1

Politicians and journalists constructing political news.

A factorial survey with political journalists to analyze source selection patterns.

Debby Vos

Paper prepared for ECPR Graduate Conference, 3-5 July 2014, Innsbruck

INTRODUCTION

In modern politics, mass media play a crucial role in connecting voters to politicians since citizens

often rely only on news media to get informed about their representatives. Being visible in news

media is essential to acquire political success and influence political processes. Politicians thus have a

crucial interest in gaining favorable news coverage (Tresch, 2009). Politicians compete to get media

access and this study analyzes who wins this media battle by taking both politicians and journalists

into account. This is a relevant approach as the construction of political news results from a co-

production between newsmakers and journalists (Wolfsfeld & Sheafer, 2006). Politicians try to move

past media gates to convey their message to citizens and peers. However, journalists do not simply

transmit messages coming from politicians, but decide themselves which issues and sources to select

and how to report on them (Althaus, 2003).

In a broader perspective, the selection of politicians as news sources fits within the traditional

paradigm of media gatekeeping, which concerns the process through which numerous events and

actors are reduced to the few stories appearing in news media. The keyword during this process is

selection: selection decisions at multiple points in time define what gets into the news, what is left

out and how the news output eventually looks like (Shoemaker & Vos, 2009). Scholars describe

media gatekeeping as a hierarchical process with influential factors on multiple levels. Five levels of

influences have been identified: individuals, routine practices, media organizations, social institutions

and societies (Reese, 2007).

We examine which politicians get selected as a news source by journalists and explain the

selection patterns by looking at the first two levels: individual journalists and routine practices. On

the one hand, journalists are led by routinized practices of news work. Gatekeeping is not a random

process but results from systematic selection mechanisms that lead to a rather uniform news menu

across news media (Shoemaker & Vos, 2009). News values are such mechanisms guiding journalists

during news selection. We apply news values to politicians and consider their characteristics and

activities as assets that can guide them through the media gates. This leads to our first research

question: which characteristics and activities of politicians matter to get selected as a news source by

journalists?

2

On the other hand, the supply of politicians who have certain news values is only one side of

the coin. As stated before, politicians present themselves to journalists to gain coverage but

ultimately journalists have the final verdict. Journalists tend to follow news values but their personal

characteristics and beliefs might also shape news selection (Shoemaker, Eichholz, Kim, & Wrigley,

2001). We are interested whether journalists judge the newsworthiness of politicians according to

their own background and preferences or whether they are guided solely by routine selection

mechanisms as one professionalized group. With our second research question we examine if

journalists’ personal characteristics and beliefs influence their selection of politicians as a news

source?

Media gatekeeping research has a well-established and enduring tradition. White (1950) was

the first to connect the gatekeeping concept to communication theory and media gatekeeping has

been studied extensively ever since. Media gatekeeping typically is studied by employing content

analyses. However, a fuller understanding of the gatekeeping process is best accomplished by

combining content analyses with other research methods, such as surveys, interviews, observations

or experiments (Shoemaker & Vos, 2009, p. 81). We employ an experimental survey with 73 Flemish

journalists as an innovative method to examine news coverage of politicians. This method

complements existing research and offers some advantages.

First of all, conducting a vignette survey with journalists allows us to analyze news selection of

politicians on a disaggregate media level. Whereas most content analyses examine one or more

media outlet(s) and draw conclusions about news organizations (for example Fogarty, 2012; Midtbø,

2011; Waismel-Manor & Tsfati, 2011), we can look specifically at those persons who create news

items in the first place: individual journalists. Since political news results from the intertwined

relation between politicians and journalists, we need to include both on a micro-level to analyze

interactions between them. For example, some studies conclude that news from female reporters

consists of a greater number of female news sources (Aday & Devitt, 2001; De Swert & Hooghe,

2010). We can verify if this holds true when examining journalists’ selection decisions instead of

analyzing news output, where also editorial and organizational decisions come into play.

Second, experiments are an appropriate means to investigate unconscious processes that are

hard to observe or explain verbally. News selection is such a process that occurs in a rather

unconscious and routinized manner. Yet, experimental research in media gatekeeping studies is rare

(for exceptions see Hudson, 1992; Patterson & Donsbach, 1996). By means of a factorial survey, we

ask journalists to judge press releases from politicians, which contain carefully manipulated features

of the politician. The judgments of the journalists will most likely be less subject to social desirability,

because they are probably not aware of the manipulations in the vignettes (Wallander, 2009).

3

Third, the numerous content analyses in gatekeeping research imply that the news product is

investigated. This news product is only the very last step in the news making process, after the news

discovery and the news selection (McManus, 1994). We deviate from this dominant perspective by

examining the selection process, which comes much earlier in the news making process. Moreover,

whereas content analyses examine what becomes news, we can analyze which events and people do

not make it into the news. As Shoemaker & Vos (2009, p. 80) put it clearly: “[…] studying what does

not become news is perhaps more revealing of the decision-making process than is studying only the

news product.” With content analyses, it is hard to know the “population” of events where news gets

selected from, whereas we can control all incoming information and then examine who journalists

judge as newsworthy enough for covering.

THEORY AND HYPOTHESES

Media gatekeeping happens on several levels, all having their own influence on news selection

(Reese, 2007). This study focuses on the individual level and routine practices to explain news

coverage of politicians. On the level of routine practices, journalists share basic definitions of

newsworthiness. They select and make news based upon a generalized set of routine practices,

although these rules are not consciously acknowledged or written down (Shoemaker & Vos, 2009).

The theory of news values explains how ground rules of newsworthiness operate during the daily

news production. Several news values, such as ‘continuity’, ‘reference to something negative’ and

‘reference to persons’, determine which information moves through the news gates (Galtung &

Ruge, 1965; Harcup & O’Neill, 2001). One news value is ‘reference to elite people’: events concerning

elite people have more chance to become news because their actions are more consequential than

the activities of others (Galtung & Ruge, 1965, p. 68). However, it is not clear who these ‘elite people’

exactly are and if they all manage to become news sources. As Harcup and O'Neill (2001, p. 271)

state: “’Elite people’ is too broad a category to shed much light on what makes news in our current

cultural climate”.

