Effects of political scandal news coverage
on political support
Tea Miettinen, 10602062
Master’s thesis
Master’s programme Communication Science
University of Amsterdam
Supervisor: Sanne Kruikemeier
1
Abstract
The number of political scandals reported by the media is increasing. Still, little is known
about how media coverage of political scandals affects political support. In this cross-national
survey experiment media coverage of sex, financial and power scandals, is used to examine the
effects. Political support is operationalized as trust in the political actors, the competence and
character reputations of politicians and citizens’ satisfaction with democracy.
The results suggest that the news coverage of political scandals has negative effects on
political support within the electorate. These perceptions are significant in evaluating a
politician’s performance and character. Decreased support is the most evident in a case of power
scandals, but sex and financial scandals also decrease support significantly. It remains unclear,
however, if political scandals have implications that go beyond the evaluations of politicians to
evaluations of the political institutions and overall satisfaction of the electorate with democracy.
Moreover, a cross-national comparison showed no significant differences suggesting that the
citizens in Finland, the Netherlands and Sweden with their similar media and political systems
may respond to scandal in a similar manner. This study indicates the importance of scandal type
differentiation while highlighting the role of the media coverage on political support.
2
News coverage and political scandals
It is becoming increasingly difficult to ignore the role of media in modern democratic
societies. Media is often referred as the fourth estate and Strömbäck (2008) viewed that politics are mediatized to the degree that we should worry about the “independence of politics and
society from the media” (p.228). The media are not only a platform and a mediator for politics,
but a definer of our political culture (Zaller, 1988). Investigative journalism, feature of fourth
estate journalism, has often revealed power, sex and financial scandals. The number of political
scandals discussed in the media has increased (Welch, 2007; Kantola & Vesa, 2013) to the extent
that political scandals are becoming a norm in Western democracies (Tumber & Waisbord, 2004;
Welch, 2007). The news values of tabloids, for example, emphasis on individuals and drama
(Downey & Stanyer, 2013; Ekström & Johansson, 2008; Kantola & Vesa, 2011, Tumber, 2004,
Waisbord, 2004) and the increased competition of media outlets with 24-hours reporting cycles
(Thompson, 2000; Welch 2007) are believed to contribute to the increased reporting on political
scandals.
By its very nature, the media coverage of scandals plays an important role in defining a
scandal, its length (van Dalen & Skovsgaard 2011; Esser & Hartung 2004; Kantola & Vesa
2013; Tumber & Waisbord 2004, Waisbord 2004) and influencing the course of public opinion
(Shah et al. 2002). More specifically, a political scandal is an intense public communication
about a real or an imagined political defect (Esser & Hartung 2004), which captures a great deal
of attention in the media, but also disappears rapidly (Tumber & Waisbord 2004).
To date, most studies on political scandals focus on describing political scandals in
specific countries (for cross-national comparison, see for example Sanders & Canel 2006; van
3 coverage of political scandals. Despite increased number of political scandals throughout Europe
(see for example: Chalaby. 2004; Jiménez 2004; Kantola & Vesa 2013), empirical studies on
implications of scandals on democracy are still absent (for similar criticism, see Bhatti et al.
2013).No previous study has investigated the effects of political scandals in a cross-national
manner.
The purpose of this study is to shed light on the effects of political scandals on political
support in a European context. Political support is understood as “attitudes towards political
parties, politicians and democracy” (Maier, 2011, p. 283). An empirical study in a form of a
cross-national experiment survey was carried out to answer the following research question:
To what extent does the news coverage of different political scandals affect citizens’ political support?
This paper first gives an overview of the theories on political scandals and their effects
that the hypotheses are based on. Then the research method of experimental surveys is described.
The third section of the paper presents the results of this study. Thereafter, the research question
is answered and the implications of this paper are discussed. It concludes with some
recommendations for future research and a short discussion of limitations of this research.
Theory
Political scandals and media coverage
A literature review reveals different conceptual definitions of political scandals.
Thompson (2000), for instance, points out “political scandals are struggles over symbolic power,
where reputation and trust are at stake” (p. 245). Kantola and Vesa (2013) argue that scandals
4 political scandals are transgressions of the legal or the moral standards of a society. (van Dalen
& Skovsgaard, 2011; Esser & Hartung, 2004; Funk, 1996; Thompson, 2000; Waisbord, 2004). However, before a politicians’ misbehaviors are turned into scandals, the media come
into play. The media are the most important information source for citizens to become informed
about politicians’ misbehavior (Goode, 1994; Kantola & Vesa 2011; Thompson 2000). The
media determine which scandals are covered and how they are framed (Chalaby, 2004;
Costas-Pérez, Solé-Ollé & Sorribas-Navarro, 2012; Ekström & Johansson, 2008; Esser & Hartung,
2004; Goode, 1994; Jiménez, 2004; Shah el al., 2002; Tumber &Waisbord, 2004; Waisbord,
2004). The public rarely knows about politicians’ misconduct, if the media have not reported it
(Costas-Pérez, Solé-Ollé & Sorribas-Navarro, 2012; Downey & Stanyer, 2013; Tumber, 2004;
Waisbord, 2004). In other words, media coverage is a precondition for the disclosure of a
political scandal (Thompson, 2000) and thus all the political scandals are media scandals
(Waisbord, 2004).
