party politics 2012

7/29/2019 Party Politics 2012

1/20

http://ppq.sagepub.com/Party Politics

http://ppq.sagepub.com/content/18/2/173The online version of this article can be found at:

DOI: 10.1177/1354068810380097

2012 18: 173 originally published online 16 December 2010Party PoliticsClaes H. de Vreese

David N. Hopmann, Christian Elmelund-Prstekr, Erik Albk, Rens Vliegenthart andParty media agenda-setting : How parties influence election news coverage

Published by:

http://www.sagepublications.com

On behalf of:

Political Organizations and Parties Section of the American Political Science Association

can be found at:Party PoliticsAdditional services and information for

http://ppq.sagepub.com/cgi/alertsEmail Alerts:

http://ppq.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

http://ppq.sagepub.com/content/18/2/173.refs.htmlCitations:

What is This?

- Dec 16, 2010OnlineFirst Version of Record

- Feb 21, 2012Version of Record>>

at University of Bucharest on December 4, 2012ppq.sagepub.comDownloaded from
http://ppq.sagepub.com/http://ppq.sagepub.com/http://ppq.sagepub.com/http://ppq.sagepub.com/content/18/2/173http://ppq.sagepub.com/content/18/2/173http://www.sagepublications.com/http://www.apsanet.org/~pop/http://www.apsanet.org/~pop/http://ppq.sagepub.com/cgi/alertshttp://ppq.sagepub.com/cgi/alertshttp://ppq.sagepub.com/cgi/alertshttp://ppq.sagepub.com/subscriptionshttp://ppq.sagepub.com/subscriptionshttp://www.sagepub.com/journalsReprints.navhttp://www.sagepub.com/journalsReprints.navhttp://www.sagepub.com/journalsPermissions.navhttp://ppq.sagepub.com/content/18/2/173.refs.htmlhttp://ppq.sagepub.com/content/18/2/173.refs.htmlhttp://ppq.sagepub.com/content/18/2/173.refs.htmlhttp://online.sagepub.com/site/sphelp/vorhelp.xhtmlhttp://online.sagepub.com/site/sphelp/vorhelp.xhtmlhttp://online.sagepub.com/site/sphelp/vorhelp.xhtmlhttp://ppq.sagepub.com/content/early/2010/12/01/1354068810380097.full.pdfhttp://ppq.sagepub.com/content/early/2010/12/01/1354068810380097.full.pdfhttp://ppq.sagepub.com/content/18/2/173.full.pdfhttp://ppq.sagepub.com/content/18/2/173.full.pdfhttp://ppq.sagepub.com/http://ppq.sagepub.com/http://ppq.sagepub.com/http://online.sagepub.com/site/sphelp/vorhelp.xhtmlhttp://ppq.sagepub.com/content/early/2010/12/01/1354068810380097.full.pdfhttp://ppq.sagepub.com/content/18/2/173.full.pdfhttp://ppq.sagepub.com/content/18/2/173.refs.htmlhttp://www.sagepub.com/journalsPermissions.navhttp://www.sagepub.com/journalsReprints.navhttp://ppq.sagepub.com/subscriptionshttp://ppq.sagepub.com/cgi/alertshttp://www.apsanet.org/~pop/http://www.sagepublications.com/http://ppq.sagepub.com/content/18/2/173http://ppq.sagepub.com/


2/20

Article

Party mediaagenda-setting: Howparties influence electionnews coverage

David N. Hopmann,Christian Elmelund-Prstekr and

Erik AlbkUniversity of Southern Denmark, Denmark

Rens Vliegenthart and

Claes H. de VreeseUniversity of Amsterdam, The Netherlands

Abstract

Political parties have substantial influence on which issues the news media cover duringelection campaigns, while the media have limited influence on party agendas. However,we know little about why some parties are more successful than others in passing themedias gates and being covered on sponsored issues. On the basis of contentanalyses of election news coverage (812 news stories) and press releases published bypolitical parties (N 334) during the 2007 national election campaign in Denmark,

we analyse which parties were successful in appearing in the news on issues on whichthey published press releases. Using Sartoris notion of relevant parties, we concludethat the more relevant parties have more success, that there is a positive spillovereffect from other parties press releases, but also a negative interaction effectbetween a partys own and other parties press releases. The results are discussedwith respect to their generalizability and arising challenges for future research.

Corresponding author:

David N. Hopmann, Centre for Journalism, Department of Political Science, University of Southern Denmark,

Campusvej 55, DK-5230 Odense M, Denmark

Email: [email protected]

Party Politics18(2) 173191

The Author(s) 2012Reprints and permission:

sagepub.co.uk/journalsPermissions.nav

DOI: 10.1177/1354068810380097ppq.sagepub.com

http://ppq.sagepub.com/http://ppq.sagepub.com/http://ppq.sagepub.com/http://ppq.sagepub.com/


3/20

Keywords

agenda-building, agenda-setting, content analysis, election campaigns, news coverage

Paper submitted 26 August 2009; accepted for publication 28 December 2009

Introduction

While media scholars and political scientists have studied the process of agenda-

setting for more than four decades, there has been less focus on why some parties

are more effective than others when it comes to influencing the issue agenda of the

media. Such knowledge is important primarily because extant research shows that

the media agenda has a substantial influence on which policy issues the public per-ceives as important (McCombs and Shaw, 1972; Weaver et al., 2004). Since party

strategists know this, they try to influence media coverage (Asp, 1983; Brandenburg,

2003), and multiple studies indicate that parties can indeed have substantial influ-

ence over the media agenda (Brandenburg, 2002; Walgrave and van Aelst, 2006).

Moreover, we know that voters associate different parties with different policy

issues, which, in turn, may affect voting behaviour (Budge and Farlie, 1983). One

important causal factor behind associating different parties with different issues is

the extent to which parties are covered in the media in relation to the different issues

(Walgrave and de Swert, 2007; Walgrave et al., 2009).It is therefore important for both the political scientist and the political parties to

gain a better understanding of the factors that condition the ability of parties to put

their preferred issues on the media agenda. This ability to influence the media

agenda undoubtedly varies across parties (Asp, 2006; Brandenburg, 2002; Harris

et al., 2005; Semetko et al., 1991), but we do not know why. The present article

is thus an important step towards providing new insights in this regard. Drawing

on datasets measuring daily television news coverage and political parties press

releases during the 2007 Danish national election campaign, we analyse whether

and, if so, why parties have differing success in appearing in the news on the issuesthey sponsor themselves.