In this paper, we narrow down the broad group of elite people and specifically focus on

political elites. We examine which features of politicians make journalist decide to report on them.

We regard the characteristics and activities of politicians ‘forces’ that facilitate or constrain

politicians’ movement through news gates (Shoemaker et al., 2001) and thus see them as news

values of politicians. We suppose that journalists all employ these selection mechanisms to judge the

newsworthiness of politicians and examine which characteristics and activities of politicians matter

to get selected as a news source by journalists.

4

In spite of journalists applying implicitly shared news norms, these norms are not binding and

reporters’ subjective beliefs and personal background might influence their judgments. Their

predispositions can lead to selective attention, selective perception and selective retention of

information (Donsbach, 2004). As Weaver and Wu (1998) put it simply: “The major assumption is

that journalists’ backgrounds and ideas have some relationship to what is reported (and how it is

covered) in the various news media around the world”. It is not only interesting to look at which

features of politicians are relevant for journalists, but also whether all reporters apply them in a

similar way. We analyze whether journalists’ personal characteristics and beliefs influence their

selection of politicians as a news source? Finding uniformity in newsworthiness judgments implies

that journalists mainly carry out a set of routine procedures without personal motives. If journalists

on the other hand differ in their selection decisions, the micro-level of journalists should matter to

explain political news coverage.

Features of politicians as news values

A whole range of features of politicians might determine if they become a news source, but we need

to choose those most interesting for us to include in the experimental survey. After all, in an

experimental setting it is important to include only few variables of interest and keep other

confounding factors as constant as possible. The decision of which features of politicians to analyze is

based upon three considerations. First, content analyses of political news often result in inconsistent

findings, where our experimental design might bring some clarification, as is the case for gender for

instance. Second, some possibly relevant features of politicians lack substantial research, such as the

specialization of politicians, which we now include in our experimental survey. Lastly, we are also

interested in interaction effects between politicians and journalists, for example between the party

of the politician and the political preference of the journalist. It is thus necessary to include features

of politicians where we might expect an interaction to occur.

The considerations above lead to five features of politicians worthwhile analyzing: party

attachment, gender, specialization, type of communication and type of action. These features

determine politicians’ newsworthiness and can be divided in politicians’ structural characteristics on

the one hand and their active engagement on the other. Party attachment and gender are fixed

treats whereas politicians can consciously decide how and when to communicate about an issue,

whether or not it is about their issue specialization, and how they act politically on that issue.

Party affiliation is a feature that cannot be ignored when analyzing news selection of Belgian

political actors, since political parties are highly present in Belgian politics and news coverage.

Belgium is considered to be a partitocracy with strong parties (De Winter, 1996) and news items

typically mention the party of the politician each time he gets covered. More importantly, including

5

parties has a clear theoretical relevance. The so-called ‘political balance’ or ‘partisan media bias’ has

been studied frequently to determine if certain political parties and its members are favored in news

coverage (Hopmann, Van Aelst, & Legnante, 2011). The strength of a party can influence the amount

of news coverage its members receives. First of all, media attention allocated to politicians is

supposed to be proportional to the electoral strength of their party. To secure balanced reporting,

journalists can apply the number of seats a party has gained in parliament, with politicians from

electoral stronger parties receiving more coverage as a result (Van Aelst, Maddens, Noppe, & Fiers,

2008). Second, government and opposition parties attract different media attention. Journalists are

inclined to favor politicians from government parties – the so-called ‘incumbency bonus’ – because

their decisions and actions are more consequential and thus more relevant. Government can actually

do something while opposition members can merely talk and criticize. Moreover, government

members are strong news sources because they have good access to news media, can offer exclusive

information and are the official sources for informing about political decisions (Hopmann, Elmelund-

Praestekaer, Albaek, Vliegenthart, & de Vreese, 2012; Midtbø, 2011).

H1a: Politicians from electoral stronger parties get selected more often by journalists.

H1b: Politicians from government parties get selected more often by journalists.

Gender is a second structural characteristic that might influence news coverage. The study of news

coverage of female politicians is a relatively young, but well-studied research field (Larson, 2001).

Scholars indicate the existence of a double gender bias in political news: female politicians receive

less news coverage (Heldman, Carroll, & Olson, 2005; Kahn, 1994; Niven, 2005) and also different

coverage (e.g. Aday & Devitt, 2001; Ross, 2004) compared to their male colleagues. Yet, some recent

studies fail to find gender differences in visibility in news coverage (e.g. Atkeson & Krebs, 2008;

Kittilson & Fridkin, 2008; Semetko & Boomgaarden, 2007). This might signify that women’s growing

presence in politics results gradually in a more equal coverage. In Belgium, female politicians have

conquered a firm position in politics and female representation is quite high: 41% of current federal

parliamentarians is female (Belgische Federale Overheidsdiensten, 2011,

http://www.belgium.be/nl/). We thus expect female politicians are being judged equally newsworthy

compared to their male colleagues.

H2: Male and female politicians get selected equally by journalists.

6

Parties are often associated with a particular issue they ‘own’. Consistent attention and policy action

on specific issues by parties makes citizens come to view the party as competent at handling those

issues, thereby creating stable and long-term issue-ownership (Petrocik, 1996). Hayes (2008, p. 380)

suggests that journalists as well are guided by partisan heuristics when reporting about politics: “[…]

party expectations influence journalists’ choices of which candidate statements to report and which

to ignore”. We take this a step further and apply the issue-ownership mechanism to individual

politicians, since we are interested in politicians and not parties as such. Most politicians who have

an elected mandate specialize in one or a few policy domains and become experts on it. That way,

they have their own issue-ownership. We propose that political journalist are driven by politicians’

specialization, in the same way they follow issue ownership heuristics. This also relates to the

concept of ‘thematic relevance’: the extent to which a political actor’s position can be linked to a

public issue (Wolfsfeld & Sheafer, 2006, p. 339). When a politician has knowledge about and

experience with the topic at hand, journalists will judge him as more relevant and thus newsworthy.