Moreover, Thomson (2000) argues that trust, reputation and symbolic power are essential
tools in politics. The trustworthiness of politicians and political institutions is the corner stone of
representative democracy, where the electorate chooses representatives based on their
reputations. Trust and reputation together construct the symbolic power that politicians use to
govern the people. The media coverage of a political scandal can threaten the democratic
political order by eroding the symbolic power of political actors with regards to trust and
reputation. Hence, a political scandal can change the status quo with negative implications on
5
Different forms of political scandals
Previous research shows that there are different forms of political scandals. Thompson
(2000) introduced three categories or forms of political scandals. He made a conceptual
distinction between sex, financial and power scandals, and they constitute the forms most
frequently discussed in political scandal literature (see for instance, Ekström & Johansson, 2008;
Tumber 2004).
The first form, the sex scandal, arises in the private sphere of a politician’s life. Most the
sex scandals involve a transgression of moral norms, but a transgression of law can also take
place, for instance, when a sexual affair involves a minor. The second form, the financial
scandal, most often takes place in the public sphere. Financial scandals often involve a violation
of the rules and laws of governing and the allocations of economic sources entrusted to public
figures. Bribery and other types of corruption are all forms of financial misbehavior. Also a
politician’s misuse of finances for private gain is a financial scandal. The last form, the power
scandal, is “the ‘purest’ form of political scandals” (Thompson, 2000, p.121). Power scandals
involve transgressions of rules, laws or norms that lie at the heart of a democracy. The key
characteristic of a power scandal is the exposition of “hidden [not democratic, transparent] forms of power” (idem, p. 197) and the ways individuals acquire or exercise such power as in the case
of the Watergate scandal (Thompson, 2000).
Taken together, there are different definitions and conceptualizations of political
scandals. Recent research has proposed that the different scandals reflect different spheres (see
table 1). Van Dalen & Skovsgaard (2011) used examples of scandals that either took place in the
6 private sphere whereas power scandals were inherently related to the public sphere. Public
sphere scandals are expected to have greater impacts because they are public in nature.
Table 1.
Types of scandals per spheres
Spheres of political scandals Types of scandals
Public sphere Power scandals
Private sphere Sex scandals
Both spheres Financial Scandals
Consequences of political scandals
News coverage of political scandals can trigger various consequences. Citizens,
politicians themselves, parties and governments can all be affected by a political scandal (Bowler
& Karp 2004; Esser & Hartung 20004; Maier, 2011; Miller, 1999; Thompson, 2000). Funk
(1996) found that political scandals have a negative impact on candidate evaluation (Funk,
1996). Competence ratings of a politician, especially, are affected by a scandal (Doherty,
Dowling & Miller, 2011; Funk, 1996). Maier (2011) stated that political support is decreased by every scandal because “citizens develop negative views of parties and politicians” across party
lines. In their study of the U.K. and the U.S., Bowler & Karp (2004) pointed out that scandals not
only negatively affect individual politicians, but also institutions and political processes as whole
(Bowler & Karp, 2004). In addition, the news coverage of political scandals affects citizens and
lead to increased distrust in authorities (Costas-Pérez, Solé-Ollé & Sorribas-Navarro, 2012;
Schudson, 2004), increased cynicism (Bowler & Karp, 2004), lower regard for the legitimacy of
7 efficacy (Bowler & Karp, 2004). This thesis will focus on the effects of news coverage of
political scandals on citizens’ political support.
As mentioned before, scandals take place in either the public or private sphere and under different circumstances. “[We] cannot assume that the consequences of these [three] different
types of scandal will be the same” (Thomson, 2000, p.233). Still, neither Thompson nor other
scholars have examined the effects of sex, financial and power scandals separately. Most of the
political scandal research uses descriptive case studies or aggregated data so the effects of
different types of scandals are rarely compared (for exceptions see, Doherty, Dowling & Miller,
2011; Funk, 1996; Waisbord, 2004). Moreover, Maier (2011) notes that “more research is needed on the types of political scandals and its effect on political attitudes.” (p.294). In this
thesis, the effects of coverage of political scandals are examined separately for each type of
scandal to better understand the specific effects of different types of political scandals.