Even though this article presents a one-country study, the ultimate goal is to iden-

tify explanations that are likely to apply to parties in other multiparty systems with

similar patterns of party competition. In addition, although the interplay between the

media and politicians may be different during elections and non-election times (van

Aelst and de Swert, 2009; Walgrave and van Aelst, 2006), the study of election

campaigns is of central importance in representative democracies and therefore

constitutes an obvious point of departure when illustrating parties varying

agenda-setting capacities.The article proceeds in four sections: first, the theoretical framework and a number of

hypotheses are presented. Second, the empirical data and the applied methods are

described. Third, the findings are reported and, finally, the wider implications of these

findings are discussed.

174 Party Politics 18(2)



4/20

Theoretical framework and hypotheses

The agenda-setting approach and efforts to set the media agenda

The basic idea of agenda-setting studies is that the salience of issues on the media agendainfluences the salience of issues on the public agenda, that is, in the minds of voters. A

number of studies have shown that this assumption in general holds true (for an overview,

see, e.g., Weaver, 1996; Weaver et al., 2004). We also know that the salience of issues can

influence the party choices of voters (Brandenburg, 2002; Green and Hobolt, 2008; Klein-

nijenhuis and de Ridder, 1998). Given this knowledge, it is not surprising that political

parties try to control or at least influence the media agenda. This process has been

described as parties media agenda-setting or agenda-building (Brandenburg, 2002;

Dearing and Rogers, 1996; Walters et al., 1996; Weaver et al., 2004).

Empirically, studies have repeatedly found high correlations between the agendas of

political parties and the media during election campaigns: in a study of the 1979 Swedish

parliamentary election, Asp (1983: 339) finds that: The parties (..) seem to be powerful

agenda setters for the news media. Analysing the British general election of 1983,

Semetko et al. (1991) found a strong correlation between British party and television

news issue agendas. Similarly, the medias issue coverage during the 2002 Irish election

was quite close to the political parties agendas (Brandenburg, 2005). In sum, such stud-

ies confirm Walgrave and van Aelsts (2006) conclusion that political parties are quite

successful in shaping the media agenda during election campaigns, while the media have

limited or no power to influence the issue agendas of the political parties. Thus, our first

hypothesis argues that on the aggregated level parties influence the media agenda duringelection campaigns more than the media influence the party campaign agendas (H1).

Parties may differ, however, in their capacities to set the media agenda. Following the

increasing number of studies on political agenda-setting looking for contingent factors

(e.g. Vliegenthart and Walgrave, 2008; Walgrave and van Aelst, 2006), we look for factors

that allow some parties to be more influential vis-a-vis the media than others. In multiparty

systems with several parties, intuitively it seems implausible that all parties are equally

successful in influencing the medias issue coverage. The literature provides some evi-

dence that this, indeed, is not the case: Asp (2006) shows that the agenda of the news cov-

erage during the 2006 Swedish national election campaign was closer to the conservativeparties than to the socialist parties. Brettschneider (2002) shows that issue coverage during

the 2002 German Bundestagelection campaign was more in line with the incumbent left-

wing parties agenda than with the liberalconservative opposition agenda.

A closer look at these studies reveals a shortcoming. As Brandenburg (2004: 7) notes,

debates of salience manipulation are lacking (...) a clear understanding of the mechan-

ism by which political actors actually influence public salience. Most studies analyse

correlations between the party and the media agendas; they do not analyse causal

mechanisms (Asp, 1983: 351). An exception, and thus valuable contribution, is Branden-

burgs (2002: 40) study of television and press issue coverage during the 1997 Britishgeneral election: as he emphasizes, [t]he problem with previous research into party

media relationship is that campaigns have not been treated as processes but as events

(italics in original). Using time-series analysis, Brandenburg shows that, on the one hand,

the major parties [Labour and Conservatives] were the driving forces behind the overall

Hopmann et al. 175



5/20

agenda dynamics rather than being driven, at least not by the media (Brandenburg,

2002: 46). On the other hand, the (smaller) Liberal Democratic Party had no success with

influencing the media agenda.

Studies on Irish and German election news coverage also reported by Brandenburg

(2004: 25) show similar results. In Germany, the then incumbent redgreen government

and the major opposition party exerted significant influence on the news agenda of the

public television station ARD. In Ireland, it was the dominant Fianna Fail that set the

media agenda. Moreover, in Ireland the media agenda apparently also had an effect

on the party agendas (Brandenburg, 2004: 22).

Notwithstanding the important insights from Brandenburgs and others studies,

the majority of works mentioned are primarily empirical, which means that the party

differences are under-theorized and explained post hoc. Thus, the following subsection

theorizes about the varying ability of parties to set the media agenda. A graphical over-

view of the theoretical model we present and subsequently test empirically is shown

in Figure 1.

The relevance of partiesStudies have shown that incumbent parties experience a bonus in terms of media atten-

tion, as is the case in for example The Netherlands (van Praag and van der Eijk, 1998),

Italy (Djerf-Pierre and Weibull, 2005), Spain (Dez-Nicolas and Semetko, 1999), the

United Kingdom (Brandenburg, 2006), Denmark (Hopmann et al., 2010) and Germany

(Schneider et al., 1999). While parties in power act, those in the opposition merely talk

and, hence, incumbent parties are more newsworthy (Schonbach and Semetko, 1996,

2000). Accordingly, when incumbent parties are more present in election news because

they are incumbent, it might be assumed that incumbent parties also have most success in

appearing in the news on issues they publish press releases about (H2A).Notwithstanding this intuitive logic, Brandenburg (2002, 2004) suggests that focusing

simply on incumbency might be too simple: first, obviously it cannot explain why (some)

also oppositional parties appear to be successful. Second, the incumbency effect might

be smaller in times of election campaigns because incumbent parties are not only the

Own press releases

Other parties'

press releases

Appearances in TV news

on issues from

press releases

Party relevance

+

++

Figure 1. Model of how parties influence their own and other parties appearances on televisionnews during election campaigns by publishing press releases




6/20

commanders in chief but also ordinary parties in open competition with other parties

that are able to gain office (cf. van Aelst and de Swert, 2009). Finally, looking at incum-

bency alone is to ignore that the size, the political power (Powell, 2000) and the goals

(Strm, 1990) of parties vary greatly in most multiparty systems. Unlike the Americancase, in multiparty systems not every party competes for governmental office, and a

range of parties have never held office; our case, Denmark, is no exception (Skjveland,

2003). Thus, some parties are more relevant than others in both the government forma-

tion process and everyday law-making.

In his seminal work on party systems, Sartori (2005: 10710) formulates two attri-

butes defining so-called relevant parties: they need to have coalition or blackmailing

potential. Thus, small parties that are redundant to form a government, and/or are

anti-system, can be seen as largely irrelevant. Following in Sartoris footsteps, political

scientists have designed more sophisticated measures by which to count the number of

relevant parties in a given political system (for an overview, see Blau, 2008). For our

present purpose, we subscribe to the parsimonious classification of Sartori and argue that

when some parties are more relevant than others in the political system the media may

reflect these differences: with limited space in the news bulletins and a wide range of

parties to cover, journalists may simply pick up the agendas of the most relevant parties.

