to specifically pay attention to the association
between media skepticism and attitude change after the scandal.5
A preliminary analysis of descriptive data shows that individuals displayed more negative evalua-
tions of the US political system, government, and Donald Trump after the scandal (comparing data
from W1 and W2). On aggregate terms, individuals trusted less the U.S political systems and the
government, but a different pattern emerged from the question that measures support for Trump.
Overall, respondents showed less favorable feelings towards Trump, but this happened even if
individuals moved towards both extremes (extremely favorable and unfavorable) (Table 1). The
extent to which changes in evaluations can be attributed to the Trump–Ukraine scandal is easier to
observe at the individual level, where we can control both for previous evaluations and alternative
explanations with an exhaustive battery of controls. Strictly controlling for remaining sources of
variation is key to sustain that the observed effects are due to the scandal, since we do not have
specific questions about it.
Our cross-sectional and lagged models show a consistent relationship between media skepti-
cism and political evaluations (Tables 2 and 3). Individuals who rank higher on media skepticism
display lower trust toward the US political system and the government and more favorable views
of Donald Trump. The connection between media skepticism and trust in institutions taps well into
the relationship between media and political trust. The coefficient in the evaluations of Trump was
expected attending to his discourse toward mainstream media agencies (Lischka, 2019). However,
this data cannot shed light on the relationship between media skepticism and attitudes after the
scandal, for which we rely on our autoregressive model. In the autoregressive model we find that,
unsurprisingly, political evaluations in the first wave are the best predictor of political evaluations
in the second wave (Table 4). Still, higher levels of media skepticism significantly relate to more
favorable views of Trump after the scandal took place (see also online appendix, Table OA.2).
However, the effect of media skepticism is not significant for trust towards the political system and
the government. We see a consistent effect of identifying as independent on the partisan scale
among the controls included. As compared to Republicans (our reference category), independents
show more negative evaluations in the three models, which may correlate with independents hav-
ing a not so strong apriorist opinion on politicians and institutions that makes them more likely to
react to scandals.6 Democrats rank lower than Republicans in their evaluations of Trump after the
Table 1. Average evaluations of political institutions/actors.
Variable (min–max) Wave 1 Wave 2
Mean (SD) Mean (SD)
US political system (1–10) 3.96 (2.52) 3.77 (2.41)
Government (1–10) 4.39 (2.47) 4.37 (2.46)
Donald Trump (0–10) 5.06 (3.9) 4.89 (4.06)
8 International Political Science Review 00(0)
Table 2. Cross-sectional model. Evaluations of US political system, government, and Trump.
Predictors Dependent variable
US political system
(Wave 1)
(Wave 1)
(Wave 1)
Gender (female as reference) −0.017 (0.127) −0.233* (0.130) −0.399** (0.181)
Age −0.202** (0.081) 0.020 (0.083) 0.152 (0.116)
Education −0.043 (0.142) 0.164 (0.145) −0.451** (0.202)
Income −0.324** (0.144) −0.085 (0.147) −0.538*** (0.206)
Race (White as reference) −0.262* (0.157) −0.182 (0.160) 0.339 (0.224)
Satisfaction with life 0.078*** (0.029) 0.117*** (0.030) 0.129*** (0.042)
Trust towards people 0.333*** (0.030) 0.311*** (0.030) −0.056 (0.043)
Dissatisfaction gov. policies −0.060* (0.035) −0.012 (0.036) −0.236*** (0.050)
News consumption 0.476*** (0.043) 0.382*** (0.044) 0.127** (0.062)
Democrat −0.346** (0.160) −0.282* (0.164) −4.904*** (0.229)
Independent −0.312* (0.160) −0.259 (0.164) −2.855*** (0.229)
Media skepticism −0.125*** (0.031) −0.157*** (0.032) 0.295*** (0.044)
Constant 2.475*** (0.445) 2.410*** (0.454) 6.591*** (0.637)
Observations 1,067 1,071 1,068
R20.337 0.285 0.463
Adjusted R20.329 0.277 0.457
Residual std. error 1.999 (df = 1054) 2.045 (df = 1058) 2.854 (df = 1055)
F statistic 44.590*** (df = 12; 1054) 35.122*** (df = 12; 1058) 75.713*** (df = 12; 1055)
Estimates are coefficients (ordinary least squares regression) with standard errors in parentheses.
*p < 0.10. **p < 0.05. ***p < 0.01.
Table 3. Lagged model. Evaluations of US political system, government, and Trump.