The consequences of different types of political scandals
Sex scandal. As discussed above, the media coverage of different types of scandals may
have dissimilar effects on political support. Sex scandals often take place in the private sphere
(van Dalen & Skovsgaard 2011) and thus can be viewed to belong to the private life of a
politician. Then, a sex scandal affects one’s character reputation, but influences the competence
rating, the public self, of a politician to a lesser extent (Doherty, Dowling & Miller, 2011; Funk,
1996; Miller, 2011). For instance, Bhatti et al. (2013) and Funk (1996) concluded that a sex
scandal had a negative influence on the trustworthiness of a politician while changes in
competence rating were reported to be barely significant (Doherty, Dowling & Miller, 2011).
Miller (2011) also found that sex scandals did not lower government trust or approval ratings.
8 severe as the other types of scandals, but will still show negative effects on trustworthiness as
well as character reputation. Because government trust or approval ratings are not expected to
decrease in the case of a sex scandal it is also predicted that the decrease in trust in other political
institutions will be insignificant.
The following hypothesis is proposed:
H1: A sex scandal (vs. no scandal) has a negative effect on trustworthiness of a politician and his character reputation.
Financial scandal. Financial scandals can take place both in public and private domains
(van Dalen & Skovsgaard, 201). In the private sphere, Funk (1996) found that a politician
involved in a tax evasion scandal was evaluated more negatively than a politician involved in a
sex scandal. A financial scandal especially affected trustworthiness in a negative manner (Funk
1996). Financial scandals were also found to have negative implications on reputation of
politicians, but even more on competence ratings of politicians (Doherty, Dowling & Miller,
2011; Funk, 1996). Bowler & Karp (2004) noted that financial corruption scandals do not only
negatively affect trust on politicians, but also trust on political institutions through the negative
mediating effect of a scandal on diffuse support. However, satisfaction with democracy can
increase, if the rascals are thrown out and wrongdoings of elites exposed as a result of a scandal
(Miller, 1999; Thompson, 2000, Tumber & Waisbord, 2004). Thus, Maier (2011) and Miller
(2011) found that political scandals can even have slightly positive effects on satisfaction with
democracy. Besides, financial scandals typically transgress laws in the public sphere; the effects
on financial political scandals are then greater than of sex scandals. Based on this discussion, I
9 H2: A financial scandal (vs. no scandal) has a negative effect on trustworthiness of a
politician and on character and competence reputation, even more so than a sex scandal. H3: A financial scandal (vs. no scandal) has a negative effect on trustworthiness political
institutions.
H4: A revelation of a financial scandal (vs. no scandal) has a positive effect on satisfaction ratings of democracy.
Power scandal. Power scandals occur in the public sphere (van Dalen & Skovsgaard,
2011). As public events, power scandals not only affect the reputation and trust of a politician,
but also trust in political institutions (Bowler & Karp, 2004, Maier, 2011). Power scandals can
lead to a questioning the status of democracy because these scandals represent a misuse or abuse
of power given to the politician by the electorate and reveals of use of “hidden powers”
(Thompson, 2000), even though Maier (2011) concluded that satisfaction with democracy was
not eroded after exposure to such scandals (see also Miller, 2011). Doherty, Dowling & Miller
(2011) concluded that when a politician in a case of sex or financial scandal also abuses power,
the evaluations of a politician’s personal character and competence were even lower than without
the abuse of power. Based on previous literature, it is therefore predicted that power scandals
might have a greater negative impact on political attitudes and trust in democracy than sex or
financial scandals. Another rational for the fact that financial scandals might affect political
attitudes to a smaller extent than power scandals, is the fact that, for instance, corruption, is often
seen as a part of policy-making (Waisbord, 2004).
10 H5: A power scandal (vs. no scandal) has a negative effect on trustworthiness of a politician
and on character and competence reputation, even more so than the sex and the financial scandal.
H6: A power scandal (vs. no scandal) has the most negative effect on trustworthiness of political institutes.
H7: A power scandal (vs. no scandal) has the most negative effect on democracy satisfaction. “By looking at political scandals, one may learn something about the normative and
cultural bases of a society” (Esser & Hartung, 2004, p. 1042). Likewise, Thompson (2000) notes
that political scandals are different in different national contexts (see also Jiménez, 2004). What
is counted as the private sphere in one country can be in the public sphere in another (van Dalen
& Skovsgaard, 2011; Downey & Stanyer, 2013). Thus, the experimental design of this research
is to test if there are cross-national differences in how the media coverage of different political
scandals affect political support. The sub research question is as follows:
SUB-RQ1: Do the consequences of political scandals vary from one country to another?