In fact, this line of thought jibes well with Bennetts indexing theory of press

politics relations (1990). According to this theory, the news media open the gates wider

or close them more tightly as they perceive potentially decisive challenges or a lack of

challenges to the most powerful institutional players and their agendas (Bennett et al.,

2007: 49). Therefore, we suggest the following hierarchy: parties with coalition potentialare more successful in influencing the media agenda than parties with blackmailing poten-

tial. The latter group of parties is, nevertheless, still more successful in influencing the

media agenda than irrelevant parties (H2B) (for similar reasoning, see also Brandenburg,

2003: 113).

The competition between parties

Parties success in appearing on issues they focus on themselves may not simply reflect

their general newsworthiness, as H2A and H2B imply: both hypotheses ignore that anygiven party competes against other parties for attention from the same journalists. The

second set of hypotheses therefore takes the party competition perspective.

Extant research tells us that journalists are concerned, especially during election cam-

paigns, with achieving more politically balanced news coverage by including candidates

from different parties in their news reports (cf. Donsbach and Klett, 1993: 65; Walgrave

and van Aelst, 2006: 98). Hence, if party A publishes a press release on a given issue and

subsequently is interviewed on this issue, journalists may also turn to party B or party C

to balance the statements by party A. Thus, we expect to find a spillover effect from other

parties press releases (H3A).Having said this, more parties focusing on the same issue may also have the

opposite effect. In their study on the effects of news on voters perceptions of

issue-ownership, Walgrave et al. (2009) argue that voters can be more affected by

news items containing only one politician claiming ownership of a given issue than

Hopmann et al. 177



7/20

on balanced news where one partys claims are countered by another party. That is,

possible effects on the audience of parties appearances are cancelled out by a bal-

ance between the parties. If we adopt this reasoning for the present study, the prox-

imate audience of the political parties is the gate-keeping journalists. Hence, itfollows that when too many parties publish press releases on the same issues on

a given day, the success with influencing journalists by an individual party

diminishes. Parties efforts to influence media coverage using press releases may

have the characteristics of a zero-sum game: as depicted in Figure 1, we hypothesize

that there is a negative interaction effect on a partys media appearances on a

specific issue of the number of press releases by that party on that issue and other

parties press releases on the same issue (H3B).

Data, case and methodThe media agenda

For all 20 days of the 2007 national election campaign, we conducted a content analysis

of all political news stories raised in the four major evening news bulletins on Danish

television (DR1: 18.30, 21.00; TV2: 19.00, 22.00; N 812). Television news bulletins

were chosen because televised news is the publics main source of political information

(Lund, 2001; Mazzoleni and Schulz, 1999). We included all political news stories

regardless of the focus, that is, both election news and other types of political news stor-

ies. Given the very short duration of Danish election campaigns, journalists refer to the

upcoming election in virtually all news stories.

For each news story, we coded the day of the campaign, the length, the dominant

issue and all actors featuring.1 To test hypotheses 2 to 4, we use the daily number of

times a party is mentioned together with an issue as the dependent variable. In order

to have a reasonable number of news stories each day, we first aggregate the results

for all four news bulletins.2 Second, the range of small issues categories is collapsed

into nine border issue categories: (1) economy and employment, (2) welfare and social

politics, (3) value politics (such as immigration and foreign aid), (4) education and

science, (5) environment and infrastructure, (6) the European Union, (7) defence and

terrorism, (8) campaign (i.e. voter and campaign trail stories; the definition followsprevious research, see DAngelo et al., 2005; Esser and DAngelo, 2006) and, finally,

(9) other issues. The last two categories are not included in the analysis: it is difficult

to conceive how and why the news media would copy issues within the campaign

category. If the television media responds to, for example, press releases on photo

opportunities, they most likely do not report on them, but simply use the information

to put their cameramen to work.3 We also exclude other issues because it is a resi-

dual category which by definition contains all kinds of issues hence, it makes no

sense to compare it across different issue agendas.

Party agendas

Here, we study the agendas of the different parties by using their press releases as pub-

lished during the election campaign. Press releases are analysed for a number of reasons.




8/20

First, a number of studies capture party agendas using party election manifestos (e.g.

Klingemann et al., 2006), but this source does not allow dynamic analysis of the devel-

opments in the course of the campaign. It is therefore necessary to choose a source of

communication that is used every day, which narrows the possible choices to campaign

ads, speeches, letters-to-the-editor and press releases. Ads (typically televised) are fre-

quently used in the Anglo-American literature (Kaid, 1999), but televised political adsare prohibited in many European countries, including Denmark (Kaid and Holtz-

Bacha, 1995). Moreover, small parties often cannot afford ads in the print media, which

would invalidate an analysis drawing solely on this source. Speeches are extremely dif-

ficult to get, since no central archive exists not even within the individual parties.

Finally, the issues of parties letters-to-the-editor can be biased by newspaper editors

who may have good reason to publish letters about issues covered by their newspapers.

None of the mentioned problems applies to press releases.

Second, journalists have limited resources and work under considerable time pressures

(Brandenburg, 2002; Neveu, 2002; Plasser, 2005). Hence, journalists rely on routine andreadily accessible channels of news material, such as party press releases (McLeod et al.,

2002; Sigal, 1973). Shea and Burton (2001: 173) emphasize that [t]he most important

tool for selling a candidate is the news release. (...) This form of communication (...) plays

a central role in any campaigns earned media strategy. Similarly, Walters et al. (1996: 9)

note that political campaigns (...) require many vehicles to transmit issue and image sal-

ience to voters. (...) Properly used, releases have the potential to amplify themes and

images stressed by the campaign. The sheer number of press releases published by all

parties indicates that these are indeed considered important channels of communication

in election campaigns (Table 1). The two minor parties New Alliance (NA) and theChristian Democrats (KD) published only five press releases each, so we are forced

to exclude them from the following analyses. The remaining major parties published at

least 40 press releases each, that is, on average at least two a day, indicating that these

parties treat this channel of communication as important.

Table 1. Numbers of press releases in different issue categories by the parties during the 2007campaign

Economy Welfare Values Education Environment EU Defence N

DF 5 8 16 2 4 3 2 40K 8 10 20 5 9 0 0 52KD 1 4 0 0 0 0 0 5NA 0 2 0 0 2 0 1 5RV 6 12 13 3 27 1 0 62S 12 11 10 2 17 0 3 55SF 5 27 8 7 22 0 2 71V 6 18 12 3 4 1 0 44N 43 92 79 22 85 5 8 334

DFDanish Peoples Party; KConservatives; KDChristian Democrats; NANew Alliance; RV SocialLiberals; S Social Democrats; SF Socialist Peoples Party; V Liberals. No data were available for the RedGreen Alliance (Enhedslisten).