Predictors Dependent variable
US political system
(Wave 2)
(Wave 2)
(Wave 2)
Gender (female as reference) −0.348* (0.210) −0.475** (0.222) −0.424 (0.297)
Age −0.314** (0.149) 0.003 (0.157) −0.156 (0.211)
Education 0.046 (0.222) 0.346 (0.235) −0.526* (0.315)
Income −0.547** (0.233) −0.284 (0.247) −0.381 (0.332)
Race (White as reference) 0.020 (0.273) 0.411 (0.290) 0.188 (0.384)
Satisfaction with life 0.050 (0.046) 0.120** (0.049) 0.177*** (0.066)
Trust towards people 0.209*** (0.051) 0.152*** (0.054) 0.004 (0.072)
Dissatisfaction gov. policies −0.127** (0.059) 0.003 (0.062) −0.292*** (0.082)
News consumption 0.350*** (0.075) 0.391*** (0.079) −0.076 (0.105)
Democrat −0.260 (0.271) −0.396 (0.286) −4.978*** (0.382)
Independent −0.648** (0.263) −0.777*** (0.278) −3.155*** (0.372)
Media skepticism −0.095* (0.050) −0.202*** (0.053) 0.403*** (0.072)
Constant 4.609*** (0.810) 3.460*** (0.857) 7.511*** (1.145)
Observations 432 431 418
R20.199 0.189 0.505
Adjusted R20.176 0.165 0.490
Residual std. error 2.055 (df = 419) 2.171 (df = 418) 2.854 (df = 405)
F statistic 8.689*** (df = 12; 419) 8.106*** (df = 12; 418) 34.413*** (df = 12; 405)
Estimates are coefficients (ordinary least squares regression) with standard errors in parentheses.
*p < 0.10. **p < 0.05. ***p < 0.01.
Marcos-Marne et al. 9
scandal, even when previous evaluations are accounted for, which is in agreement with previous
findings (von Sikorski et al., 2020). Other controls have no robust effect on the evaluations of the
political system, the government, and Trump.7
While we see the effect of media skepticism moderating reactions towards the politician
involved in the scandal (supporting H1a), we find no effect of media skepticism on changing evalu-
ations towards political institutions (against H1b). However, the existence of a gap in the percep-
tion of Trump as a function of media skepticism, even when previous attitudes are controlled for,
can have two main different interpretations. First, the gap may appear via the buffering effect of
media skepticism concerning attitude change (i.e., attitudes are overall more negative after the
scandal, which is more transparent for individuals who display lower levels of skepticism towards
media). Second, the gap may present because individuals who are highly skeptical of media, far
from reacting negatively to the scandal, see Trump in a more positive light after he is ‘attacked’ by
the mainstream media they strongly distrust.
To test which of these mechanisms was more prevalent, we first used the cross-sectional model
to predict attitudes towards Trump for individuals at maximum and minimum levels of media
skepticism (predicted attitudes were 5.98 and 8.94, respectively, z = −5.02, p < .000, two-tailed).8
In a second step, we used the autoregressive model to predict attitudes towards Trump for individu-
als displaying maximum and minimum levels of media skepticism who had the average predicted
attitudes in the first wave. New predicted values were 5.15 and 9.19, respectively (z = −5.95,
p < .000, two-tailed), which suggests the existence of a polarization effect taking place at the
Table 4. Autoregressive model. Evaluations of US political system, government, and Trump.
Predictors Dependent variable
US political system
(Wave 2)
(Wave 2)
(Wave 2)
US political system (W1) 0.555*** (0.042)
Government (W1) 0.660*** (0.041)
Trump (W1) 0.878*** (0.026)
Gender (female) −0.332* (0.177) −0.349** (0.175) −0.016 (0.150)
Age −0.074 (0.126) 0.110 (0.124) −0.126 (0.106)
Education 0.024 (0.187) 0.221 (0.185) −0.200 (0.159)
Income −0.244 (0.197) −0.083 (0.194) −0.243 (0.167)
Race (White) 0.138 (0.230) 0.343 (0.228) 0.203 (0.194)
Satisfaction with life −0.018 (0.039) 0.018 (0.039) 0.031 (0.033)
Trust towards people 0.018 (0.045) −0.056 (0.044) −0.011 (0.036)
Dissatisfaction gov. policies −0.101** (0.049) −0.023 (0.049) −0.094** (0.042)
News consumption 0.162** (0.064) 0.211*** (0.063) −0.035 (0.053)
Democrat −0.012 (0.229) −0.062 (0.226) −0.492** (0.232)
Independent −0.400* (0.223) −0.368* (0.220) −0.351* (0.205)
Media skepticism −0.018 (0.043) −0.069 (0.043) 0.091** (0.037)
Constant 2.663*** (0.700) 1.476** (0.684) 1.458** (0.604)
Observations 430 431 418
R20.437 0.500 0.874
Adjusted R20.419 0.484 0.870
Residual std. error 1.728 (df = 416) 1.707 (df = 417) 1.440 (df = 404)
F statistic 24.796*** (df = 13; 416) 32.076*** (df = 13; 417) 216.101*** (df = 13; 404)
Estimates are coefficients (ordinary least squares regression) with standard errors in parentheses.
*p < 0.10. **p < 0.05. ***p < 0.01.