H3: Politicians who communicate about their issue specialization get selected more often

by journalists.

It is not only important to look at the issue politicians communicate about, but also at the timing of

their communication. They can try to initiate a debate and get initial control about the topic or on

the contrary they can respond to an ongoing mediatized debate, thereby ‘surfing the wave’. The

modern political process can be perceived of as a sequence of cycles in which leaders, publics and

mass media concentrate on a small number of public issues for a limited period of time (Wolfsfeld &

Sheafer, 2006, p. 335). Media generate ‘hot topics’ and politicians are attentive to it: by talking about

hot topics they try to surf the current wave of attention. This relates to continuity as a news value:

once an event has become headline news it remains in the media spotlight for a while because it has

become familiar and easy to interpret (Harcup & O’Neill, 2001, p. 263).

H4: Politicians who react on an ongoing debate get selected more often by journalists.

Next to raising relevant issues, politicians are supposed to take action upon societal problems. They

have a wide scope of activities to pursue. Firstly, politicians can try to set the political agenda by

pronouncing their personal standpoint and vision. Secondly, they can employ parliamentary

questions to voice their concerns, control government and prepare legislative acts (Bailer, 2011;

Wiberg, 1995). Thirdly, initiating and passing legislation are central in representative democracies.

Politicians want to pass bills through legislation to realize their policy goals, to demonstrate to

7

citizens that they are caring and acting about salient issues and to acquire a good reputation amongst

colleague politicians (Anderson, Box-Steffensmeier, & Sinclair-Chapman, 2003).

All these actions have their inherent political relevance, but are also applied to gain visibility

and name recognition. Taking actions can be inspired by publicity-seeking motives (Bailer, 2011;

Midtbø, 2011). We indeed propose that journalists are led by these types of action when selecting

news sources, although not all actions might be considered as equally newsworthy. We differentiate

between the symbolic and the substantial politic agenda, where issue positioning and asking

parliamentary questions concern the symbolic agenda, and bill proposals are a means to influence

the substantial agenda (Walgrave & Van Aelst, 2006). We expect that journalists are mainly

interested in actions directed to the substantial policy agenda since it has more consequential impact

on society and thus is more pertinent to cover. Moreover, drawing up bill proposals shows a good

deal of effort and knowledge put forward by the politician and stands out from the abundance of

parliamentary questions. In the year 2012-2013, to give an example, 4499 oral questions and 5288

written questions were asked compared to only 530 bill that were proposed (Kamer van

volksvertegenwoordigers, 2013).

H5: Politicians who undertake a substantial action get selected more often by journalists.

Features of journalists influencing selection decisions

Journalists in general select political news sources based upon the newsworthiness of politicians.

However, it is interesting to not only look at journalists as one homogenous group, but rather

differentiate among them along socio-demographic characteristics and personal beliefs. This way, we

can examine whether their personal background has an effect on their news work practices, or

whether they are entirely socialized into the newsroom and closely follow news values and working

routines developed over time.

To begin with, news decisions can reflect in some way journalists’ political beliefs. Previous

research concludes that partisan bias is hard to find when looking at news media as a whole, but this

does not mean that “[…] every reporter and every newspaper is unbiased. Quite the opposite: a wide

variety of data […] indicates that specific newspapers and specific reporters and editors can show

substantial (and substantive) ideological bias” (D’Alessio & Allen, 2000, p. 148). Reporters tend to be

more liberal than the news organization they work for (Peiser, 2000) so the effect of political bias

might be ruled out by the combination of more left leaning reporters and more right leaning owners

of media organizations. For that reason, we investigate partisan bias by looking at the one-on-one

relation between politicians’ party affiliation and journalists’ political preference.

8

Socio-demographics might also influence professional decisions. Journalists have diverse

beliefs and priorities consistent with their gender, age, experience and education, which then leads

to a different socialization into the workplace. Scholars have analyzed reporting styles according to

gender (Hanitzsch & Hanusch, 2012; Meeks, 2013; Rodgers & Thorson, 2003), but diverse viewpoints

exist as to whether female journalists report differently. The gender model contends that men and

women socialize differently in the workplace, because they have diverse beliefs and priorities. The

job model, on the other hand, asserts that socialization is a function of the work environment.

Women and men incorporate the prevailing rules and structures of the newsroom and develop

similar working routines (Rodgers & Thorson, 2003). Van Zoonen (1998, p. 36) lists the use of sources

as one area where men and women can differ in their reporting: women are inclined to select more

often female news sources. With our design, we can examine interaction effects between the gender

of the journalist and the gender of the politician.

In addition, we also investigate their years of experience as a journalist and the degree in

which they are specialized in political news coverage. Within the notion of socialization, it is plausible

that senior journalists have fully incorporated the working routines of the newsroom and are more

acquainted with professional norms of journalism (Cassidy, 2008). As a result, they might construct

political news differently compared to less experienced journalists only just entering the profession

of news worker. Journalists’ topical specialization can also influence news selection. Journalists with

the same specialization are more homogenous than generalist journalists and tend to apply similar

working routines to a larger extent (Peiser, 2000).

RQ: Do journalists’ personal characteristics and beliefs influence their selection of

politicians as a news source?