Method
A cross-national survey experiment was conducted to examine to what extent the effects
of news coverage of political scandals differ between the most-similar countries and types of
political scandals. Bhatti et al. (2013) stated that it is difficult to study the consequences of
political scandals using observational methods, because scandals are in general one-time events
taking place in a specific country. To tackle this problem, as political scandals are not likely to
be the same in different countries at the same time, I have designed a scenario study. The
scenarios do not represent real-life events or actions, but were designed for this study to avoid
11 otherwise affect the outcome (Bhatti et al 2013; Fischele, 2000). The scenarios were built on the
three types of scandals: sex, power and financial as introduced by Thompson (2000). Each
participant was asked to read one of the scenarios and answer the written questions. All the
participants received the survey in English order to avoid translation bias. The survey was
distributed to the sample population via Facebook and direct personal contact May 14th - June 4th
2014.
Selection of research units
The most-similar system research emphasizes cultural differences between countries that
at the first glaze look alike (Dimitrova & Strömbäck, 2005). The countries chosen for this
research were Finland, the Netherlands and Sweden. All three have media system characterized
as the Democratic Corporatist Model, watch-dog journalistic culture and public service
broadcaster tradition. (Brants & van Praag 2006; Hallin & Mancini, 2004, for types of
journalistic culture see also van Dalen & Skovsgaard, 2011).
Finland, the Netherlands and Sweden are all members of the European Union as well as
being politically similar; they all are Western European countries with democratic, multi-party
systems. The voting turnout in parliamentary elections is typically fairly high in all three
countries with Sweden topping the list with circa 83 percent voter turnout. Finland and the
Netherlands both have a voter turnout of circa 71 percent (IDEA, 2014).
While it might be possible to choose more similar countries (i.e. the Scandinavian
monarchies) the goal of this study was to examine the cross-national differences in culturally and
politically similar Western European countries. These countries meet those requirements.
12 the Netherlands and Sweden, are similar as monarchies while two of the countries, Finland and
Sweden belong geographically and culturally to the Nordic countries.
Characteristics of the participants
The participants were students from higher educational institutions of Finland, the
Netherlands and Sweden who voluntary participated in this experiment and signed an informed
consent form. The student sample was chosen because of the expected difficulties in finding a
representative sample in all the three countries with the means available. Moreover, results of
this student sample can be generated to the whole population as in cross-national studies a
student sample gives an indication of cross-cultural differences (Flere & Lavric, 2008).
The total number of completed surveys was 248 (N = 248). Due to technical problems
one participant did not access the article so this participant was deleted from the sample (N =
247). Twenty respondents were excluded from the sample as they were not students bringing the
final number of surveys used to 217. Of the respondents 69 were Dutch, 70 were Finnish, 64
were Swedish and 14 from other nationalities. The age of the students ranged from 18 to 49 (M =
23.78, SD = 3.59) 63 were male (29%) and 154 female (71%)
Study design
All the participants read a brief vignette (see Appendix A for vignettes) containing a short
biography of a fictitious minister involved in one of three scandals: sex, financial or power
scandal or a control condition containing no reference to a scandal. The news headline and a sentence stating Minister’s wrongdoing were manipulated (see Table 2 for treatments). Before
carrying out this experimental survey a manipulation test was performed to show that the text
13 A 1x4 factorial between subjects design was used study differences between the survey
results. The participants were randomly assigned to read an article about a) an extramarital affair
(sex scandal), b) misuse of a state credit card (financial scandal), c) giving secret orders to the
secret service (power scandal) or d) only the biography (control condition). The participants were
asked to imagine that this hypothetical event took place in their respective country.
Table 2.
Political scandal treatments
Type of scandal Type of treatment Sex scandal
(Extramarital affair)
News Headline: Minister caught having an extramarital
affair
Body text: The minister has been having an affair for the
past two years with a 29 year-old actress (n =59)
Financial scandal
(Misuse of a state credit card)
News Headline: Minister caught using his state credit card
for private expenses.
Body text: the minister has for the past two years been using his state credit card to pay for his family’s vacations.
(n = 55)
Power scandal
(Secret orders to the secret
service)
News Headline: Minister caught giving secret orders to the
secret service
Body text: the minister has for the past two years been
giving orders to the secret service to hack intranets of
opposition parties. (n = 59)
14 The respondents were asked to carefully read the news article provided. After reading an
article, the respondents were asked to answer the survey questions, which measured their
political support. The article was visible on their computer screen for 10 seconds or longer. Once
a participant chose to proceed through the survey questions, he or she could not go back to the
article. At the end of the survey, the participants were debriefed and reminded that the characters
and the news articles both were fictional and created for this study.
Measurements
The media has an increasingly important role in the modern societies. Thus, it is plausible
to assume that the media coverage influences political support. In this study media coverage of
political scandals was used to study scandal perceptions based on different types of scandals and
different countries. The scales used in this study have been used previously in political scandals
research and thus respond to the conceptualization of political scandals (See Appendix B
question wording). The survey also included demographic questions and variables to measure participants’ media behavior and political interest.