10 International Political Science Review 00(0)
extreme values of media skepticism (i.e., while individuals ranking very low on media skepticism
worsened their perception of Trump after the scandal, individuals who were highly skeptical of
media even improved their perception slightly). The polarization effect, however, disappeared as
more moderate values of media skepticism were taken as reference. For example, attitudes towards
Trump did not improve among individuals who ranked 7 or 8 on the scale of media skepticism,
supporting the buffering effect explanation among larger sectors of the population.9
To test the combined effect of media skepticism and partisan identity, we ran an additional
model that included an interaction term between these two variables. The interaction term was not
significant, suggesting the effect of media skepticism on political evaluations after a scandal did
not vary across categories of partisan identity (Table 5). A graphical representation presented in the
online appendix (Figure OA.3) shows that slopes for predicted results run largely parallel and con-
fidence intervals overlap consistently, further supporting this claim. This does not mean that parti-
san identity was irrelevant to explain trust towards political institutions or, even more importantly,
specific political actors. While the slopes were not significantly different, intercepts varied as a
function of partisan identity (in all three models for independents, in favorable perceptions of
Trump for Democrats too). The effect of the individual terms existed but, recurring to our
Table 5. Autoregressive model with interactions.
Predictors Dependent variable
US political system
(Wave 2)
(Wave 2)
(Wave 2)
US political system (W1) 0.559*** (0.042)
Government (W1) 0.662*** (0.041)
Trump (W1) 0.877*** (0.026)
Gender (female as reference) −0.315* (0.177) −0.337* (0.175) −0.011 (0.150)
Age −0.079 (0.126) 0.106 (0.124) −0.126 (0.106)
Education 0.016 (0.187) 0.219 (0.185) −0.199 (0.159)
Income −0.254 (0.197) −0.093 (0.195) −0.250 (0.168)
Race (White as reference) 0.129 (0.230) 0.340 (0.229) 0.201 (0.194)
Satisfaction with life −0.016 (0.039) 0.021 (0.039) 0.034 (0.034)
Trust towards people 0.017 (0.045) −0.059 (0.044) −0.013 (0.036)
Dissatisfaction gov. policies −0.106** (0.050) −0.025 (0.049) −0.094** (0.042)
News consumption 0.162** (0.065) 0.214*** (0.063) −0.030 (0.053)
Democrat 0.406 (0.694) 0.566 (0.686) 0.106 (0.590)
Independent −0.910 (0.747) −0.456 (0.740) −0.130 (0.631)
Media skepticism −0.014 (0.070) −0.038 (0.070) 0.133** (0.061)
Democrat × media skepticism −0.071 (0.097) −0.098 (0.095) −0.092 (0.082)
Independent × media skepticism 0.081 (0.102) 0.020 (0.100) −0.027 (0.086)
Constant 2.663*** (0.855) 1.233 (0.844) 1.122 (0.723)
Observations 430 431 418
R20.440 0.502 0.875
Adjusted R20.420 0.484 0.870
Residual std. error 1.727 (df = 414) 1.707 (df = 415) 1.441 (df = 402)
F statistic 21.668*** (df = 15; 414) 27.902*** (df = 15; 415) 187.076*** (df = 15; 402)
Estimates are coefficients (ordinary least squares regression) with standard errors in parentheses.
*p < 0.10 **p < 0.05 ***p < 0.01.
Marcos-Marne et al. 11
reasoning above, the final effect responded to the sum of its parts (partisan identity and media
skepticism, with no moderating effect). Thus, we found no evidence to support H2a and H2b.10
Discussion and conclusion
This study sought to clarify the type of individual level antecedents that influence the effects of
political scandals over political elected officials and institutions. In doing that, we aimed to under-
stand better how the governed perceive political misbehavior and react to it, which has important
implications for the stability of governments and the overall perception of the political system.
Here, we focused on the relationship between media skepticism and attitude change after a political
Our theoretical expectation was that individuals who rank higher on media skepticism would be
less likely to be exposed to (mainstream) media news and less likely to accept their arguments.
This would affect, in turn, their reaction to political scandals. Our results supported this theoretical
proposition. According to our empirical evidence, individuals who ranked higher on media skepti-
cism displayed more favorable views of Trump after the Ukraine scandal, even when previous
evaluations of Trump were controlled for. This effect remained after controlling for important
predictors of political evaluations and different patterns of news consumption and it seemed
explained by a combination of buffering and backfire effects, with a clear predominance of the
former (the latter was only visible when the highest possible levels of media skepticism were
While we found no significant contribution of media skepticism on changing evaluations
towards the US political system or the government, we think this can demonstrate a general lack of
short-term spillover effects derived from the scandal. In this vein, and provided that the effects of
scandals go beyond the political officials involved, we would expect a similar effect of media
skepticism on support for political institutions. While our results for media skepticism focus on the
evaluation of the politician initially splashed in the political scandal, our argument could be applied
to the evaluation of political institutions whenever accessibility–diagnosticity mechanisms favor
the appearance of spillover effects (Roehm and Tybout, 2006). Overall, we speculate the lack of
spillover effects could be influenced by the strong personal implication of Trump in the scandal,
the fatigue effects derived from the exposure to a series of misbehaviors from the president
(Collinson, 2019a; Kumlin and Esaiasson, 2012), and/or the short time passed between the scandal
becoming known and our survey being fielded. However, further studies should specifically expand
our study to test these alternative ad hoc theoretical explanations. These would be of particular
importance to unravel the associations between media skepticism and overall political trust, mainly
when new technologies and social media are candidates to increase the prevalence of media skepti-
cism (Cozzens and Contractor, 1987).