Hopmann et al. 179



9/20

In their war rooms, parties monitor the campaign activities, issues and media

appearances of their opponents in order to be able to respond immediately (Farrell and

Webb, 2000; Gould, 1998; Jnsson, 2006). Obviously, a party cannot completely ignore

issues that are sponsored by other parties. Therefore, a party will also publish pressreleases on issues it would like to avoid highlighting in general. However, when still pub-

lishing press releases on such issues it frames them to its own advantage (Sides, 2007:

467). A party issuing a press release on a non-preferred issue probably aims at presenting

it with its own slant rather than letting other parties express their opinions on it unop-

posed if that was not the goal, it would not make much sense to publish a press release

on the issue. Hence, the definition of a party-sponsored issue is an issue on which a party

released a press release.

Each press release is coded the same way as individual news stories in the televised

broadcasts, that is, we coded which party published the release, the day of the campaign

and the dominant issue.4 Again we collapsed minor issues into the nine categories

described above and we excluded the campaign and the other issues categories.

The case and its political context

Apart from data availability, the 2007 Danish national election is chosen as our case

because a large number of parties competed in this campaign. At the same time, how-

ever, in Denmark there are two government nucleuses (Skjveland, 2003). We are

therefore able to see whether several parties with similar roles in the party system are

equally successful in influencing media coverage.On 24 October 2007, with 20 days notice, the then prime minister, Anders Fogh Ras-

mussen, announced that the next national election would be held on 13 November. Two

government alternatives were running against each other: on the one side, the incumbent

right-wing government consisting of the Liberals (V) as major partner in the coalition,

and the prime ministers party, and the Conservatives (K) as the minor partner. This gov-

ernment is a minority coalition supported by the Danish Peoples Party (DF). The pos-

sible alternative government composition presented to the voters consisted of the Social

Democrats (S) and the Social Liberals (RV). This coalition would also have been a

minority government, in this case their supporting parties are two left-wing parties (theSocialist Peoples Party [SF] and the minor RedGreen Alliance [EL]). An overview of

the parties included in our study, and their status within the political system in the 2007

Danish national election, is given in Table 2 (for more details on the 2007 election, see

Kosiara-Pedersen, 2008).

With the information given in Table 2, we can specify H2A and H2B: H2A assumes

that incumbent parties are successful in appearing in the news on issues they sponsor.

Hence, we expect the Liberals (V) and the Conservatives (K) to be most successful.

H2B takes a more detailed approach, arguing that there is a difference between non-

relevant and different types of relevant parties. Following Sartori (2005), all six partiesin our analysis are relevant. In Denmark, no anti-system party is represented in parlia-

ment and all other parties in the analysis are involved in government formation. H2B can

therefore be specified stating that parties with coalition potential (V, K, S, RV) are more

successful than parties with blackmailing potential only (DF, SF).




10/20

Statistical considerationsFor each party, the datafile resembles a pooled time-series structure with time-points

(days) nested within issues. Specifically, we test for each party separately whether it

appears in the evening news on the same issues (the dependent variable) it and other par-

ties published press releases on during the day. To establish the most appropriate analy-

tical technique that takes into account the nested structure as well as the possible

dynamic structure of the series, several steps have to be taken (see also Wilson and But-

ler, 2007).

The first question that needs to be addressed is whether the dependent variable resem-

bles a normal distribution. This is not the case. Rather, we are dealing with highlyskewed count data with a fairly large variance, making a negative binomial regression

a viable option (Long, 1997). However, results demonstrate that results from this type

of analysis largely resemble the same analysis using a linear regression. Since results

from the latter have a more straightforward interpretation, we chose to report those

analyses.

The second question that needs to be addressed is whether the series are stationary,

that is, whether the mean of each medium-level series is unaffected by a change of time

origin and thus the expected value is the same for all time-points. The relevant Fisher

tests indicate that all our series are stationary and do not have to be differenced.In addition, we find heterogeneity in the series which indicates the presence of panel-

specific (in our case issue-specific) differences in the dependent variable that are not cap-

tured by the independent variables in the model. From a substantial point of view, this

finding is no surprise: there are substantial differences across issues with some issues

receiving more party attention than others. Other reasons than partymedia dynamics are

likely to account for those variations (e.g. classification of issues). Taking into account

this heterogeneity, we have to choose between a fixed-effects or a random-effects anal-

ysis. The Hausman test indicates that both types of analysis the fixed effects resem-

bling a regression with dummies for each of the issues and the random effectsresembling a regression for which the intercept is allowed to randomly differ across

issues yield largely similar results. In that case, random effects are more efficient.

Finally, for most of the parties, the dependent variable has some moderate first-order

autocorrelation, which we handle by including a lagged dependent variable as an

Table 2. Parties running in the 2007 Danish national elections

Vote-share Change 0705 Status prior to election

Liberals (V) 26.2 2.7 Government (PMs party)Conservatives (K) 10.4 0.1 GovernmentDanish Peoples Party (DF) 13.9 0.6 Supporting government*

Social Democrats (S) 25.5 0.3 Government alternativeSocial Liberals (RV) 5.1 4.1 Government alternativeSocialist Peoples Party (SF) 13.0 7.0 Supporting gov. alternative*

*Government coalitions normally do not have a parliamentary majority of their own; hence, they needparliamentary support from other parties.

Hopmann et al. 181



11/20

independent variable. Overall, this results in pooled linear regressions with random

effects and a lagged dependent variable. To test our hypotheses, we execute the analysis

for each party in three steps. First, we look at a model that includes the lagged dependent

variable and the partys own press releases. Second, we add the number of press releases

other parties published on the issue. Third, we include an interaction term between the

partys own press release and the other parties press releases.

Findings

It is asked how successful parties are in appearing in the evening news on the issues of

their press releases published the same day. As a first step, we analyse the hypotheses

that parties as such are able to influence the media agenda, and that the media have little

impact on the party agendas (H1). Table 3 (left column) gives the results when all press

releases of all parties are pooled: first, no autocorrelation is found in the media agenda,

indicating little stability; therefore no lagged term of the dependent variable is included

in the model. This finding supports Brandenburgs (2002: 43) conclusion that agenda

stability and continuity appears to be a short-term process during campaigns. Second,a significant effect on the evening news bulletins of all press releases published the same

day is found. That is, the issue agenda of the parties press releases apparently has a sys-

tematic effect on the issue agenda in the evening news.

Results for the opposite causal direction are shown in the right column of Table 3:

first, over time, parties are more stable in their issue focus than the evening news is; tests

revealed that including a lagged term for their own issue focus is appropriate. Second, no

significant effect of the media agenda in the news on parties issue agenda in their press

releases the following day is found. Summing up, we find empirical support for H1: par-

ties indeed influence the media agenda, whereas the media do not influence the partiescampaign agendas.