Trustworthiness of a politician. Trustworthiness of a politician was measured on a
7-point Ohanian1 (1990) bipolar scale using three items (i.e. untrustworthy-trustworthy; M= 2.96,
SD=1.27). The factorial analysis revealed that the items measuring trust are of the same dimension (EV=3.79, explained variance 42.7%, Cronbach’s α.885).
Competence ratings of a politician. The competence rating of a politician was measured
using a 7-point Ohanian (1990) bipolar scale with three items (i.e. incompetent-competent; M =
15 4.28, SD = 1.25.) The factorial analysis ensured that the items load on one dimension (EV = 1.07, explained variance 81.1%, Cronbach’s α .861).
Character ratings of a politician. The character reputation of a politician was measured
with four items (i.e. cold person- warm person; M=3.20, SD= .822) using a 7-point bipolar scaled adapted from Funk’s (1996) work2. The factorial analysis ensured the items were on same
dimension (EV = 3.56, explained variance 71.12%, Cronbach’s α .898).
Trustworthiness of political institutions. Trust on political institutions was measured
on a 7-point scale (1=strongly disagree and 7= strongly agree) using four institutions (i.e. “After
reading about Minister Johannes Lind, I’m more inclined to trust the government of my
country”; M = 3.53, SD = .95). The factorial analysis shows that the items load on the same
dimension (EV = 2.63, explained variance 65.76, Cronbach’s α .817).
Democracy satisfaction. Democracy satisfaction refers to respondents’ satisfaction with
democracy in their countries after reading their article. Satisfaction was measured on a 7-point
scale (1 = strongly disagree, 7 = strongly agree) based on Maier’s (2011) scale3.
Differences between the countries. Changes in political support were controlled for
differences between the countries. In one variable, Finnish, Dutch, Swedish or other nationality
was used.
2 Funk (1996) reported average alpha .86 for personality ratings.
16
Analyses
To test the hypotheses as to how the media coverage of different types of political
scandals affects trust in a politician or political institutions as well as the competence and
character reputation ratings of a politician and satisfaction with democracy, analyses of variance
were performed. A MANOVA analysis was used to examine changes in trust on a politician and
his competence and character reputations. The MANOVA was chosen since several independent
variables were tested at once while controlling for correlation between the independent variables.
An ANOVA was used to examine changes in the degrees of trust on political institutions and
satisfaction with democracy.
Results
Impacts of a sex scandal news coverage
To test Hypothesis 1; a sex scandal news coverage has a negative effect on
trustworthiness of a politician and his character reputation, a MANOVA analysis was conducted (Wilks’ λ = .000, F (3, 513.669) = 3.900, p < .000, η2 = .12). As hypothesized, a post hoc
comparison presented a significant decrease in trust on a politician in the case of a sex scandal
when compared with the control group (MsexT = 3.23, SD = 1.19, p < .000 vs. McontrolT = 4.24, SD =
.92, p < .000, see all the means in the Appendix C). Also the character reputation was
significantly lower when compared to the control group (MsexR = 3.23, SD = .90, p < .000 vs.
McontrolR = 4.34, SD = .82, p < .000). In summary, the news coverage of the sex scandal led to a
decreased degree of trust on a politician and lowered his character reputation. Therefore, these
17
Impacts of a financial scandal news coverage
Hypothesis 2 states that a financial scandal has an even more of a negative effect on the
trustworthiness of a politician and on a politician’s character and competence reputation than a
sex scandal when compared with the control group. A MANOVA analysis was conducted to test the hypothesis (Wilks’ λ = .000, F (3, 513.669) = 3.900, p < .000, η2 = .12). A post hoc
comparison showed that a financial scandal decreased the trustworthiness of a politician even
more than a sex scandal (MfinancialT = 2.45, SD = .92, p < .000) vs. MsexT = 3.23, SD = 1.19, p <
.000). When involved in a financial scandal, the character reputation of a politician was lower
than of a politician involved in a sex scandal (MfinancialR = 2.82, SD = 1.02, p < .000) vs. MsexR =
3.23, SD = .90, p < .000). A post hoc comparison of a politician’s competence rating also
indicated that a politician, who was exposed in a financial scandal was perceived as a less
competent than the politician in the control condition (MfinancialC = 3.95, SD = 1.32, p < .000 vs.
McontrolC = 5.05, SD = 1.03, p < .000). Overall, these results indicate that the financial scandal
showed decreased perceived trustworthiness of a politician as well as lowering his character and
competence reputation more than the sex scandal. Thus, the hypothesis H2 is accepted.
Next, a one-way between subjects ANOVA (F (3, 213) = 1.827, p < .143) was conducted
to compare effects of a financial scandal news coverage on trust on the political institutions.
There were no significant changes in the degree of trust reported. Hypothesis 3 is rejected.