Albeit the findings presented help clarify the connection between media skepticism, political
trust, and the effects of political scandals, the study is not immune to limitation. First, we rely on
a quasi-natural experimental condition and online panel survey data. Although this has important
methodological advantages regarding controlling for previous levels of trust towards actual poli-
ticians and institutions, further studies based on sole experimental conditions may further help
to shed light on the relationship between media skepticism, political scandals, and political
evaluations from citizens. Particularly important in this regard is that we cannot guarantee our
results are solely explained by the Trump–Ukraine scandal, even if this was certainly a major
issue around his person (as shown above), because we lack explicit questions about it. Our strat-
egy has focused on considering most relevant predictors of political evaluations together with
alternative explanations for attitudinal changes, but we must recognize Donald Trump´s
12 International Political Science Review 00(0)
presidency was well known for different controversies of varying intensity that may have also
affected public assessments of his figure.11 Overall, we are far from claiming to have the final
word on the extent to which media skepticism buffers reactions to political scandals. Nevertheless,
we believe our design will contribute to this discussion by complementing experimental
approaches that are better equipped to identify causality but struggle more with replicating actual
political conditions.
Second, we are focusing our analysis on a particular scandal in the United States. This posits
advantages in terms of controlling for potential confounders in the relationship under study. Still,
more comparative studies will be necessary to know whether media skepticism moderates reac-
tions to political scandals beyond the United States and Trump’s presidency. We believe two main
aspects need acknowledgment before extrapolating our results to other cases: first, the particular
relation of Donald Trump with mainstream media (Grynbaum, 2017), including the fact that Trump
himself may have fueled media skepticism with his discourses (Meeks 2019); second, the fact that
the scandal under consideration in our research involves political figures of the highest level (both
Trump and Biden), which could make it more likely that citizens will become interested in it. We
speculate these factors may indeed affect the overall aggregated reaction to the scandal due to the
number of people exposed to it combined with the prevalence of media skeptics, but the individual
association between media skepticism and attitudinal reaction should not be determined by them.
However, only future studies can determine empirically if our theoretical link between media skep-
ticism and reaction to political scandals holds in different settings that are less shaped by media
skepticism and involve lesser-known political figures.
The results of our research have implications for our understanding of whether and how
political scandals influence public opinion. However, they also evidence significant practical
challenges that transcend academic discussions. Empirical data show that distrust towards media
has taken an unprecedented magnitude (Dahlgren, 2018). This can be related to gaps between
expectations and outcomes received from media and further connected to the increasing number
of individuals highly involved in controversial topics who perceived the media as biased
(Engelke et al., 2019). It goes without saying that healthy democratic societies display critical
attitudes also towards media, and some manipulative behaviors from media indeed deserve criti-
cal judgments (van Dijk, 2017). However, extended skepticism towards media can challenge
some of the functions traditionally assigned to them, such as informing and promoting public
debates. Although we are hardly the first one to refer to this trend, we show empirically that
media skepticism affects responses to political scandals, which in turn may also influence mech-
anisms of accountability through voting decisions and, relatedly, to overall levels of political
The authors would like to thank three anonymous reviewers for their constructive comments, which helped us
to improve the manuscript.
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publica-
tion of this article: Late stages of this work at the Democracy Research Unit (DRU) have benefited from the
support of the Spanish National Research Agency’s Program for the Generation of Knowledge and the
Scientific and Technological Strengthening Research + Development Grant PID2020-115562GB-I00. The
last author is funded by the ‘Beatriz Galindo Program’ from the Spanish Ministry of Science, Innovation &
Universities, and the Junta de Castilla y León.
Marcos-Marne et al. 13
Hugo Marcos-Marne
Pablo González-González
Homero Gil de Zúñiga
Supplemental material
Supplemental material for this article is available online.
1. Acknowledging that media skepticism can be related to very different ways of getting informed, we
included a correlation table of the construct with the individual items of news consumption included in
our survey (online appendix, Table OA.2).
2. Controlling for alternative sources of information is crucial in our research, as we aim to isolate the
effect of lower/biased exposition to mainstream media information about the scandal caused by media
3. For example, for TV news, respondents were asked five questions: ‘In the past month, how often did
you get news from the following media sources? 1. Network TV news (e.g., ABC, CBS, NBC), 2. Local
television news (cf. local affiliate stations), 3. MSNBC cable news, 4. CNN cable news, 5. FOX cable
4. In our baseline survey we have the following distribution for our sociodemographic controls. Age: 7%
between 18 and 22 years old; 32.3% between 25 and 35 years old; 39.8% between 36 and 55 years old;
28.1% 56 or older. Gender: 53.1% female; 46.6% male; 0.22% other. Education: 40.5% Bachelor’s
degree or more. Race: 74% white; 26% other. Income: 58% above the median income.