Turning to the main question of the present article, we test the remaining hypotheses

formulated above. Looking at the first model in Table 4 (first column for each party)

including each partys own press releases, we find in all instances positive coefficients

Table 3. Influence on the agenda in evening news bulletins of the party agendas and vice versa;generalized least squares fixed-effects model (unstandardized coefficients)

Issue agenda in the media

Issue agenda in

party press releases

Issue agenda in party pressreleases

0.209 (0.082)*

Issue agenda in the mediat1 0.059 (0.058)Issue agenda in party press

releasest1

0.121 (0.069)

Constant 3.074 (0.197)** 2.003 (0.224)**F-value 6.37 2.82

**0.01 significance level, *0.05 significance level (one-sided). N 140 (20 days campaign 7 issue categories)

for the analysis in the left column, N 133 (19 days 7 issue categories) for the analyses in the right column.Robust standard errors in parentheses.




12/20

confirming the basic assumption that parties are successful in influencing the media

agenda not only at the aggregate party-system level (as shown above) but also at the

individual party level. With regard to the relevance of parties, this finding clearly shows

that not only incumbent parties (Liberals, Conservatives) exert a significant impact onnews coverage. We also find significant results for other parties, thus H2A on incum-

bency remains unsupported.

The more elaborate H2B on the relevance of parties suggests that a hierarchy exists:

the more relevant a party is, the more success it will have with its press releases. There

are some indications: first, we find positive and significant effects for three out of four

parties with coalition potential (Social Democrats, Liberals and Conservatives). More-

over, it seems that the big brother in each government coalition alternative has the most

influence on the media coverage (Social Democrats and Liberals, respectively). Second,

significant results are found for only one of the two parties with less relevance, that is,

blackmailing potential only (Socialist Peoples Party, but not Danish Peoples Party).

The Socialist Peoples Party had enormous momentum throughout the 2007 election

campaign gradually altering the partys position in the political system from a blackmai-

ler to a party with coalition aspirations (comparable to the Socialist Left in the rather

similar Norwegian political landscape; see Sitter, 2006). Possibly, the media saw this

change coming during the election campaign. That said, the effects for the Socialist Peo-

ples Party are much smaller than for the Social Democrats and the Liberals. In sum,

keeping in mind that we compare six parties only, we find partial support for H2B.

Next, testing the effects of party competition, we present a second set of models in

Table 4 (second column for each party). In this model we include a term for other partiespress releases, and as expected once more we find positive coefficients in all instances.

This finding clearly supports H3A: there is indeed a spillover effect from other parties,

implying that the more other parties emphasize an issue, a given party will also be cov-

ered in the news on this issue even though it did not publish press releases on the issue.

One might also note that the coefficients are significant in three out of the four cases

when looking at parties with coalition potential (Social Democrats, Social Liberals and

Liberals), but none when looking at blackmailing parties, which corroborates the

assumption that journalists pay more attention to more relevant parties (H2B).

Finally, we hypothesized a negative interaction effect between parties own press releasesand other parties press releases (H3B). This effect would imply that the effects of publishing

press releases have the characteristics of a zero-sum game: if everybody shouts, single voices

drown. The appropriate interaction term is included in the third model in Table 4 (the third

column of each party). We find negative signs in all but one instance (significant for Social

Democrats, Social Liberals and Conservatives). The one exception is the Prime Ministers

party (the Liberals), which does not fit into this general picture. Here, we find a positive coef-

ficient possibly reflecting the especially exposed position of this party.

Figure 2 illustrates the negative interaction effect between the partys own press

releases and other parties releases for the Social Democrats. Of course, the results differfor the other parties, but in general they show a similar pattern. The graph shows the pos-

itive effect of other parties press releases when the party itself does not issue any

releases on the issue under consideration: the more press releases, the larger the intercept

in Figure 2 when other parties do not issue any press release on the issue either, the

Hopmann et al. 183



13/20

Table 4. Three sets of models on the effects on appearing on television evening news in relation tospecific issues of publishing press releases on these issues (standard errors in parentheses)

Socialist government alternative

Social Democrats Social Liberals

Own releases 0.532**(0.162)

0.316*(0.168)

0.690**(0.281)

0.041(0.093)

0.095(0.100)

0.217(0.164)

Other parties releases 0.182**(0.053)

0.254**(0.068)

0.101**(0.033)

0.153**(0.039)

Own releases Other parties releases

0.123*(0.074)

0.100**(0.042)

Media attentiont1 0.109(0.082)

0.036(0.082)

0.007(0.083)

0.066(0.082)

0.006(0.082)

0.030(0.082)

Constant 0.455**(0.135)

0.217(0.147)

0.114(0.159)

0.246**(0.083)

0.126(0.089)

0.041(0.095)

R2 0.100 0.177 0.194 0.006 0.075 0.115

Bourgeois government

Liberals Conservatives

Own releases 0.621**(0.161)

0.515**(0.168)

0.048(0.352)

0.237*(0.131)

0.183(0.139)

0.619**(0.267)

Other parties releases 0.114*

(0.057)

0.061

(0.064)

0.057

(0.050)

0.102*

(0.054)Own releases

Other parties releases 0.139*

(0.076) 0.116*

(0.061)Media attentiont1 0.191**

(0.076)0.177**

(0.076)0.209**

(0.077)0.278**

(0.081)0.244**

(0.086)0.239**

(0.085)Constant 0.578**

(0.145)0.380*

(0.174)0.470**

(0.179)0.330**

(0.114)0.252*

(0.133)0.157

(0.141)R2 0.180 0.205 0.225 0.122 0.131 0.155

Supporting parties (blackmailing potential)

Socialist Peoples Party Danish Peoples Party

Own releases 0.144*(0.080)

0.139(0.086)

0.188(0.138)

0.118(0.130)

0.100(0.131)

0.141(0.204)

Other parties releases 0.005(0.030)

0.015(0.037)

0.027(0.029)

0.032(0.034)

Own releases Other parties releases

0.016(0.034)

0.015(0.055)

Media attentiont1 0.097(0.090)

0.096(0.090)

0.092(0.091)

0.276**(0.074)

0.266**(0.075)

0.266**(0.075)

Constant 0.186**

(0.072)

0.179*

(0.083)

0.163*

(0.091)

0.200**

(0.083)

0.150

(0.099)

0.140

(0.106)R2 0.051 0.051 0.053 0.106 0.112 0.112

**0.01 significance level, *0.05 significance level (one-sided). To correct for the fact that parties tend to appearon the same issues consecutively, a lagged term for each partys appearances on the evening news on a givenissue is included. N 133 observations (19 days 7 issue categories).