DATA AND METHOD

The factorial survey method

The factorial survey approach is a method intended to determine the underlying principles behind

human judgments of social objects. The main component are vignettes: short, carefully constructed

descriptions of a person, an object or a situation that are representing a systematic combination of

characteristics. Respondents typically judge multiple vignettes to measure their belief or intended

behavior concerning each presented vignette scenario. These vignettes consist of various dimensions

which in turn all have several levels. The total vignette population is obtained by a full factorial

9

combination of all levels (Atzmüller & Steiner, 2010). By letting each of the dimensions vary

independently with respect to its levels, factor orthogonality is obtained across dimensions. This

means that respondents only need to judge a randomly or systematically drawn sample of the

vignette universe to disentangle the unique effect of dimensions that are normally very highly

correlated. However, factor orthogonality also implies that respondents may be presented with

descriptions of situations that seldom occur in the real world. Therefore, it is important to exclude

implausible combinations and pay attention to the degree of realism of the vignettes (Wallander,

2009). In our case, no implausible combinations are present and we ask a control question to

measure the degree of realism of the vignette scenarios.

Operationalization of dimensions

Based upon our theoretical expectations, we include five dimensions in our vignettes, each with two

or four levels (see table 1).

Dimension Levels

Gender Male - Female

Party Green - Socialist - Liberal - Flemish Nationalists

Issue Defense - Fiscality - Judiciary - Immigration

Type of communication Initiate - React

Type of action Standpoint - Oral question - Written question - Bill proposal

Table 1: Operationalization of dimensions

We operationalize the gender dimension by including names of real male and female Dutch-speaking

members of Belgian federal parliament. This advances the degree of realism and thus increases

ecological validity. The choice for which MPs specifically to include1 was firstly based upon the

combination of gender and party: we selected one male and one female MP of each of the four

parties in the vignettes, resulting in eight MPs appearing each in one of the eight vignettes randomly

presented to respondents. Moreover, four of the selected MPs have a clear specialization on one of

the four issues in the vignettes whereas the other four are rather generalists2. To control for an

effect of using real MPs, we measured their perceived newsworthiness by asking respondents: “How

1 The MPs are: Wouter Devriendt and Meyrem Almaci (Greens), Renaat Landuyt and Caroline Gennez

(Socialists), Mathias De Clercq and Carina Van Cauter (Liberals), Theo Francken and Sarah Smeyers (Flemish Nationalists). 2 The determination of specialization was based upon interviews with federal MPs where they were asked in

which issues they are specialized, with the support of the European Research Council (Advanced Grant ‘INFOPOL’, N° 295735) and of the Research Fund of the University of Antwerp (Grant N° 26827). Moreover, we selected only those MPs with a clear distinctive specialization that political journalists are aware of according to our assessment. This resulted in the following specializations: Devriendt – defense, Almaci – fiscality, Landuyt – judiciary, Francken – migration & defense.

10

often to the following actors succeed in making it into the news?” The respondents each had to judge

eight politicians - randomly chosen from a group of 50 Belgian politicians amongst which also our

eight MPs - on a five-point Likert scale (1 = never, 2 = rarely, 3 = sometimes, 4 = often, 5 = always).

The mean scores range between 2,7 (Mathias De Clercq) and 3,9 (Theo Francken), so all of them are

well-known by political journalists and appear regularly – but not always – in news media. We also

include this measure as a control variable in our analyses.

Secondly, we include one left government party (Socialists), one right government party

(Liberals), one left opposition party (Greens) and one right opposition party (Flemish Nationalists).

This way, we can easily test hypothesis H1b while keeping the politicians’ ideology constant.

Furthermore, we selected consciously four issues: national defense (removal of nuclear bombs on

Belgian ground), fiscality (regulation of salaries and bonuses of bank managers), judiciary

(replacement of outdated prisons), and asylum and migration (residence permits for under aged

asylum seekers). None of these topics are clearly owned by one of the four parties, but four of the

MPs have a clear specialization concerning one of these issues. This way, we are able to focus on the

effect of MP specialization and rule out the influence of issue-ownership. Concerning the type of

communication, initiating is operationalized by MPs who “want to reopen the debate on”, whereas

MPs who react state that they “want to react to the recent news about”. Lastly, we manipulate

intended action. They could “advocate for” (personal standpoint), “ask an oral question during the

next plenary session” (oral question), “draw up a written question” (written question) or “submit a

law proposal” (bill proposal).

Two variables on the MP level were constructed afterwards. Electoral strenght of the party

consists of the amount of seats in Federal Parliament after the 2010 federal elections (N-VA: 27, sp.a:

13, Open VLD: 13, Groen: 5). Whether the MP communicates about his matching specialization was

defined by combining the MP who communicates in the vignette with the topic he communicates

about.

Next to the five manipulated dimensions, we also hold certain dimensions constant. First of all,

we do not manipulate the function of politicians, but consistently present MPs. Previous studies have

shown the strong effect of political position on news coverage (e.g. Cook, 1986; Midtbø, 2011;

Schoenbach, Ridder, & Lauf, 2001), so we need to hold it constant to test other possibly influential

features of politicians. Second, all vignette texts are presented as press releases coming from one

federal MP. Press releases are still a commonly used by MPs to solicit news coverage as they are easy

to set up and a quick manner to communicate actions and policy stances to journalists (Fogarty,

2008). Third, every press release is negatively formulated, urges for actions from government or a

cabinet member, and contains a quote from the MP. This way, the press releases are conceived more

realistic and have already a certain degree of newsworthiness.

11

Taking all manipulated and constant dimensions together leads to a vignette with five manipulated

dimensions, as the example below shows (a man of the Greens reacts on a defense issue and asks a

written question):

“The B61-nucleair bombs do not need to be modernized, but rather destroyed”,

responds Green Member of Parliament Wouter Devriendt on the recent news about the

modernization of the nuclear weapons stored in Kleine Brogel. “Nuclear weapons are

dangerous and useless. Moreover, the modernization, the storage, the maintenance and

the surveillance of the nuclear bombs are extremely expensive. The government needs to

undertake action to commence and finish a complete nuclear disarmament.” Devriendt

wants to gain clarity about the measures concerning nuclear weapons in Belgium by

asking a written question to the authorized cabinet member.

By multiplying all levels (2 x 4 x 4 x 2 x 4), we become a vignette universe of 256 vignettes. We draw a

half faction factorial sample which allows for statistically efficient estimations of all direct effects and

first and second order interactions, resulting in 128 vignettes that are presented to respondents.