5. An additional model that used intra-personal change as dependent variable was also conducted. The
direction of the association between media skepticism and political attitudes after the scandal remains,
but the coefficient is not significant (see online appendix, Table OA.3). This may be attributed to the
stringency of the model as well as its tendency to inflate error variances. Thus, we decided to focus our
interpretation on the results of the autoregressive model above.
6. To further differentiate the effects of partisan identity and media skepticism, we repeated the analysis
using a dichotomous indicator of partisanship (Republican/not Republican). Results indicate, first, that
whenever media skepticism is at the lowest point (0), identifying as Republican has no effect on the eval-
uation of Trump; second, that higher levels of media skepticism are associated with better evaluations of
Trump, even for individuals who do not identify as Republican (0 being independents and Democrats);
third, there is no significant conditional effect between our two key variables.
7. The coefficient of news consumption on trust towards the system and the government could be explained
by individuals more interested in politics consuming more information and being more supportive of
political institutions.
8. All the remaining variables were set at the mean/mode.
9. Only 10% of our respondents were located at the extremes of the scale, almost all of them in the most
skeptical pole (see online appendix, Figure OA.1).
10. Non-significant results were also obtained if partisan identity was coded as a dichotomous variable
(Republican/not Republican).
11. See Collinson (2019b) for a summary of some controversies taking place in between the two waves of
our survey.
Baker, Reg, Stephen J. Blumberg, J. Michael Brick, Mick P. Couper, Melanie Courtright, J. Michael Dennis,
Don Dillman, Martin R. Frankel, Philip Garland, Robert M Groves, Courtney Kennedy, Jon Krosnick
14 International Political Science Review 00(0)
and Paul J Lavrakas (2010) Research Synthesis: AAPOR report on online panels. Public Opinion
Quarterly 74(4): 711–781. Available at:
Barabas, Jason and Jennifer Jerit (2010) Are Survey Experiments Externally Valid? American Political
Science Review 104(2): 226–242. DOI: 10.1017/S0003055410000092.
Blitzer, Ronn (2021) Ukraine’s Zelensky ‘Offended’ by Media Reporting That Trump ‘Put Pressure’ on him.
Fox News. Available at:
call-pressure (accessed 18 February 2021).
Brody, Richard A and Catherine R Shapiro (1989) Policy Failure and Public Support: The Iran-contra affair
and public assessment of President Reagan. Political Behavior 11(4): 353–369. Available at: http://
Carlin, Ryan E, Gregory J Love and Cecilia Martínez-Gallardo (2015) Cushioning the Fall: Scandals, eco-
nomic conditions, and executive approval. Political Behavior 37(1): 109–130. DOI: 10.1007/s11109-
Collinson, Stephen (2019a) It’s Nearly Impossible to Track All the Scandals Swirling Around Trump. CNN.
Available at:
Collinson, Stephen (2019b) Trump’s Controversies Show his Presidency is Really All About himself. CNN.
Available at:
Cozzens, Michael D and Noshir S Contractor (1987) The Effect of Conflicting Information on Media
Skepticism. Communication Research 14(4): 437–451. DOI: 10.1177/009365087014004004.
Dahlgren, Peter (2018) Media, Knowledge and Trust: The deepening epistemic crisis of democracy. Javnost
- The Public 25(1–2): 20–27. DOI: 10.1080/13183222.2018.1418819.
Dancey, Logan (2012) The Consequences of Political Cynicism: How cynicism shapes citizens’ reactions to
political scandals. Political Behavior 34(3): 411–423. DOI: 10.1007/s11109-011-9163-z.
Dilanian, Ken and Tom Winter (2020) Here’s What Happened When NBC News Tried to Report on the
Alleged Hunter Biden Emails. NBC News. Available at:
tion/here-s-what-happened-when-nbc-news-tried-report-alleged-n1245533 (accessed 18 February 2021).
Ditto, Peter H and David F Lopez (1992) Motivated Skepticism: Use of differential decision criteria for
preferred and nonpreferred conclusions. Journal of Personality and Social Psychology 63(4): 568–584.
Engelke, Katherine M, Valerie Hase and Florian Wintterlin (2019) On Measuring Trust and Distrust in
Journalism: Reflection of the status quo and suggestions for the road ahead. Journal of Trust Research
9(1): 66–86. DOI: 10.1080/21515581.2019.1588741.
Ernst, Nicole, Rinaldo Kühne and Werner Wirth (2017) Effects of Message Repetition and Negativity on
Credibility Judgments and Political Attitudes. International Journal of Communication 11: 3265–3285.
Garrett, R Kelly (2009) Politically Motivated Reinforcement Seeking: Reframing the selective exposure
debate. Journal of Communication 59(4): 676–699. DOI: 10.1111/j.1460-2466.2009.01452.x.
Garrett, R Kelly, Bruce Bimber, Homero Gil de Zuniga, François Heinderyckx, John Kelly and Marc Smith
(2012) Transnational Connections| New ICTs and the Study of Political Communication. International
Journal of Communication 6: 18. Available at:
(accessed 16 December 2021).
Geiß, Stefan (2017) Scandalization. In The International Encyclopedia of Media Effects. Wiley, 1–11.