14/20

predicted media attention is close to zero (line with c), while if other parties release atotal of six press releases it is close to 2 (line with ~). Furthermore, the slope of the var-

ious lines indicates the effects of issuing press releases for different levels of other par-

ties press releases. When the other parties release no press releases on the issue, each

press release results in a 0.69 higher predicted number of media appearances on that

issue. For the case where other parties release three press releases, this effect is only

0.32, while for six press releases by other parties, the effect is even negative: 0.05.

DiscussionOn the aggregated level, our data support Brandenburgs (2002, 2004) finding that in

election times party agendas affect the issue agenda in the news media but not vice versa.

The fact that Brandenburgs and our study investigate both different countries (Germany,

Ireland, the UK, Denmark) and different media outlets (press, television) renders this

conclusion quite robust. But these findings do not in and of themselves explain why

some parties are more successful in influencing the media agenda than other parties.

Therefore, it was asked how to explain that some parties are more successful in

appearing on television news on issues they focus on themselves. First, the findings show

that the more press releases a party publishes on a specific issue, the more often this partywill appear on this issue in the evening television news. This finding seems to hold more

for parties with coalition potential than for parties with blackmailing potential only, but

the results are mixed. Second, if other parties publish press releases on a specific issue,

this, too, raises the chances of being covered on this issue, especially with respect to

0 1 2 3 4 5

0

1

2

3

4

No. of press releases

No.ofappearancesinthene

ws

Figure 2. Effects on appearing on television evening news related to specific issues of publishingpress releases on these issues moderated by the number of press releases by other parties: crepresents 0 press releases by other parties, r represents 3 releases, ~ represents 6 releases(results for Social Democrats)

Hopmann et al. 185



15/20

more relevant parties. Third, the more press releases are published by other parties on a

specific issue the less the effect a given party will experience by publishing yet another

press release, leading eventually to an overall negative effect of an increased number of

press releases.Obviously, the media are not independent actors acting autonomously but are

engaged in an interaction with political actors who are their central sources in election

campaign coverage. Inspired by the emerging research focusing on the contingent factors

that determine the size of the mass medias influence on politics and vice versa, we set

out to identify factors that condition a partys success in influencing media coverage: we

find that systemic party characteristics indeed make a difference. It does make a differ-

ence whether a party has coalition potential rather than having blackmailing potential

only. Moreover, it even seems to make a difference which party is the larger coalition

member.

These findings have major implications for democracy in broader terms as well as

how parties can set the media agenda. First, it raises the important question of how inclu-

sive media coverage can and should be. If more relevant parties have more success with

influencing the media agenda and hence influence the public agenda in their favoured

direction we might be observing a self-sustaining circle because you are relevant you

have influence and because you have influence you are relevant.5 Depending on which

ideal of democracy one adheres to, such a dynamic can be seen as problematic as it

potentially disfavours new, emerging political voices (see Stromback, 2005). Second, the

findings help us to understand by which dynamics the media agenda is shaped and whose

agenda the public effectively is following when influenced by media coverage. Duringelection periods, the media indeed appear to be receptive to the issue attention of the par-

ties rather than setting the agenda for the parties. From a practical perspective, party stra-

tegists can take advantage of this fact. Moreover, our results indicate that if too many

parties publish press releases on the same issue, publishing yet another press release does

not help gain media attention; party resources can be used elsewhere.

We expect our empirical conclusions to hold true in other multiparty systems with

similar patterns of party competition: first, there is no reason to assume that journalists

in general use press releases much differently in other countries: it is not just Danish

journalists who work under considerable time and resource pressure (e.g. Neveu,2002; Plasser, 2005). Likewise, journalists in many other countries are, as mentioned

above, concerned with objectivity, political neutrality and hence aim at balancing their

news reports (cf. Donsbach and Klett, 1993; Splichal and Sparks, 1994).

Notwithstanding, the study in hand has limitations: first, we studied news coverage of

an election campaign. Recent research shows that campaign news is more balanced than

routine news (van Aelst and de Swert, 2009; Walgrave and van Aelst, 2006), which

might affect the possibility for parties to manipulate the media agenda in a negative

direction for incumbents and a positive direction for other parties. Second, to develop

a more comprehensive understanding of the presspolitics nexus, future studies mightlook at second-level agenda-setting and the parties ability to influence not only issues

in the media but also the framing of these issues (Stromback and Nord, 2006). Next,

from our study, one cannot infer that, for example, the Danish Peoples Party in gen-

eral was less successful in being covered on self-sponsored issues, only that this was




16/20

the case with respect to the partys press releases. Hence, future studies should try to

account for the fact that the mediaparty relationship also involves other communica-

tion channels than press releases. Finally, future studies should also try to account for

exogenous factors such as real-world events: one could argue that simultaneous issueattention by parties is not caused by their strategic reactions to one another, but by a

third factor or external shocks such as developments outside the political sphere, for

example, the publication of new unemployment figures or an incident in a war. Nev-

ertheless, from the journalists point of view, this does not change much, as press

releases, for whatever reason published, still signal that parties want to speak on spe-

cific issues and therefore suggest themselves to the media as interviewees.

In sum, we have shown that, during election times, parties have a substantial effect on

the media agenda, but also why some parties appear to be more successful than others in

setting the media agenda. The next steps are to compare our results with those of other

countries across time and to include more detailed information on party communication

and media content.

Notes

1. Testing for inter-coder reliability yielded the following results (Krippendorffs alpha values):

0.81 for issue, 0.96 for length and 0.96 for actor affiliation. The length variable is log-

transformed.

2. Rerunning the full model presented in Table 4 does not indicate substantial differences between

the two broadcasting stations, that is, it is not the case that the overall results are driven by one

of them. This finding is in line with previous research on Danish television broadcasters con-

cluding that the coverage is similar at both stations (Hopmann et al., 2010).

3. That said, the results presented here are essentially the same when the campaign category is

included in the analyses.

4. The inter-coder reliability was 0.73 (Krippendorffs alpha) for the issue coding.

5. Similarly, it could be suggested that parties have more success with publishing press releases on

issues which they hold ownership. Following findings from Hayes (2008) for several US pres-

idential election campaigns, journalists are more sceptical towards parties when they trespass

on other parties issues. For each party included in our analysis, we selected the issue to whichthey devoted most attention in their party manifestos and assessed whether the effects of press

releases were larger for that issue. Although we find that parties are more covered on issues they

own, there is no interaction effect between issue-ownership and success with influencing news

coverage with press releases on owned issues. Apparently, the issue-ownership logic does not

influence journalists in their decisions about what to cover, as it is argued to influence voters, at

least in a multiparty context as Denmark, in which issue-ownership can be considered to be

more divided and diffuse.