Hypothesis 4 predicted that a revelation of a financial scandal has a positive effect on the
degree of satisfaction with democracy in a participant’s country when compared to the control
group. However, using a one-way between subjects ANOVA (F (3, 213) = .192, p < .902) no
significant changes were found in the degree of satisfaction with democracy. Therefore H4 is
18
Impacts of a power scandal news coverage
A revelation of a power scandal was expected to have the greatest negative impact on
trustworthiness of a politician as well as his character and competence reputations. As a result of a MANOVA analysis (Wilks’ λ = .000, F (3, 513.669) = 3.900, p < .000, η2= .12), a post hoc
comparison revealed that the participants reading about the power scandal trusted the politician
the least (MpowerT = 2.33, SD = .97, p < .000 vs. MfinancialT = 2.45, SD = .92, p < .000) or vs. MsexT = 3.23, SD = 1.19, p < .000) and also rated his character (MpowerR = 2.67, SD = .83, p < .000 vs.
MfinancialR= 2.82, SD = 1.02, p < .000) or vs. MsexR = 3.23, SD = .90, p < .000) and competence
(MpowerC = 3.76, SD = 1.14, p < .000 vs. MfinancialC = 3.95, SD = 1.32, p < .000) lower than the ones
exposed to a sex or a financial scandal. Taken together, these results suggest that the power
scandal lowers trustworthiness, competence and character ratings of a politician more than the
other scandal types. Thus, H 5 is confirmed.
As with the financial scandal a one-way between subjects ANOVA (F (3, 213) = 1.827, p
< .143) did not show significant changes in the degree of trust on political institutions. Hence,
H6 is rejected.
Hypothesis 7, a power scandal (vs. no scandal) has the most negative effect on
democracy satisfaction, is also rejected by an ANOVA analysis (F (3, 213) = .192, p < .902) that
found no significant changes in the degree of satisfaction with democracy.
Country differences
The results were controlled with a country dummy to examine to which extent the effects
of news coverage differ between the most similar countries, Finland, the Netherlands and
Sweden. A MANOVA, with all the depended variables using a moderation analysis, showed no significant differences (Wilk’s λ = .365, F (45000, 884.322) = 1.06, p <.365).
19
Discussion
The main objective of this study was to examine to what extent the news coverage of
different political scandals affect political support. A literature search revealed little about
cross-national differences in impacts of political scandals. A cross-cross-national survey embedded
experiment was used to study expected differences. Overall, these results show that political
scandals of different types clearly affect political support to a different extent. While all three
forms of scandals surveyed showed significant negative effects on the evaluation of a politician
they did not, however, bear any direct significant effects in trust on political institutions or
satisfaction with democracy. Power scandals affected political support the most, as expected. Thus, Thomson’s (2000) notion of power scandals as the purest form of political scandals
appears to be supported. As power scandals abuse the representative power, which lies at the
heart of democracies, the impact of exposure spreads further than the impacts of sex or financial
scandals. These results provide further support for Doherty, Dowling & Miller (2011)that a
scandal that includes the misuse of power has decisive negative consequences.
Van Dalen & Skovsgaard (2011) made a distinction between scandals taking place in the
public and the private sphere. This distinction proved very useful as the results of this study
indicate that the media coverage of scandals taking place in the private sphere (i.e. an
extramarital affair) did not result in as large decrease in ratings as the coverage of a power
scandal taking place in the public sphere (i.e. secret orders to the secret service) while the
financial scandal taking place either in the private or the public sphere, (i.e., misuse of a state
credit card), fell in the middle range of the data.
Surprisingly, trust on political institutions did not seem to significantly decrease as a
20 they expose political scandals. At the same time, the importance and visibility of the media in the
contemporary Western European societies are increasing. Even though the media are said to
increasingly report on political scandals as a consequence of the adoption of tabloid norms (see
for instance Ekström & Johansson, 2008), these results suggest that individuals are more inclined
to trust media after coverage of political scandal stories. Another possibility is that the media still
are still viewed by citizens as the “fourth estate” and a trusted watchdog of government even
though scholars may state otherwise. This research also found significant evidence for impacts of
news coverage especially when it comes to evaluate an individual politician. This contradicts a
common limitation of survey studies and other studies analyzing the impact of news coverage,
where media effects are expected to be low and statistically insignificant (Maier, 2011). Taken
together, the power of influence of the media in modern politics is apparent and Strömbäck’s
(2008) concern of mediatization of the political appears relevant.