Available at: (accessed
18 February 2021).
Gil de Zúñiga, Homero, Brigitte Huber and Nadine Strauß (2018) Social Media and Democracy. El Profesional
de la Información 27(6): 1172. Available at:
EPI/article/view/epi.2018.nov.01 (accessed 18 May 2021).
Gil de Zúñiga, Homero, Pablo González-González and Manuel Goyanes (2021) Pathways to Political
Persuasion: Linking online, social media, and fake news with political attitude change through political
discussion. American Behavioral Scientist.
Grynbaum, Michael (2017) Trump Strategist Stephen Bannon Says Media Should ‘Keep Its Mouth Shut.’ The
New York Times. Available at:
Marcos-Marne et al. 15
Kepplinger, Hans Mathias, Stefan Geiss and Sandra Siebert (2012) Framing Scandals: Cognitive and emotional
media effects. Journal of Communication 62(4): 659–681. DOI: 10.1111/j.1460-2466.2012.01653.x.
Kumlin, Staffan and Peter Esaiasson (2012) Scandal Fatigue? Scandal elections and satisfaction with democ-
racy in Western Europe, 1977–2007. British Journal of Political Science 42(2): 263–282. DOI: 10.1017/
Lischka, Juliane A (2019) A Badge of Honor? Journalism Studies 20(2): 287–304. DOI:
McCarthy, Tom (2019) Support for Impeachment Surges Amid Trump–Ukraine Scandal | Donald Trump. The
Guardian. Available at:
trump-ukraine (accessed 18 February 2021).
Maier, Jürgen (2010) The Impact of Political Scandals on Political Support: An experimental test of two theo-
ries. International Political Science Review 32(3): 283–302. DOI: 10.1177/0192512110378056.
Meeks, Lindsey (2019) Defining the Enemy: How Donald Trump frames the news media. Journalism &
Mass Communication Quarterly 97(1): 211–234. Available at:
full/10.1177/1077699019857676 (accessed 23September 2021).
Mettler, Suzanne and Robert C. Lieberman (2020) Four Threats: The Recurring Crises of American
Democracy. New York: St. Martin’s Press.
Minkus, Lara, Emanuel Deutschmann and Jan Delhey (2019) A Trump Effect on the EU’s Popularity?
The U.S. presidential election as a natural experiment. Perspectives on Politics 17(2): 399–416. DOI:
Nakashima, Ellen (2019) Whistleblower Complaint About President Trump Involves Ukraine, According to
Two People Familiar With the Matter. Available at:
Nyhan, Brendan (2015) Scandal Potential: How political context and news congestion affect the president’s
vulnerability to media scandal. British Journal of Political Science 45(2): 435–466. DOI: 10.1017/
Praino, Rodrigo and Daniel Stockemer (2021) The Electoral Consequences of Scandals: A meta-analysis.
Parliamentary Affairs. Available at:
gsaa068/6299982 (accessed 24 September 2021).
Praino, Rodrigo, Daniel Stockemer and Vincent G Moscardelli (2013) The Lingering Effect of Scandals in
Congressional Elections: Incumbents, challengers, and voters. Social Science Quarterly 94(4): 1045–
1061. DOI: 10.1111/ssqu.12046.
Pramuk, Jacob (2019) Whistleblower complaint is out: It alleges Trump abused power to influence 2020 elec-
tion. Available at:
Roehm, Michelle L and Alice M Tybout (2006) When Will a Brand Scandal Spill Over, and How Should
Competitors Respond? Journal of Marketing Research 43(3): 366–373. DOI: 10.1509/jmkr.43.3.366.
Stockemer, Daniel and Rodrigo Praino (2019) The Good, the Bad and the Ugly: Do attractive politicians get
a ‘break’ when they are involved in scandals? Political Behavior 41(3): 747–767. DOI: 10.1007/s11109-
Stroud, Natalie Jomini (2010) Polarization and Partisan Selective Exposure. Journal of Communication
60(3): 556–576. DOI: 10.1111/j.1460-2466.2010.01497.x.
The New York Times (2019) Trump’s Call With the Ukrainian President. Available at: https://www.nytimes.
The Washington Post (2019) The Washington Post Announces Plans to Expand its Investigative Journalism.
Available at:
Thompson, John (2000) Political Scandal: Power and Visibility in the Media Age. Cambridge: Cambridge
Polity Press.
Tsfati, Yariv (2003) Media Skepticism and Climate of Opinion Perception. International Journal of
Public Opinion Research 15(1): 65–82. Available at:
doi/10.1093/ijpor/15.1.65 (accessed 19 February 2021).
16 International Political Science Review 00(0)
Tsfati, Yariv and Joseph N Cappella (2003) Do People Watch What They do not Trust? Exploring the asso-
ciation between news media skepticism and exposure. Communication Research 30(5): 504–529. DOI:
van Dijk, Teun A (2017) How Globo Media Manipulated the Impeachment of Brazilian President Dilma
Rousseff. Discourse & Communication 11(2): 199–229. DOI: 10.1177/1750481317691838.
von Sikorski, Christian (2018) Political Scandals as a Democratic Challenge| The Aftermath of Political
Scandals: A Meta-Analysis. International Journal of Communication 12: 25. Available at: https://ijoc.
von Sikorski, Christian, Raffael Heiss and Jörg Matthes (2020) How Political Scandals Affect the Electorate.