References

Asp, Kent (1983) The Struggle for the Agenda, Communication Research 10: 33355.

Asp, Kent (2006)Rattvisa nyhetsmedier Partiskheten under 2006 ars medievalrorelse. Goteborg:

JMG Arbetsrapport nr. 42, Goteborgs Universitet.

Hopmann et al. 187



17/20

Bennett, W. Lance (1990) Toward a Theory of PressState Relations in the United States,Journal

of Communication 40: 10325.

Bennett, W. Lance, Regina G. Lawrence and Steven Livingston (2007) When the Press Fails

Political Power and the News Media from Iraq to Katrina. Chicago, IL: University of ChicagoPress.

Blau, Adrian (2008) The Effective Number of Parties at Four Scales, Party Politics 14: 16787.

Brandenburg, Heinz (2002) Who Follows Whom? The Impact of Parties on Media Agenda For-

mation in the 1997 British General Election Campaign, Harvard International Journal of

Press/Politics 7: 3454.

Brandenburg, Heinz (2003) Agenda Building. A Time-Series Analysis of the Changing Issue Prio-

rities of Parties and Media During the 1997 General Election Campaign in the UK. Trondheim:

NTNU Press.

Brandenburg, Heinz (2004) Manipulating the Dimensions: A Comparative Study of Campaign

Effects on Media Agenda Formation. Paper presented at the ECPR Joint Sessions of Work-

shops, University of Uppsala.

Brandenburg, Heinz (2005) Political Bias in the Irish Media: A Quantitative Study of Campaign

Coverage during the 2002 General Election, Irish Political Studies 20: 297322.

Brandenburg, Heinz (2006) Party Strategy and Media Bias: A Quantitative Analysis of the 2005

UK Election Campaign, Journal of Elections, Public Opinion and Parties 16: 15778.

Brettschneider, Frank (2002) Die Medienwahl 2002: Themenmanagement und Berichterstat-

tung, Aus Politik und Zeitgeschichte 2002: 3647.

Budge, Ian and Dennis J. Farlie (1983) Explaining and Predicting Elections: Issue Effects and

Party Strategies in Twenty-Three Democracies. London: George Allen & Unwin.

DAngelo, Paul, Matthew Calderone and Anthony Territola (2005) Strategy and Issue Framing:

An Exploratory Analysis of Topics and Frames in Campaign 2004 Print News, Atlantic Jour-

nal of Communication 13: 199219.

Dearing, James W. and Everett M. Rogers (1996) Agenda-setting. Thousand Oaks, CA: Sage.

Dez-Nicolas, Juan and Holli Semetko (1999) Los programas de noticias de television y las cam-

panas electorales de 1993 y 1996: Propiedad, contenido e influencia, in Alejandro Munoz

Alonso and Juan Ignacio Rospi (eds) Democracia mediatica y campanas electorales, pp.

151201. Barcelona: Ariel Communicacion.

Djerf-Pierre, Monika and Lennart Weibull (2005) Par Condicio? Politisk partiskhet i italienskamedier. Stockholm: Studieforbundet Naringsliv och Samhalle.

Donsbach, Wolfgang and Bettina Klett (1993) Subjective Objectivity. How Journalists in four

Countries Define a Key Term of their Profession, Gazette 51: 5386.

Esser, Frank and Paul DAngelo (2006) Framing the Press and Publicity Process in U.S., British,

and German General Election Campaigns, Press/Politics 11: 4466.

Farrell, David and Paul Webb (2000) Political Parties as Campaign Organizations, in Russel

Dalton and Martin Wattenberg (eds) Parties without Partisans. Political Change in Advanced

Industrial Democracies, pp. 10229. Oxford: Oxford University Press.

Gould, Philip (1998) The Unfinished Revolution How the Modernisers Saved the Labour Party.London: Little, Brown and Company.

Green, Jane and Sara Binzer Hobolt (2008) Owning the Issue Agenda: Party Strategies and Vote

Choice in British Elections, Electoral Studies 27: 46076.




18/20

Harris, Phil, Donna Fury and Andrew Lock (2005) The Evolution of a Campaign: Tracking Press

Coverage and Party Press Releases through the 2001 UK General Election, Journal of Public

Affairs 5: 99111.

Hayes, Danny (2008) Party Reputations, Journalistic Expectations: How Issue Ownership Influ-ences Election News, Political Communication 25: 377.

Hopmann, David Nicolas, Erik Albk and Claes H. de Vreese (2010) Incumbency Bonus in Elec-

tion News Coverage Explained: The Logics of the Political System and the Media Market,

Journal of Communication.

Jnsson, Rasmus (2006) Den professionelle politiske kommunikation krver nye journalistiske

perspektiver, in Peter Bro, Rasmus Jnsson and Ole Larsen (eds) Politisk journalistik og kom-

munikation: Forandringer i forholdet mellem politik og medier, pp. 4366. Frederiksberg:

Samfundslitteratur.

Kaid, Lynda Lee (1999) Political Advertising: A Summary of Research Findings, in Bruce

Newman (ed.) The Handbook of Political Marketing, pp. 42338. Thousand Oaks, CA:

Sage.

Kaid, Lynda Lee and Christina Holtz-Bacha (eds) (1995) Western Democracies: Parties and Can-

didates on Television. London: Sage.

Kleinnijenhuis, Jan and Jan de Ridder (1998) Issue News and Electoral Volatility, European

Journal of Political Research 33: 41337.

Klingemann, Hans-Dieter et al. (2006) Mapping Policy Preferences II: Estimates for Parties,

Electors, and Governments in the Eastern Europe, European Union, and OECD 19902003.

Oxford: Oxford University Press.

Kosiara-Pedersen, Karina (2008) The 2007 Danish General Election: Generating a Fragile Major-

ity, West European Politics 31: 10408.

Long, J. Scott (1997) Regression Models for Categorical and Limited Dependent Variables, 1st

edn. London: Sage.

Lund, Anker Brink (2001) Forskning om medier og demokrati. Kbenhavn: Danske Dagblades

Forening.

Mazzoleni, Gianpietro and Winfried Schulz (1999) Mediatization of Politics: A Challenge for

Democracy? Political Communication 16: 24761.

McCombs, Maxwell and Donald Shaw (1972) The Agenda-Setting Function of Mass Media,

Public Opinion Quarterly 36: 17687.McLeod, Douglas M., Gerald M. Kosicki and Jack M. McLeod (2002) Resurveying the Bound-

aries of Political Communications Effects, in Jennings Bryant and Dolf Zillmann (eds)

Media Effects Advances in Theory and Research, pp. 21567. Mahwah, NJ: Lawrence Erl-

baum Associates.

Neveu, Erik (2002) Four Generations of Political Journalism, in Raymond Kuhn and Erik Neveu

(eds) Political Journalism: New Challenges, New Practices, pp. 2243. London: Routledge.