Survey and respondents

A typical vignette study consists of two components: 1) a vignette experiment as the core element,

and 2) a traditional survey to measure additional respondent-specific characteristics (Atzmüller &

Steiner, 2010). Each respondent first got a set of eight vignettes randomly drawn from the sample of

128 vignettes. They judged the newsworthiness of the press release on a 7-point Likert scale (1 =

definitely not, 7 = definitely) based upon the question: “Would you make a news story based upon

this information?” With a second judgment, we check the newsworthiness of the MP in the press

release on the same 7-point scale: “If you would make a news story, would you mention [name MP]

in the news story?”

Afterwards, several follow-up questions were asked. Firstly, the respondents judged the

degree of realism of the vignettes on a 10-point Likert scale (1 = totally unrealistic, 10 = totally

realistic). Overall, the respondents assess the press releases as rather realistic (mean = 5,9; SD =

1,86). This measurement is also included as a control variable in the analyses. Secondly, we gauged

which issues they find salient, which parties they associate with issues, and how often certain

politicians succeed in becoming a news source. To end, we measured respondents’ socio-economic

background and political beliefs.

12

The survey was distributed as an online survey. Our target group consists of political

journalists, but it is not straightforward to define who those journalists exactly are. In a first step, we

consulted the ‘journalistendatabank’ (http://www.avbb.be), which gathers information on more than

5000 journalists in Belgium. We searched for Flemish speaking journalists specialized in politics which

resulted in a list of 167 Flemish political journalists. However, when looking more into detail, the list

was not exhaustive: journalists who haven’t indicated their language for example, were not included

in the list. Hence, the total group of Flemish political journalists is probably larger than 167. In a

second step, we checked the websites of news organizations as well as actual newspapers and news

broadcasts to find additional political journalists. We interpreted political journalist broadly by also

selecting those journalists who write sometimes about politics.

Eventually, 293 journalists were contacted in March and April 2014 by a first e-mail, two

reminder e-mails and phone calls. Of those, 167 journalists did not enter the survey (57%), 26

journalists indicated that they were currently not working as a journalist or that they never report on

politics (9%), 25 journalists entered the survey, but did not finish it (8%) and 75 journalists did fill in

the survey (26%). However, two journalists pointed out in the follow-up question that none of their

articles contains political actors, so we exclude them from analyses, leaving us with 73 respondents.

Two recent surveys with journalists in Belgium yield similar response rates: Raeymaeckers, Paulussen

and De Keyser (2012) convinced 31% of Belgian journalists to participate and Raeymaeckers et al.

(2013) got response of 33% of Flemish journalists. Our group of respondents (see table 2) resembles

to a very large extent the respondents of Raeymaeckers et al. (2013), except for our political

journalists being higher educated than Flemish journalists in general.

Variable Measurement N Mean SD Min Max

Sex 1 = male, 2 = female 73 1,32 0,46 1 2

Ideology 0 = totally left, 10 = totally right 60 4,15 1 0 7

Journalistic experience In years 68 17,50 11 1 43

Political specialization Amount of news stories with political

actors on 10 last news stories 59 6,51 3 1 10

Table 2: Descriptives of respondents

Analyses

Since each respondent rated eight vignettes, the observations in our dataset are not independent

from each other. We account for this clustering by employing multilevel regression models. We take

each judgment on a single vignette as one case, leading to a total of 584 cases on the lowest level of

vignettes, coming from 73 respondents on the highest level.

13

RESULTS

The dependent variable in our analyses is respondents’ judgment about the newsworthiness of the

press release coming from a politician, which is normally distributed as can be seen in figure 1a

(range = 1 – 7; mean = 3,9; SD = 1,71). We checked with a second rating question whether the

politician in the press release would indeed become a news source if the respondent would make a

news story. As figure 1b shows, this variable is positively skewed (range = 1 – 7; mean = 5,83; SD =

1,46). We can conclude that if journalists select a press release for coverage, they would also include

the politician who sends the press release as a news source. Consequently, it is plausible to use the

first, normally distributed rating as our dependent variable.

Figure 1a and 1b: Distribution of ratings of press releases (1a) and ratings of politicians (1b)

First of all, we test whether we can verify our theoretical hypotheses about politicians. Model one

shows the direct effects of characteristics and activities of politicians on selection by journalists. To

begin with, we test whether party affiliation influences politicians’ newsworthiness, and look at

parties’ electoral strength and government versus opposition parties. H1a gets confirmed: politicians

from parties with a larger faction in parliament get selected more often. Concerning H1b, there is a

clear positive effect of being a government MP, which approaches significance level (p = 0,07). Purely

statistically, we cannot conclude that journalists select MPs belonging to governments parties more

often than MPs from opposition parties, but there is a clear trend in that direction. With H2, we test

differences in the selection of male and female MPs, but as expected women do not become a news

source less often (p = 0.61). Our third hypothesis concerns specialization. We think that MPs who

communicate about their own issue specialization are considered as more newsworthy. This

050

10

015

0

Fre

qu

en

cy

0 2 4 6 8PressreleaseJudgment

050

10

015

020

025

0

Fre

qu

en

cy

0 2 4 6 8PoliticianJudgment

14

expectation is not confirmed however (p = 0.97)3. With H4, we want to gauge the type and timing of

politicians’ communication. We believe that politicians who try to ride the wave by responding to an

ongoing debate are more likely to become a news source than politicians who try to initiate a debate,

which indeed gets confirmed in the analysis. Our last hypothesis examines the influence of the type

of action. Consistent with our expectation, introducing a bill to influence the substantial agenda has a

significant positive effect on journalists’ selection decisions and this effect is bigger than actions with

a more symbolic meaning. Moreover, it appears that asking oral questions has the smallest effect on

getting selected, whereas stating a personal standpoint and asking a written question take the

middle position. To conclude, we point to the insignificant effect of the perceived newsworthiness of

the MPs (p = 0.51) which indicates that the effects above are not a consequences of the real MPs

chosen.