Tracing the eroding and spillover effects of scandals with a panel study. Political Psychology 41(3):
549–568. DOI: 10.1111/pops.12638.
Wagner, Markus, Jessica Tarlov and Nick Vivyan (2012) Partisan Bias in Opinion Formation on Episodes
of Political Controversy: Evidence from Great Britain. Political Studies 62(1): 136–158. DOI:
Wolf, Zachary (2019) The Whistleblower Complaint. Available at:
Author biographies
Hugo Marcos-Marne is an Assistant Professor at the University of Salamanca, and member of the Democracy
Research Unit (DRU) at the same institution. His research focuses on public opinion, electoral behavior,
populism, and national identities.
Pablo González-González is a doctoral student in Political Science at the University of Salamanca, and a
member of the Democracy Research Unit (DRU). His research interests revolve around quantitative methods
applied to political communication, social media, extremism, and political behavior.
Homero Gil de Zuñiga currently serves as Distinguished Research Professor at University of Salamanca,
where he directs the Democracy Research Unit (DRU). He is also a Professor at Pennsylvania State University,
and Senior Research Fellow at Universidad Diego Portales, Chile. Overall, the work that takes place at the
DRU deals with the development of theoretically driven research, contributing to shed a social scientific light
over social media and new technologies effects in the arena of Political Communication.
... As trust in journalists and their reporting is-partially due to incongruent role perceptions (Abdenour et al., 2020)-declining globally (Edelman, 2018;Peters & Broersma, 2013), skepticism toward media is rising (Tsfati, 2010;Tsfati & Cappella, 2003). In general, skepticism is a useful trait for audiences, leading them to not blindly trust every piece of information and to question what they read (Marcos-Marne et al., 2022). This is particularly important following the rise of fake news and alternative information. ...
... While a base amount of skepticism engages readers with news content and consequently may curtail disinformation campaigns, high levels of media skepticism have turned from a helpful trait to democracy to a threat (Marcos-Marne et al., 2022). Users become less likely to believe anything they read coming from traditional sources and turn to alternative media outlets instead (Tsfati, 2010;Tsfati & Cappella, 2003). ...
Full-text available
News organizations require sustainable business models with reliable revenue streams to perform their intended function in liberal democracies. Since the digital transition, however, many citizens have considerably shown high reluctance to pay for digital content, consequently jeopardizing journalists’ watchdog role and substantially eroding media independence. Based on a two-wave panel survey from the United States, this study seeks to investigate the antecedents of readers’ paying behavior for news, focusing on the role of social media news use and media skepticism. Findings reveal that users’ frequency of social media news use is associated with future paying behavior, while higher levels of media skepticism significantly inhibit readers’ payment. The analyses also show that while those reporting low media skepticism are initially more inclined to pay for news, high levels of social media news use function to largely overcome this gap. The study argues that the role of social media in delivering news should be redefined in light of the new paid content strategies implemented by news organizations.
Full-text available
There is a vast research tradition examining the antecedents that lead people to be politically persuaded. However, political opinion and attitude change in social media has received comparatively scarce attention. This study seeks to shed light on this strand of the literature by theoretically advancing and empirically testing a structural equation model linking online social media, and fake news exposure, with political discussion and political persuasion in social media. Drawing on autoregressive causal tests from two waves of US survey panel data collected in 2019 and 2020, our results indicate that online, social media fake news, and political discussion are all positive predictors of individual political attitude change. Furthermore, structural equation tests reveal that online and social media news lead individuals to be exposed to fake news, which, in turn, predict higher levels of political discussion, ultimately facilitating political persuasion in the social media realm. Limitations and further suggestions for future research are also included in the study.
Full-text available
Political scandals are highly relevant for political decision-making and democratic processes more generally. While most prior research employed experimental and cross-sectional survey studies, we tested the effects of a political scandal in the context of the 2017 Austrian Parliamentary Elections using panel data (N = 559, both waves). Importantly, we used a unique data set collected before and just after a major scandal broke in the final election phase. Drawing on a motivated reasoning perspective, attribution theory, and the inclusion/ exclusion model, our results revealed a scandal-eroding effect particularly damaging a candidate's own base of supporters and leaving followers in disappointment. The findings also showed a negative scandal-spillover effect for candidate supporters high in scandal knowledge decreasing political trust toward other politicians. Importantly, the results revealed that negative candidate evaluations are not a necessary precondition for negative spillover effects on political trust more generally.
Full-text available
Slowly approaching the second quarter of the 21st century, research on social media and its effects over democracy has quickly permeated across various fields in social sciences, particularly political communication. Based on accumulated evidence in this strand of literature, this paper briefly summarizes several established research areas. First, it highlights how social media facilitates individuals’ political expression. Second, it sheds light on how this enhanced communicative behavior has reinvigorated a more participatory society in democracy. Third, it examines social media’s democratic deficit mechanisms in regards to stagnant political knowledge acquisition. Finally, the paper offers a theoretical explanation to this lack of knowledge acquisition through the lenses of the ‘news-finds-me’ perception theory. Overall, the present manuscript explains specific ways in which social media helps and hinders the advancement of healthy democracies, thereby also providing promising guidelines for future research.