Plasser, Fritz (2005) From Hard to Soft News Standards? How Political Journalists in Different

Media Systems Evaluate the Shifting Quality of News, Harvard International Journal of

Press/Politics 10: 4768.Powell, G. Bingham (2000) Elections as Instruments of Democracy: Majoritatian and Propor-

tional Visions. New Haven, CT: Yale University Press.

Sartori, Giovanni (2005) Parties and Party Systems A Framework for Analysis. Colchester:

ECPR Press.

Hopmann et al. 189



19/20

Schneider, Melanie, Klaus Schonbach and Holli Semetko (1999) Kanzlerkandidaten in den Fern-

sehnachrichten und in der Wahlermeinung Befunde zum Bundestagswahlkampf 1998 und

fruheren Wahlkampfen, Media Perspektiven 1999: 26269.

Schonbach, Klaus and Holli Semetko (1996) Wahlkommunikation, Journalisten und Wahler:Funf Thesen zum Bundestagswahlkampf 1990 mit einem internationalen Vergleich und

einem ersten Blick auf 1994, in Heinrich Oberreuter (ed.) Parteiensystem am Wendepunkt?

Wahlen in der Fernsehdemokratie, pp. 15364. Munchen: Olzog Verlag.

Schonbach, Klaus and Holli Semetko (2000) Gnadenlos professionell: Journalisten und die

aktuelle Medienberichterstattung in Bundestagswahlkampfen 19761998, in Hans Bohrmann

et al. (eds) Wahlen und Politikvermittlung durch Massenmedien, pp. 6989. Wiesbaden: West-

deutscher Verlag.

Semetko, Holli et al. (1991) The Formation of Campaign Agendas: A Comparative Analysis of

Party and Media Roles in Recent American and British Elections. Hillsdale, NJ: Lawrence Erl-

baum Associates.

Shea, Daniel M. and Michael John Burton (2001) Campaign Craft The Strategies, Tactics, and

Art of Political Campaign Management. London: Praeger.

Sides, John (2007) The Consequences of Campaign Agendas, American Politics Research 35:

46588.

Sigal, Leon V. (1973) Reporters and Officials The Organization and Politics of Newsmaking.

London: D. C. Heath.

Sitter, Nick (2006) Norways Storting Election of September 2005: Back to the Left? West Eur-

opean Politics 29: 57380.

Skjveland, Asbjrn (2003) Government Formation in Denmark 19531998. Arhus: Forlaget

Politica.

Splichal, Slavoko and Colin Sparks (1994) Journalists for the 21st Century: Tendencies of Profes-

sionalisation among First-Year Students in 22 Countries. Norwood: Ablex Publishing.

Strm, Kaare (1990) A Behavioural Theory of Competitive Political Parties, American Journal

of Political Science 34: 56598.

Stromback, Jesper (2005) In Search of a Standard: Four Models of Democracy and Their Norma-

tive Implications for Journalism, Journalism Studies 6: 33145.

Stromback, Jesper and Lars W. Nord (2006) Do Politicians Lead the Tango? A Study of the Rela-

tionship between Swedish Journalists and their Political Sources in the Context of ElectionCampaigns, European Journal of Communication 21: 14764.

van Aelst, Peter and Knut de Swert (2009) Politics in the News: Do Campaigns Matter? A Com-

parison of Political News During Election Periods and Routine Periods in Flanders (Belgium),

Communications 34: 14968.

van Praag, Philip and Cees van der Eijk (1998) News Content and Effects in an Historic Cam-

paign, Political Communication 15: 16583.

Vliegenthart, Rens and Stefaan Walgrave (2008) The Contingency of Intermedia Agenda Setting.

A Longitudinal Study in Belgium, Journalism & Mass Communication Quarterly 85: 86077.

Walgrave, Stefaan and Knut de Swert (2007) Where Does Issue Ownership Come From?From the Party or from the Media? IssueParty Identifications in Belgium, 19912005,

Press/Politics 12: 3767.

Walgrave, Stefaan and Peter van Aelst (2006) The Contingency of the Mass Medias Political

Agenda Setting Power: Toward a Preliminary Theory, Journal of Communication 56: 88109.




20/20

Walgrave, Stefaan, Jonas Lefevere and Michiel Nuytemans (2009) Issue Ownership Stability

and Change: How Political Parties Claim and Maintain Issues Through Media Appearances,

Political Communication 26: 15372.

Walters, T. N., L. M. Walters and Roger Gray (1996) Agenda Building in the 1992 PresidentialCampaign, Public Relations Review 22: 924.

Weaver, David (1996) Media Agenda-Setting and Elections: 1970s1990s, in David L.

Paletz (ed.) Political Communication Research: Approaches, Studies, Assessments, vol.

2, pp. 21124. Norwood: Ablex Publishing.

Weaver, David, Maxwell McCombs and Donald L. Shaw (2004) Agenda-Setting Research:

Issues, Attributes, and Influences, in Lynda Lee Kaid (ed.) Handbook of Political Communi-

cation Research, pp. 25781. London: Lawrence Erlbaum Associates.

Wilson, Sven E. and Daniel M. Butler (2007) A Lot More to Do: The Sensitivity of Time-Series

Cross-Section Analyses to Simple Alternative Specifications, Political Analysis 15: 10123.

Author Biographies

David N. Hopmann is Assistant Professor in Media and Politics at the Centre for Journalism,

Department of Political Science, University of Southern Denmark. He works on the production

of media content and how it affects news consumers. His most recent articles appear in Journal

of Communication and Political Communication.

Christian Elmelund-Prstekr is Assistant Professor at the University of Southern Denmark.

He is currently engaged in the Danish Local Election Program, where his primary focus is local

agenda dynamics and the concept of personalization. He is also working on the project BeyondClass Politics, which investigates the evolution in processes of political reform throughout the

post-war era. His recent works have been published in European Political Science Review and the

International Journal of Press/Politics.

Erik Albk is Professor in Journalism and Political Science at the University of Southern Den-

mark. He is currently working on projects on Political Journalism in Comparative Perspective

and Morality Politics in Comparative Perspective. He was chairman of the Danish Social Science

Research Council 20012005.

Rens Vliegenthart is Assistant Professor in Political Communication at the Amsterdam School of

Communication Research (ASCoR), University of Amsterdam. His research interests include pol-iticsmedia relations, election campaigns, media effects on public opinion, social movements and

time-series analysis.

Claes H. de Vreese is Professor of Political Communication and Director of the Amsterdam

School of Communication Research at the University of Amsterdam. His research has been pub-

lished in more than 70 articles in international peer-reviewed journals, and his work is funded by

Danish, Dutch and European Science Foundations.

Hopmann et al. 191

party politics 2012

Documents