Secondly, we want to examine whether journalists judge newsworthiness of politicians as one

homogeneous group or whether their personal characteristics and beliefs influence to some extent

their selection decisions. With model 2 we check whether individual journalists make other

judgments in general whereas model 3 makes the explicit connection between journalists’ features

and politicians’ features. The models show that journalists do not differ among each other when it

comes to judging newsworthiness. They select the same politicians regardless whether they are male

or female (p = 0.74), more right-leaning (p = 0.69), more experienced as a journalist (p = 0.56) or

more specialized in political news (p = 0.11). The interactions between politicians and journalists do

not yield any significant effect either, indicating that journalists as one professional group apply the

same selection mechanisms.4 We also included journalists’ perception of the realism of the vignettes

as a control in both models and this does has a significant effect. When the respondents perceive the

vignettes as being more realistic, they rated its newsworthiness higher.

3 With a follow-up question, we measured which parties journalists link to certain issues. The results show that

they indeed connect some parties to certain issues in the vignettes. Consequently, we tested whether the non-significant effect of specialization could be explained by the perceptions of issue-ownership. This appears to be not the case, since the effect of issue-ownership on selection is insignificant (β = 0.20, SD = 0.14, p = 0.17). 4 Next to individual features of journalists, we also looked at the level of the news organization to examine

whether journalists working for television news select differently than newspaper journalists. Similar to the individual features, they don’t yield significant results (respectively β = -0.11, SD = 0.44, p = 0.80; β = -0.14, SD 0.42, p = 0.74) and thus cannot explain variance between journalists.

15

Model 1 Model 2 Model 3

Level 1 (politician)

Size party 0.02* (0.01)

0.02 (0.01)

Government party (Ref. = opposition party) 0.30 (0.17)

0.22 (0.21)

Gender (Ref. = male) 0.07 (0.14)

-0.28 (0.43)

Specialization (Ref. = not specialization) 0.01 (0.18)

0.31 (0.44)

Type communication (Ref. = initiate) 0.21* (0.11)

0.22 (0.13)

Type action (Ref. = personal standpoint)

Oral question -0.31* (0.15)

-0.29 (0.19)

Written question -0.20 (0.15)

-0.16 (0.19)

Law proposal 0.39** (0.15)

0.45* (0.19)

Level 2 (journalist)

Gender

0.13 (0.40) -0.25 (0.62)

Ideology

-0.04 (0.11) -0.06 (0.12)

Journalistic experience

0.01 (0.02) 0.01 (0.02)

Political specialization

0.09 (0.05) 0.10 (0.06)

Interaction effects (level 1*level 2)

Gender*Gender

0.25 (0.31)

Party*Ideology

0.03 (0.05)

Specialization*Specialization

-0.07 (0.06)

Controls

Perceived newsworthiness (politician) -0.18 (0.28)

-0.07 (0.38)

Realism of vignettes (journalist)

0.20* (0.09) 0.20* (0.09)

Intercept 3.39*** 2.12 2.30

N (journalists) 73 54 54

N (vignettes) 584 432 432

Residual variance

Level journalists 1.07 1.04 1.04

Level politicians 1.29 1.40 1.36

Rho 0.41 0.35 0.37

Table 3: Multilevel linear regression (random effects) with news selection by journalists as dependent variable. Note: Standard errors in parentheses. * p ≤ 0.05, ** p ≤ 0.01, *** ≤ 0.001.

CONCLUSIONS AND DISCUSSION

Politicians and journalists are the two main actors in constructing political news. Their relationship is

intertwined and symbiotic as both control key resources. Politicians are in control of exclusive

information and are the official sources who grant legitimacy to news stories. Journalists on the

other hand act as gatekeepers and define the extent to which politicians get visibility in news media

16

(Strömback & Nord, 2006). The connection between journalists and news sources is often described

as being a ‘tango’: “The relationship between sources and journalists resembles a dance, for sources

seek access to journalists, and journalists seek access to sources. Although it takes two to tango,

either sources or journalists can lead, but more often than not, sources do the leading” (Gans, 1979,

p. 239). This quote fits our findings: political news sources prevail journalists’ personal preferences in

the negotiation of newsworthiness.

We conducted a factorial survey with 73 Flemish political journalists to examine news

coverage of politicians. We are interested firstly in which politicians succeed to become news sources

and secondly in selection decisions by individual journalists as being the first gatekeepers. Over time,

routine practices have developed by which journalists all apply news values in a similar manner. We

tested which news values of politicians guide journalist in their source selection and if these working

mechanisms are also influenced by journalists’ personal characteristics and beliefs. We conclude that

journalists as a profession indeed follow certain news values typically for politicians and could not

find differences in judgments between individual journalists, which has implications for both

journalists and politicians.

First of all, our results support the notion of a similar socialization in the newsroom regardless

ones gender, ideological preference or experience. Political journalists are a homogenous group of

highly professionalized workers who are not influenced by their own ideas and preferences. Looking

back at the media gatekeeping model (Reese, 2007), it appears that routine practices exert more

influence on news construction than individual characteristics of journalists. The traditional starting

point of White’s study was the individual reporter who published stories based upon his “own set of

experiences, attitudes and expectations” (White, 1950, p. 386). Later on, scholars contradicted the

idea of highly subjective news workers by showing that news decisions are primarily based on the

professional norms of journalism and media routines (Cassidy, 2008). Our study adds to this

conception and to theories about media logic. The media logic implies that news values become

more determining in news reporting and journalists have increasingly common views about what is

inherently relevant and interesting for the public (Van Aelst et al., 2008).