Full-text available
This article represents the first attempt to examine the effects of political scandals via meta-analysis. Seventy-eight studies, collectively including more than 54,000 participants, were identified and examined. A quantitative analysis revealed that the number of studies has steadily increased. Research predominantly stems from North America and Europe, and more than two-thirds of studies are based on student samples. Publication outlets are mostly political science and psychology journals, whereas communication journals play only a minor role. A qualitative analysis shows that two central outcome variables are frequently studied (evaluation of politicians/electoral consequences). Overall, studies generally reveal negative evaluative effects for politicians. However, five central moderators (candidate characteristics, behaviors, prior attitudes, context, and scandal type) significantly influence scandal effects. It is also apparent that research has largely neglected to precisely conceptualize the major independent variable in scandal-effects studies: news coverage and its intensity. Central research gaps are identified, and avenues for future research are discussed.
Full-text available
Did the election of Donald Trump affect the popularity of the European Union (EU) in Europe? Theoretically, both a positive rally effect (due to a perceived external threat) and a negative domino effect (due to resignation among Europhiles and/or reinforcement among europhobe nationalists) are thinkable. We treat Trump's unexpected victory as an external shock and use a Eurobarometer survey that was conducted in all EU-28 member states four days prior to (control group) and six days after the election (treatment group) as source material for a natural experiment. The analysis reveals that the election of Trump caused a significant increase in the EU's popularity in Europe immediately after the election. This "Trump effect" is considerable in size, roughly equivalent to three years of education. Gains in popularity were particularly high among respondents who perceived their country as economically struggling and, surprisingly, among the political right, suggesting that Trump's victory broadened and ideologically diversified the EU's base of support.
Full-text available
In general, politicians involved in scandals of various natures are punished by voters. Good-looking politicians, on the contrary, are rewarded by voters. Almost fifty years of empirical research has shown that ill-informed voters will use the physical attractiveness of candidates, as well as readily-available information on scandal allegations involving candidates running for office, as a heuristic shortcut to determine their voting behaviour. This article represents the first attempt to link the existing literature on the electoral effects of scandals with the existing literature of the electoral impact of candidate attractiveness. Using data on U.S. House of Representatives elections between 1972 and 2012, we find that candidate attractiveness mitigates the negative electoral effects of involvement in scandal; this implies that attractive politicians do get a “break” when involved in scandals. Of all type of scandals, we also find that candidate attractiveness has the largest moderating role if the incumbent is embroiled in a sex scandal.
As the number of scandals involving politicians in office rises worldwide, the number of studies dedicated to analysing these scandals and their consequences rises as well. In this article, we try to summarise this emerging literature focusing on quantitative studies that use scandal as an independent variable to model its influence on politicians’ electoral results. The analysis finds that scandal-ridden politicians tend to get fewer votes at the ballot box, are more likely to lose elections, and are less likely to win re-election. It also finds that the link between scandal and turnout is unclear; some models indicate that scandals depress turnout, while others report an increase in turnout.
Donald Trump has been a vocal critic of the news media. This study content analyzes how Trump tweeted about the news media as the Republican nominee through his first year in office. Results revealed that Trump privileged conservative media via praise and media appearance frames, mentions, and retweets, while denigrating nonconservative and general media via attack and bias frames. In addition, Trump employed attack frames more as president than nominee. Finally, Trump’s attack and bias frames generated more retweets and favorites than other frames. Given Trump’s influence, such framing could strongly contribute to the public’s polarized perceptions of the news media.
The rapid advancement of research on trust and distrust in the news media and the plethora of methodological approaches that accompany it leads us to critically reflect the status quo and make suggestions for the road ahead. Following a brief overview of conceptual definitions of trust and distrust as well as of related concepts used in journalism studies, we turn to our main endeavour by presenting measurements used in the field. We identify difficulties in measuring both trust and distrust in journalism and offer suggestions for dealing with them. Specifically, we focus on four main issues: the concept drawn upon for measurement, the employed research design, the object of investigation, and the items and dimensions of measurement. Rather than presenting a finished solution, we hope to advance the methodological consolidation of the field and contribute to the ongoing scholarly debate.
The argument has two related parts: firstly, we are witnessing an “epistemic” crisis in public spheres that threatens to undermine political agency. This crisis has to do with the massive amounts and speed of information, the processes by which we construct knowledge, as well as the new forms of knowledge deriving from digital technologies. Many developments in information technologies benefit democracy, but there is a growing concern about cognitive dilemmas. Secondly, in the present tumultuous juncture of Western democracies, dominated by the populist revolt, traditional distrust of media has turned into an assault on basic Enlightenment premises, eroding shared understandings of reality and compatible discourse. “Knowledge” becomes legitimated via emotionality. Critical rationality and progressive politics must engage more with these developments.