This research used the variable of Maier (2011) to examine satisfaction with democracy,
but found no significant direct effects. Instead of concluding that all the forms of political
scandals have similar effects on the degree of satisfaction with democracy, I would encourage to
find new, better ways to measure these effects. The interpretations of such terms as “democracy”
and “satisfaction” vary from person to person. Therefore any conclusions drawn from their
subjective application may indicate greater distinctions between people than this research design
was able to discern. Instead of asking about the satisfaction with democracy, other variables, for
example, trust on politicians and political institutions, could be mediating the degree of
satisfaction with democracy. In that case, this study indicates that slight changes in satisfaction
21 Besides examining the differences between the scandal types, this research also
investigated country differences. The countries studied, Finland, the Netherlands and Sweden,
were chosen based on the most similar design as presented in the method section. Even though
the countries chosen for this study offered an interesting case, the results did not show any
significant cross-national differences in scandal perception. However, these results implicate that
a possible exogenous factor, the history of real political scandals of each country, might not be
significant. Even though the trends in this cross-national comparison can be generalized, it would
be fruitful to apply a most different design in order to examine if cross-national differences are
present when compared with other political and media systems.
Taken together, this study was the first comparing effects of political scandal subtypes in
a cross-national manner. This research extends our knowledge of political scandals and their
effects on political support. Since political scandals affect symbolic power through decreased
trust and lowered character and competence reputations, media coverage of political scandals
might affect the level and legitimacy of democracy. Then, we should not underestimate the
power of news coverage on the course of democracy.
Limitations and future research
This research shows that Thomson’s (2000) breakdown of scandal types, sex, financial
and power scandals, is still very useful. With the limited small sample size of 217 caution must
be applied as the findings might not be transferable. Likewise, only 29% of respondents were
men so the female respondents are clearly dominant. Therefore, sex of respondents might be a
cofounding variable. Although statistical tests show significance and high internal validity,
22 the result. In addition this study should be replicated by using other news articles as vignettes to
assure a high external reliability.
In this study vignettes were designed not to show a minister’s party identification and the
multiparty political system of all the three countries study probably decreased the subjectivity of
partisan attachment. Future research studies, however, would do well to control for political
affiliation especially in presidential political systems or in highly politically polarized countries
(i.e. the USA).
The role of the media in defining political scandals should be further studied. Ekström and Johansson’s (2008) notion of “talk scandal” in which scandals originate within the media
deserves more attention. In better understand the role media plays in modern societies and in
forming the public opinion, we have to understand how media works to inform and frame
politics, and political scandals in particular. Additional cross-national studies using qualitative
research methods are needed to provide a bigger behind the scenes picture of how the media
coverage and political scandals together define the political landscape.
Appendices
Appendix A
Vignette with scandal treatments.
Breaking News: Minister caught [having an extramarital affair / using his
state credit card for private expenses / giving secret orders to the secret
service]
Reuters, Monday 28th of April 2014
Johannes Lind, the 44-years old minister, may be facing the end of his political career. Sources close to him have leaked an email conversation revealing that the minister [has been having an affair for the
23
to pay for his family’s vacations / has for the past two years been giving orders to the secret service to hack intranets of opposition parties.]
Johannes Lind was appointed as a minister after the last election. He began his political career almost fifteen years ago and has served two terms as a member of parliament. Before his ministerial duties he was a deputy chair of the committee on EU Affairs.
Minister Lind lives with his wife Maria and their two children Emma, 13, and Daniel, 7. Today, at 3 pm, he will give a press conference amid the biggest scandal of his career.
Follow the live coverage.
Vignette without a scandal treatment.
In the spotlight: Minister Johannes Lind
Reuters, Monday 28th of April 2014Johannes Lind, the 44-years old minister. He was appointed as a minister after the last election.
Johannes Lind began his political career almost fifteen years ago and has served two terms as a member of parliament. Before his ministerial duties he was a deputy chair of the committee on EU Affairs. Minister Lind lives with his wife Maria and their two children Emma, 13, and Daniel, 7.
Continue reading.
Appendix B Question wording.
Trustworthiness of a politician. How would you describe Minister Lind based on the
article you read? A 7-point bipolar scale. Unqualified-qualified, untrustworthy-trustworthy,
unreliable-reliable.
Competence of a politician. How would you describe Minister Lind based on the article
you read? A 7-point bipolar scale. Unskilled-skilled, undependable-dependable,
24
Character reputation. How would you describe Minister Lind based on the article you
read? A 7-point bipolar scale. Cold person-warm person, dishonest-honest, untruthful-truthful,
negative-positive, bad-good.
Trust in political institutions. Please, indicate the extent to which you agree or disagree
on the following statements. A 7-point scale from “strongly disagree” to “strongly agree”. After
reading about Minister Johannes Lind I’m more inclined to trust the government in my country /
political parties in my country / the media in my country / the legal system in my country.
Satisfaction with democracy. Based on the article you read, please indicate the extent to
which you agree or disagree on the following statements. I am satisfied with level of democracy
of my country. A 7-point scale from “strongly disagree” to “strongly agree”.
Appendix C
Table 3.