Secondly, politicians themselves can enhance their newsworthiness by actively engaging in

their communication and actions. Next to their party attachment, which is a fixed characteristic, MPs

can steer their news coverage by paying attention to their timing of communication and their

political actions. Journalists select politicians who react on an ongoing debate rather than politicians

who try to put an issue on the agenda themselves. This relates to literature on the political and the

media agenda: which agenda determines the other one? Our results suggest that journalists prefer

information already on the media agenda instead of incorporating issues originated from the political

agenda. News media are not passive transmitters of the political agenda, but have their own

17

professional interests. They do not simply reflect political realities, but rather actively choose

mediatized issues and transform them into news stories that are relevant and culturally resonant

(Sheafer & Wolfsfeld, 2009).

Surprisingly, journalists do not judge politicians communicating about their own field of

expertise as being more newsworthy. Several explanations come to mind. First, the dimension of

issue specialization might not have been operationalized clearly enough. The vignettes scenarios did

not state explicitly when an MP was communicating about his expertise. Moreover, this implies that

journalists themselves do not recognize specialized MPs. With an ever faster news cycle, most

journalists are generalists themselves instead of specialized in politics. However, our results do not

indicate that journalists who indeed are political experts select sources differently than generalist

journalists. A second plausible reason is that journalists keep their public in mind and suppose that

readers and viewers do not have any knowledge about parliamentarians’ field of expertise. As a

consequence, politicians talking about their specialization are not perceived as more relevant the

audience. A third explanation originates from the real political world: specialist MPs might not exist

(anymore). MPs can focus somewhat more on a limited amount of issues, but not necessarily

become an expert in that field.

To end, we reflect on the use of experimental designs to study news selection. Most media

gatekeeping studies are based upon content analyses of the media product. We focused on the

selection phase where journalists choose which events and persons to report on. This way, we add to

previous research on news coverage of politicians by clarifying inconsistent results or validating

earlier findings. Concerning gender for example, it remains unclear whether female politicians in

present day politics still receive less coverage. Our findings support the idea that news coverage does

not depend on the gender of a politician but on a range of other elements. This might indicate that

gender differences might not occur when journalists select political news sources, but rather arise

during the editing phase. Additionally, we find support for earlier findings that the way politicians

communicate (Sellers & Schaffner, 2007) and the activities they undertake (e.g. Midtbø, 2011;

Tresch, 2009) can raise their newsworthiness.

18

REFERENCES

Aday, S., & Devitt, J. (2001). Style over substance: newspaper coverage of Elizabeth Dole’s

presidential bid. The International Journal of Press/Politics, 6(2), 52–73.

Van Aelst, P., Maddens, B., Noppe, J., & Fiers, S. (2008). Politicians in the news: media or party logic?

Media attention and electoral success in the Belgian election campaign of 2003. European

Journal of Communication, 23(2), 193–210.

Althaus, S. L. (2003). When news norms collide, follow the lead: New evidence for press

independence. Political Communication,, 20(4), 381–414.

Anderson, W. D., Box-Steffensmeier, J. M., & Sinclair-Chapman, V. (2003). The Keys to Legislative

Success in the U.S. House of Representatives. Legislative Studies Quarterly, 28(3), 357–386.

Atkeson, L. R., & Krebs, T. (2008). Press coverage of mayoral candidates: the role of gender in news

reporting and campaign issue speech. Political Research Quarterly, 62(2), 239–252.

Atzmüller, C., & Steiner, P. M. (2010). Experimental Vignette Studies in Survey Research.

Methodology: European Journal of Research Methods for the Behavioral and Social Sciences,

6(3), 128–138.

Bailer, S. (2011). People’s Voice or Information Pool? The Role of, and Reasons for, Parliamentary

Questions in the Swiss Parliament. The Journal of Legislative Studies, 17(3), 302–314.

doi:10.1080/13572334.2011.595123

Cassidy, W. P. (2008). Traditional in Different Degrees: The Professional Role Conceptions of Male

and Female Newspaper Journalists. Atlantic Journal of Communication, 16(2), 105–117.

doi:10.1080/15456870701840020

Cook, T. (1986). House Members as newsmakers: the effects of televising congress. Legislative

Studies Quarterly, 11(2), 203–226.

D’Alessio, D., & Allen, M. (2000). Media bias in presidential elections: a meta-analysis. Journal of

Communication, 50(4), 133–156.

Donsbach, W. (2004). Psychology of news decisions. Factors behind journalists’ professional

behavior. Journalism, 5(2), 131–157.

Fogarty, B. J. (2008). The strategy of the story: media monitoring legislative activity. Legislative

Studies Quarterly, 33(3), 445–469.

Fogarty, B. J. (2012). Local newspapers, House members, and source usage. The Social Science

Journal, 49(3), 275–283.

Galtung, J., & Ruge, M. (1965). The structure of foreign news: the presentation of the Congo, Cuba

and Cyprus crises in four Norwegian newspapers. Journal of International Peace Research,

2(1), 64–91.

19

Gans, H. J. (1979). Deciding what’s news: a study of CBS evening news, NBC nightly news, newsweek

and Time. New York: Pantheon Books.

Hanitzsch, T., & Hanusch, F. (2012). Does gender determine journalists’ professional views? A

reassessment based on cross-national evidence. European Journal of Communication, 27(3),

257–277.

Harcup, T., & O’Neill, D. (2001). What is news? Galtung and Ruge revisited. Journalism Studies, 2(2),

261–280.

Hayes, D. (2008). Party Reputations, Journalistic Expectations: How Issue Ownership Influences

Election News. Political Communication, 25(4), 377–400.

Heldman, C., Carroll, S., & Olson, S. (2005). “She brought only a skirt”: print media coverage of

Elizabeth Dole’s bid for the republican presidential nomination. Political Communication,

22(3), 315–335.

Hopmann, D. N., Van Aelst, P., & Legnante, G. (2011). Political balance in the news: a review of

concepts, operationalizations and key findings. Journalism, 13(2), 240–257.

Hopmann, D. N., Elmelund-Praestekaer, C., Albaek, E., Vliegenthart, R., & de Vreese, C. H. (2012).

Party media agenda-setting: how parties influence election news coverage. Party Politics,

18(2), 173–191.

Hudson, T. J. (1992). Consonance in Depiction of Violent Material in Television News. Journal of

Broadcasting & Electronic Media, 36, 411.

Kahn, K. F. (1994). The distorted mirror: press coverage of women candidates for statewide office.

The Journal of Politics, 56(1), 154–173.

Kamer van volksvertegenwoordigers. (2013). Belgische Kamer van volksvertegenwoordigers. Gewone

zitting 2012-2013: activiteitenverslag. (p. 93). Brussel: Kamer van volksvertegenwoordigers.

Kittilson, M. C., & Fridkin, K. (2008). Gender, candidate portrayals and election campaigns: a

comparative perspective. Politics & Gender, 4(3), 371–392.

Larson, S. G. (2001). American women and politics in the media: A review essay. Political Science and

Politics, 34(2), 227–230.

McManus, J. H. (1994). Market-driven journalism : let the citizen beware? Thousand Oaks Calif: Sage.

Meeks, L. (2013). He Wrote, She Wrote Journalist Gender, Political Office, and Campaign News.

Journalism & Mass Communication Quarterly, 90(1), 58–74.

Midtbø, T. (2011). Explaining media attention for Norwegian MPs: a new modelling approach.

Scandinavian Political Studies, 34(3), 226–249.

Niven, D. (2005). Gender bias? Media coverage of women and men in Congress. In S. Tolleson-

Rinehart & J. Josephson (Eds.), Gender and American politics. Women, men, and the political

process (Vol. 2, pp. 264–283). New York: Sharpe.

20

Patterson, T. E., & Donsbach, W. (1996). News decisions: journalists as partisan actors. Political

Communication, 13(4), 455–468.

Peiser, W. (2000). Setting the journalist agenda: Influences from journalists’ individual characteristics

and from media factors. Journalism & mass communication quarterly, 77(2), 243–257.

Petrocik, J. R. (1996). Issue Ownership in Presidential Elections, with a 1980 Case Study. American

Journal of Political Science, 40(3), 825–850.

Raeymaeckers, K., Heinderyckx, F., De Vuyst, S., Libert, M., De Maeyer, J., De Dobbelaer, R., Le Cam,

F., et al. (2013). De Belgische journalist in 2013: een zelfportret. (p. 58). Gent: Academia

Press.

Raeymaeckers, K., Paulussen, S., & De Keyser, J. (2012). A survey of professional journalists in

Flanders (Belgium). In D. H. Weaver & L. Willnat (Eds.), The global journalist in the 21st

century. New York, NY, USA: Routledge.

Reese, S. (2007). Journalism research and the hierarchy of influences model: a global perspective.

Brazilian Journalism Research, 3(2), 29–42.

Rodgers, S., & Thorson, E. (2003). A socialization perspective on male and female reporting. Journal

of Communication, 53(4), 658–675.

Ross, K. (2004). Women framed: the gendered turn in mediated politics. In K. Ross & C. Byerly (Eds.),

Women and media. International perspectives (pp. 60–80). Oxford: Blackwell Publishing.

Schoenbach, K., Ridder, J. D., & Lauf, E. (2001). Politicians on TV news: getting attention in Dutch and

German election campaigns. European Journal of Political Research, 39(4), 519–531.

Sellers, P. J., & Schaffner, B. F. (2007). Winning coverage in the U.S. Senate. Political Communication,

24(4), 377–391.

Semetko, H., & Boomgaarden, H. (2007). Reporting Germany’s 2005 Bundestag election campaign:

was gender an issue? The Harvard International Journal of Press/Politics, 12(4), 154–171.

Sheafer, T., & Wolfsfeld, G. (2009). Party systems and oppositional voices in the news media: A Study

of the Contest over Political Waves in the United States and Israel. The International Journal

of Press/politics, 14(2), 146–165.

Shoemaker, P., Eichholz, M., Kim, E., & Wrigley, B. (2001). Individual and routine forces in

gatekeeping. Journalism & Mass Communication Quarterly, 78(2), 233–246.

Shoemaker, P. J., & Vos, T. P. (2009). Media gatekeeping. An integrated approach to communication

theory and research (2nd ed., pp. 75–87). New York: Routledge.

Strömback, J., & Nord, L. W. (2006). Do politicians lead the tango? A study of the relationship

between Swedish journalists and their political sources in the context of election campaigns.

European Journal of Communication, 21(2), 147–164.

21

De Swert, K., & Hooghe, M. (2010). When do women get a voice? Explaining the presence of female

news sources in Belgian news broadcasts (2003-2005). European Journal of Communication,

25(1), 69–84.

Tresch, A. (2009). Politicians in the media: determinants of legislators’ prescence and prominence in

Swiss newspapers. The International Journal of Press/Politics, 14(1), 67–90.

Waismel-Manor, I., & Tsfati, Y. (2011). Why do better-looking Members of Congress receive more

television coverage? Political Communication, 28(4), 440–463.

Walgrave, S., & Van Aelst, P. (2006). The contingency of the mass media’s political agenda setting

power: toward a preliminary theory. Journal of Communication, 56(1), 88–109.

Wallander, L. (2009). 25 years of factorial surveys in sociology: A review. Social Science Research,

38(3), 505–520.

Weaver, D. H., & Wu, W. (1998). The global journalist: News people around the world. Hampton Press

Cresskill, NJ.

White, D. M. (1950). The gatekeeper: A case study in the selection of news. Journalism quarterly,

27(4), 383–390.

Wiberg, M. (1995). Parliamentary questioning: control by communication? Parliaments and majority

rule in Western Europe, 179–222.

De Winter, L. (1996). Comparing similar countries: Italy and Belgium. Res Publica, 48(2), 215–236.

Wolfsfeld, G., & Sheafer, T. (2006). Competing actors and the construction of political news: the

contest over waves in Israel. Political Communication, 23(3), 333–354.

Van Zoonen, L. (1998). One of the girls? The changing gender of journalism. In C. Carter, G. Branston,

& S. Allan (Eds.), News, gender and power (pp. 33–46). London: Routledge.