Means of politician evaluation
Type of scandal Trust on politician Competence rating Character reputation No treatment (control) 4.24 5.04 4.34 Sex scandal 3.23 4.55 3.23 Financial scandal 2.45 3.95 2.83 Power scandal 2.22 3.76 2.67
References
Bhatti, Y., Hansen, K. M., Olsen, A. L. (2013) Political hypocrisy: The effect of political
25 Bowler, S., Karp, J. A. (2004). Politician, scandals, and trust in government. Political Behavior,
26, 271-287.
Brants, K, van Praag, P. (2006). Signs of media logic: half a century of political communication
in the Netherlands. Javnost-the Public, 13, 25-40.
Chalaby, J.K. (2004). Scandal and the rise of investigative reporting in France. American
Behavioral Scientist, 47, 1194-1207.
Costas-Pérez, E., Solé-Ollé, A., Sorribas-Navarro, P. (2012). Corruption scandals, voter
information, and accountability. European Journal of Political Economy, 28, 469-484.
Dimitrova, D.V., Strömbäck, J. (2005). Mission accomplished? Framing of the Iraq War in the
elite newspapers in Sweden and the United States. Gazette, 67, 399-417.
Doherty, D., Dowling, C.M, Miller, M.G. (2011). Are financial or moral scandals worse? It
depends. Political Science & Politics. 44(4), 749-757.
Downey, J., Stanyer, J. (2013). Exposing politicians’ peccadilloes in comparative context:
explaining the frequency of political sex scandals in eight democracies using fuzzy set
qualitative comparative analysis. Political Communication, 30:3, 495-509.
Ekström, M., Johansson, B. (2008). Talk scandals. Media Culture Society, 30, 61-79.
Esser, F., Hartung, U. (2004). Nazis, pollution, and no sex: political scandals as a reflection of
26 Flere, S. & Lavric, M. (2008). On the validity of cross-cultural studies using student samples.
Field Methods, 20:4, 399-412.
Funk, C.L. (1996). The impact of scandal on candidate evaluations: an experimental test of the
role of candidate traits. Political Behavior, 8(1), 1-24.
Fischele, M. (2000). Mass response to the Lewinsky scandal: motivated reasoning or Bayesian
updating? Political Psychology, 21(1), 135-159.
Gordon & Bruner II (2009). Marketing Scales Handbook. A complication of multi-item measures
for consumer behavior & advertising research. GCBII Productions, Carbondale, Illinois, United States.
Goode, E., Ben-Yehuda, N. (1994). Moral panics: culture, politics, and societal construction.
Annual Review Sociology, 20, 149-171.
Hallin, D. C., Mancini, P. (2004). Comparing media systems. Three models of media and
politics. Cambridge University Press, Cambridge, United Kingdom.
Jiménez, F. (2004). The politics of scandal in Spain: morality plays, social trust, and the battle
for public opinion. American Behavioral Scientist, 47, 1099-1121.
Kantola, A., Vesa, J. (2013). Mediated scandals as social dramas: transforming the moral order
in Finland. Acta Sociologica, 56, 295-308.
Maier, J. (2011). The impact of political scandals on political support: an experimental test of
27 Miller, A. H. (1999). Sex, politics, and public opinion: what political scientists really learned
from the Clinton-Lewinsky scandal. Political Science and Politics, 32(4), 721-729.
Sanders, K. José Canel, M. (2006). A scribbling tribe: reporting political scandals in Britain and
Spain. Journalism. 7, 453-476).
Shah, D.V., Watts, M.D., Domke, D., Fan, D. P. (2002). News framing and cueing of issue regimes. Explaining Clinton’s public approval in spite of scandal. Public Opinion
Quarterly, 66, 339-370.
Schudson, M. (2004). Notes on scandal and Watergate legacy. American Behavioral Science, 47,
1231-1238.
Strömbäck, J. (2008). Four phases of mediatization: an analysis of the mediatization of politics.
The International Journal of Press/Politics, 13, 228-246.
Thompson, J.B. (2000). Political Scandal: Power and visibility in the media age. Polity Press,
Cambridge, United Kingdom.
Tumber, H. (2004). Scandal and media in the United Kingdom: from Major to Blair. American
Behavioral Scientist, 47, 1122-1137.
Tumber, H., Waisbord, S. (2004). Introduction: Political scandals and media across democracies.
American Behavioral Scientist, 47, 1031-1039.
Van Dalen, A., Skovsgaard, M. (2011). Er en politisk skandale en politisk skandale? – danske
28 Voter turnout (n.d.). IDEA. Retrieved from http://www.idea.int/vt/countryview.cfm?id=197s
Waisbord, S.R. (2004). Scandals, media, and citizenship in contemporary Argentina. American
Behavioral Scientist, 47, 1072-1098.
Welch, S.E. (2007). Political scandal and politics of exposure: from Watergate to Lewinsky and
beyond. Politics and ethics review, 3, 181-199.
Zaller, J.R. (1998). Monica Lewinsky’s contribution to political sciences. Political Science and